Whole Surface Image of Mycoplasma mobile, Suggested by Protein Identification and Immunofluorescence Microscopy

Abstract

Mycoplasma mobile, a freshwater fish pathogen featured with robust gliding motility, binds to the surface of the gill, where it then colonizes. Here, to obtain a whole image of its cell surface, we identified the proteins exposed on the surface using the following methods. (i) The cell surface was labeled with sulfosuccinimidyl-6-(biotinamido) hexanoate and recovered by an avidin column. (ii) The cells were subjected to phase partitioning using Triton X-114, and the hydrophobic proteins were recovered. (iii) The membrane fraction was analyzed by two-dimensional gel electrophoresis. These recovered proteins were subjected to peptide mass fingerprinting, and a final list of 36 expressed surface proteins was established. The ratio of identified proteins to whole surface proteins was estimated through two-dimensional gel electrophoresis of the membrane fraction. The localization of three newly found proteins, Mvsps C, E, and F, has been clarified by immunofluorescence microscopy. Integrating all information, a whole image of the cell surface showed that the proteins for gliding that were localized at the base of the protrusion of flask-shaped M. mobile account for more than 12% of all surface proteins and that Mvsps, surface variants that were localized at both parts other than the neck, account for 49% of all surface proteins.

INTRODUCTION

Mycoplasmas are Gram-positive related bacteria of the class Mollicutes, all of whose members lack a peptidoglycan layer (31). They are among the smallest microorganisms and are parasitic or commensal to higher organisms; many have been identified as pathogens of humans, animals, or plants. To survive in nature, mycoplasmas infect the host by binding to host cells, vary their own surface in order to evade the host immune system (8), and transport nutrients and ions (31). These crucial activities depend on the functions of surface-exposed proteins. Many studies focused on specific mycoplasmal surface proteins, including Gli521, Gli349, and MvspI of Mycoplasma mobile (1, 28), P1 adhesin of Mycoplasma pneumoniae (25), and Mycoplasma arthritidis-derived mitogen (MAM) (20). On the other hand, many proteomic studies of mycoplasma species have been performed (5, 6, 10, 11, 18, 21). However, “whole surface images,” based on the integration of protein identification, quantification, and localization, have not been produced.

Among all mycoplasma species, M. mobile, a freshwater fish pathogen (17, 36–38), is one of the best studied for its whole surface image. Its cell surface is differentiated into three parts: head, neck, and body (19). The proteins involved in gliding and adhesion, designated Gli123, Gli349, and Gli521, are localized at the neck (19, 26, 27, 34, 39, 40). Sixteen Mvsps, named MvspA through MvspP, are thought to be related to surface variation (16), of which four have been known to localize specifically at the head and body rather than the neck (19, 42). Recently, we found that MvspI, the largest one, disappears quickly and reversibly through the binding of an antibody to the distal end of molecule sticking out from the cell membrane, in a novel mechanism, designated as “mycoplasmal antigen modulation” (2, 42).

The M. mobile genome consists of 777,079 bp coding 635 protein coding sequences (CDSs), of which 557 (88%) have been validated to express proteins by proteogenomic mapping (16). The functions of the predicted proteins were analyzed by a COG (clusters of orthologous group) category match, where 45 proteins (7.1%) belong to a COG category consisting of a cell envelope, biogenesis, and an outer membrane, and 10 proteins (1.6%) belong to a COG category consisting of cell motility and secretion (16). However, the expression, abundance, and subcellular localization have not been examined experimentally. In the present study, we isolated the fractions of surface proteins, identified the surface proteins, determined the abundance ratio and subcellular localization, and then suggested a whole surface image of the cell by integrating all of the obtained information.

MATERIALS AND METHODS

Strains, culture conditions, and antibodies.

M. mobile strain 163K (ATCC 43663) was grown at 25°C in Aluotto medium (3, 24). Cells were cultured to reach an optimal density at 600 nm of 0.07. Monoclonal antibodies were raised against the whole cell surface of M. mobile, isolated previously, and screened for the reactivity to surface proteins in the present study (19, 34).

Biotinylation of cell surface proteins.

The cells were biotinylated by sulfosuccinimidyl-6-(biotinamido) hexanoate (Sulfo-NHS-LC-Biotin; Thermo Scientific, Waltham, MA) (14, 33). One liter of cultured cells was centrifuged at 12,000 × g for 10 min at 4°C and then washed three times with ice-chilled PBS-1 medium consisting of 75 mM sodium phosphate (pH 7.3) and 68 mM NaCl, suspended in a 10-ml solution of Sulfo-NHS-LC-Biotin in PBS-1, and kept for 30 min on ice. To quench the reaction and remove the excess biotin reagent, the cells were washed three times with 0.1 mM glycine (pH 6.1).

Identification of surface proteins.

The biotinylated surface proteins were isolated and identified, referring to the procedure of a previous study (33). The cell membrane fraction was isolated through osmotic lysis as described previously (30), with slight modifications. Cells from 1 liter of culture were labeled as described above when necessary and suspended in 1 ml of 4 M glycerol containing 0.5 mM phenylmethylsulfonyl fluoride, mixed with 1 ml of 5 M NaCl, and incubated at 37°C for 10 min. The cell suspension was dispersed rapidly from a syringe into 100 ml of water kept at 37°C and then incubated for 15 min. The cell membranes were collected by centrifugation at 34,000 × g for 30 min and washed with PBS-2 medium consisting of 100 mM sodium phosphate and 150 mM NaCl (pH 7.0). The cell membranes suspended in 10 ml of PBS-2 were solubilized with 2% Zwittergent 3-14 for 1 h at 4°C with head-over-head mixing. The insoluble fraction was removed by centrifugation at 100,000 × g and 4°C for 1 h. The soluble fraction was diluted to a 1-mg/ml protein concentration using PBS-2 and subjected to 1 ml of avidin-agarose matrix (Thermo Scientific) equilibrated in PBS-2. The matrix was washed 10 times with 1 ml of PBS-2, and the trapped proteins were eluted using glycine-HCl (pH 2.7) and concentrated using Centrifugal Ultracel-3K filters (Millipore, Bedford, MA). Separately, the hydrophobic proteins were collected by Triton X-114 phase partitioning as described previously (4). To detect biotin labeling, the proteins developed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) were transferred onto a nylon sheet, treated by streptavidin conjugated with horseradish peroxidase (HRP; GE Healthcare, Milwaukee, WI), and visualized by chemiluminescence. For peptide mass fingerprinting (PMF), the protein bands or spots were excised manually and treated as described previously (25, 26, 35).

Quantification of surface protein.

Two-dimensional gel electrophoresis was performed as described previously (29). Briefly, the cell membrane was lysed by a lysis buffer. The insoluble fraction was removed by centrifugation at 100,000 × g and 4°C for 30 min, and a 20-μl soluble fraction containing 30 μg of protein was subjected to an isoelectric gel and then to SDS-PAGE. To quantify the protein amounts, the gel was stained by Coomassie brilliant blue (CBB) staining, scanned by a transparent scanner (GT9800F; Epson, Nagano, Japan), and analyzed by ImageJ 1.37v (http://rsb.info.nih.gov/ij/) (2).

Microscopy.

Mycoplasma cells were bound to glass, fixed chemically, stained, and observed as previously described (19, 34, 39, 42) using 3.3 pM antibodies or 10 μg of streptavidin/ml conjugated with Cy3 (GE Healthcare). The cell fractions were observed as they were for the fixed cells on glass slides.

Sequence analysis.

The transmembrane segment, signal peptide, hydropathicity, and lipoprotein were predicted by SMART 6 (http://smart.embl-heidelberg.de/), SignalP 3.0 Server (http://www.cbs.dtu.dk/service/SignalP/), ProtParam (http://web.expasy.org/protparam/), and DOLOP (http://www.mrc-lmb.cam.ac.uk/genomes/dolop/), respectively.

RESULTS

Labeling cell surface proteins.

To identify cell surface proteins systematically, we labeled the surface proteins by using Sulfo-NHS-LC-Biotin, a hydrophobic biotinylation reagent (33). Whole cells suspended in PBS were treated with various concentrations of Sulfo-NHS-LC-Biotin. After cell lysis, proteins were separated by SDS-PAGE and blotted onto a nylon sheet, and biotinylated proteins were detected using streptavidin-HRP (Fig. 1A). The proteins were biotinylated according to the concentration of Sulfo-NHS-LC-Biotin used. The relative intensities of signals among proteins did not agree with the CBB-stained band intensities, suggesting that the proteins exposed on the cell surface (surface proteins) were labeled preferentially. To visualize the biotinylation of surface proteins, the labeled cells were bound to glass, fixed chemically, and stained by streptavidin-Cy3 (Fig. 1B). The results showed that the surface proteins were biotinylated uniformly and that biotinylation had no significant influence on the glass binding of cells, suggesting that the biotinylation occurred uniformly on the surface protein without obvious cell damage. Therefore, we expected that the surface proteins can be identified systematically by using this reagent.

Fig 1.

Biotinylation of cell surface proteins. (A) Cells were biotinylated by various concentrations of Sulfo-NHS-LC-Biotin indicated in millimolar concentrations above the panels, proteins were separated by SDS–10% PAGE (lane “P”), and biotin residues were revealed after blotting by streptavidin-HRP (lanes “B”). The profile plots of three lanes are shown in the right panel. (B) Fluorescence microscopy of biotinylated cells stained with streptavidin-Cy3. The concentrations of Sulfo-NHS-LC-Biotin used are indicated at the top in millimolar concentrations. Bar, 5 μm.

Identification of biotinylated proteins from membrane fraction.

The surface proteins were biotinylated and fractionated through the membrane fraction by procedure I shown in Fig. 2A. The cells were biotinylated, osmotically lysed, and separated into membrane and soluble fractions by centrifugation. The membrane isolation process was examined microscopically (Fig. 2B). The images of membranes were less dense than those of intact cells, and no structure was found in the soluble fraction, suggesting that the cells were fractionated successfully. The isolated cell membranes were then solubilized using 2% Zwittergent 3-14. The soluble fraction was separated from the insoluble fraction by centrifugation and subjected to an avidin column. The trapped proteins were eluted by glycine-HCl (pH 2.7), named the “B-membrane” fraction, and developed by SDS-PAGE (Fig. 3A). Thirty-six bands were analyzed by PMF, i.e., excised, and digested by trypsin, and the digests were subjected to mass spectrometry. Thirty-one bands were identified for their CDSs, but five bands could not be identified. Since the unidentified proteins are thought to have small molecular sizes based on the results of SDS-PAGE, they should be difficult to identify by PMF, because of fewer features in the mass of their digests. Seven bands were shown to contain the following pairs of proteins coded by MMOBs (CDSs of M. mobile): 3300-0260, 3670-3290, 0360-0150, 2080-6090, 5840-6080, 6080-2610, and 6070-3360, and 24 proteins were identified as isolated bands. These results show that the proteins were well isolated by the present procedure, probably because of small number of CDSs on the M. mobile genome. Finally, 36 proteins were identified (see Table S1 in the supplemental material), including proteins that were previously reported as surface proteins and the components of a cytoskeletal “jellyfish” structure (19, 26). To examine whether or not proteins inside the cell are biotinylated and contaminated the B-membrane fraction, B-soluble fraction were separated by SDS-PAGE and blotted onto a nylon sheet, and biotinylated proteins were detected by using streptavidin-HRP. Three bands, each marked by a dot in Fig. 3A, were found to be biotinylated with less efficiencies, ranging from 17 to 38% of that of surface proteins. These results suggest that some cytoplasmic proteins are also biotinylated and that the soluble proteins may contaminate the B-membrane fraction. To determine which proteins contaminated the B-membrane, the soluble fraction from the biotinylated cells was subjected to the avidin column, recovered as a “B-soluble” fraction, and separated by SDS-PAGE (Fig. 3A). Twenty protein bands were found at the running position corresponding to those identified in the B-membrane fraction and were analyzed by PMF. Two bands contained the following pairs of proteins coded by MMOBs: 0730-1010 and 5840-4220. Two bands were derived from the identical CDS, MMOB3210. Finally, 18 proteins were identified (see Table S2 in the supplemental material). The proteins coded by MMOBs 2810, 2680, 1130, 4530, 2240, 5840, 2610, and 5830 were also found in the B-membrane fraction (Fig. 3A), and these proteins are related to protein synthesis, phosphorylation, glycolysis, and protease, suggesting that these eight soluble proteins contaminated the B-membrane fraction.

Fig 2.

Fractionation of surface proteins. (A) Schematic of fractionation procedures. The labeled cells were osmotically lysed and fractionated (I) or applied to phase partitioning (II). In the osmotic lysis procedure (I), the membrane fraction was isolated and solubilized by a detergent, Zwittergent 3-14, and the resulting soluble fraction was subjected to avidin column. The soluble protein fraction was also subjected to the avidin column. The protein fractions isolated from the membrane of biotinylated cells and the soluble fractions were named B-membrane and B-soluble, respectively. In the phase partitioning procedure (II), the labeled cells were lysed by a detergent, and the soluble fraction was fractionated into hydrophobic and hydrophilic fractions by the phase partitioning method using Triton X-114. (B) Phase-contrast microscopy of cell fractions. Cell, membrane, and soluble fractions marked by an asterisk in panel A were observed. Fractions marked by an asterisk or a dagger were analyzed by SDS-PAGE in Fig. 3. Bar, 5 μm.

Fig 3.

Identification of surface proteins. The fractions obtained as shown in Fig. 2A were subjected to SDS-PAGE of the three polyacrylamide concentrations indicated at the bottom. The protein bands identified by PMF are marked by MMOB numbers. The size standards are indicated to the right of the gel images with solid triangles. The fractions are shown on the top. (A) Protein fractions obtained through procedure (I). The proteins in the B-membrane and B-soluble fractions were identified as shown in Tables S1 and S2 in the supplemental material, respectively. The proteins commonly found in both B-membrane and B-soluble fractions are marked by open triangles. The bands marked by “X” could not be identified. The bands marked by dots are referred to in the text. (B) Protein fractions obtained through procedure (II). The hydrophobic fraction was analyzed by SDS-PAGE (lanes P) and then blotted by streptavidin-HRP (lanes B). The biotinylated proteins in the hydrophobic fraction were identified by PMF.

The profiles of the isolated membrane were similar to those of the cell lysate in 6 and 10% gels, but not in 15% gel, suggesting that most of the major high-molecular-weight proteins are exposed to the surface. In the 15% gel, some bands in the cell lysate were found in neither the membrane fraction nor the soluble fraction, suggesting that such small proteins are sensitive to the protease activity of lysed cells.

Identification of hydrophobic proteins on cell surface.

Hydrophobic proteins on the cell surface might be occluded in micelles when the membrane fraction was solubilized by Zwittergent 3-14, and might fail to be trapped by the avidin column. To address this concern, biotinylated cells were fractionated through the hydrophobic fraction by procedure II as shown in Fig. 2A, and the fraction of hydrophobic proteins was recovered. Phase partitioning is a method for fractionating all proteins from cells into TX-114, aqueous, and insoluble layers through the phase transition of Triton X-114 by shifting the temperature (4). In the present study, to identify surface proteins among the hydrophobic proteins, the isolated protein fractions were subjected to blotting and examined for biotinylation (Fig. 3B). Avidin blotting revealed 15 bands, of which 2 were not detected by CBB; thus, 13 bands were excised and analyzed by PMF. Two bands were derived from the identical CDS, MMOB5430, and one band contained proteins coded by MMOBs 5430 and 5080. Finally, 13 proteins were identified, including a protein coded by MMOB 0540, which was also found in the B-soluble fraction, along with 6 proteins coded by MMOBs—4250, 0910, 3370, 0150, 6030, and 3330—which were also found in the B-membrane fraction. Finally, six proteins coded by MMOBs–3280, 3940, 5740, 5430, 5090, and 5080–were newly found as surface proteins (see Table S3 in the supplemental material). These proteins might fail to bind to the avidin column in the fractionation of the B-membrane, because they might have been occluded in micelles.

List of surface proteins.

In the present study, the surface proteins were identified as the proteins found in the B-membrane (see Table S1 in the supplemental material) and hydrophobic fractions (Table S3), excluding proteins found in the B-soluble fraction (see Table S2 in the supplemental material). Then, 34 proteins were identified as surface proteins (Table 1). Based on the features of amino acid sequences, the surface proteins were classified into four categories: Mvsps, components of gliding machinery, transporters, and others, consisting of 11, 3, 4, and 16 proteins, respectively.

Table 1.

Surface proteins^a

^aProteins are ordered according to monoisotopic mass calculated from the whole CDS sequence. Proteins identified from fractions other than B-membrane are shaded.

^bThe normalized probability that a protein in a database is the protein being analyzed based on data, experimental conditions, and other background information.

^cSequence ratio of peptides detected in the PMF relative to the whole CDS.

^dThe grand average of hydropathicity (GRAVY) of the linear polypeptide sequence is calculated as the sum of hydropathy values of all amino acids divided by the number of residues in the sequence.

^eTM, transmembrane.

^fFull length of amino acid sequence including the processed part.

^gProtein predicted to be a lipoprotein is marked by L.

^hSubcellular localization was clarified in previous papers (19, 42), and this study, and proteins identified in the jellyfish structure in a previous paper (26) are indicated as “jellyfish.”

ⁱOrder of mass amount in all surface proteins.

^jMass ratio in all surface proteins.

^kThe origin of identification indicated by the number of tables listing the protein.

^l—, Not determined.

Whole image of M. mobile cell surface.

To estimate the proportion of surface proteins that we identified, we adopted quantifying the CBB-stained signals of membrane fraction detected in SDS-PAGE. Since many proteins have similar migrating speeds in SDS-PAGE, here we subjected the membrane fraction to two-dimensional electrophoresis to obtain isolated signals (Fig. 4). Then, the resulting 66 spots were subjected to PMF. Thirteen CDSs were identified from more than two spots. For example, five spots were from MMOB4530 underlined in Fig. 4, and three of them had similar migrating speeds in SDS-PAGE, whereas the others had slower ones. Finally, 45 proteins were identified, including 18 and 7 identified as surface and soluble proteins, respectively, in Fig. 3 (see Table S4 in the supplemental material). Of the newly identified proteins, 18 were considered intracellular proteins because their annotations are involved in cytoplasmic metabolism or protein synthesis, suggesting that these proteins are so abundant that they contaminate the membrane fraction. Two of the newly identified proteins, MvspB and spermidine/putrescine ABC transporter (PotD), were considered to be exposed to the surface, by virtue of their functions–that is, as a surface variant and a polyamine binding protein, respectively (Table 1). These two proteins could not be found in either the B-membrane or hydrophobic fraction, suggesting that they may be particularly difficult to elute from the avidin column in the process of protein purification. Then, these proteins were added to the list of surface proteins and a final list of 36 expressed surface proteins was established (Table 1). Sixteen proteins identified as surface proteins in Fig. 3 could not be found by two-dimensional electrophoresis, suggesting that those proteins are expressed at lower levels and cannot be identified without purification. The relative masses of 20 surface proteins were estimated through spot intensity (Fig. 4), and the ratios of four categories—Mvsps, components of gliding machinery, transporters, and others—were shown to be 49, 12, 12, and 27%, respectively (Fig. 5A).

Fig 4.

Two-dimensional gel electrophoresis of membrane fraction. The membrane fraction was subjected to isoelectric electrophoresis, with pIs ranging from 3 to 10.5, and then to SDS–10% PAGE (upper) and SDS–15% PAGE (lower). The gel areas with pIs ranging from 4.6 to 10.5 are shown. The protein spots identified as surface proteins in Fig. 3 are marked by solid triangles, and those found in both B-membrane and B-soluble fractions as shown in Fig. 3A are marked by open triangles. The proteins found only in the B-soluble fraction as shown in Fig. 3 and the newly found proteins are marked by gray triangles. Separate spots derived from the identical CDS are numbered i to v. The spots with underlined codes are referred to in the text.

Fig 5.

Ratio and distribution of proteins on the cell surface. (A) Mass ratio of four categories of surface proteins. The identified proteins were classified into four groups, based on their amino acid sequences, as either Mvsps, gliding proteins, transporters, or others, and the mass ratios of these protein groups were estimated from two-dimensional gel electrophoresis as shown in Fig. 4. (B) Localization of surface proteins visualized by immunofluorescence. The cells were bound to glass, fixed, and stained by a monoclonal antibody. Proteins detected by antibody are shown on the left, and enlarged images are shown on the rightmost panel. The antibodies recognize the common structures between Mvsps E and F and between Mvsps N and O in the panels marked “Mvsps E&F” and “Mvsps N&O,” respectively. Bar, 5 μm. (C) Distribution of surface proteins on the cell surface. The proteins on the surface are indicated by the circles colored according to panel A. The direction of gliding is indicated by an arrow. Protein localization shown by antibodies is presented.

Subcellular localization of surface proteins.

To examine the subcellular localization of identified proteins on the surface, we performed immunofluorescence microscopy using previously isolated monoclonal antibodies. To identify the antibodies targeted to the surface proteins, 27 antibodies raised to whole M. mobile cells were subjected to Western blotting against the B-membrane and hydrophobic fractions. Finally, two antibodies were shown to recognize MvspC and both Mvsps E and F, respectively. Since Mvsps E and F have 90.4% identity in their amino acid sequences, the antibody should recognize the common structure. Immunofluorescence microscopy using these new antibodies showed that MvspC is localized mostly at the head and slightly at the body, while Mvsps E and F are localized at the head of the cell (Fig. 5B). We integrated the ratio and subcellular localization of surface proteins and then proposed a whole image of the cell surface (Fig. 5C). M. mobile forms a “head,” which is a membrane protrusion at a cell pole, a “neck” at the base of the protrusion, and a “body” in the other part. The gliding proteins, Gli123, Gli349, and Gli521 are localized at the neck. Mvsps E, F, N, and O are localized at the head, MvspK at the body, and Mvsps C and I at the head and body, respectively; these Mvsps account for 49% of all surface proteins.

DISCUSSION

In the present study, we obtained a whole surface image of M. mobile by identifying, quantifying, and localizing surface proteins, based on the merits of small cell size and reduced genome (16, 23).

Surface proteins and abundance ratio.

Thirty-six surface proteins were identified, including 10 lipoproteins and 23 proteins featuring TM segments. Five proteins were previously identified as components of a cytoskeletal “jellyfish” structure, a finding consistent with the assumption that such a structure supports the gliding machinery under the cell membrane (26). Two proteins coded by MMOBs 2080, and 3670 did not belong to any of the above categories (see Table S3 in the supplemental material), but these proteins were biotinylated preferentially, suggesting that they are localized near the cell membrane. Figure 3 and Table S1 in the supplemental material showed that the intracellular proteins can be labeled partially by Sulfo-NHS-LC-Biotin, suggesting that small fraction of the reagent can penetrate the cell membrane. MMOB 2080 codes the beta-subunit of F1 ATP synthetase plausibly functioning on the membrane (41), and MMOB 3670 codes EF-G, which may be involved in the synthesis of membrane proteins (13), suggesting that the reagent penetrated the lipid bilayer and labeled the membrane proteins facing to the cell inside. According to the abundance ratio of surface proteins obtained here, Mvsps account for 49% of all surface proteins. In general, proteins with a molecular mass over 200 kDa reach their own pI position less efficiently than lighter proteins in two-dimensional gel electrophoresis, resulting in the underestimation of the protein amount (32). Thus, the protein amount of MvspI, with a molecular mass of 221 kDa (2, 19, 42), may be underestimated, which would suggest that Mvsps account for more than 49% of all surface proteins. This fact may suggest that surface variation is critical for mycoplasma survival, because the half of limited surface space is assigned to the surface variation. In previous studies, to obtain antibodies reactive to the surface structure of M. mobile cells, our group immunized mice by intact M. mobile cells and obtained a total of 12 antibodies reactive to the cell surface, and 3 and 8 antibodies were concluded to react to gliding proteins and Mvsps, respectively (19, 34). These observations are consistent with the high abundance ratio of surface proteins obtained in the present study.

In previous studies, we succeeded in the identification of Gli521, Gli349, and Gli123 proteins involved in the gliding machinery through the SDS-PAGE of cell lysate because these proteins are abundant (34, 39, 40). However, we could identify only Gli521 in the two-dimensional gel electrophoresis in the present study (Fig. 4). The features of gliding proteins such as flexible and filamentous for Gli349 may affect isoelectric focusing of these proteins in the first dimension of two-dimensional electrophoresis (1, 22). Therefore, the actual proportion of gliding proteins may be more than the estimation suggested in the present study (Fig. 5A).

The genomes of Mollicutes commonly feature reduced sizes, ranging from 580 to 964 kb for Mycoplasma mobile (16), Mycoplasma pulmonis (7), Mycoplasma genitalium (12), Mycoplasma pneumoniae (9), and Ureaplasma urealyticum (15); high contents of envelope proteins related to lipoprotein, cytadherence, and polysaccharides, ranging from 3.1 to 7.8% of total proteins; and transport and binding proteins ranging from 3.9 to 10.6% of total proteins (7, 16). These common features may suggest that the actual whole image of Mollicute cell surfaces does not differ much from that of M. mobile, i.e., half of surface proteins may be involved in surface variation to evade the host immune system.

Subcellular localization of Mvsps.

To date, seven Mvsps have been localized on the surface, and none of them are localized at the neck (Fig. 5) (19, 42). The proteins involved in gliding should form a meshwork at the cell neck and exclude other proteins, considering the neck surface area and the amounts of gliding proteins (40). Therefore, Mvsps should be localized on the cell surface at parts other than the neck. However, the specific localization of Mvsps, i.e., specific localization at the head, body, or both head and body, cannot be explained by this scenario. Possibly, the de novo Mvsps are inserted into different sites of the cell membrane according to their signal sequences. We searched for features of sequences linked to subcellular localization but failed to find any (data not shown). More information about subcellular localization may be needed in order to determine such sequences. Recently, we reported that MvspI, localizing at the head and body, disappears quickly and reversibly through antibody binding; we named this phenomenon “mycoplasmal antigen modulation” (42). Possibly, the specific localization of Mvsps on the M. mobile surface is related to this novel mechanism of surface variation.