Mycoplasma genitalium mg200 and mg386 genes are involved in gliding motility but not in cytadherence


  • Oscar Q. Pich,

    1. Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular. Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain.
    Search for more papers by this author
  • Raul Burgos,

    1. Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular. Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain.
    Search for more papers by this author
  • Mario Ferrer-Navarro,

    1. Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular. Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain.
    Search for more papers by this author
  • Enrique Querol,

    1. Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular. Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain.
    Search for more papers by this author
  • Jaume Piñol

    Corresponding author
      *E-mail; Tel. (+34) 93 581 1278; Fax (+34) 93 581 2011.
    Search for more papers by this author

*E-mail; Tel. (+34) 93 581 1278; Fax (+34) 93 581 2011.


Isolation and characterization of transposon-generated Mycoplasma genitalium gliding-deficient mutants has implicated mg200 and mg386 genes in gliding motility. The proposed role of these genes was confirmed by restoration of the gliding phenotype in deficient mutants through gene complementation with their respective mg386 or mg200 wild-type copies. mg200 and mg386 are the first reported gliding-associated mycoplasma genes not directly involved in cytadherence. Orthologues of MG200 and MG386 proteins are also found in the slow gliding mycoplasmas, Mycoplasma pneumoniae and Mycoplasma gallisepticum, suggesting the existence of a unique set of proteins involved in slow gliding motility. MG200 and MG386 proteins share common features, such as the presence of enriched in aromatic and glycine residues boxes and an acidic and proline-rich domain, suggesting that these motifs could play a significant role in gliding motility.


Mycoplasmas are genomically streamlined, wall-less microorganisms, phylogenetically related to Gram-positive bacteria and are among the most common pathogens of humans, animals and plants. Mycoplasma genitalium, a leading cause of chlamydia-negative, non-gonococcal urethritis (Tully et al., 1981; Taylor-Robinson and Horner, 2001), is the smallest, self-replicating cell known with a genome size of 580 kb (Fraser et al., 1995). With just 517 genes, it exhibits limited biosynthetic capabilities, is highly fastidious and requires complex media to grow. Nonetheless, M. genitalium is considered one of the most suitable models to achieve a complete understanding of the cellular biology of a single, replicating cell (Roberts, 2004). However, behind this apparent simplicity, M. genitalium (in fact all Mollicutes) present several features typical of eukaryotic organisms (Trachtenberg, 1998). M. genitalium cell membrane hides a complex cytoskeleton that shapes and polarizes cells. In this way, cells show a differentiated and complex tip structure, known as terminal organelle, that imparts a flask-shape appearance. The terminal organelle, extensively studied in Mycoplasma pneumoniae, is comprised of a complex network of adhesins and adherence-related proteins and it mediates mycoplasma parasitism of host target cells (Krause, 1998). M. genitalium cells have the ability to locomote across solid surfaces with the tip structure providing unidirectional polarity (Taylor-Robinson and Bredt, 1983). Interestingly, neither specialized motility structures, such as flagella or pili, nor genes homologous to other bacterial motility systems have been identified in M. genitalium (Fraser et al., 1995) or other motile mycoplasmas (Dandekar et al., 2000; Chambaud et al., 2001; Papazisi et al., 2003; Jaffe et al., 2004a). For this reason, it is thought that M. genitalium locomotes by a novel and completely unknown mechanism.

The ability for a cell to move across surfaces without the aid of obvious motility organelles is currently designated gliding motility. Gliding has been described for several eubacteria and it has been widely studied in the Gram-negative bacterium Flavobacterium johnsoniae (McBride, 2001). At least 10 genes, most of them encoding proteins with sequence homology to ABC transporters and lipoproteins, have been implicated in gliding motility (McBride, 2004). Still, the gliding mechanics of F. johnsoniae cells remain unknown. Also, gliding motility has been described in several Mycoplasma species, including M. pneumoniae, M. pulmonis, M. gallisepticum and M. mobile (Kirchhoff, 1992). In general, movement of mycoplasmas consists mainly of circles and narrow bends, interrupted by short resting periods (Radestock and Bredt, 1977). The average speed is different for each Mycoplasma species, ranging from 0.03–0.05 µm s−1 (M. gallisepticum) to 2.0–4.5 µm s−1 (M. mobile). Three coding regions involved in gliding motility of M. mobile have been recently identified and characterized (Uenoyama et al., 2004; Seto et al., 2005a; Uenoyama and Miyata, 2005a), based upon nonsense mutations in gliding mutants of M. mobile (Miyata et al., 2000). These three genes are coded tandemly in the genome of M. mobile and are conserved in M. pulmonis (Chambaud et al., 2001) but not in other gliding mycoplasmas already sequenced, suggesting the existence of alternative gliding mechanisms in mycoplasmas (Jaffe et al., 2004a). However, gliding motility proteins of M. pneumoniae (P1 and P30) and M. mobile (Gli349, Gli521 and Gli123) appear also implicated in glass binding and haemadsorption, highlighting a close relationship between gliding motility and cytadherence machineries (Uenoyama et al., 2004; Hasselbring et al., 2005; Seto et al., 2005a,b; Uenoyama and Miyata, 2005a).

The use of M. genitalium as a model to study gliding motility has several advantages. In contrast to other mycoplasmas, M. genitalium can be genetically modified, either by transposition or homologous recombination (Reddy et al., 1996; Dhandayuthapani et al., 1999). This represents a major benefit through the construction and isolation of knock-out mutants. Further, gliding motility is believed to be involved in mycoplasma pathogenesis (Kirchhoff et al., 1984) and, as stated above, M. genitalium has been linked to numerous genitourinary as well as extragenitourinary human pathologies (Baseman et al., 1988). Moreover, through comparative genomic and proteomic assessments (Himmelreich et al., 1997), M. genitalium is closely related to M. pneumoniae, the causative agent of tracheobronchitis and walking pneumonia (Krause and Taylor-Robinson, 1992). Thus, identification of genes involved in gliding motility could contribute to a better understanding of mycoplasma-mediated infections and disease progression.

Despite the genome availability of M. genitalium since 1995, there are still more than one hundred genes with unassigned roles. The experimental annotation of genes with unknown function and/or hypothetical genes is a necessary step to understand the biology of the smallest, self-replicating microorganism. In this study we describe methodology that permits the identification and isolation of MTnTetM438-generated M. genitalium gliding-deficient (GD) mutants (Pich et al., 2006). Further, we show through mutant characterization that two genes, mg200 and mg386, and their respective proteins, MG200 and MG386 directly contribute to gliding motility, providing new insights to the knowledge of mycoplasma virulence and pathogenicity.


Isolation of MTnTetM438-generated GD mutants of M. genitalium

It has been previously shown that M. mobile cells, embedded in medium containing a low concentration of agar, form non-compact colonies with undefined contours surrounded by large numbers of satellite microcolonies. This colony morphology reflects the gliding ability of M. mobile (Miyata et al., 2000), and these culture conditions have been successfully applied to isolate M. mobile gliding motility mutants. However, this culture method is not reliable when cells exhibit a lower gliding activity, as is the case of M. genitalium. We have found that non-compact colonies, surrounded by satellite microcolonies, can be observed when surface-attached M. genitalium wild-type (WT) cells are subsequently covered with SP-4 medium containing 0.5% low melting point agarose (Fig. 1A). These culture conditions have been used to isolate different GD mutants after mutagenesis with MTnTetM438 transposon. This transposon is a modified version of Tn4001 that confers tetracycline resistance and gives rise to random and stable insertions in the M. genitalium genome (Pich et al., 2006). Most colonies derived from MTnTetM438 transformed cells demonstrate the non-compact morphology described above but, among 12 000 colonies screened, 25 colonies exhibited a compact morphology. We considered the latter colonies to be GD. These colonies were subcloned three times and 18 cultures retained the compact colony morphology, suggesting that they were pure and genetically stable. These 18 strains were selected for further analyses.

Figure 1.

Colony morphology of WT and GD mutant cells previously attached to cell culture dishes and covered with SP-4 medium containing 0.5% low melting point agarose. (A) WT, (B) clone 12, (C) clone 5, (D) clone 15, (E) clone 25, (F) clone 10 and (G) clone 1. All pictures are shown at the same magnification.

Identification of transposon insertion points in GD mutants

Insertion points of MTnTetM438 were determined in the 18 GD mutants by sequencing corresponding genomic DNAs using Tc upstream and Tc downstream primers present in the MTnTetM438 transposon (Table 1). Sixteen MTnTetM438 insertions were located in the mg386 coding region while the remaining two were located in the mg200 coding region (Fig. 2). As expected, both transposon orientations were observed. Transposon insertion points identified in the mg386 coding region were randomly distributed along the coding region, although three transposon insertions were located between bases 486789–486784, suggesting the presence of a preferential insertion spot.

Table 1.  MTnTetM438 insertion point in the 18 GD mutant clones isolated.
GD mutantInsertion pointaDisrupted gene
  • a. 

    The insertion point refers to the NC 000908 GeneBank genome sequence from M. genitalium.

  • Symbol → (plus strand) or ← (minus strand) denotes the sense of MTnTetM438 insertions.

 1485097 ←mg386
 2238166 →mg200
 3486787 →mg386
 4486455 →mg386
 5487458 →mg386
 7486623 →mg386
 9486784 ←mg386
10485775 ←mg386
12237771 →mg200
13486789 →mg386
14485771 →mg386
15486716 ←mg386
16486081 ←mg386
17484239 ←mg386
22485307 ←mg386
23488029 →mg386
24486492 →mg386
25486323 →mg386
Figure 2.

Schematic representation of respective locations of MTnTetM438 insertions along coding regions of (A) mg386 M. genitalium bases 484702–489552 and (B) mg200 M. genitalium bases 237346–239151.

Identification of MG386 and MG200 polypeptides by SDS-PAGE

Most MTnTetM438 insertions were expected to disrupt either mg386 or mg200 coding regions (Hutchison et al., 1999). This was demonstrated by analysing and comparing SDS-PAGE protein profiles of GD mutants and WT strain. In mutants with insertions in the mg200 coding region (mg200 GD mutants), a 87 kDa protein band was missing (Fig. 3A). This band was excised from the WT lane, digested with trypsin and analysed by mass spectrometry. On the basis of the peptide mass fingerprint, this band was identified as MG200 (68.5 kDa) using MASCOT software (Fig. 3C). The mass observed for this protein was higher than expected, a feature that has been previously described for several proline-rich mycoplasma proteins (Ogle et al., 1992; Proft et al., 1995; 1996; Dirksen et al., 1996). In mutants with insertions in the mg386 coding region (mg386 GD mutants), a low electrophoretic mobility band was missing (Fig. 3B). This band was identified as MG386 by the method described above (Fig. 3D). In addition, a new band exhibiting a higher electrophoretic mobility was observed in each mg386 GD mutant. The mass observed for the new bands correlates well with the mass expected for the different MG386-truncated products (Fig. 3B). Moreover, the mass observed for the full-length MG386 polypeptide and for each of the truncated polypeptides was also higher than expected, suggesting common features in the MG200 and MG386 amino acid sequence that leads to abnormal electrophoretic migration of these two proteins. Finally, no pleiotropic effects derived from mg386 and mg200 disruption were found in the respective SDS-PAGE protein profiles.

Figure 3.

Identification of MG200 and MG386 polypeptides in SDS-PAGEs.
A. Lanes 1 and 2 correspond to 10% SDS-PAGE protein profile of WT strain and GD mutant clone 12 respectively. Arrow indicates band corresponding to MG200.
B. Lanes 1, 2, 3, 4, 5, 6, 7 and 8 correspond to 5, 8 or 10% SDS-PAGE protein profiles of WT and GD mutant clones 1, 22, 10, 16, 24, 13 and 5 respectively. Arrows indicate bands corresponding to MG386 full-length or truncated polypeptides.
C. MG200 MS spectrum with a MASCOT Score = 190 with 26 matching peptide masses covering 43% of protein sequence.
D. MG386 MS spectrum with a MASCOT Score = 229 with 32 matching peptide masses covering 28% of protein sequence.

Characterization of GD mutants

Colony morphology. The 18 GD mutants were further analysed for colony morphology. Colonies derived from all GD mutants developed in SP-4 medium containing 0.5% low melting point agarose were clearly smaller than those derived from the WT strain. Individual colonies derived from both mg200 GD mutants exhibited rough surfaces and granulate appearance (Fig. 1B). In contrast, colonies derived from mg386 GD mutants were flat and shiny (Fig. 1C–F), with the single exception of GD mutant clone 1 (Fig. 1G) which showed colonies less compact. The transposon insertion in GD mutant clone 1 is located near the 3′ end of the mg386 coding region, which is compatible with residual MG386 gliding activity.

Microcinematography. We further examined cell motility of several GD mutants and WT strain by microcinematography. WT cells glide individually and in close proximity to neighbouring cells, as previously described for M. mobile (Rosengarten and Kirchhoff, 1987). In addition, cells change the direction of movement when approaching other mycoplasmas in order to avoid collisions. A social communication appears to exist, because higher cell concentrations stimulate gliding activities of the entire population. Noteworthy, resting periods were present but infrequent under our study conditions. Most WT cells (72.3%) exhibited a movement consisting mainly of circles, while 25.3% of the population demonstrated erratic movement (gliding without defined orientation) and 2.4% did not move (Fig. 4D). WT cells exhibited a mean gliding speed of 0.163 µm s−1. In contrast, most of cells from mg200 GD mutant clone 12 were completely unable to move (95.2%). Motile cells of this mutant clone (4.8%) exhibited either erratic or circular movements (Fig. 4A) at a significantly reduced gliding speed of 0.013 µm s−1. The majority of cells from mg386 GD mutant clone 5 were motionless (79.3%). Among the motile cells (20.7%), there were cells showing circular or erratic motion (6.8%) as well as cells showing a trembling movement (13.9%, Fig. 4B). Cells from mg386 GD mutant clone 5 showing erratic or circular movements exhibited a mean gliding speed of 0.048 µm s−1. In contrast, the proportion of non-motile cells in mg386 GD mutant clone 1 was significantly reduced (52.7%) when compared to clone 5. Among the motile cells (47.3%), there were cells showing circular or erratic movements (20.7%) and also cells showing a trembling movement (26.6%, Fig. 4C). Cells from mg386 GD mutant clone 1 showing erratic or circular movements exhibited a mean gliding speed of 0.047 µm s−1. Interestingly, mg200 and mg386 GD mutants revealed a high degree of aggregation and many of the cells were interconnected by thin filaments (Fig. 5A), particularly abundant and prominent in mg386 GD mutants cells. Such filaments were rarely present in the WT strain (Fig. 5B). Further, we observed formation and retraction of these filaments within 100 s (Fig. 5C–F). These filaments may play a role in trembling movement described for mg386 GD mutants. For example, cells appear to be loosely attached to the culture dish by a thin filament and they spin vigorously around that fixed point, generating a Brownian-like or trembling movement. Eventually, individual cells contact and finally adhere to the plastic surface, and the trembling movement ceases.

Figure 4.

Cinematographic studies of WT and GD mutant clones. In the top row there are schematic representations showing cell tracks of GD mutant clone 12 (A), GD mutant clone 5 (B), GD mutant clone 1 (C) and WT (D). The location of each cell at different consecutive frames was marked over a 2 min interval, resulting in 60 points per cell. 1, cells showing erratic movement; 2, cells showing circular movement; those cells showing a Brownian-like trembling movement are indicated with a cross symbol (x); and 3, non-motile cells. In the bottom row there are the distributions of gliding speeds exhibited by motile cells from these clones.

Figure 5.

Individual cells from (A) GD mutant clone 5 and (B) WT strain attached to cell culture dishes and visualized by phase contrast microscopy. Panels C–F represent sequential frames spaced by 34 s, showing the retraction of a single filament. Arrows show cell-to-cell filaments, which were observed in all GD mutants isolated.

Haemadsorption activity (HA). All GD mutants were derived from WT cells, which display HA-positive and surface attachment capabilities. As each mutant colony was isolated from surface-attached transformed cells, we expected to detect a cytadherence-proficient phenotype in GD mutants. This feature was assessed by comparing the relative HA activity of representative mg200 and mg386 GD mutants with WT strain (Fig. 6). Although all GD mutant clones and WT strain demonstrated similar HA capacities, a little spot in the colony centre of mg200 and mg386 GD mutants was absent of erythrocytes. Beyond this particular feature, the HA assay confirms the cytadherence-proficient phenotype of the GD mutants.

Figure 6.

Qualitative HA assessment of (A) GD mutant clone 12, (B) clone 5, (C) WT strain and (D) class I HA negative mutant, using human erythrocytes.

Complementation assay

Reintroduction by transposition (MTnmg200Gm) of the mg200 WT coding region into GD mutant clone 12 restored the gliding-proficient phenotype. In the same way, reintroduction by transposition (MTnmg386Gm) of the mg386 WT coding region in GD mutant clones 1 and 5 corrected the gliding deficiency. In contrast, we failed to detect cell gliding when different GD mutants were electroporated in the presence of pMTnGm alone. This observation indicates that transient expression of the transposase gene did not restore gliding movement in mg200 or mg386 mutants by the re-mobilization of previously transposed MTnTetM438. We further analysed protein profiles of several clones recovered from the complementation assays by SDS-PAGE. The 87 kDa band corresponding to MG200 was restored in the protein profile of mg200-complemented GD mutant clone 12 (data not shown). Similarly, the band corresponding to full-length MG386 was restored in protein profiles of mg386-complemented GD mutant clones 1 and 5 (Fig. 7). These results confirm the involvement of mg200 and mg386 coding regions in M. genitalium gliding motility.

Figure 7.

SDS-PAGE (8%) protein profiles of WT strain and two different mg386 GD mutant clones before and after complementation with mg386 WT copy. Lane 1, WT strain; lanes 2 and 4, GD mutant clones 1 and 5 respectively; lanes 3 and 5, mg386-complemented GD mutant clones 1 and 5 respectively. Arrows indicate bands corresponding to truncated or full-length MG386 polypeptides in mg386-complemented GD mutant clones 1 and 5.

It was noteworthy that colonies from mg386-complemented GD mutant clone 5 were consistently smaller than those from mg386-complemented GD mutant clone 1 or WT strain and were detected 3 days later. One possible explanation for these differences is the transcriptional status of the mg385 coding region. As stated above, MTnTetM438 was inserted in the first half of the mg386 coding region in GD mutant clone 5, while it is located near the mg386 3′ end in GD mutant clone 1. Polar effects derived from MTnTetM438 transposition are more probable when the transposon insertion is proximal to the 5′ end of the target gene (Pich et al., 2006). This possibility was tested by RT-PCR amplification of the 861 bp long fragment that extends from the 3′ end region of the mg386 coding region to the mg385 end. We were able to amplify this fragment from WT strain and GD mutant clone 1 but not from GD mutant clones 5, 9 and 13 (Fig. 8). This result indicates that mg386 and mg385 are cotranscribed and comprise an operon and also that the absence (or reduced levels) of mg385 transcript in GD mutants 5, 9 and 13 is consistent with the location of the transposon insertion as described earlier. The impaired growth in mg386-complemented GD mutant clone 5 is in accordance with the current annotation of the M. pneumoniae MG385 orthologue (MPN566) as Glycerophosphoryl diester phosphodiesterase (Dandekar et al., 2000). This enzyme hydrolyses deacylated phospholipids to glycerol-3-phosphate and corresponding alcohols and is involved in energy production or conversion.

Figure 8.

Transcriptional analysis of mg385 gene by RT-PCR. Lanes 2, 4, 6, 8, 10 represent RT-PCRs performed for WT and GD mutant clones 1, 9, 13 and 5 respectively. Lanes 1, 3, 5, 7 and 9 are the corresponding RT-PCR negative controls.


Surface-attached mycoplasmas display a unique gliding motility and lack genes and structures already implicated in the mobility of other prokaryotes. Based upon genomic and proteomic comparisons, no common motility pathway appears to exist among the gliding mycoplasmas (Fraser et al., 1995; Dandekar et al., 2000; Chambaud et al., 2001; Papazisi et al., 2003; Jaffe et al., 2004a). Therefore, we chose to examine gliding motility in M. genitalium, the smallest self-replicating cell and mucosal pathogen, in order to gain an understanding of its biology and in vivo survival mechanisms. Further, we hoped that analysis of GD M. genitalium cells would provide us with additional insights that might uncover the genetic versatility of this minimal genetically streamlined cell.

In the current study we characterized 18 MTnTetM438 GD mutants which displayed compact colony morphology, in contrast to WT colonies that appear as non-compact colonies with satellite microcolonies. Identification of GD mutants was accomplished using SP-4 medium supplemented with 0.5% low melting point agarose. The use of treated cell culture dishes enhances mycoplasma cell motility and allows M. genitalium cells to develop non-compact colonies with microcolony satellites. A similar approach has been recently described to demonstrate involvement of P30 adhesin from M. pneumoniae in gliding motility (Hasselbring et al., 2005). Presently, mycoplasma genes associated with gliding activity have also been implicated in cytadherence (Romero-Arroyo et al., 1999; Uenoyama et al., 2004; Seto et al., 2005a,b; Uenoyama and Miyata, 2005a). However, as it has been already suggested (Hasselbring et al., 2004), proteins other than adhesins and cytadherence-associated components of the terminal attachment structure are also expected to contribute to the gliding activity of mycoplasmas. Our success in isolating GD mutants from surface-attached and HA-positive WT cells, and the retention of this phenotype in mg200 and mg386 GD mutants, clearly demonstrates the existence of an additional set of motility-related proteins.

In all GD mutants isolated, MTnTetM438 was inserted either in mg386 or mg200 coding regions. Interestingly, these two genes were not previously considered to be dispensable for M. genitalium growth under laboratory conditions (Hutchison et al., 1999; Glass et al., 2006), further confirming our suggestion (Pich et al., 2006) that the list of nonessential genes of M. genitalium previously reported is incomplete. Direct observation of individual cells from specific mg386 and mg200 GD mutants confirmed the GD phenotype. However, the degree of gliding deficiencies between both mutants could be distinguished based upon colony shape and texture, erratic and trembling movement, and per cent of immobile cell populations. These differences suggest distinct roles of MG386 or MG200 in the M. genitalium gliding motility mechanism. Involvement of mg200 and mg386 genes in M. genitalium gliding motility was clearly demonstrated by restoring the gliding-proficient phenotype after the introduction of the respective WT copies into selected GD mutants.

MG200 exhibits weak homology to DnaJ/CbpA near the C-terminal region, and possesses a J-like domain in the N-terminal region, which is characteristic of chaperone-related functions. Two other coding regions of M. genitalium, mg019 and mg002, encode proteins showing a certain degree of identity to DnaJ (32% in a 366-amino-acid overlap and 44% in a 61-amino-acid overlap respectively). Although we cannot discard the involvement of MG200 in the proper folding of mycoplasma proteins, its participation in M. genitalium gliding motility seems clear. For this reason, we propose to annotate mg200 gene as gmpA, where gmp is the acronym for gliding motility protein. The mg386 coding region encodes the orthologue of the P200 protein of M. pneumoniae that is present in the Triton X-100 insoluble fraction (Proft et al., 1996). The direct involvement of mg386 in M. genitalium locomotion has been demonstrated in the current study, allowing us to annotate this gene as gmpB. Based on bidimensional electrophoretic studies, it has been suggested that levels of MG386 significantly increase when M. genitalium cells reach stationary phase (Wasinger et al., 2000). This is consistent with a stress response to nutrient depletion and the search for a more optimal environment through gliding motion. As mg386 and mg385 comprise an operon, expression of MG385 could be upregulated in a similar way, providing energy necessary for gliding motility (Jaffe et al., 2004b; Uenoyama and Miyata, 2005b). However, the polypeptide identified in the work of Wasinger et al. (2000) do not correspond to the full-length mg386 product and consequently, upregulation of MG386 in the stationary phase remains to be confirmed.

It is noteworthy that MG200 and MG386 orthologues are present in M. pneumoniae (MPN119 and MPN567 respectively) and M. gallisepticum (MGA_1228 and MGA_0205 respectively), which are classified as slow gliding mycoplasmas. In contrast, orthologues of Gli349, Gli521 or Gli123, which are designated gliding motility proteins in M. mobile, are also present in M. pulmonis. Both M. mobile and M. pulmonis are fast gliding mycoplasmas. Therefore, it appears that specific and unique sets of proteins are involved in slow versus fast gliding mycoplasmas. Moreover, the three slow gliding mycoplasmas fall in a branch of the phylogeny different from that of the two fast gliding mycoplasmas (Maniloff, 2002), suggesting that gliding motility has evolved twice independently in the mycoplasmas (Jaffe et al., 2004a). Interestingly, orthologues of MG386 and MG200 also possess enriched in aromatic and glycine residues (EAGR) boxes and acidic and proline-rich (APR) domains, a feature shared with the cytadherence-accessory protein HMW1 (Balish et al., 2001). The presence of EAGR boxes in the HMW1 protein suggests a possible role of this cytadherence-associated protein in M. genitalium gliding motility. The presence of APR domains in proteins with EAGR boxes, which seem exclusive of Mycoplasma proteins involved in gliding motility, suggests a relationship between these two motifs and gliding motility (Fig. 9).

Figure 9.

Schematic representation of the sequence alignment of MG200, MG386 and MG312 orthologues from M. genitalium, M. pneumoniae and M. gallisepticum respectively. The alignment was constructed using the macaw program (Schuler et al., 1991) and manually edited to highlight EAGR boxes. A vertical black line in the alignments denotes the presence of conserved amino acids (Blosum-62 matrix).

Acidic and proline-rich domains are a common feature of several mycoplasma cytoskeletal proteins (HMW1, HMW3 and P65). Proline-rich domains are present in numerous eukaryotic proteins and seem to be involved in protein–protein interactions (Kay et al., 2000). A particular set of these proteins is involved in cell motility by regulating actin polymerization of membrane protrusions (Holt and Koffer, 2001). It is noteworthy that the presence of small membrane protrusions or ‘naps’ has been described in the terminal organelle of several gliding mycoplasmas, including M. genitalium (Baseman et al., 1982; Shimizu and Miyata, 2002). Moreover, mycoplasma gliding motility is inhibited by cytochalasin B (Maniloff and Chaudhuri, 1979), a drug known to impede eukaryotic cell motility. Although the existence of actin-like proteins in M. pneumoniae has been proposed (Rosenbusch et al., 1976; Gobel et al., 1981; Mayer, 2003), such proteins remain unidentified. However, subcellular location and relative abundance of DnaK (Regula et al., 2001), a member of the actin superfamily of proteins, suggest an actin-like role of this protein in mycoplasmas. Further, the possibility that mycoplasma gliding motility shares features with eukaryotic cell movement provides the impetus to define mechanisms of mycoplasma gliding machinery at genetic and biochemical levels. Such studies could provide fundamental insights into how cells orient themselves during in vivo residence and respond to signals that trigger mobility activities.

Experimental procedures

Culture conditions and plasmids

Mycoplasma genitalium G37 strain and GD mutants were grown in SP-4 broth (Tully et al., 1979) at 37°C under 5% CO2 in tissue culture flasks (TPP). For colony formation, mycoplasma cultures were plated in SP-4 medium supplemented with 0.8% agar (Difco). To test for gliding activity of M. genitalium, cells previously attached to cell culture dishes (Corning) were covered with SP-4 broth containing 0.5% low melting point agarose (Iberlabo). To isolate Tc-or Gm-resistant strains, SP-4 medium was supplemented with Tc 2 µg ml−1 (Roche) or Gm 100 µg ml−1 (Invitrogen) respectively. Escherichia coli XL1-Blue strain was used for the amplification of the following plasmids: pBE, pMTnGm and pMTnTetM438 (Pich et al., 2006) and pMTnmg200Gm and pMTnmg386Gm as constructed for this study.

Isolation of GD mutants

Gliding-deficient mutants of M. genitalium were generated by random transposition of MTnTetM438 minitransposon. For this purpose, WT cells were electroporated in the presence of pMTnTetM438 plasmid as previously described (Pich et al., 2006). After 2 h of incubation at 37°C, cells were resuspended in 37.5 ml of SP-4 medium and aliquots of 750 µl (∼250 cfu) were dispensed in cell culture dishes. After 2 h of incubation, SP-4 medium was removed and the attached cells were washed twice with PBS ×1. Finally, cells were covered with 0.5% low melting point agarose SP-4 medium supplemented with Tc. After 12 days of incubation, colonies exhibiting a non-disperse phenotype were picked and propagated in SP-4 medium supplemented with Tc. Isolated clones that retained the GD phenotype after three consecutive passages were expanded, filtered (0.45 µm) to remove aggregates, and recultured in 0.5% low melting point agarose SP-4 medium. Resultant colonies exhibiting a non-disperse morphology were considered pure and stable and consequently selected for further analyses.

Genomic DNA isolation and sequencing reactions

Genomic DNA from GD mutants was isolated using E.Z.N.A. Bacterial DNA Kit (Omega Bio-tek). Sequencing with fluorescent dideoxynucleotides was performed by using Big Dye 3.0 Terminator Kit (Applied Biosystems) and Tc upstream (5′-GGTAGTTTTTCCTGCATCAACATG) and Tc downstream (5′-CGTCGTCCAAATAGTCGGATAG) primers, following recommendations of the manufacturer.

SDS-PAGE and MALDI-MS analyses

The separation of total mycoplasma cell proteins by SDS-PAGE followed standard procedures. Comassie-stained protein bands of interest were excised from acrylamide gels, destained and trypsin-digested using In-Gel DigestZP Kit (Millipore), according to the manufacturer. One microlitre of sample was mixed with the same volume of a solution of a-cyano-4-hydroxy-trans-cinnamic acid matrix (0.3 mg ml−1 in ethanol: acetone 6:3), and spotted onto a 600 µm AnchorChip MALDI target plate (Bruker, Bremen, Germany) and allowed to air-dry at room temperature. MALDI-mass spectra were recorded in the positive ion mode on an Ultraflex time-of-flight instrument. Ion acceleration was set to 25 kV. All mass spectra were externally calibrated using a standard peptide mixture containing angiotensin II (1046.54180), angiotensin I (1296.68480), substance P (1347.73540), bombesin (1619.82230), renin substrate (1758.93261), adrenocorticotropic hormone 1–17 (2093.08620), adrenocorticotropic hormone 18–39 (2465.19830) and somatostatin 28 (3147.47100). Spectra were also calibrated internally using the autolysis products of trypsin at m/z 842.50 and m/z 2211.10. For PMF analysis, the MASCOT search engine (Matrix Science, London, UK) was used with the following parameters: one missed cleavage permission, 50 ppm measurement tolerance. Positive identifications were accepted with P-values higher than 0.07. In the searches, methionine residues modified to methionine sulphoxide were allowed wherever necessary.


Characterization and measurement of the gliding motility of WT strain and GD mutants were performed by microcinematography. For this purpose, individual cells from different dilutions were filtered and grown in cell culture dishes with SP-4 medium. After 16 h of incubation, the medium was removed, and attached cells were overlaid with a few drops of fresh medium and a coverslip. Cell movement was examined at 37°C using an Axioplan (Zeiss) microscope, and images were captured at intervals of 2 s for a total of 5 min with a Hamamatsu ORCA-ER CCD (charge-coupled device) digital camera controlled by Aquacosmos software. Particular movements of ∼700 individual cells from each WT and mutant strain were analysed by microcinematography in order to determine the cell percentage exhibiting different gliding motilities. Gliding speed was monitored by measuring the track described for 25 individual cells from eight different microcinematography fields for 2 min. To obtain a comparable mean gliding speed from each strain studied, all samples contained similar cell numbers.

Construction of pMTnmg200Gm and pMTnmg386Gm plasmids

The mg200 WT coding region was amplified using MG200C5′ (5′-CTCGAGTAGTATTTAGAATTAATAAAGTATGGCTGAA CAGAAACGTGATTATTATG-3′) and MG200C3′ (5′-CC CGGGCTAACTAATGGGTTCTTGGGAGAGG-3) primers. MG200C5′ primer incorporates a XhoI site (underlined) and a 22 nucleotide region (in bold) previously described as a functional promoter in M. genitalium (Pich et al., 2006). MG200C3′ primer includes a SmaI site. The resulting PCR product was cloned into EcoRV-digested pBE, excised with XhoI and SmaI and included in a ligation reaction containing XhoI–SmaI-digested pMTnGm, creating pMTnmg200Gm. The mg386 WT coding region was amplified using MG386C3′ (5′GGGCCCTAGTATTTAGAATTAATAAAGTATGCCAAAAA CAACAAAGAATAAAAAC) and MG386C5′ (5′-CCCGGGC TATTTTTTATCATTACTGCCAAAAAC) primers. MG386C3′ primers incorporate an ApaI site (underlined) and the functional promoter described above. MG386C5′ primer includes a SmaI site (underlined). By the same process described above, the resulting PCR product was used to create pMTnmg386Gm.

Complementation of GD mutants

mg200 and mg386 coding regions were reintroduced by transposition in the genome of several GD mutants. In this way, GD mutant clones 1 and 5 were electroporated in the presence of pMTnmg386Gm, and GD mutant clone 12 was electroporated in the presence of pMTnmg200Gm. The three GD mutants were also electroporated in the presence of pMTnGm plasmid to determine the frequency of reversion to a WT phenotype, which could arise as a result of the mobilization of the original transposon to a new site, mediated by the transient expression of the tnp gene present in the plasmids used. Stability of the GD phenotype during the electroporation experiment was monitored by electroporating GD mutant cells without plasmid. After 2 h of incubation at 37°C, both transformations were passed through a 0.45 µm low protein binding filter, and different dilutions were plated by the procedure described above.

Haemadsorption activity

For qualitative assessment of the HA, different dilutions of WT strain, GD mutant clones 5 and 12 cells and a class I HA negative mutant (Mernaugh et al., 1993), were passed through a 0.45 µm low protein binding filter and spread onto SP-4 plates. WT strain and class I mutant were spread onto SP-4 medium and incubated at 37°C for 8 days. GD mutants were spread onto SP-4 medium and incubated for 10 days in order to obtain colonies of comparable size. Then, colonies were flooded with 2 ml of diluted (1:50) human erythrocytes, incubated at 37°C for 1 h, washed three times with PBS ×1 and observed by light microscopy (LeicaMZFLIII). Pictures were taken using a LeicaDC500 camera connected to the microscopy.

RNA manipulation

RNA from WT strain and GD mutant clones 1, 5, 9 and 13 was isolated using RNAaqueous kit (Ambion) and treated with DNAase I (New England BioLabs). cDNAs were obtained by using the SuperScript First-Strand Synthesis System Kit (Invitrogen) and mg385-5′ primer (5′-CTAATA AATTTGTGCAGTAATTTG). PCR fragments comprising the last 150 bp of the mg386 coding region and the full (711 bp) mg385 coding region were amplified by using mg385-5′ and mg385-3′ (5′-GTATTCCAGTCTGAGGAAACTG) primers. RT-PCR negative controls were performed by omitting Reversal Transcriptase enzyme in the RT reaction.


This work was supported by grant BFU2004-06377-C02-01 to E.Q. R.B. acknowledges a FPU predoctoral fellowship from Ministerio de Educación y Ciencia. O.Q. acknowledges a predoctoral fellowship from CeRBa (Centre de Referència en Biotecnologia). Many thanks are given to Dr J.B. Baseman (UTHSCSA) for their helpful comments and suggestions when revising the manuscript, and the kindly gift of class I HA negative mutant. We also thank Anna Barceló (Servei de Seqüenciació UAB) for DNA sequencing.