The Chinese formula BFJD consists of 12 herbs (17), including Pyrola calliantha Andres [Ericaceae] (LuXianCao, LXC) 24 g, Prunella vulgaris L. [Lamiaceae] (XiaKuCao, XKC) 24 g, Scleromitrion diffusum (Willd.) R.J.Wang (BaiHuaSheSheCao, BHSSC) 24 g, Astragalus mongholicus Bunge [Fabaceae] (HuangQi, HQ) 15 g, Stemona japonica (Blume) Miq. [Stemonaceae] (BaiBu, BB) 12 g, Fagopyrum cymosum (Trevir.) Meisn. (JinQiaoMai, JQM) 15 g, Viola philippica var. philippica [Violaceae] (ZiHuaDiDing, ZHDD) 9 g, Polygonatum sibiricum Redouté [Asparagaceae] (ZhiHuangJing, ZHJ) 9 g, Atractylodes macrocephala Koidz. (BaiZhu, BZ) 9 g, Poria cocos (Schw.)Wolf (FuLing, FL) 9 g, Bletilla striata (Thunb.) Rchb.f. [Orchidaceae] (Bai Ji, BJ) 3 g and Lithospermum erythrorhizon Siebold & Zucc. [Boraginaceae] (Zi Cao, ZC) 9 g, qualified in quality inspection according to the Chinese Pharmacopoeia. All traditional Chinese medicine (TCM) components were purchased and provided by Shanghai traditional Chinese medicine Pharmaceutical Technology Co., Ltd. All the components were extracted together two times, adding 10 and 8 times the amount of water, respectively. The extraction time was boiling for 45 minutes and 30 minutes, respectively. The extraction solution was filtered by 80-mesh sieve, and concentrated it at 60 °C to a density of approximately 1.13-1.15 mg/ml. The lyophilized powder yield was approximately 21.8%.

Female wild type C57BL/6J mice, aged 6–8 weeks, were procured from Charles River Laboratory Animal Technology Co. Ltd. (Zhejiang, China). The animal experiments were approved by the Institutional Animals Care and Use Committee (IACUC) of Institute Pasteur of Shanghai Chinese Academy of Science (Approval No. P2022014).

The chemical components and serum and lung tissue prototype compounds of BFJD were identified by ultra-high performance liquid chromatography coupled to quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF/MS) system. To prepare the BFJD solution, 2 g of the BFJD extracts was weighed and extracted with 20 ml of 80% methanol (v/v) using ultrasonication (40 kHz) for 30 min, followed by centrifugation at 12,000 rpm for 5 min at room temperature. After extraction, the filtrate was collected using a 0.22 μm filter at a concentration of 0.1 g/ml, and subsequently, 1.0 μl of the filtrate was injected into the UPLC-Q-TOF/MS system for analysis.

Medicated and control blood samples were collected from Sprague-Dawley rats that were orally administered BFJD (2.20 g/ml, 1 ml/100 g body weight) and distilled water, respectively. Blood samples were collected at 0.5 h, 1 h, 1.5 h and 2 h post-final gavage. The samples were subsequently centrifuged at 3,000 rpm at 4°C for 10 minutes, and the upper layer of serum was collected. Methanol was added to the serum samples three times, mixed, and then centrifuged at 12,000 rpm at 4°C for 15 minutes. The supernatant was dried using nitrogen gas. The remaining residue was dissolved in 100 μl of 80% methanol (v/v) and centrifuged once more at 12,000 rpm at 4°C for 15 minutes. The collected supernatant for subsequent analysis.

For the preparation of the lung tissue solution, 1 g of control lung tissue and 5 g of medicated lung tissue were each cut into pieces and immersed in methanol at a 1:10 ratio, followed by ultrasonication for 30 min. After standing for 30 min at room temperature, the supernatant was obtained by centrifugation at 12000 rpm and 4°C about 15 min. The supernatant was then concentrated and dried to yield a residue. The residue was redissolved using 60μL and 300μL of 80% methanol (v/v), respectively, and centrifuged again at 12,000 rpm at 4°C for 15 minutes. The supernatant was then collected for subsequent analysis.

MS was performed using an AB Sciex Triple TOF^® 4,600 (AB SCIEX, Foster City, CA, USA) instrument to identify the BFJD solution. The optimized MS parameters were as follows: ion source temperature: 500°C; curtain gas: 35 psi; ion source gas 1 and 2: 50 psi; ion spray voltage: 5000 V (positive), 4500 V (negative); declustering potential: 100 V (MS and MS/MS); collision energy 10 eV; collision energy spread 20eV (MS/MS); mass range: 100–1500 m/z (MS), 50–1700 m/z (MS/MS); ion release delay: 30 ms; ion release width: 15 ms. Subsequently, to identify the serum and lung tissue prototype compounds in the BFJD, the Agilent 6545 Q-TOF (Agilent, Santa Clara, California, USA) was employed with the following operational parameters: gas temperature, 320°C; drying gas, 8 l/min; nebulizer, 35 psi; shealth gas temperature, 350°C; shealth gas flow, 11 l/min; Vcap, 4000 V; nozzle voltage, 1000 V; Fragmentor, 175 V; skimmer, 60 V; collision energy spread 20eV (MS/MS); mass range, 50–1700 m/z (MS and MS/MS).

The MRSA strain ST239 (18) was a generous donation from Professor Wenhong Zhang from Huashan Hospital Affiliated to Fudan University. The bacteria were cultured in Luria-Bertani (LB) medium at 37°C with shaking at 200 rpm and harvested by centrifugation at 3000 rpm for 5 min during the Mid-Log phase. The bacterial pellet was then washed twice with sterile phosphate-buffered saline (PBS). Subsequently, the bacterial culture was resuspended in PBS, and the colony forming units (CFU) were adjusted to the desired concentration using a Turbidimeter (BioMérieux).

The mice were inoculated with 1×10⁸ CFU of MRSA in 0.2 ml of PBS via a lateral tail vein. Twenty-eight days post-infection, the mice were divided into three groups: the Model group, the BFJD group, and the Linezolid group. After 14 days of treatment, the mice were euthanized and sampled. The bacterial load in the lungs, liver, and kidneys was determined by preparing organ homogenates in sterile PBS and plating 10-fold serial dilutions on LB agar.

The clinical dosage for an adult weighing 70 kg of BFJD is 162 g (crude drug) daily. Based on the body surface area index and lyophilized powder yield (21.8%), the animal medium dosage (4.59 g/kg/d) of BFJD is calculated to be equivalent to the clinical dosage. Linezolid was purchased from Abmole (HOU, USA) and dissolved in dimethyl sulfoxide (DMSO) with a final solution volume not exceeding 0.5%, and subsequently diluted in sterile water. The final oral concentration was 50mg/kg/d (19).

RAW264.7 macrophages were seeded into a 96-well plate at a density of 1×10⁴ cells per well. A blank group was included as a background control. After the cells adhered fully, the previous culture medium was removed, and different concentrations of BFJD solution (50-800 μg/mL) and active compounds (0-200 μM), prepared in DMEM complete medium, were added for 24 or 48 hours. Following the intervention, the culture medium was removed, and CCK-8 reagent (Beyotime, Shanghai, China) was added. The cells were incubated for 30–60 minutes, and the absorbance at 450 nm was measured using a microplate reader to calculate cell viability.

The macrophage intracellular MRSA persistence infection model was established (5, 9) to examine the impact of BFJD on macrophage-mediated clearance of intracellular MRSA. RAW264.7 macrophages were seeded at 2×10⁵ cells per well. The multiplicity of infection (MOI) was set at 20, and the phagocytosis period was established at 1 hour. Extracellular residual MRSA was eliminated through treatment with lysostaphin. The cells were then cultured for 24 hours in a CO₂ incubator to mimic the intracellular MRSA colonization environment. Subsequently, the corresponding concentrations of BFJD solution (20, 21), IFN-γ, and linezolid were added to the wells containing DMEM infection medium. The cells were cultured for an additional 24 hours. The experimental groups consisted of: model group, BFJD group (100 μg/mL, 200 μg/mL, 400 μg/mL), linezolid group (6 μg/mL), IFN-γ group (40 ng/mL), with three replicates for each group.

For the MRSA infection animal experiment, the left lung lobe was minced and digested with 50 μg/ml Liberase TM (Roche) and 1 μg/ml DNase I (Sigma) for 45 min at 37°C. Subsequently, cells were surface-stained for 30 minutes on ice with experimentally designed combinations of the following antibodies: Live-Dye eFluor™ 780e, CD45-BV510, CD11b-PE, F4/80-FITC, CD80-Brilliant Violet 421™, and CD206-APC. After staining, the cells were fixed, and flow cytometry was performed using a Beckman CytoFlex S (Beckman, USA) with CytExpert Software (v.2.0).

For the intercellular MRSA persistent infection experiment, antibodies selected were Live-Dye eFluor™ 780e, CD80-Brilliant Violet 421™, and CD206-APC. The experimental cells were collected, washed, stained, fixed, and analyzed within 48 hours using flow cytometry.

For the bone marrow-derived macrophages (BMDM) repolarization experiment, bone marrow-derived monocytes were extracted from wild-type mice and differentiated into macrophages using M-CSF (20 ng/mL). To induce M2 polarization, IL-4 and IL-13 (100 ng/mL) were added. The flow cytometry antibodies used were Live-Dye eFluor™ 780e, CD11b-PE, F4/80-FITC, CD80-Brilliant Violet 421™, and CD206-APC. The experimental cells were collected, washed, stained, fixed, and analyzed within 48 hours using flow cytometry.

Following the establishment of persistent MRSA infection, cells were sampled at different time points—1 hour, 24 hours, and 48 hours post-intracellular infection—and the concentration of the BFJD intervention was set at 400 μg/ml during the 24–48 hours of intracellular infection. Per condition, four biological replicates were performed. Total RNA was extracted by Trizol Reagent, and the RNA quality was assessed using Qubit^®3.0 Fluorometer (Life Technologies, CA, USA) and Nanodrop One spectrophotometer (Thermo Fisher Scientific Inc, USA). To establish an RNA library, the mRNA polyA-based enrichment method was performed. The RNA samples were sequenced using DNBSEQ-T7 at the Wuhan BGI Technology Service Co. Ltd. Quality-filtered RNA-seq reads were aligned to the mouse genome (GRCm38/mm10; https://useast.ensembl.org/index.html) by STAR software (version 2.7.4). Principal component analysis (PCA) was performed based on the raw count values of genes from each sample and visualized. The raw count values for each gene were then converted into normalized FPKM (Fragments Per Kilobase of exon per Million fragments mapped) values. Differentially expressed genes were identified using the DESeq2 (Ver. 1.36.0) package with a threshold of fold change (FC ≥ 1.5, adj p-value ≤ 0.05). Volcano plots were generated to illustrate these genes. Venn diagram was drawn to visualize the differentially expressed gene clusters between groups, and the reversed genes were also identified and visualized. Gene Ontology (GO) and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analysis of the gene clusters were performed using the DAVID online tool, and the bubble plots were generated. All visualizations were performed using R software (Ver. 4.2.1).

Blood samples were collected and centrifuged at 1200×g for 10 min to isolate serum. Multiple cytokines in mouse serum were measured using the Biolegend Multiplex kit (Biolengend, CA, USA) according to the manufacturer’s instructions. Flow cytometric detection was performed, and the results were analyzed and exported using the Biolegend LEGENDplex™ QOGNIT platform (https://legendplex.qognit.com/).

ROS levels were detected using ROS Assay Kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions. The fluorescence intensity was analyzed using flow cytometry or a fluorescence microscope. MitoSoX was detected using the MitoSoX Assay Kit with MitoSoX™ Red (Beyotime, Shanghai, China), also following the manufacturer’s instructions. The fluorescence intensity was analyzed by flow cytometry.

BMDM cells were transfected with the CD40 siRNA-mate plus kit (GenePharma, Shanghai, China) according to the manufacturer’s instructions. To elaborate, BMDMs were treated with 30 pmol of CD40 siRNA in serum-free medium containing the transfection reagent for 48 hours. The efficiency of CD40 expression knockdown was subsequently verified through flow cytometry analysis (22). The CD40-silenced BMDMs were then subjected to MRSA infection using the established experimental protocol. Supplementary Table S1 provides the siRNA sequence for gene knockdown.

RAW264.7 cells stably expressing NF-κB-dependent SEAP were constructed as previously described (23). Following the establishment of persistent MRSA infection, BFJD solution was combined with the NF-κB inhibitor BAY-11-7082 (10 μM) and ROS inhibitor NAC (5 mM) and added to the wells. BAY 11–7082 and NAC were purchased from MedChemExpress (NJ, USA). The cells were then cultured for an additional 24 hours. NF-κB activity was assessed using the Firefly Luciferase Reporter Gene Assay Kit (Beyotime Biotechnology, shanghai, China). Cells were lysed with the passive lysis buffer provided in the kit. Subsequently, 50ul of luciferase detection solution was added to the supernatant of each well with a BeyoGold™ 96 well microplate (Beyotime Biotechnology), and detection was performed on a full wavelength multifunctional microplate detector (Infinite^® E Plex, TECAN, Switzerland).

To determine the macrophage subtypes in the lungs, we conducted immunofluorescence staining on lung tissue sections. The process involved a series of standardized procedures, including tissue sectioning, dewaxing, antigen retrieval, endogenous peroxidase blockade, and serum blocking. Subsequently, the lung tissue sections were incubated with primary antibodies against CD68 (BA3638, Boster, 1:200), CD86 (CY5238, ABways, 1:200) and CD206 (PTG, 60143-1-Ig, 1:200) overnight at 4°C. The slices were then further incubated with HRP-conjugated goat anti-rabbit IgG for 60 minutes at 37°C, followed by fluorescein-conjugated anti-HRP for 30 minutes at 37°C. In this process, CD68 corresponds to TYR-555, CD86 to TYR-488, and CD206 to TYR-651. The nuclei were counterstained with DAPI (28718-90-3, Sigma, 1:1000) for 60 minutes at room temperature. Images were captured using an inverted fluorescence microscope (Leica, Germany).

Molecular docking simulations were performed using the AutodockTools (v.1.5.6) and Autodock Vina software (v.4.2). The structure of active compounds were downloaded from the Pubchem database, and then hydrogenated by AutodockTools. The structure CD40 (PDB ID: 5dmi) were obtained from the RCSB (https://www.pdb.org/) and introduced into AutodockTools for dehydration and hydrogenation, facilitating ligand separation. Autodock Vina was utilized to conduct molecular docking of target proteins with active compounds and to calculate their free binding energies. Results with binding energies lower than -5 kcal/mol were selected for analysis of local binding conformations and hydrogen bonding interactions using Discovery Studio Visualizer (v. 25.1) software.

The Student’s t-test or one-way analysis of variance (ANOVA) was employed for statistical analysis, and differences between groups were compared. Statistical analysis and graphs were generated using GraphPad Prism 8.0 software (GraphPad Software Inc., San Diego, CA, USA). p-value of less than 0.05 was considered statistically significant.

To explore the antibacterial effect of BFJD, we initially established acute and subacute infection model with 1×10⁸ CFU of MRSA in 0.2 ml of PBS via a lateral tail vein ( Supplementary Figure S1A ). As shown in Supplementary Figure S1B , regardless of whether in the clinical equivalent dose of BFJD (4.59 g/kg/d) or the clinical double dose of BFJD (9.18 g/kg/d) group, BFJD intervention from day 0 to day 4 after infection, and BFJD intervention from day 4 to day 12 after infection, the bacterial load did not show a significant reduction in the lung, liver and kidney, compared to the model group. Given that BFJD did not show obvious antibacterial effects in acute and subacute infection, a murine model of persistent MRSA infection were established to further validate its effects in persistent infection ( Figure 1A ). In contrast to acute and subacute infection, the bacterial load in the lungs, liver, and kidneys were significantly decreased after the treatment of BFJD, suggesting that BFJD may exert certain antibacterial effects in the MRSA persistent infection ( Figure 1B ).

To further investigate the antibacterial mechanism of BFJD, flow cytometry analysis of lung cell populations from MRSA persistent infection mice was performed to verify the immune regulation of BFJD. The gating plots and scatter plots were shown in Figures 1C, E .The statistical results ( Figure 1D ) indicated that the proportion of macrophages in mice with persistent MRSA infection was reduced compared to uninfected mice, which may be associated with macrophage death mediated by bacteria during persistent infection (24, 25). However, both BFJD and linezolid increased the proportion and quantity of macrophages when compared with infected mice. Furthermore, BFJD increased the proportion of M1 (CD80⁺CD206⁻) macrophages and enhanced the mean fluorescence intensity (MFI) of the M1 surface marker CD80 ( Figure 1F ). Given the immune regulatory effects of BFJD on macrophages in MRSA persistent infection, an immunofluorescence assay was performed for further verification. As shown in the Figure 1G , the green fluorescence (CD86) was increased upon BFJD intervention, suggesting that BFJD increased the proportion of M1 in the lung tissue. In conclusion, these results suggest that BFJD may exert its antibacterial effect in MRSA persistent infection through macrophage-mediated immune regulation.

Furthermore, we investigated the killing effect of BFJD on MRSA persister. Initially, we verified the direct activity of BFJD on MRSA growth inhibition through minimal inhibitory concentration (MIC) determination. The results showed that BFJD, even at a concentration of 1600 μg/mL, exhibited no direct inhibitory effect on MRSA. To simulate the MRSA persistent infection within macrophages, we adopted a persistent infection cell model based on previous studies (5, 9), which eliminates extracellular MRSA through lysostaphin ( Figure 2A ). At 0 h, 24 h, and 48 h post-infection, macrophages were lysed, and the intracellular bacterial load was quantified. The results showed that, although macrophages partially cleared intracellular MRSA over 48 hours, persistent colonization remained, confirming the limitations of macrophages in controlling persistent infection ( Figure 2B ). Our finding, illustrated in Supplementary Figure S2 , demonstrated that intervention with BFJD exhibited no cytotoxicity toward macrophages after 24 h or 48 h at concentrations ranging from 0 to 800 μg/mL. Subsequently, We treated macrophages with different concentrations of BFJD for 24 h and quantified the intracellular bacterial load. As depicted in Figure 2C , the intracellular bacterial burden was significantly reduced by BFJD at concentrations of 200 μg/mL and 400 μg/mL, as well as by IFN-γ (40 ng/mL). In contrast, linezolid did not promote killing of intracellular MRSA persister, not even at 3×MIC. Meanwhile, flow cytometry was performed to analyze the polarization of macrophage. The results showed that the proportion of M1 was promoted by BFJD at 100 μg/mL, 200 μg/mL, and 400 μg/mL, as well as IFN-γ (40 ng/mL), also indicated by a rightward shift in the CD80 overlay histogram ( Figures 2D, E ). Furthermore, the proportion of M2 (CD80⁻CD206⁺) was reduced by BFJD at 400 μg/mL and IFN-γ at 40 ng/mL ( Figures 2D, E ). These findings suggest that BFJD may exert its intracellular antibacterial effects by promoting M1 macrophage polarization, which is consistent with in vivo experiments.

Considering that M2 macrophage polarization may contribute to immunosuppressive effects in the context of MRSA persistent infection, we established a M2-polarized bone marrow-derived macrophage (BMDM) model to verify the effect of BFJD on macrophage repolarization. The results demonstrated that the proportion of M1 macrophages was increased by BFJD at 100 μg/mL, 200 μg/mL, and 400 μg/mL, as well as IFN-γ (40 ng/mL), also evidenced by a rightward shift in the CD80 overlay histogram. Additionally, the proportion of M2 was reduced by BFJD at 400 μg/mL and IFN-γ at 40 ng/mL, also evidenced by a leftward shift in the CD206 overlay histogram ( Figures 2F, G ). In summary, these findings suggest that BFJD promotes M1 polarization and M2-to-M1 repolarization under infection or cytokine-induced conditions, contributing to its potential effects in enhancing macrophage-mediated persistent bacterial clearance.

To further elucidate the mechanisms by which BFJD promotes the clearance of persistent MRSA infection in macrophages, RNA-seq analysis was conducted to comprehensively profile gene expression and functional alterations. Based on the results in Figures 2C–E , we observed that BFJD exhibited an optimal regulatory effect on macrophages at a concentration of 400 μg/mL, promoting the polarization of macrophages to clear intracellular bacteria. Consequently, this concentration was selected for RNA-seq analysis and subsequent experiments.

PCA indicated that all samples were within the 95% confidence interval, suggesting high within-group consistency and significant differences between groups ( Figure 3A ). A total of 3,362 differentially expressed genes (DEGs) were identified between 24h post-intracellular infection and the control, with 2,072 up-regulated and 1,290 down-regulated. Furthermore, 3,365 DEGs were determined between 48h post-intracellular infection compared to 24h post-intracellular infection, with 1,936 up-regulated and 1,429 down-regulated. DEGs were visualized using volcano plot ( Supplementary Figures S3A, B ). Intersection analysis revealed gene Cluster I, which consists of 391 reversed genes that were up-regulated at 24h but down-regulated at 48h post-intracellular infection. This cluster includes the down-regulation of inflammatory genes such as Il1b, Il6, Ccl2, etc ( Figures 3B, C ). Concurrently, GO-KEGG enrichment analysis of Cluster I further demonstrated that bacterial persistence down-regulated innate immune response, cytokine receptor binding, cytokine activity, and TNF signaling pathways ( Supplementary Figure S3D ).

To further understand the effect of BFJD on the transcriptome of macrophage with persistent infection, DEGs analysis was conducted between BFJD group and M_48h group. The DEGs were visualized using a volcano plot ( Supplementary Figure S3C ). Following BFJD intervention, 454 genes were upregulated and 391 genes were downregulated ( Figure 3D ). GO&KEGG enrichment analysis further revealed that BFJD regulated IL-1β and IL-6 production, response to ROS, NF-κB and TNF signaling pathways ( Figure 3E ), suggesting that these pathways may contribute to its intracellular antibacterial effects. Upon visualizing genes within the three most up-regulated KEGG pathways, we observed that BFJD significantly boosted the expression of M1 macrophage polarization markers, such as Cd40, Ccr2, Cxcl2, Tnf, and Il12b. The data also revealed a coordinated chemotactic response, evidenced by the simultaneous upregulation of chemokine receptors (Ccr1/2, Cxcr3) and their ligands (Ccl2/3/4/5, Cxcl2/3) ( Supplementary Figure S3E ). Intersection analysis was performed to align DEGs regulated by BFJD with Cluster I or Cluster II, which were defined as reversed genes. Subsequently, gene Clusters III (96 genes) and IV (74 genes) were identified as the key genes that were up-regulated or down-regulated by BFJD, respectively ( Figure 3D ). By screening these key genes, we found that the pro-inflammatory genes, which were suppressed in the persister state, were concentrated in Cluster IV. Through a heatmap of Cluster IV, we observed that these pro-inflammatory genes, including Il1b, Il6, Ccl2 and Cd40, were all restored after BFJD intervention ( Figure 3F ). In summary, these results indicate that MRSA persistent infection reprograms the host transcriptome, including the pro-inflammatory response and innate immunity, while BFJD may exert antibacterial effects by restoring these inflammatory responses in macrophages.

Redox regulation is associated with the immune-inflammatory response and M1 polarization (26). Based on GO analysis, we found that ROS, an effective antibacterial substances, may be closely associated with the immune effect of BFJD. To further investigate the effect of BFJD on ROS, we examined ROS expression using fluorescence microscopy ( Figure 4A ). The results indicated that MRSA infection could increase the intercellular ROS level, while BFJD further enhanced ROS production. In contrast, N-acetylcysteine (NAC), an ROS inhibitor, either alone or in combination with BFJD, reduced the ROS level ( Figure 4B ). Additionally, NAC treatment significantly promoted bacterial load and negatively impacted the antibacterial effect of BFJD ( Figure 4C ), suggesting that BFJD may enhance its antibacterial activity by promoting ROS production. As a critical intracellular signaling molecule, ROS plays a key role in regulating inflammatory signaling. Previous studies have demonstrated that ROS regulation of the NF-κB signaling pathway is essential for antibacterial immunity (27).

Thus, we employed RAW264.7 cells that stably express an NF-κB reporter gene to evaluate the effect of BFJD on NF-κB activity, using the NF-κB inhibitor BAY-11–7082 as a negative control. As shown in the Figure 4D , BFJD did not elicit NF-κB reporter gene expression in uninfected cells. However, NF-κB reporter gene expression was markedly increased by MRSA infection, and BFJD further augmented this expression. Moreover, NAC, whether administered alone or in conjunction with BFJD, inhibited NF-κB reporter gene expression, indicating that BFJD might modulate antibacterial immunity via the ROS-NF-κB signaling pathway.

To identify the key target through which BFJD activates the ROS-NF-κB pathway via cell membrane surface proteins, we analyzed the DEGs under BFJD intervention (22, 28, 29), hypothesizing that the transmembrane protein CD40 might be a critical target in macrophage-mediated clearance of intercellular MRSA infection. Transcriptomic data indicated that the expression of CD40 was significantly reduced at 48hpi compared to 24hpi. Notably, BFJD intervention substantially restored CD40 expression ( Figure 4E ), a finding corroborated by flow cytometric analysis of CD40⁺ macrophages ( Figure 4F ). Therefore, the CD40 inhibitor DRI-C21045 was applied for further validation. The bacterial load increased and negatively impacted the intracellular antibacterial effect of BFJD upon treatment with DRI-C21045 ( Figure 4G ), indicating that CD40 plays a crucial role in BFJD-mediated immune regulation. As a cell surface receptor, CD40 is known to positively regulate multiple inflammatory signaling pathways, including the ROS pathway (30, 31).Therefore, total ROS and MitoSoX levels were detected by flow cytometry. As shown in the Figures 4H, I , BFJD significantly increased the MFI of both total ROS and MitoSoX compared to the MRSA infection group, with the overlapping histograms in Supplementary Figures S4A, B ; however, these effects were reversed by treatment with DRI-C21045. Furthermore, genetic validation through siRNA-mediated CD40 knockdown in BMDMs further confirmed these findings. As clearly demonstrated in Figure 4J ; Supplementary Figure S4C , CD40 silencing abrogated BFJD-induced upregulation of CD40, along with ROS production ( Figure 4K ; Supplementary Figure S4D ) and bacterial clearance ( Figure 4L ). These consistent results from both pharmacological inhibition and genetic knockdown establish CD40 as the crucial membrane target to enhance macrophage-mediated bacterial clearance. Next, several cytokines and chemokines in persistent infection mouse serum were detected by LEGENDplex or ELISA analysis. The results demonstrated that BFJD significantly upregulated the serum levels of IL-1β, CCL-2, IL-6, and TNF-α ( Figure 4M ). Taken together, these results suggest that BFJD exerts its antibacterial effect against MRSA persistent infection through the CD40-ROS-NF-κB signaling pathway.

Given the complex constituents in TCM ingredients, the characterization of components in BFJD was achieved through UPLC-Q-TOF/MS. The chemical base peak ion chromatogram of BFJD exhibits distinct profiles in both negative and positive ion modes, as shown in Supplementary Figures S5A, B , respectively. A total of 115 chemical compounds were tentatively characterized within BFJD. Supplementary Table S2 provides comprehensive data. Furthermore, prototype compounds identified in the bloodstream and lung tissues from BFJD were selected to explore their potential as bioactive agents for targeting CD40 to exert antibacterial effect. By consulting the Natural Products HR-MS/MS Spectral Library database and relevant literature (32), a total of 26 prototype compounds were discovered, with 18 entering the bloodstream, 21 reaching the lungs, and 13 components entering both the bloodstream and the lungs. Supplementary Table S3 illustrates detailed information of the 26 compounds, and 20 commercially available compounds were selected for subsequent investigations ( Supplementary Table S4 ).

CD40 is considered crucial for the macrophage-mediated inflammatory responses during MRSA persistent infection. A total of twenty prototype compounds from BFJD were tested for their ability to promote CD40 expression using flow cytometry. Based on concentrations determined by cell viability assays ( Supplementary Figure S5C ), the primary screening identified nine compounds that could promote the expression of CD40 in macrophages ( Figure 5A ). Furthermore, molecular docking was utilized to clarify the potential binding modes of these nine compounds with CD40 (PDB ID: 5DMI). As indicated in Supplementary Table S5 , atractylenolide II and formononetin demonstrate high affinity for CD40, with binding energy values of −5.57 and −5.62 kcal/mol, respectively. Structural analysis revealed that atractylenolide II forms three hydrogen bonds with residues Glu150, Ala176, and Val152 ( Figure 5B ), whereas formononetin forms a single hydrogen bond with Pro44 ( Figure 5C ). Based on the concentrations indicated by the cell viability assays ( Figure 5D ), both atractylenolide II and formononetin demonstrated potent inhibitory effects on bacterial load. The IC50 values of these two active compounds were determined to be 9.716 μM and 46.79 μM, respectively ( Figure 5E ). In summary, these findings suggest that atractylenolide II and formononetin are key bioactive compounds in BFJD, contributing to its immunomodulatory effects and providing a foundation for further mechanistic exploration.