Mycoplasma synoviae induces upregulation of apoptotic genes, secretion of nitric oxide and appearance of an apoptotic phenotype in infected chicken chondrocytes

The role of chondrocytes in the development of infectious arthritis is not well understood. Several examples of mycoplasma-induced arthritis in animals indicate that chondrocytes come into direct contact with bacteria. The objective of this study was to analyze the interaction of an arthrogenic Mycoplasma synoviae strain WVU 1853 with chicken chondrocytes. We found that M. synoviae significantly reduces chondrocyte respiration. This was accompanied by alterations in chondrocyte morphology, namely cell shrinkage and cytoplasm condensation, as well as nuclear condensation and formation of plasma membrane invaginations containing nuclear material, which appeared to cleave off the cell surface. In concordance with these apoptosis-like events in chondrocytes, transcription was increased in several pro-apoptotic genes. Twenty-four hours after infection, strong upregulation was assayed in NOS2, Mapk11, CASP8 and Casp3 genes. Twenty-four and 72 h incubation of chondrocytes with M. synoviae induced upregulation of AIFM1, NFκB1, htrA3 and BCL2. Casp3 and NOS2 remained upregulated, but upregulation ceased for Mapk11 and CASP8 genes. Increased production of nitric oxide was also confirmed in cell supernates. The data suggests that chicken chondrocytes infected with M. synoviae die by apoptosis involving production of nitric oxide, caspase 3 activation and mitochondrial inactivation. The results of this study show for the first time that mycoplasmas could cause chondrocyte apoptosis. This could contribute to tissue destruction and influence the development of arthritic conditions. Hence, the study gives new insights into the role of mycoplasma infection on chondrocyte biology and development of infectious arthritis in chickens and potentially in humans.


Introduction
Species from the genus Mycoplasma have long been recognized as important pathogens that cause diseases of the respiratory tract, urogenital tract and joints in a variety of animal species, including humans. Mycoplasma synoviae (M. synoviae) is a common pathogen of chickens and turkeys. It usually colonizes the respiratory tract, although this condition can lead to the development of systemic infection and/or infectious synovitis [1][2][3][4]. Experimental infectious synovitis has been associated with the hemagglutination-positive phenotype of M. synoviae [5]. The condition is presented as an acute joint swelling accompanied by increased volume of synovial fluid, infiltration with macrophages, B and T cells, lining cell hyperplasia and fibroblastic proliferation. In the acute phase of infection, live M. synoviae have been detected in synovial fluids of infected birds, suggesting that M. synoviae can come into direct contact with chondrocytes as is the case with M. pulmonis [6] and M. artritidis [7]. M. synoviae has been reported capable of invasion into non-phagocytic chicken cells including chondrocytes in vitro [8].
Apoptosis is a highly regulated process of cell demise and is generally induced through activation of different components of two overlapping signaling pathways. The extrinsic pathway includes the activation of death receptors leading to activation of a family of cysteine proteases known as caspases [23]. Proteolytic cleavage of around 400 caspase substrates that have been identified so far [24] leads to phenotypic features of apoptosis. These include rounded morphology, condensed cytoplasm, fragmented organelles and formation of plasma membrane blebs that are separated from the cell into independent vesicles known as apoptotic bodies (reviewed in [25]). The intrinsic pathway is linked to stress signals from within the cell. It involves cleavage of Bid into tBid leading to activation of Bax and Bak1. They form a pore in the outer mitochondrial membrane and cause the release of proteins from the mitochondrial intermembrane space (endonuclease G, AIFM1, cytochrome c etc.) leading to execution of either caspase independent DNA cleavage (AIFM1, endonuclease G) [26] or the formation of apoptosome and activation of caspase 3 [27,28].
One of the key players in the induction of apoptosis of articular chondrocytes in humans is nitric oxide (NO), which is produced by immune cells in the synovium and pannus as well as by activated chondrocytes [29][30][31][32]. NO induces apoptosis through p38 MAP kinase (MAPK11) mediated stimulation of caspase 3 activity and accumulation of nuclear factor kB (NFkB). This leads to increased expression of p53, a signaling molecule that acts upstream of caspase 3 [33,34], as well as upregulation of pro-apoptotic Bax. Another mechanism of apoptosis in arthritic cartilage includes activation of CD95 (Fas) death receptor leading to upregulation of Fas ligand and p38 MAP kinase [35].
In this study, M. synoviae infected chicken chondrocytes (CCH) were evaluated during three days for changes in cell viability, morphology and expression of genes associated with apoptosis.

CCH cell culture
Chicken chondrocytes (CCH) were obtained from hyaline cartilage of hock joints as described previously [8]. Primary cultures of CCH were cultivated up to the sixth passage in Dulbecco Modified Eagles Medium (DMEM), supplemented with 7.5% fetal bovine serum (FBS) and 2.5% chicken serum (all from Sigma-Aldrich, Munich, Germany). Cells were incubated at 37°C in a 5% CO 2 incubator.

Jurkat cell culture
Jurkat cells (ATCC, CCL TIB 152, kindly donated by Prof Nataša Kopitar Jerala, Jožef Stefan Institute, Supports continous growth of 32D cells. [22] Slovenia) that overexpress death receptors were used as a positive control for apoptosis. Cells were cultivated in DMEM supplemented with 10% FBS and incubated at 37°C in a 5% CO 2 incubator.

Mycoplasma synoviae cultures
Cultures of M. synoviae type strain WVU 1853 (approximately 20 in vitro passages before being used in this study) were grown in Frey mycoplasma broth as described previously [8]. Bacteria in the logarithmic phase of growth were used for CCH infection. The number of colony forming units (CFU) was determined as described previously [36].

XTT-based cell viability test
To perform the cell viability test, CCH were seeded into 96 well plates (Corning, New York, USA) at density 5 × 10 5 cells/mL and incubated overnight at 37°C in a 5% CO 2 incubator prior to infection. M. synoviae broth culture in the logarithmic phase of growth was used to infect CCH in a ratio of 10-100 M. synoviae CFU per CCH. As a positive control of apoptosis, 5-fluorouracil was used (5-FU, final concentration 20 μg/mL, from Sigma-Aldrich, Munich, Germany). Non-infected CCH were used as negative controls. As additional controls of the experiment, Jurkat cells, seeded at density 5 × 10 5 / well, were subjected to the same three procedures. Tetrazolium-based dye XTT (Sigma-Aldrich, Munich, Germany) was added to wells after 0, 24, 48, 72 and 96 h of incubation. The final concentration of XTT in cell suspensions was 20%. Following 3 h of incubation at 37°C in a 5% CO 2 incubator, absorbance of the samples was measured at 492 nm. The background signal was measured at 690 nm. The test was performed in three independent experiments with triplicates of each experimental condition. Average absorbances were calculated using MS Excel. Student t-test for independent values was used for determining statistically significant differences (p < 0.05).

Phase contrast and confocal fluorescence microscopy
Morphological changes were observed by phase contrast and confocal fluorescence microscopy. CCH were seeded onto cover slides at density 5 × 10 5 cells/mL and incubated overnight. The cells were then infected with M. synoviae as described and incubated for 24 h, 48 h and 72 h followed by phase contrast microscopy examinations for morphological changes. Additionally, distribution of CD44 receptor (hyaluronan receptor; its crosslinking augments Fas expression and subsequent Fasmediated apoptosis of the cells [37]), and nuclear condensation were observed using confocal fluorescence microscopy. The medium was aspirated, and CCH were washed in sterile phosphate buffered saline (PBS). The cells were fixed using a pre-cooled mixture of acetone and methanol (both from Sigma-Aldrich, Munich, Germany) for 10 min at room temperature. A solution of 10% goat serum (Sigma-Aldrich, Munich, Germany) was used for blocking (1 h, room temperature), followed by 3 washes in PBS containing 0.5 mg/mL RNase (Sigma-Aldrich, Munich, Germany). Mouse monoclonal antibodies to chicken CD44 (DSHB, Iowa City, USA), diluted 1:100 in PBS were added for 1 h at room temperature. After washing, Alexafluor-488 goat anti-mouse IgG antibodies, diluted 1:2000 (Life Technologies -Invitrogen, Carlsbad, USA), were used as secondary antibodies. After 1 h at 37°C, unbound antibodies were washed off, and nuclei were labeled for 5 min in propidium iodide (Sigma-Aldrich, Munich, Germany) in PBS with 0.1% Triton X-100 and 0.5 mg/mL RNase (both from Sigma-Aldrich, Munich, Germany). Following washing in PBS, cover slides were mounted using Dako Fluorescence Mounting medium (Dako, Glostrup, Denmark) and viewed using a Leica TCS SP5 confocal microscope (Leica Microsystems, Mannheim, Germany).

RT-qPCR
In order to perform gene expression analysis, CCH were seeded into 75 cm 2 flasks (Techno Plastic Products, St. Louis, USA) at density 5 × 10 5 cells/10 mL and incubated overnight. The cells were then treated as described for the XTT assay and sampled by trypsinization after 24, 48 and 72 h of incubation. CCH were washed in sterile PBS followed by isolation of total RNA using RNeasy Mini Kit (Qiagen GmbH, Hilden, Germany). RNA concentration was determined spectrophotometrically using NanoVue (GE Healthcare, Waukesha, USA). Subsequently, RNA was stored at -80°C. Residual DNA was eliminated by incubating 1 μg of RNA with 1U of RNase-free DNase I in a buffer containing MgCl 2 ions (both from Thermo Fisher Scientific -Fermentas, St. Leon-Rot, Germany) for 30 min at 37°C. Following DNase inactivation at 65°C, cDNA was synthesized using a High-Capacity cDNA Reverse Transcription Kit (Life Technologies -Applied Biosystems, Foster City, USA), according to the manufacturer's instructions. cDNA was stored at -20°C immediately after reverse transcription reactions. Prior to use, primer pairs were checked for specificity in silico using NCBI Primer BLAST [38]. Primer efficiency was checked on each primer pair using dilutions of cDNA combined from all experimental conditions and timepoints.
For RT-qPCR reactions, 20 μL mixtures were made, containing 10 μL of Power SYBR Green PCR Master Mix (Life Technologies -Applied Biosystems, Foster City, USA), 0.5 μL of forward and 0.5 μL of reverse primer (final concentration 5 μM, Integrated DNA Technologies, Leuven, Belgium), 8 μL of DEPC-treated water (Qiagen GmbH, Hilden, Germany) and 1 μL of cDNA. Stratagene Mx3000P (Agilent Technologies -Stratagene, Santa Clara, USA) was used to perform RT-qPCR reactions and its MxPro software for analysis of amplification and dissociation plots. Reaction conditions were set to 10 min at 95°C (first segment, one cycle), 15 s at 95°C and 1 min at Tm of a specific primer pair (second segment, 40 cycles) followed by one cycle with 15 s at 95°C, 30 s at Tm and 15 s at 95°C (dissociation curve segment). Gene expression was analyzed for 15 genes ( Table 2, [39]), and GAPDH was used as a reference gene [40]. Gene expression values of non-infected CCH were used for gene expression calibration. Appropriate controls (no template and no reverse transcription control) were also performed in each run.
Three independent experiments were performed to collect RNA for RT-qPCR. Relative gene expression was assayed in each experiment and experimental condition separately. Three repeats of each RT-qPCR reaction were performed. Normalized relative quantities were calculated using the efficiency corrected 2 -ΔΔCq method [41]. The effect of intra-assay variation on the statistical significance of the results was reduced by log transformation of normalized relative quantities, mean centering and autoscaling as described in [42]. Statistical significance of the results was determined using the Student ttest (p < 0.05) and statistically significant results visualized through Ingenuity Pathways Analysis platform (Ingenuity Systems, Redwood City, USA).

Nitric oxide assay
In order to determine the concentration of NO, an experiment equal to XTT-based cell proliferation test was set up using CCH. After 24, 48 and 72 h of incubation, supernates were transferred to a sterile 96-well plate and assayed immediately using the Griess assay (R&D Systems, Minneapolis, USA). The concentration of NO was determined based on the standard curve and absorbances of the azo-compound NO forms in the reaction with sulfanilamide and N-1-naftyletylendiamin dihydrochloride. Absorbances were read at 550 nm (background at 630 nm). Two independent experiments with three replicates of each condition were performed and mean absorbances ± standard errors were used to test statistical significance of the results (Student t-test, p < 0.05).

M. synoviae decreased CCH and Jurkat cell respiration (viability)
We assayed the respiration of CCH by determining the relative decrease in reduction of tetrazolium salt XTT in M. synoviae infected CCH compared to the negative control. As a positive control, CCH were treated with 5-FU. Jurkat cells were used as an additional positive control. M. synoviae infection caused significant decrease in CCH respiration, indicating reduced viability. This effect was noticed 72 h and 96 h after infection (p ≤ 0.1) (Figure 1). An even stronger decrease in respiration, already after 24 h, was noticed in CCH treated with 5-FU (Figure 1). Jurkat cells responded stronger than CCH to both treatments, with a major decrease in cell respiration apparent already after 24 h in both cases (p < 0.01) (Figure 1). Fresh medium was not added to any of the cells at the time of measurements, resulting in a decrease in viability of fast-growing Jurkat cells even in negative controls after 96 h (Figure 1).

M. synoviae infected CCH exhibited apoptotic morphology and nuclear condensation
Phase contrast microscopy of infected CCH revealed dramatic changes in cell morphology ( Figure 2). Twenty-four hours after infection with M. synoviae, some CCH were shrinking. Plasma membranes appear rough but formation of membrane vesicles was not noticed. The cytoplasm of infected cells shows increased vacuolization. However, the number of cells that adhered to cover slides was roughly comparable to the control ( Figure 2). Longer exposure to M. synoviae resulted in more pronounced vacuolization of cytoplasm and formation of membrane blebs seen 48 h after infection ( Figure 2). After 72 h, cytoplasm vacuolization was present in some of the uninfected CCH, to which no fresh medium was added after seeding into plates. Still, most cells were attached to cover slides. In contrast, only a few cells infected with M. synoviae remained attached to cover slides ( Figure 2). The remaining ones exhibited excessive membrane blebbing with individual vesicles detaching from cells or cytoplasm condensation (cell shrinkage) and/or vacuolization ( Figure 2). In CCH treated with 5-FU, membrane roughness and cell shrinkage was noticed after 24 h, but with no cytoplasm vacuolization present. A similar morphology was observed 48 h after treatment, but in a larger number of cells. The number of cells still attached to the surface was lower than in the control slide and some cells presented with membrane blebs. Seventy-two hours of treatment with 5-FU caused cell death and detachment from cover slides in most cells, with remaining cells presenting similar morphological features as those infected with M. synoviae ( Figure 2).
Confocal microscopy of infected CCH labeled with antibodies to membrane receptor CD44 and propidium iodide revealed additional changes caused by M. synoviae infection (Figure 3). The nuclei of negative controls stained evenly with PI after all three timepoints, while brighter foci indicating condensed DNA were present in CCH infected with M. synoviae. Additionally, vesicles detaching from the blebbing CCH plasma membrane contained condensed DNA. Differences in CD44 receptor abundance and distribution were not obvious in CCH infected with M. synoviae. CCH treated with 5-FU appeared to express more CD44 (Figure 3).
Interactions between proteins encoded by genes of interest, as well as levels of gene upregulation, were depicted using the Ingenuity Pathways platform ( Figure 5).

Chondrocytes infected with M. synoviae produced nitric oxide
Infection of CCH with M. synoviae induced a significant increase in NO production at all tested time points (p ≤ 0.01) (Figure 6), in agreement with the strong upregulation

Discussion
In the research of rheumatoid arthritis etiology, the idea of a bacterial infection leading to arthritis is becoming increasingly popular [43][44][45]. Bacterial infections increase the production of NO, and can lead to inflammation. Inflammation then leads to chondrocyte apoptosis [30][31][32][33][34][35]. In cartilage tissue that is rich in extracellular matrix and with few cells in hard to reach places, clearance of dying cells is especially troublesome and can lead to the development and perpetuation of inflammation, which in turn leads to cartilage and bone degradation [46][47][48]. The idea of mycoplasmal infections, as the triggering factor of rheumatoid arthritis, has been revisited several times. Still, the role of mycoplasmas that have been detected in synovial fluids and serums of patients suffering from arthropathies remains unclear [49][50][51][52][53][54]. Factors impeding these studies are many. Because of their obligatory parasitism and complex nutritional demands, mycoplasmas are hard to cultivate. In addition, mycoplasmas adapt to the host in a way that induces subclinical infections that pass with little or no inflammation. Furthermore, mycoplasmal antigens, many of which are enzymes, can affect cells even after bacteria are dead and lysed, and their presence is difficult to detect by certain conventional methods [49]. Mycoplasma species are also well known to cause arthritis in certain animal species, including chickens [1][2][3][4][5][6][7]. Experimental infections of chickens with M. synoviae WVU 1853 resulted in thinning of the articular cartilage 25 days post infection [55]. Visible cartilage erosion appeared about two weeks later, presumably due to the presence of increasing numbers of heterophils. It seems likely that M. synoviae could also affect CCH. This assumption is supported by our observations that synovial fluids of chickens infected experimentally into hock  . Its effect on cell viability has, however, not been reported before this study. This is the first report of mycoplasma-induced apoptosis in chondrocytes. The XTT proliferation assay was performed on chicken chondrocytes infected with M. synoviae and a major decrease in cell respiration was noted, suggesting cell death had occurred (Figure 1). Chondrocyte morphology, evaluated under phase contrast microscope and confocal microscope combined with anti-CD44 receptor and propidium iodide staining additionally indicated cytopathology with apoptotic features in chondrocytes infected with M. synoviae (Figures  2 and 3).
RT-qPCR analysis, performed on 15 genes involved in apoptosis (Table 2, Figure 4), involved NO production in apoptosis induction. The Griess assay confirmed secretion of NO to supernates ( Figure 6). NO production is closely linked to chondrocyte apoptosis and cartilage loss in human arthropathies such as osteoarthritis and rheumatoid arthritis [30][31][32][33]. Also, NO production was shown to be strongly upregulated in chicken macrophages infected with M. synoviae or in contact with its lipoprotein MSPB [57]. Both endogenously and exogenously produced NO is capable of inducing chondrocyte apoptosis, but the analysis here was limited to NO produced by chondrocytes as a result of M. synoviae infection. As expected, Mapk11 (encoding p38B, the major inducer of NO-linked apoptosis) was upregulated as well, 24 h and 48 h after infection. It has been reported that p38 causes upregulation and stimulation of caspase 3 activity, and upregulation of NFB transcription factor [34,35]. In our study, upregulation was noticed in genes encoding both caspase 3 and NFB1, as well as in genes encoding other related apoptotic proteins (Figures 4 and  5).
Increased transcription of caspase genes is the most probable explanation for the formation of apoptotic bodies and cell shrinkage in infected CCH [25]. Although genes encoding endonuclease G, which is released from the mitochondria during apoptosis, and Bak1, a pro-apoptotic mitochondrial protein, were not upregulated, another mitochondrial nuclease, encoded by AIFM1, was strongly upregulated 72 h after infection, indicating mitochondrial involvement (Figures 4 and 5). Occurrence of Fas receptor mediated apoptosis, which has been reported in chondrocytes [35], was disregarded due to unchanged transcription of both Fas and FASLG during CCH infection with M. synoviae (Additional file 1, Figure S1). Apoptosis inhibitor, encoded by BCL2, was slightly upregulated after 72 h, but so was htrA3, encoding the inhibitor of Bcl2. The expression of another apoptotic inhibitor, encoded by XIAP, remained unaltered (Additional file 1, Figure S1).
Taken together, our results indicate that M. synoviae induces endogenic nitric oxide mediated, caspase 3 and 8 dependent apoptosis in chicken chondrocytes, which involves a loss of mitochondrial function ( Figure 5).
5-fluorouracil was used as a positive control of the experiment due to its well established apoptotic effect on other cell lines [58]. Compared to M. synoviae, it induced a stronger and faster apoptotic response in CCH (Figure 4, Additional file 1, Figure S1, Figures 2  and 3). The profile of gene expression modulations indicate a Fas receptor mediated, caspase 3 independent cell death involving mitochondrial inactivation. Interestingly, the Jurkat cell line that was used as a positive control of the viability test shows great susceptibility to M. synoviae infection (Figure 1).
Although live intracellular M. synoviae were re-cultivated in our previous study 24 and 48 h after infection, the percentage of CCH invasion was relatively low (1.2 ± 0.3 for the type strain WVU 1853) [8]. This study demonstrates a high susceptibility to cell death in infected chondrocytes. This suggests that M. synoviae affects CCH both from the outer side of the membrane, probably in a Fas independent manner, and from its location within the cell. It is also possible that mycoplasmas die in cell culture due to their complex nutritional demands, but their antigens continue to stimulate CCH and cause stress. An example of macrophage stimulation to produce NO, IL6 and IL1β by M. synoviae lipoprotein MSPB has been reported previously [57].
With the exception of caspase 3, patterns of apoptotic gene upregulation were not confirmed on the protein level due to the scarcity of the appropriate specific antibodies to chicken proteins or enzymes involved in apoptotic signal transduction cascades. Even with caspase 3, the antibody used was not specific enough to recognize the cleaved form of the protein (Additional file 2, Figure  S2). Still, correlations between demonstrated cytopathologic changes, results of the XTT tests, Griess assay and gene expression data indicate apoptosis and not necrosis as the main mechanism of cell death. As an additional confirmation, CCH infected with M. synoviae were labeled using Annexin V-FITC (AV) and propidium iodide (PI) to discriminate between live (AV -PI -), early apoptotic (AV + PI -), late apoptotic (AV + PI + ) and primary/ secondary necrotic cells (AV -PI + ). The results show an increase in the percentage of cells with phosphatidylserine exposed on the outer membrane layer, to which AV is bound, in the first two days of infection. This was followed by an increase in cells labeled only with PI, indicating secondary necrosis (Additional file 3, Figure S3).
In conclusion, this study provides the first demonstration of mycoplasma-induced apoptosis of chondrocytes. This process may lead to the development of M. synoviae-induced infectious synovitis in the chicken, a disease that in many factors resembles human infectious arthritis. It also supports the hypothesis that mycoplasma-induced arthritic conditions in animals might be useful models for understanding the role of mycoplasmas in similar human diseases. Hence, the study may have value outside the immediate interest of avian arthritis.

Additional material
Additional file 1: Figure S1: Expression of genes for which no significant alterations could be detected after exposure of CCH to M. synoviae WVU 1853. Exposure agent and time of exposure (in hours) are indicated below columns. Control (CTRL) represents in all graphs non-exposed CCH. Results are given as mean values ± standard error for three independent cell treatment experiments with three RT-qPCR replicates for each experiment. (file format: EPS).
Additional File 2: Figure S2: Immunodetection of caspase 3 protein in CCH cell lysates. CCH were infected with M. synoviae WVU 1853 for 24, 48 and 72 h. Cells were then lysed in IP lysis buffer (10 mM Tris HCl pH 7.4, 150 mM NaCl, 0.1% NP-40, 0.002 M EDTA) with 0.1% protease inhibitor cocktail (Sigma) and kept at -20°C. Protein concentration was determined by adding 160 μL of Bradford reagent (Bio-Rad) to 40 μL of cell lysate samples diluted 1:100 in miliQ water and comparing A 595 values to those obtained for bovine serum albumin (BSA) solutions with known concentrations. Cell lysates containing 60 μg of total proteins were prepared for polyacrilamide gel electrophoresis by adding 2 μL of dithiothreitol (DTT; RD Systems) and 5 μL of loading buffer (Fermentas) to 18 μL of sample. Samples were separated on a 12% polyacrylamide gel (30% acrylamide mix, ammonium persulfate, N, N, N', N'tetramethylethylenediamine, Trizma base from Sigma-Aldrich, sodium dodecyl-sulfate from Merck) and transferred onto a polyvinyl-difluoride membrane (Imobilon-P, Millipore) in electroblotting buffer (N-cyclohexyl-3-aminopropanesulfonic acid, MetOH from Sigma-Aldrich) by applying 0.8 mA/cm 2 of gel for 45 min. The membrane was blocked overnight in 3% BSA at 4°C and incubated in rabbit monoclonal antibodies to human caspase 3 or α/β tubulin (1:2000 in Tween20 solution in phosphate buffered saline pH 7.0 (Tween-PBS), both from Cell Signaling Technology, USA) for 2 h at room temperature. Following washing in 0.05% Tween-PBS, horseradish peroxidase-labeled goat anti rabbit IgG (1:5000 in 0.05% Tween-PBS, Sigma-Aldrich) were used as secondary antibodies. BM chemiluminiscence blotting substrate (Roche) was used for detection on film paper (reagents from Ilford). Note that increased synthesis of caspase 3 is evident, although the antibody failed to recognize 17 kDa fragments of cleaved enzyme.
Additional file 3: Figure S3: Flow cytometry analysis of Annexin V-FITC and propidium iodide stained non-infected CCH (a, b, c, g) and CCH infected with M. synoviae WVU 1853 (d, e, f, g). CCH were infected as described in the Materials and methods section, washed in PBS and scraped using a rubber policeman. Washed cells were resuspended in 100 μL 1 × binding buffer (10 × stock solution: 0.1 M HEPES pH 7.4, 1.4 M NaCl, 25 mM CaCl 2 ) and stained by adding 5 μL Anexin V-FITC (BD Pharmingen) and 10 μL propidium iodide (PI, 10 μg/ mL stock, Sigma-Aldrich). Cells were incubated at room temperature for 15 min, followed by addition of another 400 μL of 1 × binding buffer. In order to stain nuclei of live cells, DAPI (final concentration 3 μM, Invitrogen) was used just before flow cytometric analysis. Single color controls for Annexin V-FITC, PI and DAPI were used to set compensations. Annexin V-FITC was detected in the 536/40 nm channel after excitation with the 488 nm blue laser (50 mW). PI was detected in the 675/25 nm after excitation with the 488 nm blue laser (50 mW). DAPI was detected in the 455/25 nm channel after excitation with the 405 nm violet diode laser (100 mW). With this set up no compensation was needed. Cells were analyzed on a CyFlow Space (Partec) fitted with a 488 nm blue laser and violet diode 405 nm with FlowMax software. DAPI positive and CCH positive (based on FSC/SSC) signals were used for gating. Live cells were defined as AV -PI -, early apoptotic as AV + PI -, late apoptotic as AV + PI + , and primary/secondary necrotic cells as AV -PI + .