Deletion of the Mycoplasma genitalium MG_217 gene modifies cell gliding behaviour by altering terminal organelle curvature


  • 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
  • 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
  • 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 5811278; Fax (+34) 93 5812011.
    Search for more papers by this author

*E-mail; Tel. (+34) 93 5811278; Fax (+34) 93 5812011.


Motility is often a virulence factor of pathogenic bacteria. Although recent works have identified genes involved in gliding motility of mycoplasmas, little is known about the mechanisms governing the cell gliding behaviour. Here, we report that Mycoplasma genitalium MG217 is a novel protein involved in the gliding apparatus of this organism and it is, at least, one of the genes that are directing cells to move in narrow circles when they glide. In the absence of MG_217 gene, cells are still able to glide but they mainly move drawing erratic or wide circular paths. This change in the gliding behaviour correlates with a rearrangement in the terminal organelle disposition, suggesting that the terminal organelle operates as a guide to steer the mycoplasma cell in a specific direction. Immunogold labelling reveals that MG217 protein is located intracellular at the distal end of the terminal organelle, between the cell membrane and the terminal button. Such location is consistent with the idea that MG217 could act as a modulator of the terminal organelle curvature, allowing cells to move in specific directions.


Among the multiple forms of locomotion described in prokaryotes, gliding motility is probably one of the most striking forms. Gliding bacteria are characterized by moving smoothly on solid surfaces by means of active processes without the aid of flagella or other obvious motility organelles (McBride, 2001). Gliding motion is found in very diverse bacterial groups including myxobacteria, cyanobacteria, the Cytophaga/Flavobacterium group and some species of the genus Mycoplasma. In addition, multiple mechanisms rather than a common model have been proposed to explain such peculiar motion.

Mycoplasmas are cell wall-less prokaryotes, characterized by possessing streamlined genomes. This later feature makes these microorganisms well suited to the study of mechanisms related to gliding motility. To date, gliding motility has been reported for nine mycoplasma species including Mycoplasma mobile, Mycoplasma pulmonis and seven species from the Mycoplasma pneumoniae cluster (Hatchel and Balish, 2008). When the available genomes of these species are inspected, no homologous genes related to gliding motility in other bacteria are identified, raising the idea that mycoplasma species glide by new mechanisms different from those found in other bacteria. Miyata and co-authors found that three proteins coded in tandem on the genome (Gli349, Gli521 and Gli123) are components of the M. mobile gliding machinery (Uenoyama et al., 2004; Seto et al., 2005; Uenoyama and Miyata, 2005). Surprisingly, these proteins seem to be only conserved in M. pulmonis (Chambaud et al., 2001), suggesting the existence of alternative gliding mechanisms even among motile mycoplasmas.

Mycoplasma genitalium is the gliding bacterium (Taylor-Robinson and Bredt, 1983) with the smallest (580 kb) known genome (Fraser et al., 1995). In addition, it is a human pathogen capable of colonizing the genitourinary tract and it is considered the main causative agent of non-chlamydial, non-gonococcal urethritis in men (Horner et al., 1993). M. genitalium is closely related to M. pneumoniae and recent studies performed in both species have begun to elucidate the components of their gliding machineries (Hasselbring et al., 2006; Pich et al., 2006a; Burgos et al., 2007). Interestingly, many of these components are localized at the terminal organelle (Seto et al., 2001; Seto and Miyata, 2003), a differentiated and complex structure that polarizes and confers a flask-shape appearance to the cells. The terminal organelle is defined by an electron-dense core which is part of a bacterial cytoskeleton-like structure that remains insoluble after Triton X-100 extraction (Regula et al., 2001) and includes a distal terminal button and two parallel rods (Henderson and Jensen, 2006; Seybert et al., 2006). It is striking to find this complex structure in a bacterium with a minimized genome, but it is thought that the terminal organelle is involved in important cellular processes such as cell division and cell parasitism. Furthermore, studies performed in M. pneumoniae have provided clear evidences that the terminal organelle is itself the gliding motor (Hasselbring and Krause, 2007).

Motile bacteria have chemosensory systems to control the direction of the movement (Bourret and Stock, 2002). However, no obvious homologues to bacterial chemotaxis or two component signal transduction genes have been found in the sequenced mycoplasma genomes, although chemotactic (Kirchhoff et al., 1987) and rheotactic responses (Rosengarten et al., 1988) have been described in M. mobile. In addition, specific mycoplasma genes involved in the regulation of gliding motion are not yet identified. Regarding this point, there is only available some descriptive information about the movement patterns of mycoplasma cells. For example, a characteristic trait of some gliding mycoplasmas including Mycoplasma testudinis and M. genitalium is that they move mainly in circles (Pich et al., 2006a; Hatchel and Balish, 2008), although M. genitalium can also describe non-circular tracks, often changing the direction of the movement in an apparently erratic way (Pich et al., 2006a). Interestingly, electron microscopic analyses of cells from these species reveal the presence of terminal organelles and rods with a curved appearance (Hatchel and Balish, 2008). In the present study we utilized global transposon mutagenesis to seek genes related to the regulation of the M. genitalium gliding motility. We found that in the absence of MG217 protein, movement of M. genitalium cells is erratic or describing wide circles rather than narrow circular tracks. This different behaviour correlates with a change in the disposition of the terminal organelle in relation to the cell body. These results indicate that MG217 protein is at least one of the components implicated in the curving of the terminal organelle and provide evidences that this structure is not only the motor but is also steering the gliding direction.


Isolation of gliding enhanced mutants of M. genitalium

In a previous work, we successfully isolated gliding-deficient mutants. These mutants, when cultured with an overlay of SP-4 medium containing 0.5% low-melting-point agarose (LMPA), develop colonies with a compact morphology without satellite microcolonies formation (Pich et al., 2006a). In the current study, we aimed to isolate M. genitalium mutants showing an enhanced gliding activity. We predicted that such mutants should probably develop larger colonies and more satellite microcolonies. To improve the detection of the colonies with the predicted phenotype among the wild-type (WT) colonies, we used SP-4 medium containing a higher LMPA concentration (1%) to reduce gliding. In these conditions, WT cells developed smaller and more compact colonies with a reduced number of satellite microcolonies (Fig. 1).

Figure 1.

Colony morphology of WT and MG_217 mutants. Cells from WT strain, ΔMG_217 mutant and transposon E7 mutant were attached to cell culture dishes and covered with SP-4 medium containing 0.5% or 1% low-melting-point agarose. All pictures are shown at the same magnification.

Using these culture conditions, we screened a library of mutants generated by MTnTetM438 transposon mutagenesis (Pich et al., 2006b). Most of the 15 000 colonies examined exhibited a WT colony morphology (Fig. 1) and several colonies exhibited extremely compact morphologies (data not shown) indicating a gliding-deficient phenotype. One colony exhibited the morphology expected for and enhanced gliding mutant. This colony was larger, flat and especially thin at the periphery (Fig. 1) and was picked and propagated in SP-4 medium containing tetracycline (Tc). This mutant, designated E7, retained the particular colony morphology after three consecutive subcloning passages, indicating that it was pure and genetically stable. Accordingly, when E7 mutant was cultured in SP-4 medium containing 0.5% LMPA, it developed thin and extremely large colonies (Fig. 1). The transposon insertion point in E7 mutant was then determined by sequencing the corresponding genomic DNA with Tc upstream and Tc downstream primers. The transposon insertion site was located close to the 3′ end of the MG_217 gene (base 259028 of the M. genitalium genome), the homologous of M. pneumoniae P65-coding gene.

Construction of MG_217 null mutants by homologous recombination

Because only a single mutant was isolated in the transposition experiment and the transposon insertion site was compatible with the presence of a truncated protein comprising most of the sequence (88.4%) of the MG217 protein, we aimed to obtain a well-characterized MG_217 null mutant by homologous recombination. For this purpose, the suicidal plasmid pΔMG_217 was engineered to contain the tetM438 selectable marker (Pich et al., 2006b) enclosed by the flanking regions of the MG_217 gene (Fig. 2). A double-cross-over recombination event between plasmid pΔMG_217 and the M. genitalium genome results in the deletion from bases 60 to 980 of the MG_217 gene (82.2% of the coding region, Fig. 2B). After electroporation of M. genitalium cells with pΔMG_217 plasmid, we obtained 201 Tc-resistant colonies, representing a transformation efficiency of 2.3 × 10−7 per viable cell. To confirm the presence of the expected deletion, genomic DNAs from WT and several pΔMG_217 transformants were subjected to Southern blot analyses. Genomic DNAs were digested with HindIII and probed with complementary sequence corresponding to the 5′ flanking region of the MG_217 gene (Fig. 2D). Clone 2 exhibited a hybridization pattern compatible with a single-cross-over event through the 5′ flanking region of the MG_217 gene (Fig. 2C). The hybridization pattern for the remaining clones (designated ΔMG_217 mutants) was compatible with a double-cross-over event (Fig. 2B), demonstrating replacement of the MG_217 gene by the tetM438 selectable marker.

Figure 2.

Homologous recombination at the MG_217 gene.
A. Schematic representation of possible recombination events between the pΔMG_217 suicide vector and the MG_217 gene. Hatched boxes represent the flanking regions of MG_217 gene in suicidal plasmid pΔMG_217. In black the tetM438 selectable marker. Size of fragments resulting from the HindIII (H) digestion of DNAs from WT strain and several pΔMG_217 transformants are also indicated.
B. Double-cross-over recombination event between pΔMG_217 and MG_217 gene and gene replacement by the tetM438 selectable marker.
C. Disruption of the MG_217 gene by a single-cross-over event through the 5′ flanking region of the MG_217 gene.
D. Southern blot hybridization profiles of genomic DNA from WT strain and pΔMG_217 transformants using the 5′ flanking region of the MG_217 gene as a probe.

Protein profile characterization of MG_217 mutants

Western blot analysis using an anti-MG217 monoclonal antibody confirmed the loss of MG217 protein in E7 and ΔMG_217 mutants (Fig. 3C). MG_217 encodes a protein with a predicted mass of 44.6 kDa, but as described for M. pneumoniae, it migrates with a higher apparent molecular mass of approximately 65 kDa, probably due to its unusual amino acid composition (Proft et al., 1995). The failure to detect the possible MG217 truncated form in E7 mutant (Fig. 3C), even using an anti-MG217 polyclonal antibody (data not shown), indicates that the last 43 amino acids are essential for MG217 stability.

Figure 3.

Protein profile of WT strain and MG_217 mutants.
A. Organization of the MG_217 operon in M. genitalium. The inverted solid triangle shows the insertion of the MTnTetM438 transposon in E7 mutant and the arrow on E7 indicates that transposon insertion was found in sense, inside the MG_217 ORF.
B. Eight per cent SDS-PAGE showing the protein profile of the WT strain and MG_217 mutants.
C. Western blot analyses of the WT strain and MG_217 mutants using anti-MG218, anti-MG_217 and anti-P41 (orthologue to MG218.1) antibodies.
D. Western blot analyses of the MG_217 protein levels in the WT strain, ΔMG_312 (Burgos et al., 2007), ΔMG_317, ΔMG_218 (O.Q. Pich et al., 2008), ΔMG_191, ΔMG_192 (Burgos et al., 2006) and ΔMG_217 mutants. MG_312, MG_317 and MG_218 are the respective orthologues of M. pneumoniae HMW1, HMW3 and HMW2 genes.

As MG_217 gene is co-transcribed with MG_218, MG_218.1 and MG_219 genes (Musatovova et al., 2003), disruption or deletion in MG_217 could lead to transcriptional or translational defects in MG_217 downstream genes (Fig. 3A). To rule out this possibility, we performed Western blot analyses using anti-MG218 and anti-P41 (homologue to M. genitalium MG218.1) antibodies. We found normal levels of MG218 and MG218.1 in both E7 and ΔMG_217 mutants (Fig. 3C), indicating the lack of polar effects on the MG_217 downstream genes. Interestingly, these results suggest the existence of uncovered promoter sequences and/or cryptic ribosome binding sequences close to these genes. In addition, no pleiotropic effects were observed on additional proteins, including cytadherence- and cytoskeletal-related proteins MG386, MG312, P140 and P110 (Fig. 3B). We also examined the MG217 protein levels on several cytadherence-related mutants. As it has been previously shown for M. pneumoniae P65, MG217 protein is unstable in the absence of MG312, MG317 and MG218 proteins but not in the absence of P140 (MG191) and P110 (MG192) cytadhesins (Fig. 3D). Finally, adherence of ΔMG_217 mutant was also qualitatively assessed by their ability to bind erythrocytes. Although MG217 orthologue protein in M. pneumoniae (P65) is considered a cytadherence-associated protein, colonies from ΔMG_217 mutants were uniformly covered by erythrocytes (data not shown), indicating that MG217 is dispensable for cytadherence.

Cinematographic studies and gliding behaviour characterization

Colony morphology of ΔMG_217 mutants was indistinguishable from that exhibited by E7 transposon insertion mutant (Fig. 1), indicating that this morphology is a direct consequence of loss of MG217. To further investigate the specific implication of MG217 in this morphology, we analysed the gliding behaviour of ΔMG_217 mutants by time-lapse cinematography. Gliding velocities of ΔMG_217 mutant cells were comparable to that observed for the WT strain (Fig. 4), indicating that the expanded colony morphology observed in the MG217 mutants is not derived from an increased gliding activity. In addition, the number of non-motile cells was similar between both strains (Table 1). We also compared the gliding movement patterns of the WT and ΔMG_217 mutant cells (Fig. 5 and Table 1). As we described previously (Pich et al., 2006a), WT cells move mainly in circular tracks (74.8%) and only a little proportion of cells (14.1%) show erratic movements (defined as gliding without a defined orientation). In contrast, cells from ΔMG_217 mutant exhibited a higher proportion of cells describing erratic movements (54.2%) with the concomitant reduction of cells exhibiting circular paths (31.8%). In addition, we found that the diameters of the circular tracks of ΔMG_217 mutant were larger than those of the WT strain (Fig. 6). This result provides a feasible explanation to the expanded colony morphology observed for the ΔMG_217 mutants, as gliding cells drawing erratic tracks or wide circular tracks have a better chance for moving away from their original positions.

Figure 4.

Distribution of gliding speeds exhibited by the WT strain (grey bars) and ΔMG_217 mutant (white bars).

Table 1.  Terminal organelle and gliding movement parameters from WT and ΔMG_217 mutant cells.
StrainTrack movementSense of the circular motionDirection of curving of the terminal organelle
WT (%)74.814.111.1366439.260.8
ΔMG_217 mutant (%)31.854.214.0376337.162.9
Figure 5.

Motility tracks from cells of the WT strain and ΔMG_217 mutant. Top row images were obtained by merging 10 consecutive frames separated by 6 s each, showing the cell tracks after 1 min of observation. Bottom row images are the result of merging 20 consecutive frames separated by 6 s (2 min of observation). To visualize the direction of the track movement, the initial and final images were merged in different colour channels. The red label indicates the cell position at 0 s and the blue label the cell position at 60 s (top row) or 120 s (bottom row). With the numbers 1 and 2 are labelled cell tracks showing circular and erratic movements respectively. The arrows indicate changes on the cell gliding direction alternating left and right turns. All pictures are shown at the same magnification.

Figure 6.

Diameter of circular tracks from gliding cells of the WT strain and ΔMG_217 mutant. Diameters were measured from tracks of 164 and 68 isolated cells of WT strain and ΔMG_217 mutant respectively.

Cell morphology of the WT strain and ΔMG_217 mutants was further analysed by scanning electron microscopy (SEM). In agreement with a recent report (Hatchel and Balish, 2008), it was noticeable the existence of a tilt between the longitudinal axes of both the terminal organelle and the cell body (Fig. 7). Interestingly, although ΔMG_217 mutant cells showed normal terminal organelles, the inclination of this structure was clearly reduced (Fig. 7A–F). To obtain precise data about this finding, the tilt angle between the terminal organelle axis and the cell body axis was quantified for both strains. For the WT cells, this angle shows a normal distribution with an average value of 120°, whereas cells from ΔMG_217 mutant exhibit a distribution strongly biased for values close to 180° (Fig. 7G). These results strongly suggest a direct implication of MG217 protein in the inclination existing between the terminal organelle and the cell body. We further analysed the direction of curving (right or left) of the terminal organelle (Table 1). In both strains, this direction of curving was well correlated with the sense of the circular paths (clockwise or counterclockwise, Table 1).

Figure 7.

Analysis of the values obtained when measuring the tilt angle present between the terminal organelle axis and the cell body axis.
A–F. Scanning electron micrographs of the WT strain (A, C and D) and ΔMG_217 mutant (B, E and F). The main cell body axis and the terminal organelle axis are in black lines. The tilt angle between both axes is noted.
G. Distribution of tilt angles between the terminal organelle and the cell body axis from the WT and ΔMG_217 mutant. The tilt angle was measured in 100 isolated cells of each strain.

Cell location of MG217 protein

To determine if MG217 is cell surface-exposed or lies in the cell interior, we performed a limited proteolysis of intact M. genitalium cells using proteinase K (Fig. 8). The treated cells were then analysed by Western blot using anti-MG217 or anti-P140 antibodies as a positive control. As expected, the major cytadhesin P140 was completely cleaved after 30 min of digestion (Fig. 8C), whereas MG217 was resistant to the cleavage even after 1 h of proteinase K incubation (Fig. 8A). In contrast, when cells were treated with 0.1% Triton X-114 prior to proteinase K incubation, MG217 was completely cleaved after 5 min of digestion (Fig. 8B). These results indicate that most of MG217 is, at least, protected by the cell membrane and has an intracellular location.

Figure 8.

Partial time-dependent surface proteolysis of M. genitalium WT strain. Intact cells (A and C) or 0.1% Triton X-114-treated cells (B) were digested with proteinase K and reactions were stopped by adding 1 mM PMSF after 5, 10, 15, 30, 45 and 60 min. Protein profiles from proteolysed cells were analysed by Western blot using anti-MG_217 antibodies (A and B) and anti-P140 antibodies (C). C1 corresponds to the protein profile of untreated cells and C2 is the protein profile resulting when PMSF was added to the cells prior to proteinase K. M: apparent molecular weight pre-stained marker (Invitrogen).

The specific cellular location of MG217 protein was further investigated by cryo-immunogold labelling in thin sections of M. genitalium cells. Flask-shaped cells cut through different planes were visible in the electron micrographs (Fig. 9). No labelling was detected when control sections were treated only with the secondary antibody or when sections from ΔMG_217 mutant cells were treated with anti-MG217 monoclonal antibodies (data not shown). Immnunogold labelling of thin sections of WT cells treated with anti-MG217 monoclonal antibodies was predominantly located under the cell surface and very close to the terminal button of the electron-dense core (Fig. 9). Some gold particles were also observed in small cell sections probably representing cross-sections of the terminal organelle.

Figure 9.

Electron micrographs showing immunogold labelling of MG217 protein on ultrathin frozen sections from WT cells. Black arrows indicate the rod and white arrows show the terminal button. With two arrowheads are indicated the small cell sections probably representing cross-sections of the terminal organelle. All pictures are shown at the same magnification.


Many bacteria live in variable environments monitoring information from the surroundings. Analysis of this information followed by appropriate responses could be of major importance for survival. In pathogenic bacteria, motility is considered a significant virulence factor, contributing to the spreading of the infection and facilitating the access of the pathogen to the target cells (Jordan et al., 2007). Thus, the knowledge of how mycoplasmas control their movement can contribute to the better understanding of the infection dynamics of these microorganisms.

Previous studies performed in M. pneumoniae had shown that MG217 orthologue (P65) is a component of the terminal organelle (Jordan et al., 2001; Seto et al., 2001). However, the lack of conserved domains in the MG217-predicted sequence had made difficult to envisage the possible function of this protein. In this study, we demonstrate that MG217 is not required for cytadherence, thus providing an explanation to the failure when obtaining MG217 and P65 mutants by the screening of haemadsorption-negative mutants (Mernaugh et al., 1993; Reddy et al., 1996). On the other hand, we found that M. genitalium cells lacking MG217 exhibited morphological changes in the terminal organelle, specifically in the tilt angle between this structure and the longitudinal cell body axis. These data, taken together with the location of the MG217 at the tip structure, clearly indicate that this protein is involved in the terminal organelle curving of the M. genitalium cells.

Previous studies performed in M. pneumoniae and Mycoplasma gallisepticum, showed that its MG217 orthologues are, at least, partly surface-exposed (Proft et al., 1995; May et al., 2006). However, our limited proteolysis experiments on intact cells indicate that MG217 has an intracellular location. Although we cannot discard that a little portion or a small pool of MG217, undetectable by Western blot, could be surface-exposed, the cryo-immunogold labelling is consistent with the intracellular residence of this protein. Our first attempts to label MG217 by a standard immunogold procedure were unsuccessful (data not shown), suggesting that MG217 antigenicity was lost in the fixation procedure or perhaps this protein was not accessible to the antibody after the inclusion step. This problem was alleviated by using a cryo-immunogold procedure, with milder fixation conditions and without the resin inclusion step. On the other hand, interactions of MG217 with additional terminal organelle proteins, as the MG217 instability in the absence of MG218, MG312 and MG317 (Fig. 3D) suggests, could be also hindering the antibody accessibility. All these factors may explain the few gold particles found in our preparations. However, the label obtained was very specific and clearly shows that most of MG217 molecules are located at the most distal end of the terminal organelle, between the cell membrane and the terminal button. This location is consistent with the proposed role of the MG217 in the terminal organelle curving. In addition, such location also suggest that MG217 could be one of the constituents of the component C described by Henderson and Jensen (2006) connecting the terminal button with the inner layer of peripheral membrane proteins.

The existence of curved terminal organelles in most of the M. genitalium WT cells was recently reported by Hatchel and Balish (2008). However, our tilt angle measurements in the WT cells (120°) are somewhat different from those reported previously (160°). We do not have a definitive explanation to this divergence, but it could be derived from differences in the gliding surface used to adsorb cells for SEM analysis (glass or cell culture plastic slides). On the other hand, the small size of the terminal organelle, below the resolution of the optical microscope, precludes the in vivo monitoring of changes in the curvature of this structure, making difficult to unravel the possible significance of this particular trait. Interestingly, the analysis of MG_217 mutants has shown that changes in the curvature of the terminal organelle are accompanied with an altered gliding behaviour. In particular, we found that an increase in the tilt angle of the terminal organelle correlates with an increased frequency of erratic and wide circular trajectories (Figs 5–7). This suggests that the terminal organelle is steering the cell movement, being the gliding path determined by the tilt angle of this structure (Fig. S1). The correlation between the sense of the circular motion and the curving direction of the terminal organelle is also supporting this view (Table 1). Furthermore, several evidences such as the wide range of values (40° to 180°, Fig. 7G) obtained when measuring the tilt angle or the fact that cells can alternate left and right turns (Fig. 5) suggest that curving the terminal organelle is a complex process, and not the simple consequence of a fixed bend generated by the presence of MG217. Probably, proteins other than MG217 are also implicated in the curving of the terminal organelle as, for example, cytoskeletal elements constituting this structure. However, deletions in genes coding for terminal organelle components often lead to pleiotropic effects affecting the whole terminal organelle structure which usually produce adhesion and motile defects (Burgos et al., 2006; 2007). This may explain our failure to detect further genes related to the observed phenotype in our transposition experiments. Alternatively, some of these additional genes could be essential under laboratory growth conditions.

The existence of molecular mechanisms controlling chemotactic responses of mycoplasma cells is still unknown and needs to be investigated. However, our studies are consistent with the idea that M. genitalium cells could direct their movement by adjusting the tilt angle of the terminal organelle. Recently, it has been revealed that MG217 is a phosphorylated protein (Su et al., 2007), raising the possibility that the interpretation of environmental signals into an adequate response (i.e. redirecting the cell movement) could rely at least in part on the MG217 phosphorylation status. Although the specific role of MG217 protein in the terminal organelle curving remains to be determined, it may be possible that the phosphorylation status of MG217 may induce conformational changes that once propagated to the electron-dense core would bend the tip structure. Interestingly, Hatchel and Balish (2008) reported that the M. genitalium electron-dense core is also curved and the terminal button is found offset in the direction of this curvature. Bearing in mind that MG217 is located close to the terminal button, this observation opens the possibility that MG217 may induce a primary bend on the terminal button that afterwards is transmitted to the rod. Alternatively, we also envisage that the tip structure could be inclined by the presence of a propelling motor located in the distal end of the terminal organelle. This new mechanism may consist in a circular motion of the main adhesins P140/P110 assisted by specific proteins in the cytoskeleton below the cell membrane and the resulting movement should be perpendicular to the main axis of the tip structure (Fig. S1). The action of this rolling tip would force the lateral displacement of the terminal organelle, compelling this structure to adopt a curved appearance. In this model, MG217 protein could be involved either in the generation of the motion on this rolling tip or in transmitting its movement to the main adhesins P140 and P110. Interestingly, M. pneumoniae MG217 and P140 orthologues are cross-linked when treating cells with formaldehyde (Layh-Schmitt et al., 2000), suggesting the existence of interactions between both proteins. Furthermore, MG217 protein contains a proline-rich domain and a leucine zipper motif, both favouring protein–protein interactions.

In conclusion, this study is a significant step to the better understanding of the molecular basis of the mycoplasma cell motility, as it describes the first gene involved in the mycoplasma gliding behaviour. In addition, the phenotypic analysis of the MG_217 mutant cells has also provided valuable evidences about the role of the terminal organelle steering the movement of M. genitalium cells.

Experimental procedures

Bacterial strains and culture conditions

Mycoplasma genitalium WT strain G37 and MG_217 mutants were grown in SP-4 broth (Tully et al., 1979) at 37°C under 5% CO2 in tissue culture flasks (TPP). Tc 2 μg ml−1 was added to SP-4 medium for isolation and culture of transformants. For gliding motility assessment based on colony morphology, the obtained mutant strains were attached to cell culture dishes as previously described (Pich et al., 2006a) and covered with SP-4 medium containing either 0.5% or 1% LMPA.

Escherichia coli strain XL-1 Blue was used for plasmid amplification and it was grown as previously described (Pich et al., 2006b).

Isolation of enhanced motility mutants

To obtain a library of transposon-generated mutants, M. genitalium G37 strain was transformed by electroporation with the minitransposon MTnTetM438, as previously described (Pich et al., 2006b). Then, electroporated cells were diluted in 45 ml of SP-4 medium and aliquots of 750 μl were dispensed in cell culture dishes (Corning). After 2 h of incubation at 37°C to allow both cell adherence and expression of the Tc resistance gene, SP-4 medium was removed and the attached cells were washed once with PBS. Finally, cells were covered with SP-4 medium containing Tc and 1% LMPA. Plates were then examined after 12 days of incubation at 37°C.

Construction of pΔMG_217 plasmid and transformation

General DNA manipulations were performed by following standard procedures (Sambrook and Russell, 2001). Genomic DNA of the WT strain was isolated as previously described (Pich et al., 2006b) and used as template for PCRs. The construction of pΔMG_217 plasmid was designed as follows. A 918 bp fragment encompassing the upstream region and the first 60 bp of the MG_217 ORF was amplified by using primers 5′BEKOMG217 (5′-GTCGACGGCCTGGAGCTGCAACC) and 3′BEKOMG217 (5′-GAATTCGTTAAAAGGTTGGTTTGATGC). These primers incorporate at their ends SalI and EcoRI restriction sites (underlined) respectively. Another 976 bp fragment containing the last 139 bp of the MG_217 ORF and the corresponding downstream region was obtained by using primers 5′BDKOMG217 (5′-GGATCCTTTATTGAGAACTACATTACCC) and 3′BDKOMG217 (5′-TCTAGATTATCAACTAACTCTTGTTTGG). These primers incorporate at their ends BamHI and XbaI restriction sites (underlined) respectively. Both PCR fragments were cloned into EcoRV-digested pBE (Pich et al., 2006b), then excised with the corresponding restriction enzymes (Roche) and ligated together with a 2 kb fragment containing the tetM438 selectable marker (Pich et al., 2006b) and a SalI/XbaI-digested pBSKII (+) (Invitrogen). The tetM438 selectable marker was previously released from plasmid pMTnTetM438 (Pich et al., 2006b) by digestion with EcoRI and BamHI. Transformation of M. genitalium strain G37 with suicide plasmid pΔMG_217 was performed by electroporation as previously described (Pich et al., 2006b) and using 30 μg of DNA.

Southern blotting

Genomic DNAs from pΔMG_217 transformants were digested with HindIII and electrophoresed on 0.8% agarose gels. DNA fragments were then transferred to a nylon membrane (Roche) and probed with the sequence corresponding to the 5′ flanking region of the MG_217, previously amplified when constructing the pΔMG_217 plasmid. Probe labelling and hybridization detection were performed using the Dig DNA Labelling and detection Kit (Roche).

DNA sequencing

Sequencing with fluorescent dideoxynucleoties was performed by using the Big Dye 3.0 Terminator Kit (Applied Biosystems) and Tc upstream (5′-GGTAGTTTTTCCTGCATCAACATG) and Tc downstream (5′- CGTCGTCCAAATAGTCGGATAG) primers, following the recommendations of the manufacturer, and analysed in an ABI 3100 Genetic Analyser (Applied Biosystems).

Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western immunoblotting

Total mycoplasma cell proteins were prepared for SDS-PAGE following standard procedures. Gels were stained with Coomassie blue or transferred electrophoretically to nitrocellulose membranes and probed as previously described (Sambrook and Russell, 2001) with monoclonal antibodies anti-MG217 (Burgos et al., 2007) at a 1:500 dilution or anti-MG218 (isolated as previously described in Burgos et al., 2007) at a 1:500 dilution and anti-P41 serum (Fisseha et al., 1999) at a 1:1000 dilution.


Characterization and quantitative analyses of gliding motility of the WT strain and ΔMG_217 mutants were performed by microcinematography as previously described (Pich et al., 2006a). Gliding speed was monitored by measuring the tracks described by 50 individual cells from eight different fields for 2 min. To analyse cell movement patterns, 20 consecutive image frames captured at intervals of 6 s were merged using Adobe Photoshop CS version 8.0.1, obtaining in this way the cell motility tracks during 2 min of observation (Hatchel et al., 2006). To visualize the direction of the track movement, the initial (red) and final (blue) images were merged in different colour channels. Data of the different track movements and gliding frequencies were derived from the analysis of 220 individual cells from these merged images. The diameter of 164 and 68 circular tracks from WT cells and the ΔMG_217 mutant cells, respectively, was measured by using the Scion Image processing and analysis program. The gliding sense (clockwise or counterclockwise) from the circular tracks of approximately 100 cells was determined from both strains.

Scanning electron microscopy (SEM)

Wild-type strain and ΔMG_217 mutants were grown in cell culture Permanox chamber slides (Nunc) for 16 h. Samples were then processed as described previously (Burgos et al., 2006) and examined in a Hitachi S-570 (Tokyo, Japan) SEM. Terminal organelle tilt angle measurements were performed on 100 isolated cells in scanning micrographs by using the Scion Image processing and analysis program. The curving direction of the terminal organelle (left or right) was also determined approximately in 100 cells from both strains.

Surface proteolysis of M. genitalium

To detect possible surface-exposed regions of the MG217 protein, WT cells either intact or lysed with 0.1% Triton X-114 were digested with 80 μg ml−1 proteinase K as previously described (Burgos et al., 2007). Digestions were analysed by Western blotting using monoclonal antibodies anti-P140 (Burgos et al., 2006) and anti-MG217 (Burgos et al., 2007) at a 1:1000 and a 1:500 dilution respectively.

Immunogold labelling

Wild-type and ΔMG_217 mutant cells were grown to mid-log phase and washed twice with 0.1 M phosphate buffer pH 7.4 (PB). Cells were scrapped off and pellets were then re-suspended and fixed with 4% (v/v) formaldehyde in PB for 2 h at room temperature. Pellets were rinsed four times with PB containing 50 mM glycine, embedded in 12% (w/v) gelatine, infused with 2.1 M sucrose in PBS and frozen in liquid nitrogen. Ultrathin sections were cut with an ultracryomicrotome (Leica Ultracut UCT, Vienna) operating at −120°C. Cryosections were deposited onto Formvar Cu/Pd grids and labelled following conventional immunogold procedures. Briefly, grids were incubated for 30 min in 2% gelatine in PBS at 37°C, washed five times with PBS with 0.05 M NH4Cl and twice with PBS containing 1% BSA. Then, sections were incubated for 1 h with anti-MG217 monoclonal antibody diluted 1:5 in 1% BSA in PBS. After three washes with 0.1% BSA in PBS, the sections were incubated for 30 min at room temperature with anti-mouse IgG conjugated to 10 nm gold particles (British BioCell International, Cardiff) diluted 1:25. Grids were then washed with PBS and fixed for 5 min in PBS containing 1% glutaraldehyde, washed and contrasted with a mixture 1:9 (v/v) of 3% uranyl acetate and 2% methyl cellulose. Air-dried grids were observed in a JEOL JEM-2011 (Tokio, Japan) transmission electron microscope.


This work was supported by Grant BFU2004-06377-C02-01 and Grant BIO2007-67904-C02-01 to E.Q. R.B. acknowledges an FPU predoctoral fellowship from Ministerio de Educación y Ciencia and O.Q.P. acknowledges a predoctoral fellowship from CeRBa (Centre de Referència en Biotecnologia). We are grateful to D.C. Krause for providing P41 antisera. We also thank the staff of Servei de Microscòpia (UAB) for processing electron microscopy samples; we are especially indebted to Alejandro Sánchez-Chardi for his invaluable advice when performing transmission electron microscopy and immunogold labelling analyses.