Molecular Phylogenetic Analysis of Tryptophanyl-tRNA ...znaturforsch.com/s63c/s63c0418.pdf · Molecular Phylogenetic Analysis of Tryptophanyl-tRNA Synthetase of Actinobacillus actinomycetemcomitans

Molecular Phylogenetic Analysis of Tryptophanyl-tRNA Synthetase ofActinobacillus actinomycetemcomitansNarayanan Rajendrana,*, Rajendram V. Rajnarayananb, and Donald R. Demuthc

a Biology, MASC, Kentucky State University, Frankfort, KY-40601, USA.Fax: (5 02) 5 97Ð68 26. E-mail: [email protected]

b Chemistry, NSD, Tougaloo College, Tougaloo, MS-39174, USAc Center for Oral Health and Systemic Disease Ð School of Dentistry,

University of Louisville, Louisville, KY-40292, USA

* Author for correspondence and reprint requests

Z. Naturforsch. 63c, 418Ð428 (2008); received November 8, 2007/January 7, 2008

Aminoacyl-tRNA synthetase family enzymes are of particular interest for creating univer-sal phylogenetic trees and understanding the gene flow as these enzymes perform the basicand analogous biochemical function of protein synthesis in all extant organisms. Amongthem, tryptophanyl-tRNA synthetase (Trp-RS) plays a foremost role in phylogeny owing tothe close relationship with tyrosine-tRNA synthetase. In this study, the sequence of the geneTrp-RS was amplified using degenerated adenylation domain primers in the periodontal bac-terium Actinobacillus actinomycetemcomitans. The sequence of the cloned PCR ampliconconfirmed the adenylation domain sequence with glutamic acid residue, which is absent infive other oral bacteria used in this study as well as in a number of other bacteria describedin the database. The Trp-RS sequence analysis prevailed the identify elements such as Ross-mann-fold sequence and tRNATrp binding domains including acceptor stem and anticodon.A theoretical model of Trp-RS of A. actinomycetemcomitans was generated. Guided dockingof the ligand tryptophanyl-5�-AMP revealed a highly identical active site in comparison withthe bacterial template. The phylogenetic positioning of Trp-RS among a group of oral bac-terial species revealed that A. actinomycetemcomitans is closely related to Haemophilus influ-enzae, H. ducreyi and Pasteurella multocida.

Key words: Aminoacyl-tRNA Synthetase (AARS), Tryptophanyl-tRNA Synthetase (Trp-RS),Actinobacillus actinomycetemcomitans, Phylogenetic Tree

Introduction

The aminoacyl-tRNA synthetase (AARS) fam-ily of enzymes is of particular interest for under-standing vertical gene flow (parent to offspring)and horizontal gene transfer (between differentspecies) for creating universal phylogenetic trees,as they perform the same fundamental biochemi-cal function of ribosomal protein synthesis in allorganisms (Donoghue and Luthey-Schulten, 2003).Previous studies revealed that the identities ofsuch AARSs are pharmaceutically important asthey are potentially valuable targets of antimicro-bials (Kitabatake et al., 2002) and could lead tothe development of novel antibiotics that show nocross-resistance to other classical antibiotics (Kimand Choi, 2003).

The AARSs are essential to understand the bac-terial evolution of translation and the transitionfrom RNA to proteins (Donoghue and Luthey-Schulten, 2003) especially in oral community dueto the existence of its diversity. Out of 500 species

0939Ð5075/2008/0500Ð0418 $ 06.00 ” 2008 Verlag der Zeitschrift für Naturforschung, Tübingen · http://www.znaturforsch.com · D

of microbial inhabitants that have been recordedin the human oral vicinity, about half of them haveyet to be cultured (Harper-Owen et al., 1999) andto be identified for their peptide products as wellas for biochemical and phylogenetical significance.While there have been many studies on the 20 var-iable AARSs responsible for translation of the ge-netic message, only a few studies have targetedeach specific group of AARSs in oral bacteria forphylogenetic studies (Farahi et al., 2004; Dono-ghue and Luthey-Schulten, 2003; Woese et al.,2000). Tryptophanyl-tRNA synthetase (Trp-RS)was not investigated individually for phylogeneticanalysis among such groups of bacteria, especiallyin the periodontal bacterium Actinobacillus acti-nomycetemcomitans.

A. actinomycetemcomitans is a Gram-negative,non-motile, fermentative, oral coccobacillus yet tobe recognized in Trp-RS phylogenetic studies.Over 200 genome projects have been completed(Duncan, 2003) along with A. actinomycetemcomi-

N. Rajendran et al. · Molecular Phylogenetic Analysis of Tryptophanyl-tRNA Synthetase 419

Table I. Peptide synthetase recognition primers used in this study.

Set Primer Degenerated primer sequence 5�Ð3� (F = forward and R = reverse) Reference

1 A3 F 5� TAC ACS AGC GGS AGC ACS GG 3� MarahielA7 R 5� AVG TCS CCS GTS CKG TAC ATS C 3� et al., 1997

2 A8 F 5� CAG GTS AAG RTS MGS GGS TWC MG 3� MarahielE2 R 5� GTC SAC SRM SAR GTG GTG 3� et al., 1997

3 PS- F 5� CCA ATC GGC AAA CCA ATC TCC 3� Carniospecific R 5� GGT TTT AGT GCT TCT CCA CTA GC 3� et al., 2001

4 A2a F 5� GGA ATT CCT SAA GDC SGG CGG IGC CTA CGT SCC 3� HealyA3a R 5� GGA ATT CCC TTS GGC IKS CCG GTS GIS CCG GAG G 3� et al., 2000

5 A7 F 5� TAC CGI ACI GGI GAT CTI GTI CG 3� Turgay andT R 5� ATI GAG TCI CCI CCI GGG IAA AAG AA 3� Marahiel,

19946 A2 F 5� GGA ATT CCT CAA GGC GGG CGG IGC CTA CGT GCC CIA CCC 3� Rajendran,

A3 R 5� GGA ATT CCC TTG GGC IGG CCG GIC GIC CCG CAG GTG TAG A 3� 19997 PSF F 5� GGWCDACHGGHMANCCHAARGG 3� Schulz

PSF-2 R 5� GGCAKCCATYTYGCCARGTCNCCKGT 3� et al., 2005

a High-GC biased degenerated oligonucleotides with terminal EcoRI adaptors.

tans and other oral bacteria including Prevotellaintermedia, Fusobacterium nucleatum, Porphyro-monas gingivalis, Streptococcus mutans and Tre-ponema denticola (www.oralgen.lanl.gov). TheirTrp-RS identify elements have not yet been ana-lyzed phylogenetically. Analysis of AARSs includ-ing completely sequenced bacterial, archaeal, andeukaryotic genomes shows a complex evolution in-volving a variety of horizontal gene transfers(Wolf et al., 1999). Among them Trp-RS, the classI AARS, catalyzes tryptophan activation in theabsence of its cognate tRNA (Guo et al., 2007).Trp-RS also plays a foremost role in phylogenyowing to the close relationship with tyrosine-tRNA synthetase (Tyr-RS). In this study, an at-tempt was made to reveal the conserved motifs ofTrp-RS in A. actinomycetemcomitans and analyzethe Trp-RS phylogeny of A. actinomycetemcomi-tans with its related species.

Materials and Methods

Primer selection, PCR conditions, subcloning andDNA sequencing

Degenerated primers, derived from the sequen-ces of non-ribosomal peptide synthetases (Ta-ble I), were synthesized at Integrated DNA Tech-nologies, Coralville, IA, USA. All primers wereemployed against the genomic DNA of the perio-dontal bacteria A. actinomycetemcomitans 652,Porphyromonas gingivalis, and Fusobacterium nu-cleatum as well as the predominantly oral bacteriaStreptococcus mutans (KPSK2) and Streptococcusgordonii (M5). The bacterial samples used in the

present study were selected because of their majorrole in oral microbiota and availability from thelaboratory stocks of Prof. Demuth. The work wascarried out at the University of Louisville, KY,USA. The published PCR protocol (Rajendran,1999) was modified using Invitrogen PCR-Super-mix under the following conditions: 95 ∞C (5 min),95 ∞C (1 min), 55 ∞C (2 min), and 72 ∞C (3 min) for30 cycles in a Perkin-Elmer PCR machine. PCRamplicons were analyzed via 1% agarose gel elec-trophoresis using 100 bp DNA ladder as standardmarker (catalog # N3231S) of New England Bio-Labs, Ipswich, MA, USA. The fragments weresubcloned in pGEM-T Easy plasmid vector andtransformed into DH5α competent (E. coli) cellsaccording to the manufacturer’s protocols of Pro-mega, Madison, WI, USA. Based on IPTG/X-galselection, the clones were cultured in LB mediumwith ampicillin. Plasmids were prepared using Pro-mega mini-plasmid kit with a vacuum manifold.The plasmids of the confirmed clones were ana-lyzed via 1% agarose gel electrophoresis using 100bp DNA ladder and were sequenced using a T7primer at the Center for Genetics and MolecularMedicine (CGeMM) DNA core facility of theUniversity of Louisville, KY, USA.

Sequence analysis and construction ofphylogenetic tree

After confirmation of the DNA sequence of the1400 bp fragment of the A. actinomycetemcomi-tans clone (Pri2-Aa-1-T7), derived using primerset 2 (Table I), further computational analysis was

420 N. Rajendran et al. · Molecular Phylogenetic Analysis of Tryptophanyl-tRNA Synthetase

conducted. A homologous nucleotide sequencesearch was performed at NCBI’s non-redundantdatabase by using the BLASTX and proteinBLASTP algorithms (http://www.ncbi.nlm.nih.gov/BLAST/). The newly determined Trp-RS nucleo-tide sequence was deposited in GenBank and anaccession number was obtained using the BankIT:GenBank (www.ncbi.nlm.nih.gov/BankIT/) sub-mission program. To confirm the signature se-quence of Trp-RS in A. actinomycetemcomitans,the eMOTIFS search server at Stanford University(http://dna.stanford.edu/cgi-bin/emotif/nph-emotif-search) was used. The resulting sequences werealigned by using the CLUSTALW program (http://pir.georgetown.edu/pirwww/search/multaln.html).An initial sequence alignment was created with alist of 20 bacterial species that had homologousamino acid sequences. An alignment of Trp-RS se-quences from 10 significant bacterial species onthe basis of highest possible homology was thenprepared. Identification of the published sequen-ces of tRNATrp binding domains in Trp-RS (Jiaet al., 2002), as well as the basic Rossmann-foldsequence (Brown et al., 1997), was searched in thealignment. The ClustalX program was then usedto generate the phylogenetic tree (Thompsonet al., 1997). Theoretical homology models of Trp-RS were generated using Modeller7v7 (Marti-Renom et al., 2000) and structural analyses wereperformed using a combination of SYBYL 7.0(Tripos, St. Louis, MO, USA) and Insight II (Ac-celrys, Burlington, MA, USA). All the graphicalimages were generated using UCSF Chimera mo-lecular visualization software (Pettersen et al.,2004).

Results

Trp-RS of A. actinomycetemcomitans

Degenerative nucleotide primers encoding con-served acylation and epimerization domains ofnon-ribosomal peptide synthetase were used inPCR to amplify corresponding regions on genomicDNA from A. actinomycetemcomitans, Porphy-romonas gingivalis, Fusobacterium nucleatum,Streptococcus mutans (KPSK2) and Streptococcusgordonii. Among the studied organisms, the ge-nomic DNA of the periodontal bacterium A. acti-nomycetemcomitans yielded a 1.4 kb in the sizePCR fragment with primer set 2 (Fig. 1). The se-quence of the fragment has been submitted toGenBank and obtained the accession number

Fig. 1. Electrophoresis of PCR amplicons. Seven sets ofdegenerated primers (Table I) were employed againstthe genomic DNA of the periodontal bacterium A. acti-nomycetemcomitans 652 under the following conditions:95 ∞C (5 min), 95 ∞C (1 min), 55 ∞C (2 min), and 72 ∞C(3 min) for 30 cycles in a Perkin-Elmer PCR machine.PCR amplicons were analyzed via 1% agarose gel elec-trophoresis. This figure indicates the amplified PCRfragment of interest at the size of 1.4 kb using primerset 2 (lane 1) and unconcerned fragment using primerset 3 (lane 2) and a 100 base pair DNA ladder (lane 7).

DQ143942. The BLAST searches reveal its iden-tity as Trp-RS, and the signature sequence of Trp-RS was confirmed in A. actinomycetemcomitansby using the eMOTIFS search server, whichyielded all four tryptophanyl-tRNA synthetasesignature motifs: I (16-LTIGNYLGALRQWVKMQ-32), II (67-YLACGIDPAKSTIFIQSHV-85),III (144-VPVGEDQKQHLEITRDIAQR-163) andIV (195-KMSKSDEN-202) as depicted in Fig. 2.

Seeking the adenylation domain

When the same degenerative primer set 2 wasemployed with other oral bacteria, none of themyielded a positive amplicon under the same PCRconditions and hence yielded no fragment for Trp-RS sequence. In order to seek further informationand to confirm the observation, the nucleotidesequence obtained from A. actinomycetemcomi-tans was compared as query with other completedgenome sequences of oral pathogens by blastingat www.oralgen.lanl.gov. The resulting Trp-RS se-quences from some of the oral bacterial sequencesobtained from the genome database showed noglutamic acid (Glu, E) in the GELTIGNYLG se-quence at the residue 14Ð23 (Fig. 3B), which is aputative consensus adenylation domain sequenceof one of the NRPS core motifs, GEL-xIxGxG(VL)ARGYL (Marahiel et al., 1997).


Fig. 2. Amino acid sequence alignment of Trp-RS. The results of the BLASTX sequences were aligned by using theCLUSTALW program. These sequence alignments were compared on the basis of highest percentage (60 or more)of homologous amino acid sequences from various bacterial species. Full names of the organisms are given in TableII. A. actinomycetemcomitans and a group of closely related bacterial species unusually have Gly-Glu-Leu (GEL)residues in their putative consensus adenylation domain sequence (GELTIGNYLG) in Trp-RS.


Table II. Homology sequence search results.

Highly matched bacterial sequences Sequene GenBankwith the query Trp-RS of A. actinomycetemcomitans identity accessionidentified from the NCBI-BLAST algorithms search (%) number

Actinobacillus actinomycetemcomitans (query) 100 DQ143942a

Pasteurella multocida subsp. multocida str. PM70 88 AE006199Haemophilus influenzae Rd KW20 81 U32746Klebsiella aerogenes 81 AF308467Salmonella typhimurium LT2 79 AE008860Escherichia coli W3110 79 U38647Photorhabdus luminescens subsp. laumondii TTO1 78 BX571859Haemophilus ducreyi 35000HP 78 AE017151Yersinia pestis biovar Medievalis str. 91001 76 AE017127Vibrio cholerae O1 biovar eltor str. N16961 64 AE004329Helicobacter pylori 26695 52 AE000630Helicobacter hepaticus ATCC 51449 49 AE017145Synechococcus elongatus PCC 6301 48 AP008231Caulobacter crescentus CB15 46 AE005680Prochlorococcus marinus str. MIT 9313 45 BX572096Mesorhizobium loti MAFF303099 38 BA000012Mycoplasma genitalium G-37 36 U39693Xylella fastidiosa 9a5c 34 AE003894Xylella fastidiosa Temecula1 34 AE012559Pseudomonas syringae pv. tomato str. DC3000 33 AE016853

a Newly submitted GenBank accession number from this study.

Fig. 3. Homologous special sequence comparisons of aligned amino acids in bacteria obtained from Oralgen andNCBI-BLAST. (A) The sequence alignment of this study matching with the published sequence of tRNATrp bindingdomains in tryptophanyl-tRNA synthetase at residue 109Ð124 and 239Ð242, as well as the basic Rossmann-foldsequence at residue 195Ð199 in A. actinomycetemcomitans and other related bacterial species. (B) The residueglutamic acid (Glu or E), present in Trp-RS of A. actinomycetemcomitans and a group of closely related bacterialspecies, is apparently not present in these sequences of phylogenetically distant organisms. Instead of glutamic acidanother residue, predominantly lysine (Lys or K), is found in those sequences.


Seeking other homologous residues

Out of many homologous prokaryotic and eu-karyotic Trp-RS sequences obtained from BLASTsearch at NCBI, a group of 20 bacterial specieswas selected based on the closest identity startingwith a 33% homology cut-off value (Table II).Most AARS from these species, including A. acti-nomycetemcomitans, are expressed from singlecopy genes where tryptophan (W) is conservedin the standard bacterial Trp-RS type-I (Buddhaand Crane, 2005) at residue of 93 (Fig. 2). Otherwell conserved residues include: 40(C), 128(Y),132(M), 135/136(D, I), 144(V), 150(Q), and153(H). All of them are key residues in bindingtryptophan (Buddha and Crane, 2005). However,two copies of same synthetases, like Trp-RS type-I and auxiliary Trp-RS II, are found in some bacte-ria such as Streptomyces coelicolor and Deinococ-cus radiodurans (Buddha and Crane, 2005). Ourhomology search for auxiliary Trp-RS II found nomatches in A. actinomycetemcomitans as well as inother test bacteria. Similarly a special Trp-RS ofStreptococcus pyrogenes, the only Trp-RS that isknown to be more homologous to Trp-RS II (Bud-dha and Crane, 2005), was not found in A. actino-mycetemcomitans or in other test bacteria in thepresent study. However, an interesting observationwas made with uvrA gene expression in A. actino-mycetemcomitans. As reported earlier, when thestandard Trp-RS was inhibited, either by DNAdamage or by inhibitors such as indolmycin, an in-hibitor-resistant auxiliary Trp-RS II was induced,as in the case of Deinococcus radiodurans (Bud-dha and Crane, 2005). It was also observed thatuvrA gene expression occurs when DNA damagehappens due to ultraviolet light or other factors(Buddha and Crane, 2005). Although it was notdemonstrated in A. actinomycetemcomitans be-fore, a 250 bp fragment was amplified in thepresent study using primer set 2 in A. actinomyce-temcomitans that corresponded to the uvrA geneas revealed by BLAST search at NCBI (data notshown).

Rossmann-fold and anticodon binding domainsearch

In the current study, the Rossmann-fold (RF)sequence, at residues 195Ð199 of A. actinomyce-temcomitans and the other bacterial speciesaligned with query (Fig. 3), matches other pub-lished data on Trp-RS (Brown et al., 1997). Besides

this KMSKS (RF domain), which is a consensusmotif (Brown et al., 1997), a T(H)IGN domainthat participates in ATP binding and typifiesClass-Ic synthetases (Buddha and Crane, 2005)was identified in A. actinomycetemcomitans. Thesequence alignment of the current study (Fig. 2)matched with the published sequence of tRNATrp

binding domains in Trp-RS at residues 109Ð124and 239Ð242 (Fig. 3), indicating the presence ofconserved regions of the acceptor stem and antico-don of variant tRNATrp.

Phylogenetic positioning

Phylogenetic analysis of AARSs including com-pletely sequenced bacterial, archaeal, and eukary-otic genomes shows a complex evolution involvinga variety of horizontal gene transfers (Wolf et al.,1999). In a recent sequence alignment study ofthree Trp-RS sequences from Bacillus stearother-mophilus, Deinococcus radiodurans, and Homo sa-piens, for which crystal structures are currentlyknown, the highest sequence conservation be-tween species was noticed within the signatureKMSKS loop of Class-I synthetases (Buddha andCrane, 2005). The mechanism by which thisKMSKS domain stages the adenylation reaction isalso conserved (Buddha and Crane, 2005) in thesespecies. In the present study 10 bacterial specieswere selected from the aligned 20 bacterial species(Table II) on the basis of highest possible homol-ogy (over 60% of Trp-RS sequence homology)with A. actinomycetemcomitans. The selection ofhighest possible homology minimizes putative er-rors in evaluating the reliability of a phylogenetictree and ensures the closest possible phylogeneti-cally linked species with A. actinomycetemcomi-tans. Interestingly, the signature KMSKS loop ofClass-I synthetases is conserved in all 10 species(Fig. 3A) giving further evidence to substantiatethe proposed phylogenetic tree (Fig. 4). The mech-anism by which this KMSKS domain stages theadenylation reaction is thus likely to be conservedin A. actinomycetemcomitans and in other phylo-genetically related bacterial species as shown inthe phylogenetic tree (Fig. 4).

Discussion

Searching for peptide synthetases

Peptide synthetases in oral bacteria have bio-chemical and phylogenetic significance since thesegenes have not yet been fully investigated in most


species. In general every oral bacterial cell harbors20 AARSs responsible for the synthesis of the setof 20 canonical aminoacyl-tRNA families. Thereare two ways of forming aminoacyl-tRNA (Woeseet al., 2000). First is the direct acylation of tRNAby discriminating AARSs, which is an ATP-de-pendent reaction, and the second is indirect acyla-tion of tRNA by non-discriminating AARSs,which is a tRNA-dependent amino acid modifica-tion (Guo et al., 2007). Woese et al. (2000) statedthat, “the second indirect pathway depends on theacylation of tRNA with a precursor amino acid.This precursor amino acid is then converted whilebound to tRNA to the correct amino acid by asecond, non-synthetase enzyme. The currentknowledge about these enzymes is still far fromcomplete”. Despite their significance, its processremains unknown like that of amidotransferaseenzyme that converts Glu-tRNA to Gln-tRNA(Tumbula et al., 2000). In an other case, buildingof pyrrolysine on tRNA is not required and mightnot occur as reported earlier (Polycarpo et al.,2004; Blight et al., 2004). However, an unknownenzyme responsible for the conversion of thecharged amino acid has been reported earlier(Srinivasan and Krzycki, 2002), which indicatesthat there is an additional enzyme involved in thisprocess. Many microbial cells also contain multi-enzyme complexes that make specific protein tem-plates for a nucleic acid-independent biosynthesisof low-molecular weight peptides (Marahiel et al.,1997). These non-ribosomally produced peptidesinclude lipopeptides, depsipeptides, and peptido-lactones (Schulz et al., 2005; Rajendran et al., 1999)and are assembled from an exceedingly diversegroup of precursors including pseudo, non-proteo-genic hydroxy, N-methylated, and d-amino acids(Marahiel et al., 1997). The use of multiple sets ofdegenerated primers (Turgay and Marahiel, 1994;Marahiel et al.,1997; Rajendran, 1999; Healy et al.,2000; Carnio et al., 2001), derived from the con-served domains of non-ribosomal peptide synthe-tase (NRPS), reveals promising peptides and theirproducts in many soil bacteria. But no such exten-sive attempts were made in oral bacteria. Usingsuch degenerated primers to attempt to revealfunctional genes from oral bacteria is thereforemore valuable in terms of the presence of diversi-fied bacterial synthetases, their synthesis and theirphylogenetic importance. Keeping this in mind, weapplied these multiple sets of degenerated synthe-tase primers and found that one of the primers is

capable to amplify the Trp-RS as confirmed byPCR and molecular analysis.

Glutamic acid residue in Trp-RS ofA. actinomycetemcomitans

The present study indicates that A. actinomyce-temcomitans and a group of closely related bacte-rial species unusually have GlyÐGluÐLeu (GEL)residues in their Trp-RS sequence (Fig. 2). The res-idue glutamic acid (Glu or E) is not present in thesequences of phylogenetically distant organisms(Fig. 3B), which is apparently perceptible in thephylogenetic tree (Fig. 4). Instead of glutamic acid,another residue, lysine (Lys or K), was predomi-nantly found in these sequences (Fig. 3B). The cat-alytic role of Glu or Lys has not yet been analyzedelsewhere. Similarly the role of the adenylationdomain residue Glu, unusually it exists in A. acti-nomycetemcomitans and its closely related speciesbut not in other distantly related oral species, hasnot yet been studied.

Fig. 4. The phylogenetic tree of tryptophanyl-tRNA syn-thetase. Aligned sequences of A. actinomycetemcomitanswere compared with related species. The phylogenetictree reveals that Trp-RS of A. actinomycetemcomitans isclosely related to Haemophilus influenzae, Haemophilusducreyi and Pasteurella multocida. Bootstrap values ofmore than 800 are indicated.

Conserved residue search and analysis

During translation of the genetic code an error-free molecular recognition of 21 amino acids andtheir attachments to tRNAs that bear the appro-priate anticodon triplets is critical (Woese et al.,2000). There are two classes of AARSs, based onthe structure and reactivity, involved in the cata-lytic reaction. Class-I AARS are comprised of acatalytic Rossmann-fold domain with KMSKS


residues and a helical tRNA anticodon binding do-main (Buddha and Crane, 2005). The RF domainmakes ATP interaction possible in bacterial spe-cies. The existence of the signature motifs, whichare involved in ATP binding, such as the consen-sus motif KMSKS and a T(H)IGN domain, revealsthat the Trp-RS sequences of A. actinomycetem-comitans and other bacterial sequence in thepresent study belong to Class-Ic synthetases. Aprevious study showed that Trp-RS from variousspecies shared a high homology of residues, espe-cially within two short regions, “QFKDKS_RY-AENVNVG” and “NKAG”. These two short re-gions were predicted to bind to acceptor stem andanticodon of variant tRNATrp (Jia et al., 2002). Thepresence of such a domain in A. actinomycetem-comitans indicates the conserved nature of theRossmann-fold.

Compared to ribosomal protein synthesis, non-ribosomal protein synthesis has less specificity.The substrate activation for the ribosomal originis tRNA synthetase, whereas adenylation is thesubstrate activation domain for the non-ribosomalorigin. This is likely due to the evolutionary forceto maintain a given substrate selection, and “it isnot very strict” as predicted earlier (Doekel andMarahiel, 2001). The adenylation domain is thespecific “main gatekeeper” of NRPS that catalyzesthe activation of cognate carboxy acids as adenyl-ates by hydrolysis of ATP (Doekel and Marahiel,2001). In Fig. 3, residues 14Ð23 (GELTIGNYLG)indicate the presence of a component of the puta-tive consensus adenylation domain sequencewhich is of one of the non-ribosomal synthetasecore motifs GELxIxGxG(VL)ARGYL (Marahielet al.,1997). The primer set 2 which used in thepresent study (Table I) represents one of the ad-enylation domains of the NRPS. This adenylationdomain also happens to contain the residues ofone of the signature motifs that participates inATP binding; T(H)IGN (Buddha and Crane,2005). This was observed in A. actinomycetemcom-itans by applying that primer for the first time. Ad-enylation domains in NRPS, that are about 550amino acids in size, indicate that it is a member ofthe superfamily of adenylate-forming enzymes likeacyl-CoA-ligase. As reported earlier (Doekel andMarahiel, 2001), this domain in NRPS is a part ofa larger protein frame, but could exhibit a compa-rable catalytic activity also when expressed heter-ologously as a separate unit (Marahiel et al.,1997;Doekel and Marahiel, 2001). Searching for other

conserved motifs in the current amino acid se-quence of A. actinomycetemcomitans by using theNRPS-PKS BLAST server (http://www.nii.res.in/nrps-pks.html) yielded no matching sequence asexpected, since the query has only 340 residues.However, it could be possible that the adenylationsequence observed in A. actinomycetemcomitans isthe Trp-RS signature sequence and has no relationto NRPS. As explained elsewhere (Doekel andMarahiel, 2001) nature could divulge various alter-natives to produce variations through tRNA,NRPS, polyketide synthases (PKS), mixed peptidesynthetases (NRPS-PKS), and/or engineered pep-tide synthetases. Therefore, it is likely that the se-quence of the adenylation domain used in thepresent study, which reveals the Trp-RS identity,could have some phylogenetic relation with earlyNRPS origin, and may reveal the Trp-RS identityin other closely related species. This is under in-vestigation.

Phylogenetic analysis

Trp-RS is the smallest protein in the family ofbacterial tRNA synthetases (Jia et al., 2002) andless complex to compare. Phylogenetic analysis ofTrp-RS from various oral bacterial sequences isnot available elsewhere. It was the main rationaleto select Trp-RS for phylogenetic analysis in thepresent study. For the first time, the amino acidsequence was analyzed with the goal of phyloge-netically positioning the Trp-RS of A. actinomyce-temcomitans among other closely related bacterialspecies by using phylogenetic tree reconstructionmethods. The compatible data set, without biasfrom a stochastic effect, indicates that Trp-RS ishighly conservative in A. actinomycetemcomitansand is closely related to Haemophilus influenzae,Haemophilus ducreyi and Pasteurella multocida(Fig. 4). In a previous genome sequence study, itwas predicted that the proteins with highest ho-mologies to other organismshave orthologs in H.influenzae, approx. 43%, and in P. multocida, 41%(Duncan, 2003). In the current study, the codonusage indicates that A. actinomycetemcomitans isclosely related to H. influenzae, H. ducreyi and P.multocida (Table II). The previous study on 16Sribosomal RNA analysis between A. actinomyce-temcomitans and H. influenzae also indicated theirclose relationship (Duncan, 2003). The BLASTXsearch of the present study on Trp-RS reveals thatA. actinomycetemcomitans has 88% of gene se-quence homology with P. multocida subsp. multo-


cida str. PM70, more than with H. influenzae andH. ducreyi, and they share close phylogeneticcharacteristics as depicted in the phylogenetictree (Fig. 4).

In another approach, synthetases of other kindsclosely related to Trp-RS were searched amongthe sequences in order to find the close relation-ship between the closest AARSs. As reported ear-lier, a comparative study indicated that the phy-logeny of tRNA genes matched with that of Tyr-RS and Trp-RS and hence Tyr-RS and Trp-RS mayhave close relationship (Ribas et al., 1996). How-ever, it was later overlooked and put forward thatTyr-RS and Trp-RS have an early divergence inthe phylogeny (Brown et al., 1997). Our currentstudy observes no congruent amino acid alignmentbetween Tyr-RS and Trp-RS sequences. This find-ing hence supports the latter observation that apossible divergence could exist between these twosynthetases, and Trp-RS could putatively maintaina highly conserved nature thereafter in A. actino-mycetemcomitans and its closely related speciessuch as Haemophilus influenzae, Haemophilus du-creyi and Pasteurella multocida.

Besides the high sequence similarities amongthese groups of bacteria and the divergence fromthe nearest synthetases, the presence of glutamicacid (Glu or E) in Trp-RS of A. actinomycetem-comitans and in a group of closely related bacterialspecies (Fig. 2) reveals its phylogenetic value, sinceit is apparently not present in the sequences ofdistant organisms. The unrooted phylogenetic tree,constructed in the present study (Fig. 4), clearlydepicts this observation for the first time, and itcould be the putative reason for those closely re-lated species (Fig. 2) to have phylogenetic rela-tionships rather than distant species (Fig. 3B). In-stead of glutamic acid another residue, lysine (Lysor K), is predominantly found in the sequences ofthose distant species. This study was helpful interms of identifying Trp-RS in an oral bacteriumand guided to the further investigation on the unu-sual existence of an adenylation domain compo-nent with glutamic acid in A. actinomycetemcomi-tans and its closely related species. Furtherinvestigation in this direction will reveal more in-sight into their relationships and help to constructa universal Trp-RS phylogenetic tree.

Structural analysis of Trp-RS

In order to analyze the structural homology andconservation of the Trp-binding site, we con-

Fig. 5. Structure of Trp-RS from A. actinomycetemcomi-tans. A theoretical model of Trp-RS was generated usingModeller7v7 with a high-resolution bacterial Trp-RStemplate (PDB ID: 1I6K, 57% sequence identity). Thestructure of Trp-RS of A. actinomycetemcomitans isshown as a ribbon model. The ball-and-stick model withcarbon framework at the center represents the boundligand tryptophanyl-5�-AMP.

structed theoretical homology models of all Trp-RS sequences of Fig. 4 using Modeller7v7. A high-resolution bacterial Trp-RS X-ray crystal structure(PDB ID: 1I6K, B. stearothermophilus) with 57%sequence identity to Trp-RS A. actinomycetem-comitans was used as the structural template. Thefinal models were energy-minimized using the Dis-cover module of Insight II prior to further struc-tural analyses. Superposition of the Trp-RS models@ Cα with the template had a RMS deviation inthe range 0.15Ð0.35 A showing the high homologyof this structural domain. Guided docking(SYBYL 7.0) of the ligand tryptophanyl-5�-AMPinto the binding site of Trp-RS resulted in an iden-tical binding pose in comparison with the template(Fig. 5). A closer look at the binding site ofTrp-RS of A. actinomycetemcomitans and that ofits template revealed a highly identical active site(a ~90% identity) in comparison with the bacterial


Fig. 6. A closer look at the binding site of Trp-RS of A.actinomycetemcomitans. Guided docking (SYBYL 7.0)of the ligand tryptophanyl-5�-AMP revealed a highlyidentical active site in comparison with the bacterialtemplate. The Trp-RS A. actinomycetemcomitans isshown at the top, and the template-1I6K is shown as thebottom figure. The binding site residues are rendered asa wire frame model and the ligand as a ball and stickmodel.

Blight S. K., Larue R. C., Mahapatra A., Longstaff D.,Chang E., Zhao G., Kang P. T., Green-Church K. B.,Chan M. K., and Krzycki J. A. (2004), Direct chargingof tRNA(CUA) with pyrrolysine in vitro and in vivo.Nature 431, 333Ð335.

Brown J. R., Weiss R., and Doolittle W. F. (1997), Evi-dence for the early divergence of trptophanyl-tRNAand tyrosyl-tRNA synthetases. J. Mol. Evol. 45, 9Ð16.

Buddha R. M. and Crane B. R. (2005), Structures oftryptophanyl-tRNA synthetase II from Deinococcusradiodurans bound to ATP and tryptophan: Insightinto subunit cooperativity and domain motions linkedto catalysis. J. Biol. Chem. 280, 31965Ð31973.

Trp-RS template of A. actinomycetemcomitansand a quantitative conservation of the Trp bindingsite (Fig. 6).

Conclusion

In conclusion, the prevalence of tryptophanyl-tRNA synthetase conserved motifs of Actinobacil-lus actinomycetemcomitans and its molecular phy-logenetic analysis have been carried out. Compari-son of A. actinomycetemcomitans with other oralbacterial genomes revealed that Trp-RS has glu-tamic acid residues in its adenylation domain thatare not present in other related oral bacterial spe-cies. The constructed Trp-RS phylogenetic tree re-vealed the relationship with a group of oral spe-cies. Our oral microbiota have diversified bacterialinhabitants and this molecular phylogenetic studyon oral bacteria could support future studies toconstruct a comprehensive phylogenetic tree onTrp-RS among the available oral bacterial species.Since the aminoacyl-tRNA synthetase family ofenzymes is universal and can be found in all livingorganisms, their universality will allow a futureconstruction of a detailed phylogenetic tree ofAARSs in all extant organisms.

Acknowledgements

This work was supported by Grant # P20 RR-16481 from IDeA Networks of Biomedical Re-search Excellence (INBRE) program of the Na-tional Center for Research Resources, USA andpartially by Grant #MCB-0639356 from NationalScience Foundation (NSF), USA to NarayananRajendran. We thank Deanna James for labora-tory assistance and Bruce Griffis for correctingthe manuscript.

Carnio M. C., Stachelhaus T., Kevin P. F., and SchererS. (2001), Pyridinyl polythiazole class peptide antibi-otic micrococcin P1, secreted by foodborne Staphylo-coccus equorum WS2733, is biosynthesized nonriboso-mally. Eur. J. Biochem. 268, 6390Ð6400.

Doekel S. and Marahiel M. A. (2001), Biosynthesis ofnatural products on modular peptide synthetases.Metabolic Eng. 3, 64Ð77.

Donoghue P. O. and Luthey-Schulten Z. (2003), On theevolution of structure in amino-acyl-tRNA syntheta-ses. Microbiol. Mol. Biol. Rev. 67, 550Ð573.

Duncan M. J. (2003), Genomics of oral bacteria. Crit.Rev. Oral Biol. Med. 14, 175Ð187.


Farahi K., Pusch G. D., Overbeek R., and WhitmanW. B. (2004), Detection of lateral gene transfer eventsin the prokaryotic tRNA synthetases by ratios of evo-lutionary distances method. J. Mol. Evol. 58, 615Ð631.

Guo L. T., Chen X. L., Zhao B. T., Shi Y., Li W., XueH., and Jin Y. X. (2007), Human tryptophanyl-tRNAsynthetase is switched to a tRNA-dependent mode fortryptophan activation by mutations at V85 and I311.Nucleic Acids Res. 35, 5934Ð5943.

Harper-Owen R., Dymock D., Booth V., WeightmanA. J., and Wade W. G. (1999), Detection of uncultu-rable bacteria in periodontal health and disease byPCR. J. Clinical Microbiol. 37, 1469Ð1473.

Healy F. G., Gibson D. M., and Loria R. (2000), The txtAB genes of the plant pathogen Streptomyces acidis-cabies encode a peptide required for phytotoxin thax-tomin A production and pathogenicity. Mol. Micro-biol. 38, 794Ð804.

Jia J., Feng X., Jin Y., and Wang D. (2002), Two essentialregions for tRNA recognition in Bacillus subtilis Trp-tRNA synthetase. Biochemistry 365, 749Ð756.

Kim S. and Choi S. Y (2003), Aminoacyl-tRNA synthe-tases and inhibitors as a novel family of antibiotics.Appl. Microbiol. Biotechnol. 61, 278Ð288.

Kitabatake M., Demain A., Sakamota K., Yokoyama S.,and Soll E. (2002), Indolmycin resistance of Strepto-myces coelicolar by induced expression of one of itstwo tryptophanyl-tRNA synthetases. J. Biol. Chem.277, 23882Ð23887.

Marahiel M. A., Stachelhaus T., and Mootz H. D. (1997),Modular peptide synthetases involved in nonriboso-mal peptide synthesis. Chem. Rev. 97, 2651Ð2673.

Marti-Renom M. A., Stuart A., Fiser A., Sanchez R.,Melo F., and Sali A. (2000), Comparative proteinstructure modeling of genes and genomes. Annu. Rev.Biophys. Biomol. Struct. 29, 291Ð325.

Pettersen E. F., Goddard T. D., Huang C. C., CouchG. S., Greenblatt D. M., Meng E. C., and Ferrin T. E.(2004), UCSF chimera Ð A visualization system forexploratory research and analysis. J. Comput. Chem.25, 1605Ð1612.

Polycarpo C., Ambrogelly A., Berube A., WinbushS. M., McCloskey J. A., Crain P. F., Wood J. L., andSöll D. (2004), An aminoacyl-tRNA synthetase that

specifically activates pyrrolysine. PNAS 101, 12450Ð12454.

Rajendran N. (1999), Identification and cloning of agene locus encoding peptide synthetase of Pseudomo-nas fluorescens. Z. Naturforsch. 54c, 105Ð109.

Rajendran N., Stachelhaus T., and Marahiel M. A.(1999), Multifunctional peptide synthetase requiredfor nonribosomal peptide synthesis. In: Comprehen-sive Natural Products Chemistry, Vol. 4 (Barton D.,Nakanishi K., and Kelly J. W., eds.). Elsevier ScienceLtd, Oxford, England, pp. 198Ð220.

Ribas P. L., Fugier M., Quinn C. L., and Schimmel P.(1996), Evidence that two present-day componentsneeded for the genetic code appeared after nucleatedcells separated from Eubacteria. PNAS 93, 166Ð170.

Schulz M. E., Vukov N., Shaheen R., Andersson M.,Martlbauer E., and Scherer S. (2005), Identificationand partial characterization of the nonribosomal pep-tide synthetase gene responsible for cereulide produc-tion in emetic Bacillus cereus. Appl. Environ. Micro-biol. 71, 105Ð113.

Srinivasan G. and Krzycki J. (2002). Pyrolysine encodedby IAG in Archaea: charging of a UAG-decoding spe-cialized tRNA. Science 296, 1459Ð1462.

Thompson J. D., Gibson T. J., Plewniak F., JeanmouginF., and Higgins D. G. (1997), The ClustalX windowsinterface: flexible strategies for multiple sequencealignment aided by quality analysis tools. NucleicAcids Res. 24, 4876Ð4882.

Tumbula D. L., Becker H. D., Chang W. Z., and Soll D.(2000), Domain-specific recruitment of amide aminoacids for protein synthesis. Nature 407, 106Ð110.

Turgay K. and Marahiel M. A. (1994), A general ap-proach for identifying and cloning peptide synthetasegenes. Peptide Res. 7, 238Ð241.

Woese C. R., Olsen G. J., Ibba M., and Soll D. (2000),Aminoacyl-tRNA synthetases, the genetic code, andthe evolutionary process. Microbiol. Mol. Biol. Rev.64, 202Ð236.

Wolf Y. I., Aravind L., Grishin N. V., and Koonin E. V.(1999), Evolution of aminoacyl tRNA synthetasesanalysis of unique domain architectures and phyloge-netic trees reveals a complex history of horizontalgene transfer events. Genome Res. 9, 689Ð710.

Documents

Molecular Phylogenetic Analysis of Tryptophanyl-tRNA ...znaturforsch.com/s63c/s63c0418.pdf · Molecular Phylogenetic Analysis of Tryptophanyl-tRNA Synthetase of Actinobacillus actinomycetemcomitans