Introduction

Bovine subfertility is one of the major reasons for enormous economic losses in the dairy industry1. Impaired reproductive performance such as prolonged intervals from calving to conception can be associated with uterine diseases of the postpartum period, e.g. subclinical endometritis (SCE)2,3.

Current treatments of SCE include antimicrobials and prostaglandin (PG) F2α4,5. However, unwanted consequences of the administration of antimicrobials are residues in milk and meat6 and the potential spread of bacterial resistances7. Furthermore, the efficacy of a hormonal therapy with PGF2α remains controversial. Administration of a single dose or 2 treatments with PGF2α at 35 and 49 (±3) days in milk (DIM) did not affect the prevalence of SCE and purulent vaginal discharge8,9. The reproductive performance was neither improved by antibiotic nor PGF2α treatment9,10. Therefore, an alternative strategy for the effective treatment of SCE to improve the fertility rates without the unwanted implications of the use of antimicrobials would be of great benefit.

In human medicine, an alternative to the use of antimicrobials for the treatment of bacterial vaginosis in women is the use of lactobacilli11. Characteristics of the Gram-positive lactobacilli are the capacity to produce acetic and lactic acid, hydrogen peroxide and bacteriocins, which are rated to be beneficial for the suppression of pathogenic bacteria12.

From samples of the bovine uterus, several Lactobacillus spp. were cultivable13,14,15,16 and also detected by metagenomic pyrosequencing of the 16S rRNA gene17,18. It was shown that co-culturing of bovine endometrial epithelial cells with L. buchneri (now registered as L. buchneri DSM 32407) up to a multiplicity of infection (MOI) of 10 did not affect the viability of epithelial cells16. In addition, the mRNA expression or release of pro-inflammatory factors was not influenced for up to 6 h and 48 h, respectively. In contrast, the presence of L. ruminis and L. amylovorus provoked a pro-inflammatory response of the epithelial cells. An early study indicated that lactobacilli have an immunostimulatory effect on the endometrium19. In that study, after the intrauterine administration of two live Lactobacillus spp., an infiltration with mostly mononuclear cells into the endometrium was observed and a colonization of the endometrium by the selected lactobacilli strains for up to 12 days was noted. However, the impact of an intrauterine administration beyond the endometrial infiltration with immune cells remains unclear, especially referring to the uterine health status.

One possibility to draw conclusions on the uterine health status is the analysis of the mRNA expression of pro-inflammatory factors such as interleukins, chemokines, and enzymes of the PG synthesis. An elevation of the mRNA expression of such selected factors in case of SCE was found in several studies20,21,22,23,24. In the present study, the exact same time points of the puerperium were chosen for sampling as in a previous study where the levels of mRNA expression of such factors showed correlations with SCE24. Therefore, the mRNA expression analysis of pro-inflammatory factors seemed suitable to support the evaluation of endometrial health.

The objective of the present study was to clarify the influence of the intrauterine administration of L. buchneri DSM 32407 on the reproductive performance of clinically healthy cows (SCE or healthy). Therefore, cows were divided into two groups on days 24–30 postpartum (pp) and either the autochthonous L. buchneri DSM 32407 or a placebo was administered intrauterine. In addition, all cows with SCE were monitored for their uterine health status and the endometrial mRNA expression pattern of selected pro-inflammatory factors on a weekly basis during the following days 31–51 pp.

Materials and Methods

Preparation of L. buchneri solutions for intrauterine administration

The strain L. buchneri DSM 32407 was isolated from a uterus of a healthy cow and stored long-term at −80 °C in 15% (v/v) glycerol in MRS broth (according to DeMan, Rogosa and Sharpe; Sigma-Aldrich, Steinheim, Germany)16. This stock was used for enrichment by cultivation under aerobic conditions in MRS broth (Sigma-Aldrich) at 37 °C for 48 h until the suspension reached an optical density of one at the wavelength of 600 nm. After centrifugation for 10 min at 15000 × g and resuspension in MRS broth with 50% (v/v) glycerol, 200 µl aliquots were stored at −80 °C until further use.

36–48 h prior to the preparation of the solutions for intrauterine administration, aliquots were thawed at room temperature. To calculate the number of colony forming units (cfu)/ml of the aliquots, serial dilutions were cultivated on Rogosa SL agar (Sigma-Aldrich) under microaerophilic conditions (Anaerocult C, Merck, Darmstadt, Germany) at 37 °C. Thawed aliquots were stored at 4 °C until the preparation of the solutions for intrauterine administration.

It was observed that the number of cfu/ml was stable for up to 48 h in the thawed aliquots (stored at 4 °C) and for 8 h in the prepared solutions for intrauterine administration (stored at room temperature).

L. buchneri solutions for intrauterine administration consisted of 1.5–2 × 1010 cfu in 20 ml 0.9% (w/v) isotonic saline solution (B. Braun, Melsungen, Germany) drawn up into a 20 ml plastic syringe (Injekt, B. Braun). Bacteria were prepared at the day of administration (08:00 a.m.), transported to the farm at room temperature and administered latest on the same day after a maximum of 8 h.

Examination and enrollment of cows in this study

Lactating Holstein cows included in this study were kept at the Lehr- und Versuchsanstalt für Tierzucht und Tierhaltung e. V. in Groß Kreutz (Brandenburg, Germany), housing around 200 cows in freestall facilities with slotted floors and cubicles in accordance with the guidelines of the National Animal Welfare Legislation. Animal experimental procedures were approved by the relevant authorities of the state Brandenburg, Germany (Landesamt für Umwelt, Gesundheit und Verbraucherschutz; V3-2347-2-2012 and V3-2347-19-2013).

On days 24–30 pp (Exam 1 [E1]), cows were examined by inspection of the vulva, vaginoscopy, transrectal palpation and ultrasonography (Tringa Linear, Esaote, Köln, Germany) of the uterus and ovaries to determine their uterine health status and stage of the oestrous cycle as reported previously24.

All cows with an initial status of a clinical healthy uterus (no (muco)purulent discharge detectable in the vagina)25 were included in this study (n = 116; 30 primiparous and 86 multiparous). Cows with signs of clinical endometritis (CE) [(muco)purulent uterine discharge detectable in the vagina] at E1 were excluded from the study.

If a corpus luteum was present, cows were considered as being in the luteal phase (n = 95). The absence of a corpus luteum and also a Graafian follicle defined cows to be either prior to their first ovulation after parturition or 1–3 days after ovulation (n = 21). Cows showing signs of oestrus (presence of clear and cohesive mucus discharged from the vulva and/or the presence of a Graafian follicle) were excluded to prevent that the applied bacteria would be flushed out by the mucus.

Intrauterine administration of L. buchneri or placebo and collection of luminal endometrial epithelial samples

After the examination at E1, samples from the endometrial epithelium were obtained with the cytobrush technique from the uterine body as reported previously20,26. Three cytobrush samples were collected. The first cytobrush was used for cytological analysis by rolling the cytobrush on a clean glass microscope slide directly after sampling at the farm. The second cytobrush was collected for the isolation of total RNA and was placed in a cryotube, which was immediately immersed into liquid nitrogen and stored at −80 °C until further use. The third collected cytobrush was used for bacteriological analysis and therefore was placed in an Amies medium containing tube (Heinz Herenz, Hamburg, Germany) and transported to the laboratory at room temperature.

After the sampling, cows were randomly divided into two groups. The prepared L. buchneri solutions were administered intrauterine to cows of the first group (n = 56; Lactobacillus buchneri group [LAC]). 20 ml of isotonic saline solution 0.9% (w/v) per cow as a placebo were administered intrauterine to the cows of the second group (n = 60; placebo group [PLA]). Briefly, the metallic catheter passing the cervix for cytobrush sampling from the uterine body was hold in place after the collection of endometrial samples. A sterilized 62 cm long polytetrafluoroethylene tube with an internal diameter of 2 mm (Rotilabo, Carl Roth) was inserted through the catheter for the aseptic administration of the prepared L. buchneri solutions or the placebo.

After the transportation to the laboratory, slides for cytological analysis were prepared and evaluated as reported previously24. If the content of polymorphonuclear neutrophils (PMN) in the cytological sample was < 5%, the uterine health status of the cows was defined as healthy (H; n = 83). If the content was ≥5%, cows were classified as having SCE (n = 33)20,27. Within the LAC group (n = 56), 40 cows were classified as H and 16 cows were diagnosed with signs of SCE. Within the PLA group (n = 60), 43 cows were classified as H and 17 cows were diagnosed with SCE. Reproduction performance data of all cows treated with either L. buchneri or placebo were recorded and considered for statistical analysis.

Monitoring of uterine health status and collection of luminal endometrial epithelial samples after E1

Cows with SCE, both from the LAC and the PLA group, were further monitored for their uterine health status and sampled for mRNA expression analysis except for one cow from the PLA group because of technical reasons. This results in a distribution of 16 cows per group (LAC/SCE and PLA/SCE).

Each of these cows was examined and sampled three more times in the same manner as described above at E1 in weekly intervals on days 31–37 pp (Exam 2 [E2]), on days 38–44 pp (Exam 3 [E3]), and on days 45–51 pp (Exam 4 [E4]).

Microbiological analysis

At E1, E2 and E4, a cytobrush sample was taken for microbiological analysis from 31 cows of the LAC group (LAC/SCE n = 16 and LAC/H n = 15) and from 17 cows of the PLA group (PLA/SCE n = 15 and PLA/H n = 2), respectively. Within the PLA/SCE group, the sample at E2 could not be taken from one cow for technical reasons. The samples were analysed by aerobic and anaerobic cultivation after direct inoculation of the cytobrush with the suitable agar plates and also after the enrichment in a medium with subsequent cultivation on agar as described previously24.

Biopsy sampling and histopathological examination

Non-pregnant cows (n = 8), which were not in the phase around oestrus and without any signs of clinical endometritis, were sampled for histopathological examination. Briefly, the vulva was cleansed with dry paper towel. The biopsy instrument (Institute for the Reproduction of Farm Animals, Bernau, Germany) was inserted into the uterine body through the cervix. Endometrial tissue samples were taken from the uterine body, followed by intrauterine administration of 1.5–2 × 1010 cfu of L. buchneri in 20 ml 0.9% (w/v) isotonic saline solution (B/LAC; n = 5) or placebo (20 ml 0.9% (w/v) isotonic saline solution; B/PLA; n = 3) in each cow, respectively. The application procedure was the same as described above. Biopsy samples from each cow were taken again one week later.

Pieces of endometrial tissue were immediately immersed into tubes containing Bouin’s solution for transportation. Samples were embedded in paraffin. Two sections of 2 µm thickness each 1000 µm apart were prepared, stained with hematoxylin and eosin, and examined for the presence of immune cells, integrity of the mucosa and proliferation by an observer blinded to the administration of L. buchneri or placebo.

Breeding management

The voluntary waiting period was 60 days pp for most cows included in this study. However, 14 cows (LAC n = 9 and PLA n = 5) were inseminated earlier. Heat detection was done by observation once daily (10:00 a.m.) and by cow activity measured automatically with a respactor (X-ponder, Nedap, Groenlo, Netherlands) and a pedometer (Acto, Insentec, Marknesse, Netherlands). Cows were artificially inseminated within 24 h when heat was detected with deep frozen semen from bulls with similar fertility. Pregnancy diagnosis was performed by transrectal palpation and ultrasonography 30 days after insemination.

Isolation of total RNA and reverse transcription

Total RNA from cytobrush samples harvested from cows with signs of SCE was isolated using the RNeasy Plus Mini Kit (Qiagen, Hilden, Germany), stored at −80 °C, and the integrity of the obtained total RNA was validated as previously reported24.

For the removal of genomic DNA, a DNase treatment was performed before reverse transcription28. Single strand cDNA was generated from 100 ng total RNA per sample with the addition of 200 U RevertAid Reverse Transcriptase and 2.5 µM random hexamer primers (both Thermo Scientific, Schwerte, Germany) in a total volume of 60 µl29. For the confirmation of the absence of any genomic DNA or contaminations, samples without reverse transcriptase were also prepared as negative controls. Generated cDNA was stored in aliquots at −20 °C until further analysis.

Quantitative polymerase chain reaction (qPCR)

qPCR was performed to evaluate the mRNA expression of the candidate genes as reported previously29 following the minimum information for publication of quantitative real-time PCR experiments (MIQE) guidelines30. The genes of interest that were measured by this method are chemokine ligand 1/2 (CXCL1/2), CXCL3, CXCL5, chemokine receptor 2 (CXCR2), interleukin 1 alpha (IL1A), interleukin 1 beta (IL1B), IL6, IL8, IL10, interleukin 1 receptor antagonist (IL1RN), prostaglandin-endoperoxide synthase 2 (PTGS2), prostaglandin E2 synthase 1 (PTGES), PTGES3, prostaglandin D2 synthase (PTGDS), tumour necrosis factor (TNF), matrix metallopeptidase 1 (MMP1), and protein tyrosine phosphatase, receptor type C (PTPRC). Primer pairs were synthesized by Eurofins Genomics (Ebersberg, Germany) and details are given in Supplement Table 1. A gradient-PCR was performed to determine the optimal annealing temperature of unpublished primer pairs and obtained amplicons were subjected to commercial DNA sequencing (GATC Biotech, Konstanz, Germany) to confirm 100% homology to the published bovine sequences29.

Using the Rotor Gene 3000 (Corbett Research, Mortlake, Australia), amplification of 1 µl cDNA per sample was carried out in the presence of 0.4 µM of each primer (forward and reverse) and 5 µl 2 × SensiMix SYBR Low-ROX (Bioline, Luckenwalde, Germany) in a total reaction volume of 10 µl. Denaturation at 95 °C for 10 min was followed by a three-step amplification in 45 cycles: denaturation at 95 °C for 15 s, annealing for 20 s (temperatures depicted in Supplement Table 1), and extension at 72 °C for 30 s. Subsequently, a melting curve program (50–99 °C) with continuous fluorescence measurement confirmed specific amplification. For the generation of a standard curve, a dilution series with known concentrations of the purified amplicons was amplified simultaneously. In comparison with these standard curves, transcript amounts of specific mRNA were calculated using the Rotor Gene 6.1 software (Corbett Research).

Statistical analysis

Reproductive performance was described by proportion of cows sold/culled, proportion of cows pregnant, days to first service, conception at first service, days to conception, and services per pregnancy. Survival curves for the proportion of cows pregnant on day 200 pp in relation to the number of days pp were generated using the Kaplan-Meier survival analysis. For the comparison of the Kaplan-Meier survival curves, three different statistical tests were used. The Breslow test tends to perform best towards the early time points and the Log Rank test to the late time points of the investigated period after calving. The Tarone-Ware test tends to perform best for the middle of this period.

The Fisher’s exact test was used to calculate the incidence of cultivable bacteria in the LAC group in relation to the PLA group at E2 and E4.

The obtained mRNA expression values of the genes of interest was normalized using the geNorm tool31. Succinate dehydrogenase complex, subunit A (SDHA) and suppressor of zeste 12 homolog (SUZ12) were chosen as reference genes. Inter-run calibration was performed based on 10 inter-run calibrator samples using formula 15′32. Box plots were generated presenting the median values with 50% of all data within the box. Outliers (circles; values between 1.5- and 3.0-fold the interquartile range) and extreme values (asterisks; values beyond 3.0-fold the interquartile range) were included in statistical analysis.

Normal distribution was tested with the Shapiro-Wilk test. Neither the percentages of the PMN nor the normalized values of mRNA expression of the genes of interest or the data for reproductive performance were normally distributed. The Mann-Whitney U test was used to analyse values of the mRNA expression comparing samples from the LAC/SCE group with samples from the PLA/SCE group at E2, E3 and E4 and for the data of reproduction comparing the LAC group with the PLA group and the LAC/SCE group with the PLA/SCE group. Fold changes of the mRNA expression levels were calculated as the approximate ratio of the mean value of mRNA expression for the LAC/SCE group to the mean value of the PLA/SCE group.

All statistical evaluations and the generation of the box plots were performed using IBM SPSS Statistics 20.0 (SPSS, Chicago, USA) and the level of significance was set at P ≤ 0.05. A tendency to significance was considered when P was <0.10.

Results

Reproduction performance data

Several data of the descriptive reproductive performance considering all treated cows (SCE and H) are shown in Table 1. The proportion of cows sold/culled before pregnancy was 25% both in the LAC group and in the PLA group. Two cows of the PLA group were sold before breeding and 27 cows were culled due to problems as follows: the locomotor system (LAC group: 1 cow; PLA group: 4 cows), the udder (LAC group: 11 cows; PLA group: 7 cows), the fertility (LAC group: 2 cows), or a low milk yield (PLA group: 2 cows). The proportion of cows becoming pregnant was 75% both in the LAC and in the PLA group. The number of cows inseminated at least once was 96 in total, including 9 cows that were culled before conception (LAC group: 4; PLA group: 5).

Table 1 Descriptive reproductive performance outcomes of 116 cows as well as of 33 cows with signs of SCE after the intrauterine administration of 1.5–2 × 1010 cfu of L. buchneri DSM 32407 diluted in 20 ml isotonic saline solution 0.9% (LAC group) or 20 ml isotonic saline solution 0.9% as a placebo (PLA group) on days 24–30 pp.
Full size table

The median days to first service were lower and the first service conception rate was higher in the LAC group compared with the values of the PLA group. The LAC group had less services per pregnancy and significant shorter median days to conception compared with the data in the PLA group.

Similar differences were observed when only considering cows with signs of SCE (Table 1). 8 cows of the SCE group were culled due to problems as follows: the locomotor system (LAC/SCE group: 1 cow), the udder (LAC/SCE group: 5 cows; PLA/SCE group: 3 cows). The days to first service were similar. However, a higher proportion of cows of the LAC/SCE group was pregnant after the first service and less services per pregnancy were required compared with the PLA/SCE group. The median days to conception for cows pregnant on day 200 pp were 90 days shorter in the LAC/SCE group (P = 0.001) compared with the PLA/SCE group.

Kaplan-Meier survival analysis was performed to show survival curves for the proportion of cows pregnant on day 200 pp in relation to the number of days pp (Fig. 1A). Cows pregnant after day 200 pp were censored. The proportion of cows censored was 7.1% (3 out of 42) for the LAC group and 20% (9 out of 45) for the PLA group. Comparing these groups, cows from the LAC group became pregnant significantly earlier using the Breslow test (P = 0.035) and the Tarone-Ware test (P = 0.047), whereas the Log Rank test (P = 0.06) showed a tendency to significance. The Kaplan-Meier survival analysis is also presented for the cows with signs of SCE to show the proportion of cows pregnant on day 200 pp (Fig. 1B). The cows from the LAC/SCE group became significantly earlier pregnant compared with the PLA/SCE group using the Breslow test, the Tarone-Ware test and the Log Rank test (P = 0.001 each).

Figure 1
figure 1

Kaplan-Meier survival curves for cows pregnant at day 200 pp. (A) The curves show the proportion of cows pregnant (in %) in relation to the number of days pp for the LAC group (n = 42; solid line) and the PLA group (n = 45; dotted line). The proportion of cows censored (not pregnant at day 200 pp) was 7.1% for the LAC group and 20.0% for the PLA group. P values: Breslow test (P = 0.035), Log Rank test (P = 0.06) and Tarone-Ware test (P = 0.047). (B) The curves show the proportion of cows pregnant (in %) in relation to the number of days pp for the LAC/SCE group (n = 16; solid line) and the PLA/SCE group (n = 17; dotted line). The proportion of cows censored (not pregnant at day 200 pp) was 0.0% for the LAC/SCE group and 33.3% for the PLA/SCE group. P values: Breslow test, Log Rank test and Tarone-Ware test (P = 0.001 each).

Full size image

Uterine health status for the LAC/SCE and PLA/SCE groups at E2, E3, and E4

The number of cows in the LAC/SCE group (n = 16 at E1) dependent on their current uterine health status (H/SCE/CE) was distributed at the later time points of monitoring as follows: E2 (n = 10/5/1), E3 (n = 12/2/2), and E4 (n = 12/3/1). The distribution for the PLA/SCE group (n = 16 at E1) was as follows (H/SCE/CE): E2 (n = 13/1/2), E3 (n = 13/1/2), and E4 (n = 11/2/3).

After day 31 pp, the uterine health status of most cows was diagnosed as H in both groups. However, there were still 5 cows with signs of SCE in the LAC/SCE group in comparison to only 1 cow in the PLA/SCE group at E2. At E3, all of these cows were diagnosed as H.

Two cows of the PLA/SCE group showed continuously signs of CE at E2, E3, and E4. In both groups, all other cows showing signs of SCE or CE at E3 and/or E4 developed these diseases after being healthy before, except for one cow of the LAC/SCE group that had signs of CE at E2, SCE at E3, and was healthy at E4.

Content of PMN in endometrial cytobrush samples and microbiological analysis

At E1, the content of PMN in the cytological preparations from endometrial cytobrush samples varied from 5–73.3%. At the later time points, the majority of samples had a percentage of PMN of <5%, both in the PLA/SCE and LAC/SCE groups (Fig. 2).

Figure 2
figure 2

Percentages of PMN in bovine endometrial cytobrush samples harvested from dairy cows on days 24–30 pp (E1; n = 32), on days 31–37 pp (E2), on days 38–44 pp (E3), and on days 45–51 pp (E4). At E2, E3 and E4 cows were divided into the PLA/SCE (n = 16) and the LAC/SCE (n = 16) groups. Extreme values are diagrammed as asterisks, outliners as circles. Extreme values are not shown for one cow at E1 (73.3%) and for one cow of the PLA/SCE group at E2 (65%).

Full size image

No significant differences were observed comparing the content of PMN of the PLA/SCE group with the LAC/SCE group at E2, E3, and E4. However, the number of PMN tended to be higher (P = 0.071) in the LAC/SCE group compared with the PLA/SCE group at E2 (Fig. 2).

At E1, E2 and E4, cultivable bacteria were Trueperella pyogenes, Escherichia coli, Histophilus somni and Streptococcus uberis from few samples of both the PLA/SCE and the LAC group. Klebsiella pneumoniae was only cultivable from one sample of LAC/SCE group. The number of samples that were positive for each bacterial species for these time points are presented in Supplement Table 2.

A significant difference was only observed for H. somni at E2 (P = 0.032) with a higher number of positive samples in the PLA/SCE group compared with the LAC group.

Evaluation of the mRNA expression analysis

Endometrial epithelial mRNA expression of all selected candidate genes was detected at all investigated time points in the postpartum period. However, mRNA expression for IL1A, IL6, IL10, TNF, MMP1 and PTPRC could not be found in each sample.

At E2, E3 and E4, most of the cows (PLA/SCE and LAC/SCE) were during their luteal phase, 1–3 days after ovulation or still prior to their first ovulation. However, seven cows were found during pro-oestrus at different time points (two at E2, two at E3, and three at E4) and two cows were found during oestrus at E3. The different oestrous cycle stages did not have a statistical significant influence on the mRNA expression values.

In the following text, mainly significant differences in normalized mRNA expression are described.

mRNA expression of CXCL1/2, CXCL3, CXCL5, and CXCR2

Primers named CXCL1/2 are specific for CXCL1 as well as for CXCL233. At E4, CXCL1/2 and CXCL3 mRNA was seven- and threefold more highly expressed in the PLA/SCE group compared with the LAC/SCE group, respectively (Fig. 3A–B). In contrast, the mRNA expression pattern of CXCL5 did not show significant differences between the PLA/SCE and LAC/SCE groups at all time points (Fig. 3C). However, a tendency (P = 0.07) was observed at E2 when the CXCL5 mRNA expression was higher in the LAC/SCE group compared with the PLA/SCE group.

Figure 3
figure 3

Normalized mRNA expression of (A) CXCL1/2, (B) CXCL3, (C) CXCL5 and (D) CXCR2 in bovine endometrial cytobrush samples harvested from dairy cows on days 24–30 pp (E1; n = 32), on days 31–37 pp (E2), on days 38–44 pp (E3), and on days 45–51 pp (E4). At E2, E3 and E4 cows were divided into the PLA/SCE (n = 16) and the LAC/SCE (n = 16) groups. Bold asterisks over dotted lines indicate significant differences between the groups (P ≤ 0.05). Extreme values are diagrammed as asterisks, outliners as circles. Extreme values are not shown for CXCL1/2 obtained from three cows at E1 (17.69, 18.75, and 19.87) and from one cow of the PLA/SCE group at E4 (26.63); for CXCL3 obtained from one cow at E1 (22.67); for CXCL5 obtained from one cow of the PLA/SCE group at E2 (13.56) and E4 (21.72) and for CXCR2 obtained from one cow of the PLA/SCE group at E2 (28.62), E3 (110.76), and E4 (109.09).

Full size image

Similar to CXCL1/2 and CXCL3, the mRNA expression of CXCR2 in luminal endometrial epithelium samples was affected at E4. CXCR2 mRNA was 34-fold more highly expressed in the PLA/SCE group compared with the LAC/SCE group (Fig. 3D).

mRNA expression of IL1A, IL1B, IL6, IL8, IL10, and IL1RN

No significant differences were observed for the contents of IL1A mRNA in cytobrush samples of the PLA/SCE group compared with the samples of the LAC/SCE group during all time points (Fig. 4A). In addition, a decrease of the IL1A mRNA expression from E1 to the later time points was observed. In contrast, the mRNA expression of IL1B at E4 was higher (P = 0.05) in samples from cows of the PLA/SCE group compared with the samples obtained from the LAC/SCE group (Fig. 4B).

Figure 4
figure 4

Normalized mRNA expression of (A) IL1A, (B) IL1B, (C) IL6, (D) IL8, (E) IL10 and (F) IL1RN in bovine endometrial cytobrush samples harvested from dairy cows on days 24–30 pp (E1; n = 32), on days 31–37 pp (E2), on days 38–44 pp (E3), and on days 45–51 pp (E4). At E2, E3 and E4 cows were divided into the PLA/SCE (n = 16) and the LAC/SCE (n = 16) groups. Bold asterisks over dotted lines indicate significant differences between the groups (P ≤ 0.05). Extreme values are diagrammed as asterisks, outliners as circles. Extreme values are not shown for IL1A obtained from one cow at E1 (198.24) and from one cow of the PLA/SCE group at E3 (64.07) and E4 (94.13); for IL1B obtained from one cow of the PLA/SCE group at E2 (79.07), E3 (549.78), and E4 (223.11); for IL6 obtained from two cows of the LAC/SCE group at E2 (116.41) and E3 (13.87); for IL8 obtained from two cows at E1 (167.56 and 144.01) and from one cow of the PLA/SCE group at E3 (112.67) and E4 (143.67); for IL10 obtained from one cow at E1 (14.2) and from one cow of the PLA/SCE group at E4 (18.7) and for IL1RN obtained from one cow at E1 (20.89) and from one cow of the PLA/SCE group at E3 (30.99) and E4 (19.59).

Full size image

IL6 mRNA content was similar during the investigated period of the puerperium in all groups (Fig. 4C).

However, IL8 mRNA was about 50-fold more highly expressed in endometrial samples from cows of the PLA/SCE group compared with the LAC/SCE group at E4 (Fig. 4D).

IL10 and IL1RN transcript amount did not differ significantly between the different treatment groups (Fig. 4E–F). However, IL1RN mRNA expression tended to be higher (P = 0.07) at E4 in samples of the PLA/SCE group compared with samples of the LAC/SCE group.

mRNA expression of TNF, MMP1, and PTPRC

The mRNA expression of TNF in endometrial epithelium cells was affected at E2 by the treatment (Fig. 5A). TNF mRNA was threefold more highly expressed in samples obtained from cows of the LAC/SCE group compared with samples of the PLA/SCE group.

Figure 5
figure 5

Normalized mRNA expression of (A) TNF, (B) MMP1, (C) PTPRC, (D) PTGS2, (E) PTGES, (F) PTGES3 and (G) PTGDS in bovine endometrial cytobrush samples harvested from dairy cows on days 24–30 pp (E1; n = 32), on days 31–37 pp (E2), on days 38–44 pp (E3), and on days 45–51 pp (E4). At E2, E3 and E4 cows were divided into the PLA/SCE (n = 16) and the LAC/SCE (n = 16) groups. Bold asterisks over dotted lines indicate significant differences between the groups (P ≤ 0.05). Extreme values are diagrammed as asterisks, outliners as circles. Extreme values are not shown for TNF obtained from two cows at E1 (14.83 and 11.06) and from one cow of the PLA/SCE group at E4 (13.44); for MMP1 obtained from one cow at E1 (56.61); for PTPRC obtained from one cow at E1 (23.53) and from one cow of the PLA/SCE group at E2 (12.86); for PTGS2 obtained from one cow at E1 (138.26); for PTGES3 obtained from one cow at E1 (6.16) and for PTGDS obtained from one cow of the PLA/SCE group at E4 (23.04).

Full size image

MMP1 mRNA contents did not differ between the PLA/SCE group compared with the LAC/SCE group at all time points (Fig. 5B).

Furthermore, PTPRC mRNA was threefold more highly expressed in the endometrial samples of the PLA/SCE group compared with the samples of the LAC/SCE group at E4 (Fig. 5C).

mRNA expression of PTGS2, PTGES, PTGES3, and PTGDS

No significant differences of the mRNA expression were found in the transcript amounts in samples obtained from the LAC/SCE group compared with samples of the PLA/SCE group at E2, E3 and E4 for the analysed key enzymes of the PG synthesis as follows: PTGS2, PTGES, PTGES3, and PTGDS (Fig. 5D,G). However, the LAC/SCE group tended to have a higher PTGDS mRNA expression (P = 0.095) at E3 compared with the PLA/SCE group (Fig. 5G).

Histopathology of biopsy samples

No infiltration with immune cells was observed in all biopsy samples taken before the intrauterine administration of the L. buchneri solutions or the placebo, respectively. In addition, the integrity of the mucosa and the status of proliferation were without pathological findings (Fig. 6A,B).

Figure 6
figure 6

Biopsy samples of the endometrium of one cow from the B/LAC group. Hematoxylin and eosin staining; 2 µm thickness. (A and B) prior to the intrauterine administration of L. buchneri DSM 32407; no infiltration of immune cells, integrity of the mucosa and status of proliferation without pathological findings. (C and D) 7 days after the intrauterine administration of L. buchneri DSM 32407; moderate mainly granulocytic infiltration (arrows) and moderate multifocal degeneration.

Full size image

One week after the intrauterine administration of the L. buchneri solutions or the placebo, respectively, all biopsy samples of the B/PLA group and one sample of the B/LAC group were judged identical to the samples that were taken previously to the administration. In the B/LAC group, a minimal granulocytic infiltration was observed in one sample, a minimal lymphocytic infiltration was observed in two samples, and a moderate mainly granulocytic infiltration was observed in one sample. This sample also showed a moderate multifocal degeneration (Fig. 6C,D). The integrity of the mucosa and the status of proliferation were without pathological findings in all other samples from the B/LAC group.

Discussion

In a former study was shown that the strain L. buchneri DSM 32407 did neither influence the viability of endometrial epithelial cells nor provoke a pro-inflammatory response in contrast to other Lactobacillus strains16. Pathogenic (Trueperella pyogenes) or potential pathogenic strains (Bacillus pumilus) caused cell death even within 16–24 h33,34. Therefore, the strain L. buchneri DSM 32407 seems suitable to serve as a probiotic strain within the bovine uterus to modulate the immune response and the reproductive performance of dairy cows. The results of the present study support the hypothesis that lactobacilli improve the genital health resulting in better fertility rates, which are negatively influenced by inflammatory processes in the bovine uterus caused by pathogenic bacteria35,36,37.

A beneficial impact on reproductive performance, in particular on the interval from calving to conception, was observed after the intrauterine administration of L. buchneri DSM 32407. This might be associated to the down-regulated local immune system on days 45–51 pp, an indication for a healthy uterus at a time of the puerperium closer to first insemination24. Supporting this assumption of a positive impact on reproductive performance, one other study revealed that the presence of lactobacilli in the bovine uterus is related to a better pregnancy rate17. The percentage of cows positive for lactobacilli at 35 days in milk (DIM) was significantly higher in cows that were pregnant by 200 DIM compared with the group of cows that were not pregnant by that time. Interestingly, a similar effect was observed in a recent study that investigated the human endometrial microbiota. It showed that women with a Lactobacillus-dominated-microbiota (>90%) have higher chances of implantation, pregnancy and live birth after in vitro fertilization38. In this context, a positive effect of intrauterine lactobacilli on fertility might also be related to stimulatory effects on the blastocyst around the time of implantation. In vitro experiments showed that L. acidophilus culture supernatant positively influenced the growth and development of bovine embryos39.

Such influence on bovine genital health was additionally observed in recent studies. The weekly administration of a mixture of lactobacilli (LAB) into the vagina from two weeks before until four weeks after parturition (six treatments) decreased the occurrence of purulent vaginal discharge in dairy cows at week three pp40. In the same study, survival analysis revealed shorter calving to pregnancy intervals for the group of multiparous cows treated intravaginally with LAB. In two related studies, the same mixture of LAB were administered intravaginally only three or two times. Treated cows had a faster uterine involution and a lower incidence of uterine infections, cows with three treatments resumed ovarian cyclicity earlier and cows with two treatments had fewer days open41,42. However, a recent study showed that the number of intrauterine counted lactobacilli did not differ compared with controls when lactobacilli were intravaginal administered43. In contrast to the mentioned studies, the present study focused on the treatment of cows with signs of SCE, which have impaired reproductive performance and are difficult to detect44. In addition, the present study revealed that only one intrauterine application improved the reproductive performance tremendously. It was shown that cows, even healthy or with signs of SCE, showed a better reproductive performance after the treatment.

The improved fertility may be obtained because an intrauterine administration of L. buchneri DSM 32407 initially stimulates the local immune system. One week after the administration, at a time when the uterine health status of most cows recovered to health, cows of the LAC/SCE group tended to have higher percentages of PMN in endometrial samples. These findings are supported by an earlier study that observed the endometrial infiltration with immune cells for up to 12 days after the intrauterine administration of lactobacilli to clinically healthy cows and stated a possible stimulatory effect on endometrial defence mechanisms19.

The higher mRNA expression of TNF in the LAC/SCE group one week after the intrauterine administration of L. buchneri DSM 32407 supports these findings on a molecular basis. TNF is a pro-inflammatory cytokine that is produced by monocytes and/or macrophages infiltrating to the site of an inflammation45. It was found to be more highly expressed in the bovine endometrium during an inflammatory process, which reflects an activated immune system within the uterus46. The stimulation of cells with TNF results in the synthesis of CXCL547,48, which explains that CXCL5 mRNA contents tended to be higher in the LAC/SCE group at this time point. CXCL5 is a chemoattractant responsible for mediating neutrophil recruitment during inflammation and infection and binds to CXCR2, which is present especially on the surface of immune cells, e.g. PMN49,50.

In addition, the histopathology of endometrial biopsy samples of cows showed an infiltration with immune cells only in cows of the B/LAC group one week after the intrauterine administration. These biopsy data were supported by an in vitro study showing that L. buchneri DSM 32407 did not affect the viability of bovine endometrial epithelial cells in this short-time co-culturing experiments with a MOI up to 1016. However, this study only focused on the short-term influence of L. buchneri DSM 32407, whereas the results of the present study provide more information about the long-term effects in vivo.

Although there is an indication of an activated local immune system, the mRNA expression levels of the other investigated pro-inflammatory factors were not significantly influenced seven days after the intrauterine administration of L. buchneri DSM 32407. However, it is possible that the mRNA expression of these factors increased earlier than seven days after the administration, previous to the influx of immune cells.

In the early puerperium, an up-regulation of the local immune system for the clearance of invaded pathogenic bacteria is regarded as a physiological process51. In this context, it was observed that cows with a high PMN infiltration within the uterus during the first week pp have a better fertility compared with cows, which a have lower PMN infiltration52. Bacterial contaminations of the bovine uterus are almost inevitable with a prevalence of around 90%53 with pathogenic bacteria being the main cause for the development of endometritis54,55. The immunostimulatory effects for lactobacilli were also observed in earlier studies19,56,57 and the present study also suggests that lactobacilli are supportive for the containment of these pathogenic bacteria. It was also shown that a distinct lactobacilli strain alone or in combination with other reduced an E. coli infection and affected the pro-inflammatory reaction in bovine endometrial cells in vitro58,59.

In this context, there is indication that lactobacilli produce mitogenic and chemotactic factors. Culture supernatant of lactobacilli had strong pro-inflammatory properties, inducing the influx of PMN, the proliferation of macrophages and lymphocytes and the production of TNF by macrophages60. The competition with pathogens for limited nutrients also has to be taken into account61.

An early containment of pathogens might subsequently reduce local inflammatory processes in the endometrium and has beneficial effects on the genital health status. Although no different effects of the administration of L. buchneri DSM 32407 or the placebo on the clinical uterine health status of the LAC/SCE and PLA/SCE groups were observed, the present results of the mRNA expression analysis support this assumption. Three weeks after the intrauterine administration of L. buchneri DSM 32407 or the placebo, the mRNA of CXCL1/2, CXCL3, CXCR2, IL1B, IL8 and PTPRC was lower expressed in the LAC/SCE group compared with the PLA/SCE group. These pro-inflammatory factors were higher expressed in cows with signs of SCE and/or CE and thereby related to uterine health problems in a recent study at the same time point24 and in previous studies at earlier time points of the puerperium21,23,33. The mRNA data of the present study thereby indicates that the endometrial immune system of cows treated with L. buchneri is downregulated three weeks after its administration to the uterus. Unfortunately, the present study design does not allow a satisfying explanation on why it takes three weeks to a noticeable effect on the immune system. In future studies, it would be of great benefit to investigate the endometrial microbiota by pyrosequencing before and after the administration of L. buchneri to better understand the time-dependent changes and possible correlations to the effects on the immune system.

In conclusion, L. buchneri DSM 32407 improved the reproductive performance of cows with SCE and healthy cows. This distinct lactobacillus strain seems to first have a stimulatory effect on the local immune system one week after its intrauterine administration to dairy cows on days 24–30 pp. This might be beneficial at that time point of the puerperium, presumably supporting the elimination of pathogenic bacteria from the uterine lumen. Three weeks after the administration, the endometrial mRNA expression of several pro-inflammatory factors was down-regulated, suggesting the containment of local inflammation possibly related to the earlier containment of pathogenic bacteria, which in turn might be associated with better parameters for reproductive performance. However, to confirm this suggestion of improved fertility, a larger cohort of cows treated with L. buchneri DSM 32407 would be necessary.