Eight mares with endometritis (Zhangye, Gansu) and 12 healthy adult female rabbits (2.5–3 kg, from Lanzhou Veterinary Institute, Chinese Academy of Agricultural Sciences) were used.

We conducted clinical examination of mares that had been infertile for a long time on the farm and laboratory examination of mares suspected of having endometritis (fluid or swelling in the uterus upon examination) after rectal and ultrasound examinations. Briefly, the mare was bound, and the mare's vulva was disinfected using benzalkonium bromide (0.1% concentration). After disinfection, a double-layer sampling tube was inserted into the mare's vagina, and the inner tube was slowly inserted into the mare's uterus when the outer tube reached the cervix. Saline was injected into the uterus, and then the uterine flush was collected, followed by bacteria isolation and identification (see Supplementary material for results). The purified pathogens (E. coli, S. equi subsp zooepidemicus, and S. aureus) were inoculated into nutrient broth liquid medium containing sheep serum (Servicebio, Wuhan, China) and incubated for 12 h at 37°C on a shaker (23).

According to previous research results on intraperitoneal injection of pathogenic bacteria in mice (Preliminary test—Figure 5 in Supplementary material), 10⁹ CFU/ml bacterial solution was selected for the modeling experiment. After 2 weeks of normal feeding, the rabbits were divided into control, model, and treatment groups (n = 4 per group). Adequate environmental disinfection (UV illumination) was done. First, the experimenter disinfected the rabbits' vulva using benzalkonium bromide (0.1%) according to the principles of aseptic operation. Then, 0.5 ml of E. coli, S. equi subsp zooepidemicus, and S. aureus were mixed and injected into the rabbit uteri of the experimental and treatment groups through a catheter, and the rabbits were lifted in reverse for 3 min after the injection to prevent the bacterial fluid from flowing out. The control group was injected with the same amount of nutritional broth. A total of three injections were administered per week (Days 1, 3, and 6). During this period, the rabbits were observed for body temperature, vaginal discharge, and other physiological conditions. One week later, the rabbits were dissected and observed for gross uterine lesions. After each inoculation, the rabbits in the treatment group were injected intramuscularly with penicillin (30,000 IU/kg) and streptomycin (10 mg/kg) according to the results of the previous drug sensitivity test against these pathogenic bacteria (Preliminary test—Tables 1–3 in Supplementary material).

Venous blood was collected from rabbits in the control, model, and treatment groups at the ear margins for routine blood analysis (MindrayBC-20s, Shanghai, China), with an emphasis on the changes in their white blood cells (WBC).

The uterine secretions were collected with a cotton swab, soaked in 2 ml of PBS solution, and an equal amount of sodium hydroxide (4%) was added. After mixing and observing, a colorless result is judged as negative, and a yellowish or lemon yellow result is judged as positive. Take 2 ml of rabbit urine, add 1 ml of silver nitrate solution (5%) and mix well, then heat to boiling, the appearance of black precipitate is positive, brown or light brown precipitate is negative.

Rabbit uterine tissues were fixed in 4% formaldehyde for 24 h. After paraffin embedding, sectioning, and hematoxylin-eosin staining (H&E staining), the morphological changes of the tissues were observed under a light microscope (APExBIO, Houston, USA) at a magnification of 100 × (24).

The wax block-fixed uterine tissue sections (0.4 mm) were baked in an oven at 65°C for 2 h. The paraffin wax was melted and then removed with xylene. The sections were soaked in different concentrations of alcohol (100, 95, 85, 75%) for 5 min. The samples were removed, rinsed with distilled water for 3 min, rinsed with PBS for 5 min and then boiled with citrate solution, then cooked on low heat for 15 min, and after natural cooling, the sections were removed and rinsed with PBS for 5 min for antigen repair. The sections were incubated with H₂O₂ (3%) solution dropwise for 15 min for blocking endogenous peroxidase, rinsed with PBS for 5 min then antibody dilutions (IL-1β, IL-6, TNF-α and MPO; 1:200) were added dropwise and incubated for 12 h at 4°C. The incubated samples were removed and rinsed with PBS for 5 min, and then the secondary antibody dilution (IgG; 1:2,000) was added dropwise and incubated for 30 min at 37°C. The sections were rinsed with PBS for 5 min and then DAB color development solution was added dropwise. After color development, hematoxylin was added dropwise to re-stain the sections for 5 min. After rinsing with running water, alcohol was used to dehydrate the sections, xylene was used to clear the sections, and finally neutral resin was added dropwise to seal the sections. The fabricated sections were photographed and observed under a microscope at a magnification of 400 × . MPO and TNF-α were localized in the cell membrane or cytoplasm, IL-1β and IL-6 were localized in the cytoplasm. The proportion of positive cells in each high magnification field (400 × ) and the intensity of cell coloration were scored using the semi-quantitative integration method with reference to the scoring criteria of Fromowitz et al. (25). Color intensity score: no coloration was scored as 0, light yellow coloration was scored as 1, brown coloration was scored as 2, and tan coloration was scored as 3. The proportion of positive cells was scored: the proportion of positive cells < 1% was scored as 0, 1%−25% was scored as 1, 26%−50% was scored as 2, and >50% was scored as 3. The immunohistochemical score was obtained by adding the color intensity and the percentage of positive cells. The sum of the scores was 0–1, the expression was negative, with 1–2 being weakly positive (+), 3–4 being positive (++), and 5–6 being strongly positive (+++).

Briefly, 0.1 g of uterine tissue was weighed in a mortar, liquid nitrogen was added, and the tissue was crushed. Then, the crushed tissue was mixed with 1 ml phosphate buffer saline (PBS), centrifuged at 3,000 rpm for 10 min at 4°C, and the supernatant was used for the assay. Next, 40 μl of sample diluent was added to the octal tube of the ELISA kit (JINGMEI BIOTECHNOLOGY, Jiangsu, China), mixed with 10 μl of supernatant sample, and incubated for 30 min at 37°C; the liquid was discarded, followed by washing with washing solution and pat drying. Then, 50 μl of horseradish peroxidase (HRP) was added to the wells, incubated for 30 min at 37°C, and the liquid was discarded, followed by pat drying on absorbent paper. Next, each well was filled with washing solution, washing was repeated five times, and then 50 μl of chromogen solution A and 50 μl of chromogen solution B was added to each well, followed by incubation at 37°C for 15 min. Finally, 50 μl of termination solution was added to each well, and the optical density (OD) of each well was measured at 450 nm using a microplate reader (Molecular devices, Shanghai, China). The concentrations of these three inflammatory factors were measured in rabbits using ELISA kits for IL-1β, IL-6, and TNF-α, respectively (26, 27).

TRIzol reagent (Servicebio, Wuhan, China) was used to extract total RNA from rabbit uterine tissue (28, 29). Briefly, 0.1 g of tissue was crushed and transferred to a centrifuge tube; 1 ml of TRIzol reagent was added and lysed at 4°C for 10 min; then, 200 μl of chloroform was added and mixed well. The mixture was allowed to stand for 10 min, followed by centrifuging at 12,000 rpm for 15 min. Next, 0.5 ml of the supernatant was mixed with an equal amount of isopropanol. After allowing to stand for 5 min, the RNA precipitate was obtained by centrifugation at 12,000 rpm for 15 min, and then dissolved in ddH₂O after washing with anhydrous ethanol. The OD value (260/280) of RNA was measured using an Ultra Micro nucleic acid determinator (Pultton Technology, USA) and quantified to 200 ng/μl using ddH₂O. RNA was reverse transcribed to cDNA using M-MLV reverse transcriptase. Then, 2 μl of 5 × gDNA clean reaction mix, 1 μl of RNA sample, and 7 μl of RNase free water were mixed and reacted at 42°C for 2 min to remove genomic DNA; then, 4 μl of 5 × EVO M-MLV RT reaction mix [contains RNase inhibitor, dNTPs, Oligo Dt(18T) primer, etc.] and 6 μl RNase free water was added and reacted at 37°C for 15 min, 85°C for 5 s, and then cooled to 4°C. The primers for β-actin (control), IL-1β, IL-6, and TNF-α were designed for rabbits (Table 1), synthesized (Tsingke Biotechnology, Beijing, China), and diluted to 10 μmol/ml. Then, 1 μl of cDNA was added to 10 μl of enzyme (2 × M5 Hiper SYBR Premix EsTaq), followed by 0.5 μl each of Forward primer and Reverse primer, and 8 μl of RNase free water. Pre-denaturing was performed at 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. Results are expressed as mean relative quantification values, which were calculated using the 2^−ΔΔCT method (30).

The protein content of inflammatory factors in the uterine tissue was determined by Western blotting (31, 32). Briefly, 0.1 g of uterine tissue was crushed in a mortar and poured into a centrifuge tube. Then, 1 ml of Radio-Immunoprecipitation Assay Buffer (RIPA Lysis Buffer, Servicebio, Wuhan, China) and 10 μl of the protease inhibitor Phenylmethylsulfonyl fluoride (PMSF, Abcam, Cambridge, UK) were added, followed by lysis for 30 min. Then, the samples were centrifuged at 10,000 rpm for 10 min at 4°C, and the supernatant was collected. The concentration was measured using the protein quantitative assay kit (BCA kit, JINGMEI BIOTECHNOLOGY, Jiangsu, China), protein loading buffer was added, and the samples were denatured at 95°C for 5 min. Polyacrylamide gel electrophoresis (12% SDS–PAGE) was performed on equal amounts of denatured protein samples, and the proteins in the gel were electrically transferred to polyvinylidene fluoride membrane (PVDF membrane, Servicebio, Wuhan, China). Then, the PVDF membrane was blocked in 5% skim milk for 2 h, and transferred to the diluent of the monoclonal antibody (Abcam, Cambridge, UK) of β-actin (control), IKBα, TNF-α, and NF-kB (antibody/antibody diluent = 1/1,000). The mixture was incubated at 4°C for 12 h, washed with phosphate buffer solution containing tween (PBST), and transferred to the secondary antibody diluent (Goat anti-mouse IgG, Abcam, Cambridge, UK). After incubation at 37°C for 2 h, PBST was used for washing, and ECL luminescence reagent (JINGMEI BIOTECHNOLOGY, Jiangsu, China) was added for exposure. The results were analyzed with Image J software (image J, Media Cybernetics Co. Rockville, MD, USA).

The above trials were each conducted at least three times. Statistical analyses were performed using SPSS version 22.0 (SPSS Inc, Chicago, IL, USA). Data were analyzed by the t-test (between two groups). The statistical graphs were drawn using Prism version 5.0 (GraphPad software Inc, La Jolla, CA, USA). P < 0.05 was considered to indicate statistically significant differences.

After the rabbits were dissected, it was observed that the uterus of the rabbits in the control group was pink without symptoms such as congestion and swelling, while the uterus of the rabbits in the model group was dark red, congested, and swollen, and contained a large amount of pus. In the treatment group, the uterine congestion and swelling of the rabbits subsided (Figure 1).

The blood routine results showed that rabbits in the model group had a significant increase in the number of WBC (P < 0.01). Compared with the model group, the number of WBC in the blood of the rabbits in the treatment group decreased significantly (P < 0.01), but it increased significantly compared with the control group (P < 0.05; Figure 2).

Chemical examination of uterine secretions showed colorless in the control group, yellow in the model group, and pale yellow in the treatment group (Figure 3). Urine chemistry showed brown deposits in the control group, a large amount of black deposits in the model group, and a small amount of black deposits in the treatment group (Figure 4).

The results of H&E showed that the endometrium of rabbits in the control group was complete, the structure was clear, and there were no obvious blood cells and lymphocytes in the stroma. In the model group, the endometrium was detached from the uterine body, had increased inflammatory exudates, and a large number of red blood cells were seen in the tissue. In the treatment group, a small amount of blood exuded in the stroma (Figure 5).

Immunohistochemical results showed that the model group rabbits showed large amount of brown deposits in the uterus, the control group had basically no coloration, and the treatment group had a small amount of brown deposits (Figure 6). The mean scores of IL-1β, IL-6, TNF-α, and MPO in the uterus of control rabbits were 0.25, 1, 0.75, and 0.75, respectively, while the scores in the model group were 3.5, 5.25, 7.5, and 6.25, and the scores in the treatment group were 1.5, 1.25, 2, and 2.25. The expression of each factor in the model group was significantly increased compared with that in the control group (P < 0.01), and after the expression of each factor decreased significantly (P < 0.01) after treatment, but was still significantly higher (P < 0.05) than the control group except for IL-6 (Figure 7).

The concentrations of IL-1β, IL-6, and TNF-α in the uterus of rabbits in the model group were significantly increased (P < 0.01). Compared with the model group, the concentrations of IL-1β, IL-6, and TNF-α in the uterus of rabbits in the treatment group decreased significantly decreased (P < 0.01), but did not return to normal levels (Figure 8).

qPCR results showed that the mRNA expressions of IL-1β, IL-6, and TNF-α in the uterus of rabbits in the model group were significantly increased (P < 0.01). Compared with the model group, the expression of IL-1β and TNF-α in the treatment group was significantly decreased (P < 0.01), and the expression of IL-6 was decreased, but the difference was not significant (P > 0.05; Figure 9).

The results of western blot showed that the protein expressions of NF-kB and TNF-α in the uterus of rabbits in the model group were significantly increased (P < 0.01), and the expression of IkBα was significantly decreased (P < 0.01). Compared with the model group, the expressions of NF-kB and TNF-α in the rabbits in the treatment group were significantly decreased (P < 0.01), and the expression of IkBα was significantly increased (P < 0.05; Figure 10).