Ad-mCherry-GFP-Microtubule-associated protein 1 light chain 3 beta (LC3B) (C3011), Ad-GFP-LC3B (C3006), DiI (1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate, C1036), Hanks’ Balanced Salt Solution (with Ca²⁺& Mg²⁺, HBSS, C0219), coverslips (FCGF60), Hoechst 33258 (C1017), and Histone H3 antibody (AF0009) were from Beyotime (Shanghai, China). Fenofibrate (HY-17356), SBE-β-CD (Sulfobutylether-β-Cyclodextrin, HY-17031), DMSO (Dimethyl sulfoxide, HY-Y0320C), and Filipin complex (HY-N6716) were purchased from MedChemExpress (Monmouth Junction, NJ, USA). Anti-LAMP1 antibody (67300-1-Ig), anti- lysosomal-associated membrane protein 2 (LAMP2) antibody (66301-1-Ig), anti-TFEB antibody (13372-1-AP), anti-TFE3 antibody (14480-1-AP), anti-LC3 polyclonal antibody (14600-1-AP), anti-ATG5 antibody (10181-2-AP), anti-RAB7A antibody (55469-1-AP), anti-mouse IgG-horseradish peroxidase (HRP) (SA00001-1), and Goat anti-rabbit IgG (SA00001-2) were all from Proteintech (Chicago, IL, USA). Anti-SQSTM1/p62 antibody (23214), anti-GAPDH antibody (97166) and Anti-LC3B (E5Q2K) antibody (83506) were from Cell Signaling Technology (Danvers, MA, USA).

Two strains of M. bovis, PG45 (ATCC 25523) and WT21 (Wild type isolate) were cultured in PPLO medium (BD Biosciences) and yeast extract (BD Biosciences) broth with 20% horse serum (Solarbio, Beijing, China) and 100 IU/L penicillin (Coolaber) in 5% CO₂ at 37°C for 72 h. The culture was centrifuged (6000 × g for 30 min) and washed with phosphate-buffered saline (PBS). Bovine mammary epithelial cells (bMECs) (MAC-T; Jingma Biological Technology Co., Ltd., Shanghai, China) were cultured in DMEM (Gibco™, 11995073) supplemented with 10% fetal bovine serum (FBS, Gibco, Grand Island, NY, USA) at 37°C with 5% CO₂ in 25-cm² cell culture plates (Corning Inc., Corning, NY, USA). The bMECs were grown in the incubator for 24 h and passaged after reaching 60-70% fusion. At 24 h before the experiment, bMECs were seeded (1×10⁵ cells/mL) in 6-well plates (2 mL per well) or 96-well plates (0.1 mL per well) for M. bovis infection.

The bMECs were grown in 6-well culture plates to reach 60-70% fusion, then infected (defined as 0 h) with M. bovis (multiplicity of infection [MOI]=1:30). After culture at 37°C for 1 h, cells were washed 3 times with sterile PBS (2 mL/well) to remove nonadherent M. bovis. Then, 2 mL of DMEM containing 400 μg/mL gentamicin (Solarbio, Beijing, China) was added to each well and incubated at 37°C for 2 h to kill extracellular M. bovis (Liu et al., 2021). Cells were again washed 3 times with sterile PBS (2 mL/well) and 2 mL of DMEM with 10% FBS and 10 μg/mL gentamicin was added to each well before experiments.

Using the method described above with M. bovis-infected cells, at indicated time points, cells were rinsed 3 times with PBS, detached with sterile scrapers, and resuspended in 500 μL of sterile PBS. The suspension was mixed by passing it through a syringe and 23G needle, followed by 10-fold serial dilutions. Dilutions were plated on M. bovis PPLOA medium and incubated at 37°C with 5% CO₂ for up to 7 d. Each treatment had 3 replicate wells, with each well counted 3 times. Experiments were independently repeated 3 times.

Adenoviruses encoding GFP-LC3B and mCherry-GFP-LC3B fusion proteins (Ad-GFP-LC3B and Ad-mCherry-GFP-LC3B) are widely used for autophagy detection. These adenoviruses effectively label autophagosomes by expressing fusion proteins, facilitating real-time monitoring of autophagy. The bMECs were inoculated into 6-well culture plates as described above and cultured for 12 h to reach 40-50% fusion, when cells were infected using Ad-GFP-LC3B and Ad-mCherry-GFP-LC3B adenovirus mixed with DMEM containing 10% FBS at MOI = 1:4. Cells were grown in an incubator at 37°C and 5% CO₂ for 24 h prior to subsequent experiments. GFP-LC3B labels autophagosomes whereas mCherry-GFP-LC3B detects autophagic flux.

Nuclei and cytoplasmic proteins were extracted using nuclei and cytoplasmic protein extraction kits (Beyotime, Shanghai, China). To extract total cellular proteins, cells were washed 3 times with PBS at the indicated time points, and total proteins were extracted using RIPA lysis buffer (Beyotime) under ice-bath conditions. Collected proteins were centrifuged (12,000 × g for 20 min at 4°C), protein concentration was estimated using a BCA kit (Sigma Aldrich, B9643-1L) and samples were subjected to SDS-PAGE and transferred to PVDF (PolyVinyliDene Fluoride). Membranes were incubated with specific primary antibodies overnight at 4°C, then a peroxidase-coupled secondary antibody added, with incubation at room temperature for 1 h. Immunoreactivity was detected using the ECL detection system, and band densities analyzed with Image J (National Institutes of Health, Bethesda, MD, USA).

Cells were lysed with 1 mL of TransZol Up lysate per well, according to manufacturer’s instructions, and total RNA extracted according to the instructions of the Total RNA Extraction Kit (TransGen Biotech, Beijing, China). First-strand cDNA was synthesized by reverse transcription of RNA using One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech). RNA was reverse-transcribed using One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech). Real-time PCR was done on an Applied Biosystems StepOnePlus real time PCR (Thermo Fisher Scientific) system using SYBR Green PCR Master Mix (TransGen Biotech). Primer sequences were: Acetyl-CoA acetyltransferase 1 (ACAT1) (forward primer: AGGCGGGTGCAGGAAATAAG; reverse primer: ACCAAGTTTAGTGGCTGGCA), 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) (forward primer: GCAGAGCAATAGGCCTTGGT; reverse primer: GAGTCACAAGCACGGGGAAA) and Sterol regulatory element binding transcription factor 2 (SREBF2) (forward primer: ATCGCTCCTCCATCAACGAC; reverse primer: CCTCAGAACGCCAGACTTGT) in real-time PCR.

Infected cells were washed 3 times with PBS at designated time points, followed by addition of Buffer A for Metabolic Assay from the Amplex Red kit (Beyotime, Beijing, S0211S). Cells were gently resuspended to detach them and free cholesterol and cholesterol esters measured. Cholesterol standards were prepared according to manufacturer’s instructions, with a concentration range of 0 to 4 μg/mL (0 to 10 μM). An aliquot (50 μL) of cell lysate diluted with reaction buffer were added to the microplate, followed by addition of 50 μL of 300 μM Amplex Red working solution. The microplate was incubated at 3°C in the dark for 30 min. Fluorescence was measured using a fluorescence microplate reader (Thermo Fisher Scientific, USA) with an excitation wavelength of 560 nm and an emission wavelength of 590 nm.

M. bovis was stained with DiI (Beyotime, C1036), a red fluorescent cell membrane dye. The M. bovis pellet was resuspended in DMEM medium containing 5 μM DiI, vortexed to ensure uniform mixing, and incubated at room temperature for 20 min. The bacterial pellet was harvested by centrifugation at 8000 × g for 30 min, followed by washing twice with PBS for subsequent infection of bMECs. After treatment at designated time points, cells were washed 3 times with PBS, fixed with 4% paraformaldehyde (PFA, Solarbio) for > 4 h, permeabilized with 0.5% Triton X-100 (Sigma Aldrich, X100) for 10 min, and incubated 1 h at room temperature in PBS with 3% bovine serum albumin. Subsequently, cells were incubated overnight at 4°C with primary antibodies targeting TFEB and TFE3 (1:500). Following antibody incubation, cells were washed with sterile PBS and then incubated at room temperature with Cy3-conjugated goat anti-rabbit IgG (H+L) (Beyotime, A0516, 1:500) and Cy3-conjugated goat anti-mouse IgG (H+L) (Beyotime, A0521, 1:500) for 1 h each. Finally, nuclei were stained with Hoechst 33258 (Beyotime, C1017). Cell cholesterol was stained using Filipin complex; cell slides were immersed in a solution of 500 μg/mL Filipin complex in PBS at room temperature for 20 min. Coverslips were prepared using an anti-fluorescence quenching mounting medium (Beyotime, P0126), and observed under a Nikon A1 LFOV laser scanning confocal microscope at wavelengths of 405, 488 and 561 nm. mCherry and GFP puncta were quantified from 20 cells per sample and a minimum of 60 cells per group for statistical analysis. All experiments were performed independently at least 3 times (n ≥ 3), and data are presented as mean ± SD.

To establish a pathogenic model of M. bovis-induced mastitis, specific pathogen-free (SPF) female Balb/c mice, 6–8 wk old (SiPeiFu Laboratory Animal Technology, Beijing, China) were used. Pregnant mice at gestational day 20 were housed in sterile isolators under controlled environmental conditions, with ad libitum access to feed and water and a 12-h light/dark cycle. On postpartum day 3, mice were anesthetized via intramuscular injection of 50 mg/kg Zoletil 50 (Virbac, Carros, France). The fourth pair of mammary glands was surgically exposed by excising the nipple tip, and 100 μL of M. bovis suspension (10^6 CFU) was slowly injected through a blunt needle (30 G, 25 mm) into the mammary duct. Mice were randomly assigned to 6 groups (n = 6 per group): 2 infection groups (inoculated with the PG45 reference strain or WT21 wild-type strain); 3 treatment groups (fenofibrate alone, PG45 + fenofibrate, WT21 + fenofibrate); and 1 negative control group (sterile PBS). Nursing pups were removed 2 h before intramammary inoculation. At 12 h post-infection, mice were weighed and treated with fenofibrate (100 mg/kg) via intraperitoneal injection. Fenofibrate was first dissolved in DMSO to prepare a stock solution, then diluted with 10% stock solution and 90% SBE-β-CD in saline, followed by ultrasonic heating to ensure complete solubilization. After 36 h of treatment, mice were anesthetized via intramuscular injection of Zoletil 50 (50 mg/kg) and subsequently euthanized by CO₂ inhalation (displacement rate of 20-30% chamber volume per minute) followed by cervical dislocation to ensure death, and mammary tissues harvested to determine bacterial burden. Under sterile conditions, 0.1 g of mammary tissue was excised, homogenized in sterile tubes, and centrifuged (3000 g x 15 min). An aliquot (10 μL) of supernatant was serially diluted and plated onto PPLO solid medium. After incubation, viable colonies were counted and expressed as colony-forming units (CFU) per gram of tissue.

Mice were anesthetized and euthanized 48 h post-inoculation with M. bovis or treatment. The fourth pair of mammary glands were aseptically excised, immersed in 75% alcohol for 3 min, and fixed in 4% paraformaldehyde. Tissues were dehydrated, embedded in paraffin, sectioned (5 μm), dewaxed in xylene, rehydrated through an alcohol series, stained with hematoxylin and eosin, and viewed under an optical microscope.

Serial sections of paraffin-embedded mouse mammary tissues were deparaffinized with xylene and rehydrated through a graded ethanol series. Antigen retrieval involved boiling in citrate buffer (pH 6.0) for 15 min, then natural cooling to room temperature. Endogenous peroxidase activity was blocked by incubating in 3% H₂O₂ in methanol at room temperature for 10 min. After 3 washes, sections were blocked with 5% goat serum at 37 °C for 30 min. Primary antibodies against LC3B, LAMP1, LAMP2, and SQSTM1 were applied and incubated overnight at 4 °C in a humidified chamber. After 3 washes with PBS buffer, secondary antibodies were added and incubated. Immunoreactivity was visualized using 3,3’-diaminobenzidine (DAB) substrate solution, with development time microscopically monitored. Sections were counterstained with hematoxylin, dehydrated, and cleared. Finally, slides were mounted with neutral resin, and immunohistochemical (IHC) micrographs were captured using a light microscope.

Mammary tissues were accurately weighed, placed into 1.5 mL centrifuge tubes containing precooled PBS (pH 7.4), stainless steel beads were added and samples homogenized using a high-throughput tissue grinder (60 Hz, 2 min, twice). Homogenates were centrifuged at 5000 × g for 10 min at 4 °C, and supernatants collected. Concentrations of tumor necrosis factor (TNF-α), interleukin 1β (IL-1β), interleukin 6 (IL-6), and interleukin 10 (IL-10) in supernatants were determined using ELISA kits (Enzyme-linked Biotechnology Co., Ltd., Shanghai, China), in accordance with manufacturer’s instructions.

Data are mean ± standard deviation (SD) from ≥ 3 independent experiments. Statistical analyses were performed using SPSS 26.0 (IBM Corp., Armonk, NY, USA), with GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA) used for graphical representations. For parametric data, 1- or 2-way analysis of variance (ANOVA) was applied, followed by Tukey’s (1-way) or Bonferroni’s (2-way) post hoc tests. Significance thresholds were: *P < 0.05, **P < 0.01, and ***P < 0.001.

Western blotting was used to optimize concentration and treatment duration of fenofibrate for autophagy induction in bovine mammary epithelial cells (bMECs), using HBSS treatment as a positive control. Treatment with 50 μM fenofibrate for 9 h significantly upregulated expression of lysosomal membrane-associated protein LAMP2 and lysosomal transport and maturation protein RAB7A, compared to the previous time point. Concurrently, autophagosome-associated marker LC3B (P < 0.001) and autophagy receptor protein SQSTM1 (P = 0.008) were also increased in a time-dependent manner (Figure 1A). Furthermore, treatment with 50 μM fenofibrate for 9 h resulted in the most significant enhancement of LAMP2 (P = 0.028), RAB7A (P = 0.001), LC3B (P < 0.001), and SQSTM1 (P = 0.004) expression (Figure 1B). Therefore, 50 μM for 9 h was used for fenofibrate-induced autophagy in bMECs.

To investigate whether fenofibrate modulates autophagy in bMECs through regulation of autophagy and lysosomal activity during M. bovis infection, in vitro infection models were established using M. bovis standard strain PG45 and wild-type strain WT21. In the M. bovis-infected group, Western blotting results indicated that fenofibrate markedly upregulated expression of lysosomal membrane protein LAMP2 and LC3B (autophagy marker), but downregulated lysosomal trafficking protein RAB7A (P < 0.001) and autophagy receptor SQSTM1 (P < 0.001). As SQSTM1 binds to autophagic substrates and is degraded in lysosomes, its reduction is a key indicator of autophagic flux integrity (Figures 2A, B). To further assess effects of fenofibrate on autophagic flux, confocal microscopy was used to examine cells transfected with mCherry-GFP-LC3B. Upon autophagosome-lysosome fusion, the lysosomal environment quenches the GFP signal, leaving red fluorescence. Quantification of red and yellow puncta yielded no significant increase in red puncta in the M. bovis-infected group compared to controls, indicating impaired lysosomal cargo transport during infection. Fenofibrate treatment increased both red puncta (P < 0.001) and yellow puncta (P < 0.001) relative to controls, even with M. bovis infections (2 strains), indicating enhanced autophagic flux (Figures 2C, D).

To investigate how fenofibrate regulates autophagy in bMECs, we assessed expression and nuclear translocation TFEB and TFE3, key transcriptional regulators of lysosome biogenesis and autophagy. Fenofibrate increased cytoplasmic TFE3 expression in the M. bovis PG45-infected group (P = 0.005; Figure 3A), with minimal impact on cytoplasmic TFEB (Figure 4A). Furthermore, fenofibrate promoted nuclear accumulation of TFEB and TFE3 (P ≤ 0.011; Figures 4A, 3A). Confocal imaging corroborated these findings, with increased TFEB and TFE3 fluorescence intensity following fenofibrate exposure (Figures 4B, 3B). Notably, under M. bovis infection, both TFEB and TFE3 were predominantly localized in the cytoplasm, whereas fenofibrate enhanced their translocation into the nucleus (Figures 4B, 3B). Therefore, fenofibrate promoted TFEB and TFE3 nuclear localization, upregulating transcription of autophagy- and lysosome-related genes and enhancing autophagic activity in bMECs.

We evaluated effects of fenofibrate on intracellular free and total cholesterol, as well as expression of cholesterol-regulating genes (ACAT1, HMGCR, SREBF2) in M. bovis-infected bMECs. Infection elevated intracellular free cholesterol concentrations (P ≤ 0.044), whereas fenofibrate reduced both total and free cholesterol compared to infected controls (P ≤ 0.001; Figures 5A, B). Notably, fenofibrate upregulated transcription of ACAT1, a gene involved in esterification of free cholesterol with acyl-CoA, whereas expression of HMGCR, the rate-limiting enzyme in cholesterol biosynthesis, and its upstream transcriptional activator SREBF2 were downregulated (P ≤ 0.024; Figures 5C–E). Concurrently, fenofibrate decreased intracellular colony-forming units (CFUs) of both PG45 and WT21 strains (P ≤ 0.012; Figure 5F). Therefore, fenofibrate modulated both free and total cholesterol concentrations and regulated transcription of cholesterol-associated genes during M. bovis infection, thereby attenuating intracellular bacterial burden.

Alterations in intracellular cholesterol concentrations can inhibit maturation of autolysosomes, thereby modulating autophagic activity. In this study, intracellular free cholesterol was visualized using Filipin staining, M. bovis was labeled with red fluorescence, and the autophagy marker LC3 was monitored via GFP-LC3 adenoviral transfection. Colocalization among cholesterol, M. bovis, and LC3 was assessed using confocal laser microscopy. Infection with M. bovis PG45 or WT21 reduced colocalization between autophagosomes (green) and M. bovis (red), as indicated by a decreased proportion of yellow puncta. Concurrently, colocalization between M. bovis and free cholesterol increased, as evidenced by enhanced intensity of pink puncta. Fenofibrate diminished M. bovis-cholesterol colocalization (reduced pink puncta) and restored autophagosome-M. bovis colocalization (increased yellow puncta), thereby limiting bacterial access to host free cholesterol and suppressing M. bovis intracellular replication, as reflected by decreased red fluorescence intensity (Figure 6).

To assess in vivo therapeutic efficacy of fenofibrate against M. bovis infection, a Balb/C murine mastitis model was established. Immunohistochemistry of mammary tissues was done to determine whether fenofibrate mitigated M. bovis-induced mastitis by modulating autophagy and lysosomal activity. Compared to the control group, infection with M. bovis strains PG45 and WT21 did not significantly alter expression of the autophagy marker LC3B (P ≥ 0.405), indicating no evident autophagy. Notably, fenofibrate treatment restored LC3B (P ≤ 0.003) expression and reduced SQSTM1 in the M. bovis-infected group (P ≤ 0.002) accumulation, suggesting a reactivated autophagic flux (Figures 7A–D). Furthermore, expression of lysosomal membrane proteins LAMP1 and LAMP2 was decreased following M. bovis infection (P ≤ 0.035), whereas fenofibrate upregulated both proteins compared to the infected group (P ≤ 0.043; Figures 8A–D).

To assess impacts of fenofibrate treatment on cholesterol in mammary tissues during M. bovis infection, an Amplex™ Cholesterol Assay Kit was used. Total and free cholesterol concentration in mammary glands were significantly elevated at 48 h post-infection (hpi) compared to uninfected controls. Intraperitoneal administration of fenofibrate (100 mg/kg) for 36 h reduced both total and free cholesterol concentrations in infected tissues relative to the untreated infection group (P < 0.01; Figures 9A, B). Histopathological evaluation demonstrated pronounced inflammatory cell infiltration in the lumen of mammary alveoli at 48 hpi, with accumulation of neutrophils (yellow arrows), lymphocytes (green arrows), and plasma cells (red arrows), with the latter 2 indicative of a chronic inflammatory response. In addition, desquamated epithelial cells, presumably undergoing necrosis, were observed within the lumen of mammary alveoli. In contrast, fenofibrate-treated mice had intact mammary gland architecture and negligible inflammatory cell infiltration (Figure 9C).

To quantify bacterial burden, M. bovis was isolated from mammary tissues and colony-forming units (CFU) enumerated. At 48 hpi, PG45 and WT21 strains reached average burdens of 1.5 × 10^4 CFU/g and 2.4 × 10^4 CFU/g, respectively. Notably, fenofibrate reduced intramammary M. bovis load (P < 0.001; Figure 9D) and decreased mammary expression of pro-inflammatory cytokines TNF-α (P < 0.05; Figure 9E), IL-1β (P < 0.001; Figure 9F), and IL-6 (P < 0.001; Figure 9G), and the anti-inflammatory cytokine IL-10 (P < 0.001; Figure 9H), compared to PG45 and WT21-infected groups.