Autophagy is a host defense system that plays a vital role in resisting bacterial infections. First, we explored the correlation between E. coli infection and autophagy. After 8 h, E. coli infection reduced LC3II, ATG5, and Beclin1 expression levels and LC3II to LC3I ratio compared with the control, while LGR-1 pretreatment alleviated this reduction ( Figure 1A ). In contrast, E. coli infection increased p62 expression, a common autophagy substrate, indicating autophagic flux blockage, while LGR-1 pretreatment attenuated this increase ( Figure 1A ). Confocal laser scanning confirmed the inhibition of autophagy in MAC-T cells ( Figure 1B ). Besides, E. coli infection significantly decreased LC3 puncta while LGR-1 pretreatment increased LC3 puncta ( Figure 1B ). These results suggest that E. coli inhibits autophagy, and LGR-1 alleviates this inhibition. Interestingly, E. coli significantly reduced PINK1 and Parkin protein levels, and LGR-1 pretreatment prevented this reduction ( Figure 1A ). Furthermore, immunofluorescence revealed colocalized Parkin and MitoTracker in LGR-1 pretreated MAC-T cells, indicating the formation of mitophagosomes ( Figure 1C ). Consistent with the immunofluorescence results, TEM showed mitophagosomes in LGR-1 pretreated MAC-T cells ( Figure 1D ). TEM results also showed that E. coli can cause mitochondria swell, mitochondrial cristae disappear and vacuolization ( Figure 1D ). Collectively, our findings indicate PINK1/Parkin–mediated mitophagy was inhibited in MAC-T cells infected with E. coli.

In order to further verify the relationship between LGR-1, E. coli and mitophagy, we used promoter and inhibitor for positive and negative verification respectively. MAC-T cells were incubated with 3-MA before exposure to LGR-1 and Rapa for 12 h before exposure to E. coli. Immunoblot analysis of PINK1, Parkin, p62, LC3, ATG5, and Beclin-1 showed that mitophagy was inhibited by 3-MA and activated by Rapa. The increased levels of PINK1, Parkin and LC3II proteins, proved that Rapa promoted mitophagy. The decreased levels of PINK1, Parkin and LC3II proteins, proved that 3-MA inhibited mitophagy. E. coli decreased PINK1, Parkin, p62, LC3, ATG5, and Beclin-1 expression levels and increased p62 expression level even with Rapa pretreatment ( Figure 2A ). On the contrary, LGR-1 increased PINK1, Parkin, p62, LC3, ATG5, and Beclin-1 expression levels and decreased p62 expression level even with 3-MA pretreatment ( Figure 2B ). Similarly, E. coli significantly reduced the number of LC3 puncta ( Figure 2D ). Furthermore, the colocalization of Parkin and MitoTracker was reduced ( Figure 2E ) after E. coli infection in cells pretreated with Rapa compared with Rapa control. MitoTracker staining showed mitochondria with a clear outline and a network structure in Rapa-pretreated MAC-T cells compared with the E. coli group ( Figure 2E ). TEM also showed mitophagosomes and normal mitochondria in Rapa-pretreated MAC-T cells before exposure to E. coli ( Figure 2C ). These findings suggest that the promotion of mitophagy in MAC-T cells reduced the mitochondrial damage by E. coli and LGR-1 may play a certain protective role by reducing the inhibition of E. coli on mitophagy.

Further, we investigated the regulatory role of mitophagy on NLRP3 inflammasome and apoptosis. Immunoblot analysis showed an increase in NLRP3, ASC, and Caspase-1 p10, indicating NLRP3 inflammasome activation in cells infected with E. coli, and LGR-1 reduced this increase ( Figure 3A ). Chromatin deep staining and nuclear fragmentation, the apoptosis-related events, were observed after E. coli infection ( Figures 1B, C ). E. coli infection increased BAX and Caspase-3 p17 expression levels and BAX to Bcl-2 ratio and decreased Bcl-2 expression level; LGR-1 pretreatment reversed these phenomena ( Figure 3B ). These data suggest that LGR-1 can reduce the NLRP3 inflammasome activation and apoptosis induced by E. coli.

Autophagy negatively regulates NLRP3 inflammasome activation and apoptosis. In order to try to clarify the effect of E. coli/LGR-1 on mitophagy in MAC-T cells without considering other factors, and try to simulate the protective mechanism of LGR-1, Rapa and 3-MA were introduced. LGR-1 pretreatment prior to E. coli infection improved cell state by promoting mitophagy. Rapa displayed a potent inhibitory effect on NLRP3 and Caspase-1 activation in E. coli challenged cells ( Figure 4A ). In E. coli-infected MAC-T cells, Caspase-3 p17 and BAX expression levels and BAX to Bcl-2 ratio decreased and Bcl-2 expression level decreased by Rapa pretreatment ( Figure 4A ). Similarly, the state of the nucleus in Rapa-pretreated cells was better than in the cells infected with E. coli after Rapa pretreatment ( Figure 4B ). In the presence of the autophagy inhibitor 3-MA, NLRP3 inflammasome and apoptosis were activated, and LGR-1 pretreatment alleviated this situation ( Figure 4C ). These data indicate that activated mitophagy suppresses E. coli-induced NLRP3 inflammasome activation and apoptosis.

Furthermore, we analyzed the mitochondria in the MAC-T cells. TOM20 staining showed evenly distributed linear or short, rod-shaped, straight or curved mitochondria with clear outlines, forming a network structure throughout the cytoplasm ( Figure 5A ). Mitochondria were fragmented in the cells infected with E. coli, and LGR-1 pretreatment returned mitochondria to normal. TEM also showed severe mitochondrial damage, manifested as mitochondrial swelling, loss of cristae, and vacuolization, in cells infected with E. coli; LGR-1 alleviated the damage ( Figure 1D ). We further measured the T-AOC and ROS and SOD levels in the cells. E. coli decreased T-AOC and SOD level and increased ROS levels, and LGR-1 pretreatment alleviated these trends ( Figure 5B ). Collectively, our findings indicate that E. coli activates NLRP3 inflammasome and apoptosis by causing mitochondrial damage and oxidative stress and LGR-1 alleviates these effects.

E. coli induced mitochondrial damage led to the accumulation of ROS, which stimulate NLRP3 and apoptosis. MAC-T cells infected by E. coli showed high ROS levels ( Figure 5B ). Meanwhile, the ROS levels in E. coli-infected cells pretreated with the ROS scavenger-NAC were similar to those in the control cells ( Figure 6C ). Furthermore, LGR-1 pretreatment prevented ROS production in response to E. coli infection ( Figure 5B ), showing the effective ROS scavenging activity of LGR-1. The ROS scavenging effect of LGR-1 was further demonstrated in H₂O₂-induced MAC-T cells ( Figure 8D ). We tested the activation of NLRP3 inflammasome and apoptosis after adding NAC. Accompanying the decrease in ROS, NLRP3, Caspase-1 p10, and Caspase-3 p17 levels ( Figure 6A ) due to E. coli infection was mitigated by NAC, which indicated that E. coli-induced NLRP3 inflammasome and apoptosis activation required ROS production. Therefore, after NAC treatment, the ability of E. coli to induce ROS production is weakened, and ROS levels return to normal levels. The increase in NLRP3, ASC, Caspase-1 p10, BAX, and Caspase-3 p17 levels and BAX to Bcl-2 ratio and the decrease in Bcl-2 level in H₂O₂-treated cell suggest that H₂O₂ activates NLRP3 inflammasome and apoptosis by producing ROS ( Figure 6B ). Meanwhile, LGR-1 showed an effect similar to NAC; LGR-1 alleviated H₂O₂-induced NLRP3 inflammasome activation and apoptosis ( Figure 6B ). Flow cytometry showed less Annexin V-positive MAC-T cells after LGR-1 pretreatment compared with H₂O₂ ( Figure 8C ). Meanwhile, LGR-1 pretreatment decreased the ROS levels in cells compared with H₂O₂-induced cells ( Figure 8D ). Based on these data, we speculate that LGR-1 eliminates ROS to relieve E. coli-induced NLRP3 inflammasome activation and apoptosis.

LGR-1 attenuated NLRP3 inflammasome activation and apoptosis induced by E. coli; therefore, we explored the relationship between NLRP3 inflammasomes and apoptosis. Immunoblot analysis showed an increase in the pro-apoptotic protein (BAX and Caspase3) and decrease in the anti-apoptotic protein (Bcl-2) with E. coli exposure ( Figures 3B , 4A , 6A and 7A ). MCC950, an NLRP3 inflammasome inhibitor, was used to pretreat MAC-T cells before exposure to E. coli to inhibit NLRP3 inflammasome activation and explore the correlation between NLRP3 inflammasome and E. coli-induced apoptosis. MCC950 abolished E. coli-induced apoptosis, which was demonstrated by the decrease in BAX, BAX/Bcl-2 ratio and Caspase-3 p17 level and an increased in Bcl-2 level in immunoblot analysis ( Figure 7A ). Flow cytometry also showed less Annexin V-positive MAC-T cells after MCC950 pretreatment ( Figure 7B ). Collectively, our findings indicate that LGR-1 inhibits NLRP3 inflammasome activity to protect MAC-T cells from E. coli-induced apoptosis.

To verify whether LGR-1 eliminates ROS production through mitophagy mediated by PINK1/Parkin to inhibit the activation of NLRP3 inflammasome and apoptosis, we applied siRNA interference technology to knock down the PINK1 protein expression in MAC-T cells. Immunoblot analysis confirmed that siRNA effectively suppressed PINK1 expression ( Figure 8A ). We further added H₂O₂ as a positive control to increase ROS production and activate NLRP3 inflammasome and apoptosis. LGR-1 could not reduce H₂O₂-induced NLRP3, ASC, and Caspase-1 p10 levels after PINK1 was knocked down ( Figure 8B ). Besides, LGR-1 could not alleviate H₂O₂-apoptosis, manifested by the increase in BAX, BAX/Bcl-2 ratio, and cleaved-Caspase-3, and the decrease in Bcl-2 ( Figure 8B ). Knocking down PINK1 inhibited the ROS scavenging effect of LGR-1 ( Figure 8C ). Flow cytometry showed more Annexin V-positive MAC-T cells after silencing PINK1 compared with the H₂O₂-induced cells pretreated with LGR-1 ( Figure 8D ). These results indicate that LGR-1 exerts a beneficial effect through PINK/Parkin-mediated mitophagy.

In order to verify whether LGR-1 has obvious protective effect in vivo, we successfully established a mouse mastitis model. First, we observed the morphological and histological changes of mammary gland tissue by H&E staining. There was no significant change in the mammary glands of the CONT group and LGR-1 group. In ECOL group, the wall of mammary acinus was obviously thickened, the stroma was congested, and the acinus was filled with inflammatory cells ( Figure 9A ). And LGR-1 could significantly improve the histopathological changes caused by E. coli. Further detection of cytokines expression showed that E. coli could significantly increase IL-1 β and TNF- α, whereas LGR-1 can inhibit the production of these inflammatory factors in the breast ( Figure 9B ). Then, in order to detect whether LGR-1 can affect mitophagy, NLRP3 inflammasome and apoptosis, we detected the changes of related proteins by Western blotting. In line with the results of in vitro, E. coli can inhibit mitophagy related protein (PINK1, Parkin, ATG5, Beclin-1 and LC3II/LC3I), increase the expression of NLRP3 inflammasome (NLRP3, ASC and Caspase1 p10), and cause apoptosis, while LGR-1 can reduce these effects of E. coli ( Figures 9C–E ). Finally, TUNEL staining was performed to further explore the effect on apoptosis. The results showed that E. coli induced the accumulation of TUNEL-positive cells in the epithelial cells of breast tissue, while LGR-1 could significantly reduce the proportion of TUNEL-positive cells ( Figure 9F ).

Thirty-two females (8-10 week old) specific-pathogen-free pregnant Crl: CD1 (ICR) mice were purchased from Charles River (Beijing, China). Mice were reared in a sterile environment with 12h light and dark cycle. and ad libitum access to food and water.

This study is approved by the Animal Ethics Committee of China Agricultural University, and all animal care and experimental procedures are under the supervision of this committee.

In order to verify the protective effect of LGR-1 on E. coli induced mastitis, we established a mouse mastitis model. This mouse mastitis model was induced by E. coli as previously described (14). In short, offspring were removed 4h before intramammary inoculation. The mice were anesthetized with Zoletil 50 (55mgkg, WK001, Virbac, France) and placed in supine position under the stereoscope. Disinfect the fourth pair of mammary glands and expose the mammary duct by cutting the tip of the nipple. E. coli dissolved in 30 µL physiologic saline was slowly intraductal injected through a 100-µL syringe with a 30-gauge blunt needle. Four groups (n = 8 per group) of mice were allocated: (1) a negative control group (CONT group); (2) E. coli group (ECOL group); (3) E. coli + LGR-1 group (ECOL + LGR-1 group); (4) LGR-1 group (LGR-1 group). Before E. coli infection, mice in the ECOL + LGR-1 and LGR-1 group were inoculated with LGR-1 (2.5 × 10⁸ CFU/200 μL saline) by oral gavage for 7 consecutive days, once a day; mice in the CONT and ECOL groups were administered an equal volume of sterile physiologic saline at 10:00 AM daily. At 10:00 AM on Day 8, ECOL and ECOL + LGR-1 group mice were intraductal injected E. coli (1 × 10⁶ CFU/30 μL saline), whereas mice in the CONT and LGR-1 group received equal volume of sterile physiologic saline. After 24 hours of E. coli infection, all mice were euthanized and the mammary glands tissue were collected and frozen at -80 °C until use.

Lactobacillus rhamnosus GR-1 (LGR-1, ATCC 55826; American Type Culture Collection, Manassas, VA, United States) was grown in De Man, Rogosa, and Sharpe (MRS) broth (Oxoid, Hampshire, United Kingdom) under microaerophilic conditions at 37°C for 24 h. After overnight incubation, LGR-1 was diluted (1:1000) in fresh MRS broth and grown for approximately 12 h until mid-log phase (OD600 = 0.6).

Escherichia coli strain (serotype O111:K58, CVCC1450, the China Institute of Veterinary Drug Center, Beijing, China) was grown in Luria–Bertani (LB) broth (Oxoid, Hampshire, United Kingdom). After overnight incubation, E. coli was diluted (1:1000) in fresh LB at 37°C for 8h and 200rpm until mid-log phase (OD600 = 0.6).

MAC-T cells (a kind gift from Dr. Ying Yu in China Agricultural University) were seeded in 6- or 12-well plates (3×10⁵ and 3×10⁴ cells per well, respectively) and cultured in DMEM/Ham’s F-12 (1:1) (GE Healthcare Life Sciences HyClone Laboratories, Utah, USA) supplemented with 10% FBS (ThermoFish Scientific, Rockford, USA) and penicillin (100 Units/mL)/streptomycin (100 μg/mL) at 37°C in a 5% CO₂ incubator for 24 h. MAC-T cells were pretreated with LGR-1 (10⁵ CFU/mL; multiplicity of infection, MOI = 1) for 3 h. Then cells were washed three times with PBS and exposed to E. coli (10⁷ CFU/mL, MOI = 66). After 8 h, MAC-T cells were collected for further analysis.

MAC-T cells were treated as follows to verify the role of autophagy in LGR-1 defense against E. coli infection: MAC-T cells were (1) pretreated with 3-methyladenine (3-MA, 5 mM, S2767, autophagy inhibitor from Selleck Chemicals, Houston, USA) for 12 h washed three times with PBS, and treated with LGR-1 (MOI = 1) for 3 h, or (2) pretreated with rapamycin (Rapa, 2 μM, S1039, autophagy activator from Selleck Chemicals, Houston, USA) for 12 h, washed three times with PBS, and challenge with E. coli (MOI = 66) for 8 h.

MAC-T cells were treated as follows to verify whether ROS is the target of LGR-1 in the defense against E. coli infection: MAC-T cells were (1) pretreated with N-acetyl-L-cysteine (NAC, 5 mM, S0077, ROS scavenger from Beyotime Biotechnology, Shanghai, China) for 2 h, washed three times with PBS, and exposed to E. coli (MOI = 66) for 8 h or (2) pretreated with LGR-1 for 3 h, washed three times with PBS, and exposed to hydrogen peroxide (H₂O₂, ROS inducer; 0.5 mM) for 0.5 h.

MAC-T cells were treated as follows to verify the correlation between NLRP3 inflammasome activation and cell apoptosis during E. coli infection: MAC-T cells were pretreated with MCC950 (100 nM, HY-12815A, NLRP3 inhibitor from MedChemExpress, New Jersey, USA) for 0.5 h, washed three times with PBS, and exposed to E. coli (MOI = 66) for 8 h.

In this study, MAC-T cells in all control groups were not treated, only washed with PBS synchronously. Both untreated and treated cells were collected for protein analysis.

The mammary tissues were fixed with 4% paraformaldehyde for at least 24h. The paraffin embedded tissues were sliced into 3 μm thick slices and stained with hematoxylin and eosin (H&E) staining to observe the pathological changes. The level of mammary inflammation was scored as described previously (45).

To verify the importance of PINK1/Parkin mediated-mitophagy pathway in the protection of LGR-1, we performed gene silencing to knocked down PINK. MAC-T cells were seeded in 24 well plates. When the cells grew to 60%-80% confluence, PINK1-siRNA (si-PINK1, sense 5’-GGAGCGGUCACUGACAGAATT-3’; antisense 5’-UUCUGUCAGUGACCGCUCCTT-3’, GenePharma, Suzhou, China) diluted with Lipofectamine™ RNAiMAX (13778075, a transfection reagent from ThermoFish Scientific, Rockford, USA) was transfected into the cells. At 5h after transfection, media were removed, supplemented with DMEM and incubated at 37°C for 48 h.

At 8 h after E. coli challenged, MAC-T cells were harvested and fixed in 3% glutaraldehyde (pH = 7.4) for 48 h. Samples were treated following the standard TEM procedure (15, 46).

MAC-T cells were cultivated on cell climbing sheets for staining and treated as described above (47). For labeling the mitochondria, Mito-Tracker Red CMXRos (CMXRos, C1035, Beyotime Biotechnology, Shanghai, China) was added to living MAC-T cells at 37°C for 30 min. The treated MAC-T cells were fixed with 4% paraformaldehyde for 10 min, then incubated with 1% Triton-X-100 (T8787, Sigma-Aldrich, St. Louis, USA) for 15 min at room temperature to permeate the cell membrane, and blocked with 2% bovine serum albumin for 1.5 h at room temperature. The cells were incubated with anti-Parkin (1:200, 14060-1-AP, Proteintech Group Inc, Rosemont, IL 60018, USA) or anti-TOM 20 (1:300, 11802-1-AP, Proteintech Group Inc), at 4°C overnight. Then cells were incubated with Alexa Fluor 488 goat anti-rabbit secondary antibody (Beyotime Biotechnology, Shanghai, China) at room temperature for 1 h. DAPI (C0060, Solarbio Science&Technology Co., Ltd, Beijng, China) was used to stain cell nuclei. The cells were observed and images captured under a confocal laser scanning microscope (Nikon A1); MitoTracker Red CMXRos was detected at 555 nm, and secondary antibodies at their corresponding wavelengths (488 nm/555 nm).

The T-AOC (S0121, Beyotime Biotechnology, Shanghai, China) and SOD (S0101S, Beyotime Biotechnology, Shanghai, China) level were determined using the commercial kits, following the manufacturers’ instructions.

Intracellular ROS levels were measured by the 2,7-dichlorofluorescein diacetate (DCFH-DA, S0033S, cell-permeable fluorescent probe from Beyotime Biotechnology, Shanghai, China). Fluorescence was directly assessed by a fluorescence plate reader (BioTek Synergy H1). All the values were normalized using control. Flow cytometry was performed using a FACS Calibur system (BD Biosciences), and data were analyzed using FlowJO software (version 10.0.7).

Total RNA was extracted from mice mammary tissue by using RNAiso Plus (9108, TaKaRa, Japan), and RNA transcription was performed adopting the PrimeScriptTM RT Reagent Kit (RR047A, TaKaRa, Japan). Quantitative real-time RT-PCR was performed through using a SYBR Green PCR Master Mix (LS2062, Promega, USA). The mRNA expression of IL-1β and TNF-α was normalized to the mRNA expression of β-actin. The primer sequences are demonstrated in Table S1 and the gene expression levels were analyzed with the 2^−ΔΔCT.

MAC-T cells were lysed in RIPA buffer containing a protease/phosphatase inhibitor cocktail on ice for 30 min.). Protein (equal amounts, 20 µg) were loaded on 10% or 12% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride (PVDF) membranes (Roche). After blocking with 5% skim milk at 37°C for 1 h and the membranes were incubated with the following primary antibodies at 4°C overnight: anti-LC3 I/II (1:1000, #4108) and anti-cleaved-Caspase-3 (1:1000, #9664) from Cell Signaling Technology (Danvers, USA); anti-ATG5 (1:750, 10181-2-AP), anti-Beclin-1 (1:1000, 11036-1-AP), anti-p62/SQSTM1 (1:1000, 18420-1-AP), anti-ASC (1:1000, 10500-1-AP), anti-BAX (1:1000, 50599-2-Ig), anti-Bcl-2 (1:1000, 12789-1-AP), anti-Caspase-3 (1:1000), anti-PINK1 (1:1000, 23274-1-AP), anti-Parkin (1:1000, 14060-1-AP), anti-GAPDH (1:5000, 60004-1-AP) and anti-β-Actin (1:5000, 60008-1-AP) from Proteintech Group Inc (Rosemont, IL 60018, USA); anti-NLRP3 (1:500, AF2155) from Beyotime Biotechnology (Shanghai, China) and anti-Caspase-1 (1:1000, ab179515) from Abcam (Cambridge, UK). The immunoreactive bands were visualized with an ECL detection system (Tanon 6200 chemiluminescence imaging workstation, Tanon Science & Technology Co., Ltd. Shanghai, China). The protein bands were quantified by densitometry using ImageJ software (version 1.50).

Cell apoptosis was measured by the Annexin V-PE/7-AAD apoptosis detection kit (A213-01, Vazyme Biotech, Nanjing, China). MAC-T cells were harvested, and softly re-suspended in 1× binding buffer (100 μL). Then Annexin V-PE (5 μL) and 7-ADD (5 μL) were added to each group and incubated in the dark for 15 min. Approximately 10,000 or 20,000 cells from each group were used to analyze on a FACS Calibur system (BD Biosciences) and evaluated with the FlowJo software (version 10.0.7).

The paraffin embedded tissues were sliced into 3 μm thick slices, dewaxed and dehydrated, and then TUNEL staining (A112, Vazyme, Nanjing, China) was performed according to the manufacturer’s instructions. Briefly, the slices were washed with PBS and reacted with TdT enzyme/buffer at 37°C for 1 h, followed by DAPI (C0060, Solarbio Science&Technology Co., Ltd, Beijng, China) staining at room temperature for 5 min. Then the slices were observed and images were captured at x400 magnification using a confocal laser scanning microscope (Nikon A1).

Using Prism 7 (GraphPad) to perform statistical analysis. Qualitative data were expressed as means ± standard error of the mean (SEM; n = 3 or 6). One-way analysis of variance (ANOVA) was applied to analyze statistically significant differences at p < 0.05, followed by Tukey’s test.