Background

Mycoplasma species (class Mollicutes) are widespread parasites of man, animals, plants and insects that are considered typical surface parasites [1]. Mycoplasma species have a small genome and are the smallest self-replicating organisms [2]. Owing to their limited biosynthetic capabilities, most Mycoplasma species exhibit strict host and tissue specificities [2]. Due to lack of a cell wall, mycoplasmal adhesins must be part of their cell membrane allowing for direct contact between the mycoplasmal cell membrane and specific receptors on the host cell membrane [3]. This close interaction probably creates a micro-environment which allows for the uptake of important nutrients and accumulation of mycoplasmal metabolic end products causing damage to host cell membranes [1]. Adhesion mechanisms of M. pneumoniae [4] and M. genitalium [5] are the best studied among Mycoplasma species. In M. genitalium defective mutants of the adhesions P1 and P30 have been shown to be avirulent, hence considering them as major virulence factors, [6]. For M. conjunctivae, adhesion was shown to depend on the RGD motif (Arg-Gly-Asp) of Lipoprotein T (LppT), a protein belonging to the α-integrin binding lectin family that includes fibronectin, vibronectin, fibrillin and von Willebrand factor [7]. No adhesion molecules of Mycoplasma mycoides subsp. mycoides (Mmm) have been described yet.

Mmm is the causative agent of contagious bovine pleuropneumonia (CBPP), a severe, highly contagious respiratory disease of cattle. The disease has also been reported in Italian buffaloes (Bubalus bubalis) [8], American bison (Bison bison) and Asian yak (Bos grunniens), but never in African buffalo (Syncerus caffer) [9]. CBPP is transmitted via aerosols of infected animals and is characterized by severe inflammatory, exudative lesions at lung and pleural membranes. In calves however, infection of Mmm results mainly in swollen painful limbs (arthritis) and associated lameness, but rarely in pulmonary lesions [9]. Mmm belongs to the closely related M. mycoides cluster [10] which includes four other species and subspecies, namely M. mycoides subsp. capri (Mmc), M. capricolum subsp. capripneumoniae, M. capricolum subsp. capricolum, and M. leachii. Although Mmc and Mmm are phylogenetically closely related [11], they greatly differ in host tropism [12]. Unlike Mmm, Mmc affects small ruminants causing clinical signs including mastitis, arthritis, kerato-conjonctivitis, pneumonia and septicemia, abbreviated as ‘MAKePS’ [13]. Mmm has been isolated from sheep and goats [14] however there is no indication that it can cause disease in these species. Likewise, Mmc has also been isolated from cattle [15], but has not been reported to cause disease in cattle.

A previous study [16] demonstrated the adherence of Mmm to embryonic calf nasal epithelial (ECaNEp) cells using realtime PCR assays. Flow cytometric assays have been previously described to measure cyto-adherence of other Mycoplasma species [17,18]. The aim of this study was to set up a novel flow cytometry (FCM)-based high throughput adhesion screening for Mmm and utilize this model for characterization of adhesion among different ruminant cell lines and a panel of M. mycoides strains. This approach might reveal an in vitro binding pattern that correlates with disease outcomes observed in young and adult cattle.

Methods

Cell culture

Primary epithelial and endothelial cells were cultured from adult bovine tissues including accessory lung lobe epithelial cells (BoLEC) and aorta endothelial cells (BoAEC), bovine fetal lung epithelial cells (BoFLEC) and caprine lung epithelial cells (CaLEC) were cultured from bovine fetuses and goat lungs respectively. Additionally, previously established immortal bovine skin fibroblasts (IBoSF) were used in this study [19].

BoLEC and CaLEC were cultured using the protease digestion technique as described elsewhere [20]. Briefly, tissues were digested using 1% protease XIV (Sigma, St Louis, USA) at 4°C overnight before collecting cells. BoAEC were cultured using the same technique as described above however tissues were digested in 0.25% collagenase (Sigma, St Louis, USA) [21]. BoFLEC were cultured according to a method described by Schweizer and Peterhans (1999) [22] with slight modifications. Fetal lung tissue was digested in trypsin-EDTA (0.025%: 0.01%) in PBS at 37°C for 1 h with shaking (200 rpm). All cell lines were maintained in Dulbecco’s minimum essential medium (DMEM) supplemented with 10% inactivated fetal bovine serum (Sigma, St Louis, USA), 200 IU/ml penicillin, 150 μg/ml streptomycin, 1 μ g/ml nystatin, 2 mM L-glutamine and 0.15 M 2-mercaptoethanol.

Mycoplasma culture preparations

Mmm and Mmc strains used in this study (Table 1) were cultured in 20 ml “pleuropneumonia -like organism” (PPLO) medium (Becton Dickinson, Park, USA) supplemented with 10% horse serum (Sigma, St Louis, USA), 0.9 g/l yeast extract, 0.5% glucose and 0.03% penicillin G at 37°C for 48 and 18 h to a density of 108 and 109 cfu/ml respectively. Each Mycoplasma strain concentration was determined by the Spearman–Karber formula [23] before centrifugation. The mycoplasmas were harvested by centrifugation at 6,000 × g at 4°C for 30 min, washed once in DMEM without supplements and suspended in 10 ml of the same media. Final Mycoplasma titers were calculated using optical density at OD650 and readings plotted on a standard curve based on correspondence between OD650 and Mmm numbers as determined by TaqMan Real Time PCR.

Table 1 Mycoplasma strains used in this study*
Full size table

Polyclonal rabbit antibody production

The production of rabbit polyclonal antibodies was approved by the institutional animal care and use committee (IACUC reference number 2008.14). Mmm strain Afadé was cultured as described above and the proteins prepared by ultra-sonication, and protein concentration determined by the micro BCA according to the manufacturer’s instructions. One six weeks old rabbit was immunized twice with 1 mg of whole cell lysate at an interval of eight weeks. The rabbit was immunized by giving two immunizations with heat inactivated (60°C for 10 min) whole cell lysate (500 μl) mixed with the same volume of complete Freund’s adjuvant in the first immunization or the same volume of incomplete Freund’s adjuvant in the boost. At each immunization half of the antigen-adjuvant mixture was administered intradermally (500 μl) and half subcutaneously (500 μl). Pre-immunization serum was collected two weeks before initial immunization. Post-immunization serum was collected four weeks after the second immunization, heat inactivated at 56°C for 30 min, aliquoted and stored at -20°C until use.

The rabbit serum was validated using dot blot for all the Mycoplasma strains used in this study. Briefly, 2 μl of proteins from each Mycoplasma strain was dotted on to nitrocellulose membrane (0.45 μm pore size, Bio-rad), and allowed to dry at room temperature for 30 min. Unspecific binding sites on the blot were blocked in 5% BSA-PBS for 1 hr at room temperature. The blot was then incubated for 1 hr at room temperature with the rabbit serum diluted at 1:2,500. After 3 washes in PBS-Tween 20, the blots were incubated for 1 hr at room temperature with anti-rabbit IgG (whole molecule)-alkaline phosphatase antibody produced in goat (Sigma) at a dilution of 1:5000. After three washes, blots were developed with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (BCIP/NBT, Sigma) in alkaline phosphatase buffer (100 mM NaCl, 5 mM MgCl2, 100 mM Tris, pH 9.5).

Indirect immunofluorescent microscopy

1×105 BoLEC resuspended in 500 μl complete DMEM were seeded into a six cell well plate (Corning, New York, USA) containing sterile glass cover slips (Thermo Scientific) and cultured for 2-3 days at 37°C in 5% CO2. Cover slips were collected and washed three times in DMEM without supplements. Mmm strain Afadé was cultured as described above and 200 μl (approximately 1.5×108 mycoplasmas) was added to the cell culture and incubated for two hours. Unbound mycoplasmas were washed off three times using 500 μl DMEM without supplements, prewarmed to 37°C. Cells were fixed in 3% paraformaldehyde for 15 min at room temperature. Cyto-adherent mycoplasmas were indirectly stained using anti-Mmm Afadé polyclonal rabbit serum (100 μl, 1:100 dilution) for 1 hour on ice, followed by washing thrice with 500 μl cold DMEM without supplements and 1 hour incubation with goat anti-rabbit FITC (Sigma, St Louis, USA; 100 μl, 1:500 dilution). Cover slips were air dried and images acquired using a fluorescent microscope (Axio Imager, Carl Zeiss AG, Gottingen, Germany) at a magnification of 400x. A control experiment was done as described above only that mycoplasmas were omitted and the cells were counter stained with 4′, 6 –Diamidino-2-phenylindole (DAPI).

Flow cytometry analysis of cyto-adherence

Cell lines were cultured in 24 well plates to a density of approximately 1.5×105 cells per well. To determine the optimum minimum concentration of mycoplasmas for cyto-adherence, all the Mycoplasma strains used in this study were ten-fold serially diluted. The cyto-adherence assay was performed as described above. Unbound mycoplasmas were washed off thrice with 500 μl prewarmed DMEM without supplements. Cells with bound mycoplasmas were detached from the culture plates with 250 μl PBS/EDTA buffer and the reaction stopped after 3 min by adding 500 μl prewarmed DMEM without supplements, transferred to 1.5 ml eppendorf tubes and centrifuged at 1,000 g for 10 min. The supernatant was discarded and the cells transferred to 96 round bottom well plates (Corning, New York, USA). Experiments were performed three times in duplicate. Staining was done as described above (100 μl/well, 1:100 dilution), followed by washing thrice with 200 μl of cold DMEM without supplements and centrifugation for 2 minutes at 1000 g. Cells were incubated with goat anti- rabbit FITC (Sigma, St Louis, USA; 100 μl/well, 1:250 dilution) for 1 hour on ice. After the final wash, cells were suspended in 100 μl of FACS medium (PBS, 2% horse serum, 0.2% sodium azide and 2% formalin). Cells were analysed by flow cytometry (FACS canto II, Becton Dickinson Co., San Jose, USA) in 5 ml falcon tubes according to the manufacturer’s instructions. Three thousand signals were acquired and the results presented as dot plots. To calculate the percentage of positive cells, fluorescence threshold was set such that unstained cells without mycoplasmas would have 0% fluorescence. This threshold was kept the same for all samples of each experiment and reset for each cell type. Percentage of cells with bound mycoplasmas and mean fluorescence intensity (MFI) were calculated using flow jo software [24]. Statistical analysis of the cells with bound mycoplasmas was done using ANOVA.

Results

Specificity and kinetics of cyto-adherence

Rabbit serum directed against whole cell antigens of Mmm reacted to both Mmm and Mmc with similar efficiency (see Additional file 1: Figure S1), which justified its subsequent use for the detection of both Mycoplasma subspecies.

Microscopic analysis of adherent Mmm to BoLEC showed uneven distribution of Mmm on the cell surface. The adherent mycoplasmas appeared as aggregated bright fluorescent spots on the surface of the cells (Figure 1). Due to the small size of Mmm, it was not possible to visualize individual Mycoplasma cells under the conditions used.

Figure 1
figure 1

Indirect immunofluorescent staining of Mycoplasma mycoides subsp. mycoides ( Mmm ) bound to bovine lung epithelial cells (BoLEC). A shows BoLEC without Mmm stained with anti-Mmm (Afade) + anti-rabbit FITC and counter stained with DAPI. B shows BoLEC with adherent Mmm stained with anti-Mmm (Afade) + anti-rabbit FITC and counter stained without DAPI. The arrow indicates bright fluorescent patches of Mycoplasma clusters on the surface of an epithelial cell.

Full size image

Mmm cyto-adherence kinetics were analyzed using an indirect flow cytometry assay with appropriate controls using unstained lung epithelial cells without adherent mycoplasmas to set the fluorescence cut-off level (Figure 2). In all subsequent experiments the MFI was used to compare binding of Mmm to target cells, as this was a more representative parameter of the relative average number of mycoplasmas bound per cell. Saturation of Mmm binding to BoLEC and Mmc to CaLEC occurred 2 hours post infection (Figure 3A) at approximately 108 and 109 mycoplasmas per 1.5×105 cells (Figure 3B and C). In all subsequent experiments, mycoplasmas from each strain were used at saturating numbers.

Figure 2
figure 2

Flow cytometric analysis of Mycoplasma mycoides subsp. mycoides ( Mmm strain Afade) attachment to bovine lung epithelial cells (BoLEC). Adherent Mmm was detected by specific rabbit anti-Mmm serum followed by FITC-conjugated goat anti-rabbit-Ig (A) unstained BoLEC without Mmm; (B) stained BoLEC without Mmm; (C) BoLEC + Mmm without anti-Mmm rabbit serum; (D) BoLEC + Mmm + anti-Mmm rabbit serum without goat anti- rabbit FITC; (E) BoLEC + Mmm + pre-immunization rabbit serum + goat anti-rabbit FITC; (F) BoLEC + Mmm + anti-Mmm rabbit serum + goat anti-rabbit FITC. Vertical axis: relative fluorescence intensity. Horizontal axis: forward light scatter.

Full size image
Figure 3
figure 3

Cyto-adhesion kinetics of Mycoplasma strains used in this study. (A) Adherence of Mycoplasma mycoides subsp. mycoides (Mmm) strain Afade to cattle lung epithelial cells (BoLEC) as a function of incubation time. (B) Serial dilution (1:10) of Mmm strains and adherence to BoLEC. (C) Serial dilution (1:10) of Mycoplasma mycoides subsp. capri (Mmc) strains and adherence to caprine lung epithelial cells (CaLEC).

Full size image

Cell lines from different host tissues were cultured to assess the binding specificity of a panel of eight Mmm strains (Table 1). Maximum binding of all Mmm strains occurred on BoLEC, whereas intermediate binding occurred with BoAEC (p = 0.006) and low binding with BoFLEC (p = 2.12 × 10-10) and IBoSF (p = 7.24 × 10-11) (Figure 4).

Figure 4
figure 4

Cyto- adherence capacity of different Mycoplasma mycoides subsp. mycoides ( Mmm ) strains to different cattle cell lines, expressed as mean fluorescence intensity (MFI) of cells with bound Mycoplasma .

Full size image

Cyto-adherence of Mmm and Mmc to specific host cells was assessed by testing the binding ability of both Mycoplasma species to goat and cattle cells. We demonstrated a statistically significant difference between the cyto-adherence capacities of Mmm to BoLEC and CaLEC (p = 1.11 × 10-12), with Mmm binding more efficiently to BoLEC than CaLEC (Figure 5). On the other hand, Mmc bound more efficiently to CaLEC than BoLEC (Figure 6) however the difference was less apparent than in the case of Mmm but statistically significant (p = 0.006).

Figure 5
figure 5

Cyto- adherence capacity of different Mycoplasma mycoides subsp. mycoides ( Mmm ) strains to lung epithelial cells from cattle (BoLEC) and goats (CaLEC), expressed as mean fluorescence intensity (MFI) of cells with bound Mycoplasma .

Full size image
Figure 6
figure 6

Cyto- adherence of different Mycoplasma mycoides subsp. capri ( Mmc ) strains to lung epithelial cells from cattle (BoLEC) and goats (CaLEC), expressed as mean fluorescence intensity (MFI) of cells with bound Mycoplasma .

Full size image

Intraspecies variation of cyto-adherence capacity

The different strains of Mmm did not show much variation in the ability to adhere to BoLEC although some small statistical differences of cyto-adherence capacities could be detected for certain strains; in particular with strain Madrid which seemed to have a higher binding capacity compared to the other strains (Figure 5; see Additional file 2: Table S1). Mmc on the other hand showed relatively high intra-species variation in its capacity to adhere to CaLEC with strains PG3, Capri-L and My325 binding weakly to the cells (Figure 6; see Additional file 3: Table S2).

Discussion

The aim of this study was to set up a novel flow cytometry (FCM)-based high throughput screening for adhesion of Mmm and to utilize the model for characterization of adhesion of various strains of M. mycoides subsp. mycoides and M. mycoides subsp. capri to different ruminant cell lines. In our study, specks of fluorescence could be detected that varied in size, suggesting that Mmm were often present as clusters or aggregates of different sizes. A previous study also showed a heterogeneous distribution of M. bovis on the surface of host cells [20]. Using the flow cytometric assay we determined a statistical difference in the binding of Mmm to various bovine host tissues; the highest binding was detected with BoLEC, intermediate binding with BoAEC and low binding with IBoSF and BoFLEC. High binding to BoLEC is expected as the main lesions and pathology caused by Mmm are found in the lung of adult cattle [25] suggesting that there might be a receptor for Mmm on lung cells that allows for this interaction. Low adherence of Mmm to fetal lung epithelial cells may suggest a lack or low abundance of a receptor for Mmm, which may be acquired through activation of the corresponding genes later in life. This result probably explains why neonatal calves that are infected with Mmm do not develop severe lung lesions [9]. This also correlates with epidemiological observations showing that CBPP generally affects adult cattle [9,12]. In a previous study, binding of Mmm to embryonic calf nasal epithelial cells (ECaNEp) was measured in order to analyze the differences in the adhesion capacity of Mmm strains with diverse degrees of virulence, as ECaNEp cells were used as a model for cytotoxicity. However, this study did not compare host cells of animals of various ages and hence measurements have been done at a relatively low range of binding capacity [16].

Moderate binding to endothelial cells might correlate with pathology seen in capillaries and lymph vessels. Mmm has been isolated from the lymph nodes of cattle with CBPP [25] and vasculitis is a common occurrence in heavily infected animals. An interaction between mycoplasmas and endothelial cells may contribute to the vasculitis. Subcutaneous inoculation of mycoplasma often causes serious necrotic edema, but not in lymphatic-poor areas, such as nose or tail tip [26] however this method of inoculation does not result in typical clinical CBPP.

According to our results, there was little significant statistical difference between the adherence capacities of the different Mmm strains tested, which suggests that all strains used expressed the relevant adhesion ligands and the differences in the virulence among these strains is not due to variation in their adhesins. This is also probably due to the fact the Mmm strains exhibit very little genetic diversity [11]. The caprine Mycoplasma (Mmc) on the other hand showed more variation in cyto-adherence patterns among strains, which could correlate with their broader genetic diversity [11]. But the clinical signs of Mmc are not confined to lung, as there are many other manifestations, abbreviated as ‘MAKePS’ [13]. Adherence to other cell types might show different patterns and assessing the correlations between cyto-adherence of Mmc strains and different cell types needs further investigation. Furthermore, there is a possibility that the rabbit anti-Mmm serum used in this study might have reacted differently with individual Mmc strains. Mmm cyto-adherence assays revealed relevant host species specificity: Mmm adheres to bovine lung epithelial cells, but very little to caprine lung epithelial cells. This suggests a correlation between the capacity to adhere to target cells, in this case to cattle cells, and the development of disease specifically in cattle, suggesting the presence of host specific adhesion factors. However, Mmc binds to lung cells from both species. This needs further investigation since Mmc has never been reported as causing disease in cattle [15] but other factors may influence the disease outcome. The cyto-adherence of Mmm supports previous reports from other Mycoplasma species that adherence is tissue and organ specific; however it is not uncommon to isolate mycoplasmas from tissues or hosts other than their natural habitats [27].

Conclusion

Cyto-adherence of Mmm to mammalian host cells was studied using an indirect, flow cytometric assay. The data show that strong binding of Mmm is specific to lung epithelial cells from adult cattle, but not fetuses. This correlates with the severe in vivo clinical signs observed in lung tissue of infected, adult cattle, but which are mostly absent in infected, very young calves. An intermediate binding to endothelial cells correlates with pathological signs detected in capillaries and lymph vessels. Binding was weak to other tissue cells and to lung cells from goats, suggesting that a strong binding of Mmm with their target cells might be the basis of the species and tissue specificity of CBPP, and be a requisite step in the development of the disease. Blocking adhesion in vivo might represent a valuable target to prevent colonization of the lung by Mmm.