23
http://www.diva-portal.org This is the published version of a paper published in Nucleic Acids Research. Citation for the original published paper (version of record): Moukadiri, I., Garzón, M., Björk, G., Armengod, M. (2014) The output of the tRNA modification pathways controlled by the Escherichia coli MnmEG and MnmC enzymes depends on the growth conditions and the tRNA species. Nucleic Acids Research, 42(4): 2602-2623 http://dx.doi.org/10.1093/nar/gkt1228 Access to the published version may require subscription. N.B. When citing this work, cite the original published paper. Permanent link to this version: http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-87654

MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

httpwwwdiva-portalorg

This is the published version of a paper published in Nucleic Acids Research

Citation for the original published paper (version of record)

Moukadiri I Garzoacuten M Bjoumlrk G Armengod M (2014)

The output of the tRNA modification pathways controlled by the Escherichia coli MnmEG and

MnmC enzymes depends on the growth conditions and the tRNA species

Nucleic Acids Research 42(4) 2602-2623

httpdxdoiorg101093nargkt1228

Access to the published version may require subscription

NB When citing this work cite the original published paper

Permanent link to this versionhttpurnkbseresolveurn=urnnbnseumudiva-87654

The output of the tRNA modification pathwayscontrolled by the Escherichia coli MnmEG andMnmC enzymes depends on the growth conditionsand the tRNA speciesIsmaıl Moukadiri1 M-Jose Garzon1 Glenn R Bjork2 and M-Eugenia Armengod13

1Laboratory of RNA Modification and Mitochondrial Diseases Prıncipe Felipe Research Center 46012-ValenciaSpain 2Department of Molecular Biology Umea University S90187 Sweden and 3Biomedical ResearchNetworking Centre for Rare Diseases (CIBERER) (node U721) Spain

Received July 31 2013 Revised and Accepted November 5 2013

ABSTRACT

In Escherichia coli the MnmEG complexmodifies transfer RNAs (tRNAs) decoding NNANNG codons MnmEG catalyzes two different modi-fication reactions which add an aminomethyl (nm)or carboxymethylaminomethyl (cmnm) groupto position 5 of the anticodon wobble uridineusing ammonium or glycine respectively IntRNAGln

cmnm5s2UUG and tRNALeucmnm5UmAA however

cmnm5 appears as the final modification whereasin the remaining tRNAs the MnmEG products areconverted into 5-methylaminomethyl (mnm5)through the two-domain bi-functional enzymeMnmC MnmC(o) transforms cmnm5 into nm5whereas MnmC(m) converts nm5 into mnm5 thusproducing an atypical network of modificationpathways We investigate the activities and tRNAspecificity of MnmEG and the MnmC domains theability of tRNAs to follow the ammonium or glycinepathway and the effect of mnmC mutations ongrowth We demonstrate that the two MnmCdomains function independently of each other andthat tRNAGln

cmnm5s2UUG and tRNALeucmnm5UmAA are sub-

strates for MnmC(m) but not MnmC(o) Synthesisof mnm5s2U by MnmEG-MnmC in vivo avoidsbuild-up of intermediates in tRNALys

mnm5s2UUU We alsoshow that MnmEG can modify all the tRNAs via theammonium pathway Strikingly the net output of theMnmEG pathways in vivo depends on growth

conditions and tRNA species Loss of any MnmCactivity has a biological cost under specificconditions

INTRODUCTION

Transfer RNAs (tRNAs) are highly and diverselymodified and each has a unique pattern of modification(1ndash3) These modifications are post-transcriptionallyintroduced at precise positions by specific enzymes andplay important roles in folding stability identity and inthe functions of tRNAs

Modifications at the wobble position of the anticodoncontribute to the accuracy and efficiency of protein syn-thesis (134) Changes in the modification levels of thewobble position affect the synthesis of specific proteinsand also lead to complex phenotypes through unknownmechanisms (35ndash10)

Wobble modifications are often complex and require fortheir synthesis the participation of several enzymes (3)Even though all of the enzymes involved in a specificmodification pathway are known it is unclear how theiractions are modulated and coordinated to produce thefinal modification Progress in this area may lead to abetter understanding of tRNA biology

Proteins MnmE and MnmG (formerly TrmE andGidA respectively) are evolutionary conserved from bac-teria to eukaryotic organelles In Escherichia coli theyare involved in the modification of the wobble uridine

of tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG

tRNALeucmnm5UmAA tRNAArg

mnm5UCU and tRNAGlymnm5UCC all

To whom correspondence should be addressed Tel +34 963289680 Fax +34 963289701 Email armengodcipfesCorrespondence may also be addressed to Ismaıl Moukadiri Tel +34 963289681 Fax +34 963289701 Email IsmailMoukadiriuves

The authors wish it to be known that in their opinion the first two authors should be regarded as Joint First Authors

2602ndash2623 Nucleic Acids Research 2014 Vol 42 No 4 Published online 30 November 2013doi101093nargkt1228

The Author(s) 2013 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby30) whichpermits unrestricted reuse distribution and reproduction in any medium provided the original work is properly cited

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

of which read NNAG codons of the split codon boxes

with the exception of tRNAGlymnm5UCC which reads codons

of the glycine family box (11) MnmE and MnmG aredimeric and form a functional a2b2 heterotetramericcomplex (MnmEG) in which both proteins are inter-dependent (1213) MnmE is a GTP- and tetrahydrofolate(THF)-binding protein whereas MnmG is a FAD- andNADH-binding protein (14ndash18) The MnmEG complexcatalyzes the addition of the aminomethyl (nm) andcarboxymethylaminomethyl (cmnm) groups to position 5of the wobble uridine using ammonium and glycine re-spectively [Figure 1 (13)] Both reactions require GTP andFAD as well as NADH if FAD is limiting in the in vitroreaction Moreover a THF derivative likely methylene-THF serves as the donor of the methylene carbon that isdirectly bonded to the C5 atom of U34 However thedetailed mechanism of the basic reaction has not yetbeen demonstrated (1113) According to current modelsGTP hydrolysis by the MnmE G-domain which is locateddistant from the MnmEG active center leads to structuralrearrangements in the MnmEG complex which are essen-tial for tRNA modification (1119)

The wobble uridine (U34) of tRNAs modifiedby MnmEG may be further modified by other enzymes

at position 5 or other positions In tRNALysmnm5s2UUU

tRNAGlumnm5s2UUC and tRNAGln

cmnm5s2UUG thiolation atposition 2 of the U34 is performed by MnmA (20)

whereas in tRNALeucmnmUmAA TrmL methylates the 20-OH

group of the U-ribose (21) Therefore some tRNA sub-strates for the MnmEG complex are also substrates forMnmA or TrmL

Moreover in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC the resultingproducts of MnmEG activity are not the final modifica-tions because these tRNAs carry the methylaminomethyl(mnm) group at position 5 of U34 Formation of themnm5-final group is mediated by the action of the two-domain bi-functional enzyme MnmC (22ndash26) Theoxidoreductase activity of the C-terminal domain heredesignated as MnmC(o) is responsible for the FAD-dependent deacetylation that transforms cmnm5U intonm5U whereas the methyltransferase activity of theN-terminal domain designated as MnmC(m) catalyzesthe SAM-dependent methylation that transforms nm5Uto mnm5U (Figure 1)

The crystal structure of MnmC consists of two globulardomains interacting with each other (27) The structurealso reveals that the two catalytic centers of MnmC(o)and MnmC(m) face opposite sides of the protein thusfavoring a model in which the two domains couldfunction in a relatively independent manner In factwhen two MnmC mutant proteins possessing a catalytic-ally dead MnmC(o) or MnmC(m) domain were mixedpartial recovery of mnm5-group synthesis was observed(25) However considering the relatively hydrophilicnature of the domain interface the possibility that con-formational changes within the entire MnmC protein mayoccur in vivo and affect its functioning cannot beexcluded A previous attempt to separately express both

MnmC domains revealed that MnmC(m) was solublewhereas MnmC(o) produced inclusion bodies suggestingthat MnmC(m) is required for the correct folding or struc-tural stability of MnmC(o) (25)How the activities of the MnmEG and MnmC enzymes

are organized and modulated remains unclear WhethertRNAs preferentially utilize one of the two MnmEGpathways the glycine or the ammonium pathway(Figure 1) in vivo depending on metabolic circumstanceshas not been investigated The fact that only cmnm5 butnot nm5 or mnm5 has been detected to date at position 5

of U34 in tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA suggeststhat both tRNAs do not use the ammonium pathway andare not substrates for MnmC(o) Accumulation of thenm5s2U nucleoside which may have a dual origin[Figure 1 (13)] has not been observed in totaltRNA purified from wild-type E coli cells (23)

Therefore the mnm5s2U synthesis in tRNALysmnm5s2UUU

and tRNAGlumnm5s2UUC appears to be organized to prevent

nm5s2U accumulation However the presence or absenceof cmnm5s2U as an intermediate in mnm5s2U biosynthesisis difficult to determine from nucleoside analysis of bulktRNA due to the natural occurrence of cmnm5s2U in

tRNAGlncmnm5s2UUG Biosynthetic tuning of mnm5s2U to

avoid the accumulation of intermediates could beachieved by either kinetic tuning of the activities ofMnmEG and MnmC which requires that the sequentialmodifications are performed at similar or increasing ratesor selective degradation of partially modified tRNAsA previous steady-state kinetic analysis of the activitiesof the full MnmC protein indicated that the MnmC(m)-dependent reaction (nm5

mnm5) occurs faster than theMnmC(o)-dependent reaction (cmnm5

nm5) or at asimilar rate at very high substrate (tRNA) concentrations(26) However this study did not take into account theefficiency of the reactions performed by MnmEG(Figure 1) which is crucial to understand the organizationof the mnm5s2U biosynthetic processModifications located within or adjacent to the anti-

codon are important for stabilization of codonndashanti-codon pairing as well as for restricting the dynamicsof the anticodon domain and shaping its architecture(14) Considering the network of pathways leading tomnm5s2U production (Figure 1) it is important to de-termine whether mutations affecting the differentactivities of MnmC [MnmC(o) or MnmC(m)] havedistinct biological consequences and to what extent anaccumulation of modification intermediates affects bac-terial biologyIn this study we investigated the activities and the

tRNA specificity of the MnmEG and MnmC enzymesin vivo and in vitro analyzed how the U34 modificationstatus is influenced by genetic and physiological condi-tions in bulk and specific tRNAs and assessed the effectsof mnmC mutations on bacterial growth Our study wasfacilitated by the cloning and separate expression of thetwo MnmC domains MnmC(o) and MnmC(m) Incontrast to a previous report (25) we demonstrate thatMnmC(o) can fold independently of MnmC(m) and thatthe separate domains exhibit similar kinetic properties

Nucleic Acids Research 2014 Vol 42 No 4 2603

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

(s2)U34

U34

(MnmA)

MnmEbullMnmG

cmnm5(s2)U34 nm5(s2)U34

mnm5(s2)U34

MnmC(o)

MnmC(m)

CH2-THFGTP

FADNADH

FAD

SAM

N

NH

O

S

OOR

O HO R

CH2NH2

N

NH

O

OOR

O HO R

S

CH2NHCH2COOH

N

NH

O

OOR

O HO R

S

N

NH

O

S

OOR

OHO R

CH2NHCH3

25

CH2NHCH2COOH

N

NH

O

OOR

O HO R

0

(cmnm5)U

CH2NHCH2COOH

N

NH

O

OOR

OCH3O R

0

(cmnm5)Um

TrmL

tRNALeucmnm5UmAA

Figure 1 Synthesis of mnm5(s2)U in E coli The MnmEG complex formed by proteins MnmE and MnmG is active on position 5 of the wobble

uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA tRNAArgmnm5UCU and tRNAGly

mnm5UCC MnmE is a three-domainprotein that binds GTP and methylenetetrahydrofolate (CH2-THF) whereas MnmG is a FAD- and NADH-binding protein The MnmC(o) activityof the bi-functional enzyme MnmC transforms cmnm5(s2)U to nm5(s2)U and the MnmC(m) activity of MnmC transforms nm5(s2)U to the final

product mnm5(s2)U Thiolation at position 2 of the wobble uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlncmnm5s2UUG is mediated by

the protein MnmA Modifications occur at the 5- and 2-positions independent of each other TrmL methylates the 20-OH group of the U-ribose in

tRNALeucmnm5UmAA (inset panel)

2604 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

to those of the full protein We also show that

tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA are substrates forMnmC(m) but not for MnmC(o) Our data suggest thatMnmEG andMnmC are kinetically tuned to produce only

the fully modified nucleoside mnm5U in tRNALysmnm5s2UUU

We demonstrate that all the tRNA substrates of MnmEGare modified in vitro through the ammonium pathwayHowever the net output of the ammonium and glycinepathways of MnmEG in vivo depends on growth condi-tions and tRNA species Finally we demonstrate that theloss of any MnmC activity has a biological cost underspecific conditions

MATERIALS AND METHODS

Bacterial strains plasmids primers media andgeneral techniques

Escherichia coli strains and plasmids are shown in Table 1A list of the oligonucleotides and primers used inthis work is provided in Supplementary Table S1Transduction with phage P1 was performed as previouslydescribed (28) Deletion of the mnmC(o) coding regionwas performed by targeted homologous recombination(29) using the primers MnmC(o)-F and MnmC(o)-RThe MnmC(o)-F primer introduced a TAA stop codon

Table 1 Strains and plasmids used in this study

Strain or plasmid Description References

Eschirichia coli strainsDEV16 (IC4385)a thi-1 rel-1 spoT1 lacZ105UAG valRmnmE-Q192X (1448)BW25113 (IC5136)a lacIq rrnBT14 lacZWJ16 hsdR514 araBADAH33 rhaBADLD78 (29)TH178 (IC5255)a mnmA1 fadRTn10 (45)MG1655(IC5356)a F- D TouatiTH48 (IC6017)a ara (lac-proB) nalA argEam rif thi fadLTn10 (23)TH49 (IC6018)a TH48 mnmC(m)-G68D (23)TH69 (IC6019)a TH48 mnmC-W131stop (23)IC4639 DEV16 mnmE+ bgl (Sal+) (12)IC5241 MG1655 mnmGTn10 [TetR] (12)IC5358 MG1655 mnmEkan (43)IC5397 P1 (IC5255) x IC4639 [IC4639 mnmA-Q233stop] M VillarroyaIC5827 BW25113 mnmEkan NBRP- JapanIC5854 BW25113 trmLkan (21)IC5937 IC4639 mnmAkan [or IC4639 DmnmA] (39)IC5975 BL21-DE3 mnmGkan (17)IC6010 BW25113 mnmCKan [or BW25113 DmnmC] (13)IC6166 IC5975 carrying pIC1446 (17)IC6222 P1 (IC6010) x IC4639 [IC4639 DmnmC)] This studyIC6374 IC4639 trmLkan [IC4639 DtrmL] (21)IC6411 P1 (IC5241) x IC6374 (trmLkan) [IC4639 DtrmL DmnmG] (21)IC6424 IC5358 carrying pIC684 (14)IC6587 P1 (TH49) x IC4639 [IC4639 mnmC(m)-G68D] This studyIC6588 P1 (TH49) x IC5937 [IC4639 DmnmA mnmC(m)-G68D] This studyIC6589 P1 (IC6010) x IC5397 [IC4639 mnmA-Q233stop DmnmC] This studyIC6629 BW25113 mnmC(o)cat [or BW25113 DmnmC(o)] This studyIC6725 P1 (IC6629) x IC5937 [IC4639 DmnmA DmnmC(o)] This studyPlasmids

pIC684 GST fusion of mnmE cloned in pGEX-2T (14)

pIC1083 Eschirichia coli tRNALysmnm5s2UUU cloned in pUC19 (34)

pIC1253 flag-mnmC cloned in pBAD-TOPO (13)pIC1339 flag-mnmC(o) cloned in pBAD-TOPO This studypIC1340 flag-mnmC(m) cloned in pBAD-TOPO This study

pIC1394 Eschirichia coli tRNACysGCA cloned into pUC19 SmaI site This study

pIC1395 E schirichia coli tRNAArgmnm5UCU cloned into pUC19 SmaI site This study

pIC1446 his-mnmG cloned in pET15b (17)

pIC1535 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBAD-TOPO This study

pIC1536 Eschirichia coli tRNAGlumnm5s2UUC cloned into pUC19 SmaI site This study

pIC1537 Eschirichia coli tRNAGlymnm5UCC cloned into pUC19 SmaI site This study

pIC1550 Eschirichia coli tRNALeucmnm5UmAA cloned into pUC19 SmaI site This study

pIC1577 pBSKrna (32)

pIC1617 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRV site This study

pIC1664 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRIPstI sites This study

pIC1665 Eschirichia coli tRNALeucmnm5UmAA cloned into pBSKrna EcoRIndashPstI site This study

pIC1677 his-mnmC(m) cloned in pBAD-TOPO This study

pIC1714 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBSKrna EcoRIndashPstI site This study

aName in our collection between brackets

Nucleic Acids Research 2014 Vol 42 No 4 2605

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 2: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

The output of the tRNA modification pathwayscontrolled by the Escherichia coli MnmEG andMnmC enzymes depends on the growth conditionsand the tRNA speciesIsmaıl Moukadiri1 M-Jose Garzon1 Glenn R Bjork2 and M-Eugenia Armengod13

1Laboratory of RNA Modification and Mitochondrial Diseases Prıncipe Felipe Research Center 46012-ValenciaSpain 2Department of Molecular Biology Umea University S90187 Sweden and 3Biomedical ResearchNetworking Centre for Rare Diseases (CIBERER) (node U721) Spain

Received July 31 2013 Revised and Accepted November 5 2013

ABSTRACT

In Escherichia coli the MnmEG complexmodifies transfer RNAs (tRNAs) decoding NNANNG codons MnmEG catalyzes two different modi-fication reactions which add an aminomethyl (nm)or carboxymethylaminomethyl (cmnm) groupto position 5 of the anticodon wobble uridineusing ammonium or glycine respectively IntRNAGln

cmnm5s2UUG and tRNALeucmnm5UmAA however

cmnm5 appears as the final modification whereasin the remaining tRNAs the MnmEG products areconverted into 5-methylaminomethyl (mnm5)through the two-domain bi-functional enzymeMnmC MnmC(o) transforms cmnm5 into nm5whereas MnmC(m) converts nm5 into mnm5 thusproducing an atypical network of modificationpathways We investigate the activities and tRNAspecificity of MnmEG and the MnmC domains theability of tRNAs to follow the ammonium or glycinepathway and the effect of mnmC mutations ongrowth We demonstrate that the two MnmCdomains function independently of each other andthat tRNAGln

cmnm5s2UUG and tRNALeucmnm5UmAA are sub-

strates for MnmC(m) but not MnmC(o) Synthesisof mnm5s2U by MnmEG-MnmC in vivo avoidsbuild-up of intermediates in tRNALys

mnm5s2UUU We alsoshow that MnmEG can modify all the tRNAs via theammonium pathway Strikingly the net output of theMnmEG pathways in vivo depends on growth

conditions and tRNA species Loss of any MnmCactivity has a biological cost under specificconditions

INTRODUCTION

Transfer RNAs (tRNAs) are highly and diverselymodified and each has a unique pattern of modification(1ndash3) These modifications are post-transcriptionallyintroduced at precise positions by specific enzymes andplay important roles in folding stability identity and inthe functions of tRNAs

Modifications at the wobble position of the anticodoncontribute to the accuracy and efficiency of protein syn-thesis (134) Changes in the modification levels of thewobble position affect the synthesis of specific proteinsand also lead to complex phenotypes through unknownmechanisms (35ndash10)

Wobble modifications are often complex and require fortheir synthesis the participation of several enzymes (3)Even though all of the enzymes involved in a specificmodification pathway are known it is unclear how theiractions are modulated and coordinated to produce thefinal modification Progress in this area may lead to abetter understanding of tRNA biology

Proteins MnmE and MnmG (formerly TrmE andGidA respectively) are evolutionary conserved from bac-teria to eukaryotic organelles In Escherichia coli theyare involved in the modification of the wobble uridine

of tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG

tRNALeucmnm5UmAA tRNAArg

mnm5UCU and tRNAGlymnm5UCC all

To whom correspondence should be addressed Tel +34 963289680 Fax +34 963289701 Email armengodcipfesCorrespondence may also be addressed to Ismaıl Moukadiri Tel +34 963289681 Fax +34 963289701 Email IsmailMoukadiriuves

The authors wish it to be known that in their opinion the first two authors should be regarded as Joint First Authors

2602ndash2623 Nucleic Acids Research 2014 Vol 42 No 4 Published online 30 November 2013doi101093nargkt1228

The Author(s) 2013 Published by Oxford University PressThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (httpcreativecommonsorglicensesby30) whichpermits unrestricted reuse distribution and reproduction in any medium provided the original work is properly cited

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

of which read NNAG codons of the split codon boxes

with the exception of tRNAGlymnm5UCC which reads codons

of the glycine family box (11) MnmE and MnmG aredimeric and form a functional a2b2 heterotetramericcomplex (MnmEG) in which both proteins are inter-dependent (1213) MnmE is a GTP- and tetrahydrofolate(THF)-binding protein whereas MnmG is a FAD- andNADH-binding protein (14ndash18) The MnmEG complexcatalyzes the addition of the aminomethyl (nm) andcarboxymethylaminomethyl (cmnm) groups to position 5of the wobble uridine using ammonium and glycine re-spectively [Figure 1 (13)] Both reactions require GTP andFAD as well as NADH if FAD is limiting in the in vitroreaction Moreover a THF derivative likely methylene-THF serves as the donor of the methylene carbon that isdirectly bonded to the C5 atom of U34 However thedetailed mechanism of the basic reaction has not yetbeen demonstrated (1113) According to current modelsGTP hydrolysis by the MnmE G-domain which is locateddistant from the MnmEG active center leads to structuralrearrangements in the MnmEG complex which are essen-tial for tRNA modification (1119)

The wobble uridine (U34) of tRNAs modifiedby MnmEG may be further modified by other enzymes

at position 5 or other positions In tRNALysmnm5s2UUU

tRNAGlumnm5s2UUC and tRNAGln

cmnm5s2UUG thiolation atposition 2 of the U34 is performed by MnmA (20)

whereas in tRNALeucmnmUmAA TrmL methylates the 20-OH

group of the U-ribose (21) Therefore some tRNA sub-strates for the MnmEG complex are also substrates forMnmA or TrmL

Moreover in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC the resultingproducts of MnmEG activity are not the final modifica-tions because these tRNAs carry the methylaminomethyl(mnm) group at position 5 of U34 Formation of themnm5-final group is mediated by the action of the two-domain bi-functional enzyme MnmC (22ndash26) Theoxidoreductase activity of the C-terminal domain heredesignated as MnmC(o) is responsible for the FAD-dependent deacetylation that transforms cmnm5U intonm5U whereas the methyltransferase activity of theN-terminal domain designated as MnmC(m) catalyzesthe SAM-dependent methylation that transforms nm5Uto mnm5U (Figure 1)

The crystal structure of MnmC consists of two globulardomains interacting with each other (27) The structurealso reveals that the two catalytic centers of MnmC(o)and MnmC(m) face opposite sides of the protein thusfavoring a model in which the two domains couldfunction in a relatively independent manner In factwhen two MnmC mutant proteins possessing a catalytic-ally dead MnmC(o) or MnmC(m) domain were mixedpartial recovery of mnm5-group synthesis was observed(25) However considering the relatively hydrophilicnature of the domain interface the possibility that con-formational changes within the entire MnmC protein mayoccur in vivo and affect its functioning cannot beexcluded A previous attempt to separately express both

MnmC domains revealed that MnmC(m) was solublewhereas MnmC(o) produced inclusion bodies suggestingthat MnmC(m) is required for the correct folding or struc-tural stability of MnmC(o) (25)How the activities of the MnmEG and MnmC enzymes

are organized and modulated remains unclear WhethertRNAs preferentially utilize one of the two MnmEGpathways the glycine or the ammonium pathway(Figure 1) in vivo depending on metabolic circumstanceshas not been investigated The fact that only cmnm5 butnot nm5 or mnm5 has been detected to date at position 5

of U34 in tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA suggeststhat both tRNAs do not use the ammonium pathway andare not substrates for MnmC(o) Accumulation of thenm5s2U nucleoside which may have a dual origin[Figure 1 (13)] has not been observed in totaltRNA purified from wild-type E coli cells (23)

Therefore the mnm5s2U synthesis in tRNALysmnm5s2UUU

and tRNAGlumnm5s2UUC appears to be organized to prevent

nm5s2U accumulation However the presence or absenceof cmnm5s2U as an intermediate in mnm5s2U biosynthesisis difficult to determine from nucleoside analysis of bulktRNA due to the natural occurrence of cmnm5s2U in

tRNAGlncmnm5s2UUG Biosynthetic tuning of mnm5s2U to

avoid the accumulation of intermediates could beachieved by either kinetic tuning of the activities ofMnmEG and MnmC which requires that the sequentialmodifications are performed at similar or increasing ratesor selective degradation of partially modified tRNAsA previous steady-state kinetic analysis of the activitiesof the full MnmC protein indicated that the MnmC(m)-dependent reaction (nm5

mnm5) occurs faster than theMnmC(o)-dependent reaction (cmnm5

nm5) or at asimilar rate at very high substrate (tRNA) concentrations(26) However this study did not take into account theefficiency of the reactions performed by MnmEG(Figure 1) which is crucial to understand the organizationof the mnm5s2U biosynthetic processModifications located within or adjacent to the anti-

codon are important for stabilization of codonndashanti-codon pairing as well as for restricting the dynamicsof the anticodon domain and shaping its architecture(14) Considering the network of pathways leading tomnm5s2U production (Figure 1) it is important to de-termine whether mutations affecting the differentactivities of MnmC [MnmC(o) or MnmC(m)] havedistinct biological consequences and to what extent anaccumulation of modification intermediates affects bac-terial biologyIn this study we investigated the activities and the

tRNA specificity of the MnmEG and MnmC enzymesin vivo and in vitro analyzed how the U34 modificationstatus is influenced by genetic and physiological condi-tions in bulk and specific tRNAs and assessed the effectsof mnmC mutations on bacterial growth Our study wasfacilitated by the cloning and separate expression of thetwo MnmC domains MnmC(o) and MnmC(m) Incontrast to a previous report (25) we demonstrate thatMnmC(o) can fold independently of MnmC(m) and thatthe separate domains exhibit similar kinetic properties

Nucleic Acids Research 2014 Vol 42 No 4 2603

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

(s2)U34

U34

(MnmA)

MnmEbullMnmG

cmnm5(s2)U34 nm5(s2)U34

mnm5(s2)U34

MnmC(o)

MnmC(m)

CH2-THFGTP

FADNADH

FAD

SAM

N

NH

O

S

OOR

O HO R

CH2NH2

N

NH

O

OOR

O HO R

S

CH2NHCH2COOH

N

NH

O

OOR

O HO R

S

N

NH

O

S

OOR

OHO R

CH2NHCH3

25

CH2NHCH2COOH

N

NH

O

OOR

O HO R

0

(cmnm5)U

CH2NHCH2COOH

N

NH

O

OOR

OCH3O R

0

(cmnm5)Um

TrmL

tRNALeucmnm5UmAA

Figure 1 Synthesis of mnm5(s2)U in E coli The MnmEG complex formed by proteins MnmE and MnmG is active on position 5 of the wobble

uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA tRNAArgmnm5UCU and tRNAGly

mnm5UCC MnmE is a three-domainprotein that binds GTP and methylenetetrahydrofolate (CH2-THF) whereas MnmG is a FAD- and NADH-binding protein The MnmC(o) activityof the bi-functional enzyme MnmC transforms cmnm5(s2)U to nm5(s2)U and the MnmC(m) activity of MnmC transforms nm5(s2)U to the final

product mnm5(s2)U Thiolation at position 2 of the wobble uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlncmnm5s2UUG is mediated by

the protein MnmA Modifications occur at the 5- and 2-positions independent of each other TrmL methylates the 20-OH group of the U-ribose in

tRNALeucmnm5UmAA (inset panel)

2604 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

to those of the full protein We also show that

tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA are substrates forMnmC(m) but not for MnmC(o) Our data suggest thatMnmEG andMnmC are kinetically tuned to produce only

the fully modified nucleoside mnm5U in tRNALysmnm5s2UUU

We demonstrate that all the tRNA substrates of MnmEGare modified in vitro through the ammonium pathwayHowever the net output of the ammonium and glycinepathways of MnmEG in vivo depends on growth condi-tions and tRNA species Finally we demonstrate that theloss of any MnmC activity has a biological cost underspecific conditions

MATERIALS AND METHODS

Bacterial strains plasmids primers media andgeneral techniques

Escherichia coli strains and plasmids are shown in Table 1A list of the oligonucleotides and primers used inthis work is provided in Supplementary Table S1Transduction with phage P1 was performed as previouslydescribed (28) Deletion of the mnmC(o) coding regionwas performed by targeted homologous recombination(29) using the primers MnmC(o)-F and MnmC(o)-RThe MnmC(o)-F primer introduced a TAA stop codon

Table 1 Strains and plasmids used in this study

Strain or plasmid Description References

Eschirichia coli strainsDEV16 (IC4385)a thi-1 rel-1 spoT1 lacZ105UAG valRmnmE-Q192X (1448)BW25113 (IC5136)a lacIq rrnBT14 lacZWJ16 hsdR514 araBADAH33 rhaBADLD78 (29)TH178 (IC5255)a mnmA1 fadRTn10 (45)MG1655(IC5356)a F- D TouatiTH48 (IC6017)a ara (lac-proB) nalA argEam rif thi fadLTn10 (23)TH49 (IC6018)a TH48 mnmC(m)-G68D (23)TH69 (IC6019)a TH48 mnmC-W131stop (23)IC4639 DEV16 mnmE+ bgl (Sal+) (12)IC5241 MG1655 mnmGTn10 [TetR] (12)IC5358 MG1655 mnmEkan (43)IC5397 P1 (IC5255) x IC4639 [IC4639 mnmA-Q233stop] M VillarroyaIC5827 BW25113 mnmEkan NBRP- JapanIC5854 BW25113 trmLkan (21)IC5937 IC4639 mnmAkan [or IC4639 DmnmA] (39)IC5975 BL21-DE3 mnmGkan (17)IC6010 BW25113 mnmCKan [or BW25113 DmnmC] (13)IC6166 IC5975 carrying pIC1446 (17)IC6222 P1 (IC6010) x IC4639 [IC4639 DmnmC)] This studyIC6374 IC4639 trmLkan [IC4639 DtrmL] (21)IC6411 P1 (IC5241) x IC6374 (trmLkan) [IC4639 DtrmL DmnmG] (21)IC6424 IC5358 carrying pIC684 (14)IC6587 P1 (TH49) x IC4639 [IC4639 mnmC(m)-G68D] This studyIC6588 P1 (TH49) x IC5937 [IC4639 DmnmA mnmC(m)-G68D] This studyIC6589 P1 (IC6010) x IC5397 [IC4639 mnmA-Q233stop DmnmC] This studyIC6629 BW25113 mnmC(o)cat [or BW25113 DmnmC(o)] This studyIC6725 P1 (IC6629) x IC5937 [IC4639 DmnmA DmnmC(o)] This studyPlasmids

pIC684 GST fusion of mnmE cloned in pGEX-2T (14)

pIC1083 Eschirichia coli tRNALysmnm5s2UUU cloned in pUC19 (34)

pIC1253 flag-mnmC cloned in pBAD-TOPO (13)pIC1339 flag-mnmC(o) cloned in pBAD-TOPO This studypIC1340 flag-mnmC(m) cloned in pBAD-TOPO This study

pIC1394 Eschirichia coli tRNACysGCA cloned into pUC19 SmaI site This study

pIC1395 E schirichia coli tRNAArgmnm5UCU cloned into pUC19 SmaI site This study

pIC1446 his-mnmG cloned in pET15b (17)

pIC1535 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBAD-TOPO This study

pIC1536 Eschirichia coli tRNAGlumnm5s2UUC cloned into pUC19 SmaI site This study

pIC1537 Eschirichia coli tRNAGlymnm5UCC cloned into pUC19 SmaI site This study

pIC1550 Eschirichia coli tRNALeucmnm5UmAA cloned into pUC19 SmaI site This study

pIC1577 pBSKrna (32)

pIC1617 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRV site This study

pIC1664 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRIPstI sites This study

pIC1665 Eschirichia coli tRNALeucmnm5UmAA cloned into pBSKrna EcoRIndashPstI site This study

pIC1677 his-mnmC(m) cloned in pBAD-TOPO This study

pIC1714 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBSKrna EcoRIndashPstI site This study

aName in our collection between brackets

Nucleic Acids Research 2014 Vol 42 No 4 2605

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 3: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

of which read NNAG codons of the split codon boxes

with the exception of tRNAGlymnm5UCC which reads codons

of the glycine family box (11) MnmE and MnmG aredimeric and form a functional a2b2 heterotetramericcomplex (MnmEG) in which both proteins are inter-dependent (1213) MnmE is a GTP- and tetrahydrofolate(THF)-binding protein whereas MnmG is a FAD- andNADH-binding protein (14ndash18) The MnmEG complexcatalyzes the addition of the aminomethyl (nm) andcarboxymethylaminomethyl (cmnm) groups to position 5of the wobble uridine using ammonium and glycine re-spectively [Figure 1 (13)] Both reactions require GTP andFAD as well as NADH if FAD is limiting in the in vitroreaction Moreover a THF derivative likely methylene-THF serves as the donor of the methylene carbon that isdirectly bonded to the C5 atom of U34 However thedetailed mechanism of the basic reaction has not yetbeen demonstrated (1113) According to current modelsGTP hydrolysis by the MnmE G-domain which is locateddistant from the MnmEG active center leads to structuralrearrangements in the MnmEG complex which are essen-tial for tRNA modification (1119)

The wobble uridine (U34) of tRNAs modifiedby MnmEG may be further modified by other enzymes

at position 5 or other positions In tRNALysmnm5s2UUU

tRNAGlumnm5s2UUC and tRNAGln

cmnm5s2UUG thiolation atposition 2 of the U34 is performed by MnmA (20)

whereas in tRNALeucmnmUmAA TrmL methylates the 20-OH

group of the U-ribose (21) Therefore some tRNA sub-strates for the MnmEG complex are also substrates forMnmA or TrmL

Moreover in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC the resultingproducts of MnmEG activity are not the final modifica-tions because these tRNAs carry the methylaminomethyl(mnm) group at position 5 of U34 Formation of themnm5-final group is mediated by the action of the two-domain bi-functional enzyme MnmC (22ndash26) Theoxidoreductase activity of the C-terminal domain heredesignated as MnmC(o) is responsible for the FAD-dependent deacetylation that transforms cmnm5U intonm5U whereas the methyltransferase activity of theN-terminal domain designated as MnmC(m) catalyzesthe SAM-dependent methylation that transforms nm5Uto mnm5U (Figure 1)

The crystal structure of MnmC consists of two globulardomains interacting with each other (27) The structurealso reveals that the two catalytic centers of MnmC(o)and MnmC(m) face opposite sides of the protein thusfavoring a model in which the two domains couldfunction in a relatively independent manner In factwhen two MnmC mutant proteins possessing a catalytic-ally dead MnmC(o) or MnmC(m) domain were mixedpartial recovery of mnm5-group synthesis was observed(25) However considering the relatively hydrophilicnature of the domain interface the possibility that con-formational changes within the entire MnmC protein mayoccur in vivo and affect its functioning cannot beexcluded A previous attempt to separately express both

MnmC domains revealed that MnmC(m) was solublewhereas MnmC(o) produced inclusion bodies suggestingthat MnmC(m) is required for the correct folding or struc-tural stability of MnmC(o) (25)How the activities of the MnmEG and MnmC enzymes

are organized and modulated remains unclear WhethertRNAs preferentially utilize one of the two MnmEGpathways the glycine or the ammonium pathway(Figure 1) in vivo depending on metabolic circumstanceshas not been investigated The fact that only cmnm5 butnot nm5 or mnm5 has been detected to date at position 5

of U34 in tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA suggeststhat both tRNAs do not use the ammonium pathway andare not substrates for MnmC(o) Accumulation of thenm5s2U nucleoside which may have a dual origin[Figure 1 (13)] has not been observed in totaltRNA purified from wild-type E coli cells (23)

Therefore the mnm5s2U synthesis in tRNALysmnm5s2UUU

and tRNAGlumnm5s2UUC appears to be organized to prevent

nm5s2U accumulation However the presence or absenceof cmnm5s2U as an intermediate in mnm5s2U biosynthesisis difficult to determine from nucleoside analysis of bulktRNA due to the natural occurrence of cmnm5s2U in

tRNAGlncmnm5s2UUG Biosynthetic tuning of mnm5s2U to

avoid the accumulation of intermediates could beachieved by either kinetic tuning of the activities ofMnmEG and MnmC which requires that the sequentialmodifications are performed at similar or increasing ratesor selective degradation of partially modified tRNAsA previous steady-state kinetic analysis of the activitiesof the full MnmC protein indicated that the MnmC(m)-dependent reaction (nm5

mnm5) occurs faster than theMnmC(o)-dependent reaction (cmnm5

nm5) or at asimilar rate at very high substrate (tRNA) concentrations(26) However this study did not take into account theefficiency of the reactions performed by MnmEG(Figure 1) which is crucial to understand the organizationof the mnm5s2U biosynthetic processModifications located within or adjacent to the anti-

codon are important for stabilization of codonndashanti-codon pairing as well as for restricting the dynamicsof the anticodon domain and shaping its architecture(14) Considering the network of pathways leading tomnm5s2U production (Figure 1) it is important to de-termine whether mutations affecting the differentactivities of MnmC [MnmC(o) or MnmC(m)] havedistinct biological consequences and to what extent anaccumulation of modification intermediates affects bac-terial biologyIn this study we investigated the activities and the

tRNA specificity of the MnmEG and MnmC enzymesin vivo and in vitro analyzed how the U34 modificationstatus is influenced by genetic and physiological condi-tions in bulk and specific tRNAs and assessed the effectsof mnmC mutations on bacterial growth Our study wasfacilitated by the cloning and separate expression of thetwo MnmC domains MnmC(o) and MnmC(m) Incontrast to a previous report (25) we demonstrate thatMnmC(o) can fold independently of MnmC(m) and thatthe separate domains exhibit similar kinetic properties

Nucleic Acids Research 2014 Vol 42 No 4 2603

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

(s2)U34

U34

(MnmA)

MnmEbullMnmG

cmnm5(s2)U34 nm5(s2)U34

mnm5(s2)U34

MnmC(o)

MnmC(m)

CH2-THFGTP

FADNADH

FAD

SAM

N

NH

O

S

OOR

O HO R

CH2NH2

N

NH

O

OOR

O HO R

S

CH2NHCH2COOH

N

NH

O

OOR

O HO R

S

N

NH

O

S

OOR

OHO R

CH2NHCH3

25

CH2NHCH2COOH

N

NH

O

OOR

O HO R

0

(cmnm5)U

CH2NHCH2COOH

N

NH

O

OOR

OCH3O R

0

(cmnm5)Um

TrmL

tRNALeucmnm5UmAA

Figure 1 Synthesis of mnm5(s2)U in E coli The MnmEG complex formed by proteins MnmE and MnmG is active on position 5 of the wobble

uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA tRNAArgmnm5UCU and tRNAGly

mnm5UCC MnmE is a three-domainprotein that binds GTP and methylenetetrahydrofolate (CH2-THF) whereas MnmG is a FAD- and NADH-binding protein The MnmC(o) activityof the bi-functional enzyme MnmC transforms cmnm5(s2)U to nm5(s2)U and the MnmC(m) activity of MnmC transforms nm5(s2)U to the final

product mnm5(s2)U Thiolation at position 2 of the wobble uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlncmnm5s2UUG is mediated by

the protein MnmA Modifications occur at the 5- and 2-positions independent of each other TrmL methylates the 20-OH group of the U-ribose in

tRNALeucmnm5UmAA (inset panel)

2604 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

to those of the full protein We also show that

tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA are substrates forMnmC(m) but not for MnmC(o) Our data suggest thatMnmEG andMnmC are kinetically tuned to produce only

the fully modified nucleoside mnm5U in tRNALysmnm5s2UUU

We demonstrate that all the tRNA substrates of MnmEGare modified in vitro through the ammonium pathwayHowever the net output of the ammonium and glycinepathways of MnmEG in vivo depends on growth condi-tions and tRNA species Finally we demonstrate that theloss of any MnmC activity has a biological cost underspecific conditions

MATERIALS AND METHODS

Bacterial strains plasmids primers media andgeneral techniques

Escherichia coli strains and plasmids are shown in Table 1A list of the oligonucleotides and primers used inthis work is provided in Supplementary Table S1Transduction with phage P1 was performed as previouslydescribed (28) Deletion of the mnmC(o) coding regionwas performed by targeted homologous recombination(29) using the primers MnmC(o)-F and MnmC(o)-RThe MnmC(o)-F primer introduced a TAA stop codon

Table 1 Strains and plasmids used in this study

Strain or plasmid Description References

Eschirichia coli strainsDEV16 (IC4385)a thi-1 rel-1 spoT1 lacZ105UAG valRmnmE-Q192X (1448)BW25113 (IC5136)a lacIq rrnBT14 lacZWJ16 hsdR514 araBADAH33 rhaBADLD78 (29)TH178 (IC5255)a mnmA1 fadRTn10 (45)MG1655(IC5356)a F- D TouatiTH48 (IC6017)a ara (lac-proB) nalA argEam rif thi fadLTn10 (23)TH49 (IC6018)a TH48 mnmC(m)-G68D (23)TH69 (IC6019)a TH48 mnmC-W131stop (23)IC4639 DEV16 mnmE+ bgl (Sal+) (12)IC5241 MG1655 mnmGTn10 [TetR] (12)IC5358 MG1655 mnmEkan (43)IC5397 P1 (IC5255) x IC4639 [IC4639 mnmA-Q233stop] M VillarroyaIC5827 BW25113 mnmEkan NBRP- JapanIC5854 BW25113 trmLkan (21)IC5937 IC4639 mnmAkan [or IC4639 DmnmA] (39)IC5975 BL21-DE3 mnmGkan (17)IC6010 BW25113 mnmCKan [or BW25113 DmnmC] (13)IC6166 IC5975 carrying pIC1446 (17)IC6222 P1 (IC6010) x IC4639 [IC4639 DmnmC)] This studyIC6374 IC4639 trmLkan [IC4639 DtrmL] (21)IC6411 P1 (IC5241) x IC6374 (trmLkan) [IC4639 DtrmL DmnmG] (21)IC6424 IC5358 carrying pIC684 (14)IC6587 P1 (TH49) x IC4639 [IC4639 mnmC(m)-G68D] This studyIC6588 P1 (TH49) x IC5937 [IC4639 DmnmA mnmC(m)-G68D] This studyIC6589 P1 (IC6010) x IC5397 [IC4639 mnmA-Q233stop DmnmC] This studyIC6629 BW25113 mnmC(o)cat [or BW25113 DmnmC(o)] This studyIC6725 P1 (IC6629) x IC5937 [IC4639 DmnmA DmnmC(o)] This studyPlasmids

pIC684 GST fusion of mnmE cloned in pGEX-2T (14)

pIC1083 Eschirichia coli tRNALysmnm5s2UUU cloned in pUC19 (34)

pIC1253 flag-mnmC cloned in pBAD-TOPO (13)pIC1339 flag-mnmC(o) cloned in pBAD-TOPO This studypIC1340 flag-mnmC(m) cloned in pBAD-TOPO This study

pIC1394 Eschirichia coli tRNACysGCA cloned into pUC19 SmaI site This study

pIC1395 E schirichia coli tRNAArgmnm5UCU cloned into pUC19 SmaI site This study

pIC1446 his-mnmG cloned in pET15b (17)

pIC1535 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBAD-TOPO This study

pIC1536 Eschirichia coli tRNAGlumnm5s2UUC cloned into pUC19 SmaI site This study

pIC1537 Eschirichia coli tRNAGlymnm5UCC cloned into pUC19 SmaI site This study

pIC1550 Eschirichia coli tRNALeucmnm5UmAA cloned into pUC19 SmaI site This study

pIC1577 pBSKrna (32)

pIC1617 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRV site This study

pIC1664 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRIPstI sites This study

pIC1665 Eschirichia coli tRNALeucmnm5UmAA cloned into pBSKrna EcoRIndashPstI site This study

pIC1677 his-mnmC(m) cloned in pBAD-TOPO This study

pIC1714 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBSKrna EcoRIndashPstI site This study

aName in our collection between brackets

Nucleic Acids Research 2014 Vol 42 No 4 2605

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 4: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

(s2)U34

U34

(MnmA)

MnmEbullMnmG

cmnm5(s2)U34 nm5(s2)U34

mnm5(s2)U34

MnmC(o)

MnmC(m)

CH2-THFGTP

FADNADH

FAD

SAM

N

NH

O

S

OOR

O HO R

CH2NH2

N

NH

O

OOR

O HO R

S

CH2NHCH2COOH

N

NH

O

OOR

O HO R

S

N

NH

O

S

OOR

OHO R

CH2NHCH3

25

CH2NHCH2COOH

N

NH

O

OOR

O HO R

0

(cmnm5)U

CH2NHCH2COOH

N

NH

O

OOR

OCH3O R

0

(cmnm5)Um

TrmL

tRNALeucmnm5UmAA

Figure 1 Synthesis of mnm5(s2)U in E coli The MnmEG complex formed by proteins MnmE and MnmG is active on position 5 of the wobble

uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA tRNAArgmnm5UCU and tRNAGly

mnm5UCC MnmE is a three-domainprotein that binds GTP and methylenetetrahydrofolate (CH2-THF) whereas MnmG is a FAD- and NADH-binding protein The MnmC(o) activityof the bi-functional enzyme MnmC transforms cmnm5(s2)U to nm5(s2)U and the MnmC(m) activity of MnmC transforms nm5(s2)U to the final

product mnm5(s2)U Thiolation at position 2 of the wobble uridine (U34) in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlncmnm5s2UUG is mediated by

the protein MnmA Modifications occur at the 5- and 2-positions independent of each other TrmL methylates the 20-OH group of the U-ribose in

tRNALeucmnm5UmAA (inset panel)

2604 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

to those of the full protein We also show that

tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA are substrates forMnmC(m) but not for MnmC(o) Our data suggest thatMnmEG andMnmC are kinetically tuned to produce only

the fully modified nucleoside mnm5U in tRNALysmnm5s2UUU

We demonstrate that all the tRNA substrates of MnmEGare modified in vitro through the ammonium pathwayHowever the net output of the ammonium and glycinepathways of MnmEG in vivo depends on growth condi-tions and tRNA species Finally we demonstrate that theloss of any MnmC activity has a biological cost underspecific conditions

MATERIALS AND METHODS

Bacterial strains plasmids primers media andgeneral techniques

Escherichia coli strains and plasmids are shown in Table 1A list of the oligonucleotides and primers used inthis work is provided in Supplementary Table S1Transduction with phage P1 was performed as previouslydescribed (28) Deletion of the mnmC(o) coding regionwas performed by targeted homologous recombination(29) using the primers MnmC(o)-F and MnmC(o)-RThe MnmC(o)-F primer introduced a TAA stop codon

Table 1 Strains and plasmids used in this study

Strain or plasmid Description References

Eschirichia coli strainsDEV16 (IC4385)a thi-1 rel-1 spoT1 lacZ105UAG valRmnmE-Q192X (1448)BW25113 (IC5136)a lacIq rrnBT14 lacZWJ16 hsdR514 araBADAH33 rhaBADLD78 (29)TH178 (IC5255)a mnmA1 fadRTn10 (45)MG1655(IC5356)a F- D TouatiTH48 (IC6017)a ara (lac-proB) nalA argEam rif thi fadLTn10 (23)TH49 (IC6018)a TH48 mnmC(m)-G68D (23)TH69 (IC6019)a TH48 mnmC-W131stop (23)IC4639 DEV16 mnmE+ bgl (Sal+) (12)IC5241 MG1655 mnmGTn10 [TetR] (12)IC5358 MG1655 mnmEkan (43)IC5397 P1 (IC5255) x IC4639 [IC4639 mnmA-Q233stop] M VillarroyaIC5827 BW25113 mnmEkan NBRP- JapanIC5854 BW25113 trmLkan (21)IC5937 IC4639 mnmAkan [or IC4639 DmnmA] (39)IC5975 BL21-DE3 mnmGkan (17)IC6010 BW25113 mnmCKan [or BW25113 DmnmC] (13)IC6166 IC5975 carrying pIC1446 (17)IC6222 P1 (IC6010) x IC4639 [IC4639 DmnmC)] This studyIC6374 IC4639 trmLkan [IC4639 DtrmL] (21)IC6411 P1 (IC5241) x IC6374 (trmLkan) [IC4639 DtrmL DmnmG] (21)IC6424 IC5358 carrying pIC684 (14)IC6587 P1 (TH49) x IC4639 [IC4639 mnmC(m)-G68D] This studyIC6588 P1 (TH49) x IC5937 [IC4639 DmnmA mnmC(m)-G68D] This studyIC6589 P1 (IC6010) x IC5397 [IC4639 mnmA-Q233stop DmnmC] This studyIC6629 BW25113 mnmC(o)cat [or BW25113 DmnmC(o)] This studyIC6725 P1 (IC6629) x IC5937 [IC4639 DmnmA DmnmC(o)] This studyPlasmids

pIC684 GST fusion of mnmE cloned in pGEX-2T (14)

pIC1083 Eschirichia coli tRNALysmnm5s2UUU cloned in pUC19 (34)

pIC1253 flag-mnmC cloned in pBAD-TOPO (13)pIC1339 flag-mnmC(o) cloned in pBAD-TOPO This studypIC1340 flag-mnmC(m) cloned in pBAD-TOPO This study

pIC1394 Eschirichia coli tRNACysGCA cloned into pUC19 SmaI site This study

pIC1395 E schirichia coli tRNAArgmnm5UCU cloned into pUC19 SmaI site This study

pIC1446 his-mnmG cloned in pET15b (17)

pIC1535 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBAD-TOPO This study

pIC1536 Eschirichia coli tRNAGlumnm5s2UUC cloned into pUC19 SmaI site This study

pIC1537 Eschirichia coli tRNAGlymnm5UCC cloned into pUC19 SmaI site This study

pIC1550 Eschirichia coli tRNALeucmnm5UmAA cloned into pUC19 SmaI site This study

pIC1577 pBSKrna (32)

pIC1617 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRV site This study

pIC1664 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRIPstI sites This study

pIC1665 Eschirichia coli tRNALeucmnm5UmAA cloned into pBSKrna EcoRIndashPstI site This study

pIC1677 his-mnmC(m) cloned in pBAD-TOPO This study

pIC1714 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBSKrna EcoRIndashPstI site This study

aName in our collection between brackets

Nucleic Acids Research 2014 Vol 42 No 4 2605

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 5: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

to those of the full protein We also show that

tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA are substrates forMnmC(m) but not for MnmC(o) Our data suggest thatMnmEG andMnmC are kinetically tuned to produce only

the fully modified nucleoside mnm5U in tRNALysmnm5s2UUU

We demonstrate that all the tRNA substrates of MnmEGare modified in vitro through the ammonium pathwayHowever the net output of the ammonium and glycinepathways of MnmEG in vivo depends on growth condi-tions and tRNA species Finally we demonstrate that theloss of any MnmC activity has a biological cost underspecific conditions

MATERIALS AND METHODS

Bacterial strains plasmids primers media andgeneral techniques

Escherichia coli strains and plasmids are shown in Table 1A list of the oligonucleotides and primers used inthis work is provided in Supplementary Table S1Transduction with phage P1 was performed as previouslydescribed (28) Deletion of the mnmC(o) coding regionwas performed by targeted homologous recombination(29) using the primers MnmC(o)-F and MnmC(o)-RThe MnmC(o)-F primer introduced a TAA stop codon

Table 1 Strains and plasmids used in this study

Strain or plasmid Description References

Eschirichia coli strainsDEV16 (IC4385)a thi-1 rel-1 spoT1 lacZ105UAG valRmnmE-Q192X (1448)BW25113 (IC5136)a lacIq rrnBT14 lacZWJ16 hsdR514 araBADAH33 rhaBADLD78 (29)TH178 (IC5255)a mnmA1 fadRTn10 (45)MG1655(IC5356)a F- D TouatiTH48 (IC6017)a ara (lac-proB) nalA argEam rif thi fadLTn10 (23)TH49 (IC6018)a TH48 mnmC(m)-G68D (23)TH69 (IC6019)a TH48 mnmC-W131stop (23)IC4639 DEV16 mnmE+ bgl (Sal+) (12)IC5241 MG1655 mnmGTn10 [TetR] (12)IC5358 MG1655 mnmEkan (43)IC5397 P1 (IC5255) x IC4639 [IC4639 mnmA-Q233stop] M VillarroyaIC5827 BW25113 mnmEkan NBRP- JapanIC5854 BW25113 trmLkan (21)IC5937 IC4639 mnmAkan [or IC4639 DmnmA] (39)IC5975 BL21-DE3 mnmGkan (17)IC6010 BW25113 mnmCKan [or BW25113 DmnmC] (13)IC6166 IC5975 carrying pIC1446 (17)IC6222 P1 (IC6010) x IC4639 [IC4639 DmnmC)] This studyIC6374 IC4639 trmLkan [IC4639 DtrmL] (21)IC6411 P1 (IC5241) x IC6374 (trmLkan) [IC4639 DtrmL DmnmG] (21)IC6424 IC5358 carrying pIC684 (14)IC6587 P1 (TH49) x IC4639 [IC4639 mnmC(m)-G68D] This studyIC6588 P1 (TH49) x IC5937 [IC4639 DmnmA mnmC(m)-G68D] This studyIC6589 P1 (IC6010) x IC5397 [IC4639 mnmA-Q233stop DmnmC] This studyIC6629 BW25113 mnmC(o)cat [or BW25113 DmnmC(o)] This studyIC6725 P1 (IC6629) x IC5937 [IC4639 DmnmA DmnmC(o)] This studyPlasmids

pIC684 GST fusion of mnmE cloned in pGEX-2T (14)

pIC1083 Eschirichia coli tRNALysmnm5s2UUU cloned in pUC19 (34)

pIC1253 flag-mnmC cloned in pBAD-TOPO (13)pIC1339 flag-mnmC(o) cloned in pBAD-TOPO This studypIC1340 flag-mnmC(m) cloned in pBAD-TOPO This study

pIC1394 Eschirichia coli tRNACysGCA cloned into pUC19 SmaI site This study

pIC1395 E schirichia coli tRNAArgmnm5UCU cloned into pUC19 SmaI site This study

pIC1446 his-mnmG cloned in pET15b (17)

pIC1535 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBAD-TOPO This study

pIC1536 Eschirichia coli tRNAGlumnm5s2UUC cloned into pUC19 SmaI site This study

pIC1537 Eschirichia coli tRNAGlymnm5UCC cloned into pUC19 SmaI site This study

pIC1550 Eschirichia coli tRNALeucmnm5UmAA cloned into pUC19 SmaI site This study

pIC1577 pBSKrna (32)

pIC1617 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRV site This study

pIC1664 Eschirichia coli tRNALysmnm5s2UUU cloned in pBSKrna EcoRIPstI sites This study

pIC1665 Eschirichia coli tRNALeucmnm5UmAA cloned into pBSKrna EcoRIndashPstI site This study

pIC1677 his-mnmC(m) cloned in pBAD-TOPO This study

pIC1714 Eschirichia coli tRNAGlncmnm5s2UUG cloned into pBSKrna EcoRIndashPstI site This study

aName in our collection between brackets

Nucleic Acids Research 2014 Vol 42 No 4 2605

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 6: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

at the end of the MnmC(m) coding region which islocated in the 50-terminal region of the mnmC geneDeletion of the mnmC(o) coding sequence was confirmedby PCR and DNA sequencing The resulting strain wasnamed IC6629 DNA sequences coding for the MnmC(o)(250ndash668 aa) and MnmC(m) domains (1ndash250 aa) fusedto the C-terminal end of a Flag-epitope were amplified bypolymerase chain reaction (PCR) from genomic DNA ofE coli MG1655 using the specific primer pairs Flag-MnmC(o)FFlag-MnmC(o)R and Flag-MnmC(m)FFlag-MnmC(m)R respectively The amplicons werecloned into pBAD TOPO TA for expression undercontrol of the AraC-PBAD system The resultingplasmids were transformed into a mnmCkan strainunless otherwise specified A DNA fragment encodingthe MnmC(m) domain fused at its N-terminal end to a6x-His-epitope was cloned into the same vector afterPCR amplification with the His-MnmC(m)RHis-MnmC(m)F primer pair LBT (LB broth containing40mgl thymine) and LAT (LBT containing 20 g ofDifco agar per liter) were used for routine culturesand plating of E coli respectively unless otherwisespecified When required antibiotics were added at thefollowing final concentrations 100 mgml ampicillin125mgml tetracycline 25 mgml chloramphenicol and80 mgml kanamycin Cell growth was monitored bymeasuring the optical density of the cultures at 600 nm(OD600)

Growth rate determinations and competition experiments

To determine bacterial growth rates overnight cultureswere diluted 1100 in fresh LBT medium and incubatedat 37C with shaking The doubling time was measured bymonitoring the optical density of the culture at 600 nmSamples were taken from exponentially growing culturesafter at least 10 generations of steady-state growth Thegrowth rate was calculated as the doubling time of eachculture in the steady-state log phase by linear regressionCompetition experiments were performed as previouslyreported (21) Briefly the reference strain (BW25113)and the strain to be tested (TH48 TH49 or TH69) weregrown separately to stationary phase by incubation at37C Equal volumes of both strains were mixed and asample was immediately taken to count viable cells onLAT plates with and without the antibiotic required toestimate the content of each strain (tetracycline toidentify strains of the TH48 background) Six cycles of24-h growth at 37C (with shaking) were performedby diluting mixed cultures of 11000 in LBT mediumAfter the sixth cycle the mixed culture was analyzedfor its strain content as before The final ratio wascalculated as the ratio of the number of colony formingunits (CFU) per milliliter recovered on LAT supple-mented with tetracycline versus CFU per milliliteron LAT

Purification of recombinant proteins

Routinely a single colony was inoculated into 5ml LBTsupplemented with adequate antibiotics and grown over-night at 37C The pre-culture was then inoculated (1100

dilution) into 1000ml LBT plus antibiotic and grown untilthe OD600 reached 05 Protein expression was inducedby the addition of 02 arabinose (Flag-tagged proteins)The cultures were grown for an additional 3ndash4 h at 30Cwith gentle shaking harvested by centrifugation (4500gfor 15min at 4C) washed with TBS buffer (50mMTrisndashHCl pH 74 150mM NaCl and 5mM MgCl2) andstored at 20C For enzyme purification frozen cellswere thawed on ice resuspended in 10ml of lysis buffer(TBS containing 1mM PMSF and 1mM EDTA) and dis-rupted by sonication The lysate was centrifuged at10 000g for 45min at 4C and the supernatant wasmixed with 03ndash06ml of pre-equilibrated anti-Flagagarose resin (ANTI-FLAG M2 Affinity Gel A2220-Sigma) and incubated at 4C for 1 h with mild shakingInstructions from the manufacturer were followed forwashing and eluting Flag-tagged proteins The fractionscontaining the proteins were pooled concentrated withAmicon Ultra-15 30k devices in TBS buffer and storedin 50-ml fractions with 15 glycerol at 20C The His-MnmC(m) domain was purified using Clontechrsquos TALONMetal Affinity Resin according to the manufacturerrsquos in-structions The MnmE and MnmG proteins were purifiedas previously described (1417) Protein concentrationswere determined with a NanoDrop spectrophotometer at280 nm The purity of all enzymes was gt95 as estimatedby sodium dodecyl sulphatendashpolyacrylamide gel electro-phoresis (SDSndashPAGE) and coomassie blue staining

FAD cofactor analysis

FAD was released from recombinant proteins by heatingat 75C for 5min in the dark and analyzed by high-performance liquid chromatography (HPLC) BrieflyHPLC separation was achieved with a Synergi 4uFusion column (25 cm 46mm 5 mm) using a gradientof 5mM ammonium acetate pH 65 to acetonitrile 50and water 50 over 35min at a flow rate of 06mlminFlavins were detected by fluorescence emission (525 nm)using a Waters 474 scanning fluorescence detector set at450 nm for excitation

Stability of MnmC recombinant proteins

Strain IC6010 carrying pIC1253 pIC1339 or pIC1340 wasgrown overnight in LBT with ampicillin at 37COvernight cultures were diluted 1100 in the samemedium supplemented with 01 L-arabinose (inducerof the AraC-PBAD system) and incubated at 37C for2 h To halt the expression of the MnmC recombinantproteins the cells were recovered by centrifugation at3000g for 10min washed once with fresh LBT mediumresuspended in the same volume of pre-warmed LBT con-taining ampicillin and 1 glucose (repressor of theAraC-PBAD system) and incubated at 37C Sampleswere taken several times after the addition of glucoseThe cells were lysed by brief ultrasonication and thelysate was centrifuged at 10 000g for 10min at 4C Thesoluble fractions (50ndash150mg of total proteins) wereanalyzed by western blotting using anti-Flag and anti-GroEL antibodies

2606 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 7: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

Gel filtration analysis of protein interactions

To investigate whether Flag-MnmC(o) interacts withFlag-MnmC(m) each protein was mixed at a final concen-tration of 5 mM in a final volume of 100 ml in TBS buffercontaining 5mM DTT and 3 glycerol and incubated atroom temperature for 2 h The same procedure was usedto study the interaction of the full MnmC protein withMnmE or MnmG The samples were analyzed by gel fil-tration using a Superdex 75 HR (MnmC domains) or aSuperdex 200 HR (MnmCMnmEMnmG) column at aflow rate of 07 or 03mlmin respectively in TBS Buffercontaining 5mM DTT Gel filtration markers were usedto calibrate the columns The proteins were detected byultraviolet (UV) absorbance at 280 nm

Surface plasmon resonance evaluation of theMnmC(o)ndashMnmC(m) interaction

Surface plasmon resonance (SPR)-based kinetic analysiswas used to determine the affinity between the recombin-ant MnmC(o) and MnmC(m) proteins An anti-Hismonoclonal antibody (Roche 100 ngml in 10mMsodium acetate pH 45) was immobilized onto a CM-5sensor chip (Biacore AB Uppsala Sweden) at 7000 res-onance units (RUs) using an amine coupling kit (BiacoreAB) according to the manufacturerrsquos instructions His-MnmC(m) at 2 mM in TBS buffer containing 0005Tween-20 surfactant was immobilized by capturing(600 RUs) Subsequently various concentrations ofFlag-MnmC(o) protein in TBS buffer were passed overthe sensor chip at a flow rate of 30 mlmin at 25C andthe interactions were monitored for 1min The sensorsurface was washed with TBS buffer to detect dissociationand then regenerated with a pulse of 5mM NaOH (10 s at60 mlmin) The data were evaluated with BiaEvaluation31 Software (Biacore AB Uppsala Sweden)

Isolation of bulk tRNA from E coli and reverse-phaseHPLC analysis of nucleosides

Total tRNA purification and analysis of nucleosides byreverse-phase HPLC were performed as described previ-ously (133536) HPLC analysis was monitored at appro-priate wavelengths to achieve optimal adsorption of thetarget nucleosides 314 nm for thiolated nucleosides and254 nm for non-thiolated nucleosides The nucleosideswere identified according to their UV spectra (35) andby comparison with appropriate controls

Isolation of specific chimeric and native tRNAs

The E coli tRNALysmnm5s2UUU tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA genes were cloned into pBSKrna (37)

digested with either EcoRV (to produce chimeric tRNAsin which the E coli tRNA is inserted into a human cyto-

solic tRNALys3 sequence that is used as a scaffold) or

EcoRI and PstI to produce a tRNA without the scaffold(overexpressed lsquonativersquo tRNAs) The overproduction oftRNAs in strains transformed with pBSKrna derivativeplasmids was performed as described previously (37)Specific chimeric tRNAs were purified from bulk tRNAby the Chaplet Column Chromatography method (38)

using a biotinylated DNA probe that is complementary

to the scaffold human cytosolic tRNALys3 moiety

(common to all chimeric tRNAs) The probe wasimmobilized on a HiTrap Streptavidin HP column Thesame approach was used to purify overexpressed lsquonativersquoand true native tRNAs using biotinylated DNA probesthat are complementary to the specific sequence of eachE coli tRNA (Supplementary Table S1)

In vitro transcription of E coli tRNAs

The E coli genes encoding tRNACysGCA tRNAArg

mnm5UCU

tRNAGlumnm5s2UUC tRNAGly

mnm5UCC and tRNALeucmnm5UmAA

were PCR-amplified from genomic DNA using lsquoventrsquopolymerase and the primers Cys-FCys-R Arg-FArg-RGlu-FGlu-R Gly-FGly-R and Leu-FLeu-R respect-ively The amplicons were cloned into a SmaI-linearizedpUC19 plasmid to produce pIC1394 pIC1395 pIC1536pIC1550 and pIC1537 The E coli gene encoding

tRNAGlncmnm5s2UUG was PCR-amplified from genomic

DNA using expand-long polymerase and the primersGln-F and Gln-R and cloned into pBAD-TOPO The

plasmid pIC1083 (containing the tRNALysmnm5s2UUU gene)

a derivative of pUC19 was a gift from Dr Tamura (34)Unmodified E coli tRNAs were prepared by in vitro tran-scription from BstNI-digested plasmids (pIC1083pIC1394 and pIC1395) and HindIII-digested plasmids(pIC1535 pIC1536 pIC1550 and pIC1537) using theRiboprobe T7 transcription kit (PROMEGA) and2ndash5 mg of each digested plasmid as a DNA template in a50-ml reaction mix

In vitro and in vivo activity of the recombinant MnmC(o)and MnmC(m) proteins

To analyze the in vitro activity of the recombinantproteins total tRNA (40mg) from a mnmC-W131stop ormnmC(m)-G68D mutant was incubated in a 200-mlreaction mixture containing 50mM TrisndashCl (pH 80)50mM ammonium acetate 05mM FAD or 05mM S-Adenosyl-L-Methionine (SAM) 5 glycerol and 2 mMof the purified protein [Flag-MnmC Flag-MnmC(o) orFlag-MnmC(m)] for 40min at 37C tRNA was recoveredby phenol extraction and ethanol precipitation and sub-sequently treated with nuclease P1 and E coli alkalinephosphatase The resulting nucleosides were analyzedby reverse-phase HPLC as described (1330) For in vivocomplementation studies overnight cultures of strainsmnmC-W131stop or mnmC(m)-G68D containingpBAD-TOPO or derivative plasmids (pIC1340 andpIC1339) were diluted 1100 in 100ml LBT containing02 arabinose and grown to an OD600 of 05 The cellswere recovered by centrifugation at 4500g for 15min at4C Bulk tRNA was obtained and analyzed by HPLC asdescribed above

Kinetic analysis of MnmC- and MnmEG-catalyzedmodifications

To assay MnmC(o) or MnmC(m) activity the reactionmixture (100 ml) contained 50mM TrisndashHCl (pH 8)

Nucleic Acids Research 2014 Vol 42 No 4 2607

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 8: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

50mM ammonium acetate 3 glycerol 2mM NaCl73 mM MgCl2 tRNA (05ndash5 mM) Flag-tagged protein(25 nM) and 100 mM FAD or SAM (depending on theMnmC activity to be determined) The mixtures werepre-incubated at 37C for 3min before addition ofproteins All reactions were performed at 37C The reac-tions were halted after 60ndash90 s of incubation by adding100ml of 03M sodium acetate pH 52 To assay theammonium-dependent MnmEG activity with respect totRNA concentration the reaction mixture (100 ml) con-tained 100mM TrisndashHCl pH 8 100mM ammoniumacetate 5mM MgCl2 5 glycerol 5mM DTT 05mMFAD 2mM GTP 1mM methylene-THF 10 mg bovineserum albumin (BSA) and tRNA (01ndash2 mM) Thereaction was initiated by the addition of 01 mMMnmEMnmG complex obtained as previously described(13) and stopped after 2min of incubation at 37C byadding 100ml 03M sodium acetate pH 52 In all casestRNA was finally recovered by phenol extraction andethanol precipitation and treated with nuclease P1 andE coli alkaline phosphatase The resulting nucleosideswere analyzed by reverse-phase HPLC The area of thesynthesized nucleoside was calculated and its amountwas extrapolated from standard curves that wereprepared using chemically synthesized nucleosides (AMalkiewicz University of Lodz Poland) over a range of0ndash250 ng Experiments were performed in triplicate Todetermine the Vmax and Km values the data were fittedto the MichaelisndashMenten equation using nonlinear regres-sion (GraphPad Prism v40)

Analysis of the tRNA substrate specificity of MnmC(o)and MnmC(m) in vitro

To analyze the specificity of MnmC(o) and MnmC(m)in vitro the tRNA (10ndash15 mg) was first modified usingthe MnmEG complex through the ammonium andglycine pathways as described previously (13) Themodified tRNA was phenolized ethanol precipitated andincubated with 2 mM MnmC(o) or MnmC(m) domains in200ml (total volume) of MnmC buffer containing 50mMTrisndashHCl (pH 80) 50mM ammonium acetate 05mMFAD [for the MnmC(o) assay] or 05mM SAM (for theMnmC(m) assay) and 5 glycerol After incubating for40min at 37C the tRNA was recovered by phenolizationand ethanol precipitation and the tRNA was then treatedwith nuclease P1 and E coli alkaline phosphatase Theresulting nucleosides were analyzed by reverse-phaseHPLC as described above

Acid resistance assay

The acid resistance experiments were performed essen-tially as previously reported (39) Briefly strains weregrown in LBT containing 04 glucose to stationaryphase The cultures were then diluted 11000 into EGmedium [minimal E medium containing 04 glucose(40)] pH 20 supplemented or not with 07mM glutam-ate Samples were obtained at 0 1 2 3 and 4 h post acidchallenge and spotted on LAT plates

RESULTS

General roles of the MnmEG and MnmC enzymes in themodification status of tRNAs

In a previous study we demonstrated that MnmEG cata-lyzes two different reactions in vitro and produces nm5 andcmnm5 using ammonium and glycine respectively as sub-strates (13) We also reported the presence of cmnm5s2Uand nm5s2U in total tRNA purified from an mnmC nullmutant However it has been suggested that the forma-tion of these intermediates is sensitive to growth condi-tions and specific strains which might explain why theywere not detected in other studies (232641)

To obtain a clear picture of the functional activity ofenzymes MnmEG and MnmC during exponential growth(mid-log phase OD600 of 06) in LBT we analyzed theHPLC profile of bulk tRNA isolated from several strainsand different genetic backgrounds (Table 2 andSupplementary Figure S1) In the HPLC analysis absorb-ance was monitored at 314 nm to maximize the detectionof thiolated nucleosides We consistently detected a smallamount (10ndash20) of cmnm5s2U in the wild-typestrains of three different backgrounds TH48 BW25113and MG1655 (Table 2) which may originate

from tRNAGlncmnm5s2UUG or might be an intermediate of

the final product mnm5s2U that is present in

tRNALysmnm5s2UUU and tRNAGlu

mnm5s2UUC Notwithstandingmnm5s2U was the major final product (80ndash90)We were unable to identify nucleosides mnm5U

and cmnm5Um (which are present in tRNAArgmnm5UCU

tRNAGlymnm5UCC and tRNALeu

cmnm5UmAA) when tRNA hydrol-ysates were monitored at 254 nm as a comparison of chro-matograms of total tRNA purified from wild-type strainsand their mnmE or mnmG derivatives did not allow thedetection of any peaks attributable to those nucleosides inthe wild-type chromatogram

Table 2 Relative distribution of nucleosides in bulk tRNA purified

from exponentially growing strains

Strain Relative distribution ()of nucleosidesa

nm5s2U mnm5s2U cmnm5s2U s2U

TH48 background

wt 83plusmn5 17plusmn5mnmC-W131stop 31plusmn1 69plusmn1mnmC(m)-G68D 85plusmn5 15plusmn5BW25113 background

wt 83plusmn3 17plusmn3DmnmG or DmnmE 100DmnmC 27plusmn4 73plusmn4DmnmC(o) 14plusmn3 86plusmn3MG1655 background

wt 85plusmn3 15plusmn3DmnmG 100

atRNA was purified and degraded to nucleosides for HPLC analysisThe percentage of nucleosides represents the distribution of the peakarea of each nucleoside compared to the sum of the peak areas of thetwo nucleosides considered Each value represents the mean of at leasttwo independent experiments wt wild-type

2608 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 9: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

As expected the intermediates cmnm5s2U and nm5s2Uwere observed in mnmC null mutants (carrying mnmC-W131stop or DmnmC mutations) these nucleosides weretypically distributed in a 7030 ratio suggesting that theMnmEG enzyme preferably uses the glycine pathwayunder the growth conditions used in these experiments(Table 2) This conclusion was consistent with thecmnm5s2Umnm5s2U ratio observed in the DmnmC(o)mutant (8614) which clearly exhibited a MnmC(o)

MnmC(m)+ phenotype Moreover the relative distribu-tion of cmnm5s2U and nm5s2U in the mnmC(m)-G68Dmutant was 1585 suggesting that most of thecmnm5s2U formed by the glycine pathway is convertedto nm5s2U by the MnmC(o) activity present in thisstrain and that the remaining cmnm5s2U could proceedat least partially from tRNAGln

cmnm5s2UUG Taken togetherthese results clearly support the idea that MnmEG usesammonium or glycine to modify tRNAs in vivo and thatMnmC(m) functions independently of MnmC(o) [seestrain DmnmC(o) in Table 2] to transform tRNAsmodified by MnmEG via the ammonium pathway(Figure 1) which raises the possibility that certaintRNAs may be substrates for MnmC(m) but not forMnmC(o)

Expression and purification of the MnmC(o) andMnmC(m) domains

In order to study the functional independence of the twoMnmC domains we decided to express them separatelyThe full mnmC gene and its two domains mnmC(o)(encoding for amino acids 250ndash668 of MnmC) andmnmC(m) (encoding for amino acids 1ndash250 of MnmC)were N-terminal Flag-tagged by PCR and cloned intothe pBAD-TOPO expression vector Recombinantprotein expression was induced in the E coli mutantmnmCkan (DmnmC) by adding 02 arabinose TheFlag-tagged proteins MnmC MnmC(o) and MnmC(m)were purified close to homogeneity and exhibitedapparent molecular masses of 78 48 and 34 kDa respect-ively (Figure 2A)

The Flag-MnmC protein and its C-terminal domain[Flag-MnmC(o)] were yellow and their spectra showedmaximum absorption peaks at 375 and 450 nmindicating the presence of a flavin derivative Theputative cofactor was identified as FAD by its retentiontime (Figure 2B) Therefore the isolated MnmC(o)domain contains non-covalently bound FAD whichsuggests that it is correctly folded To our knowledgethis is the first report describing the expression and puri-fication of the soluble form of MnmC(o) A previousattempt to purify this domain was unsuccessful becauseoverexpression of a somewhat different recombinantMnmC(o) produced inclusion bodies which led to the hy-pothesis that MnmC(o) is incapable of folding on its ownand requires the presence of MnmC(m) (25) Our datahowever demonstrate that MnmC(o) can fold independ-ently of MnmC(m) Interestingly the in vivo stability ofthe separated domains was significantly lower than that ofthe full protein (Figure 2C) suggesting that the physical

interaction of both domains in the full protein confersgreater stabilityThe full MnmC protein and the separate domains

MnmC(o) and MnmC(m) behaved as monomericproteins when subjected to gel filtration chromatography(Figure 2D) However a mixture of MnmC(o) andMnmC(m) displayed the same elution profile as the fullprotein indicating that the separate domains interactin vitro (Figure 2D) This interaction was furtherexplored by kinetic analysis using surface plasmon reson-ance (SPR) Flag-MnmC(o) was injected at different con-centrations onto the immobilized His-MnmC(m) ligand(Figure 2E) The apparent equilibrium constant KD was87plusmn15nM which is indicative of a strong interactionbetween MnmC(o) and MnmC(m)Taking into consideration the functional relationships

of MnmC with the MnmEG complex we exploredwhether MnmC interacts with members of the complexie MnmE or MnmG A final concentration of 5 mMFlag-MnmC was mixed with 5 mM His-MnmG or MnmE andthe samples were analyzed by gel filtration As shown inFigure 2F no interaction between the full MnmC proteinand MnmE or MnmG was observed The same negativeresult was obtained by SPR (data not shown)

In vitro and in vivo determination of the enzymaticactivities of MnmC(o) and MnmC(m)

To analyze whether the recombinant MnmC(o) andMnmC(m) proteins are functionally active we firstassessed their tRNA modifying capability in vitro usingtotal tRNA purified from exponentially growing culturesas a substrate The in vitro activity of the recombinantMnmC(o) protein was investigated by incubating tRNApurified from the null mutant mnmC-W131stop (nm5s2Ucmnm5s2U ratio amp 3169 Table 2) with 2 mM Flag-MnmC(o) domain and 05mM FAD As shown inFigure 3A a major part of cmnm5s2U was convertedto nm5s2U which demonstrates that Flag-MnmC(o)is catalytically active in the in vitro assay although asmall part of cmnm5s2U probably proceeding fromtRNAGln

cmnm5s2UUG remained unaltered When tRNApurified from the mnmC(m)-G68D mutant (nm5s2Ucmnm5s2U ratio amp 8515 Table 2) was incubated with2 mM Flag-MnmC(m) and 05mM SAM nm5s2U wasmodified to mnm5s2U demonstrating the capability ofthe recombinant protein to perform the methylationreaction in vitro (Figure 3B) As expected the smallpeak corresponding to cmnm5s2U (according to its reten-tion time and spectrum) remained unchanged after thein vitro reaction mediated by MnmC(m) (Figure 3B)Then we assessed the capability of the recombinant

MnmC(o) and MnmC(m) proteins to modify tRNAin vivo The strain with the null mutation mnmC-W131stop (IC6019) was transformed with pBAD-TOPOand its derivative pIC1339 expressing MnmC(o) whereasstrain mnmC(m)-G68D (IC6018) was transformed withpBAD-TOPO and its derivative pIC1340 expressingMnmC(m) In the resulting strains which were grownto exponential phase in the presence of the arabinoseinducer the recombinant MnmC(o) protein catalyzed

Nucleic Acids Research 2014 Vol 42 No 4 2609

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 10: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

C

0 10acute 20acute 30acute 40acute 50acute 60acute

Gro

EL

MnmC

MnmC(o)

MnmC(m)

0

10

20

30

40

Volume85 90 95 100 105 110 115 ml

Mn

mC

(m)

Mn

mC

(o)

Mn

mC

a b c d

A28

0(m

AU

)

Mn

mC

(o)

Mn

mC

(m)

+

BSA

67kDa

Ovalbumin

43 kDa

Chymotry

psinogen

25 kDa

BA

A28

0(m

AU

)

Mn

mE

Mn

mG

Mn

mC

Volume

β-Amyla

se

224 kDa

Aldolase

158 kDa BSA

67 kDa

0

50

100

150

110 120 130 140 150 ml

01 microM

025 microM

05 microM

1 microM

2 microM

E

Time (s)

Res

pons

e

D

F

a b c dInp

ut

3442

52kDa

MnmC(o) extract

Minutes196 20 204 208 212 216 220 224

Flu

ores

cenc

e

0

20

40

60

80

100

120

140

160

180

200FAD and FMN

FA

D

FM

N

Mn

mC

Mn

mC

(o)

Mn

mC

(m)

260

135

95

72

52

42

34

kDa

26

Ka (1Ms) kd (1s) KD (M) Chi2

9934E+4 0007602 7652E-8 299

Figure 2 Characterization of the MnmC recombinant proteins (A) Coomassie blue staining of SDSndashPAGE containing the purified Flag-MnmCFlag-MnmC(o) and Flag-MnmC(m) proteins used in this work (B) HPLC analysis of the Flag-MnmC(o) extract (blue line) and markers (red line)FMN flavin mononucleotide (C) Half-life of the Flag-MnmC proteins over time as determined by tracking their decline after the addition of glucoseto cultures of IC6010 (DmnmC) transformed with the plasmids pIC1253 pIC1339 or pIC1340 which express Flag-MnmC Flag-MnmC(o) and Flag-MnmC(m) respectively GroEL was used as a loading control Protein levels were detected by western blotting (D) Gel filtration analysis of thepurified Flag-MnmC proteins [full MnmC protein green MnmC(o) blue and MnmC(m) black] and a Flag-MnmC(o)Flag-MnmC(m) mix (red)The elution positions of the size markers are indicated on the top Elution fractions andashd from the chromatography of the Flag-MnmC(o)Flag-MnmC(m) mix were pooled for further analysis Inset SDSndashPAGE of elution fractions andashd from the Flag-MnmC(o)Flag-MnmC(m) mix Fractions(500 ml) were precipitated with trichloroacetic acid before loading The gel was stained with coomassie blue The marker molecular masses areindicated on the left (E) SPR analysis of the MnmC(o)ndashMnmC(m) interaction Flag-MnmC(o) was injected into a solution passing over a sensorchip containing the immobilized His-MnmC(m) (600 RU) Representative sensorgrams for various concentrations of Flag-MnmC(o) are shown (F)Gel filtration analysis of the purified MnmC protein (blue line) and a mix of MnmC with MnmG (black line) or MnmE (red line) The elutionpositions of the size markers are indicated on the top Proteins were detected by UV absorbance at 280 nm

2610 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 11: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

the conversion of cmnm5s2U into nm5s2U although apeak of remnant cmnm5s2U was observed (Figure 3C)and nm5s2U was fully transformed to mnm5s2U byMnmC(m) (Figure 3D) Similar results were obtained inthe absence of arabinose (data not shown) This findingindicates that recombinant MnmC(o) and MnmC(m) areexpressed and accumulate to some extent in the absence ofthe inducer although it was not possible to detect them bywestern blotting with an anti-Flag antibody (data not

shown) The fact that both proteins were capable of mod-ifying tRNA despite their very low concentrationssuggests that these proteins have high activity in vivo

Kinetic analysis of the MnmC- and MnmEG-dependentreactions

To explore whether the recombinant MnmC(o) andMnmC(m) proteins display similar kinetic properties to

In vitro activity of MnmC(o)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

tRNA from mnmC-W131stop (a)

(a) + Purified MnmC(o) domain

nm5s2U cmnm5s2U

s2CA31

4 nm

In vitro activity of MnmC(m)

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

1

2

3

4

5

6

(b) + Purified-MnmC(m) domain

tRNA from mnmC(m)-G68D (b)

nm5s2U mnm5s2U

s2Ccmnm5s2U

A31

4 nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC(m)-G68D + empty pBAD

mnmC(m)-G68D + pBAD-mnmC(m)

cmnm5s2Us2C

nm5s2U mnm5s2U

In vivo activity of MnmC(m) A

314

nm

Minutes26 27 28 29 30 31 32 33 34 35 36 37

0

2

4

6

8

mnmC-W131stop + empty pBAD

mnmC-W131stop + pBAD-mnmC(o)

nm5s2U

s2C

cmnm5s2U

In vivo activity of MnmC(o)

A31

4 nm

A B

C D

Figure 3 In vitro and in vivo activity of the recombinant proteins MnmC(o) and MnmC(m) (A) and (B) HPLC analysis of total tRNA from a nullmnmC mutant (IC6019 panel A) and an mnmC(m)-G68D mutant (IC6018 panel B) before (solid black line) and after in vitro incubation (dotted redline) with purified MnmC(o) and MnmC(m) recombinant proteins respectively (C) and (D) HPLC analysis of total tRNA extracted from IC6019(panel C) and IC6018 (panel D) transformed with pBAD-TOPO (solid black lines) or a pBAD-TOPO derivative expressing Flag-MnmC(o) and Flag-MnmC(m) respectively (dotted red lines) Absorbance was monitored at 314 nm to maximize the detection of thiolated nucleosides Small variationsin elution times between upper and lower panels were probably due to negligible variations in buffers prepared in different days Note thatnucleosides were identified by both elution times and spectra (data not shown)

Nucleic Acids Research 2014 Vol 42 No 4 2611

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 12: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

those exhibited by the entire MnmC protein we con-ducted steady-state kinetic experiments The conditionsused for each reaction were similar to enable a comparisonof the kinetic constants and were based on conditionsknown to optimize activity (2326) The substrate for the

assays was a chimeric version of E coli tRNALysmnm5s2UUU

expressed from pBSKrna (3742) which facilitates over-production and purification of recombinant RNA and hasbeen successfully used by our group in previous studies

(2133) Here the tRNALysmnm5s2UUU gene was inserted

into the EcoRV site of the region encoding the scaffold

tRNA (a human cytosolic tRNALys3 lacking the anticodon

region) The resulting chimeric tRNA essentially con-

tained the complete E coli tRNALysmnm5s2UUU sequence

fused at its 50- and 30-ends to the truncated anticodon

stem of human cytosolic tRNALys3 producing an RNA

of 170 nt (Supplementary Figure S2)The FAD-dependent-oxidoreductase activity of the re-

combinant MnmC(o) and MnmC proteins was comparedby determining the conversion of cmnm5s2U to nm5s2U onthe cmnm5s2U-containing chimeric tRNALys

mnm5s2UUU (chi-tRNALys

mnm5s2UUU) purified from an mnmC null mutant(IC6010) The SAM-dependent methyltransferase activityof MnmC(m) and MnmC was compared by followingthe conversion of nm5s2U to mnm5s2U on the nm5s2U-containing chi-tRNALys

mnm5s2U extracted from anmnmC(m)-G68D mutant (IC6018) All proteins displayedMichaelisndashMenten kinetics with respect to varying concen-trations of chi-tRNALys

mnm5s2UUU (SupplementaryFigure S3) As shown in Table 3 the enzymatic activityof each isolated MnmC domain had kinetic parameters(kcat and Km) similar to those exhibited by the entireprotein which supports the notion that the MnmCdomains function independentlyThe kcat constants of the MnmC activities were similar

to those of previous studies (2326) although the Km

values were higher than those reported by Pearson andCarell (26) most likely due to the nature of the tRNAsubstrate used in our experiments In any case our dataindicated that MnmC(m) displayed a catalytic efficiency(kcatKm) 2-fold higher than that of MnmC(o) (01 versus005 Table 3) which is in agreement with the proposalthat synthesis of the final modification mnm5(s2)U isfavored by kinetic tuning of the MnmC activities thusavoiding accumulation of the nm5(s2) intermediate (26)We also analyzed the catalytic parameters of the

ammonium-dependent reaction mediated by theMnmEG complex whose activity precedes that of

MnmC(m) in the mnm5(s2)U synthesis (Figure 1) Weused conditions known to be appropriate for assayingthe MnmEG activities in vitro (13) and included thepresence of Mg2+ (5mM) which is required for GTP hy-drolysis by MnmE (43) In line with this the assay bufferdiffered from that of the MnmC reactions where a rela-tively low concentration of Mg2+ was used (73mM) asMnmC activities are known to be severely inhibited at5mM of Mg2+ (23) Under the proper in vitro conditionsfor each enzyme the modification reaction mediated byMnmEG (incorporation of nm5 into U34) displayed an5-fold lower catalytic efficiency than that of theMnmC(m)-mediated reaction (Table 3) However thiscomparison should be handled with caution because theconditions used in the MnmEG and MnmC(m) assayswere different and the Km values may change accordingto the buffer conditions In addition the structure of thechimeric tRNA used as a substrate could affect the cata-lytic cycle of each enzyme in a different manner

Therefore in order to gain information on the effi-ciency of the mnm5s2U assembly-lines in vivo wedecided to explore the modification status of the native

tRNALysmnm5s2UUU in several strains The presence of modi-

fication intermediates in the tRNA purified from a wild-type strain would be indicative of the MnmEG andMnmC activities not being kinetically tuned HPLC

analysis of native tRNALysmnm5s2UUU purified during

the exponential phase revealed no accumulation of thecmnm5s2U intermediate in the wild-type strain(Figure 4A) whereas this nucleoside was predominant inthe tRNA extracted from the mnmC strain (Figure 4B)Moreover we observed no nm5s2U in the native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4A) despite this intermediate accumulated in

tRNALysmnm5s2UUU of the mnmC(m)-G68D mutant

(Figure 4C) These results suggest that tRNALysmnm5s2UUU

containing cmnm5s2U or nm5s2U at the wobble positionis stable and consequently that the disappearance of theseintermediates in the wild-type strain is probably due totheir conversion into mnm5s2U by MnmC activitiesThus it appears that MnmEG and MnmC are kineticallytuned to produce only the final mnm5 group in

tRNALysmnm5s2UUU

Notably the intermediate s2U resulting from theactivity of MnmA (Figure 1) was only observed in aDmnmG strain which is defective in the MnmEG activity(Figure 4D)Moreover we did not detect the non-thiolated

intermediates cmnm5U and nm5U when tRNALysmnm5s2UUU

Table 3 Kinetic parameters with respect to chi-tRNALysmnm5s2UUU for reactions catalyzed by MnmC MnmC(o) MnmC(m) and MnmEG

Reaction Km (mM) Vmax (nmoles min1mg1) kcat (s1) kcatKm (s1mM-1)

MnmC(FAD) (cmnm5s2U nm5s2U) 157plusmn34 457plusmn12 059plusmn002 0038MnmC(o) domain(FAD) (cmnm5s2U nm5s2U) 61plusmn21 486plusmn69 039plusmn005 0064MnmC(SAM) (nm

5s2U mnm5s2U) 44plusmn11 365plusmn77 046plusmn010 0105MnmC(m) domain(SAM) (nm

5s2U mnm5s2U) 42plusmn11 895plusmn179 052plusmn010 0124MnmEG(NH4) (s

2U nm5s2U) 06plusmn02 59plusmn03 0012plusmn0001 0020

The values are the meanplusmnSD of a minimum of three independent experiments

2612 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 13: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

hydrolysates were monitored at 254 nm (data not shown)These data suggest that the coordination of the MnmA-and MnmEGC-dependent pathways is tightly coupled to

synthesize mnm5s2U on native tRNALysmnm5s2UUU without a

build-up of intermediates

tRNA substrate specificity of MnmC(o) andMnmC(m) in vivo

Escherichia coli tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA

contain cmnm5 at the wobble uridine instead of themnm5 found in the remaining MnmEG substrates(httpmodomicsgenesilicoplsequenceslisttRNA)

These data suggest that tRNAGlncmnm5s2UUG and

tRNALeucmnm5UmAA are not substrates for MnmC(o)

Moreover they raise the question of what happens withthe ammonium pathway of the MnmEG complex whichtheoretically produces nm5U in all its tRNA substrates(Figure 1) Interestingly the analysis of gln1 tRNA fromSalmonella enterica serovar Typhimurium indicated that80 of the molecules contained cmnm5s2U whereas theremaining 20 contained mnm5s2U (44) In this case wespeculate that mnm5s2U might be synthesized by MnmC[using the MnmC(m) activity] from the nm5s2U generatedvia the ammonium-dependent MnmEG pathwayTherefore we decided to investigate whether a fraction

of tRNAGlncmnm5s2UUG and tRNALeu

cmnm5UmAA in E colicontains mnm at position 5 in the wobble uridine andwhether both tRNAs function as substrates forMnmC(m) but not MnmC(o) (Figure 1)To address this question we constructed a series

of tRNA expression plasmids by inserting the coding

sequences of tRNAGlncmm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU into EcoRIPstI-digested pBSKrna (37)

thus we eliminated the tRNA scaffold-encoding region(ie the DNA region encoding the human cytoplasmic

tRNALys3 ) present in the vector (Supplementary

Figure S2) Selected strains were transformed with the re-sulting plasmids We then examined the HPLC profile ofthe overexpressed lsquonativersquo tRNAs purified from culturesgrown overnight to the stationary phase It should benoted that the expression of the tRNA genes in these con-structs was under the control of the strong constitutive lpppromoter (37) As mutations leading to reduced expres-sion might be selected over time because overexpressionultimately slows growth the fresh transformation of cellswith the pBSKrna derivatives every time and overnightincubation of the transformed cells prior to the purifica-tion of the cloned target (in this case tRNAs) are highlyrecommended (37)The HPLC analysis of tRNALys

mnm5s2UUU (Figure 5A)used herein as a control indicated that (i) mnm5s2Uwas predominant in a wild-type strain (ii) both nm5s2Uand cmnm5s2U accumulated in the mnmC null strain and(iii) nm5s2U prevailed in the mnmC(m)-G68D strainTaken together these results indicate that bothMnmEG-MnmC(o)-MnmC(m) and MnmEG-MnmC(m)

pathways operate on tRNALysmnm5s2UUU and that they

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14A

314

nm

mnm5s2U s4U

wt

Minutes28 30 32 34 36 38 40 42 44 46

0

1

2

3

4

5

6

7

A31

4 nm

s4Unm5s2U

mnmC(m)-G68D

Minutes

28 30 32 34 36 38 40 42 440

5

10

15

20

A31

4 nm

cmnm5s2U

nm5s2U

ΔmnmC s4U

A

B

C

D

A31

4 nm

Minutes

28 30 32 34 36 38 40 42 440

2

4

6

8

10

12

14s4U

s2U

ΔmnmG

Figure 4 HPLC analysis of native tRNALysmnm5s2UUU purified from dif-

ferent strains at exponential phase Native tRNALysmnm5s2UUU was purified

from the wild-type (A) DmnmC (B) mnmC-G68D (C) and DmnmG (D)strains and subjected to HPLC analysis which was monitored at314 nm The relevant nucleosides at position 34 (mnm5s2U s2Ucmnm5s2U and nm5s2U) and position 8 (s4U) of tRNALys

mnm5s2UUU areindicated

Nucleic Acids Research 2014 Vol 42 No 4 2613

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 14: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

converge to produce the final modification mnm5s2U inthe wild-type strainAn analysis of overexpressed tRNAGln

cmnm5s2UUG(Figure 5B) indicated that cmnm5s2U and nm5s2Uaccumulated at a ratio of 9010 when tRNA waspurified from the mnmC-W131stop mutant Despite thelower proportion of nm5s2U the presence of this nucleo-tide indicates that the MnmEG complex was indeed able

to modify tRNAGlncmnm5s2UUG through the ammonium

pathway Nucleosides cmnm5s2U and mnm5s2U were

found at a 9010 ratio when tRNAGlncmnm5s2UUG was

purified from the wild-type strain Detection of

mnm5s2U in tRNAGlncmnm5s2UUG clearly indicates that this

tRNA is a substrate for MnmC(m)It should be noted that both cmnm5s2U and nm5s2U

accumulated in tRNALysmnm5s2UUU purified from the

mnmC-W131stop mutant whereas cmnm5s2U practically

Minutes29 30 31 32 33 34 35 36 37

0

2

4

6

8

nm5s2U

mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNALys

A31

4 nm

Minutes29 30 31 32 33 34 35 36 37

0

1

2

3

nm5s2U mnm5s2U cmnm5s2U

Wild type mnmC(m)-G68DmnmC-W131stop

pBSK-tRNAGln

A31

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

2

4

cmnm5U

mnm5U

Xnm5U

X

pBSK-tRNALeu

Wild type ΔtrmL

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

5

10

15

20

cmnm5U

nm5U mnm5U

Markers

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19 20

0

1

2

3

4

5

6 ΔtrmL

Native tRNALeu

ΔtrmLΔmnmG

cmnm5U

A25

4 nm

A B

C D

E

Figure 5 In vivo specificity of MnmC(o) and MnmC(m) for substrate tRNAs Representative HPLC profiles of tRNALysmnm5s2UUU (A)

tRNAGlncmnm5s2UUG (B) and tRNALeu

cmnm5s2UmAA (D) expressed from pIC1664 pIC1714 and pIC1665 respectively The strains used in panels A andB were IC6017 (WT) IC6018 [mnmC(m)-G68D] and IC6019 (mnmC-W131stop) The strains used in panel D were IC5136 and IC5854 HPLC

profiles of commercial markers are shown in panel C Representative HPLC profiles of native tRNALeucmnm5UmAA purified from strains DtrmL (IC6374)

and DtrmLDmnmG (IC6411) are shown (E)

2614 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 15: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

disappeared when tRNALysmnm5UUU was purified from the

mnmC(m)-G68D strain (Figure 5A) This result wasexpected considering that cmnm5s2U was transformedinto nm5s2U by the MnmC(o) activity present in themnmC(m)-G68D strain Interestingly the HPLC profiles

of tRNAGlncmnm5s2UUG purified from the mnmC(m)-G68D

and mnmC-W131stop strains were similar (Figure 5B)In both cases cmnm5s2U was much more abundantthan nm5s2U suggesting that the MnmC(o) activitypresent in the mnmC(m)-G68D strain did not function

in tRNAGlncmnm5s2UUG

Overexpression of tRNAs often causeshypomodification In fact the hypomodified nucleosides2U resulting from the action of MnmA was relatively

abundant in the overexpressed tRNALysmnm5s2UUU and

tRNAGlncmnm5s2UUG (data not shown) Nevertheless s2U

accumulated at a similar proportion in all tested strainsso that the ratios of the nucleosides of interest (mnm5s2Ucmnm5s2U and nm5s2U) were not greatly affected When

the tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG hydrolysateswere monitored at 254 nm the non-thiolated nucleosidesnm5U mnm5U cmnm5U were hardly detected (data notshown) Therefore we think that the different HPLC

pattern of tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG in themnmC(m)-G68D strain cannot be attributed to thepresence of hypomodified nucleosides Rather it reveals

that tRNAGlncmnm5s2UUG is not a substrate for MnmC(o)

We also analyzed the HPLC profile of tRNALeucmnm5UmAA

purified from a trmL strain in which the methylation ofthe ribose in the wobble uridine is impaired due to thetrmL mutation (21) We adopted this approach to

examine the modification of tRNALeucmnm5UmAA because we

were unable to distinguish nucleosides cmnm5Um nm5Um

and mnm5Um in our HPLC chromatograms However wewere able to identify the peaks corresponding to cmnm5Unm5U and mnm5U using proper markers (Figure 5C) As

shown in Figure 5D overexpressed tRNALeucmnm5UmAA

purified from a trmL strain contained cmnm5UHowever no traces of mnm5U or nm5U were detectedin this tRNA Similar results were obtained when

analyzing native non-overexpressed tRNALeucmnm5UmAA

purified in the exponential phase (Figure 5E)

Altogether these data suggest that tRNALeucmnm5UmAA is

not modified by the ammonium-dependent MnmEGpathway and that it is not a substrate for MnmC(o)

tRNA substrate specificity of MnmC(o) and MnmC(m)in vitro

To further explore the specificity of MnmC(o) and

MnmC(m) for tRNAGlncmnm5s2UUG tRNALeu

cmnm5UmAA and

tRNALysmnm5s2UUU we performed in vitro modification reac-

tions using in vitro synthesized tRNAs as substrates Thesemodification reactions included two steps First substratetRNA was modified by the MnmEG complex throughthe ammonium (Figure 6A B E and F) or the glycine(Figure 6C and D) pathways The resulting tRNA carryingnm5U or cmnm5U (solid black lines) was subsequently used

as a substrate to characterize the activity of theMnmC(m) orthe MnmC(o) domains respectivelyThe data in Figure 6 indicate that MnmC(m)

catalyzed the conversion of nm5U into mnm5U in both

tRNALysmnm5s2UUU (panel A) and tRNAGln

mnm5s2UUG (panelB) whereas MnmC(o) catalyzed the conversion of

cmnm5 into nm5 in tRNALysmnm5s2UUU (panel C) but not

in tRNAGlncmnm5s2UUG (panel D) These results support the

idea that tRNAGlncmnm5s2UUG does not function as a sub-

strate for MnmC(o)Surprisingly we observed that MnmEG catalyzed the

formation of nm5U from in vitro synthesized

tRNALeucmnm5UmAA and in turn nm5U was converted into

mnm5U through MnmC(m) (Figure 6E) This resultcontradicts the data provided in Figure 5D and E whereno traces of nm5U or mnm5U were observed in the HPLC

analysis of the tRNALeucmnm5UmAA purified from a trmL

strain Considering that the experiments shown inFigure 6E were performed using an in vitro synthesized

tRNALeucmnm5UmAA (therefore lacking modifications) we

thought that the presence of some modified nucleoside(s)

in the in vivo synthesized tRNALeucmnm5UmAA (Figure 5D

and E) could hinder the recognition of this substrate byMnmEG However when we used overexpressed

tRNALeucmnm5UmAA purified from a double DtrmLDmnmG

mutant as a substrate in the in vitro modification assaythe MnmEG-mediated synthesis of nm5s2U was onceagain observed (Figure 6F) Moreover the MnmEG-

modified tRNALeucmnm5UmAA was a good substrate for

the in vitro synthesis of mnm5 through MnmC(m)(Figure 6F) Altogether these results suggest that theammonium pathway of MnmEG is ineffective on

tRNALeucmnm5UmAA in vivo even though it efficiently functions

on this tRNA in the in vitro assayWe also assessed the activity of MnmC(o) on the

tRNALeucmnm5UmAA purified from a trmL strain and on the

in vitro transcribed tRNALeucmnm5UmAA previously modified

by MnmEG in vitro via the glycine pathway (ie ontRNA molecules carrying cmnm5U) The synthesis ofnm5 from cmnm5 was not observed under any condition(data not shown) Therefore we concluded that

tRNALeucmnm5UmAA is a substrate in vitro for MnmC(m)

(Figure 6E and F) but not MnmC(o)Notably MnmEG also catalyzed the ammonium-

dependent synthesis of nm5U in tRNAGlumnm5s2UUC

tRNAArgmnm5UCU and tRNAGly

mnm5UCC obtained by in vitrotranscription (Supplementary Figure S4) We suspectthat modification of these tRNAs follows a pattern

similar to that observed in tRNALysmnm5s2UUU because all

these tRNA species appear to contain mainly the mnm5

group of U34 However this hypothesis should beexamined in future studies

Synthesis of nm5U and cmnm5U is modulated by growthconditions and the tRNA species

The overexpressed tRNALysmnm5s2UUU and tRNAGln

cmnm5s2UUG

purified from the mnmC-W131stop mutant during the

Nucleic Acids Research 2014 Vol 42 No 4 2615

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 16: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

stationary phase exhibited different cmnm5s2Unm5s2Uratios (3565 and 9010 respectively Figure 5Aand B) Moreover the ratio in the overexpressed

tRNALysmnm5s2UUU (3565) differed from that found in

the total tRNA obtained from the mnmC-W131stop

strain during exponential growth (6931 Table 2)These results suggest that the glycine pathway (whichproduces cmnm5s2U) could be less effective on

tRNALysmnm5s2UUU in the stationary phase These observa-

tions prompted us to carefully investigate the efficiency of

Minutes

10 11 12 13 14 15 16 17 18 19

0

1

2

3

4Glycine pathway of tRNAGln

tRNAGln + MnmEbullMnmG (d)

(d) + MnmC(o) domain

nm5U cmnm5U

X

A25

4 nm

Minutes10 11 12 13 14 15 16 17 18 19

0

1

2

3

4tRNALys + MnmEbullMnmG (c)

(c) + MnmC(o) domain

Glycine pathway of tRNALys

nm5U cmnm5UA25

4 nm

Ammonium pathway of tRNA Gln

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

SAM

tRNAGln + MnmEbullMnmG (b) (b) + MnmC(m) domain

A25

4 nm

Ammonium pathway of tRNALys

Minutes10 11 12 13 14 15 16 17 18 19

0

2

4

6

8

nm5U mnm5U

tRNALys + MnmEbullMnmG (a) (a) + MnmC(m) domain

SAM

A25

4 nm

Minutes

10 11 12 13 14 15 16 17 18 19 20

0

2

4

6

8

10

mnm5U nm5U

tRNA Leu from ΔtrmLΔ mnmG

(f) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (f)

A25

4 nm

Ammonium pathway of tRNALeu

(e) + MnmC(m) domain

tRNALeu + MnmEbullMnmG (e)

nm5U

Minutes10 11 12 13 14 15 16 17 18 19 20

00

05

10

15

20

mnm5U

A25

4 nm

A B

C D

E F

Figure 6 In vitro specificity of MnmC(o) and MnmC(m) for substrate tRNAs (AndashD) The modification reactions were performed using in vitrosynthesized tRNAs in two steps First the substrate tRNA was modified by the MnmEG complex via the ammonium (A) and (B) or the glycinepathway (C) and (D) and the resulting tRNA (solid black line) carrying nm5U or cmnm5U was used as a substrate to examine the activity of the

MnmC(m) or MnmC(o) domain (dashed red line) respectively (E) HPLC analysis of in vitro synthesized tRNALeucmnm5UmAA after in vitro modification

by MnmEG (solid black line) and MnmC(m) (dashed red line) (F) Overexpressed tRNALeucmnm5UmAA purified from the strain IC6411 (DtrmLDmnmG)

served as the substrate in the ammonium-dependent reaction catalyzed by MnmEG The resulting tRNA (black line) was used as a substrate for theMnmC(m)-mediated reaction (red line)

2616 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 17: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

the glycine and ammonium pathways in total tRNA andspecific native tRNAs during the growth curve

First we determined the modification status of totaltRNA in the wild-type strain and mnmC mutants In thewild-type or mnmC(m)-G68D strain (Figure 7 panels Aand B) we could not identify which pathway contributesto the synthesis of the final modification (mnm5s2U ornm5s2U respectively) because the activity of MnmC(o)converges both pathways by transforming cmnm5s2Uinto nm5s2U (Figure 1) However the strain carrying anmnmC-W131stop or mnmC(o) null mutation (Figure 7panels C and D) revealed that the level of cmnm5s2Uand nm5s2U or mnm5s2U in bulk tRNA depends on thegrowth phase nm5s2U (or mnm5s2U) accumulates asoptical density increases whereas cmnm5s2U exhibits theopposite trend and becomes the minor component at anOD600 of 20 As the data in Figure 7 represent therelative distribution of each nucleoside with respect tothe sum of the peak areas of the two nucleosides it isimportant to point out that the net decrease in thecmnm5s2U peak area along the growth curve was con-comitant with the net increase in the nm5s2U ormnm5s2U area

Interestingly when the DmnmC(o) strain was grown inminimal medium instead of LBT cmnm5s2U was consist-ently observed as the major intermediate irrespective ofthe growth phase (Table 4) Therefore the growth

conditions (growth medium and growth phase) affect thepathway responsible for modification at position 5 of U34We subsequently analyzed the behavior of native

tRNAs in both the exponential and stationary phases(Table 5 and Supplementary Figure S5) During the expo-nential growth of the DmnmC strain cmnm5s2U was themajor intermediate accumulated in nativetRNALys

mnm5s2UUU (94) but this nucleoside became theminor intermediate (4) in the stationary phase Incontrast HPLC analysis revealed that cmnm5s2U wasconsistently the major intermediate present in nativetRNAGln

cmnm5s2UUG (gt90) irrespective of the growthphase Altogether these results indicate thattRNAGln

cmnm5s2UUG functions different fromtRNALys

mnm5s2UUU and total tRNA when the DmnmCstrain is grown to stationary phase in LBT (ie ingrowth conditions that favor the ammonium pathway)Therefore the accumulation of cmnm5 or nm5 dependson the specific characteristics of the tRNA

Importance of MnmC-mediated modifications for cellgrowth

The limited data available regarding the effect of mnmCmutations on growth rate are contradictory No differ-ences between mnmC and wild-type strains were initiallyreported (2245) However more recently Pearson andCarell (26) measured the growth rate of an mnmC-

Wild type

0 2 4 60

20

40

60

80

100

cmnm5s2U

mnm5s2U

OD600

n

ucle

osid

es

mnmC(m)-G68D

0 2 4 60

20

40

60

80

100

nm5s2U

cmnm5s2U

OD600

n

ucl

eosi

des

mnmC- W131stop

0 2 4 60

20

40

60

80

100

cmnm5s2U

nm5s2U

OD600

n

ucle

osid

es

BA

C ΔmnmC(o)

0 2 4 60

20

40

60

80

100

mnm5s2U

cmnm5s2U

OD600

n

ucle

osid

es

D

Figure 7 Synthesis of nm5U34 and cmnm5U34 is depending on the growth phase HPLC analysis of total tRNA purified from TH48 (A) TH49 (B)TH69 (C) and IC6629 (D) during the growth cycle The strains were grown in LBT The percentage of nucleosides represents the distribution of thepeak area of each nucleoside compared to the sum of the peak areas of the two nucleosides considered Each time point represents the average fromtwo independent experiments The standard deviations were within plusmn20

Nucleic Acids Research 2014 Vol 42 No 4 2617

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 18: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

knockout strain from the Keio collection and observed alarger growth constant for the wild-type strain in agree-ment with the hypothesis that accumulation of intermedi-ates in the synthesis of mnm5(s2)U affects the translationprocess negatively We investigated the effect of mnmCmutations on growth rate in different genetic backgroundsand did not observe significant differences between thewild-type strain and any derivative carrying an mnmCallele [DmnmC mnmC-W131stop DmnmC(o) andmnmC(m)-G68D Table 6] We cannot explain why ourresults differ from those reported by Pearson and Carell(26) even when the same background (BW25113) wasused Perhaps the dilution method used to maintain thepre-cultures in exponential growth could explain the dif-ferent results We took great care to not dilute the pre-cultures below an OD600 of 025 and to not exceed anOD600 of 05 in order to avoid major changes under theexponential growth conditionsThe lack of effect in our hands of mnmC mutations

during the exponential growth of cells contrasts with thebehavior of mnmE- and mnmG-knockout strains whichas we have previously demonstrated grow slower than thewild-type strain (1232) These data suggest that the lackof any modification at the wobble uridine is more deleteri-ous to the cell than the presence of any intermediate ofmnm5(s2) synthesis Supporting this conclusion we alsoobserved that the mnmC mutations unlike the mnmE mu-tations did not affect resistance to acid stress of E coli(Figure 8) This resistance depends on several systems oneof which involves proteins whose translation likelydepends on tRNAs modified by MnmEG (3946)The MnmA protein catalyzes the last step of thiolation

at position 2 of U34 in tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC

and tRNAGlncmnm5s2UUG (Figure 1) Inactivation of the

mnmA gene has been reported to increase the doublingtime (3147) Interestingly combination of the mnmAand mnmC mutations had a synergic effect (Table 6) re-vealing that loss of any MnmC activity is detrimentalunder specific conditions Mutation mnmC(m)-G68Dappeared to slow down the growth rate of the mnmA-knockout strain slightly more than the DmnmC mutation

We also performed competition experiments betweenBW25113 (which is a wild-type tetracycline-sensitivestrain) and TH48 (wild-type) TH49 (mnmC(m)-G68D)and TH69 (mnmC-W131stop) which are tetracycline-resistant strains (Table 7) This type of experimentcompares not only the growth in exponential phase butalso the ability to sustain and recover from stationaryphase Although TH48 and its derivatives were out-competed by BW25113 it is clear that the impairment ofthe MnmC functions reduces the ability of the strain tocompete Curiously in these experiments mutant mnmC-W131stop was less competitive than mnmC(m)-G68D

DISCUSSION

This study aimed to further explore the activities andspecificities of the MnmEG complex and the MnmCdomains the ability of specific tRNAs to follow theammonium- or glycine-MnmEG pathway under differentgrowth conditions and the effect of mnmC mutations oncell growth

The fulfillment of the two first aims was facilitated bythe cloning and separate expression of the MnmCdomains We demonstrate for the first time that theMnmC(o) domain can fold independently of MnmC(m)and that both domains are catalytically active whenexpressed separately Importantly the domains exhibitsimilar kinetic properties to those observed in the entireMnmC protein which suggests that the two domainsoperate independently of each other MoreoverMnmC(o) and MnmC(m) differ in terms of their specifi-city for tRNA substrates Thus MnmC(o) modifies

tRNALysmnm5s2UUU but not tRNAGln

cmnm5s2UUG and

tRNALeucmnm5UmAA In contrast all three tRNAs

(tRNALysmnm5s2UUU tRNAGln

cmm5s2UUG and tRNALeucmnm5UmAA)

serve as substrates for MnmC(m) not only in vitro but

also in vivo at least in the cases of tRNALysmnm5s2UUU and

tRNAGlncmm5s2UUG These data strongly support the notion

that the MnmC(o) and MnmC(m) domains function inde-pendently The presence of mnm5s2U in in vivo synthesized

Table 4 Reprogramming of the U34 modification depends on the growth conditions

Nucleoside distributiona OD600 nm (LBT) OD600 nm (MM)

06 09 2 26 5 (ON) 06 17 3 (ON)

cmnm5s2U 86plusmn3 41plusmn4 23plusmn7 22plusmn7 32plusmn1 89plusmn1 96plusmn1 100mnm5s2U 14plusmn3 59plusmn4 77plusmn7 78plusmn7 68plusmn1 11plusmn1 4plusmn1 0

aRelative distribution () of nucleosides Total tRNAs were purified throughout the growth cycle from strain DmnmC(o) (IC6629) growing in LBTor minimal medium (MM YM9 supplemented with 04 glucose) and analyzed by HPLC The OD600 of the overnight (ON) cultures was differentaccording to the growth medium

Table 5 Reprogramming of U34 modification also depends on the

tRNA species

Specific tRNAs Growth Phase LBTa

cmnm5s2U nm5s2U

tRNALysmnm5s2UUU Exponential 94plusmn1 6plusmn1

Stationary 4plusmn2 96plusmn2tRNAGln

cmnm5s2UUG Exponential 99plusmn1 1plusmn1Stationary 93plusmn2 7plusmn2

aRelative () distribution of the nucleosides in accordance with thegrowth phase and the tRNA species Native tRNAs were purifiedfrom strain DmnmC (IC6010) during the exponential (OD600 04) orstationary phase (OD600 3)

2618 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 19: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

tRNAGlncmnm5s2UUG suggests that the nomenclature of this

tRNA should be changed to tRNAGlnethcTHORNmnm5s2UUG

Sequence searches of complete genomes have revealedthat the distribution of the putative orthologs of the E colibi-functional MnmC protein are conserved only ing-proteobacteria with some additional members in otherbacterial classes (24) Interestingly in several genomespotential orthologs of a single domain have been identified(24) but the biological function has been experimentallydemonstrated only in the case of the Aquifex aeolicusDUF752 protein an MnmC(m) homolog (41) It isunclear from the phylogenetic analysis whether the inde-pendent putative orthologs represent the ancestral orderived versions of the bi-functional MnmC enzyme(24) The ability of the E coli MnmC(o) and MnmC(m)domains to function independent of each other as shownin this work suggests that the origin of the full MnmCprotein present in g-Proteobacteria (24) likely occurred bydomain fusion The MnmC(o)ndashMnmC(m) fusion in E colisuggests a functional advantage of the spatial proximity ofboth domains The fusion could increase the stability ofthe resulting protein In fact our results indicate that theentire MnmC protein is substantially more stable than theseparate domains (Figure 2C) or other RNA modificationproteins such as RsmG or MnmG which have half-livesof 17 and 31min respectively (48) The possibility thatthe fusion also facilitates the functional cooperationbetween the domains during modification of tRNAs fol-lowing the lsquocanonicalrsquo MnmEG-MnmC(o)ndashMnmC(m)pathway (cmnm5Unm5Umnm5U) cannot be ruled

out A better understanding of how the MnmC proteinfunctions will entail determining whether the binding ofa tRNA molecule to one domain exerts a negative allo-steric effect on the other domain In this respect the struc-tural characterization of MnmCndashtRNA complexes couldprove helpfulBiosynthesis of complex modifications like mnm5s2U

cmnm5s2U cmnm5Um and mnm5U requires the partici-pation of at least two specific enzymes (Figure 1)Our in vivo data suggest that MnmEG and MnmCactivities are kinetically tuned to produce the final modi-

fication mnm5s2U on native tRNALysmnm5s2UUU because no

intermediates (cmnm5s2U and nm5s2U) were observed in awild-type strain (Figure 4) Apparently the intermediatesare not subject to active degradation as they were shownto be stable in strains DmnmC and mnmC(m)-G68DNotably neither s2U nor the non-thiolatednucleosides cmnm5U and nm5U were found in native

tRNALysmnm5s2UUU purified from the wild-type strain

(Figure 4 data not shown) Therefore we believe thatthe MnmEGndashMnmC and MnmA pathways are alsocoordinated by tuning the activities and relative abun-dances of the tRNA-modifying enzymes Modificationsof U34 by MnmEG and MnmA occur independently ofeach other [(3049) Figure 4D] but whether the presenceof the s2 group facilitates the modification of position 5 in

tRNALysmnm5s2UUU tRNAGlu

mnm5s2UUC and tRNAGlnethcTHORNmnm5s2UUG

by modulating the electron density distribution in theuridine ring remains unclearThere are very few data available on the identity deter-

minants required by TrmL Methylation of the ribose

Table 6 Growth rates of mnmC mutants in different genetic backgrounds

Background Doubling time (min)a

wt mnmCb mnmC(m)c DmnmC(o) DmnmA mnmA DmnmA DmnmADmnmCd mnmC(m)c DmnmC(o)

TH48 34plusmn1 33plusmn1 32plusmn1 nd nd nd nd ndBW25113 28plusmn1 30plusmn1 nd 28plusmn1 40plusmn1 nd nd 58plusmn3IC4639 28plusmn1 29plusmn1 30plusmn1 nd 56plusmn1 73plusmn2 78plusmn2 nd

aDoubling-time values of strains (wild-type or carrying the indicated mutations) are the meanplusmnSD of at least two independent experimentsbThe null mnmC alleles were mnmC-W131stop (TH48 background) and DmnmC (IC4639 and BW25113 backgrounds)cThe mnmC(m) point mutation was mnmC(m)-G68D nd not donedThe mnmA point mutation was mnmA-Q233stop

WT

Glutamate

ΔΔmnmE

ΔmnmC

ΔmnmC(o)0 1 2 3 4 0 1 2 3 4 (hours)

_ +

LBT

-aga

r

Strains

Figure 8 The effect of mnmE and mnmC mutations on acid resistanceStrains IC5136 (wild-type) IC5827 (DmnmE) IC6010 (DmnmC) andIC6629 [DmnmC(o)] were grown in LBT containing 04 glucose to sta-tionary phase and then diluted 11000 into EG pH 20 medium supple-mented or not with glutamate The samples were spotted after 0 1 2 3and 4 h post acid challenge on LAT plates

Table 7 Growth competition assay

Straina Ratiob at mix time Ratiob after six24-h cycles

Wild-type (TH48) 056plusmn0007 0021plusmn0003mnmC(m)-G68D (TH49) 053plusmn009 00020plusmn00006mnmC-W131stop (TH69) 056plusmn004 00002plusmn00001

aBW25113 (tetracycline-sensitive strain) and the indicated TH strain(tetracycline-resistant strain) were mixed 11 at the start of theexperimentbThe ratio was calculated as CFUml recovered on LAT supplementedwith tetracycline versus CFUmL recovered on LAT

Nucleic Acids Research 2014 Vol 42 No 4 2619

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 20: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

20-OH group by this enzyme occurs as a late step in

tRNALeuCmAA maturation given that the enzyme fails to

methylate tRNALeuCmAA without the prior addition of

N6-(isopentenyl)-2-methyladenosine (ms2i6A) at position37 (21) Nucleoside ms2i6A is also present in

tRNALeucmnm5UmAA Hence it is feasible that TrmL also

requires ms2i6A as a positive identity determinant to

modify tRNALeucmnm5UmAA Here our data indicate that

MnmEG modifies tRNALeucmnm5UmAA in the absence of

20-O-methylation (Figures 5D and E 6E and F) but wedo not know whether TrmL requires the presence of the

cmnm5 group to act on tRNALeucmnm5UmAA More experi-

ments are needed to unravel the order of action of themodifying enzymes working on U34 and the identityelements required by each enzymeAnother interesting question concerning tRNA matur-

ation is whether physical interactions among tRNA modi-fication enzymes contribute to the efficiency of complexpathways In Aquifex aeolicus an interaction betweenthe E coli MnmC(m) homolog protein DUF752 andthe A aeolicus MnmE and MnmG proteins has beenobserved (41) In E coli we did not detect interaction ofMnmC with MnmE or MnmG by either fast-performanceliquid chromatography (FPLC) analysis (Figure 2F) orSPR (data not shown) However we cannot rule out thepossibility of the MnmEG complex being able to interactwith MnmC in vivo It appears reasonable to hypothesizethat the spatial clustering of tRNA-modifying enzymeswithin the cell can optimize the coordination of theirwork Therefore one challenging aim for future studiesis to determine whether tRNA modifying enzymes are spa-tially organized as observed for other enzymes involved inRNA biology such as the components of the RNAdegradosome (50)The use of strains carrying the DmnmC(o) or DmnmC

mutation has allowed us to study the ability of MnmEG touse the ammonium or the glycine pathway under differentgrowth conditions and thus to detect the reprogrammingof U34 base modification in both bulk and specific tRNAs(Tables 4 and 5 Figure 9) The analysis of bulk tRNAindicates that when DmnmC(o) or DmnmC strains aregrown in LBT the glycine pathway prevails over theammonium pathway in the exponential phase ascmnm5s2U is the most abundant nucleoside (Figure 9A)However a gradual change in the output of bothpathways is observed along the growth curve so that thenucleosides produced by the ammonium pathway becomeprevalent during entry into the stationary phase(Figure 9B) In contrast when cells are grown inminimal medium the glycine pathway is prevalent inboth the exponential and stationary phases (Figure 9A)A preference by MnmEG for the ammonium or theglycine pathway to catalyze tRNA modification maydepend on the availability of these substrates which inturn depends on the growth medium and cell metabolismThe nature of tRNA species also appears crucial for

the net output of the MnmEG pathways (Table 5 and

Figure 9) Thus whereas tRNALysmnm5s2UUU follows the

pattern observed in bulk tRNA and is preferably

modified via the ammonium pathway during the transition

to the stationary phase tRNAGlnethcTHORNmnm5s2UUG fails to be ef-

fectively modified through this pathway as indicated bythe low accumulation of nm5s2U in the DmnmC mutantMoreover the ammonium pathway proves ineffective in

modifying tRNALeucmnm5UmAA at any phase of growth as no

nm5U or mnm5U is detected in a trmL strain (Figure 5D

and E) Therefore neither tRNALeucmnm5UmAA nor

tRNAGlnethcTHORNmnm5s2UUG is a good substrate for the ammonium

pathway in vivo even though both tRNAs appear ratherwell modified by this pathway in vitro (Figure 6B and EFigure 9C and D)

Why the behavior of tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA is different from that of

tRNALysmnm5s2UUU remains unresolved The use of

ammonium or glycine by MnmEG could be regulatedby interactions of this complex with its tRNA substrates

and unknown factors in vivo tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA may be poor substrates for the

ammonium pathway and thus out-competed by other

tRNAs in vivo Alternatively tRNAGlnethcTHORNmnm5s2UUG and

tRNALeucmnm5UmAA carrying nm5 or mnm5 but not cmnm5

may be unstable and specifically degraded by some nucle-ase(s) (551) Additional experiments are required to testthese hypotheses and to elucidate the mechanismunderlying the selection by MnmEG of the glycine orthe ammonium pathway

Our results highlight the crucial role of MnmC(o) andMnm(m) in the biosynthesis of mnm5(s2)U in accordancewith growth conditions (Figure 9) During the exponentialphase in LBT and at any phase in minimal medium theoxidoreductase activity of MnmC(o) is required to drivethe intermediate generated by the more active glycinepathway [cmnm5(s2)U] toward the MnmC(m)-dependent

synthesis of the final modification in tRNALysmnm5s2UUU and

presumably in tRNAGlumnm5s2UUC tRNAArg

mnm5UCU and

tRNAGlymnm5UCC (Figure 9A) In contrast MnmC(o) plays

a minor role when the ammonium pathway prevailsie during the transition to stationary phase of culturesgrowing in LBT In this case the methylase activity ofMnmC(m) functions fairly well alone [ie without the par-ticipation of MnmC(o)] to drive the production of thefinal modification mnm5(s2)U by transforming the inter-mediate nm5(s2) generated by the more active ammoniumpathway (Table 4 and Figure 9B) Thus the incorporationof MnmC(o) and MnmC(m) into the enzyme content ofE coli extends the adaptability of this bacterium tochanging growth conditions How our findings in E coliapply to other bacteria especially those lacking certainMnmC activity is an issue that merits future research

The lack of modifications at the wobble uridine due tothe presence of mnmE or mnmG mutations leads toimpaired growth high sensitivity to acidic pH anddefects in translational fidelity(1235464852ndash55) Hereour functional studies suggest that MnmC impairment isless detrimental than the inactivation of the MnmEGcomplex given that mnmC mutations neither reduce the

2620 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 21: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

growth rate in rich medium (Table 6) nor affect resistanceto acidic pH (Figure 8) Nevertheless impairment of anyMnmC activity has a biological cost Thus mnmC muta-tions diminish the ability of cells to compete (Table 7)suggesting that fully modified tRNAs and therefore theactivities of MnmC are required for efficient translationof mRNAs involved in the stationary phase stressresponse In addition our data indicate that mnmC muta-tions slow the growth rate of a mnmA strain revealing asynergistic effect of mutations mnmC and mnmA AsmnmC mutations alone do not affect the growth rate we

hypothesize that the presence of the MnmA-dependentthiolation at position 2 of U34 in a subset of tRNAs cancompensate for the lack of MnmC modifications atposition 5 during exponential growthIn an mnmA background accumulation of nm5U

produced by the mnmC(m)-G68D mutation appearsslightly more detrimental than accumulation of cmnm5

caused by a null mnmC mutation (Table 6) In line withthis it has been demonstrated that the reading of anamber codon by an ochre-suppressing derivative of

tRNALysmnm5s2UUU (supG) is also more affected by the

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

MnmC(o)

Mnm

C(m

)

Gly

s2U34

U34

MnmA

MnmEG

cmnm5s2U34 nm5s2U34

mnm5s2U34

Gly

nm5s2U34

mnm5s2U34

s2U34

U34

MnmA

MnmEG

cmnm5s2U34

Mnm

C(m

)

NH4Gly

U34

MnmEG

cmnm5U34nm5U34

mnm5U34

Mnm

C(m

)

Gly

MnmC(o)

Mnm

C(m

)

Bulk tRNA MM- LogP and StatPBulk tRNA LBT- LogPtRNALys LBT- LogP

Bulk tRNA LBT- StatPtRNALys LBT- StatP

tRNAGln LBT- LogP and StatP tRNALeu LBT- LogP and StatP

NH4

NH4

BA

DC

NH4

Figure 9 Summary scheme of the reprogramming of U34 base modification in both bulk and specific tRNAs The main and minor modificationpathways are indicated by thick and thin arrows respectively pathways whose activity was only observed in vitro are indicated by dotted arrows

Activity of TrmL on tRNALeucmnm5UmAA has been omitted in panel D Labels in each panel indicate the type of tRNA (specific or bulk tRNA)

following the pathways and the growth conditions under which the tRNA was obtained LogP logarithmic phase StatP stationary phase MMminimal medium LBT LBT medium

Nucleic Acids Research 2014 Vol 42 No 4 2621

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 22: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

mnmC(m)-G68D mutation than by the null mnmC-W131stop mutation although this effect is dependent onthe codon context (56) Therefore it appears thatcmnm5(s2) confers the cell some advantage overnm5(s2)U to translate specific mRNAs during exponentialgrowthStrikingly in the competition experiments (Table 7) the

mnmC-W131stop mutation appears to be more disadvan-tageous than the mnmC(m)-G68D allele which reveals anopposite trend to that observed in the growth rate deter-mination experiments (Table 6) During the transition tothe stationary phase nm5(s2)U is more abundant inmnmC(m)-G68D than in mnmC-W131stop (Figure 7panels B and C) Therefore the slightly better ability ofthe mnmC(m)-G68D mutant to compete might indicatethat nm5(s2)U is more effective than cmnm5(s2)U intranslating mRNAs required to sustain stationary phasestress conditions Thus it appears that the advantagesoffered by each intermediate nm5(s2)U and cmnm5(s2)Udepend on the nature of the mRNAs to be translatedwhich in turn depend on the growth conditionsObviously the synthesis of the final modificationmnm5s2U is more advantageous to the cell as the strainsexpressing a fully active MnmC protein grow better understressing conditions (ie those related to lack of MnmAand competition with other bacteria) than those strainslacking any MnmC activity In conclusion adaptation ofMnmEG to use ammonium or glycine together with theacquisition of MnmC(o) and MnmC(m) activities provideE coli with significant advantages to synthesize mnm5s2Uand to survive under different conditions

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online

ACKNOWLEDGEMENTS

We are very grateful to M Villarroya S Prado R Ruiz-Partida and E Knecht for much valuable advice We alsothank the National BioResource Project (NIG Japan) forproviding some of the E coli strains used in this study

FUNDING

The Spanish Ministry of Economy and Competitiveness[BFU2007-66509 and BFU2010-19737] the GeneralitatValenciana [ACOMP2012065 PROMETEO2012061to M-EA] the Swedish Science Research Council[project BU-2930] and Carl Trygger Foundation (toGRB) a pre-doctoral fellowship from the SpanishMinistry of Economy and Competitiveness (to M-JG)IM was supported by short-term fellowships from theMinistry of Science and Innovation and GeneralitatValenciana during his stays at the Umea UniversityFunding for open access charge Spanish Ministry ofEconomy and Competitiveness [BFU2010-19737]

Conflict of interest statement None declared

REFERENCES

1 BjorkGR and HagervallTG (2005) Transfer tRNAmodification (posting data) Escherichia Coli and SalmonellaCellular and Molecular Biology ASM Press Washington DC

2 GrosjeanH (2009) Nucleic acids are not boring long polymers ofonly four types of nucleotides a guided tour DNA and RNAModification Enzymes Structure Mechanism Function andEvolution Landes Biosciencies Texas USA pp 1ndash18

3 El YacoubiB BaillyM and de Crecy-LagardV (2012)Biosynthesis and Function of Posttranscriptional Modifications ofTransfer RNAs Annu Rev Genet 46 69ndash95

4 AgrisPF (2008) Bringing order to translation the contributionsof transfer RNA anticodon-domain modifications EMBOReports 9 629ndash635

5 PhizickyEM and HopperAK (2010) tRNA biology charges tothe front Genes Dev 24 1832ndash1860

6 ZehariaA ShaagA PappoO Mager-HeckelAM SaadaABeinatM KarichevaO MandelH OfekN SegelR et al(2009) Acute infantile liver failure due to mutations in theTRMU gene Am J Hum Genet 85 401ndash407

7 SasarmanF AntonickaH HorvathR and ShoubridgeEA(2011) The 2-thiouridylase function of the human MTU1(TRMU) enzyme is dispensable for mitochondrial translationHum Mol Genet 20 4634ndash4643

8 GhezziD BaruffiniE HaackTB InnvernizziFMelchiondaL DallabonaC StormTM PariniRBurlinaAB MeitingerTM et al (2012) Mutations of themitochondrial-tRNA modifier MTO1 cause hypertrophiccardiomyopathy and lactic acidosis Am J Hum Genet 901079ndash1087

9 ChanCT PangYL DengW BabuIR DyavaiahMBegleyTJ and DedonPC (2012) Reprogramming of tRNAmodifications controls the oxidative stress response by codon-biased translation of proteins Nat Commun 3 937

10 ChenC HuangB EliassonM RydenP and BystromAS(2011) Elongator complex influences telomeric gene silencing andDNA damage response by its role in wobble uridine tRNAmodification PLoS Genet 7 e1002258

11 ArmengodME MoukadiriI PradoS Ruiz-PartidaRBenitez-PaezA VillarroyaM LomasR GarzonMJMartinez-ZamoraA MeseguerS et al (2012) Enzymology oftRNA modification in the bacterial MnmEG pathway Biochimie94 1510ndash1520

12 YimL MoukadiriI BjorkGR and ArmengodME (2006)Further insights into the tRNA modification process controlledby proteins MnmE and GidA of Escherichia coli Nucleic AcidsRes 34 5892ndash5905

13 MoukadiriI PradoS PieraJ Velazquez-CampoyABjorkGR and ArmengodME (2009) Evolutionarily conservedproteins MnmE and GidA catalyze the formation of twomethyluridine derivatives at tRNA wobble positions NucleicAcids Res 37 7177ndash7193

14 CabedoH MacianF VillarroyaM EscuderoJCMartınez-VicenteM KnechtE and ArmengodME (1999) TheEscherichia coli trmE (mnmE) gene involved in tRNAmodification codes for an evolutionarily conserved GTPase withunusual biochemical properties EMBO J 18 7063ndash7076

15 ScrimaA VetterIR ArmengodME and WittinghoferA(2005) The structure of the TrmE GTP-binding protein and itsimplications for tRNA modification EMBO J 24 23ndash33

16 OsawaT InanagaH and NumataT (2009) Crystallization andpreliminary X-ray diffraction analysis of the tRNA-modificationenzyme GidA from Aquifex aeolicus Acta Crystallogr Sect FStruct Biol Cryst Commun 65 508ndash511

17 ShiR VillarroyaM Ruiz-PartidaR LiY ProteauAPradoS MoukadiriI Benıtez-PaezA LomasR WagnerJet al (2009) Structure-function analysis of Escherichia coliMnmG (GidA) a highly conserved tRNA-modifying enzymeJ Bacteriol 191 7614ndash7619

18 MeyerS ScrimaA VerseesW and WittinghoferA (2008)Crystal structures of the conserved tRNA-modifying enzymeGidA implications for its interaction with MnmE and substrateJ Mol Biol 380 532ndash547

2622 Nucleic Acids Research 2014 Vol 42 No 4

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from

Page 23: MnmC enzymes depends on the growth conditions and the tRNA ...umu.diva-portal.org/smash/get/diva2:710872/FULLTEXT01.pdf · (s2)u 34 u34 (mnma) mnme•mnmg cmnm5(s2)u 34 nm 5(s2)u

19 PradoS VillarroyaM MedinaM and ArmengodME (2013)The tRNA-modifying function of MnmE is controlled bypost-hydrolysis steps of its GTPase cycle Nucleic Acids Res 416190ndash6208

20 IkeuchiY ShigiN KatoJ NishimuraA and SuzukiT (2006)Mechanistic insights into sulfur relay by multiple sulfur mediatorsinvolved in thiouridine biosynthesis at tRNA wobble positionsMol Cell 21 97ndash108

21 Benıtez-PaezA VillarroyaM DouthwaiteS GabaldonT andArmengodME (2010) YibK is the 2rsquo-O-methyltransferase TrmLthat modifies the wobble nucleotide in Escherichia colitRNA(Leu) isoacceptors RNA 16 2131ndash2143

22 HagervallTG and BjorkGR (1984) Genetic mapping andcloning of the gene (trmC) responsible for the synthesis of tRNA(mnm5s2U)methyltransferase in Escherichia coli K12 Mol GenGenet 196 201ndash207

23 HagervallTG EdmondsCG McCloskeyJA and BjorkGR(1987) Transfer RNA(5-methylaminomethyl-2-thiouridine)-methyltransferase from Escherichia coli K-12 has two enzymaticactivities J Biol Chem 262 8488ndash8495

24 BujnickiJM OudjamaY RooversM OwczarekS CailletJand DroogmansL (2004) Identification of a bifunctional enzymeMnmC involved in the biosynthesis of a hypermodified uridine inthe wobble position of tRNA RNA 10 1236ndash1242

25 RooversM OudjamaY KaminskaKH PurtaE CailletJDroogmansL and BujnickiJM (2008) Sequence-structure-function analysis of the bifunctional enzyme MnmC that catalysesthe last two steps in the biosynthesis of hypermodified nucleosidemnm5s2U in tRNA Proteins 71 2076ndash2085

26 PearsonD and CarellT (2011) Assay of both activities of thebifunctional tRNA-modifying enzyme MnmC reveals a kineticbasis for selective full modification of cmnm5s2U to mnm5s2UNucleic Acids Res 39 4818ndash4826

27 KitamuraA SengokuT NishimotoM YokoyamaS andBesshoY (2011) Crystal structure of the bifunctional tRNAmodification enzyme MnmC from Escherichia coli Protein Sci20 1105ndash1113

28 MillerJH (1992) A Short Course in Bacterial Genetics Alaboratory manual and handbook for Escherichia coli and relatedbacteria Cold Spring Harbor Laboratory Press Cold SpringHarbor NY

29 DatsenkoKA and WannerBL (2000) One-step inactivation ofchromosomal genes in Escherichia coli K-12 using PCR productsProc Natl Acad Sci USA 97 6640ndash6645

30 ElseviersD PetrulloLA and GallagherPJ (1984) Novel E colimutants deficient in biosynthesis of 5-methylaminomethyl-2-thiouridine Nucleic Acids Res 12 3521ndash3534

31 NilssonK LundgrenHK HagervallTG and BjorkGR(2002) The cysteine desulfurase IscS is required for synthesisof all five thiolated nucleosides present in tRNA fromSalmonella enterica serovar typhimurium J Bacteriol 1846830ndash6835

32 Martinez-VicenteM YimL VillarroyaM MelladoMPerez-PayaE BjorkGR and ArmengodME (2005) Effects ofmutagenesis in the switch I region and conserved arginines ofEscherichia coli MnmE protein a GTPase involved in tRNAmodification J Biol Chem 280 30660ndash30670

33 Benıtez-PaezA VillarroyaM and ArmengodME (2012) TheEscherichia coli RlmN methyltransferase is a dual-specificityenzyme that modifies both rRNA and tRNA and controlstranslational accuracy RNA 18 1783ndash1795

34 TamuraK HimenoH AsaharaH HasegawaT andShimizuM (1992) In vitro study of Ecoli tRNA(Arg) andtRNA(Lys) identity elements Nucleic Acids Res 20 2335ndash2339

35 GehrkeCW and KuoKC (1989) Ribonucleoside analysis byreversed-phase high-performance liquid chromatographyJ Chromatogr 471 3ndash36

36 EmilssonV and KurlandCG (1990) Growth rate dependence oftransfer RNA abundance in Escherichia coli EMBO J 94359ndash4366

37 PonchonL BeauvaisG Nonin-LecomteS and DardelF (2009)A generic protocol for the expression and purification of

recombinant RNA in Escherichia coli using a tRNA scaffoldNat Protoc 4 947ndash959

38 SuzukiT and SuzukiT (2007) Chaplet column chromatographyisolation of a large set of individual RNAs in a single stepMethods Enzymol 425 231ndash239

39 GongS MaZ and FosterJW (2004) The Era-like GTPaseTrmE conditionally activates gadE and glutamate-dependent acidresistance in Escherichia coli Mol Microbiol 54 948ndash961

40 VogelHJ and BonnerDM (1956) Acetylornithinase ofEscherichia coli partial purification and some properties J BiolChem 218 97ndash106

41 KitamuraA NishimotoM SengokuT ShibataR JagerGBjorkGR GrosjeanH YokoyamaS and BesshoY (2012)Characterization and structure of the Aquifex aeolicus proteinDUF752 a bacterial tRNA-methyltransferase (MnmC2)functioning without the usually fused oxidase domain (MnmC1)J Biol Chem 287 43950ndash43960

42 PonchonL and DardelF (2007) Recombinant RNA technologythe tRNA scaffold Nat Methods 4 571ndash576

43 YamanakaK HwangJ and InouyeM (2000) Characterizationof GTPase activity of TrmE a member of a novel GTPasesuperfamily from Thermotoga maritima J Bacteriol 1827078ndash7082

44 ChenP CrainPF NasvallSJ PomerantzSC and BjorkGR(2005) A lsquolsquogain of functionrsquorsquo mutation in a protein mediatesproduction of novel modified nucleosides EMBO J 241842ndash1851

45 MarinusMG MorrisNR SollD and KwongTC (1975)Isolation and partial characterization of three Escherichia colimutants with altered transfer ribonucleic acid methylasesJ Bacteriol 122 257ndash265

46 KanjeeU and HouryWA (2013) Mechanisms ofacid resistance in Escherichia coli Annu Rev Microbiol 6765ndash81

47 KrugerMK PedersenS HagervallTG and SorensenMA(1998) The modification of the wobble base of tRNAGlumodulates the translation rate of glutamic acid codons in vivoJ Mol Biol 284 621ndash631

48 Benıtez-PaezA VillarroyaM and ArmengodME (2012)Regulation of expression and catalytic activity of Escherichia coliRsmG methyltransferase RNA 18 795ndash806

49 SullivanMA CannonJF WebbFH and BockRM (1985)Antisuppressor mutation in Escherichia coli defective inbiosynthesis of 5-methylaminomethyl-2-thiouridine J Bacteriol161 368ndash376

50 BandyraKJ BouvierM CarpousisAJ and LuisiBF (2013)The social fabric of the RNA degradosome Biochim BiophysActa 1829 514ndash522

51 PhizickyEM and AlfonzoJD (2010) Do all modificationsbenefit all tRNAs FEBS Lett 584 265ndash271

52 BregeonD ColotV RadmanM and TaddeiF (2001)Translational misreading a tRNA modification counteracts a +2ribosomal frameshift Genes Dev 15 2295ndash2306

53 YimL Martinez-VicenteM VillarroyaM AguadoCKnechtE and ArmengodME (2003) The GTPase activity andC-terminal cysteine of the Escherichia coli MnmE protein areessential for its tRNA modifying function J Biol Chem 27828378ndash28387

54 UrbonaviciusJ QianQ DurandJM HagervallTG andBjorkGR (2001) Improvement of reading frame maintenance isa common function for several tRNA modifications EMBO J20 4863ndash4873

55 BrierleyI MeredithMR BloysAJ and HagervallTG (1997)Expression of a coronavirus ribosomal frameshift signal inEscherichia coli influence of tRNA anticodon modification onframeshifting J Mol Biol 270 360ndash373

56 HagervallTG and BjorkGR (1984) Undermodification in thefirst position of the anticodon of supG-tRNA reducestranslational efficiency Mol Gen Genet 196 194ndash200

Nucleic Acids Research 2014 Vol 42 No 4 2623

at Um

ea University L

ibrary on April 8 2014

httpnaroxfordjournalsorgD

ownloaded from