|Home | About | Journals | Submit | Contact Us | Français|
Mycoplasma hyorhinis (strain MCLD) was recently isolated from a melanoma cell culture. Growth of MCLD was considerably improved by 24 serial passages in a modified Hayflick's mycoplasma medium. Transmission electron microscopy showed that MCLD exhibits a polymorphic appearance, with ovoid or elongated cells frequently harboring an electron-dense core at one of the poles. Adherence of M. hyorhinis to melanoma cells followed saturation kinetics. Furthermore, although M. hyorhinis has been considered to remain attached to the surface of the host cells, we show for the first time, qualitatively by confocal laser scanning microscopy and quantitatively by a gentamicin resistance assay, that MCLD is able to invade melanoma cells. The ingested mycoplasmas were randomly distributed in the cytoplasm, tending to concentrate near the plasma membrane. Both adherence to and invasion of melanoma cells by M. hyorhinis strain MCLD were dramatically enhanced by mild proteolytic digestion with proteinase K (2.5 μg/mg cell protein for 2.5 min at 37°C) that affected the surface-exposed proteins of this organism, mainly the major 47-kDa lipoprotein. We suggest that the intracellular location of M. hyorhinis strain MCLD is a privileged niche, which may explain the survival of M. hyorhinis in tissue cultures. The enhanced binding to and invasion of melanoma cells by protease treatment may be due to either the activation or the enhanced exposure of an adhesin(s) on the mycoplasmal cell surface.
Mycoplasmas (class Mollicutes) are the smallest self-replicating bacteria. These bacteria lack a rigid cell wall and are parasites, exhibiting strict host and tissue specificities (2, 22). Many mycoplasmas are pathogenic to humans and animals and are frequent contaminants of cell cultures (24). Almost all human and animal mycoplasmas depend on adhesion to host cells for subsequent colonization and infection (20, 24). In these mycoplasmas, adhesion is the major virulence factor, and adherence-deficient mutants are avirulent (2, 20). The lack of a cell wall has forced mycoplasmas to develop sophisticated molecular mechanisms to enable their prolonged existence within the host, usually without causing major harm. Mycoplasma hyorhinis was first isolated from the respiratory tract of young pigs and has been implicated in pleuritis, peritonitis, pericarditis, arthritis, and otitis media in swine (9, 18). Interest in M. hyorhinis has recently increased after the detection of this organism in human gastric cancer tissues, suggesting a possible association between M. hyorhinis and tumorigenesis (12, 14). Another important property of M. hyorhinis is its effectiveness in contaminating cell cultures, which impinges on many aspects of biological research (16). A mycoplasma identified as M. hyorhinis has recently been identified in the melanoma cell line LB33mel A1 (to be referred to as strain MCLD) (11). Although M. hyorhinis is considered a typical extracellular microorganism that is able to adhere to epithelial cells (16, 20), ultrastructural studies performed with engulfed MCLD have also revealed mycoplasmas within melanoma cells (Feinstein, personal communication). Furthermore, elevated expression of a ligand to CD99 on the cell surface of melanoma cells infected with MCLD (11) and a marked increase in calpastatin levels within infected neuroblastoma cells (N. S. Kosower and E. Elkind, personal communications) strongly suggest that the interaction of this strain with cell cultures may upregulate genes. These results prompted us to develop an experimental system to further understand the mechanisms underlying the interaction of this organism with host melanoma cell cultures.
M. hyorhinis strain MCLD (melanoma cell line derived), which was isolated in our laboratory (11), was used throughout this study. In some experiments, Mycoplasma fermentans strain JER from our culture collection and Mycoplasma penetrans strain GTU-54-6A1 (kindly provided by S.-C. Lo, Armed Forces Institute of Pathology, Washington, DC) were also utilized. The mycoplasmas were grown for 24 to 48 h at 37°C in a modified Hayflick's medium (13). The medium for M. hyorhinis was supplemented with 10% heat-inactivated fetal calf serum whereas M. fermentans and M. penetrans were grown in medium that contained 5% heat-inactivated horse serum (Biological Industries, Beit Haemek, Israel). For metabolic labeling, the bacteria were grown in a medium containing 0.3 μCi of [9,10(n)-3H]palmitic acid, oleic acid, or thymidine (53.0 Ci/mmol; New England Nuclear) per ml. The mycoplasmas were harvested at the mid-exponential phase of growth (A595 of 0.08 to 0.12; pH 6.8) by centrifugation for 20 min at 12,000 × g, washed once, and resuspended in a buffer solution containing 0.25 M NaCl, 10 mM MgCl2, and 10 mM Tris-HCl adjusted to pH 7.5 (TN buffer). The number of viable mycoplasmas was determined by plating and was expressed as the number of CFU per milliliter. The melanoma cell line LB33mel A1 (kindly provided by O. Mandelboim, The Lautenberg Center for General and Tumor Immunology, The Hebrew University, Jerusalem, Israel), the epithelial cell line HeLa-229 and the lymphocyte cell line CD4+ Molt-3 (American Type Culture Collection, Rockville, MD) were grown in flasks containing Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal calf serum (Biological Industries), 2 mM l-glutamine, 100 U of penicillin per ml, and 100 μg of streptomycin per ml. The flasks were incubated at 37°C in a 5% CO2 atmosphere. The red blood cell types AB, A, B, and O were kindly provided by the Hadassah Hospital blood bank (Jerusalem, Israel).
Adhesion of M. hyorhinis to melanoma cells or HeLa cells was determined in a reaction mixture containing 106 eukaryotic cells and [3H]palmitate-labeled M. hyorhinis (50 μg of cell protein) in 1 ml of DMEM containing 10 mM MgCl2 (27). The adhesion mixtures (in duplicate) were incubated for up to 6 h in a 5% CO2 atmosphere. The nonadhering mycoplasmas were removed by washing the eukaryotic cells twice with 2.5 ml of phosphate-buffered saline (PBS). The washed cells were trypsinized for 15 min at 37°C and resuspended in PBS, and aliquots were transferred to scintillation vials containing 4 ml of scintillation liquor (Quicksafe A). The radioactivity was measured in a liquid scintillation counter (Kontron), and results are expressed as counts per minute (cpm). Adhesion of M. hyorhinis to Molt-3 cells or red blood cells was determined in a reaction mixture containing 106 cells and [3H]palmitate-labeled M. hyorhinis (100 μg of cell protein) in PBS. The adhesion mixtures (in duplicate) were incubated in a rotating bath. After incubation for 6 h, 100-μl samples were pipetted on the surface of 1 ml of Ficoll in a 1.5-ml plastic microcentrifuge tube and centrifuged at 12,000 × g for 2 min. Under these conditions the host cells passed through the Ficoll, forming a pellet, while the nonadhering mycoplasmas remained in the upper part of the Ficoll tube. The tips of the tubes containing the pellet (host cells and adhered mycoplasmas) were cut off, and radioactivity was counted.
The lipoprotein fraction of M. hyorhinis membranes was obtained by the Triton X-114 fractionation method (4). In brief, membranes (1 mg/ml) were incubated in 1 ml of Tris-buffered saline (TBS) containing 1% Triton X-114 at 4°C for 1 h with gentle agitation. After centrifugation at 4°C for 30 min at 12,000 × g, the supernatant containing the soluble proteins was subjected to three cycles of phase fractionation including incubation at 37°C for 5 min for micelle formation, followed by centrifugation at room temperature for 3 min at 12,000 × g for phase separation, resulting in an upper aqueous phase and a lower detergent phase. The detergent phase was collected and adjusted to 1 ml with TBS. The Triton X-114 was removed from the samples by gentle agitation of the detergent phase with beads (Bio-Beads SM-2; Bio-Rad) at 4°C for 2 h (29).
Invasion of M. hyorhinis strain MCLD was determined by a bacteriological assay (1, 27) and by the confocal laser scanning microscopy (CLSM) of immunofluorescent preparations (27). In the bacteriological assay, the viability of intracellularly grown MCLD was estimated by utilizing a gentamicin protection assay (1). The sensitivity of the mycoplasmas to gentamicin (MIC) was determined using the Etest procedure (AB Biodisk, Solna, Sweden). For immunofluorescence staining, M. hyorhinis-infected melanoma cells grown on coverslips were fixed at room temperature for 10 min with 4% paraformaldehyde (PFA) in PBS. After three washings with PBS, the cells were permeabilized by incubation for 3 min with 0.2% Triton X-100 in PBS containing 1% bovine serum albumin (PBS-BSA solution), washed twice with PBS, and blocked for 20 min with 2% horse serum. The cultures were then overlaid for 60 min at room temperature with a fluorescein-labeled donkey polyclonal anti-M. hyorhinis antiserum (NIAID, Bethesda, MD) diluted 1:200 in PBS-BSA solution. Nonbound antibody was removed by rinsing the coverslips three times with PBS. The coverslips were then mounted with a solution containing 90% glycerol, 7% PBS, and 3% 1,4-diazabicyclo-(2,2)-octane as an antifading agent. The specificity of immunostaining was evaluated by utilizing nonspecific antibodies (nonimmune donkey serum). Immunofluorescent samples were analyzed using a Zeiss 410 laser scanning confocal microscope (Zeiss, Germany) equipped with an argon ion laser tuned at 488 nm.
Total protein content was determined by the method of Bradford (5) using bovine serum albumin as the standard. Cell and membrane proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 11.5% acrylamide gels. Mass spectrometry analysis of polypeptide bands extracted from polyacrylamide gels was performed as described previously (28). Proteolysis of intact M. hyorhinis, M. penetrans, and M. fermentans was carried out by incubating the mycoplasmas (0.5 mg cell protein/ml) with 2 to 20 μg/ml of either proteinase K, papain, trypsin, chymotrypsin, or pronase E (all products of Sigma) at 37°C for 5 to 30 min. The digested mycoplasmas were washed twice with TN buffer, and their adhesion to melanoma or HeLa cells was determined. For further specific identification of the Mycoplasma species utilized, PCR was performed using the primers Myco23s (CGGTACTAGTTCACTATCGG) and Myco16sRev (GGGTGAAGTCGTAACAAGG), designed to amplify the variable spacer between the conserved 23S and 16S rRNA mycoplasma genes. The PCR product was sequenced and compared with the GenBank database.
M. hyorhinis pellets (1 mg/ml) or monolayers infected for 1 h with M. hyorhinis (multiplicity of infection [MOI] of 100) were fixed with 2% PFA and 2% glutaraldehyde (in 0.2 M cacodylate buffer, pH 7.4) for 30 min in the cold. Processing of the samples by osmification, dehydration, and embedding in Epon was performed as previously described (28). The samples were then sectioned using an LKB-3 ultramicrotome and observed with a Tecnai 12 electron microscope (Phillips, Eindhoven, The Netherlands) equipped with a MegaView II charge-coupled-device camera.
Many M. hyorhinis strains grow very poorly in conventional mycoplasma medium (16) or are even noncultivable (26). The yields obtained after 48 h of incubation were as little as 3.5 ± 1.0 mg/liter. After 24 serial passages in a medium where the horse serum component was replaced by fetal calf serum, we were able to obtain, after 48 h of incubation, cell yields of 13.0 ± 1.5 mg/liter. These yields were sufficient to analyze cell surface components and to establish a system for studying the interactions of MCLD with melanoma cells. M. hyorhinis strain MCLD was shown by transmission electron microscopy (TEM) images to have ovoid to round cells (Fig. (Fig.1A)1A) with electron-dense poles. Some of the poles were not clearly associated with the mycoplasma cells and may represent pole structures that were detached from the cells. Electron-dense structures have been described previously in other Mycoplasma species and in some cases were found to play a major role in the adhesion to host cells (2, 24). Interestingly, in a melanoma cell culture contaminated by M. hyorhinis strain MCLD, mainly elongated mycoplasmas with electron-dense regions were detected (Fig. (Fig.1B),1B), but it is difficult to assess whether the mycoplasmas use the electron-dense regions for association with the host melanoma cells. Elongated, flask-shaped forms are common in mycoplasmas such as Mycoplasma pneumoniae, Mycoplasma genitalium, Mycoplasma pulmonis, and Mycoplasma gallisepticum (15).
Mycoplasmas lack many of the biosynthetic pathways found in ordinary bacteria. The dependence on an external supply of nutrients provides a tool for the specific labeling of cellular components. None of the mycoplasmas tested so far are able to synthesize or modify long chain fatty acids, and thus they depend on an exogenous supply of fatty acids in the growth medium (23). Thus, M. hyorhinis strain MCLD was metabolically labeled with [3H]palmitate or [3H]oleate. The labeling intensity with palmitic acid was two to three times higher than the labeling intensity obtained with oleic acid. Both radioactive fatty acids were incorporated into membrane lipids, with 98% of the radioactivity recovered in the polar lipid fraction representing biosynthetically labeled lipids (data not shown).
Binding of MCLD to melanoma cells was measured as a function of time. The extent of binding was calculated from the amount of radioactivity associated with the host cells after incubation for various periods of time. Binding of MCLD to melanoma cells was observed within 1 h of incubation and increased linearly during the next 5 h (Fig. (Fig.2).2). Similar results were obtained in control experiments utilizing M. fermentans. However, whereas M. fermentans markedly adhered to Molt-3 lymphocytes, the binding of M. hyorhinis to Molt-3 was very low or nonexistent (data not shown). Neither M. hyorhinis nor M. fermentans adhered to the red blood cell preparations tested (data not shown). Metabolic labeling of MCLD by growing the bacteria in the presence of [3H]thymidine obtained similar results, excluding the possibility that the radioactivity detected in the melanoma cells is due to exchanged radiolabeled lipids between MCLD and host cells. Pretreatment of melanoma cells with 4% PFA for 15 min at 37°C reduced the binding of native M. hyorhinis by 20%. However, pretreatment of M. hyorhinis with 4% PFA reduced the binding of the mycoplasmas to native melanoma cells by 80% (Fig. (Fig.2),2), suggesting that binding depends on the full native tertiary structure of the mycoplasma's cell surface proteins. The ability to adhere was not influenced by pretreating M. hyorhinis for 15 min at 37°C with the uncoupler carbonyl cyanide m-chlorophenylhydrazone (25 μM), the ATPase inhibitor dicyclohexylcarbodiimide (20 μM), the K+ ionophore valinomycin (10 μM), or chloramphenicol (100 μg/ml). These results suggest that the proton or electrochemical gradient across the cell membrane and the protein synthesis of M. hyorhinis do not play a role in increasing the competence of the mycoplasmas for binding. To identify possible receptor molecules on the cell surface of melanoma cells, simple carbohydrates were tested for binding inhibition activity. The carbohydrates tested were d-glucose, d-fructose, d-mannose, sucrose, mannitol, lactose, rafinose, melibiose, sorbitol, ribitol, arabinose, maltose, ramnose, d-xylose, and l-xylose (all products of Sigma). None of the carbohydrates tested at a concentration of 1% had any inhibitory activity. It has previously been shown that sialic acid residues on the cell surface of host cells play a major role in the binding of M. pneumoniae to red blood cells (21). Nonetheless, the ability of MCLD to adhere to melanoma cells was neither inhibited by sialic acid-containing molecules, such as siallyllactose (0.1 mg/ml) or fetuin (1.25 mg/ml), nor by pretreating the melanoma cells with neuraminidase (10 units/ml for up to 24 h at 37°C), which is known to cleave terminal sialic acid residues, suggesting that MCLD binding is independent of sialic acid. Since plasminogen (Plg) markedly facilitated the binding of M. fermentans to HeLa cells (27), the effect of Plg on M. hyorhinis binding to melanoma cells was studied. However, M. hyorhinis binds Plg to the same extent as M. fermentans, as determined by the immunoblot assay (27), but the maximal binding of MCLD to melanoma cells was not affected by preincubation (1 h at 37°C) with 25 μg/ml Plg (data not shown).
Although M. hyorhinis is a parasite associated with the cell surface of host cells (20), early light microscopy and electron microscopy observations showed that under certain conditions M. hyorhinis may be visualized being engulfed in membrane vesicles, leading to conflicting interpretations (13). The question remains as to whether M. hyorhinis bacteria are intracellular or whether they are at the bottom of crypts formed by the invagination of the cell membrane. The ability of MCLD to grow and survive within melanoma cells was therefore studied. The gentamicin resistance assay has previously been introduced to mycoplasma studies to determine quantitatively the number of intracellular bacteria (1). In this assay, intracellular bacteria are shielded from the antibiotic effect due to the limited penetration of gentamicin into the host eukaryotic cells. With eubacteria, the number of intracellular bacteria is determined by washing the host cells from the antibiotic, lysing them with a mild detergent to release the intracellular bacteria, and counting the colonies. Since mycoplasmas are as susceptible to detergent lysis as the host cells, we plated dilutions of the infected host cells directly on solid mycoplasma medium without lysing them beforehand. Therefore, each mycoplasma colony obtained represents an infected host cell rather than a single intracellular bacterium. The gentamicin procedure was successfully adapted to M. penetrans, M. fermentans, and M. gallisepticum systems (1, 25, 27) although these bacteria have a low susceptibility to gentamicin. With M. penetrans, gentamicin concentrations as high as 200 to 400 μg/ml were utilized for invasion assays, and in some cases the susceptibility to gentamicin was further increased by adding low concentrations of Triton X-100 to the invasion medium (1). As the sensitivity of M. hyorhinis to gentamicin was very high (MIC of 0.5 μg/ml compared to MIC of >256 μg/ml for M. penetrans), the intracellular location of M. hyorhinis strain MCLD in melanoma cells was therefore determined by the gentamicin protection assay using a low gentamicin concentration (10 μg/ml). Table Table11 shows that, whereas binding of radiolabeled M. hyorhinis was observed with both viable and PFA-treated melanoma cells, invasion determined by the colony counting procedure revealed a significant CFU count with native melanoma cells but a very low or nonexistent CFU count with the PFA-treated cells. We assume that the binding to the PFA-treated melanoma cells represents adhering bacteria, whereas the binding to the native melanoma cells represents both adhering and invading bacteria. The low number of CFU obtained with PFA-treated melanoma cells may represent noninvasive M. hyorhinis bacteria that survive the gentamicin treatment, apparently due to being shielded by the melanoma cells. Intracellularly located bacteria were detected as early as 1 h postinfection. Up to 24 h postinfection the intracellular M. hyorhinis did not cause obvious damage to the host cells, as shown by the trypan blue viability assay, but as the infection proceeded, the number of intracellular bacteria increased; at a late stage of infection (>48 h) host cell integrity was disrupted, and cell death followed (data not shown).
To further establish the notion that M. hyorhinis bacteria do invade melanoma cells, immunofluorescence staining followed by analysis with CLSM was performed. Melanoma cells were infected by M. hyorhinis, fixed with PFA, permeabilized by treatment with Triton X-100, and incubated with a fluorescent donkey polyclonal nonspecific anti-M. hyorhinis antiserum. The results clearly demonstrated that in native melanoma cells infected by M. hyorhinis, both surface and intracellular foci of fluorescence corresponding to intracellular and extracellular mycoplasmas are seen (Fig. (Fig.3A).3A). The amounts of intracellularly located M. hyorhinis increased during the first 24 h of infection (data not shown). Invasion was temperature dependent; thus, at 4°C, fluorescence was predominantly associated with the cell surface of the host cells (data not shown). Invasion was not detected in control PFA-treated melanoma cells infected with M. hyorhinis, where numerous foci of fluorescence corresponding to extracellular, bound mycoplasmas were seen (Fig. (Fig.3B)3B) or when native melanoma cells were infected with PFA-treated M. hyorhinis (data not shown). To obtain three-dimensional information on the cellular location of the fluorescence, a series of optical sections was made through the infected melanoma cells (Fig. (Fig.4).4). The results clearly demonstrated surface binding and intracellularly located MCLD. The cytoplasmic pool of the ingested mycoplasmas was organized in bright fluorescent aggregates that were randomly distributed, tending to concentrate near the plasma membrane. There is no doubt that invasion is associated with adhesins that mediate the interaction of the bacteria with the host cell (24, 27). It is also likely that surface molecules that facilitate the adhesion process will have an effect on the invasion. Nevertheless, adherence to the surface of host cells is not sufficient to trigger events that lead to invasion. The signals generated by the interaction of host cells with invasive mycoplasmas have yet to be investigated.
Proteolytic modification of bacterial surface protein(s) is an emerging theme in the study of bacterial pathogenicity. It has been suggested that, because mycoplasmas lack an external cell wall, they rely on surface proteins that act as adhesins for direct interactions with host cells (2, 22, 24). Thus, pretreatment of intact M. fermentans or M. penetrans with proteases has been shown to reduce binding to host cells by 70% and 90%, respectively (27, 29). Surprisingly, proteolysis of MCLD by proteinase K (2.5 to 20 μg/ml for 5 to 30 min) resulted in enhanced binding to melanoma cells in comparison to the marked binding inhibition induced by proteinase K treatment of M. penetrans or M. fermentans (Fig. (Fig.5).5). Similar results were obtained with HeLa cells (data not shown). Treatment of MCLD with proteinase K (20 μg/ml for 30 min at 37°C) in the presence of the commonly used protease inhibitor phenylmethylsulfonyl fluoride (PMSF) produced no significant enhancement effect (data not shown). Likewise, no enhancement was obtained when MCLD was treated with proteinase K preheated to 75°C for 15 min before use (data not shown). To determine whether the proteolytic enhancement was specific to proteinase K, alternative proteolytic treatments were tested. Enhanced binding was also observed when MCLD was treated with papain, but treatment with trypsin or chymotryspin (20 μg/ml up to 60 min) had only a small effect (data not shown). The enhanced binding of MCLD to melanoma cells was observed even after a mild proteolytic treatment (2.5 μg/ml proteinase K for 5 min at 37°C). The effect of mild proteolysis on the binding to native or PFA-treated melanoma cells is shown in Fig. Fig.6.6. Pretreatment of melanoma cells with 4% PFA for 15 min at 37°C reduced the binding of both untreated and proteinase K-treated MCLD by 20% with no major change in the binding kinetics. Proteinase K treatment of MCLD also markedly enhanced the invasion of melanoma cells by the bacteria, as shown by the gentamicin resistance assay (Table (Table1)1) as well as by the immunofluorescence staining followed by CLSM analysis (data not shown).
Figure Figure77 shows protein profiles of MCLD prior to and after proteolysis by proteinase K. The figure shows that most protein bands affected by proteolysis were detected in the membrane lipoprotein fraction (Fig. (Fig.7,7, lanes II). Lipoproteins are abundant mycoplasmal cell membrane compounds, consistent with the many putative lipoprotein-encoding genes identified in the mycoplasma genomes that have been sequenced so far (6). The lipoproteins are of great importance in the biology and pathogenesis of mycoplasmas with regard to both specific interactions with host cells (3, 8) and the broader modulin activities associated with them (8, 19). In M. hyorhinis the lipoproteins contain a lipoylated amino-terminal cysteinyl residue but not an N-acyl group (19), and a prototype family of seven genes encoding the variable surface lipoproteins of this organism was characterized (7). The most abundant lipoproteins in MCLD were detected as 83-kDa, 75-kDa, 47-kDa, and 44-kDa protein bands (p83, p75, p47, and p44, respectively). Upon proteolysis, the most pronounced effect was the decreased intensity of p47, which, according to mass spectrometry, was homologous to the p46 lipoprotein of Mycoplasma hyopneumoniae (10). A small decrease in the intensity of p83 and the disappearance of the fine band corresponding to p75 were also noticed. The decreased intensity of p47 was associated with a concomitant increase in the intensity of p44. Mass spectrometry of p47 and p44 suggested that the two protein bands are homologous, differing by a 3-kDa peptide present at the C terminus of p47. The possibility that the enhancement of the binding to and invasion of melanoma cells by protease-treated MCLD is due to either the activation of an adhesin(s) or to enhanced exposure of an adhesin(s) merits further investigation.
The help and advice of H. Rechnitzer and the excellent technical assistance of A. Katzenell are greatly appreciated.
Editor: B. A. McCormick
Published ahead of print on 16 November 2009.