Mouse models and treatments

All animal experiments were performed following the guidelines outlined by the ARRIVE (Animal Research: Reporting of In Vivo Experiments)⁶⁷. The experimental protocols received approval from the Animal Welfare and Use Committee of Nanfang Hospital, affiliated with Southern Medical University, ensuring compliance with ethical standards for animal research.

Cdh5Cre mice (C57BL/6 background, strain#017968)⁶⁸, AdipoqCre mice (C57BL/6 background, strain#010803)⁶⁹, Lyz2Cre mice (C57BL/6 background, strain#004781)³³, Rosa-tdTomato reporter mice (C57BL/6 background, strain#007909)⁷⁰, and Rosa-iDTR mice (C57BL/6 background, strain#007900)⁷¹ were obtained from the Jackson Laboratory. Areg^flox/flox (Areg^f/f) mice (C57BL/6 background, strain#T017891) were sourced from GemPharmatech Co., Ltd (Jiangsu, China). Egfr ^flox/flox (Egfr ^f/f) mice (C57BL/6 background, strain#NM-CKO-00133), Cx3cr1Cre mice (C57BL/6 background, strain#NM-KI-200079), and Ly6gCre mice (C57BL/6 background, strain#NM-KI-200219) were obtained from Shanghai Model Organisms Center Co., Ltd (Shanghai, China). Wild-type mice with C57BL/6 genetic background were obtained from the Experimental Animal Center of Southern Medical University.

All mice were bred and maintained in the specific-pathogen-free animal facilities of Southern Medical University, with unrestricted access to food and water. The housing conditions were strictly controlled, involving a stable temperature maintained at 23 ± 2 °C, relative humidity fixed at 50%, and a 12-hours light-dark cycle. Both male and female mice were selected for all procedures. To ascertain the genotypes of the mice, genomic DNA was extracted from tail clips and analyzed using polymerase chain reaction (PCR). The primer sequences used for genotyping are listed in Table S1.

The implant-associated S. aureus osteomyelitis mouse model was developed according to our previously described methodology with minor modifications²⁷. Briefly, mice aged 10–12 weeks were anesthetized via intraperitoneal injection of tribromoethanol (125 mg/kg). The right hind limb was shaved and aseptically prepared, followed by blunt dissection through fascia and muscle tissue to expose the femoral surface. A unicortical defect was then created in the third trochanter using a 27-gauge needle. Through this defect, we injected 3 µL of either S. aureus suspension (2 × 10⁵ CFU/mL) or PBS (sham control) into the femoral medullary using a precise microliter syringe. Subsequently, we insert a sterile, self-tapping steel screw (1.75 mm cap, 1 mm thread with 0.15 mm pitch, 2 mm length) through the same entry point, advancing it in a distal orientation along the femoral medullary cavity. The incision site was then closed with 5–0 surgical sutures. All mice received daily intraperitoneal injection of gentamicin (20 mg/kg) from the first postoperative day, as detailed in our recent work^36,72. Right femoral samples were harvested at day 14 post-infection, when infected femurs displayed profound changes in bone remodeling characteristic of the subacute phase of osteomyelitis^27,73. To assess chronic-stage pathological changes, femoral samples from osteomyelitis models using AdipoqCre-iDTR mice and Cx3cr1Cre-Areg^f/f mice were additionally collected at day 28 post-infection, corresponding to the established chronic phase of osteomyelitis^27,28.

For cell ablation experiments, AdipoqCre-iDTR mice received daily intraperitoneal injections of either phosphate-buffered saline (PBS, control) or diphtheria toxin (50 µg/kg, Aladdin, #D381867) for two consecutive days prior to surgery. This regimen continued every other day for two weeks following the surgery. In rapamycin treatment experiments, both AdipoqCre-tdTomato mice and Cdh5Cre-tdTomato mice were administered a single intraperitoneal dose of either PBS (vehicle control) or rapamycin (4 mg/kg, Selleck, #S1039) one day before surgery, followed by every other-day dosing postoperatively. For gefitinib treatment, S. aureus osteomyelitis mice were intraperitoneally injected daily from post-infection day 5 with either gefitinib (17 mg/kg/day; YEASEN, #52717ES60, China) or the same volume of vehicle (10% DMSO plus 90% corn oil).

To evaluate the bacterial burden in infected femurs, we aseptically removed all soft tissue from the femurs and recorded their weights. The bones were then homogenized in PBS (1 mL per 0.1 g femoral tissue) using a tissue homogenizer. Bone homogenates were serially diluted to 10⁻⁴, with 100 μL aliquots plated onto tryptic soy agar (TSA) for bacterial culture. For biofilm analysis on implants, we carefully removed screws, rinsed them twice with PBS to eliminate planktonic bacteria and tissue debris, and subsequently subjected them to ultrasonic treatment (40 kHz, 300 W) in 1 mL PBS for 10 minutes to detach the S. aureus in biofilm. The solicited suspension was diluted 100-fold, and 100 μL aliquots were plated onto TSA. Following 16 hours of incubation at 37 °C, we enumerated bacterial colonies to determine bacterial CFU.

To assess antibiotic distribution, we intraperitoneally administered FITC-conjugated gentamicin (20 mg/kg, Ruixibio, #R-694). Infected femurs were harvested 1 hour after injection and immediately embedded in optimal cutting temperature (OCT) compound (Biosharp Life Sciences, #BL557A) for cryosection. Sections of 15 μm thickness were cut and subsequently imaged with a confocal laser scanning microscope (Zeiss LSM980, Germany) with SoftMax Pro 6 software (version 6.5.1). The relative distribution of FITC-gentamicin was quantified by calculating the FITC⁺ area percentage in both intra-abscess and extra-abscess regions using ImageJ software (FIJI version, NIH, USA).

In the S. aureus osteomyelitis model, each experimental group comprised five mice for individual assays. A total sample sizes of 25–30 mice per group was used to enable multiple analytical endpoints, including CFU quantification, immunofluorescence, histopathological analysis, perfusion measurements, and micro-computed tomography (micro-CT) analysis. Prior to sample collection, mice were euthanized using carbon dioxide inhalation followed by cervical dislocation.

Bacterial culture and quantification

The methicillin-sensitive S. aureus clinical isolate used in this research was cultured from a chronic osteomyelitis patient²⁷. Strain was accurately identified using the Phoenix 100 automated bacterial identification system (BD Diagnostics, USA). Unless otherwise stated, this clinical isolate served as the primary strain for infection experiments in this study. To evaluate the general pathogenic effects of S. aureus in bone marrow, we additionally employed the reference strain S. aureus subsp. aureus ATCC 25923 (BioKONT Co. Ltd, China). All strains were preserved in tryptic soy broth (TSB) containing 50% glycerol and stored at −80°C until experimental use.

A single colony of the S. aureus strain was inoculated into sterile TSB medium and cultured at 37°C for 16 to 18 hours with constant agitation (180 rpm) in a shaking incubator. The overnight bacterial culture was then pelleted by centrifugation and washed twice with PBS. The optical density (OD) of the bacterial suspension was adjusted to an OD600 of 0.5 using a spectrophotometer (SpectraMAX i3x, Molecular Devices, USA) with SoftMax Pro 6 software (version 6.5.1), corresponding to a bacterial concentration of approximately 1 × 10⁸ CFU/mL as verified by serial dilution plating on a TSA plate. All microbiological procedures were conducted in a biosafety level 2 (BSL-2) setting, adhering strictly to the biosafety protocols established by the Biosafety Committee of Nanfang Hospital, Southern Medical University.

Vessel perfusion assays

To evaluate bone marrow microvascular perfusion, mice received a slow tail vein injection of 100 µL of fluorescein-conjugated lectin (0.75 mg/mL; Vector Laboratories, #FL-1171) administered over 2 minutes. Eight minutes after injection, the mice were euthanized, and the right femurs were immediately harvested. The excised femurs were fixed in 4% paraformaldehyde for 24 hours at 4 °C, followed by decalcification in a 0.5 M ethylenediaminetetraacetic acid (EDTA) (pH 8.0) for 48 hours at 4 °C. Subsequently, the samples were soaked in 20% and then 30% sucrose solutions, with each step lasting for 24 hours, before being embedded in OCT compound for cryosection.

To assess fluorescein-lectin perfusion in bone vessels during osteomyelitis pathogenesis and rapamycin treatment experiments, we utilized Cdh5Cre-tdTomato reporter mice to characterize its activity. Given that both Emcn and Cdh5 are established endothelial cell markers^74,75, and that over 90% of Emcn⁺ cells in mouse tibias and femurs are Cdh5⁺ endothelial cells⁷⁶, we labeled vessels with Emcn in femoral sections from osteomyelitis models established using AdipoqCre-iDTR mice, Lyz2Cre-Areg^f/f mice, Cx3cr1Cre-Areg^f/f mice, and AdipoqCre-Egfr^f/f mice.

The processed femurs were sectioned (15 µm) and incubated with anti-Endomucin (Santa Cruz Biotechnology, #sc-65495, 1:100) overnight at 4 °C. Following washing, sections were incubated with an Alexa Fluor 594-conjugated goat anti-rat IgG antibody (Abcam, #ab150160, 1:400) at room temperature (RT) for 1 hour. Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI). The sections were imaged using a laser scanning confocal microscope (Zeiss LSM980, Germany) with the ZEN blue software (version 3.8). Quantitative analysis was conducted using ImageJ software (FIJI version, NIH, USA). We quantified blood vessels within 0.1–0.3 mm² region surrounding the abscess, measuring the relative percentages of intravascular lumen area, Cdh5⁺Lectin⁺ co-localized area and Emcn⁺Lectin⁺ co-localized area to the total field of view.

Immunofluorescence

Immunofluorescence staining was conducted following standard protocols. Mouse right femurs were fixed overnight, followed by decalcification for seven days and sucrose protection treatment as aforementioned procedures, femurs were embedded in OCT compound and stored at −80°C until ready for sectioning.

Frozen sections of 15 µm thickness were warmed to 37 °C for 20 minutes, washed with 0.1% PBST and PBS, blocked with 10% goat serum solution in PBST for 1 hour at room temperature, then incubated with primary antibody overnight at 4°C. A panel of primary antibodies was employed: Emcn (Santa Cruz, #sc-65495, 1:100), αSMA (CST, #19245, 1:400), Col-I (Abcam, #ab260043, 1:400), F4/80 (Bio-Rad, MCA497, 1:100), AREG (Santa Cruz, #sc-74501, 1:50), S. aureus (Thermo Fisher Scientific, #PA1-7246, 1:200), YAP (CST, #14074, 1:200), Adiponectin (Proteintech, #66239-1-Ig, 1:100), p-mTOR (CST, #5536, 1:200), Perilipin (CST, #9349, 1:200), EGFR (CST, #4267, 1:200). The subsequent day, sections were incubated with fluorescent secondary antibodies (1 hour, RT): Alexa Fluor 488-conjugated goat anti-rabbit IgG (CST, #4412, 1:500), Dylight 488-conjugated goat anti-rat IgG (Abbkine, #A23240, 1:200), Alexa Fluor 594-conjugated goat anti-rat IgG (Abcam, #ab150160, 1:400), and Alexa Fluor 594-conjugated goat anti-mouse IgG (CST, #8890, 1:400). Nuclei were counterstained with DAPI. Images were acquired using a Zeiss LSM980 laser confocal microscope with the ZEN blue software (version 3.8). Quantitative analysis of positively stained cells was performed in a 0.1–0.3 mm² region adjacent to the abscess margin using ImageJ (FIJI).

Histological analysis and Masson’s trichrome staining

Paraffin-embedded femurs were sectioned coronally at 4 µm thickness for histochemical analysis. Hematoxylin and Eosin (H&E) staining was conducted using a commercial kit (Beyotime, #C0105, China) following the manufacturer’s protocol. To evaluate the histopathological alterations in infected femurs, we tailored Smeltzer’s scoring system⁷⁷. The adjusted criteria incorporated key features of both clinical osteomyelitis⁷⁸ and a mouse model of implant-associated S. aureus osteomyelitis²⁷. The assessment included a summation of scores across four parameters: intraosseous acute inflammation, characterized by small clusters of neutrophils with possible intramedullary abscess formation; intraosseous chronic inflammation, featuring extensive neutrophilic infiltrates with potential intramedullary fibrosis; periosteal reaction, encompassing periosteal inflammation that may include subperiosteal abscess formation; and bone necrosis, incorporating areas of bone death with or without sequestra present. Each parameter was scored on a scale from 0 (normal) to 4 (most severe manifestations).

Additionally, collagen deposition was visualized using Masson’s Trichrome staining (Solarbio, #G1346, China) according to manufacturer’s instructions. This technique specifically stained collagen fibers blue while counterstaining muscle fibers red. Images were digitally scanned using a Teksqray SQS-1000 scan system (Shengqiang Technology Co., Ltd.) with SlideScan software (version 1.1.10). Utilizing ImageJ (FIJI) software, endorsed by NIH, USA, we quantified the proportion of collagen-positive area relative to the total bone marrow area within each microscopic field.

Scanning electron microscopy analysis

SEM was employed to examine the morphological features and quantify the adherent S. aureus population on self-tapping screw surfaces. Prior to dissecting the right femurs from the experimental mice, the screws were carefully removed and gently rinsed twice with PBS to eliminate non-adherent cells and tissue debris, followed by overnight fixation at 4 °C in a mixed fixative containing 2.5% glutaraldehyde and 4% paraformaldehyde.

After fixation, a graded ethanol dehydration process was performed, culminating in 100% ethanol. The screws were then transferred to a vacuum freeze dryer (Labgene, Switzerland) for 48 hours to ensure complete dehydration. Subsequently, the screws were mounted onto aluminum stubs using conductive carbon adhesive tape and sputter-coated with gold to optimize electron conductivity. Finally, high-resolution imaging was performed using a Phenom Pharos G2 SEM (Phenom World, Netherlands) at 15 kV acceleration voltage. Bacterial adhesion patterns and screw surface topography were visualized using Phenom Pro Suite software (version 3.2). S. aureus adherence to the screw surface was quantified by enumerating bacterial cells in three randomly selected 1500 μm² fields per screw using ImageJ (FIJI, NIH).

Micro-CT analysis

The right femurs of both mock- and S. aureus-infected mice were scanned using a high-resolution micro-CT system (SkyScan1276, Bruker, Germany) operated with manufacturer’s acquisition software (version 1.1). All scans were performed at a voxel resolution of 6 µm with the following parameters: 145 mA current, 55 kV voltage, and 400 ms integration time. Three-dimensional (3D) reconstruction was performed using NRecon software (version 1.7.5.0, SkyScan). For quantitative analysis, CTAn software (version 1.15.4.0, SkyScan) was employed. Image visualization was conducted using DataViewer (version 1.5.6.2, SkyScan) for 2D analysis and CTvox (version 3.3.0, SkyScan) for 3D rendering. Cortical bone parameters, including cortical bone loss and reactive bone formation, were evaluated across 1200 consecutive slices (7.2 mm) spanning the screw implantation site. For calibration of bone mineral density (BMD), we utilized ceramic reference samples. Trabecular microarchitecture was evaluated in a 1.2 mm-long region beginning 0.444 mm distal to the growth plate. Multiple parameters were analyzed for the distal femur, including bone mineral density (BMD), the ratio of bone volume to total volume (BV/TV), trabecular number (Tb. N), trabecular thickness (Tb. Th), and trabecular pattern factor (Tb. Pf).

Flow cytometry analysis

All flow cytometric analyses were carried out using a Beckman CytoFLEX Flow Cytometer with CytExpert Software (version 2.5.0.77, Beckman Coulter) and analyzed using FlowJo software (version 10.4, BD LifeSciences San Jose, CA, USA).

For flow cytometric analysis of primary mouse BMpreA, following 3 days of adipogenic differentiation of bone marrow stromal cells (as described in the Cell culture subsection), cells were treated with 10 μg/mL brefeldin A for 4 hours. They were subsequently detached using Trypsin-EDTA solution and fixed/permeabilized with BD Cytofix/Cytoperm solution (BD Cytofix/Cytoperm, #51-2090KZ) on ice for 10 minutes. After washing with flow buffer (0.1% BSA in PBS), cells were resuspended at 1 × 10⁶ cells /100 μL flow buffer and incubated with rabbit anti-SCF (Proteintech, #26582-1-AP, 1:100) and goat anti-Adiponectin (R&D, #AF119, 1:200) antibodies at RT for 1 hour. Next, cells were incubated with Alexa Fluor® 647 conjugated goat anti-rabbit IgG polyclonal Antibody (HuaBio, #HA1106, 1:200) and FITC-conjugated donkey anti-goat IgG H&L (Bioss, #bs-0294D-FITC, 1:200) secondary antibody at RT for one hour. Dead cells and debris were excluded based on forward scatter (FCS) and side scatter (SSC) profiles.

To determine whether myofibroblasts can modulate the local effector immune response, we harvested whole femoral bone marrow cells from DT-treated AdipoqCre-iDTR mice and DT-treated iDTR control mice, with and without S. aureus-induced femoral osteomyelitis. To identify AREG-producing cell populations during S. aureus infection, we collected bone marrow cells from infected femurs and sham-implanted controls. All cell suspensions were filtered through a 70-μm strainer and treated with ACK red cell lysis buffer (LEAGENE, #CS0001, China). Cells were resuspended at 1 × 10⁶ cells/100 μL in flow buffer containing Zombie Aqua dye (BioLegend, #423101, 1:100) and incubated at RT for 15 minutes. After washing, cells were blocked with anti-mouse CD16/32 antibody (BioLegend, #101319, 1:50) on ice for 10 minutes, followed by staining with fluorochrome-conjugated antibodies on ice for 30 minutes. For intracellular ROS measurement, cells were incubated with 10 μM dihydroethidium (YEASEN, #50102ES02) in flow buffer at RT for 1 hour following antibody incubation.

The antibodies used in this study include: FITC anti-mouse CD11b (BioLegend, #101205, 1:200), APC/Fire 750 anti-mouse F4/80 (BioLegend, #123151, 1:40), PerCP/Cyanine5.5 anti-mouse Ly6G (BioLegend, #127616, 1:80), APC/Cyanine7 anti-mouse Ly6C (BioLegend, #128025, 1:80), anti-mouse CD11c PE (eBioscience, #12-0114-81, 1:40), PerCP/Cyanine5.5 anti-mouse MHCII (BioLegend, #107625, 1:250), FITC anti-mouse CD3 (Biolegend, #100203, 1:50), PE anti-mouse CD4 (Biolegend, #100407, 1:80), APC anti-mouse CD8a (Biolegend, #100711, 1:80), BV421 anti-mouse CD86 (Biolegend, #105031, 1:20), PerCP anti-mouse CD19 (Biolegend, #115531, 1:80), and Amphiregulin antibody Alexa Fluor 647 (Santa Cruz, #SC-74501 AF647, 1:100).

RNA-sequencing data analysis

We revisited a set of published transcriptomic data from both mock- and S. aureus-infected mouse femurs, taken at day 14 post-surgery (GEO: GSE166522)²⁷. Using the DESeq2 package (version 1.16.1)⁷⁹, we identified genes as differentially expressed (DEGs) based on criteria of a Benjamini-Hochberg adjusted p value below 0.05 and an absolute fold change of 2 or more. For additional enrichment analyses, including Gene Ontology (GO) analysis and GSEA, were conducted using the clusterProfiler package (version 4.8.1)⁸⁰. To graphically represent the enriched biological processes uncovered by the GO analysis, we used the ggplot2 package (version 4.3.0)⁸¹. Only GO items with an adjusted p value below 0.05 were deemed to represent significant enrichments. We also generated Venn diagrams with the VennDiagram package (version 1.6.20)⁸² to delineate shared DEG profiles across various signaling pathways. For the GSEA analysis and subsequent data visualization, we employed the GseaVis package (version 0.0.8) (https://github.com/junjunlab/GseaVis) and enrichplot package (version 1.20.0) (https://bioconductor.org/packages/enrichplot). We established the significance thresholds of values as an absolute NES of at least 1, a p value under 0.05, and a Benjamini-Hochberg false discovery rate below 0.25.

Cell culture

The mouse 3T3-L1 preadipocyte cell line (ATCC, stock no. CL-173) was cultured in a humidified incubator at 37 °C with a 5% CO₂ atmosphere. Cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM, GIBCO) supplemented with 10% bovine calf serum and 1% penicillin/streptomycin. Cells were seeded at 5 × 10⁵ per well in a six-well plate for quantitative PCR (qPCR) assay and at 1 × 10⁶/60 mm dish for western blot assay.

For primary mouse bone marrow preadipocytes (BMpreA) culture, the bone marrow was flushed from the tibias and femurs of 4–5-week-old C57BL/6 mice using α-Minimum Essential Medium (αMEM, GIBCO) and filtered through 70-μm cell strainer. For initial expansion, cells were seeded at 40–50 × 10⁶ per 100 mm dish and maintained in growth media consisting of αMEM supplemented with 20% fetal bovine serum (FBS, ThermoFisher, #10091148). Upon reaching 80–90% confluence, cells were trypsinized and replated at 0.5 × 10⁶/well in a 6-well-plate for qPCR analysis and at 3 × 10⁶/100 mm dish for western blot analysis. Cells were cultured to confluence in growth medium before being switched to an adipogenic medium (Cyagen, #MUXMX-90031, China) for 3 days. To determine the bone marrow Adipoq⁺ preadipocytes phenotype, cells were characterized using flow cytometry with antibodies against Adiponectin and SCF (please refer to the subsection of Flow cytometry analysis).

3T3-L1 preadipocytes and primary BMpreA were cultured in DMEM growth medium containing either 100 ng/mL amphiregulin (AREG; Peprotech, #315-36) or 100 ng/mL heparin-binding EGF-like growth factor (HBEGF; Glpbio, #GP20274). For inhibitor studies, the following concentrations were employed: 5 μM gefitinib (YEASEN, #52717ES60), 1 μM rapamycin (Selleck, #S1039), 0.5 μM PD0325901 (Selleck, #S1036), 1 μM SCH772984 (Selleck, #S7101), 3 μM ruxolitinib (Selleck, #S1378), 2.5 μM stattic (MedChemExpress, #HY-13818), 5 μM AZD5363 (Selleck, #S8019), 0.1 μM EVP4593 (Apexbio, #A4217), and 1 μM verteporfin (MedChemExpress, #HY-B0146).

During infection experiments, 3T3-L1 preadipocytes or primary BMpreA were exposed to either S. aureus suspension at a multiplicity of infection (MOI) of 10, or an equal volume of PBS (control) for 30 minutes. Following this procedure, cells were washed twice with PBS to remove non-adherent bacteria. To eliminate residual extracellular S. aureus, cells were treated with bactericidal medium containing lysostaphin (20 µg/mL; Bioss, #D10379) and gentamicin (50 µg/mL; Sigma, #E003632) for 1 hour at 37 °C. Following treatment, cells were maintained in fresh complete medium for 24 and 48 hours before harvesting for total RNA extraction.

Preparation of BMDMs, neutrophils, and monocytes

In order to culture primary BMDMs, bone marrow cells were harvested from the femurs and tibias of 6–8 weeks old C57BL/6 mice. The cells were then dispersed through gentle pipetting and filtered through a 70 µm cell strainer. Red blood cells were lysed using ACK lysis buffer. Afterwards, the remaining bone marrow cells were plated at 1 × 10⁶/well in 6-well-plate for qPCR analysis and 5 × 10⁶/100 mm dish for conditioned medium collection. Cells were maintained in RPMI 1640 medium supplemented with 10% FBS and either 30% L929 cell-conditioned medium or 20 ng/mL recombinant macrophage colony-stimulating factor for 7 days to induce macrophage differentiation.

To isolate neutrophils and monocytes, bone marrow cells were flushed from tibias and femurs of 8–10-week-old C57BL/6 mice, gently resuspended, and layered onto a three-tiered Percoll gradient (72%, 62%, and 52%) in 15 mL tube. Following centrifugation at 1540 × g for 35 minutes at 24°C, neutrophils were harvested from the 72%/62% interface, while mixed populations of monocytes and lymphocytes were collected from 62%/52% interface. The mixed cell population was then carefully overlaid onto fresh 60% Percoll and centrifuged under similar conditions. Post-centrifugation, monocytes were collected from above the 60% layer and washed twice with PBS. Isolated bone marrow neutrophils and monocytes were seeded at 5 × 10⁶/well and 2 × 10⁶/well, respectively, in six-well-plate, and maintained in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin/streptomycin.

In the in vitro experiment, primary BMDMs, neutrophils, and monocytes were exposed to S. aureus suspensions (MOI = 10) or PBS control for 1 hour. Following infection, extracellular bacteria were eliminated through one-hour incubation in bactericidal medium containing 20 µg/mL lysostaphin and 50 µg/mL gentamicin. Cells were then washed three times with PBS to remove bacterial debris and antibiotics, followed by maintenance in fresh medium with 1% penicillin/streptomycin. Total RNA was extracted at designated time points (6, 12, and 24 hours post-infection) for subsequent analysis.

To evaluate the effect of intracellular and extracellular S. aureus on the AREG mRNA level in macrophages, BMDMs were challenged with S. aureus (MOI = 10) for 1 hour with or without 10 μM cytochalasin D (phagocytosis inhibitor; YEASEN, #53215ES03). Following infection, extracellular bacteria were eliminated by one-hour treatment with 20 µg/mL lysostaphin and 50 µg/mL gentamicin. Cells were then maintained for 24 hours in RPMI 1640 medium supplemented with 10% FBS and 1% penicillin/streptomycin before total RNA extraction.

To investigate the impact of AREG on EGFR and mTOR signaling in 3T3-L1 preadipocytes, cells were treated with recombinant AREG (Perprotech, #315-36, 100 ng/mL) and collected at designated time points (0, 15, 30, and 60 minutes post-infection) for protein analysis. To explore the involvement of the EGFR/mTOR pathway in YAP activity, cells were pretreated for 1 hour with either gefitinib or rapamycin prior to 6 hours of stimulation with AREG or S. aureus-MФ CM. Following treatments, cells were harvested for cytoplasmic/nuclear fractionation and western blot analysis.

Preparation of macrophage-derived conditioned medium and treatments

Primary BMDMs were exposed to S. aureus suspension (MOI = 10) or PBS control for 1 hour, then treated with lysostaphin and gentamicin to eliminate any extracellular bacteria. After treatment, BMDMs were maintained in fresh medium containing 1% penicillin/streptomycin for 24 hours. The resulting culture supernatants from both vehicle-treated (uninfected) and S. aureus-infected were collected, centrifuged at 1000 × g for 15 minutes, and sterile-filtered through 0.22 µm filter to ensure complete bacterial removal. These filtered supernatants were then mixed 1:1 with preadipocytes growth medium to prepare control conditioned medium (Vehicle-MФ CM) and S. aureus-infected conditioned medium (S. aureus-MФ CM), respectively.

To explore the influence of macrophage-secreted factors induced by S. aureus on myofibroblast marker gene expression in 3T3-L1 preadipocytes or primary BMpreA, cells were incubated in either Vehicle-MФ CM or S. aureus-MФ CM for 48 hours. Subsequently, total RNA was extracted, and the mRNA levels of key myofibroblast markers (Acta2, Tagln, Col1a1, Lox, and Loxl2) were measured using qPCR.

RT-qPCR

Total RNA from cells was extracted utilizing RNAiso PLUS reagent (Takara, 9109). According to the manufacturer’s protocol, reverse transcription and quantitative PCR were conducted on the QuantStudio 5 system (Applied Biosystems, USA) with QuantStudio Design and Analysis software (version 1.3) using the Evo M-MLV RT Premix Kit (Accurate Biology, #AG11706) and SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biology, #AG11701). For normalization purposes, the Actb gene served as the endogenous control. Relative quantification of target gene expression was carried out employing the 2^-∆∆CT method. Detailed sequences of the primers used for RT-qPCR in this research are provided in Table S2.

Enzyme-linked immunosorbent assay (ELISA)

Proceeding as described, BMDMs were maintained for 24 hours post-extracellular bacterial removal with lysostaphin and gentamicin. Subsequently, the culture supernatants from both uninfected and S. aureus-infected primary mouse BMDMs were harvested. These supernatants were then centrifuged (1000 × g, 15 minutes) and filtered through 0.22 µm membranes to eliminate any residual contaminants. Post-filtration, the supernatants were stored at −80 °C for subsequent analyses.

Soluble AREG and HBEGF concentrations in cell-free supernatants were quantified using Mouse AREG ELISA Kit (Biorbyt, #orb774868) and Mouse HBEGF ELISA Kit (Biorbyt, #orb775164) according to manufacturer’s protocols. Optical density values were measured at 450 nm using a SpectraMax i3x microplate reader (Molecular Devices, USA) with SoftMax Pro 6 software (version 6.5.1). Analyte concentrations were determined using CurveExpert 1.4 software (Hyams Development, USA) through optimized standard curve fitting based on serial dilutions of reference standards.

Cytoplasmic and nuclear protein extraction

Cytoplasmic and nuclear fractions were isolated using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, #P0027) following the manufacturer’s protocols. Collected cells were resuspended in pre-chilled cytoplasm extraction buffer and incubated on ice for 10 minutes. After 30-second vigorous vortex, samples were centrifuged at 13,523 ×g for 5 minutes at 4 °C. The resulting supernatant was carefully collected as the cytoplasmic fraction. The remaining pellet was resuspended in nuclear extraction buffer and subjected to 10 cycles of 1-minute vortex and 2-minute ice incubation. Finally, samples were centrifuged at 13,523 ×g for 10 minutes at 4 °C to obtain the nuclear fraction supernatant.

Immunocytochemistry

3T3-L1 preadipocytes were seeded at a density of 5 × 10⁴ per well in 35 mm confocal dishes and treated with either 100 ng/mL AREG or S. aureus-MФ CM for predetermined durations (0, 0.5, 6, and 12 hours). For BMpreA, cells were trypsinized after 3 days of adipogenic differentiation, replated at identical density (5 × 10⁴ per well) in confocal dishes, and treated with AREG or S. aureus-MФ CM for 6 hours. Following treatment, cells were washed twice with PBS and fixed with 4% paraformaldehyde for 15 minutes. Cells were permeabilized with 0.5% Triton X-100 in PBS for 20 minutes, blocked with 10% goat serum for 60 minutes. Cells were then incubated at 4 °C overnight with anti-YAP primary antibody (CST, #14074, 1:200), followed by 1 hour RT incubation with Alexa Fluor 594-conjugated goat anti-rabbit IgG secondary antibody (CST, #8889, 1:400) and subsequent DAPI nuclear counterstaining. Confocal imaging was carried out using a Zeiss LSM980 microscope (Germany) with the ZEN blue software (version 3.8). The relative YAP fluorescence intensity in each cell (normalized to controls) and the number of nuclear YAP⁺ puncta per cell (denoted as N. nuclear YAP⁺ puncta) were quantified using ImageJ (FIJI) software (NIH, USA). For this quantitative assessment, 20 cells from each treatment group were randomly selected and analyzed.

Western blot

Post-treatments, cells were washed twice with ice-cold PBS and lysed in RIPA buffer containing 1% protease inhibitor and 1% phosphatase inhibitor at 4 °C for 30 minutes. Protein concentrations were determined using a BCA Protein Assay Kit (Beyotime, P0010). Equal protein amounts were separated by SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes. Membranes were blocked with 5% skim milk for 90 minutes at RT and incubated with primary antibodies overnight at 4 °C. The following day, membranes were incubated with horseradish peroxidase-linked secondary antibodies (1 hour, RT). Primary antibodies employed in the analysis were p-EGFR (HuaBio, #ET1606-44, 1:2000), EGFR (HuaBio, #ET1604-44, 1:2000), p-mTOR (Proteintech, #67778-1-lg, 1:10,000), mTOR (Proteintech, #66888-1-lg, 1:10,000), p-YAP (CST, #13008, 1:3000), YAP (CST, #14074, 1:3000), Lamin B1 (Proteintech, #66095-1-Ig, 1:20,000), and β-Actin (Proteintech, #66009-1-Ig, 1:20,000). Protein bands was visualized using the BLT GelView 6000 Pro chemiluminescence imaging system (Biolight Biotechnology, China) with the manufacturer’s acquisition software (version 1.0.0.5) following ECL substrate application. Band intensities were quantified by densitometric analysis using ImageJ (FIJI) software provided by NIH, USA.

Statistical analysis

Statistical analyses were conducted using IBM SPSS Statistics software version 20. Data were graphically presented using GraphPad Prism software (version 9.4.0). The results were expressed as means ± standard errors of the mean (SEM), where ‘n’ stands for the number of mice or cells evaluated in each independent experiment. The normality of the data was assessed using either the Shapiro-Wilk test or the Kolmogorov-Smirnov test. For comparisons between two independent groups, the unpaired two-tailed Student’s t test was utilized. For multi-group comparisons, one-way analysis of variance followed by either Fisher’s LSD (for equal variance) or Dunnett’s T3 (for unequal variance) post hoc test was applied. A p value of

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.