Intrauterine bacteriological samples were collected using the cytobrush technique from Holstein-Friesian cows on a commercial dairy farm in Schleswig-Holstein (Germany), as described previously (Wagener et al., 2014b, 2015). The strains of T. pyogenes used in this study were isolated from cows on day 15 postpartum (pp). One T. pyogenes strain (TP2) was isolated from a cow that showed vaginal discharge with more than 50% pus on day 21 pp and was classified accordingly as a cow developing CE. Another T. pyogenes strain (TP5) was isolated from a cow that showed clear vaginal discharge on day 21 pp and was classified accordingly as a healthy cow. The bacteria were cultivated in 5 ml brain heart infusion (BHI) broth (Fluka, Steinheim, Germany), supplemented with 5% heat-inactivated (HI) fetal calf serum (FCS; Biochrom, Berlin, Germany) at 37°C for 48 h under aerobic conditions. The bacteria were harvested by centrifugation at 2,000 × g for 10 min and washed once with Dulbecco's phosphate-buffered saline (PBS; Sigma-Aldrich, Deisenhofen, Germany). Both strains (TP2, TP5) were stored in aliquots containing 80% (v/v) bacteria suspended in PBS and 20% (v/v) glycerol at −80°C until further use. Some aliquots were thawed and plate counting was performed on sheep blood agar (Oxoid, Hampshire, United Kingdom) to determine the number of colony-forming units (CFU)/ml.

Heat-inactivated (HI) bacteria and bacteria-free filtrates (BFF) of both T. pyogenes strains (TP2, TP5) were prepared by enriching them from a glycerol stock in BHI supplemented with 5% FCS for 48 h at 37°C. Bacteria were harvested by centrifugation at 2,000 × g, washed with PBS and incubated in epithelial cell culture medium [Dulbecco's Modified Eagle Medium (DMEM)/Ham's F-12 medium containing 10% FCS (both Biochrom)] at 37°C and 5% CO₂ until the medium changed color to yellow. To heat inactivate the bacteria, a portion of this bacterial suspension was incubated at 63°C for 30 min, centrifuged at 2,000 × g for 10 min, washed once with PBS, re-suspended in PBS, and stored in aliquots at −80°C until further use. To obtain the BFF, another portion of the bacterial suspension was centrifuged at 2,000 × g for 10 min. The supernatant was filtered through a 0.22 μm filter (Rotilabo syringe filters, Carl Roth, Karlsruhe, Germany) and stored in aliquots at −80°C until further use. Aliquots of the bacterial suspension were used for plate counting on sheep blood agar (Oxoid) to determine the number of CFU/ml for HI bacteria or equivalents to BFF. A sample of the prepared HI bacteria and BFF was cultured on sheep blood agar plates and incubated at 37°C for 48 h. Absence of bacterial colonies indicated complete inactivation and absence of any live bacteria.

Both T. pyogenes strains (TP2 and TP5), reactivated from glycerol stocks, were pre-grown in 5 ml BHI broth supplemented with 5% HI FCS at 37°C under aerobic conditions for 48 h. The optical density of each bacterial suspension was measured at the wavelength of 600 nm (OD₆₀₀) and bacteria were diluted into 50 ml BHI broth supplemented with 5% HI FCS to obtain a starting OD₆₀₀ of 0.05, which was confirmed by OD₆₀₀ measurement. Bacterial cultures were incubated at 37°C under aerobic conditions and OD₆₀₀ was measured every 2 h over a period of 24 h. The experiment was repeated independently five times.

In addition, samples from three of the independent growth experiments were taken after 6, 12, and 24 h for transcriptional analyses of selected virulence genes, as described in detail below. Cells were harvested by centrifugation at 2,000 × g for 10 min and washed once with PBS. The resulting bacterial cell pellets were stored at −80°C until further use.

FTIR spectroscopy was employed to generate bacterial metabolic fingerprints, as described previously (Ehling-Schulz et al., 2005). Briefly, bacteria were grown as lawns on tryptic soy agar (Oxoid) for 24 h at 30°C. FTIR measurements were carried out using a Tensor 27 FTIR spectrometer (Bruker Optics, Billerica, MA, USA) coupled to a HTS-XT microplate adapter. Data analysis of FTIR spectra was performed using the OPUS software (version 5.5; Bruker Optics). First derivatives of the original spectra were calculated and spectral windows of 3,030 to 2,830, 1,350 to 1,200, and 900 to 700 cm⁻¹ were used with a weight factor of 1 and a repro-level 30 for the cluster analysis. Hierarchical cluster analysis (HCA) was performed as described previously (Wagener et al., 2014a). Dendrograms were calculated using Ward's algorithm and Streptococcus uberis as an outgroup. To identify the spectral regions with the most significant differences between TP2 and TP5, normalized average spectra of the three independent FTIR measurements were calculated for both strains, and the average spectra of TP2 and TP5 were compared.

Bacterial genomic DNA was extracted using the RTP Bacteria DNA Mini Kit (Stratec, Berlin, Germany) according to the manufacturer's instructions. For this purpose, cell pellets obtained from the growth curve experiments of the two different T. pyogenes strains were used. DNA yield was estimated by spectrophotometry at a wavelength of 260 nm.

A conventional PCR was carried out, as described previously (Gärtner et al., 2016), using 150 ng of genomic bacterial DNA. Primer pairs were designed using Primer-BLAST (Ye et al., 2012) and synthesized with Eurofins Genomics (Ebersberg, Germany). The amplification was performed using cycling conditions as follows: 10 min at 95°C, 40 cycles of 15 s at 95°C, 30 s at the specific annealing temperature for each primer pair (Table 1) and 30 s at 72°C, followed by a final extension step at 72°C for 4 min. PCR without DNA served as a negative control. The identity of each amplicon was further confirmed by DNA sequencing (GATC Biotech, Konstanz, Germany). Based on sequence data, specific primers for fimA, fimC, fimE, and fimG (Table 1) were designed for the subsequent quantitative PCR.

Bovine endometrial epithelial cells were isolated and cultured as described previously (Betts and Hansen, 1992; Gärtner et al., 2015). Briefly, apparent healthy uteri from non-pregnant cows were obtained from a local abattoir. Small pieces of the endometrium were cut from various loci in the intercaruncular regions. The tissue mass was chopped very finely before being digested in an enzyme solution composed of 150 U/ml collagenase (Sigma-Aldrich), 150 U/ml hyaluronidase (Sigma-Aldrich), 200 U/ml penicillin (Biochrom), and 200 μg/ml streptomycin (Biochrom) in Hank's Balanced Salt Solution (Biochrom) for 2 h at 37°C. After centrifugation, the cell pellet obtained was washed in epithelial cell culture medium (DMEM/Ham's F-12 medium containing 10% FCS, 55 μg/ml gentamicin, and 1.4 μg/ml amphotericin B; all from Biochrom). Then, the cells were seeded in 25 cm² flasks (Corning, Corning, NY, USA) and incubated in a humidified incubator at 37°C and 5% CO₂ for 18 h to allow selective attachment of stromal cells. After this time, the medium containing unattached cells was re-seeded to obtain a pure epithelial cell population (Sheldon et al., 2010).

In the first passage, endometrial epithelial cells were cultured until they reached more than 80% confluence. At that time, they were seeded into 24-well plates at a density of 1 × 10⁵ cells in 0.5 ml culture medium or in 6-well plates at a density of 3 × 10⁵ cells in 3 ml culture medium.

After the endometrial epithelial cells reached confluence in the second passage, they were prepared for co-culture experiments by removing the medium and washing twice with PBS.

Immunocytochemistry against pan-keratins was performed as described previously (Miessen et al., 2012). After the second passage, a 100% pure epithelial cell population was observed.

Whole cow blood collected at the slaughterhouse was used for PBMC isolation by density gradient centrifugation using Ficoll-Paque plus (GE Healthcare, Uppsala, Sweden). The isolated PBMCs were pre-incubated with epithelial cell culture medium at a density of 5 × 10⁶ cells/ml for 24 h in a humidified incubator at 37°C and 5% CO₂ to adjust the cells to culture conditions.

After the 24 h incubation, the PBMCs were washed twice with PBS and suspended in antibiotic-free epithelial cell culture medium before co-culturing with endometrial epithelial cells and/or T. pyogenes.

In experiment 1, endometrial epithelial cells (n = 5 cows) were co-cultured in passage 2 in 24-well plates with bacterial strain TP2 or TP5 in the form of live bacteria at a multiplicity of infection of 1 (MOI = 1), HI bacteria at MOI = 1, or BFF equivalent to MOI = 1.

In experiment 2, endometrial epithelial cells were co-cultured (n = 5 cows) in 6-well plates with live TP2 at MOI = 1 and/or heterogeneous PBMCs at a ratio of 1:1 to endometrial epithelial cells.

Different treatments were applied using antibiotic-free epithelial cell culture medium. Epithelial cells cultured with epithelial cell culture medium without antibiotics served as controls. After the indicated time points in Figures 5, 6, the medium was removed, and the epithelial cells were washed twice with PBS and lysed with RLT lysis buffer (Qiagen, Hilden, Germany). Control epithelial cells were also lysed at 0 h. Cell lysates were stored at −80°C until further use. In addition, the viability of the epithelial cells (n = 3) was monitored in 24-well plates for all treatments and corresponding controls by Trypan blue staining up to 72 h of co-culture, as described previously (Gärtner et al., 2016). Different optical fields were considered, pictures were taken and analyzed using ImageJ version 1.51 (National Institutes of Health, USA). The number of viable (unstained) cells was calculated as the percentage of the total number of cells.

In experiment 3, the hypothesis that PBMCs can influence the growth of T. pyogenes in lower concentrations was tested. PBMCs (n = 3 cows) were seeded after pre-incubation in 6-well plates at a density of 1 × 10⁷ cells in antibiotic-free epithelial cell culture medium and co-cultured with TP2 at MOI = 0.1. After 2, 4, and 6 h, PBMCs were harvested by centrifugation at 850 × g for 5 min and washed once with PBS. The cell pellets were stored at −80°C until use. In addition, the viability of the PBMCs (n = 3) co-cultured with TP2 at two different MOI (0.1 and 1) was monitored up to 48 h using a Neubauer chamber and Trypan blue exclusion.

Total RNA was isolated from lysed co-cultured endometrial epithelial cells, as described previously (Gärtner et al., 2016). Lysed PBMCs were subjected to total RNA extraction using the InviTrap Spin Cell RNA Mini Kit (Stratec) according to the manufacturer's instructions.

Bacterial total RNA was extracted using the RNeasy Plus Mini Kit (Qiagen) according to the manufacturer's protocol with some modifications. Briefly, a frozen bacterial cell pellet was suspended in 100 μl 10 mM Tris-HCl—1 mM EDTA buffer (pH 8). Then, 500 μl lysis buffer RLT (Qiagen) containing 1% of 2-mercaptoethanol (Carl Roth) was added to the cell suspension and this was transferred in its entirety to a 2-ml screw cap microtube (Sarstedt, Nümbrecht, Germany) containing 300 mg of 0.1 mm silica spheres (Lysing Matrix B; MP Biomedicals, Eschwege, Germany). The bacterial cells were lysed by shaking the tube using a Fastprep machine (FP120, Savant Instruments, New York, USA) at maximum speed for 30 s for 4 cycles with 2 min cooling on ice between consecutive cycles. The cell debris was removed by centrifugation at 11,000 × g for 1 min and the supernatant containing total RNA was loaded on an RNeasy spin column.

The total RNA yield from eukaryotic and bacterial cells was estimated spectrophotometrically at 260 nm. RNA integrity and quality was verified using an RNA 6000 Nano LabChip kit on an Agilent 2100 Bioanalyzer (both Agilent Technologies, Waldbronn, Germany).

Reverse transcription (RT) was performed as described previously (Odau et al., 2006). Briefly, 150 ng total RNA from eukaryotic or bacterial cells was treated with DNAse I (Fermentas, St. Leon-Roth, Germany) in a first step to remove any genomic DNA contamination (Huang et al., 1996). In the second step, first-strand cDNA was synthesized from total RNA using 2.5 μM random hexamers, 0.66 mM dNTPs, 1 × RT buffer and 200 U RevertAid reverse transcriptase (all from Fermentas) in a total volume of 60 μl. Each sample was stored as 20 μl aliquots at −20°C until further use. Reactions without reverse transcriptase were run in parallel to the RT to confirm the absence of any genomic DNA or contamination.

qPCR was carried out as described previously (Odau et al., 2006) following the minimum information for publication of quantitative real-time PCR experiments (MIQE) guidelines (Bustin et al., 2009). Briefly, 1 μl cDNA was amplified in the presence of 1x SensiMix Low-ROX (Bioline, Luckenwalde, Germany) and 0.4 μM of each primer (Tables 1, 2) in a total volume of 10 μl using the Rotor Gene 3000 (Corbett Research, Mortlake, Australia). The qPCR was performed with the following temperature profile: an initial denaturation step at 95°C for 10 min, followed by a cycling step (45 cycles, each cycle consisting of 15 s denaturation at 95°C, 20 s annealing at temperature specified for each gene (Tables 1, 2), and 30 s extension at 72°C), followed by a melting curve program (temperature ramp starting from 50 to 99°C with continuous fluorescence monitoring), and a final cooling step at 40°C for 1 min.

To quantify the content of each expressed gene of interest, a serial dilution of PCR products with a known concentration of the target gene was amplified simultaneously with the samples, generating a standard curve. In relation to the standard curve generated, concentrations of target genes were calculated using the Rotor Gene 6.1 software (Corbett Research). The specificity of amplification of target genes was confirmed by melting point analysis using the Rotor Gene 6.1 software (Corbett Research) and by sequencing (GATC Biotech) of the obtained amplicons, which showed a 100% homology to the published sequences or to the bacterial sequences already obtained in the study.

To normalize the qPCR data obtained, the mRNA expression values of pro-inflammatory factors generated in mammalian cells were divided by an accurate normalization factor. This factor was calculated using geNorm (Vandesompele et al., 2002) for the most stable expressed reference genes in the endometrial epithelial cells [18S ribosomal RNA (18S rRNA), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and succinate dehydrogenase complex, subunit A (SDHA)], in PBMCs [18S rRNA, beta actin (ACTB), and SDHA], and in bacteria [16S rRNA and chromosome segregation protein (smc)].

Due to the large number of samples of endometrial epithelial cells co-cultured with different strains of T. pyogenes, the target genes were amplified in more than one run. Therefore, the normalized values obtained from these cells were inter-run calibrated by dividing using an inter-calibration factor, which was calculated by the geometric averaging of the normalized mRNA expression values of 10 inter-run calibrator samples (Hellemans et al., 2007).

The normalized values obtained for pro-inflammatory factor mRNA expression in the endometrial epithelial cells co-cultured with T. pyogenes and/or PBMCs, and PBMCs and the normalized mRNA expression values of virulence genes in different T. pyogenes strains were log-transformed.

Normalized inter-run calibrated values and normalized log-transformed values in controls and treatments at each time point were scaled relative to the mRNA expression value at 0 h, set equal to one. These relative values were used to generate bar charts, presenting means ± SEM.

An independent t-test was undertaken to compare the differences between the growth rate and normalized mRNA expression of the virulence genes of two T. pyogenes strains (TP2 vs. TP5) at different time points. One-way repeated measures analysis of variance (ANOVA) with the Bonferroni post-hoc test was used to compare the differences between the normalized mRNA expression of virulence factors at three different time points within each T. pyogenes strain. The Wilcoxon signed-rank test was used to calculate statistical differences between the treatments with different T. pyogenes strains. For this, each treatment was compared with the control at the same time point, and the same treatments of each strain (TP2 vs. TP5) were also compared. General linear model multivariate ANOVA with the Bonferroni post-hoc test was used for multiple comparisons of different treatments with T. pyogenes and/or PBMCs. A paired t-test was performed to compare the effect of T. pyogenes on PBMCs (untreated vs. treated).

All statistical calculations were undertaken using SPSS version 22 (SPSS, Chicago, USA). Values of P ≤ 0.05 were considered to be significant; P-values of 0.05–0.1 were considered to indicate tendencies.

T. pyogenes strain TP2 grew faster than TP5 during the first hours (Figure 1), reaching a significant difference at 6 h (P = 0.043). However, both strains reached their plateau phase of growth after 18 h. In addition, TP2 grew as a homogeneous suspension during the entire 24 h period in contrast to TP5, which grew up to 14 h rather as a cloudy suspension and thereafter as a homogeneous one.

FTIR spectroscopy was used as a high-resolution vibrational spectroscopic technique to generate metabolic fingerprints of the two T. pyogenes strains included in the study (Figure 2). FTIR spectral analysis of the average spectra generated from the multiple independent measurements of the strains revealed the most pronounced differences in the metabolic profiles of TP2 and TP5 in the amide/protein region (wavenumber 1,800 to 1,500 cm⁻¹) and polysaccharide region (wavenumber 1,200 to 900 cm⁻¹). The most prominent differences between the two strains were located at wavenumber 1,658 cm⁻¹ (amide I region), 1,544 cm⁻¹ (N = O bonds), and in the polysaccharide region at wavenumber 1,090 to 980 cm⁻¹.

All the virulence genes investigated, except nanP, were present in the genome of the two T. pyogenes strains (data not shown). The nanP gene was only present in the bacterial genomic DNA of TP5, not of TP2.

According to the BLASTN analysis, the partial sequences of all virulence factors showed homologies to the sequences (GenBank accession numbers in parenthesis), as follows: cbpA 100% (AY223543.1), fimA 98% (CP007003.1 and CP007519.1), fimC 99% (CP007003.1 and CP007519.1), fimE 96–99% (CP012649.1, CP007003.1, and CP007519.1), fimG 98–99% (CP007003.1, CP007519.1, and CP012649.1), nanH 93–98% (AF298154.1, CP007003.1, and CP007519.1), nanP 97–100% (CP007003.1, CP012649.1, and AY045771.1), and PLO 97–99% (KJ150329.1, CP007519.1, and AB027461.1).

Different mRNA expression of most virulence genes was noted between the two T. pyogenes strains (Figure 3). In detail, transcripts of PLO were detected during all time points with no significant differences between the two strains (Figure 3A). However, PLO mRNA expression was time-dependent in TP2 but not in TP5. mRNA expression of PLO in TP2 was significantly—20- and 7-fold—higher after 24 h compared with 6 and 12 h, respectively.

The cbpA gene was present in the genomes of both strains. However, cbpA mRNA expression was only detected in TP2, not in TP5 (Figure 3B), and showed a tendency for decreased mRNA expression after 24 h compared with 6 h (P = 0.057).

nanH mRNA expression was observed in TP2 and TP5 at all time points (Figure 3C). The transcription level of nanH in TP5 was similar during the 24 h cultivation. In contrast, nanH mRNA expression in TP2 increased 3-fold from 6 to 12 h, reaching a 4-fold lower level after 24 h. However, this finding failed to reach statistical significance (P > 0.1). After 6 and 12 h, nanH transcription in TP2 was significantly higher (P = 0.004 and 0.014, respectively) compared with TP5. nanH mRNA expression was 2-fold higher in TP2 than in TP5 after 24 h, but this difference did not reach statistical significance (P = 0.068).

Nearly, the same transcript amounts of nanP were detected in TP5 at the selected time points, but none in TP2 not containing this gene in its bacterial genome (Figure 3D).

Transcripts of fimA were detected in TP2 and TP5 at all time points investigated, with no significant differences between the two strains (Figure 3E). mRNA expression of fimC was observed during the entire cultivation time in both the T. pyogenes strains investigated (Figure 3F). A time-dependent fimC mRNA expression was noted for TP2, but not for TP5. The amount of fimC mRNA in TP2 was lower after 24 h (4-fold) compared with 6 h (P = 0.027) and 12 h (P = 0.057). Nearly the same fimC transcript amounts were detected in TP2 and TP5 after 6 and 12 h. However, mRNA expression of fimC after 24 h tended to be about 3-fold higher in TP5 compared with TP2, but failed to reach statistical significance (P = 0.051).

Similar to cbpA, the gene of fimE was present in the genomes of both strains. However, fimE transcripts were only detected in TP2, not in TP5 (Figure 3G). In addition, fimE mRNA expression was time-dependent and tended to increase (P = 0.06) from 6 h to 2-fold higher values at 12 h and decreased about 4-fold (P = 0.002) after 24 h compared with 12 h.

Transcripts of fimG were detected in TP2 and TP5 (Figure 3H). After 6 h, fimG mRNA expression was significantly higher in TP2 compared with TP5. No differential fimG mRNA expression was noted between the two strains after 12 and 24 h. The mRNA expression pattern of fimG in TP2 decreased with cultivation time, with about 15- and 4-fold lower fimG mRNA expression after 24 h compared with 6 and 12 h, respectively.

The effect of live bacteria, BFF (representing exotoxins), and HI bacteria (representing endotoxins) of both T. pyogenes strains on bovine endometrial epithelial cells was evaluated (Figure 4A). Similar cytotoxicity on endometrial epithelial cells was elicited by the live form of each of the two strains of T. pyogenes, which caused death in >90% of the cells within 16 h. This was not observed after 8 h, when >95% cells were still viable in treated and control epithelial cells (data not shown). In contrast, HI bacteria and BFF of the two strains did not elicit a cytotoxic effect on endometrial epithelial cells up to 72 h.

The viability of bovine endometrial epithelial cells co-cultured with the T. pyogenes strain TP2 isolated from a cow developing CE in the presence/absence of PBMCs was investigated (Figure 4B). More than 90% of the endometrial epithelial cells were dead after 16 h of co-culture with live TP2 alone, or with TP2 and PBMCs together, compared with the control. The presence of PBMCs did not affect the viability of endometrial epithelial cells compared with the control.

In addition, the viability of bovine PBMCs co-cultured with T. pyogenes strain TP2 at different MOI (0.1 and 1, respectively) was evaluated. After 8 h, the percentage of live PBMCs co-cultured with TP2 recovered at different MOI was lower compared with the percentage of recovered PBMCs incubated with control culture medium (data not shown). After 16 h, 100% of PBMCs were dead in the presence of TP2 at an MOI of 0.1 and 1 compared with control cells (data not shown).

The mRNA expression of PTGS2 in endometrial epithelial cells was influenced significantly by the presence of TP2 but not by TP5 (Figure 5A). In detail, a significantly (3-fold) higher amount of PTGS2 mRNA was observed in epithelial cells co-cultured with live TP2 after 8 h compared with the control. The PTGS2 mRNA expression tended to decrease in the presence of live TP5 after 2 h (P = 0.08). In addition, the presence of live TP2 resulted in significantly higher PTGS2 mRNA expression of endometrial epithelial cells, 1.5- and 2-fold after 2 and 8 h respectively, compared with live TP5. Endometrial epithelial cells co-cultured with TP2 BFF after 6 h showed a lower (P < 0.05) PTGS2 mRNA expression compared with untreated cells. Furthermore, transcript amounts of PTGS2 tended to decrease (P = 0.08) in the presence of HI or BFF TP5 after 6 h.

The presence of live TP2 did not affect CXCL3 mRNA expression, with similar transcript amounts in endometrial epithelial cells compared with non-treated controls during the 8 h time period (Figure 5B). However, in presence of live TP5, significantly lower CXCL3 mRNA expression was observed compared with controls after 2 h. Therefore, the mRNA expression of CXCL3 in endometrial epithelial cells 2 h co-cultured with live TP2 was higher (P < 0.05) compared with live TP5.

During the 8 h co-culturing period, IL1A mRNA expression in endometrial epithelial cells was not significantly affected by the presence of live or HI TP2 and TP5, respectively (Figure 5C). However, lower IL1A mRNA expression (P < 0.05) was observed after 6 h in the presence of BFF TP5 compared with control cells, but not for BFF TP2.

IL6 showed differences in the mRNA expression pattern in endometrial epithelial cells co-cultured with TP2 or TP5 at different time points (Figure 5D). After 2 h, IL6 mRNA expression was significantly lower in the presence of live TP5 in comparison with untreated cells, but this was not observed for TP2. Significantly higher mRNA expression of IL6 was detected after 8 h in the presence of live TP2 and live TP5 compared with control cells. Endometrial epithelial cells co-cultured with HI TP2 showed slightly lower (P < 0.05) IL6 mRNA expression after 6 h compared with untreated cells. In contrast to the other pro-inflammatory factors, a lower mRNA expression of IL6 in endometrial epithelial cells was observed in the presence of BFF TP2 compared with BFF TP5 after 8 h.

IL8 mRNA expression in endometrial epithelial cells was similar in the presence of TP2 in the form of HI or BFF compared with control cells at each time point during the 8 h co-culture period (Figure 5E). However, IL8 mRNA expression after 8 h tended to be higher (P = 0.08) in epithelial cells in the presence of live TP2 compared with the control. In addition, only a numerical decrease (P = 0.08) of IL8 mRNA expression was observed after 2 h in presence of live TP5 in comparison with untreated cells. Endometrial epithelial cells showed a significantly higher IL8 mRNA expression in presence of live and HI TP2 compared with live and HI TP5 after 2 h, respectively.

To monitor the absence/influence of any contamination with immune cells, mRNA expression of CD45 was evaluated. Transcripts of CD45 were not detected in almost all endometrial epithelial cell samples co-cultured with PBMCs. The melting curve analysis showed no or non-specific amplification except for six samples co-cultured with PBMCs, the threshold cycle value of which was >38 (data not shown).

Nearly, the same transcript amounts of PTGS2 were noted in endometrial epithelial cells co-cultured with PBMCs compared with controls during different time points (Figure 6A). Endometrial epithelial PTGS2 mRNA expression was 4-fold higher in presence of TP2 alone after 8 h compared with untreated cells controls but failed to reach a significant difference (P = 0.096). However, endometrial epithelial cells in the presence of TP2 and PBMCs showed significantly higher PTGS2 mRNA expression (9-fold) after 8 h compared with controls.

The mRNA expression of CXCL3 in endometrial epithelial cells was not affected significantly by the presence of PBMCs alone compared with untreated cells throughout the entire time course of co-culturing (Figure 6B). There was significantly higher CXCL3 mRNA expression in endometrial epithelial cells co-cultured with TP2 and PBMCs compared with the untreated controls (P < 0.001), but not with TP2 alone (P = 0.17), after 4 h. However, greatly higher mRNA expression of CXCL3 (50- and 90-fold) was observed in endometrial epithelial cells after 8 h co-cultured with TP2 compared with untreated cells and cells co-cultured with PBMCs, respectively. This effect was further amplified in endometrial epithelial cells in the presence of TP2 and PBMCs, which showed a 90- to 170-fold increase of CXCL3 mRNA expression in comparison with controls.

There was no significant effect in the presence of TP2 or PBMCs, either alone or combined, on IL1A mRNA expression in endometrial epithelial cells after 2 h of co-culture (Figure 6C). After 4 h, epithelial cells co-cultured with TP2 and PBMCs showed an 8-fold higher IL1A mRNA expression compared with untreated cells. This increase failed to reach statistical significance (P = 0.06) due to the high variability between the animals. However, mRNA expression of IL1A in endometrial epithelial cells after 8 h was significantly higher with TP2 alone (35- and 60-fold), or TP2 and PBMCs (90- and 140-fold), when compared with untreated cells and cells co-cultured with PBMCs, respectively.

The IL6 mRNA expression pattern was similar in treated endometrial epithelial cells and the controls after 2 h (Figure 6D). Although the mRNA expression of IL6 did not increase in response to TP2 alone after 4 h, higher mRNA expression in epithelial cells (P < 0.05) was observed in the presence of TP2 and PBMCs compared with untreated cells. After 8 h, TP2 alone, or TP2 and PBMCs together induced higher IL6 mRNA expression in endometrial epithelial cells compared with the controls. In addition, IL6 mRNA expression in endometrial epithelial cells co-cultured with TP2 and PBMCs was 4-fold higher (P < 0.05) in comparison with epithelial cells co-cultured with TP2 alone after 8 h.

IL8 mRNA expression exhibited no differences between treated endometrial epithelial cells and controls after 2 h (Figure 6E). In comparison with the control cells, higher IL8 mRNA expression was observed in epithelial cells co-cultured with TP2 and PBMCs after 4 h. After 8 h, endometrial epithelial cells co-cultured with TP2 showed a 30-fold increase in IL8 mRNA expression (P < 0.001); this increase was further amplified to 70-fold in the presence of PBMCs.

PBMCs co-cultured with the T. pyogenes strain TP2 at MOI = 0.1 showed higher mRNA expression of some selected pro-inflammatory factors. In detail, PBMCs co-cultured with TP2 showed a 2-fold increase (P < 0.05) of PTGS2 mRNA expression as early as after 2 h in comparison with the controls (Figure 7A), but not after 4 or 6 h.

The presence of TP2 affected the mRNA expression of CXCL3 in PBMCs (Figure 7B). CXCL3 mRNA was three times more highly expressed (P < 0.05) in PBMCs co-cultured with TP2 in comparison with the controls after 4 h. However, this increase was not observed at 2 or 6 h.

Nearly, the same transcript amount of IL1A was detected in PBMCs co-cultured with TP2 and in the controls after 2 and 4 h (Figure 7C). However, a significantly (5-fold) higher IL1A mRNA expression was observed in PBMCs co-cultured with TP2 in comparison with untreated cells after 6 h.

mRNA expression of IL6 did not change significantly in PBMCs co-cultured with TP2 in comparison with controls at all investigated time points (Figure 7D).

Similar to PTGS2, higher IL8 mRNA expression (P < 0.05) was noted in PBMCs treated with TP2 in comparison with controls as early as after 2 h (Figure 7E). Such higher mRNA expression was not observed at later time points.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.