SGM Journals
Research Article

Identification and distribution of genetic markers in three closely related taxa of the Mycoplasma mycoides cluster: refining the relative position and boundaries of the Mycoplasma sp. bovine group 7 taxon (Mycoplasma leachii)

Florence Tardy, Laure Maigre, François Poumarat, Christine Citti

  • 1AFSSA-Lyon, 31 Avenue Tony Garnier, 69364 Lyon Cedex 07, France
  • 2UMR 1225 INRA, ENVT Ecole Nationale Vétérinaire, 23 Chemin des Capelles, BP 87614, 31076 Toulouse Cedex 3, France
  • Correspondence
    Florence Tardy
    f.tardy{at}afssa.fr

    • Microbiology 2009; 155(11):3775 · https://doi.org/10.1099/mic.0.030528-0

    View at publisher PubMed

    Abstract

    Mycoplasmas belonging to the Mycoplasma mycoides phylogenetic cluster are all important ruminant pathogens that are genetically closely related but differ in terms of severity and prevalence of the associated diseases. They are distributed among six taxa, the description of which has recently been amended. In the present study, DNA fragments that diverge between the type strains of three taxa were enriched using suppression subtractive hybridization. Of the three taxa, two were representative of the well-established species M. mycoides and M. capricolum, while the third one, Mycoplasma sp. bovine group 7 (Mbg7), has only recently been proposed as a separate species, Mycoplasma leachii. Specific DNA fragments were further characterized by sequencing and used as markers to assess the genetic diversity within and between taxa. The data indicate that the selected markers are unequally distributed within their own taxon but also across taxa. The patterns observed suggest the occurrence of a genetic continuum of strains within the M. mycoides cluster that may compromise the boundaries between taxa and, in turn, diagnosis outcomes. For Mbg7, the overall nature and distribution of the markers indicate a rather homogeneous group that is distinct from the M. capricolum and M. mycoides species and might be considered as a genomic chimera between these two species.

    • ↵†These authors contributed equally to this work.

    • Two supplementary tables, with details of the mycoplasma clones used in the study and the overall results of the SSH experiments, are available with the online version of this paper.

    Edited by: J. Renaudin

    INTRODUCTION

    Members of the genus Mycoplasma are wall-less bacteria that belong to the class Mollicutes. These organisms, which are regarded as the smallest replicating cells, have evolved from more classical bacteria by a reduction of their genome size and hence a reduction of their metabolic pathways. Mycoplasmas are widespread in nature and several are well-known pathogens of humans and animals (Razin et al., 1998). In veterinary medicine, the so-called Mycoplasma mycoides cluster is of particular interest because it comprises only ruminant-pathogenic mycoplasmas. More specifically, it is divided into six taxa, five of which include mycoplasmas responsible for animal diseases listed by the World Organisation for Animal Health (OIE).

    The six taxa that constitute the M. mycoides phylogenetic cluster are genetically closely related and share a number of common phenotypic traits (Cottew et al., 1987; Pettersson et al., 1996). However, they differ in terms of their prevalence worldwide, the severity of the associated diseases and their economic impact (Bergonier et al., 1997; Frey, 2002; Thiaucourt & Bölske, 1996). For instance, two taxa are responsible for diseases associated with high mortality and are currently restricted to Africa, Asia and the Middle East: Mycoplasma mycoides subsp. mycoides biotype Small Colony (MmmSC), the aetiological agent of contagious bovine pleuropneumonia (CBPP), and Mycoplasma capricolum subsp. capripneumoniae (Mccp), responsible for contagious caprine pleuropneumonia (CCPP). Mycoplasma capricolum subsp. capricolum (Mcc) and two other taxa, namely Mycoplasma mycoides subsp. mycoides biotype Large Colony (MmmLC) and Mycoplasma mycoides capri (Mmc), recently grouped under one unique subspecies, Mmc (Brown & Bradbury, 2008; Manso-Silvan et al., 2009; Vilei et al., 2006), cause contagious agalactia syndrome (CA) in goats. This syndrome includes mastitis, arthritis, keratoconjunctivitis and pneumonia, and varies from mild to severe, occasionally causing death of the infected animal. CA is distributed worldwide and is considered enzootic in several countries of the Mediterranean Basin (Bergonier et al., 1997). It is important to note that two other species that do not belong to the M. mycoides cluster are responsible for CA; these are the phylogenetically remote Mycoplasma agalactiae (considered as the classic CA agent in sheep) and the close relative Mycoplasma putrefaciens. Finally, members of the sixth taxon of the M. mycoides cluster, namely Mycoplasma sp. bovine group 7 (Mbg7), are less known and not so well described. They have been associated with polyarthritis in calves (Hum et al., 2000; Shiel et al., 1982; Simmons & Johnston, 1963) and with mastitis (Connole et al., 1967), pneumonia (Alexander et al., 1985) or abortion (Hum et al., 2000) in cows. In the 1960s, experimental reproduction of the disease confirmed that the PG50 type strain is highly pathogenic (Connole et al., 1967; Simmons & Johnston, 1963) but the unravelling of its virulence factors has just started (Bower et al., 2003; Djordjevic et al., 2003).

    In addition, the geographical distribution of Mbg7 has been poorly documented. Most of the outbreaks were reported to occur in Australia (Alexander et al., 1985; Connole et al., 1967; Hum et al., 2000; Shiel et al., 1982; Simmons & Johnston, 1963) but cases probably occurred elsewhere (America, Europe, Africa or India) since several strains originating from these countries were included in comparative studies (Djordjevic et al., 2001; Manso-Silvan et al., 2007). Since 2000, there have been no case reports concerning Mbg7 in Europe (Ayling et al., 2004; Poumarat et al., 2006; Thomas et al., 2002) but this observation has to be taken with care because identification of Mbg7 strains is often impaired owing to serological cross-reactivity between Mcc strains and anti-Mbg7 sera and the lack of a specific PCR assay (Le Grand et al., 2004). Recently, the severity of an outbreak in Australia (Hum et al., 2000) remotivated interest in this taxon. Several isolates recovered from this outbreak were further studied and were shown to derive from a unique clone that (i) rapidly disseminated within and among its hosts during the outbreak but (ii) was different from that of other clinical episodes, based on RFLP profiles (Djordjevic et al., 2001).

    The species status of the Mbg7 taxon has long been debated, notably because of low numbers of clinical cases and isolates. In agreement with DNA/DNA hybridization results (Christiansen & Ernø, 1982), several studies suggest that Mbg7 is a distinct species that has diverged from M. capricolum while sharing the bovine host and several antigenic, biochemical and genetic traits with MmmSC (Djordjevic et al., 2003; Monnerat et al., 1999; Salih et al., 1983). Other investigations, mostly based on sequence analyses (Harasawa et al., 2000; Manso-Silvan et al., 2007; Pettersson et al., 1996; Thiaucourt et al., 2000) but also on SDS-PAGE patterns (Costas et al., 1987; Olsson et al., 1990), counter this view and propose that Mbg7 is only another subspecies of M. capricolum. This debate officially ended by the recent proposal to rename Mbg7 as the new species Mycoplasma leachii (Manso-Silvan et al., 2009). Regardless of their designation, members of the former ‘bovine group 7’ or of the new M. leachii deserve to be further studied, as a better knowledge of this taxon is needed for the future identification of new isolates using specific, unambiguous criteria.

    In this study, a fine genetic analysis was undertaken to define (i) whether intra-taxon variation of Mbg7, Mcc and MmmLC isolates compromises the boundaries between taxa and (ii) whether differences at the DNA level reflect the apparent difference in epidemiology existing between on the one hand the goat pathogens MmmLC and Mcc and on the other hand the bovine Mbg7 taxon that seems to have re-emerged in Australia. For this purpose, DNA sequences that diverge between the type strains of Mbg7, MmmLC and Mcc [respectively PG50, Y-Goat (YG) and California Kid (CK)] were enriched using suppression subtractive hybridization (SSH) and further analysed. Their distribution was then assessed in a panel of field isolates and type strains to define the level of inter- and intra-taxon variation. Overall, this analysis sheds new light on the genetic make-up of the three taxa MmmLC, Mcc and Mbg7. It confirms the status of Mbg7 as a species and further suggests that it may represent a chimera between the M. mycoides and M. capricolum species.

    METHODS

    Bacterial strains and culture conditions.

    Escherichia coli used for DNA manipulations was grown at 37 °C in Luria–Bertani broth supplemented with antibiotics when necessary.

    Mycoplasmas were grown at 37 °C under 5 % CO2 in PPLO broth (Difco) supplemented as previously described (Poumarat et al., 1992). Three type strains were used for SSH experiments, namely Mcc CK (ATCC 27343), MmmLC YG and Mbg7 PG50 (NCTC 10114). The distribution of the DNA sequences identified in this study was assessed in a set of mycoplasma clones derived from field isolates of MmmLC, Mcc and Mbg7 from various origins and clinical contexts (see Supplementary Table S1, available with the online version of this paper). These clones were previously described (Maigre et al., 2008) with the exception of five supplementary Mbg7 strains kindly provided by F. Thiaucourt (UMR15 CIRAD-INRA, Montpellier, France): two from Australia (#063 and #064), one from India (#9733), one from Germany (#941006) and one from Nigeria (#940923). Strain B144P, representative of group L according to the classification of Al-Aubaidi & Fabricant (1971), a group described as very close to Mbg7 (Askaa et al., 1978), was also included as a reference. Several other type strains that represent mycoplasma species usually or occasionally isolated from small ruminants in Europe were also used. These were (i) two strains that are other members of the M. mycoides cluster, namely MmmSC PG1 (NCTC 10114) and Mmc PG3 (NCTC 10137); (ii) three strains that belong to the Spiroplasma phylogenetic group but not to the M. mycoides cluster, namely M. putrefaciens strain KS1 (NCTC 10155), Mycoplasma cottewii strain VIS (NCTC 11732) and Mycoplasma yeatsii strain GIH (NCTC 11730); and (iii) six other type strains that represent the ‘Hominis’ phylogenetic group, namely Mycoplasma auris strain UIA (NCTC 11731), Mycoplasma arginini strain G230 (NCTC 10129), Mycoplasma conjunctivae strain HRC581 (NCTC 10147), Mycoplasma adleri (ATCC 27948), M. agalactiae strain PG2 (NCTC 10123) and Mycoplasma ovine/caprine serogroup 11 strain 2-D.

    SSH and SSH library constitution.

    An SSH experiment was carried out, using Mcc California Kid (CK) genomic DNA as the driver and Mbg7 PG50 (PG50) genomic DNA as the tester (SSH-C). Experimental methodology was identical to that previously described (Maigre et al., 2008). The hybridization was performed at 45 °C and tester-specific sequences were amplified by PCR and cloned into E. coli using the pGEM-T-Easy plasmid. The presence in E. coli libraries of cloned DNA sequences was assessed by PCR, directly on individual colonies, using the N24 and J24 primers that correspond to specific parts of the adapters. DNA inserts with an estimated size >100 bp were sequenced. Duplicated inserts or sequence chimeras displaying an internal Sau3AI restriction site were excluded from the rest of the study.

    DNA extraction and PCR assays.

    Genomic DNA was extracted from mycoplasmas grown in liquid media using the DNeasy tissue kit (Qiagen). Plasmids were purified from E. coli using a QIAprep Spin Miniprep kit (Qiagen). Experimental conditions and oligonucleotides used for PCR assays were as previously described (Maigre et al., 2008).

    Dot-blot DNA hybridization and pattern analysis.

    For dot-blot hybridizations, genomic DNAs spotted onto a nylon membrane were hybridized with heat-denatured probes corresponding to N24-J24 PCR products labelled with the ECL Direct Nucleic Acid Labelling system (GE Healthcare Life Sciences), as described in a previous study (Maigre et al., 2008). Hybridization signals were revealed using detection systems according to the manufacturer's instructions and were analysed, when needed, with the Volume Circle Tool of the Diversity Database program (Bio-Rad). Briefly, an identical circular boundary was created around each spot and the local background was subtracted. For each membrane, dots corresponding to the tester and driver DNA were used as standards and were assigned concentrations of 100 and 1, respectively. The unknown volume concentration of individual spots was calculated from the volume regression curve and was set as the hybridization score. The scores were split into three categories (0–25 %, 25–50 % and 50–100 %), which were represented by different levels of grey for visual interpretation of the results (Fig. 1⇓). A hierarchical clustering of strains based on the hybridization scores was performed using the R-software statistical package (). The distance matrix was calculated using the Euclidean method and the strains were agglomerated using the average linkage method (R-Development_Core_Team, 2009).

    Figure image not available in archive
    Fig. 1.

    Dot-blot hybridization patterns of PG50-, YG- and CK-specific DNA fragments against genomic DNA from Mcc (in yellow), MmmLC (in blue) and Mbg7 (in orange) field strains. The B114P strain and several type strains were also added as references. Hybridization signals were quantified using the Diversity Database program (see Methods) and the resulting intensity was converted to a grey level according to the left bottom corner scale. The symbols *, ° and # refer to the same categories as in Tables 1, 2 and 3.

    Sequence analyses.

    Redundancy of SSH-selected sequences was monitored using the CAP3 program (Huang & Madan, 1999). Nucleotide and protein sequence comparisons with databases were performed using blastn and blastx programs (without filter for low-complexity regions) through the National Center for Biotechnology Information (NCBI) resource (), where Mcc CK and MmmSC PG1 whole-genome sequences are available in the nr/nt databases (GenBank accession no. NC_007633 and NC_005364, respectively). The genome of MmmLC strain GM12 is also available as four contigs with accession numbers AAZK1000001–AAZK1000004 in the whole genome sequence shotgun (wgs) database. These contigs cover the whole genome with the exception of a 60 bp region (Lartigue et al., 2007). Matches were considered significant when displaying an E-value ≤10−10 and only the best hit obtained in each database was selected (except for two very close E-scores which were then both mentioned).

    RESULTS AND DISCUSSION

    Selection of specific DNA fragments

    DNA fragments (i) that are present in Mcc California Kid (CK) but absent (or greatly divergent) from MmmLC Y-Goat (YG) (SSH-A) and (ii) that are present in YG or in Mbg7 PG50 (PG50) but absent (or greatly divergent) from CK (SSH-B and C, respectively) were isolated by SSH (for the overall results see Supplementary Table S2). The two SSH experiments, SSH-A and SSH-B, were performed in a previous study to identify taxon-specific markers for diagnosis (Maigre et al., 2008). The resulting tester-specific fragments were further analysed in the present study. A third SSH, SSH-C, was carried out here and resulted in a library of 86 DNA fragments ranging from 100 to 1000 bp, of which 34 were shown to specifically react in dot-blot hybridization with DNA from PG50 but not with DNA from CK. Based on sequence analysis, one duplicated DNA fragment and six DNA fragment chimeras displaying an internal Sau3AI restriction site were further excluded. The tester-specific sequences (23, 35 and 27 for CK, YG and PG50 respectively) were compared to sequences present in nr/nt or wgs databases using the blastn algorithm and best hits are presented in Tables 1⇓, 2⇓ and 3⇓.

    Table 1.

    Summary of blastn analyses of DNA fragments derived from the California Kid library

    The Seq. column shows the designation of the CK DNA fragments derived from SSH-A. The Size column shows fragment length in base pairs. The nr/nt and the wgs databases were searched independently. The E-value column refers to the E-value obtained for the best match given from the blastn alignment. When two very close E-scores were obtained, they are both mentioned. ns, non-significant E-scores [an arbitrary cut-off was set at E=1×10−10 (1E–10)]; na, not annotated.

    Table 2.

    Summary of blastn analyses of DNA fragments derived from the Y-Goat library

    The Seq. column shows the designation of the YG-DNA fragments derived from SSH-B. * indicates DNA fragments involved in sugar metabolism. Other footnotes and abbreviations are as in Table 1.

    Table 3.

    Summary of blastn analyses of DNA fragments derived from the PG50 library

    The Seq. column shows the designation of the PG50 DNA fragments derived from SSH-C. For abbreviations see Table 1. * indicates DNA fragments involved in sugar metabolism.

    Divergences between Mcc CK and MmmLC YG

    The 23 Mcc CK-specific sequences (SSH-A) could be readily identified because of the CK genome sequence being available in GenBank (accession no. NC_007633). Except for one non-coding DNA fragment (A1), all corresponded to annotated genes (see second column of Table 1⇑), with a majority (18/22) predicted to encode membrane-associated proteins. One of these (A51) was previously described and used to develop an Mcc-specific PCR assay (Maigre et al., 2008). Interestingly, 6/23 of the subtracted sequences were encoded within a putative integrative conjugative element (ICE), ICEC. As for other bacteria (Burrus & Waldor, 2004), these conjugative elements may be a source and a motor for promoting divergence between the species M. capricolum and M. mycoides and within each of these species. As expected, alignment scores of the CK-specific DNA fragments with wgs reads of MmmLC GM12 strain (accession nos AAZK1000001–AAZK1000004 in the wgs database) were low, confirming the divergence of the selected sequences between M. capricolum and M. mycoides (Table 1⇑, third column). This divergence was nicely illustrated by the selection of a fragment (A6) corresponding to the arcA gene, which encodes an arginine deiminase that is important for the anaerobic catabolism of arginine and, within the M. mycoides cluster, only Mcc was shown to be able to hydrolyse this amino acid (Pettersson et al., 1996).

    Of the 35 MmmLC YG-specific fragments derived from SSH-B, most matched sequences of the wgs database (see Table 2⇑, third column) that correspond to MmmLC GM12, a strain from the USA considered as the biochemical and serological equivalent of YG (DaMassa et al., 1992). Approximately half of the YG-specific sequences corresponded to hypothetical products with predicted membrane-associated functions because they contained transmembrane domains, lipobox domains or domains associated with transport functions. Only three fragments (B7, B15 and B57) displayed no significant blastn or blastx hit with GM12, although they were predicted as coding sequences. When analysed against the nr/nt databases (which do not contain MmmLC GM12 sequences), the B15 fragment of YG matched part of an insertion sequence (IS) element, ISMmy2A, described for another MmmLC strain, while the B7 and B57 fragments obtained no significant hits. As expected, in the nr/nt databases, the most significant blast hits (19/35) were obtained with the genome of MmmSC PG1 (PG1) (GenBank accession no. NC_005364), a mycoplasma of the M. mycoides cluster that is the closest relative to MmmLC, and the least significant hits (4/35) were obtained with the driver (CK) genome. Interestingly, a number of YG-specific fragments matched MmmLC GM12 and MmmSC PG1 sequences that are predicted to encode products involved in sugar metabolism, suggesting that this metabolic function diverged between the M. mycoides species and Mcc CK. However, as already suggested (Abu-Groun et al., 1994), the M. mycoides species is not homogeneous in terms of polysaccharide metabolism. Here three YG-specific fragments (B11, B38 and B79), most likely involved in starch utilization, were shown to be absent in PG1 but present in close proximity to one another on the GM12 chromosome, suggesting that they occur as an operon that is lacking in MmmSC PG1. Several other differences were observed between blast results obtained with MmmSC PG1 and MmmLC GM12, some of which reflected differences in annotation (see for instance fragments B67, B10, B29 and B37), while others showed true genetic differences. For instance, the B68 fragment matched a DNA adenine methylase in GM12 but had no significant hit in other mycoplasmas, not even in MmmSC PG1, which is not surprising for a product linked to specific R–M (restriction–modification) systems.

    Nature of DNA fragments that diverge between Mbg7 PG50 and Mcc CK

    Only two of the PG50-specific DNA fragments obtained from the third SSH experiment (SSH-C) matched the few Mbg7 sequences that were already available in databases. The first fragment, C26, corresponds to a gene encoding a major immunodominant lipoprotein P67 that was proposed as a target for a taxon-specific PCR (Frey et al., 1998). The second, C55, corresponds to gtsA, which is involved in glycerol transport but also indirectly in hydrogen peroxide production, a phenomenon that was suggested in MmmSC to play a role in cell toxicity (Pilo et al., 2005). The gtsA gene belongs to a locus that was shown to differentiate highly virulent African/Australian MmmSC strains from low-virulence European ones (Vilei & Frey, 2001). Both fragments correspond to genes previously characterized as being divergent from those of mycoplasmas of the M. capricolum species but highly similar to their MmmSC homologues (Djordjevic et al., 2003; Frey et al., 1998) and, accordingly, their E-score values obtained with MmmSC PG1 were almost identical to those obtained with PG50. In addition, for 21 out of the remaining 25 PG50-specific DNA fragments the most significant blast hits in the nr/nt databases were obtained with PG1 sequences (Table 3⇑). The four remaining fragments (C16, C78, C80, and C82) matched sequences of the driver DNA, CK, although they were divergent enough at the nucleotide level to be successfully selected by SSH. One (C80) corresponds to an oligopeptide ABC transporter and gave an E-score nearly identical to that obtained with the PG1 homologue. The three others did not match at all with MmmSC PG1 genome sequences and correspond to lipoproteins or to membrane proteins that are found across the M. mycoides cluster. Interestingly, these data indicate that the outcome of SSH-C contrasts with that of SSH-B, in which almost half (16/35) of the MmmLC YG-specific fragments showed no significant blast hits with MmmSC PG1 sequences. Since these two subtractive hybridizations, SSH-B and SSH-C, were both conducted using Mcc CK as the driver, this indicates that PG50-specific sequences analysed here (which were not subtracted by CK) may be closer to MmmSC PG1 than YG-specific sequences (not subtracted by CK). This is further supported by the overall high E-scores obtained between PG50-specific fragments and the MmmLC GM12 strain. Overall, about half (14/27) of the PG50-specific DNA fragments corresponded to peptides involved in conserved metabolic functions, with several products (7/14) associated with sugar transport or metabolism. Interestingly, part of the PTS system already hit by the YG-specific DNA fragment (B12) was also recovered as a PG50-specific fragment (C19). The remaining 13 DNA fragments presented significant E-values with hypothetical coding sequences, among which nine were predicted to be membrane-associated.

    Sequences selected by SSH are characteristic of the M. mycoides cluster

    To evaluate the genetic diversity existing within and between the Mcc, MmmLC and Mbg7 taxa, the distribution of the CK-, YG- and PG50-specific sequences isolated by SSH (see above) was assessed using a collection of clinical isolates (see Supplementary Table S1), several of which were previously described (Maigre et al., 2008). This was performed by dot-blot hybridization using each tester-specific sequence as a probe and results were expressed as an intensity score attributed to each hybridization spot with the help of the Diversity Database program. In this experiment, several additional type strains were included that represent mycoplasma species usually or occasionally isolated from small ruminants. More specifically, several belong to the M. mycoides cluster (MmmSC and Mmc) or are related members of the same Spiroplasma phylogenetic group (M. putrefaciens, M. cottewii and M. yeatsii), while others belong to the remote Hominis phylogenic group (M. auris, M. arginini, M. conjunctivae, M. adleri, M. agalactiae and Mycoplasma serogroup 11). The B144P strain representative of an unassigned group of bovine mycoplasmas, the serogroup L (Al-Aubaidi & Fabricant, 1971), was also included because it was shown to be up to 94 % homologous to PG50 (Askaa et al., 1978). The distribution of the tester-specific DNA fragments among this collection of strains is presented in Fig. 1⇑, where the different levels of grey indicate the hybridization intensity.

    As expected, none of the CK-, YG- or PG50-specific sequences hybridized with type strains of the Hominis group and only a few (6/85) gave a positive signal with M. yeatsii, M. cottewii or M. putrefaciens type strains. These data confirmed that most of these probes are specific of strains belonging to the M. mycoides cluster and react only poorly with closely related taxa. DNA from the B144P strain hybridized with 21/27 PG50-specific fragments, in agreement with the important relatedness between PG50 and B144P as inferred by whole-genome DNA/DNA hybridization data (Askaa et al., 1978). In the past, the classification of the B144P strain as a member of Mbg7 varied depending upon the methodology used. While distinguishable from the Mbg7 strains based on the P67-PCR, a PCR that claims to be specific to the Mbg7 taxon (Frey et al., 1998), or by sequence analysis of a conserved locus (Thiaucourt et al., 2000), the B144P strain fell into the Mbg7 group based on rpoB (Vilei et al., 2006) and fusA (this study) sequences and on restriction analysis (Djordjevic et al., 2001).

    Interestingly, the Mmc type strain PG3 reacted more with probes derived from PG50 (21/27) than with those derived from YG (15/35), underlining the divergence between PG3 and YG despite their 75–94 % DNA/DNA hybridization (Askaa et al., 1978; Christiansen & Ernø, 1982). Furthermore, MmmSC PG1 is more recognized by PG50-specific (24/27) than by YG-specific (15/35) fragments. This correlates with in silico analyses which showed that after subtraction of sequences shared with CK, PG50 appears closer to PG1 than YG.

    Distribution of selected sequences within the M. mycoides cluster and perimeter of the Mbg7 taxon

    When considering hybridization patterns with field strains, one of the most striking results is the high homogeneity of the Mbg7 taxon, which contrasts with the important intra-taxon polymorphisms of Mcc and particularly of MmmLC. Indeed, 23/27 PG50-specific probes were present in all Mbg7 strains while only 9/23 CK- and 2/35 YG-specific fragments hybridized with all Mcc and MmmLC field strains, respectively. This observation suggests that genes are highly conserved among Mbg7 strains, as shown recently by sequence analysis of several housekeeping genes (Manso-Silvan et al., 2007) but does not give any indication of gene synteny. On the other hand, the polymorphism between several Mbg7 strains as revealed by RFLP analysis could result from gene rearrangements rather than from gene divergence (Djordjevic et al., 2001). In our hands, only the bovine strain #064 isolated in Australia in 1982 was shown to differ from the other Mbg7 strains by the lack of signal with four PG50-specific probes that encode enzymes involved in sugar metabolism (C32 and C45), a conserved ABC transporter (C80) and a putative lipoprotein (C78). Interestingly, these four probes belong to the group of six that distinguished B144P from PG50. Strong intra-taxon genetic relatedness was described for two other taxa of the M. mycoides cluster, namely MmmSC and Mccp (Kusiluka et al., 2001), which are both known as the most severe mycoplasma pathogens of ruminants. Whether this reflects the preferential spreading of a highly pathogenic clone or a high selection pressure that would limit genetic diversity within taxa is an interesting question that remains to be answered.

    As shown in Fig. 1⇑, a majority of the CK, YG and PG50 fragments selected as ‘specific’ by SSH also recognized strains outside of their own taxon. Indeed, only 7/23 and 9/35 of specific fragments for CK and YG, respectively, do not cross-react outside of their taxon and none of the PG50. Interestingly, among the seven CK fragments that were taxon-specific, five were shown to be located in genes encoded within the ICE of CK (ICEC) (Table 1⇑). Hybridization data indicate that ICEC-like genes were present in 50–75 % of Mcc strains, regardless of their history, but that they were found neither in Mbg7 nor in MmmLC strains (Fig. 1⇑). This was further confirmed by Southern blot using PCR probes targeting two other genes of the ICEC, namely MCAP_0559 (traE) and MCAP_0565 (traG). No hybridization with non-Mcc strains was obtained except for the #064 Mbg7 strain that hybridized with the MCAP_0565 probe (data not shown). This strain was already shown to be divergent from the highly homogeneous group of Mbg7 strains (see above). These results suggest that ICEs contribute to strain genetic variability within a given taxon but the question of whether these elements are involved in inter-taxa divergence has yet to be assessed.

    In contrast, some fragments were present in nearly all field strains, regardless of their identification. This is the case for B11 and B79, which encode proteins involved in starch degradation. We considered that the high mol% A+T of subtracted sequences, which was rather constant between probes, was not responsible for potential non-specific reactions. This ‘out-of-the-taxon’ distribution of fragments contributes to the blurring of boundaries between taxa. The case of the B15 probe is different as it is part of an MmmLC IS (see Table 2⇑). It is evenly distributed among strains as is IS1296. The copy number of IS1296 is very variable in MmmLC and Mcc (see Supplementary Table S1). In contrast, Mbg7 strains showed a stable number (2) of IS1296 copies (Frey et al., 1995; Supplementary Table S1) and did not hybridize with B15. These data suggest that the genome of Mbg7 is quite stable.

    Relative positioning the Mbg7 taxon

    The PG50-specific fragments were the most widely distributed among other taxa: 14/27 recognized at least one Mcc strain (except CK) and 26/27 recognized most if not all MmmLC strains. Reciprocally, the Mbg7 strains were shown to hybridize with a comparable proportion of sequences specific to CK (36 %) and YG (31 %). These results confirm in vitro DNA/DNA hybridization data that showed PG50 to be equidistant from YG (57±7 % homology) and CK (58±8 % homology) (Christiansen & Ernø, 1982), but do not support the hypothesis that Mbg7 might be a subspecies of M. capricolum (Harasawa et al., 2000; Manso-Silvan et al., 2007; Pettersson et al., 1996; Thiaucourt et al., 2000). One striking example of the proximity between MmmLC and Mbg7 is the presence of the B68 probe, which corresponds to a DNA methylase involved in self-recognition in almost all MmmLC and Mbg7 strains but not in Mcc strains or any other type strains of small ruminant mycoplasmas.

    Inter-strains distances were further inferred by a hierarchical clustering of strains as a function of their hybridization patterns using all the 85 taxon-specific probes. The resulting dendrogram confirmed the species status of the Mbg7 lineage, as recently reported (Manso-Silvan et al., 2009), but also its relative proximity with the MmmLC group and with the type strain of MmmSC, PG1 (Fig. 2⇓). The overall proximity between Mbg7 and MmmSC taxa should be further confirmed using other MmmSC strains. Altogether these analyses indicate that Mbg7 might be considered as a homogeneous taxonomic entity, most likely as a chimera between the species M. capricolum and M. mycoides.

    Figure image not available in archive
    Fig. 2.

    Hierarchical clustering of strains as obtained from the overall hybridization scores. Names of the type strains are listed in Methods. The length of vertical branches indicates the distance between strains according to their hybridization profile. Branches are drawn in yellow for strains of the Mcc taxon, blue for strains of the MmmLC taxon and orange for strains of the Mbg7 taxon.

    Conclusion

    In this study, genetic markers that diverge between the M. capricolum and M. mycoides species were selected. Many of these encode products involved in sugar metabolism and/or associated with the membrane. Distribution of these markers is not homogeneous within a given taxon and may occur across taxa, blurring taxon boundaries and thus challenging strain identification. Mbg7 (now M. leachii) appears to be a chimera between M. capricolum and M. mycoides and represents a perfect example of the genetic continuum existing between strains across the species frontiers. Whether this is the result of divergent evolution from a common ancestor or of genomic exchanges between strains sharing the same habitat remains to be elucidated. Massive horizontal gene transfer (HGT) has been demonstrated between ruminant mycoplasmas belonging to remote phylogenetic groups (Sirand-Pugnet et al., 2007) but is more difficult to assess between closely related organisms such as the mycoplasmas of the M. mycoides cluster. Recently, however, HGT has been reported in an MmmLC strain, where the 5′-end of the lepA gene showed sequence characteristics of M. capricolum, while the 3′-end clearly derived from an M. mycoides sequence (Manso-Silvan et al., 2007). In the case of the Mbg7 taxon, the consistent low-copy IS profiles and the absence of ICEC do not support horizontal genetic exchange, though Mbg7 strains could harbour other mobile genetic elements that are still to be identified.

    Acknowledgments

    We are very grateful to L. Manso-Silvan and F. Thiaucourt (UMR15 CIRAD-INRA, Montpellier) for kindly supplying several Mbg7 isolates.

    References

    1. Abu-Groun, E. A. M., Taylor, R. R., Varsani, H., Wadher, B. J., Leach, R. H. & Miles, R. J. (1994). Biochemical diversity within the ‘Mycoplasma mycoides cluster’. Microbiology 140, 2033–2042.
    2. Al-Aubaidi, J.-M. & Fabricant, J. (1971). Characterization and classification of bovine mycoplasma. Cornell Vet 61, 490–518.
    3. Alexander, P. G., Slee, K. J., McOrist, S., Ireland, L. & Coloe, P. J. (1985). Mastitis in cows and polyarthritis and pneumonia caused by Mycoplasma species bovine group 7. Aust Vet J 62, 135–136.
    4. Askaa, G., Ernø, H. & Ojo, M. O. (1978). Bovine mycoplasmas: classification of groups related to Mycoplasma mycoides. Acta Vet Scand 19, 166–178.
    5. Ayling, R. D., Bashiruddin, S. E. & Nicholas, R. A. (2004). Mycoplasma species and related organisms isolated from ruminants in Britain between 1990 and 2000. Vet Rec 155, 413–416.
    6. Bergonier, D., Berthelot, X. & Poumarat, F. (1997). Contagious agalactia of small ruminants: current knowledge concerning epidemiology, diagnosis and control. Rev Sci Tech 16, 848–873.
    7. Bower, K., Djordjevic, S. P., Andronicos, N. M. & Ranson, M. (2003). Cell surface antigens of Mycoplasma species bovine group 7 bind to and activate plasminogen. Infect Immun 71, 4823–4827.
    8. Brown, D. R. & Bradbury, J. M. (2008). International commitee on systematics of procaryotes – subcommittee on the taxonomy of mollicutes. Minutes of the meetings, 6 and 11 July 2008, Tianjin, PR China. Int J Syst Evol Microbiol 58, 2987–2990.
    9. Burrus, V. & Waldor, M. K. (2004). Shaping bacterial genomes with integrative and conjugative elements. Res Microbiol 155, 376–386.
    10. Christiansen, C. & Ernø, H. (1982). Classification of the F38 group of caprine mycoplasma strains by DNA hybridization. J Gen Microbiol 128, 2523–2526.
    11. Connole, M. D., Laws, L. & Hart, R. K. (1967). Mastitis in cattle caused by a Mycoplasma sp. Aust Vet J 43, 157–162.
    12. Costas, M., Leach, R. H. & Mitchelmore, D. L. (1987). Numerical analysis of PAGE protein patterns and the taxonomic relationships within the ‘Mycoplasma mycoides cluster’. J Gen Microbiol 133, 3319–3329.
    13. Cottew, G. S., Breard, A., DaMassa, A. J., Ernø, H., Leach, R. H., Lefèvre, P. C., Rodwell, A. W. & Smith, G. R. (1987). Taxonomy of the Mycoplasma mycoides cluster. Isr J Med Sci 23, 632–635.
    14. DaMassa, A. J., Wakenell, P. S. & Brooks, D. L. (1992). Mycoplasmas of goats and sheep. J Vet Diagn Invest 4, 101–113.
    15. Djordjevic, S. R., Forbes, W. A., Forbes-Faulkner, J., Kuhnert, P., Hum, S., Hornitzky, M. A., Vilei, E. M. & Frey, J. (2001). Genetic diversity among Mycoplasma species bovine group 7: clonal isolates from an outbreak of mastitis, and abortion in dairy cattle. Electrophoresis 22, 3551–3561.
    16. Djordjevic, S. P., Vilei, E. M. & Frey, J. (2003). Characterization of a chromosomal region of Mycoplasma sp. bovine group 7 strain PG50 encoding a glycerol transport locus (GtsABC). Microbiology 149, 195–204.
    17. Frey, J. (2002). Mycoplasmas of animals. In Molecular Biology and Pathogenicity of Mycoplasmas, pp. 73–90. Edited by S. Razin & R. Herrmann. New York: Kluwer Academic/Plenum Publishers.
    18. Frey, J., Cheng, X., Kuhnert, P. & Nicolet, J. (1995). Identification and characterization of IS1296 in Mycoplasma mycoides subsp. mycoides SC and presence in related mycoplasmas. Gene 160, 95–100.
    19. Frey, J., Cheng, X., Monnerat, M. P., Abdo, E. M., Krawinkler, M., Bolske, G. & Nicolet, J. (1998). Genetic and serological analysis of the immunogenic 67-kDa lipoprotein of Mycoplasma sp. bovine group 7. Res Microbiol 149, 55–64.
    20. Harasawa, R., Hotzel, H. & Sachse, K. (2000). Comparison of the 16S–23S rRNA intergenic spacer regions among strains of the Mycoplasma mycoides cluster, and reassessment of the taxonomic position of Mycoplasma sp. bovine group 7. Int J Syst Evol Microbiol 50, 1325–1329.
    21. Huang, X. & Madan, A. (1999). CAP3: A DNA sequence assembly program. Genome Res 9, 868–877.
    22. Hum, S., Kessell, A., Djordjevic, S., Rheinberger, R., Hornitzky, M., Forbes, W. & Gonsalves, J. (2000). Mastitis, polyarthritis and abortion caused by Mycoplasma species bovine group 7 in dairy cattle. Aust Vet J 78, 744–750.
    23. Kusiluka, L. J. M., Ojeniyi, B., Friis, N. F., Kokotovic, B. & Ahrens, P. (2001). Molecular analysis of field strains of Mycoplasma capricolum subspecies capripneumoniae and Mycoplasma mycoides subspecies mycoides, small colony type isolated from goats in Tanzania. Vet Microbiol 82, 27–37.
    24. Lartigue, C., Glass, J. I., Alperovich, N., Pieper, R., Parmar, P. P., Hutchison, C. A., III, Smith, H. O. & Venter, J. C. (2007). Genome transplantation in bacteria: changing one species to another. Science 317, 632–638.
    25. Le Grand, D., Saras, E., Blond, D., Solsona, M. & Poumarat, F. (2004). Assessment of PCR for routine identification of species of the Mycoplasma mycoides cluster in ruminants. Vet Res 35, 635–649.
    26. Maigre, L., Citti, C., Marenda, M., Poumarat, F. & Tardy, F. (2008). Suppression-subtractive hybridization as a strategy to identify taxon-specific sequences within the Mycoplasma mycoides Cluster: design and validation of a M. capricolum subsp. capricolum-specific PCR assay. J Clin Microbiol 46, 1307–1316.
    27. Manso-Silvan, L., Perrier, X. & Thiaucourt, F. (2007). Phylogeny of the Mycoplasma mycoides cluster based on analysis of five conserved protein-coding sequences and possible implications for the taxonomy of the group. Int J Syst Evol Microbiol 57, 2247–2258.
    28. Manso-Silvan, L., Vilei, E. M., Sachse, K., Djordjevic, S. P., Thiaucourt, F. & Frey, J. (2009). Mycoplasma leachii sp. nov. as a new species designation for Mycoplasma sp. bovine group 7 of Leach, and reclassification of Mycoplasma mycoides subsp. mycoides LC as a serovar of Mycoplasma mycoides subsp. capri. Int J Syst Evol Microbiol 59, 1353–1358.
    29. Monnerat, M. P., Thiaucourt, F., Poveda, J. B., Nicolet, J. & Frey, J. (1999). Genetic and serological analysis of lipoprotein LppA in Mycoplasma mycoides subsp. mycoides LC and Mycoplasma mycoides subsp capri. Clin Diagn Lab Immunol 6, 224–230.
    30. Olsson, B., Bolske, G., Bergstrom, K. & Johansson, K. E. (1990). Analysis of caprine mycoplasmas and mycoplasma infections in goats using two-dimensional electrophoresis and immunoblotting. Electrophoresis 11, 861–869.
    31. Pettersson, B., Leitner, T., Ronaghi, M., Bölske, G., Uhlèn, M. & Johansson, K. E. (1996). Phylogeny of the Mycoplasma mycoides cluster as determined by sequence analysis of the 16S rRNA genes from the two rRNA operons. J Bacteriol 178, 4131–4142.
    32. Pilo, P., Vilei, E. M., Peterhans, E., Bonvin-Klotz, L., Stoffel, M. H., Dobbelaere, D. & Frey, J. (2005). A metabolic enzyme as a primary virulence factor of Mycoplasma mycoides subsp. mycoides small colony. J Bacteriol 187, 6824–6831.
    33. Poumarat, F., Longchambon, D. & Martel, J.-L. (1992). Application of dot immunobinding on membrane filtration (MF dot) to the study of relationships within “M. mycoides cluster” and within “glucose and arginine-negative cluster” of ruminant mycoplasmas. Vet Microbiol 32, 375–390.
    34. Poumarat, F., Le Grand, D., Mercier, P., Tardy, F., Gaurivaud, P. & Calavas, D. (2006). VIGIMYC, le réseau français d'épidémio-surveillance des mycoplasmoses des ruminants, bilan 2003–2005. Le Nouveau Praticien Vétérinaire 3, 22–26.
    35. Razin, S., Yogev, D. & Naot, Y. (1998). Molecular biology and pathogenicity of mycoplasmas. Microbiol Mol Biol Rev 62, 1094–1156.
    36. R-Development_Core_Team (2009). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing.
    37. Salih, M. M., Erno, H. & Simonsen, V. (1983). Electrophoretic analysis of isoenzymes of mycoplasma species. Acta Vet Scand 24, 14–33.
    38. Shiel, M. J., Coloe, P. J., Worotniuk, B. & Burgess, G. W. (1982). Polyarthritis in a calf associated with group 7 mycoplasma infection. Aust Vet J 59, 192–193.
    39. Simmons, G. C. & Johnston, L. A. Y. (1963). Arthritis in calves caused by Mycoplasma sp. Aust Vet J 39, 11–14.
    40. Sirand-Pugnet, P., Lartigue, C., Marenda, M., Jacob, D., Barré, A., Barbe, V., Schenowitz, C., Mangenot, S., Couloux, A. & other authors (2007). Being pathogenic, plastic, and sexual while living with a nearly minimal bacterial genome. PLoS Genet 3, e75
    41. Thiaucourt, F. & Bölske, G. (1996). Contagious caprine pleuropneumonia and other pulmonary mycoplasmoses of sheep and goats. Rev Sci Tech 15, 1397–1414.
    42. Thiaucourt, F., Lorenzon, S., David, A. & Breard, A. (2000). Phylogeny of the Mycoplasma mycoides cluster as shown by sequencing of a putative membrane protein gene. Vet Microbiol 72, 251–268.
    43. Thomas, A., Ball, H., Dizier, I., Trolin, A., Bell, C., Mainil, J. & Linden, A. (2002). Isolation of Mycoplasma species from the lower respiratory tract of healthy cattle and cattle with respiratory disease in Belgium. Vet Rec 151, 472–476.
    44. Vilei, E. M. & Frey, J. (2001). Genetic and biochemical characterization of glycerol uptake in Mycoplasma mycoides subsp. mycoides SC: its impact on H2O2 production and virulence. Clin Diagn Lab Immunol 8, 85–92.
    45. Vilei, E. M., Korczak, B. M. & Frey, J. (2006). Mycoplasma mycoides subsp. capri and Mycoplasma mycoides subsp. mycoides LC can be grouped into a single subspecies. Vet Res 37, 779–790.