JOURNAL OF BACTERIOLOGY, Apr. 1992, p. 2606-26110021-9193/92/082606-06$02.00/0Copyright © 1992, American Society for Microbiology

Vol. 174, No. 8

Evolutionary Relationships of a Plant-Pathogenic MycoplasmalikeOrganism and Acholeplasma laidlawii Deduced from

Two Ribosomal Protein Gene SequencesPYUNG-OK LIM1 AND BARBARA B. SEARS',2*

Genetics Program' and Department ofBotany and Plant Pathology,2Michigan State University, East Lansing, Michigan 48824

Received 26 July 1991/Accepted 14 February 1992

The families within the class MoUicutes are distinguished by their morphologies, nutritional requirements,and abilities to metabolize certain compounds. Biosystematic classification of the plant-pathogenic mycoplas-malike organisms (MLOs) has been difficult because these organisms have not been cultured in vitro, and hencetheir nutritional requirements have not been determined nor have physiological characterizations beenpossible. To investigate the evolutionary relationship of the MLOs to other members of the class MoUlicutes, asegment of a ribosomal protein operon was cloned and sequenced from an aster yellows-type MLO which ispathogenic for members of the genus Oenothera and from Acholeplasma laidlawii. The deduced amino acidsequence data from the rp122 and rps3 genes indicate that the MLOs are more closely related toA. laidlawii thanto animal mycoplasmas, confirming previous results from 16S rRNA sequence comparisons. This conclusionis also supported by the finding that the UGA codon is not read as a tryptophan codon in the MLO andA. laidlawii, in contrast to its usage in Mycoplasma capricolum.

The taxonomic position of plant-pathogenic mycoplasma-like organisms (MLOs) has been uncertain since they werefirst recognized more than 20 years ago (2). This is primarilybecause MLOs have not been cultured in vitro (7), and thusthey have not been characterized biochemically nor havetheir nutritional requirements been determined. BecauseMLOs lack a cell wall, they have been tentatively classifiedas members of the class Mollicutes (16), although it has notbeen clear whether they truly resemble the sterol-requiringmycoplasmas or the non-sterol-requiring acholeplasmas.

Recently, we have shown that an aster yellows-typeMLO, which is pathogenic for members of the genusOenothera, has a low G+C content in its genome (18); this isconsistent with the A+T-rich genomes of members of theclass Mollicutes (12). The genome size of the MLO resem-bles those of the highly degenerate animal mycoplasmas (9),but the 16S rRNA sequence data (8) point to a closerrelationship to Acholeplasma laidlawii and a more distantrelationship to the spiroplasmas and animal mycoplasmas.To supplement these studies, we have cloned and sequenceda segment from a ribosomal protein gene cluster from arepresentative of the MLOs and from A. laidlawii. Thesedata are valuable because the primary sequences of proteingenes do not have the base-pairing constraints that slow theevolution of rRNA genes. Furthermore, confirmation of anevolutionary tree by using different molecular data rendersunlikely the possibility of horizontal transfer.We were also interested in examining the codon usage of

the protein genes, because this is another trait that differen-tiates non-sterol-requiring acholeplasmas and sterol-requir-ing mycoplasmas. Although UGA was formerly consideredto be a universal termination codon in prokaryotes andeukaryotes, animal mycoplasmas and spiroplasmas areknown to use UGA as a tryptophan codon (1, 4, 13, 22) in

* Corresponding author.

addition to the standard UGG tryptophan codon. In con-trast, in A. laidlawii, only the UGG codon specifies tryp-tophan (19). Prior to the initiation of this study, the onlysequence data available for MLOs was from the rRNAoperon (8), and thus, the nature of the MLO genetic codewas unknown.

MATERIALS AND METHODS

Bacterial strains and plasmids. Escherichia coli DH5aL wasused as the host for cloning experiments with pUC18. TheA.laidlawii strain was provided by Terry Moser (MichiganState University Veterinary Clinic). Plasmid pMC1088 (13),which was kindly provided by S. Osawa (Nagoya Univer-sity, Nagoya, Japan), contains a segment of the ribosomalprotein gene operon of Mycoplasma capricolum (Fig. 1).

Preparation of plasmid and chromosomal DNAs. PlasmidDNA was prepared by the alkaline method described byManiatis et al. (11). Total MLO DNA was isolated frominfected Oenothera leaf tip cultures according to the methodof Sears et al. (18). Total A. laidlawii DNA was isolated bya method described previously (23), except that the lyso-zyme treatment was omitted and the pronase was replacedwith proteinase K.

Cloning of ribosomal protein genes from an MLO. Themethods were essentially those of Maniatis et al. (11). Aheterologous probe, pMC1088, was used to identify a 2.7-kbHindIII fragment in Southern blots of the Oenothera MLO.DNA fragments of this size were gel purified and ligated intopUC18, which had been digested with HindIII. Clones wereselected by colony hybridization and Southern blot hybrid-ization with the pMC1088 probe under low-stringency wash-ing conditions (Fig. 1).

Cloning of ribosomal protein genes from A. laidlawii. Afterthe MLO sequence was determined, two conserved sites(Fig. 2) bracketing the rp122 and rps3 genes were chosen forthe synthesis of oligonucleotides for use as primers in the

2606

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

EVOLUTIONARY RELATIONSHIPS OF MLO AND A. LAIDLAWII 2607

S3'ILI LI6L2*1 114124LS1 S14 61L8SSeSScY Adk

H X H

B MLO IL2 S3| L162.7kb-Hindlill

4-

FIG. 1. (A) Plasmid pMC1088 containing a segment of the ribo-somal protein gene operon of the M. capncolum. (B) Map of the 2.7-kbHindIII fragment of the Oenothera MLO. HindIII (H) and XbaI (X)sites are indicated. Arrows show the direction of sequencing.

polymerase chain reaction (PCR). For the PCR, 10 ng ofA.laidlawii DNA was used, with 200 ,uM each deoxynucleotidetriphosphate and 0.05 U of Amplitaq (Perkin-Elmer Cetus)per ,ul. The amplification condition consisted of 30 cycles of

1 min at 95°C, 3 min at 42°C, and 3 min at 72°C, followed by10 min at 72°C. After the PCR amplification, 15 U of T4 DNApolymerase was added to 100 ,ul of the PCR mixture to fill inthe ends and this reaction mixture was incubated at 37°C for20 min (3). The gel-purified PCR product was cloned intopUC18 digested with HincII.DNA sequencing. Sequencing of double-stranded plasmid

DNA was performed with Sequenase (U.S. Biochemicals).dITP was used to resolve ambiguities resulting from com-pression in G+C-rich regions. For sequencing, universalforward and reverse primers were used, as were syntheticoligonucleotides made by facilities at Michigan State Uni-versity operated by the Department of Biochemistry and C.R. Somerville. DNA sequence identity and similarity wereassessed by using the "gap" option of the University ofWisconsin Genetics Computer Group Program.

Tree construction. The phylogenetic tree was constructedwith the "phylogenetic analysis using parsimony" (PAUP)program written by David L. Swofford (University of Illi-nois). After the deduced amino acid sequences were aligned,

rpsl9 ich

rpl22

AAAACATGGTAGGACATAAGTTAGGTGAATTTTCCCCTACACGTACTTACCGCG GACAcacAAAGAGGACATAAGTTAGGTGAATTTGCTCCTACACGTACATTCCGTGGACACAAAAAAGGA |

7059

.10 AAAATAAAATAAATAATrLGgA&AG5ATAACT ATGGAAACCAAAAAGCCCAAAGCGATTG 140ach .GAAaTT..AAGAGATAG ...... G AA .TaA.TT ATGGAA ......... GCTAAAGCAATTG 109

alo CTAGAAAAGTTTCAATCGCCCCTCGAAAAGCACGTTTAGTTGTTGATTTAATTCGAGGAAAAAATATTGC 210ach GAAAAACCATTAGAATCGCTCCTAGGAAAGTACGTCTTGTCGTCGATTTAATTCGTGGTAAGAATGTTAA 179

slo ACAAGCTCAAGCCATTTTAACTTTTACCCCTAAAGTAGCTGCTCCCGTTATTTTAAAACTTTTAAACAGT 280ach GGAAGCACAAGCTATTCTAATGTTTACACCTCGTGGTGCATCACCAGTTATTGCAAAAGTCTTAGACTCA 249

slo GCTGTTTCCAATGCGTTAATAA?XTAATTAAACCGCGAACAACTTTATGTTAAAGAAGTTTTTGTCA 350ach GCAATCGCG&ATCGAACACACAACTTAAATTTAAATTTAGAAAACTTATTTGTTAAAGAAGTATGGGCTA 319

ala ACGAAGGTTTGCGTTTAAAACGTATGTTTCCAAGAGCTAAAGGTTCTGGTGATATGATTAAAAAAGAAC 420ach ACGAATCAATTACTATGAAAAGAATGTTACCACGCGCTAAAGGTAGTGGACATTTAATcAGA&AACGTAc 389

slo CAGCCCT CCTTCTAGCACAAACTTGCAAACATCAAAgQAQgAAGAAcAAA rTOGG 490ach ATCACACATTACTGTGG. ....TTGTCG .CAGAAAggGAGTAA TTAAAGC ATGGG 438

lo! TCAAAAAACTAA TCCTAACGGCTTAAGATTAGGCATTATTAGAA pAATCT s560ich ACAAAAAACAAA.TCCAAACGGACTTAGATTAGGTATCATCCGTA CAAAT TGCTGA 508

alo

ach

alo

ach

alo

ach

alo

ach

rps3

alo

ach

alo

ach

alo

ach

ala

ach

ala

ich

sloach

rpl 16 alinchl

GATAAAGAAATTCCTAAT L'TTAAAGAAGATTTTTTAATTCGTAAACTAATCAATAATTTTACTAAAACAAAAAOATGTACCAGCAEtAAGAATTAGAGAATTCCTAAACGAGAACTTTAGTA

AAAGTGCTATCAGTCAAATTGACATTGAACGCCTAAAAGAAAAAAATAAAAACCGTATCACTATTTCTGTAAcLGGGTGTATCTCAAATTGAGATTGAACGCGTTAAAGCTAAGTCTAAAGAAAGAGTTAcAATTAAAcT

CCACACCGCTAAACCAGGCGTTATTATTGGAAAAGATGGCGATACACGCAACAAATTAGTTGCCAAACTCTTATGsTZCAAACCTGGTATTGCATTAGGTAAAGAAGCATCAGTTAAAAATAAAGCAGTTTCTAATTTA

GAATACTTAACTAAAAAAGAAGTTATCTTAAATATTATTGAAGTAAGACGCCCTGAAAAAGTTGCGGTTT

TAATTGCTCAIIOi~TGGCTGAACAACTAGAAATCGTATGTTTTTCCGCCGTGTTCAAA AATGGCAATTAGTTGCTCA7*TTGCAGAACAATTAGAAAACCGTGCTTCATTCAGACGTGCTCAAAAAATGGCTAT

CCAAAAAGCCCTAAAAGCTGGTGCCAAAGGAGTAAAAACTTTAATTTCTGGTCGTTTGGGTGGTGCTGAATCAACGTGCCCTTAAATCAGGCGCTAAAGGTATTAGAACTTTAGTATCTGGTCGTTTAGGTGGAGCAGAA

630578

700648

770718

840788

910858

980928

ATAGCTCGTAGCGAAGGACATGCCGAAGGCAGAGTTCCTCTACACACTCTAAGAGCAGACATCGATTACG 0050ATGGCTAGAAGCGAAGGTTATTCAGAAGGACGCGTGCCTCTACATACATTAAGAGCAGACGTAGACTATG

CTGCTGXTTAAGCTCACTACTTA3iGTTTTAGGA&TTAAAGTATGGATTTTCCACGGTGAAGTTTTCAACAGCAGAAGCAAGTACTACTTAATATCTTAGGTATTAAAGTATGGATTTATCATGGTGAAGTATT

998

11201068

ACCAGGACAAACCATTCTAGACACTAGAAAACCGTTT.. .GCTTCCCAATCTTCTAACACTCCTAACAGA 1187ACCAGGACAATCTATTTTAGACACAAGAAAACCTTTTGAAGCTGGTAATCAAAGACGTGGTCAAAAACGT 1138

CGCCCTCGCAATTTCAAAGG&OCAA.CAATAATC 1221CGTCCACGTAATGACCAACCAGTCAAAGACCTCAACAAAGAAAAAGAAATAGCAAGGAWACTAATTACT 1208

ATGTTAATGCCAAAAAGAACTA ATATCGTCGTCGTCCTCACATGTTAATGCCAAAAAGAACTAA ATATCGT

12631238

FIG. 2. Nucleotide sequence of the ribosomal protein genes (3' region of rpsl9, entire rp122 and rps3, and 5' region of rp116) from the MLO(mlo) and A. laidlawii (ach). Whole and partial reading frames are boxed. Codons in shaded boxes represent positions corresponding to UGAtryptophan codons in M. capricolum. Putative Shine-Dalgarno sequences are underlined. PCR primer regions are shown by the broken lines.

VOL. 174, 1992

A pMCB 1088

II

I

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

2608 LIM AND SEARS

(A)

m1o ME KKPKA'RKW R KRLV DLIRGK IA _AIL V 45ach ME... AKA GKTIRIA R V VLDLIRGK KEA AILMM G 42myc ME.. ..AKAIKLSMIRI RKM LV D IFSK SVA ATLKNLN D 42bst ... A V~RR .[RUIAR L I LIRGK GF I RH pK 42eco ME... T KHRH S RL A LIRGK KSlELYNKK 42

mlo AA IL A NAA N KL EWK ION LLRLF 90ach ASAS DVI- A,NTH N L LFKWAN:ITMKRL 8 7myc AEIKLA NA GMEADKIKT IFiNIEGIPTL KRF 87bst A P I EK 1 A NKNSEH DMDVNN LIVISQAYIVIDIEGIPTLKF 87eco VJA N H RGADIDCPyKtTKIFVDSMI 87

mlo R M KKRTSHI LVTIjTISTNLQTS KjEJPQS1SKN* 130ach IR LIRKRTSHI VVVA ........ R ...... * 112myc RH RAYEIKRTS vIVV.D.......EK .... * 112bst IRAM RASAINKRTSHI IVV.S......... KKE..P...* 114eco RAK RAfRILKRTSHIVV.D ........ R .....* 111(B)

mlo MGQK PNGLRGIIRELE 45ach MGQmNPNGLRLGIIR E KR IAgD AL,1-01WBREEFLNE 45myc MGQK|RSPN|9LRLGIV R YAEKDQY KWZQ ALFK 45bst MGQ KIVGPLRGII ESR E . YADLVH EDLKI YJNK 44eco MGQKIVLPN I1R LG I K T FADNLD K YLTK 45

mlo FTJKSA I IER EK RI NSVHTAK II..GK GDTR 88ach NFSK S IEIER AKS ERVIK LYVSKP IAL..G EASVK 88myc LLD S I IERTTKD.LTLFIK..TARPAI LGQEGKNIEKI 87bst RLQD AVSV IER. .AARVNVfFI ..HTA VI. .G GSE. 82eco ELAK SS WV R.PAKS. IRVrIJ. .HTA IVI..GK GED. 83

mlo NLVAR.LKELT D.N L VKNS IjLIj'A LENR 131

achK

VSN.LEYLT E.VI VRR V VLVA IA LENR 131myc VLAVR VKN. . K LI R I IKSPDARSTLVAIRWjIIGE QUNRI 130bst VEALR LTQLT .EHI IV IKK LD LVAjEjNIJARQLENR| 126eco VEKLR VADIAGVP.AQI IA VRK LD I DS TS LSJ! 127

mlKMFFR I L AKG EISGRLGGAIARSMGHAE R 176ach ASFR A I ALKS AKGIRTtVSGRLGGA S S|EGIR 176myc ASFR AKL ALK AKGIITIVSGRLGGM ARTEGLE S 175bst VSF IA R TG VM RLGGA I S SH E4T 171eco VMF K R AKGIE SGRLGGAEEIRR 1 7 2

mlo VPLHTLRADIDYVEAI TTY KTVWI HGEV PI IL 221ach VPLHTLRA DY AEAE TY KVWIYHGEV P IL 221myc VPLTL NIDY YEATY KVWILNHGEV ........ FK 212bst VPLHTLRADIDYjTAAE TTY KGMKVWIRGEP ...... TK 210eco IVPLHTLRADIDY TSIE TTY GKGKVWIK GGMAAVEQPX 217

mia 1 *MPj! rA1QS!SNTPNFPPRflFe§GjN............. ; 253ach .IKPjEJNQRRGQKRPR DQPVKDLNKEKEIARRT YMNAkkk* 266myc.IK.. ERMNNijQIMA.KPRTNIG R ......* 234bst .NKAE........... EG(.* 219eco tPAA.. .QPKKQQR. KK. * 234

FIG. 3. Alignment of the amino acid sequences of the rp122 (A) and rps3 genes (B) from the Oenothera MLO (mlo), A. laidlawii (ach), M.capricolum (myc), B. stearothernophilus (bst), and E. coli (eco). Boxes indicate amino acids found in at least four organisms. Shaded boxesindicate the regions in which the amino acid sequences are the same in and unique to the MLO-A. laidlawii pair, the MLO-M. capncolumpair, the MLO-B. stearothermophilus pair, or the MLO-E. coli pair. Three variable regions (lines above sequence) in the amino acid sequenceof the rps3 gene (B) are indicated. Asterisks indicate termination codons.

a heuristic analysis with the simple algorithm was con-ducted, with gaps excluded from consideration.

Nucleotide sequence accession numbers. The sequence datahave been communicated to GenBank and have been as-signed accession numbers M74770 (MLO) and M74771 (A.laidlawii).

RESULTS

Cloning, sequencing and DNA alignments. We have cloneda 2.7-kb HindIll DNA fragment from an MLO which ispathogenic for members of the genus Oenothera (Fig. 1) byusing a heterologous probe that contains a segment of aribosomal protein gene operon of M. capricolum (13). By

using this clone, the entire rp122 and rps3 genes weresequenced.

Since the analogous ribosomal genes fromA. laidlawii hadnot been analyzed, we cloned a 1.2-kb fragment from A.laidlawii by using PCR and then sequenced it. This sequencealigns well with the sequence of the MLO clone (Fig. 2),confirming that the targeted ribosomal protein genes wereindeed cloned. The nucleotide sequences for the rp122 geneof the MLO andA. laidlawii are 62% identical. The rps3 genesequences are 67% identical.

In both the MLO and A. laidlawii, ribosomal binding sites(Shine-Dalgarno sequences) are found upstream of the initi-ation codons for the rp122, rps3, and rp116 genes, but nopromoterlike sequence is present. For all but one of the

J. BACTERIOL.

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

EVOLUTIONARY RELATIONSHIPS OF MLO AND A. LAIDLAWII 2609

TABLE 1. Comparisons of amino acid identities and similaritiesof deduced sequences from ribosomal proteins L22 and S3Y

Sequence % Identity (% similarity") among sequences from:source Mlo Ach Myc Bst Eco

Mlo 60 (79) 50 (60) 50 (69) 48 (67)Ach 64 (80) 46 (63) 54 (77) 49 (67)Myc 43 (62) 46 (63) 53 (66) 53 (68)Bst 54 (72) 58 (75) 49 (67) 50 (68)Eco 46 (65) 48 (68) 46 (60) 58 (74)

a Sequences in Fig. 3 were compared. Numbers at the upper right indicatethe relatedness based on the deduced rp122 ribosomal protein sequence;numbers at the lower left indicate the relatedness based on the deduced rps3ribosomal protein sequence. Abbreviations: Mlo, Oenothera MLO; Ach, A.Iaidlawii; Myc, M. capricolum; Bst, B. stearothermophilus; Eco, E. coli.

b Includes conservative amino acid substitutions.

genes, AUG is used as the initiation codon. The exception isthe rps3 gene of the MLO. For this gene, the initiation codonappears to be GUG, which is the most frequently encoun-tered alternative (5). In Bacillus stearothermophilus, GUG isalso used as the initiation codon of the rps3 gene (6).The genes for the rpsl9, rpl22, rps3, and rpll6 ribosomal

proteins are organized very tightly in both the MLO and A.laidlawii, with very short intergenic regions and/or overlap-ping genes (Fig. 2).Comparison of amino acid sequences. Figure 3 shows an

alignment of the deduced amino acid sequences of the rp122(Fig. 3A) and rps3 (Fig. 3B) genes of the Oenothera MLO, A.laidlawii, M. capricolum (13), B. stearothermophilus (6), andE. coli (24). The length of the rpl22 gene (111 codons) in A.laidlawii is almost the same as those of M. capricolum (111codons), B. stearothermophilus (113 codons), and E. coli(110 codons). However, in the MLO, the rpl22 gene is 129codons long. The rps3 genes of M. capricolum and E. coliencode a protein of 233 amino acids. The rps3 genes of theMLO and A. laidlawii are 252 and 265 codons long, respec-tively. In B. stearothermophilus, the rps3 gene is 218 codonslong. Divergence in the amino acid sequences occurs mainlyin three regions, as indicated in Fig. 3B.By using the "gap" option of the University of Wisconsin

Genetics Computer Group Program, we calculated aminoacid identity and similarity, which also includes conservativeamino acid substitutions. As shown in Table 1 (right upperhalf), the rp122 gene of the MLO is most similar to that ofA.laidlawii. In the pairwise comparisons, the rp122 amino acidsequence of the MLO is equally diverged from the analogousgenes in M. capricolum and B. stearothermophilus, withslightly less identity to the rpl22 gene of E. coli. Surprisingly,the pairwise comparisons indicate that the A. laidlawii rpl22gene is least similar to that of M. capricolum.The left lower half of Table 1 shows the amino acid

sequence identity and similarity of the rps3 genes amongthese five organisms. The MLO is most similar to A.laidlawii, which is consistent with the rp122 sequence data.Even in the variable regions (Fig. 3B), a high degree ofsimilarity in the sequences of these two organisms is ob-served. The rps3 amino acid sequences of both the MLO andA. laidlawii are more similar to those of B. stearothermophi-lus and E. coli than to that of M. capricolum.

Phylogenetic trees were constructed by using the PAUPprogram. The minimum tree length was obtained with E. colias the out-group. The trees derived from the two ribosomalprotein genes are shown in Fig. 4. Although the evolutionaryrates of the two protein genes are different, both trees show

rp122

t. capricolumB. stearothermophilus

E. coli

T mlo

A. laidlawli

rps3

mlo

I I A. laidlawifLt. capricolum

B. stearothermophilusE. coil

FIG. 4. Phylogenetic trees of deduced amino acid sequencesfrom the rp122 and rps3 genes. Branch lengths are proportional toevolutionary distances.

that the MLO is closely related to A. laidlawii and that thesetwo bacteria are monophyletic with M. capricolum, althoughtheir divergence is ancient. As expected, the members of theclass Mollicutes cluster with B. stearothermophilus.Codon usage. The codon usages in the rpl22 and rps3 genes

of the Oenothera MLO, A. laidlawii, M. capricolum, B.stearothermophilus, and E. coli were compared (Table 2).About 80% of the codons in the MLO and A. laidlawii haveA or U at the third position. This is true for an even higherfraction of codons (91%) in M. capricolum (13), whosegenome has a lower G+C content (25%) than the MLO(29.5%) and A. laidlawii (30 to 32%). In contrast, the G+Ccontent of the genomes of B. stearothermophilus (6) and E.coli (12) are 52 and 50%, respectively, and 44 to 50% of thecodons have A or U at the third position. The A and Urichness of codons of members of the Mollicutes is alsoevident at the first position. In the three members of theMollicutes studied, the frequency of codons with A or U atthe first position is 51 to 57%, but in B. stearothermophilusand E. coli, the occurrence of A or U at the first position is38%. Unlike the situation with the first and third positions,the frequency of A or U at the second codon position; ofmembers of the Mollicutes (62 to 63%) is similar to those ofB. stearothermophilus (60%) and E. coli (59%).The sequence of the rps3 ribosomal protein gene from M.

capricolum includes five UGA codons in the reading frame(13). In the MLO and A. laidlawii, no UGA codon is foundin either the rp122 or rps3 genes for a total of 381 and 377codons, respectively (Table 2). Nor was any UGA codonfound when we examined several additional nearby openreading frames from the MLO (10). In the five positionscorresponding to the UGA codon in the M. capricolum rps3gene, three UGGs are found in the MLO, B. stearothernno-philus, and E. coli sequences and two UGGs are present inthe A. laidlawii sequence (Fig. 2).

DISCUSSION

We have determined the DNA sequences for two ribo-somal protein genes from a nonculturable plant-pathogenicMLO and A. laidlawii. The comparisons of deduced aminoacid sequences from five bacteria indicate that the plant-pathogenic MLOs are more similar toA. laidlawii than toM.

VOL. 174, 1992

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

2610 LIM AND SEARS

TABLE 2. Codon usage in the Oenothera MLO, A. laidlawii, M. capricolum, B. stearothermophilus,and E. coli ribosomal protein genesa

Second position of codon and no. of times each codon for designated amino acid is used:First U C A G Third

position AA AA AA AA positionml al mc bs ec ml al mc bs ec ml al mc bs ec ml al mc bs ec

U Phe 7 4 7 0 2 Ser 8 5 1 0 7 Tyr 2 7 8 4 0 Cys 1 0 1 00 UPhe 3 2 0 3 4 Ser 2 0 0 3 2 Tyr 1 2 0 5 5 Cys 0 0 0 00 CLeu 20 22 25 3 0 Ser 3 10 10 1 0 Stop 2 2 22 1 Stopb 0 0 5 0 1 ALeu 3 0 0 8 1 Ser 0 0 0 5 1 Stop 0 0 0 0 Trp 3 3 0 3 4 G

C Leu 4 5 0 3 1 Pro 7 6 3 0 4 His 2 3 2 5 3 Arg 9 13 7 10 19 ULeu 1 1 0 4 0 Pro 2 0 0 1 0 His 5 2 1 3 5 Arg 6 2 0 17 12 CLeu 7 3 3 0 Pro 3 7 7 3 1 Gln 15 12 8 7 1 Arg 4 1 0 2 0 ALeu 0 0 0 2 19 Pro 1 0 0 7 7 Gln 00 0 2 10 Arg 0 1 0 4 0 G

A Ile 22 21 19 13 10 Thr 13 8 13 0 7 Asn 17 12 18 3 1 Ser 4 3 5 00 UIle 7 7 7 15 17 Thr 6 1 0 2 9 Asn 14 9 3 9 9 Ser 3 2 0 3 6 CIle 2 1 1 0 0 Thr 3 8 6 3 0 Lys 40 33 39 24 23 Arg 9 13 20 2 0 AMet 6 7 8 6 8 Thr 0 0 0 11 1 Lys 3 4 1 6 12 Arg 0 4 0 0 0 G

G Val 19 14 24 8 15 Ala 19 15 22 8 23 Asp 8 4 13 4 6 Gly 9 6 9 5 13 UVal 2 5 0 10 2 Ala 7 2 0 7 2 Asp 4 6 0 10 12 Gly 6 1 0 7 10 CVal 4 9 7 3 12 Ala 4 17 11 6 6 Glu 20 26 19 21 18 Gly 8 7 10 5 0 AVal 2 2 1 9 4 Ala 1 3 0 15 6 Glu 1 3 0 6 2 Gly 1 0 0 5 0 G

a The identities of the first, second, and third positions of each codon and the numbers of codons present in the reading frames of the L22 and S3 ribosomalprotein genes are indicated. Abbreviations: AA, amino acid; ml, Oenothera MLO; al, A. laidlawii; mc, M. capnicolum; bs, B. stearothermophilus; ec, E. coli.

b UGA codon specifies tryptophan in M. capricolum.

capricolum, confirming previous data in which the 16SrRNA gene sequence of the MLO was searched for thepresence of signature oligonucleotides (8). Surprisingly, thesequence identity of the rps3 gene between the MLO and M.capricolum was less than that of the MLO and B. stearo-thermophilus or E. coli. Similar low levels of identity wereobserved for both the rp122 and rps3 genes of A. laidlawiiand M. capricolum. These data could be interpreted toindicate that the MLO and A. laidlawii are more closelyrelated to B. stearothennophilus and E. coli than to M.capricolum. However, by using the PAUP program, whichconducts a multiway analysis and excludes invariant anduninformative characters, the phylogenetic trees show thatthe MLO andA. laidlawii group with M. capricolum (Fig. 4).Nonetheless, the deep branching indicates that the clade thatcontains the MLO andA. laidlawii diverged from the myco-plasmas early in the evolution of members of the Mollicutes.The length of the branches is consistent with a rapid rate ofevolution within the class Mollicutes, as suggested by pre-vious 5S and 16S rRNA sequence data (17, 20, 21).The G+C content of the genome of members of the

Mollicutes is very low. Thus, a strongly biased mutationpressure replacing GC pairs with AT pairs in DNA musthave occurred during their evolution (15). The effect of thisso-called "AT pressure" on the usage of amino acid codonsis seen in Table 2. The GC to AT substitution in genomes ofmembers of the Mollicutes has occurred mainly at the thirdcodon position but also at the first position. Since functionalconstraints of the protein should limit the change in thecoding sequence, and since silent codon positions are thethird base and sometimes the first base, this explains whyGC to AT base substitutions have occurred more frequentlyat those positions.According to the codon usage data summarized in Table 2,

it appears that, unlike the animal mycoplasmas and spiro-plasmas, the MLO and A. laidlawii probably do not use

UGA as a tryptophan codon. No UGA codon was found inthe reading frames, and the UGG tryptophan codon wasused three times in both the MLO and A. laidlawii genes. Incontrast, in M. capricolum, all of the tryptophan codons areUGA in these genes (13, 14). The codon usage data for A.laidlawii are consistent with previous results showing thatthis organism contains a single tRNAccATrP (19). In theMLO, of the five termination codons that we have observed(this work and reference 11), all are UAA. In A. laidlawii,two stop codons are UAA and one is UAG. Since oursequence data are somewhat limited, it is difficult to con-clude whether UGA is used as a stop codon in the MLO andA. laidlawii, but we expect that its use for termination isinfrequent or rare.

In conclusion, our amino acid comparisons and the codonusage data confirm our earlier results showing that plant-pathogenic MLOs are more closely related to A. laidlawiithan to the mycoplasmas. According to the 16S rRNAsequence comparisons (20), members of the Mollicutes fitfive phylogenetic groups (the hominis, pneumoniae, spiro-plasma, anaeroplasma, and asteroleplasma groups). Theanaeroplasma group includes anaeroplasmas and mostacholeplasmas. Since our data indicate a close relationshipbetween the MLO and A. laidlawii, MLOs probably belongin this group. In order to determine more precisely theappropriate phylogenetic position of MLOs within thisgroup, comparison of sequences from other MLOs and fromother organisms belonging to the anaeroplasma group wouldbe required.

ACKNOWLEDGMENTS

This research was supported by National Science Foundationgrant BSR-89068279 and the Michigan State Agricultural Experi-ment Station.

J. BACTERIOL.

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

EVOLUTIONARY RELATIONSHIPS OF MLO AND A. LAIDL4WII 2611

REFERENCES1. Bove, J. M., P. Carle, M. Garnier, F. Laignet, J. Renaudin, and

C. Saillard. 1988. Molecular and cellular biology of spiroplas-mas, p. 235-346. In R. F. Whitcomb and J. G. Tully (ed.), Themycoplasmas, vol. V: spiroplasmas, acholeplasmas, and myco-plasmas of plants and arthropods. Academic Press, Inc., NewYork.

2. Doi, Y., M. Teranaka, K. Yora, and H. Asuyana. 1969. Myco-plasma or P.L.T. group-like organisms found in the phloemelements of plants infected with mulberry dwarf, potatowitches' broom, aster yellows or Paulowinia witches' broom.Annu. Phytopath. Soc. Jpn. 33:259-266.

3. Huang, M. N., and K. A. High. 1990. Efficient subcloning ofDNA fragments amplified by crude oligonucleotides. BioTech-niques 9:710-711.

4. Inamine, J. M., K.-C. Ho, S. Loechel, and P.-C. Hu. 1990.Evidence that UGA is read as a tryptophan codon rather than asa stop codon by Mycoplasma pneumoniae, Mycoplasma geni-talium, and Mycoplasma gallisepticum. J. Bacteriol. 172:504-506.

5. Kozak, M. 1983. Comparison of initiation of protein synthesis inprocaryotes, eucaryotes, and organelles. Microbiol. Rev. 47:1-45.

6. Krtmer, W. J., T. Hatakeyama, and M. Kimura. 1990. Nucle-otide sequences of Bacillus stearothernophilus ribosomal pro-tein genes: part of the ribosomal S10 operon. Biol. Chem.Hoppe-Seyler 371:631-636.

7. Lee, I. M., and R. E. Davis. 1986. Prospects for in vitro cultureof plant-pathogenic mycoplasma-like organisms. Annu. Rev.Phytopathol. 24:339-354.

8. Lim, P.-O., and B. B. Sears. 1989. 16S rRNA sequence indicatesthat plant-pathogenic mycoplasmalike organisms are evolution-arily distinct from animal mycoplasmas. J. Bacteriol. 171:5901-5906.

9. Lim, P.-O., and B. B. Sears. 1991. The genome size of aplant-pathogenic mycoplasmalike organism resembles those ofanimal mycoplasmas. J. Bacteriol. 173:2128-2130.

10. Lim, P.-O., and B. B. Sears. 1991. DNA sequence of theribosomal protein genes rp12 and rpsl9 from a plant-pathogenicmycoplasma-like organistn. FEMS Microbiol. Lett. 84:71-74.

11. Maniatis, T., E. F. Fritsch, and J. SambrooL 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

12. Muto, A., and S. Osawa. 1987. The guanosine and cytosine

content of genomic DNA and bacterial evolution. Proc. Natl.Acad. Sci. USA 84:166-169.

13. Ohkubo, S., A. Muto, Y. KiAwauchi, F. Yamao, and S. Osawa.1987. The ribosomal protein gene cluster of Mycoplasma capn-colum. Mol. Gen. Genet. 210:314-322.

14. Osawa, S., and T. H. Jukes. 1989. Codon reassignment (codoncapture) in evolution. J. Mol. Evol. 28:271-278.

15. Razin, S. 1985. Molecular biology and genetics of mycoplasmas(Mollicutes). Microbiol. Rev. 49:419-455.

16. Razin, S., and E. A. Freundt. 1984. The mycoplasmas, p.740-793. In N. R. Krieg and J. G. Holt (ed.), Bergey's manualof systematic bacteriology, vol. 1. The Williams and WilkinsCo., Baltimore.

17. Rogers, M. J., J. Simmons, R. T. Walker, W. G. Weisburg,C. R. Woese, R. S. Tanner, I. M. Robinson, D. A. Stahl, G.Olsen, R. H. Leach, and J. Maniloff. 1985. Construction of themycoplasma evo)utionary tree from SS rRNA sequence data.Proc. Natl. Acad. Sci.- USA 82:1160-1164.

18. Sears, B. B., P.-O. Un, N. Holland, B. C. Kirkpatrick, andK. L. Kloimparens. 1989. Isolation and characterization of DNAfrom a mycoplasmalike organism. Mol. Plant-Microbe Interact.2:175-180.

19. Tanaka, R., A. Muto, and S. Osawa. 1989. Nucleotide sequenceof tryptophan tRNA gene in Acholeplasma laidlawii. NucleicAcids Res. 17:5842.

20. Weisburg, W. G., J. G. Tully, D. L. Rose, J. P. Petzel, H.Oyaizu, D. Yang, L. Mandelco, J. Sechrest, T. G. iA*rence, J.Van Etten, J. Maniloff, and C. R. Woese.,1989. A phylogeneticanalysis of the mycoplasmas: basis for their classification. J.Bacteriol. 171:6455-6467.

21. Woese, C. R., J. Maniloff, and L. B. Zablen. 1980. Phylogeneticanalysis of the mycoplasmas. Proc. Natl. Acad. Sci. USA77:494-498.

22. Yamao, F., A. Muto, Y. Kawauchi, M. Iwami, S. Iwagami, Y.Azumi, and S. Osawa. 1985. UGA is read as tryptophan inMycoplasma capncolum. Proc. Natl. Acad. Sci. USA 82:2306-2309.

23. Yee, T., and M. Inouye. 1981. Reexamination of the genome sizeof myxobacteria, including the use of a new method for genomesize analysis. J. Bacteriol. 145:1257-1265.

24. Zurawski, G., and S. M. Zurawski. 1985. Structure of theEscherichia coli S10 ribosomal protein operon. Nucleic AcidsRes. 13:4521-4526.

VOL. 174, 1992

on January 24, 2021 by guesthttp://jb.asm

.org/D

ownloaded from