Nucleic Acids Research, 1994, Vol. 22, No. 1 41-46

dutA RNA functions as an untranslatable RNA in thedevelopment of Dictyostelium discoideum

Hiderou Yoshida, Hiroshi Kumimoto and Koji OkamotoDepartment of Botany, Faculty of Science, Kyoto University, Kyoto 606-01, Japan

Received October 7, 1993; Revised and Accepted December 6, 1993 DDBJ accession no. D16417

ABSTRACT

dutA is a gene specifically expressed during thedevelopment of Dictyostelium discoideum. Towardunderstanding Its possible role In development, weIsolated and characterized the gene and Its completecDNA. We found that dutA is encoded by the nucleargenome as a single copy gene without Introns. Inaddition, the following unique and Interesting featuresoldutA RNA (1322 nt) emerged: (1) It has no sustainedORFs (MAX = 126 nt) (2) it Is extremely AU-rich (83%)(3) it contains peculiar sequence motifs (largepalindromes, long AU-stretches and GC-clusters) (4) Itis localized In the cytoplasm but completely absentfrom ribosomes. These features suggest that dutA RNAfunctions without being translated into protein.Disruption of the dutA gene did not cause phenotyplcchanges, suggesting that the function of dutA isredundant.

INTRODUCTION

There is a class of RNA (structural RNA) which exhibits itsfunction by itself without being translated into protein, forexample, rRNA, tRNA or RNAs in ribozymes (1, 2) andspliceosomes (3, 4). Recently, several new structural RNAs havebeen reported. Mouse H19 RNA (5, 6) is induced duringembryogenesis and has tumor-suppressor activity. Human X1STRNA (7) is thought to be responsible for the inactivation of theX chromosome. However, their action mechanisms are stillunclarified.

We previously identified a gene (previously called DC6) inDictyostelium discoideum, whose expression is strictly regulatedin development through the interaction between cells, and thusmay be implicated in the developmental process (8). In the presentstudy, we cloned the gene, characterized its complete cDNA,examined the subcellular localization of its RNA and disruptedthe gene. The peculiar sequence (such as no sustained ORF, longpalindromes) and the unexpected localization of its RNA (thecomplete absence from ribosomes) suggest that this RNA is notan mRNA but a structural RNA. We thus refer this gene to dutA(development-specific but untranslatable RNA). Our preliminaryresults suggest that cognate sequences are widespread amongorganisms.

MATERIALS AND METHODS

General methods

The strain Ax2 of D. discoideum was used and all manipulationsof cells were performed according to Sussman (9). We usuallyused the cells (T(15) cells), which developed for 15hr insuspension (20 mM Na2HPO4/KH2PO4 pH7.0, 2 mM MgSO4).Transformation of Ax2 cells was carried out by the calciumphosphate method (10).

DNA and RNA manipulations were performed as describedby Maniatis et al. (11). DNA labeling and cDNA libraryconstruction were done with a multiprime DNA labeling system(Amersham) and a You-Prime cDNA synthesis kit (PharmaciaLKB), respectively. Filter hybridization with DNA probes wascarried out for 12 hr at 42 °C in the presence of 50% formamideand 5XSSPE, and washed at 37°C or 50°C in 0.1 XSSC and0.1% SDS.

PCR was performed with a thermal cycler B-641 (KURABO)according to the manufacture's instruction. For RT-PCR, poly(A)RNA was reverse-transcribed and resultant cDNA was amplifiedwith Tth polymerase (TOYOBO) according to Myers and Gelfand(12).

Isolation of genomk clonesGenomic clone EE2900 was isolated from genomic mini-library.Nuclear DNA of Ax2 cells was digested with EcoRI andseparated in 0.8% TAE-agarose gel. 2.3-4.3 kb fragments ofdigested DNA were collected and inserted into plasmid pUCl 18.This library was screened with labeled C6A probe (a previouslyisolated cDNA clone). ME900 were isolated by inverse PCR(13). Nuclear DNA which was digested with Mbol and ligatedto be circularized was used as a template, and amplification wascarried out with primer H and I (Fig. 3).

Fractionation of cellular componentsDeveloped Ax2 cells (in the slug stage) were mildly lysed in thelysis buffer (50 mM Hepes (pH 7.5), 1% Triton X-100, 10%sucrose and 5 mM MgSO4). The lysate was fractionated bydifferential centrifugation, i.e., at 400 g for 5 min and then at2,000 g for 5 min. First precipitate (PI: cell debris) and secondprecipitate (P2: nuclear fraction) were resuspended in lysis buffer.Second supernatant (S2: cytosol-organelle fraction) was furtherfractionated by 15—40% sucrose density gradient centrifugationat 100,000 g for 3 hr.

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018

42 Nucleic Acids Research, 1994, Vol. 22, No. 1

H ET T

X BT T

EE2900-ME900-

Fragmei"[ = -EX889-0EL15OOL

ED211 -XE2000

C E

— S I napping

-C6A-C6B

Pr iraers

cOHA c lones

1.0kb •dutA RNA

Figure 1. (A) Physical map of the dutA gene derived from genomk Southernanalysis. The locations of genomic clones and their fragments used in this studyare shown in lines. The enzymes shown are: E, EcoRI; M, Mbol; X, Xhol; B,Bgin. (B) A strategy for the determination of dutA RNA sequence. C6A is apreviously-isolated cDNA clone. C6B and C6C were the longest cDNA clonesisolated from oligo(dT)-primed cDNA library and primer E-primed cDNA library,respectively. C6D was cloned by RT-PCR method with primer A and C. 5'-regionof dutA RNA was determined by SI nuclease mapping.

rVEfr|T G C A + - + -

M-3i

••

Figure 2. SI nuclease protection assay. Poly(A) RNA of t l 3 cells (DEV) orvegetative cells (VEG) was hybridized with the labeled ncmcoding strand of ED211,digested with 30 units of SI nuclease (+) or undigested ( - ) (reaction volume= 200 /J), and separated on a 6% acrylamide-urea gel. Protected fragmentsare indicated by dots. A sequencing ladder of ED211 was run in parallel.

Construction of the disruption vector and the antisense vector

The disruption vector was made as follows: EE2900 was insertedinto pUC119 at the EcoRI site and a plasmid in which theupstream region in EE2900 was located near the BamHI site wasselected (PUC-EE2900S). EE800 fragment was blunted withKlenow fragment and inserted into pUC119 at the HincII siteand a plasmid in which the upstream region in EE800 was locatednear the PstI site (pUC-EE800S). BamHI-PstI fragmentcontaining EE800 was excised from pUC-EE800S and insertedinto pUC-2900S at the BamHI-PstI site (pUC-EE800S-

aaatttattaatttsatttattattttttctttttttttttttttttttttttttttttt

tiatsaaaagatacacctcaaaaacatttcagttgactgtattaactttgttgtgatatt

attttcatatttttttatcataaatgaaattattaagaaactcactgattgaacgtttct[TATAn TT T

agat t t tassataa ta taaaaaaatgaat tcaaataa t tagt tc t taTTTTGAAAATGAA

AAAAAAGTGGAATTTAATTAATTAAAI11111111IAI1111111IAI1111111111IG

GAAATMfTGGCTATG(aCCAAATG6AAAATAAAAAAGGAAAAAAAnAAAATAAGAATT

AAATTTATAATAAAAAAAAAAAAAAAAAAAAAAAAGAAAAAAAATAAAAnAAAAAATAA

AATAAAATAMAATAA^AMT+AAGTTGTAATGTGnATCAAAAATTTTnATTTTTTTA

nGGGGTAAAAAAAAAAAAAAAATAAAAAAAAAATAAnAAAAATTAAAAACACACATTA

BTTG(^TTTaGAT/WTAATAAAATTrTTTTTAan/WTMCT(rrTACAGATAATTTA

TTTAATTTTGAAnAAATnATTTTTTTAATAATTTTTTOkAAAACTATTTAAGAGACAA

AATTTTTTTATrnGTrATTCAAAAATGTCAAATAAATAAAAAAAAAAAAAAAAAAATAA

AAAAAAATnAATM(rrTTTAAAATGGTATA(nTTTAATAAAAAAAAAAAAAAAACCACT

ATCAHAGTACT/WAAAAAAAATAAAAAATAAAAAAAAAAAAAATAAAAAAAAATAATAA

TAAGCTATAIXATGGGACTTAGAGAAAATAAAAAAGTTTAAAAAAAAAATAAAAAATAAT

AAAAAATAAAAAAAAAATAAAAAAAAAAACTTnAAAACTCTCCTTCTACTTTTTGOTGCI priaar J =>

nCTGnAAaTAATGaAAGGTGTaCTaCCCCCTTTGCTTTGCAnACTCTCCTCAApriwr K —=>i priwr F — o

^TMT(rra5AGCTCaACGATGCT(rprimer DCCTATGCAA

pnaer Ep <=— p

CAaTTCCTATGCAAaaCTCAAATGTATAAGCACTCnTAACAAAAAAAAATATTTAA

AA^^AAAAAAAAAAGAAAAAAAAAAAATATTAAAATATTTTTTTTAAAATCAATTTTAAA

TATAAATAATAATUI11111111IIIIATAAAAAAAAGAAAGAAAAAAAGAAAAAAAAAA

AAAAAAGAAAAAAAGAATTTAAAAAAAAAAAAAAAAATAGAAAAAAAAAAGTTAAAGATG

TAATl^AAAAkAAA(^AGAAT(3GAAATCCAAAAAAAGATCTTrATAAATTGCCACTATAATCAATGTATAGATAnGTTCCGTCTGTGTGTATarrGT+AGGTCCAGTTTGCTCTTTA

<^>— priner G '

GTT/(CTTTAATnAAATCTATATTGTAATTTTTATAAAAAAAAAAAAAAgaaaaaaaaaa

acaaoaaaaaaaagaaaaaagaaaaaataataataataataataataataatacaataaa

tttaaataaaaataaagaataaaccttatagtgggtttaccaatttataaaaaaaatatt

tacatatgctttttattttttatotctttttttatattttcagtttttttatcaactcga

ttgaccactaaaaaaaaataaaaatataataaaaaaaaataaaaaaataaaaattaaaaa

tataaattttaaaaaaaacaatgatattactattttagataatgaastgtttaatttatt

-168

-108

-48

13

73

133

193

2S3

313

373

433

493

553

613

673

T33

793

853

913

973

1033

1093

1153

1213

1273

1333

1393

1453

1513

1573

1633

Flgure 3. Nucleotide sequence of the dutA gene. The coding strand is shownin the 5'-to-3' direction. Numbers on the right indicate nucleotide sequencepositionj. Initiation sites of transcription and the TATA box are marked. Nucleotidesequences that are transcribed into dutA RNA are in capital letters and those nottranscribed are in lowercase letters. Location of the oUgonucleotide primers usedin this study are shown above or below the nucleotide sequence. This sequencedata has been submitted to DDBJ, EMBL and GenBank and assigned the accessionnumber D16417.

EE2900S). Xbal-Xbal fragment containing actin 15promoter-aminoglycoside 3' phosphotransferase gene-actin 15terminator (NeoO was excised from pDNeoII and inserted intopUCl 19 at the Xbal site and a plasmid in which 5' of Neo7 waslocated near the Sail site was selected (pUC-NeorS).BamHI-Sail fragment containing Netf was excised from pUC-NeorS and inserted into pUC-EE800S-EE2900S at the Xhol(located in EE2900)- BamHI site (pUC-EE800S-NeorA-XE2000S). This plasmid was cut with EcoRI and PstI to liberatethe insert (EESOOS-NetfA-XKOOOS) and Ax2 cells weretransformed with it. To create the antisense vector, EX889fragment was inserted into plasmid pDNeoII (14) at theEcoRI-Sail site to allow the antisense RNA expression(pDNeo-EX889A).

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018

Nucleic Acids Research, 1994, Vol. 22, No. 1 43

1 2 3 4 5 6

•5 .0kb

• 1 0 U >

Figure 4. Southern analysis of nuclear DNA of the strain Ax2. Nuclear DNAwas digested with Hindin and Kpnl (lane 1), BamHI and HindHI (lane 2), Xholand PstI Oane 3), BamHI and Xhol (lane 4), BamHI and PstI (lane 5), and EcoRI(lane 6), separated on a 0.8% TBE-agarose gel, transferred to nylon membraneand hybridized with the 32P-labeled C6A DNA. The numbers on the side of theMot indicate the length of DNA fragments in kb. The difference in the hybridizationsignal intensity could be due to the poor quality of our DNA preparation (enrKhedin low molecular weight fragments).

RESULTSThe structure of the dutA gene

We have previously isolated a partial cDNA clone of dutA RNA(C6A in Fig. IB) (8). Using this clone, overlapping cDNA cloneswere isolated from an oligo(dT)-primed cDNA library and aninternally-primed cDNA library (Fig. IB). However, all thesestandard methods yielded only partially extended cDNA, probablydue to its peculiar sequences. Therefore a cDNA of the 5'-halfregion of dutA RNA was isolated by RT-PCR (12). Theinitiation site of transcription was determined by SI nucleaseprotection assay (Fig. 2). All clones isolated from the oligo(dT)library had the same 3'-terminus as C6A, indicating that thetranscription of dutA RNA terminates at this point. From these,we determined the entire sequence of dutA RNA (1332 nt) (Fig.3).

Genomic Southern analysis shows that dutA is present as asingle copy gene in the haploid genome (Fig. 4). This conclusionwas confirmed by the gene disruption experiment (see below).We also performed Southern blot analysis at lower stringencies(blots were washed in 6xSSC, 4xSSC or 2XSSC at 42°C) butno related genes were detected. Two genomic clones (EE2900and ME900) that cover the dutA gene were isolated and sequenced(Fig. 1A). Since restriction maps of these clones perfectly agreedwith that derived from genomic Southern analysis, DNArearrangement should not have occurred in the cloning process.The restriction map of dutA gene is not consistent with that ofmitochondrial DNA which was previously reported (15),indicating that dutA is not a mitochondrial but a nuclear gene.This conclusion was confirmed by Southern analysis with purifiedmitochondrial DNA (data not shown).

The cDNA sequence is completely coincident with thecorresponding region of the genomic sequence, implying that thedutA RNA is encoded by one long exon. Mung bean nucleaseprotection assay gave rise to about a 1.2-1.3 kb protectedfragment (data not shown) and this also indicates that dutA RNAis encoded by one long exon. As for a poly (A) tail, there is apoly (A) tract at the 3 '-end of cDNA sequence but it is not certainwhether this tract is posttranscriptionally added or transcribedfrom the genomic sequence, since a similar poly(A) tract ispresent at the corresponding position of the genome. The absence

(A)

L i l l JiL LKkb)

dutA MA

T.C. II

T.C. i

i.e. " I '

T'C. I

I1 I ' I" I ' I

ii i ii ii i n i nil in i i i I I II

r r M| ' i i " i"

I I I II I I I I I ! i l l l I l l l I I

TT' r i " i "I I I II II I I I I I III I I I I III I Hill

01

L32J L 7 1 J L34J Palindro

(D)

T-AT-A

A

T

T

A

•

T-A< 1

A • AA a A

T V /T T -*

A

A

•A-A

-A-A

-A-A

-T-1

-A-A

T

TI

1

C

•- 1

A AT-AA-T

A I*-rT-*T-A1-T

T |A-TA-TA-TA-TA-TA-TA-T

A-T

A-T

T

1T

TIr

Figure 5. The structure of dutA RNA. (A) The location of the initiation codonand termination codons. Three forward reading frames are shown. I and i indicatethe location of the initiation codon (I.C.) and the termination codon (T.C.),respectively. Each ORF is numbered. (B) Distribution of GC-content. AverageGC-contents in a 100 nt of window are calculated and plotted. (C) The locationof palindromes. Lower limits of length and matching are 30 nt and 75%,respectively. Numbers indicate their length. (D) The nucleoode sequences ofpalindromes.

of the consensus sequence for the polyadenylation signal on theRNA supports the latter possibility.

In the upstream non-coding region (Fig. 3), there is apresumptive TATA box (-27 bp) but neither a G-rich elementnor a CA-rich element, which are the proposed cis-regulatoryelements of transcription in D.discoideum (17). This agrees withits peculiar characteristics that the expression is not affected bycAMP (8) or DIF-1 (unpublished), which regulate the expressionof many previously investigated genes of D.discoideum.

Protein coding potential of dutA RNATo deduce the amino acid sequence of the putative dutA product,we searched the entire RNA sequence for any possible ORFs.However, multiple stop codons appear very frequently in all 6

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018

44 Nucleic Acids Research, 1994, Vol. 22, No. 1

(A) 1 2 (B) 1 2

•I »•(O

ittif i jfifj

(A) C6A

1.0kb

E E XhBT T 1 !

dutA RMA

EE800 1 Neor

tmmmmP X

RNA-coding ration

A15P-Noor-A15T

1

kH

\

1X

E

H

ET

XE2000

E«AI BXbal PHindi 11 Xh

s.

\

Bel iPstlXhol

3'

|iE

Chroaoson

DisriptionConitruct

m1 2 3 4 1 2 3 4 5 6

(E)

Figure 6. (A) The Subcellular Localization of dutA RNA. A lysate of slug cellswas fractionated into cell debris fraction, nuclear fraction (lane 1) andcytoplasm-organelle fraction (lane 2). RNA was extracted from each fraction,separated, transferred and probed with EX889. RNA from the same number ofcells was loaded. (B) The same blot was probed with Dd8 DNA. (C-E) Thecytoplasm—organelle fraction obtained in (A) was further fractionated by sucrosedensity gradient centrifugation. The bottom fraction is left and the top is right.(Q The optical density profile (wave length •= 260 nm). (D) RNA was extracted,transferred and probed with EX889. (E) The same blot was probed with actin6 cDNA of D.discoideum.

reading frames (on the sense strand and the antisense strand) andthus no sustained ORFs could be found (Fig. 5A). The longestpossible ORF identified is only 126 nt long, which is less than10% of the length of dutA RNA (the longest ORF that initiatesat a GUG codon is 156 nt long but GUG has not been reportedto be used as an initiation codon in D.discoideum). Moreover,the context of ATG in each ORF was quite different from theoptimal consensus of the initiation context among D. discoideumgenes (AAAATG: Yoshida and Okamoto, unpublished; someORFs contain complete consensus sequence but are quite short(<, 27 nt)). It is very unlikely that the sequence we determinedwas distorted by rearrangement during the cloning processbecause the sequences of all genomic and cDNA clones, whichwere independently isolated, completely accord with each other.

Moreover, dutA RNA has some other peculiar characteristics:first, the nucleotide sequence is composed largely of A and Uresidues (A = 51.1%, U = 32.1%, G = 8.6%, C = 8.2%),which frequently appear as homopolymer tracts, especially inthe 5'-region. Second, G and C residues are not randomlyscattered over dutA RNA but rather localized in two GC richdomains (Fig. 5B). Third, the sequence of dutA RNA has six

Figure 7. (A) The strategy for the disruption of dutA gene. The disruption constructhas the EE800 and XE2000 fragments interrupted by Tn903 phosphotransferasegene with the actin 15 promoter and terminator. (B) Southern blot analysis oftransformants. Genomic DNA extracted from parental strain (Ax2: lane 1 and3) and transformants (TF503: lane 2 and 4) was digested with EcoRI (lane 1and 2), EcoRI+Xbal (lane 3 and 4), separated, transferred and probed with C6A.(Q Northern blot analysis of disruptants. Cells of disrupted strains were developedon fitters for 15 hours and then cellular RNA was extracted from them, transferredand probed with C6A. (lane 1) parental strain (Ax2); (lanes 2-5) disruptant(TF5O3, TF505, TF511 and TF555); (lane 6) non-disrupted transfbrmant (TF541).The upper band in lane 6 corresponds to the endogenous dutA RNA (the gel slightlysmiled). The lower band in lane 6 could be a transcript derived from a vectorrandomly inserted near the promotor of the other gene.

large palindromes and the largest one extends over 71 nt (Fig.5C and 5D).

The above data are most consistent with the idea that duiA RNAdoes not encode protein but functions as a structural RNA.Computer searches with available data bases (by IDEAS program)could not find any other nucleotide sequences that had significantsimilarity to dutA RNA.

Subcellular localization of dutA RNAThe subcellular location of dutA RNA was analyzed to assesswhether dutA RNA is transported into the cytoplasm, and whetherit associates with the translational machinery of the cell. Cellsin the slug stage were mildly disrupted and fractionated bydifferential centrinjgation into low speed precipitate (cell debris),high speed precipitate (nuclear fraction), and high speedsupernatant (cytosol-organelle fraction). The majority of dutARNA was fractionated into the cytosol -organelle fraction ratherthan the nuclear fraction (Fig. 6A), whereas that of Dd8 RNA(small nuclear RNA of D.discoideum identified by Kaneda (18))was in the nuclear fraction as expected (Fig. 6B).

The cytosol-organelle fraction was then further fractionatedby sucrose density gradient centrifugation (Fig. 6C-E). ActinRNA was fractionated into both polysomal fractions and lighter

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018

Nucleic Acids Research, 1994, Vol. 22, No. 1 45

1 2 3 4 5 6 7

•43.0kb

W

dutA MA

£ t 1 faf. % pr Isars

F

Figure 8. Southern analysis of the genwnic DNA of S.cerevisiae. DNA wasdigested with EcoRI (lane 1), EcoRI + PstI (lane 2), EcoRI + BamHI (lane3), EcoRI + Xhol (lane 4), Dral (lane 5), Mbol (lane 6) and Haem (lane 7),separated, transferred and hybridized with the ^P-labded full length cDNA. Theblot was washed in 2xSSC at 42°C.

m1 2 3 4 5 6 7 1 2 3 4 5 6 7

fractions, while dutA RNA was found only in the lighter fraction,indicating that dutA RNA does not associate with ribosomes. Thisis consistent with the absence of protein-coding potential of dutARNA.

Disruption of dutA geneTo directly assess the function of dutA RNA in vivo, We madeantisense mutagenesis but the phenotype of antisensetransformants were all normal in development (data not shown).Then we disrupted the dutA gene by gene targeting in the haploidstrain. The gene replacement construct EE800S-NeorA-XE2000S (Fig. 7A), which contains the genomic fragmentsEE800 and XE2000 interrupted by the bacterial neomycinphosphotransferase gene (14), was introduced into Ax2 cells.Disruptants were selected from neomycin-resistant transformantsby genomic Southern hybridization (Figure 7B). By theseprocedures, we obtained 4 disruptants out of 74 transformants.In these disruptants, dutA RNA was completely lost (Fig. 7C).However, all of them showed normal morphology in developmentas far as examined. This suggests that dutA function is redundantin this organism.

Cognate sequences in other organismsNext we examined whether the sequences similar to dutA arepresent in the genomes of other organisms. As shown in Fig.8, a cognate sequence in Saccharomyces cerevisiae was detectedin the lower stringency genomic Southern blot. Cognatesequences in other organisms were also examined by using PCRand Southern hybridization. Genomic DNAs prepared from othercellular slime molds (D.mucoroides, D. rosarium andD.purpleum), S.cerevisiae, Schizosaccharomyces pombe,Drosophila melanogaster, Oryzias latipes (killifish), Musmusculus (mouse), Bos taurus (bovine), Arabidopsis thaliana andLemna paucicostata (duck weed) were used as a template andPCR was performed with 7 pairs of primers (Fig. 9A). AmplifiedDNA fragments were Southern-blotted and probed with dutA(DEL1500L). As shown in Fig. 9C—F (data of several organismsare not shown), several bands were detected in all of theorganisms examined. It is unlikely that these were amplified fromcontaminated DNA of D.discoideum because their sizes were notcompletely identical to those expected from D.discoideum (Fig.9B), and some of the pairs of primers did not yield hybridizablebands. These results suggest that dutA like sequences arewidespread from lower eukaryotes to mammals and plants.

(D) CE)1 2 3 4 5 6 7 1 2 3 4 5 6 7

H

(F)

1 2 ^3 4 5 6 7

Figure 9. (A) The location of primers for PCR study. The length of primersis 20 m. The precise location is shown in Fig. 3. (B)-(F) PCR was made withgenomic DNA of various organisms, i.e., (B) D.discoideum, (Q S.cerevisiae,(D) D.melanogaster, (E) mouse (balb/c), (F) A.thalUma. Amplified DNA wasseparated, transferred and probed with the "P-labeled full length cDNA. Blocswere washed in 0.1 xSSC at 42°C. (lane 1) primer A and C: (lane 2) primerA and D: (lane 3) primer A and E: (lane 4) primer B and C: (lane 5) primerB and D: (lane 6) primer B and E: (lane 7) primer F and G.

However, we must await further investigation to ascertain thatthose sequences are transcribed.

DISCUSSIONDoes dutA encode a protein?D.discoideum uses universal codons in nuclear and mitochondrialgenes (19). When the sequence of dutA RNA is converted intothe amino acid sequence according to the universal codons, stopcodons appear so frequently in all 3 reading frames that any

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018

46 Nucleic Acids Research, 1994, Vol. 22, No. 1

potential ORFs are too short ( ^ 126 base) to encode a protein.Moreover, dutA RNA shows extremely low GC-content (17%G/Q, compared to the minimal GC-content necessary to encodean average protein (36% G/Q. \n D.discoideum, protein codingregions are relatively GC-poor (— 38% G/C) but not as poor asdutA (19). This suggests that dutA RNA does not encode aprotein.

One possibility is that a functional ORF is generated byposttranscriptional nucleotide insertion (RNA editing) (20).However, we have never observed edited dutA RNAs in anycDNAs we cloned. Thus it is unlikely that editing makes dutARNA translatable.

Finally, the most compelling evidence for dutA RNA not beingan mRNA is the absence of dutA RNA from ribosomes in thecytoplasm. From these, we conclude that dutA RNA does notencode protein.

How does dutA exhibit its function?The absence of coding potential indicates that dutA RNA worksby itself without being translated into protein, that is, as astructural RNA. Its peculiar nucleotide sequence (AU tracts, GCislands and palindromes) might be important for the conformationof RNA, association with proteins and its function. An attemptto calculate the most stable secondary structure of dutA RNA(by MFOLD program) was unsuccessful, probably due to toomany possibilities of pairing among AU-rich regions.

Recently several new candidates for a structural RNA havebeen reported. HI9 RNA of mouse and human is induced duringthe embryogenesis, associates with a cytoplasmic particle andhas been shown to have tumor-suppressor activity (5, 21, 22).The XIST gene of human (and Xist of mouse) is expressed fromonly one of the two X chromosomes, localized in theheterochromatic Barrbody and possibly involved in theinactivation of X chromosome (7, 16). These RNAs are assumedto be structural RNAs but the molecular mechanisms by whichthey operate are still unknown. Since their length are relativelylarge (1—20 kb), their action mechanisms may be quite complex.Although the length and the subcellular localization of dutA RNAis similar to those of H19 RNA, they are not similar in theirnucleotide sequence.

Does dutA have a role in the development of D.discoideum?

Previously we found that the expression of dutA is strictlyregulated by cellular interaction during the development (8) anddutA RNA accumulates relatively abundantly ( — 0.01% ofpoly(A) RNA) (unpublished). Moreover, preliminary studiesindicate that the sequences similar to dutA are widespread amongorganisms, from lower eukaryotes to mammals and plants. Thesefacts support the notion that dutA has some important role inD.discoideum and other organisms.

Although neither antisense mutagenesis nor gene disruptioncause phenotypic changes, this does not necessarily imply thatdutA RNA is a useless RNA because there are many reports thatdisruption and antisense mutagenesis of D.discoideum genes, aswell as mammalian genes, cause no phenotypic changes (14, 23,24). This may result from the redundancy of the gene functionand it is even possible that more important genes have moreredundancy. Another possibility is that under the optimalconditions of growth and development provided in the laboratory,phenotypic defects may not be detectable. We are now tryingto isolate and disrupt a cognate gene in S.cerevisiae. Comparisonbetween the D.discoideum and S.cerevisiae RNA could enable

us to identify conserved regions or features which might beimportant for dutA function.

ACKNOWLEDGEMENTS

The authors would like to thank Drs M.Satoh, T.Iwasato,T.Yonesaki, Y.Ohshima and H.Uchiyama for kind and valuablesuggestions, and S.Suzuki and R.Uchida for help with PCRworks. We also thank Drs Y.Tanaka and W.Nellen, T.Fukasawa,M.Yamamoto and M.Okada for generous gifts of D.discoideummitochondrial DNA, plasmid PDneoII, budding yeast DNA,fission yeast DNA and fruit fly DNA and Drs K.Ozato,S.Muramatsu, T.Meshi and K.Satoh for kindly providing us withkilly fish, mouse, A.thaliana and duckweed. The kind gift ofpiperacillin (antibiotics) from Mr T.Satomi of Toyama ChemicalIndustries Inc. is highly appreciated.

REFERENCES

1. Guerrier-Takada, (1983) Cell 35, 849-857.2. Cech.T.R. and Bass.B.L. (1986) Annu. Rev. Biochem. 55, 599-629.3. Brody,E. and AbekonJ. (1985) Science 228, 963-966.4. Cech.T.R. (1990) Annu. Rev. Biochem. 59, 543-568.5. Brannan.C.I., Dees.E.C, Ingram.R.S. and Tilghman.S. (1990) Mo/. Cell.

Bid. 10, 28-36.6. Hao.Y., CrenshawJ., Moulton,T., Newcomb.E. and Tycko.B. (1993)

Nmwre 365, 764-767.7. Brown,C.J., Hendrich.B.D., RupenJ.L., Lafreniere.R.G., Xing.Y.,

Lawrence^, and Willard.H.F. (1992) Cell 71, 527-542.8. Yoshida.H., Yamada.Y. and Okamoto.K. (1991) Differentiation 46,

161-166.9. Sussman,M. (1987) In SpudichJ. A. (ed.), Methods in Cell Biology, Academic

Press, FL, Vol. 28, pp. 9-29 .10. NeUen.W., Silan.C. and Firtel.R.A. (1984) MoL Cell. BioL 4,2890-2898.11. Maniatis.T., Fritsch,E.F. and SambrookJ. (1989) Molecular Cloning: A

Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold SpringHarbor, NY.

12. Myers.T.W. and Gelfand.D.H. (1991) Biochemistry 30, 7661-7666.13. Triglia.T., Peterson.M.G. and Kemp.D.J. (1988) Nucleic Acids Res. 16,

8186.14. Witke.W., Nellen.W. and Noegel,A. (1987) EMBO J. 6, 4143-4148.15. Fukuhara.H. (1982) Biol. Cell 46, 321-324.16. Brockdorff.N., Ashworth.A., Kay.G.F., McCabe.V.M., Norris.D.P.,

Cooper.P.J., Swift,S. and Rastan.S. (1992) Cell 71, 515-526.17. Kiminel.A.R. (ed.) (1988) Molecular Biology of Dictyostelium discoideum

Development. Alan R.Liss, Inc.18. Kaneda.S., Gotoh.O., Seno.T. and Takeishi.K. (1983)/ BioL Chan. 258,

10606-10613.19. Neflen.W., Datta.S., Raymond,C, Sivertsen.A., Mann.S., Crowley,T. and

Firtel.R.A. (1987) In Spudich.J.A. (ed.), Methods in Cell Biology, 28,Academic Press, FL, Vol. 28, pp. 67-100.

20. Simpson.L. and ShawJ. (1989) Cell 57, 355-366.21. Padmis.V., Brannan.C.I. and Tilghman.S.M. (1988) EMBOJ. 1, 673-681.22. Bartolomei.M.S., Zemel.S. and Tilghman.S.M. (1991) Nature 351,

153-155.23. FJkins.T., Zinn.K., McAllister.L., Hoffinann,F.M. and Goodman.C.S.

(1990) CW/60, 565-575.24. Saga,Y., Yagi,T., Ikawa.Y., Sakakura,T. and Aizawa.S. (1992) Genes Dev.

6, 1821-1831.

Downloaded from https://academic.oup.com/nar/article-abstract/22/1/41/2358949by gueston 08 March 2018