Vagus nerve stimulation alleviates S. aureus-induced mastitis by regulating gut microbiota S24-7-PPARγ and NF-ΚB/NLRP3 signaling in mice

Abstract

Gut microbiota dysbiosis has been implicated in the pathogenesis of mastitis. While the vagus nerve exerts well-documented anti-inflammatory effects and modulates gut microbiota, its potential influence on mastitis progression via gut microbiota modulation remains unclear. To investigate this, we employed vagus nerve stimulation (VNS) in Staphylococcus aureus (S. aureus)-induced mastitis in mice. We demonstrate that VNS significantly attenuated mammary gland inflammation and restored epithelial barrier integrity following S. aureus challenge. Crucially, antibiotic depletion of the gut microbiota abrogated the protective effects of VNS, and fecal microbiota transplantation (FMT) from VNS-treated mice conferred protection against mastitis, establishing a causal role for the gut microbiota in mediating the VNS effect. Specifically, VNS markedly increased the abundance of Muribaculaceae. in the gut. Replenishment with S24-7, a representative strain of this genus, alleviated S. aureus-induced mammary gland inflammation in mice. Transcriptomic analysis revealed that S24-7 exerted its effects by activating peroxisome proliferator-activated receptor gamma (PPARγ), which subsequently suppressed the NF-κB/NLRP3 signaling pathway. Overall, our findings suggest that targeting the vagus nerve - mediated Muribaculaceae/PPARγ axis may represent a promising strategy for mastitis treatment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12974-025-03654-6.

Introduction

The dairy industry plays an important role in the global economy and the safety of milk quality is a matter of life and health for consumers. However, mastitis is one of the most common diseases in dairy cows and one of the most detrimental to the development of the dairy industry, causing a huge economic burden to dairy farms [1, 2]. Infection of the mammary gland by pathogenic microorganisms is considered to be the main cause of mastitis. S. aureus is the most common causative agent of mastitis in dairy cows [3], and its pathogenesis has not yet been fully elucidated and the effectiveness of the existing methods of prevention and treatment is limited, which has led to the lack of effective methods of prevention and treatment in clinical practice except for antibiotics, and the misuse of antibiotics can lead to problems such as bacterial resistance and drug residues. Recent studies have shown that gut microbiota dysbiosis exacerbates S. aureus-induced mastitis, manifesting as severe mammary histopathology, enhanced leukocyte infiltration, and impaired immune responses [4–6]. With further research on the gut microbiota and mastitis, it is now hypothesized that the gut-mammary axis is mediated primarily through gut microbiota metabolites, immune cells, gut bacteria, and neural pathways [6]. As the principal neural efferent pathway regulating intestinal homeostasis and immunity, the vagus nerve emerges as a critical modulator of this axis [7–9]. Our prior research revealed that surgical vagotomy disrupts microbial tryptophan metabolism, subsequently inducing mastitis in mice [10]. Complementarily, transcutaneous auricular vagus nerve stimulation (taVNS) in dairy cattle significantly reduces milk somatic cell counts (SCC) [11], a key biomarker of subclinical mastitis. These findings collectively suggest that targeted neuromodulation of vagal pathways may yield novel therapeutic strategies for mastitis management.

The vagus nerve mediates anti-inflammatory responses through multiple neuroimmune pathways. Central to this is the cholinergic anti-inflammatory pathway (CAP), wherein efferent vagal fibers release acetylcholine (ACh) that binds α7 nicotinic acetylcholine receptors (α7nAChR) on immune cells, suppressing pro-inflammatory cytokine production while enhancing anti-inflammatory mediators [12, 13]. This mechanism is exemplified by VNS reducing serum TNF-α by >50%, attenuating tissue damage, and improving survival in sepsis models [14]. Concurrently, VNS activates splenic adrenergic circuitry, inducing norepinephrine release that stimulates β2-adrenergic receptors on choline acetyltransferase-positive (ChAT+) T cells, triggering ACh release to further inhibit macrophage inflammation [15, 16]. Afferent vagal signaling also activates the hypothalamic-pituitary-adrenal (HPA) axis, increasing glucocorticoid secretion to modulate systemic inflammation [17, 18]. While VNS therapeutic applications for inflammatory disorders expand, its role in gut microbiota-mediated immunomodulation remains underexplored. Vagal afferents sense microbial metabolites, regulating intestinal barrier integrity and local immunity, while efferent fibers influence mucosal defense and enteric nervous function collectively shaping microbial ecology [19–21]. However, the precise neuro-microbial-immune circuitry through which vagally modulated gut microbiota influences systemic inflammation remains incompletely characterized.

This study hypothesizes that VNS alleviates S. aureus-induced mastitis in mice by remodeling the gut microbiota, specifically by enriching Muribaculaceae. This enrichment activates PPARγ and inhibits the NF-κB/NLRP3 signaling pathway. Our findings identify the “vagus nerve-Muribaculaceae/PPARγ” axis as a potential therapeutic target for mastitis.

Results

VNS alleviates S. aureus-induced blood-milk barrier disruption and mastitis in mice

Towards elucidating the anti-inflammatory potential of VNS in S. aureus-induced mastitis in mice, a left-sided cervical VNS protocol was implemented, followed by S. aureus perfusion of the fourth pair of mammary glands of mice. Mammary and intestinal tissue were collected 24 h post-infection (Fig. 1A). Successful vagal activation was confirmed by enhanced gastrointestinal motility (Figures S1A and S1B), elevated Ach levels in serum, mammary, and intestinal tissues (Figures S1C–S1E), and increased c-Fos + cells in the nucleus tractus solitarius (NTS) (Figures S1F and S1G). It should be noted that VNS did not induce structural damage (Figures S1H-S1K) or barrier impairment in intestinal tissues (Figures S1L-S1P).

Fig. 1.

VNS alleviatesS. aureus-induced blood-milk barrier disruption and mastitis in mice. A. Experimental Design. A random division of the mice was conducted into four groups: the control group, the VNS group, the S.aureus group and the VNS + S.aureus group (n = 6). Following a 24-hour period, the mice were sacrificed for collection of mammary and intestinal tissue. B. Representative H&E-stained images of mammary tissues from four different groups (scale bar, 50 μm). C. Histopathological score of the mammary gland. D-F. The concentrations of inflammatory cytokines, including TNF-α (D) and IL-1β (E) concentrations and MPO activity (F) in the four groups of mammary tissues were detected using the kit. G. Bacterial load of S.aureus in mammary tissue of four groups of mice. H-K. The expression levels of the TJs proteins ZO-1, Occludin and Claudin-3 in representative mammary tissues as detected by western blotting (H) and the relative intensity of the three proteins using β-actin as an internal reference (I-K) L. Assessment of the histopathological distribution of TJs proteins in the mammary gland was conducted through the utilization of immunohistochemistry, with a scale bar denoting a distance of 20μm. M-N. Serum levels of TNF-α (M) and IL-1β (N) Data are presented as the means ± SD and one-way analysis of variance (ANOVA) was performed for statistical analysis. *p < 0.05, **p < 0.01, *** p < 0.001 and **** p < 0.0001 indicate significant differences. See also Figures S1

The H&E analysis results indicate that the mammary alveolar structure in VNS mice comparable to controls (Fig. 1B). In contrast, the S. aureus group exhibited disrupted alveolar contours with neutrophils infiltration within the alveolar structure (Fig. 1B). The tissue score was significantly elevated in the S. aureus group but was mitigated by VNS intervention (Fig. 1C). This effect can be attributed to the restoration of the intact alveolar structure in the VNS + S. aureus group, as evidenced by a marked reduction in the tissue score (Fig. 1B and C). Notably, inflammatory cytokines TNF-α and IL-1β confirmed that VNS significantly alleviated S. aureus-induced mastitis in mice (Fig. 1D and E). Furthermore, Myeloperoxidase (MPO), an inflammatory marker synthesized and released by immune cells such as neutrophils and macrophages [22], was significantly increased in the S. aureus group but significantly reduced in the VNS + S. aureus group (Fig. 1F). Bacterial loads in mammary tissue were also reduced by VNS (Fig. 1G). It is also worthy of note that systemic inflammation was similarly attenuated, with serum TNF-α and IL-1β levels reduced (Fig. 1M and N). The blood-milk barrier (BMB) plays a crucial role in preventing harmful substances, toxins, and pathogenic bacteria from entering the mammary gland alveoli [23]. Consistent with expectations, the results demonstrated that VNS ameliorated S. aureus-induced impairment of the BMB, as evidenced by the restored expression of tight junction (TJs) proteins ZO-1, Occludin, and Claudin-3 (Fig. 1H and K), which was corroborated by immunohistochemistry (Fig. 1L). Taken together, these results demonstrate that VNS attenuates S. aureus-induced mastitis by limiting inflammatory damage, reducing bacterial burden, and preserving BMB integrity.

FMT from VNS mice alleviates S. aureus-induced mastitis and enhances barrier function in recipient mice

Emerging evidence demonstrates that gut microbiota modulates host immunity and neurophysiological functions through microbial metabolites, with the vagus nerve serving as a key signaling route in the gut–brain axis [7]. Given this established interplay, we investigated whether the gut microbiota mediates the therapeutic effect of VNS in S. aureus-induced mastitis (Fig. 2A). Comparative histopathological analysis revealed profound alveolar destruction and sustained neutrophilic infiltration in Abx + S. aureus versus Abx group (Fig. 2B). Critically, VNS failed to attenuate these pathological changes in microbiota-depleted hosts, with Abx + VNS + S. aureus mice showing comparable tissue scores and neutrophil densities to Abx + S. aureus counterparts (Fig. 2B and C). Consistent with these observations, no substantial disparities were observed in the activities of the mammary inflammatory cytokines TNF-α, IL-1β, and MPO activity in the Abx + VNS + S. aureus group of mice compared to the Abx + S. aureus group (Fig. 2D and F). The results for bacterial load were similar (Fig. 2G). Moreover, it was observed that VNS intervention failed to reduce serum TNF-α and IL-1β levels induced by S. aureus in mice with gut microbiota elimination (Fig. 2H and I). Furthermore, the restorative effect of VNS on the BMB was lost after the microbiota was cleared by detecting the expression of ZO-1, Occludin, and Claudin-3 proteins (Figs. 2J, Figures S2A and S2B). These results indicate that subsequent to the removal of the gut microbiota, VNS no longer exerted its original mitigating effect on mastitis and its restoring effect on the BMB. This suggests that gut microbiota plays a significant role in VNS to alleviate S. aureus-induced mastitis in mice.

Fig. 2.

FMT from VNS mice alleviates S. aureus-induced mastitis and enhances barrier function in recipient mice. A. Experimental Design. Following illustration depicts FMT. FMT was performed on pregnant mice for a period of two weeks, following a five-day pre-treatment with an antibiotic cocktail designed to eliminate the commensal microbiota. B/K. Illustrative images of H&E-stained mammary tissues are presented (scale bar, 50 μm). C/L. Histological score of the mammary gland. D-F/M-O. Concentrations of TNF-α (D/M)and IL-1β (E/N) and MPO activity (F/O) levels in the mammary gland. G/P. Bacterial load of S.aureus in mammary tissue of four groups of mice. H-I/R-S. Serum levels of TNF-α (H/R) and IL-1β (I/S). J/Q. The expression levels of TJs proteins ZO-1, Occludin, and Claudin-3 detected by Western blotting in representative breast tissues Data are presented as the means ± SD and one-way analysis of variance (ANOVA) was performed for statistical analysis. *p < 0.05, **p < 0.01, *** p < 0.001 and **** p < 0.0001 indicate significant differences. See also Figures S2

Following this, the role of gut microbiota was investigated further by means of FMT from donor mice (control and VNS groups) into Abx-treated mice, followed by S. aureus challenge. Both FMT groups exhibited attenuated mammary pathology, with reduced alveolar damage and lower tissue scores compared to the Abx + S. aureus group (Fig. 2K and L). Corroborating this finding, the results of TNF-α and IL-1β, along with MPO activities (Fig. 2M and O), further substantiated that FMT from the control and VNS groups served as a therapeutic intervention, alleviating S. aureus-induced mastitis in recipient mice. Moreover, a decline in the bacterial load of S. aureus (Fig. 2P) and serum TNF-α and IL-1β (Fig. 2R and S) was observed in the Abx + FMT^Control+S.aureus group and the Abx + FMT^VNS+S.aureus group. In addition, the expression of three TJs proteins in the mammary was significantly up-regulated after transplantation of control and VNS group colonies, respectively, in recipient mice (Figs. 2Q, Figures S2C-S2F). Therefore, these results demonstrate that gut microbiota play a protective role in VNS alleviation of S. aureus-induced mastitis in mice.

VNS increases the abundance of g_Muribaculaceae _unclassified in the mice intestine

To further explore the specific effects of VNS on gut microbiota in mice, 16 S rRNA sequencing was performed on feces. The results demonstrated that VNS treatment significantly increased alpha diversity (Shannon and Simpson indices; Figs. 3 A and 3B), though no differences were observed in species richness (Chao1 and observed species; Figs. 3 C and 3D). Venn diagram analysis showed significant differences in the composition of the gut microbiota between the two groups, with a total of 1,149 core species and 1,687 specific species in the control group and 2,329 specific species in the VNS group (Figure S3A). Principal coordinate analysis (PCoA) of operational taxonomic units (OTUs) revealed distinct microbial community structures between VNS and control mice (R = 0.4752, p = 0.002; Fig. 3E). Taxonomic profiling revealed phylum-level remodeling, with a heightened abundance of Bacteroidota observed in the VNS mice, concomitant with a diminished presence of Proteobacteria and Desulfobacterota populations in comparison to the control group (Figure S3B). At the genus level, the most markedly enriched taxon in VNS mice was g_Muribaculaceae_unclassified (Fig. 3G, S3C). To identify discriminative microbial taxa between cohorts, we conducted linear discriminant analysis effect size (LEfSe) using a log10-transformed LDA score threshold of > 3.0. This analysis revealed hierarchically enriched taxa in VNS mice, most notably the g_Muribaculaceae_unclassified, which demonstrated differential abundance across taxonomic hierarchies (Fig. 3F and J). Spearman’s correlation coefficient was utilized to analyze the data, with the result that it was found that g_Muribaculaceae_unclassified abundance was negatively associated with mammary inflammatory markers, while positively correlating with BMB integrity parameters (Fig. 3K). Overall, this finding suggests that g_Muribaculaceae_unclassified may be a potential mediator of VNS to alleviate S. aureus-induced mastitis in mice.

Fig. 3.

VNS increases the abundance of g_Muribaculaceae _unclassified in the mice intestine. A-B. Shannon (A) and Simpson (B) indices are components of the alpha diversity index. C-D. Chao1 index (C) and Obsereved_species (D) are both used to calculate species richness. E. Principal coordinate analysis (PCoA) exhibited a structural separation of the gut microbiota between the control and VNS groups, as determined by unweighted UniFrac distance (R = 0.4752, P = 0.002) F. LEfSe analysis revealed that different bacterial taxa were enriched in various groups (log10 LDA score > 3). G-J. Relative abundances of distinct genera in the gut microbiota of two groups of mice. K. The relationship between significant gut bacterial taxa and inflammatory markers was determined by means of Spearman correlation analysis. The color red is used to denote a positive correlation, whilst blue is used to denote a negative correlation. The strength of the Spearman correlation is inversely correlated with the color’s saturation Data are expressed as box plot. *p < 0.05, **p < 0.01 and *** p < 0.001 by Mann-Whitney U test. See also Figures S3

Replenishment of S24-7 alleviates BMB disruption and mastitis induced by S. aureus in mice

Furthermore, to investigate the role of g_Muribaculaceae_unclassified in VNS alleviation of S. aureus-induced mastitis in mice, a representative bacterium, Muribaculum intestinale S24-7, was orally administered to mice (Fig. 4A). S24-7 administration preserved mammary alveolar architecture (Fig. 4B), reduced histopathological scores (Fig. 4C), and suppressed local levels of TNF-α, IL-1β, and MPO activity (Fig. 4D and F). Similarly, the instillation of S24-7 significantly reduced the bacterial load of S. aureus in mammary tissue (Fig. 4G). S24-7 was shown to enhance ZO-1, Occludin, and Claudin-3 expression in a comparison with pathogen-exposed controls (Fig. 4H and K), which was corroborated by immunohistochemistry (Fig. 4L). Of particular interest is the observation that systemic inflammatory parameters exhibited a parallel to local improvements, as evidenced by reductions in serum TNF-α and IL-1β levels following S24-7 treatment (Fig. 4M and N).

Fig. 4.

Replenishment of S24-7 alleviates BMB disruption and mastitis induced by S. aureus in mice. A. Experimental Design. The mice were randomly divided into four groups: the control group, the S24-7 group, the S.aureus group and the S24-7 + S.aureus groups, and the mice in the S24-7 and S24-7 + S.aureus group were gavaged with 1 × 10¹⁰CFU/mL of bacterial solution daily for 7 days. B. H&E-stained images results showed that S24-7 significantly alleviated S.aureus-induced mammary tissue structural damage in mice (scale bar, 50 μm). C. S24-7 has been demonstrated to reduce scores of mammary tissues significantly. D-F. S24-7 reduced the levels of TNF-α (D), IL-1β (E) and MPO (F) activities induced by S.aureus. G. S24-7 reduces bacterial load of S.aureus in mammary tissues. H-K. S24-7 restores the integrity of the BMB induced by S.aureus (H), as evidenced by the significant increase in the expression of ZO-1 (I), Occludin (J), and Claudin-3 (K) proteins. L. Immunohistochemical results show that S24-7 increases TJs protein expression in the mammary gland (scale bar, 20 μm). M-N. S24-7 reduces serum TNF-α and IL-1β levels increased by S.aureus Data are presented as the means ± SD and one-way analysis of variance (ANOVA) was performed for statistical analysis. *p < 0.05, **p < 0.01, *** p < 0.001 and **** p < 0.0001 indicate significant differences

Recent studies have demonstrated that S24-7 has the capacity to degrade polysaccharides and transfer carbohydrates, as well as to metabolize and produce short-chain fatty acids (SCFAs) and other substances [24, 25]. Such functions are believed to contribute to the regulation of intestinal immune homeostasis and the maintenance of intestinal barrier health [26]. However, the precise mechanisms through which S24-7 exerts its effects remain to be fully elucidated. Building on this, we systematically evaluated the effects of both the S24-7 supernatants (S24-7 Sup) and heat-inactivated S24-7 (HI-S24-7) suspension on S. aureus-induced mastitis in mice, to further elucidate the precise protective mechanism of S24-7. Notably, both treatments partially preserved mammary tissue structure (Fig. 5A and B), mitigated inflammatory responses (Fig. 5C and E), lowered bacterial burden (Fig. 5F), and restored expression of TJs proteins (Fig. 5G and K). A similar phenomenon has been observed in serum inflammatory cytokines (Fig. 5L and M). Taken together, our results suggest that exogenous supplementation of S24-7 significantly attenuates S. aureus-induced BMB destruction and mastitis in mice, and that its protective effects are realized through multiple, synergistic mechanisms.

Fig. 5.

Replenishment of S24-7 alleviates BMB disruption and mastitis induced by S. aureus in mice. A. H&E staining images showed that both the supernatant and the heat-inactivated bacterial solution of S24-7 significantly attenuated S.aureus-induced structural damage to mouse mammary tissue (scale bar, 50 μm). B. Mammary tissue scores were reduced in the S24-7 Sup + S.aureus and HI-S24-7 + S.aureus groups of mice. C-E. TNF-α, IL-1β and MPO activity levels were reduced in groups S24-7 Sup + S.aureus and HI-S24-7 + S.aureus. F. The supernatant and heat-inactivated bacterial solution of S24-7 also reduced the bacterial load of S.aureus in mammary tissue. G-J. Western Blot results showed increased expression of TJs proteins in mammary tissue of mice in the S24-7 Sup + S.aureus and HI-S24-7 + S.aureus groups. K. Immunohistochemical results showed redistribution of ZO-1, Occludin, and Claudin-3 protein expression in mammary tissues of S24-7 Sup + S.aureus and HI-S24-7 + S.aureus groups. L-M. Supernatant and heat-inactivated bacterial solution of S24-7 also reduced S. aureus-induced serum TNF-α and IL-1β levels Data are presented as the means ± SD and one-way analysis of variance (ANOVA) was performed for statistical analysis. *p < 0.05, **p < 0.01, *** p < 0.001 and **** p < 0.0001 indicate significant differences

Replenishment of S24-7 alters S. aureus-induced transcriptome of mammary tissue in mice

To delineate the molecular mechanisms underlying the protective effects of S24-7, RNA sequencing (RNA-seq) was performed on mammary tissues. Utilizing hierarchical clustering and principal component analysis (PCA), it was evident that distinct transcriptional landscapes were exhibited by the various groups, with the S. aureus-infected group demonstrating maximal divergence from the control group (Fig. 6A). It is noteworthy that the S24-7 + S. aureus groups exhibited a tendency to align more closely with the control group, suggesting a partial reversal of infection-driven transcriptional dysregulation (Fig. 6A). The Venn diagram was employed to visualize the distribution of differentially expressed genes (DEGs) across various comparison groups. As illustrated by the Venn diagram, a comparison was made between the Control group and the S. aureus group, revealing the presence of 904 differentially expressed DEGs in the Control group, 907 DEGs unique to the S. aureus group, and 11,634 DEGs expressed in both groups. A comparison was made between the S. aureus group and the S24-7 + S. aureus group. This comparison revealed 694 differentially expressed genes in the S. aureus group and 559 differentially expressed genes in the S24-7 + S. aureus group. The aggregate of these two sets of differentially expressed genes is 11,847 (Fig. 6B). Volcano plots were constructed utilizing a threshold of |log2FC| >1.0 and an adjusted p-value < 0.05, thereby providing a direct representation of the distribution of DEGs. Subsequent analysis revealed that 1,332 genes were found to be up-regulated, whilst 1,893 genes were found to be down-regulated in the S. aureus group in comparison with the control group (Fig. 6E). Furthermore, a comparison of the S24-7 + S. aureus group with the S. aureus group revealed the upregulation of 451 genes and the downregulation of 303 genes in the former (Fig. 6F). Subsequent investigation into the role of differential DEGs was facilitated by means of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses, which were conducted to assess their function. Analysis of the results indicated that, in comparison with the control group, the S. aureus group exhibited a marked increase in leukocyte migration, myeloid leukocyte migration, granulocyte migration, neutrophil migration, and in external encapsulating structure, while there was a significant decrease in extracellular matrix (Figure S4A). Nevertheless, supplementation with S24-7 resulted in a significant down-regulation of immunoglobulin production, immune response molecular mediator production, and up-regulation of cellular matrix structural components and cell adhesion molecule binding (Figure S4B). Subsequent comparison of the differentially expressed genes with the KEGG database revealed a total of eight pathways to be significantly enriched in the S. aureus group, including the NF-κB signaling pathway (Fig. 6C). Furthermore, the results demonstrated that the PPAR signaling pathway was significantly enriched in the S24-7 + S. aureus group (Fig. 6D). Collectively, these results indicate that S24-7 modulates the transcriptomic response to S. aureus infection, potentially through PPAR-mediated mechanisms.

Fig. 6.

Replenishment of S24-7 alters S. aureus-induced transcriptome of mammary tissue in mice. A. Principal component analysis of the transcriptome of three groups of mouse mammary tissue. B. The results of Veen diagrams showed differentially expressed genes between the three groups. C-D. Signaling pathways that were significantly enriched among different groups were revealed by comparison of differentially expressed genes with the KEGG database. E-F. Volcano plot of DEGs between the three groups, blue color indicates genes with significantly lower expression levels, while red color indicates genes with significantly higher expression levels (|log2FC| >1.0, p-value < 0.05p < 0.05)

VNS alleviatesS. aureus-induced mastitis in mice by inhibiting NF-κB/NLRP3 signaling pathway through the gut microbiota S24-7/PPARγ axis

On the basis of the results of the transcriptome analysis, the expression of the PPAR gene was examined in further detail in mouse mammary tissue. The results demonstrated that the gene level of PPARγ was significantly increased by S24-7 (Figs. 7A). Furthermore, it was determined that the gene level of PPARγ underwent significant upregulation in the mammary gland of mice within the VNS group (Fig. 7B). Confirmation of this result at the protein level was provided by Western Blot analysis (Fig. 7C and F). Given the pivotal function of PPARγ in the modulation of inflammation and immune response, as well as the significance of the PPARγ, NF-κB and NLRP3 signaling pathways in the progression of diseases [27]. The protein expression of the relevant pathways in mammary tissues was further examined, and it was demonstrated that both VNS and S24-7 inhibited S. aureus-induced activation of NF-κB and NLRP3 signaling pathways (Fig. 7G and T). Collectively, these results suggest that VNS can alleviate S.aureus-induced mastitis by increasing the abundance of S24-7 bacteria in the intestine and thereby activating PPARγ, which in turn inhibits the activation of NF-κB and NLRP3 signaling pathways.

Fig. 7.

VNS alleviates S. aureus-induced mastitis in mice by inhibiting NF-κB/NLRP3 signaling pathway through the gut microbiota S24-7/PPARγ axis. A-B. Quantitative levels of the PPARγ gene in mammary tissue. C-F. Expression levels and density analysis of PPARγ protein in mammary tissue. G/N. The expression levels of p-p65, p65, p-IκB, IκB, NLRP3, ASC, IL-1β and Caspase-1 were determined by Western Blot (G). H-M/O-T. The relative intensities of p-p65, p-IκB, NLRP3, ASC, Caspase-1, and IL-1β were determined

Materials

Antibiotics including ampicillin (Cat# A5354), neomycin (Cat# N6386), metronidazole (Cat# 16677), vancomycin (Cat# V2002) were sourced from Sigma Aldrich (St. Louis, MO, USA). FITC-dextran (46944) was purchased from Sigma Aldrich. Primary antibodies against the following targets were procured from Affinity Biosciences (OH, USA): ZO-1 (# AF5145), Occludin (# DF7504), Claudin-3 (# AF0129), p-IκB (#AF2002), IκB (#AF6239), p-p65(#AF2006), p65(#5006), PPARγ(#AF6248), ASC(#DF6304), NLRP3(#DF7438), IL-1β(#AF4006), Caspase-1(#AF5418), MUC2(#DF8390), c-Fos(#AF5354) and β-actin (#AF7018). Tumor necrosis factor (TNF)-α (Cat. No. 430915) and interleukin (IL)−1β (Cat. No.432615) enzyme-linked immunosorbent assay (ELISA) kits were purchased from Biolegend (San Diego, California, USA). The myeloperoxidase (MPO) assay kit (A044-1-1) and AB-PAS staining kit (CAS:75881-23-1) were purchased from Nanjing Jiancheng BioEngineering Institute (Nanjing, China).

Animals and experimental design

A cohort of 160 specific pathogen-free (SPF) BALB/c mice (120 females, 40 males; age: 6–8 weeks; weight: 23–25 g) were sourced from Liaoning Changsheng Biotechnology Co. Ltd (Benxi, China). Animals were acclimatized for one week under SPF conditions with a 12-hour light/dark cycle, with ad libitum access to food and water. Mating pairs (3 females:1 male per cage) were established, and males were removed upon confirmation of pregnancy via vaginal plug detection. All experimental procedures strictly adhered to the NIH Guide for the Care and Use of Laboratory Animals and received prior approval from the Institutional Animal Care and Use Committee (IACUC) of Jilin University with approval number 2022-272.

This study employed four distinct experimental approaches: (1) VNS, (2) antibiotic-mediated microbiota depletion, (3) fecal microbiota transplantation (FMT), and (4) oral gavage with S24-7 bacterial supplementation.

For the VNS aim of the study, 24 pregnant BALB/c mice were assigned to four experimental groups: (1) Control: Standard housing for three weeks; (2)S. aureus challenge (S. aureus, 10⁷ CFU):S. aureus perfusion into the fourth mammary gland pair; (3) VNS only: Cervical VNS (10 Hz, 1 mA, 5 min duration) using biphasic pulses (250 μs pulse width, 50 μs inter-phase interval); (4) VNS+S. aureus: Identical VNS parameters administered prophylactically (every 2 h for 3 sessions pre- perfusion) and therapeutically (every 12 h post- perfusion).

In the antibiotic treatment experiments, 18 pregnant mice were divided into the Abx group: a mixture of 4 antibiotics treated by gavage for 5 days (200 mg/kg ampicillin, metronidazole, neomycin and 100 mg/kg vancomycin), the Abx+S. aureus group: after 5 days of Abx pretreatment, the fourth pair of mammary glands were perfused with S. aureus, and the Abx+VNS+ S. aureus group: after 5 days of Abx pretreatment, the mammary glands were subjected to VNS 2 hours prior to the perfusion of the mammary gland by S. aureus. The mice were then stimulated at 12-hour intervals and executed 24 hours later. A total of 24 mice were used in the FMT experiments, which were divided into four groups: the Abx group, the Abx+S. aureus group, the Abx+FMT^Control+S. aureus group, and the Abx+FMT^VNS+S. aureus group. The specific methods of FMT were the same as before. Following antibiotic treatment, the recipient mice were replaced antibiotics with water for one day, and then incubated with prepared fecal microbial suspension from cows in the VNS or Control groups intragastric ally daily for three consecutive days followed by one time every two days for 26 days. The Abx mice were received with equivoluminal sterile PBS. The specific methods of FMT were the same as before. In brief, fresh faeces were harvested from control and VNS animals, pooled within group, and immediately transferred into an anaerobic workstation. After dilution to 0.1 mg/mL in ice-cold sterile PBS, the slurry was cleared at 100 ×g for 5 min; the bacteria-rich supernatant was used as the inoculum.

Fifty-four pregnant mice were involved in the S24-7 supplementation experiments, which were divided into control, S. aureus group, S24-7+S. aureus group, S24-7 Sup+ S. aureus group and the HI-S24-7+S. aureus group. Mice were gavaged with S24-7 for one week at a dose of 1x 10¹⁰CFU/mL.

Bacterial growth condition and preparation of heat-inactivated and supernatant forms

Muribaculum intestinaleS24-7 (DSM 28989) was procured from the Leibniz Institute DSMZ-German Collection of Microorganisms and Cell Cultures. Cultures were maintained in pre-reduced General Anaerobic Medium (GAM; HaiBo Biologicals, Qingdao, China) under strict anaerobic conditions (5% H₂, 85% N₂, 10% CO₂) at 37 °C until stationary phase (approximately 7 days). S. aureus SA113 (ATCC35556) was obtained from Mikrobielle Genetik, University of Tubingen, Germany. S. aureus was incubated in TSB medium at 37 °C and 120 rpm overnight, and the incubation was stopped when the concentration was approximately 1×10⁷CFU/mL. Bacterial cultures were harvested during the logarithmic growth phase under anaerobic conditions. The cultures were centrifuged at 12,000 × g for 10 min at 4 °C, and the resulting supernatant was filtered through a 0.22 μm membrane to ensure sterility. For the preparation of heat-inactivated S24-7, bacterial pellets collected at the logarithmic phase were obtained by centrifugation at 12,000 × g for 15 min and washed twice with sterile PBS. The bacterial suspension was subsequently inactivated by heating at 70 °C for 30 min. To confirm the absence of viable bacteria, the heat-inactivated preparations were plated on blood agar medium (TSA supplemented with 5% defibrinated sheep blood) and incubated anaerobically for at least one week. Both the heat-inactivated bacteria and the sterile supernatant were aliquoted and stored at –80°C until use. To maintain anaerobic conditions during intragastric administration, all solutions and pre-chilled gavage needles were thawed and handled inside an anaerobic chamber. Each gavage procedure was completed within 30 seconds to minimize oxygen exposure.

Vagus nerve stimulation

Mice underwent overnight food and water deprivation prior to surgery. Mice were anesthetized using isoflurane and secured on a thermostatically controlled surgical platform to maintain normothermia throughout the procedure. The surgical site was prepared as follows: neck hair was removed, the skin was sterilized and incised, and the left cervical vagus nerve was isolated. A bipolar electrode was carefully positioned around the exposed vagus nerve. Strict aseptic technique was maintained; all instruments were sterilized prior to use. Mice were continuously maintained at normothermia postoperatively. Following full recovery from anesthesia, electrical stimulation was initiated. Stimulation electrodes were connected via a 0.2 mm output wire to a programmable electrical stimulator. Stimulation parameters were: 1 mA current intensity, 250 μs biphasic pulses, 50 μs interphase delay, 10 Hz frequency, applied for 5 minutes per session[28]. Control animals underwent identical surgical procedures, including electrode placement, but did not receive electrical stimulation.

Measurement of intestinal permeability

Following a 4-hour fasting period (with ad libitum access to water), mice received an oral gavage of 4 kDa fluorescein isothiocyanate-dextran (FITC-dextran; 200 mg/kg). After 4 hours, blood samples were collected via retro-orbital puncture under brief isoflurane anesthesia. Whole blood was centrifuged at 3,000×g for 10 min at 4 °C to obtain serum. Plasma FITC-dextran concentrations were quantified by measuring fluorescence intensity (excitation: 495 nm, emission: 525 nm) using a spectrofluorometer, with values expressed relative to a standard curve.

Gastrointestinal (GI) Transit

To measure small intestinal transit rates in mice, we conducted charcoal propulsion assays[29]. Briefly, mice were fasted for 16 hours (with ad libitum access to water) and subsequently administered 0.25 mL of an activated charcoal suspension orally. After 60 min, mice were euthanized, and the entire small intestine—from the pylorus to the ileocecal junction—was aseptically excised. The intestine was placed on blotting paper and gently straightened without stretching. The linear distance traveled by the charcoal front and the total intestinal length were measured. The small intestinal transit rate for each mouse was calculated as: small intestinal transit (%) = (distance traveled by charcoal/total intestinal length) × 100.

16S rRNA and metagenomic sequencing

Genomic DNA was isolated from control and VNS group fecal samples using the E.Z.N.A.® Stool DNA Kit (Omega Bio-tek, Switzerland). DNA concentration and integrity were assessed by 1% agarose gel electrophoresis. Purified DNA was diluted to 1 ng/μL using sterile water, with final concentrations determined spectrophotometrically for subsequent 16S rRNA gene and metagenomic sequencing. The V3-V4 hypervariable regions of the bacterial 16S rRNA gene were amplified using universal primers 515 F and 806R. Resulting amplicons were size-selected on 2% agarose gels, purified (AxyPrep PCR Cleanup Kit, Axygen Biosciences, USA), and quantified fluorometrically (Quant-iT PicoGreen dsDNA Assay Kit). Sequencing libraries were generated from amplicons. Libraries exceeding 2 nM were normalized, pooled proportionally, and denatured to single strands using NaOH. Paired-end sequencing (2×250 bp) was performed on an Illumina NovaSeq platform. Paired-end reads were merged into longer tags using FLASH software based on sequence overlap. Barcode and primer sequences were subsequently trimmed. Chimeric sequences were identified and removed with Vsearch. High-quality sequences (clean tags) were clustered into operational taxonomic units (OTUs) at 97% sequence similarity using Vsearch. Alpha and beta diversity analyses were conducted in QIIME. OTU representative sequences were taxonomically classified by alignment (BLAST) against the RDP and NCBI-16S rRNA databases.

Transcriptome analysis of the mammary gland

Total RNA was isolated from murine mammary tissues representing control, S. aureus, and S24-7+S. aureus groups using TRIzol reagent (Thermo Fisher Scientific, 15596018) following manufacturer protocols. RNA concentration and purity measurements were conducted on a NanoDrop ND-1000 spectrophotometer (NanoDrop), with integrity verified via Bioanalyzer 2100 (Agilent). Samples meeting stringent quality thresholds (>50 ng/μL concentration, RIN >7.0, total RNA >1 μg) were advanced to downstream procedures. Polyadenylated RNA underwent dual-round enrichment using oligo(dT) magnetic beads (Dynabeads, Thermo Fisher Scientific, 25–61005). Purified mRNA was fragmented (94°C, 5–7 min) employing the NEBNext® Magnesium RNA Fragmentation Module (NEB, E6150S). First-strand cDNA synthesis utilized SuperScript™ II Reverse Transcriptase (Invitrogen, 1896649) with fragmented RNA as template. Second-strand synthesis incorporated Escherichia coli DNA polymerase I (NEB, M0209), RNase H (NEB, M0297), and dUTP (Thermo Fisher Scientific, R0133). Following end-repair and 3'-adenylation, fragments (~300 ± 50 bp) were size-selected via magnetic bead purification. Strand-specific libraries were generated through UDG enzyme digestion (NEB, M0280) of uracil-containing strands. Library amplification involved: 95 °C for 3 min; 8 cycles of 98°C/15s, 60°C/15s, 72°C/30s; final extension at 72 °C for 5 min. Final libraries underwent paired-end sequencing (PE150) on the Illumina NovaSeq™ 6000 platform (LC Bio).Raw sequencing data underwent quality filtering to yield clean reads. Hisat2 (v2.1.1) aligned clean reads to the reference genome. Transcript assembly and merging employed StringTie (v2.1.6). Subsequent analyses included: Gene expression quantification, Differential expression analysis (DESeq2), Gene Set Enrichment Analysis (GSEA), Functional enrichment evaluation. Quality metrics were assessed throughout using FastQC (v0.10.1) and RSeQC (v4.0.0).

Histological analysis

Intestinal and mammary tissues collected from all experimental groups underwent fixation in 4% paraformaldehyde (48 hours) followed by standard paraffin embedding. Serial sections (4–5 μm) were prepared from these blocks. Sections were subjected to routine hematoxylin and eosin (H&E) staining. Mammary histopathology was scored according to established criteria[30], evaluating three key parameters: (1) acinar structural integrity, (2) inflammatory cell infiltration density, and (3) presence of hyperemia/edema. The scoring system applied was: Score 0: Intact acinar architecture; no inflammation, hyperemia, or edema. Score 1: Mild acinar disruption; focal inflammatory infiltrates; mild vascular changes. Score 2: Moderate acinar destruction; moderate diffuse inflammation. Score 3: Severe acinar disintegration; dense inflammatory infiltrates; marked hyperemia/edema. Intestinal sections were evaluated based on modified criteria[31], focusing on three primary features: (1) Degree of inflammatory cell infiltration within the lamina propria and submucosa, (2) Alterations in epithelial morphology (e.g., hyperplasia, ulceration), and (3) Changes in mucosal architecture (e.g., villus blunting, crypt distortion).

Cytokines assay

Mammary gland tissue samples (30 mg) were weighed and immediately placed in ice-cold phosphate-buffered saline (PBS). A 1:9 (w/v) tissue-to-buffer homogenate was prepared using a mechanical tissue homogenizer (CBCL-24, Shanghai Cebo Biological Technology Development Center, China) under constant cooling conditions. Following homogenization, the lysates were centrifuged at 12,000 × g for 10 min at 4°C. The resulting supernatant was carefully collected, aliquoted, and stored at −80°C pending analysis. Concentrations of pro-inflammatory cytokines TNF-α and IL-1β in the clarified supernatants were subsequently quantified using specific immunoassays.

MPO activity

Mammary tissue was processed according to kit manufacturer guidelines. A 10% (w/v) tissue homogenate in ice-cold PBS was prepared using a mechanical homogenizer. MPO enzymatic activity in mammary homogenates was quantified spectrophotometrically using a commercial MPO detection kit.

Mammary bacterial burden assay

Mammary gland specimens were aseptically collected from all experimental cohorts, immediately weighed, and homogenized (10% w/v) in PBS. Aliquots (50 μL) of homogenized tissue suspension were plated onto Mannitol Salt Agar. After 24 h incubation at 37 °C, characteristic yellow colonies presumptive for S. aureus were enumerated. Bacterial loads were calculated as colony forming units (CFU) per gram of tissue weight. All procedures employed sterile consumables and instruments validated by autoclaving.

Western blot

Total protein was isolated from mammary tissue using a Tissue Protein Extraction Kit (Thermo Fisher Scientific) supplemented with protease/phosphatase inhibitors. Protein concentrations were determined via BCA assay (Thermo Fisher Scientific). Equal protein aliquots were resolved on 10% or 12% SDS-polyacrylamide gels and electro transferred to PVDF membranes. Membranes were blocked with 5% non-fat dry milk in TBST for 3 h at room temperature, followed by overnight incubation at 4 °C with primary antibodies (1:1000 dilution in blocking buffer). After three 20-min TBST washes, membranes were probed with species-matched HRP-conjugated secondary antibodies (goat anti-rabbit or goat anti-mouse, 1:20,000) for 2 h at room temperature. Immunoreactive bands were visualized using an ECL substrate system. Densitometric analysis of band intensities was performed with ImageJ software (NIH), with target protein expression normalized to loading controls.

Goblet assessment by AB-PAS staining

Intestinal morphology and goblet cell distribution were evaluated using AB-PAS staining (Solarbio, China) according to established methodology. Paraffin-embedded sections underwent sequential processing: dewaxing in xylene, rehydration through graded ethanol, and distilled water rinsing (2 min). Sections were incubated in 1% AB for 15 min, followed by three 2-min aqueous washes. Tissue sections were oxidized in 0.5% periodic acid (10 min), rinsed twice with distilled water, then treated with Schiff's reagent (15 min) and washed extensively (10 min). Nuclei were counterstained with Mayer's hematoxylin (2 min), differentiated in acid-alcohol, and blued in Scott's solution (3 min) with a final 3-min rinse. After dehydration via ethanol series and xylene clearing, sections were permanently mounted with synthetic resin. Blinded histopathological assessment was performed using brightfield microscopy (Olympus BX53, Japan) with quantitative analysis of goblet cell density per 100 μm intestinal epithelium.

Immunohistochemistry

Mammary tissue sections were sequentially processed through xylene dewaxing and graded ethanol dehydration. Antigen retrieval was performed using heated sodium citrate buffer (pH 6.0). Endogenous peroxidase activity was quenched by incubating sections with peroxidase blocking reagent (SAP IHC Kit, MXB Biotechnologies) for 40 min at room temperature (RT).After three 5 min PBS washes, non-specific binding sites were blocked with normal goat serum (SAP Kit) for 40 min at RT. Sections were then incubated overnight at 4 °C with primary antibodies diluted in antibody diluent. Following PBS washes, sections were treated with HRP-conjugated goat anti-rabbit IgG secondary antibody (1:200) for 30 min at RT, then incubated with HRP polymer (SAP Kit) for 20 min. After additional PBS washes, chromogenic development was performed using DAB substrate (SAP Kit) for precisely 3 min under microscopic monitoring. The reaction was terminated by immersion in distilled water when optimal signal-to-background ratio was achieved. Sections were counterstained with Mayer's hematoxylin for 5 min, differentiated in 1% acid-alcohol, and blued in 0.02% ammonium hydroxide. Finally, sections were dehydrated through graded ethanol, cleared in xylene, and mounted with permanent neutral resin medium.

Immunofluorescence

To assess c-Fos expression, mice were transcranial perfused with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) 2 hours post VNS. Brains were extracted, post-fixed overnight in 4% PFA at 4 °C, and cryoprotected in 30% sucrose solution until saturated. Coronal sections (20 μm) encompassing the caudal nucleus tractus solitarius (NTS; anterior-posterior: −7.24 to −7.79 mm relative to bregma) were prepared using a cryostat. Sections were then re-fixed in 4% PFA for 30 min, washed in PBS, and blocked with 5% bovine serum albumin (BSA) in PBS for 1 hour. Subsequently, sections were incubated overnight at 4 °C with anti-c-Fos primary antibody. Following PBS washes, sections were incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG secondary antibody (1:500) for 1 hour at room temperature. Nuclei were counterstained with DAPI (0.5 μg/mL) for 10 min. After final PBS washes, immunofluorescence was visualized using a confocal laser scanning microscope.

RNA extraction and quantitative reverse transcription PCR (qRT-PCR)

Total RNA was isolated from freshly harvested tissue using TRIzol™ reagent (Invitrogen, Carlsbad, CA, USA), with RNA integrity verified electrophoretic ally and purity confirmed spectrophotometrically (A260/A280 >1.8). First-strand cDNA synthesis was performed with 1 μg total RNA using TransStart Tip Green qPCR SuperMix (TransGen Biotech, Beijing, China) under manufacturer-specified thermocycling conditions. Quantitative PCR amplification was conducted in triplicate 20-μL reactions containing FastStart Universal SYBR Green Master Mix (ROX) (Roche, Basel, Switzerland) using a StepOnePlus™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Gene-specific primer sequences (Table S1) were designed using NCBI Primer-BLAST with amplicon lengths between 80–150 bp. Thermal cycling parameters followed standard three-step amplification: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 sec, 60 °C for 30 sec, and 72 °C for 30 sec. GAPDH served as the endogenous normalization control. Relative transcript quantification employed the 2^-ΔΔCt method, with expression levels normalized to the control group calibrator sample. Melt curve analysis confirmed amplification specificity.

Discussion

The dairy industry faces severe economic challenges due to mastitis, with S. aureus being particularly problematic due to its high prevalence, tissue invasiveness, and association with antimicrobial resistance [1, 32]. Critically, emerging research indicates that gut microbiota significantly influences mammary gland health via systemic immune modulation, a relationship formally conceptualized as the ‘gut-mammary axis‘ [33]. Central to this axis, the vagus nerve, serving as a critical neural pathway connecting the brain and gut, functions as a principal modulator in this regulatory cascade [34]. TaVNS in cows with subclinical mastitis significantly reduced milk SCC [11], which is a key indicator of mammary inflammation. However, the mechanisms by which VNS mediates the protective effect against mastitis and its effect on the microbiota remain to be elucidated. Afferent vagus nerve fibers (80%) transmit intestinal inflammatory signals to the NTS, activating the HPA axis to promote glucocorticoid-mediated systemic immunosuppression [12, 18, 35]. Conversely, efferent fibers (20%) release ACh that binds α7nAChR on immune cells, inhibiting inflammatory cytokine release [35–37]. As a therapeutic neuromodulation technique, VNS demonstrates clinical efficacy in refractory epilepsy, depression, and neurodegenerative disorders [38, 39]. Similarly, taVNS reduces milk SCC in bovine subclinical mastitis independent of bacteriological culture outcomes [11]. To investigate the mechanisms of VNS in alleviating S. aureus-induced mastitis in mice, we established a left cervical VNS model [28]. Neurophysiological validation confirmed efferent vagal activation through increased intestinal propulsion rates [40, 41], while c-Fos immunoreactivity in the NTS indicated afferent fiber engagement [42, 43]. Critically, histopathological assessment and inflammatory cytokine profiling verified that cervical VNS surgery did not induce intestinal inflammation or compromise barrier integrity. Furthermore, VNS attenuated S. aureus-induced mastitis pathology and restored TJs protein expression, enhancing mammary epithelial barrier function. Our study introduces a novel neuro-microbiological axis in mastitis pathophysiology, demonstrating that VNS confers protection against S. aureus-induced mastitis by restructuring the gut microbiota, specifically through enrichment of Muribaculaceae. A key finding is the significant enrichment of g_Muribaculaceae_unclassified in VNS-treated mice. Correlation analyses positioned this bacterium as a potential mediator, linking its abundance to improved mammary health. Functional validation via exogenous administration of the representative strain Muribaculum intestinale S24-7 confirmed its capacity to attenuate mastitis and restore BMB integrity. Mechanistically, we identified that the protective effects of S24-7 are mediated through the activation of the PPARγ signaling pathway, leading to subsequent suppression of the pro-inflammatory NF-κB/NLRP3 axis. This work provides the first experimental evidence for a functional “vagus nerve-Muribaculaceae-PPARγ” axis in mammary gland immunology, offering new insights into the gut-mammary axis and presenting potential alternatives to conventional antibiotic therapies. It is noteworthy that the present study proposes the vagus nerve as a potential therapeutic target for mastitis. However, there are still multiple barriers to its safe and effective translation into routine clinical treatment strategies. There are no accepted guidelines for the optimization of VNS parameters, and the lack of standardized treatment protocols further limits clinical application [44]. Nevertheless, further exploration into the neuroimmune interactions underlying mastitis may yet yield novel insights into the development of effective therapeutic strategies.

The abolition of VNS-mediated protection in microbiota-depleted mice establishes the gut microbiome as a crucial intermediary in this neuro-immune pathway. This finding aligns with growing recognition of microbiota-dependent neuroimmunomodulation across various physiological systems. However, the interpretation of antibiotic ablation experiments warrants careful consideration. While antibiotic administration effectively eliminates the gut microbiota, it may concurrently alter host immune tone through direct effects on immune cell function or by inducing transient metabolic shifts. Studies have documented that certain antibiotics can modulate inflammatory responses independently of their antimicrobial effects, potentially through interactions with mitochondrial function or direct immunomodulatory properties [45–47]. Although our data demonstrate that VNS requires an intact microbiota, we cannot entirely exclude potential confounding effects of antibiotic-induced immunomodulation on mastitis susceptibility. FMT experiments further substantiate the causal role of gut microbes in VNS-mediated protection. FMT from both control and VNS groups attenuated S. aureus-induced mastitis and restored mammary barrier integrity, though VNS-FMT conferred superior protection. Through 16 S rRNA sequencing, we identified g_Muribaculaceae_unclassified as the most significantly enriched taxon in VNS-treated mice. Spearman correlation analysis demonstrated inverse associations between Muribaculaceae abundance and mammary inflammatory markers, and positive correlations with mammary barrier integrity indicators. These findings position Muribaculaceae as a primary mechanistic mediator of VNS bioactivity in mastitis resolution. This aligns with established mechanisms: VNS enhances intestinal mucin secretion and TJs assembly, creating a probiotic-permissive niche. For instance, in post-stroke models, VNS reverses blood-brain barrier damage and corrects stroke-induced dysbiosis—specifically restoring Firmicutes abundance while reducing Bacteroidetes [48]. Vagal signaling further modulates Lactobacillus colonization via Brunner’s gland regulation [49]. Our research adds to this growing body of evidence by specifically implicating Muribaculaceae as a vagally-modulated taxon relevant to mammary gland health. The mechanisms through which VNS enriches for Muribaculaceae likely involve VNS-induced modifications to the intestinal microenvironment. This is supported by prior findings demonstrating that Muribaculaceae exhibit high sensitivity to luminal osmolality, with growth inhibited under conditions of elevated salt, sugar alcohol, or polyethylene glycol concentrations [49]. VNS enhances intestinal mucin secretion and TJs proteins assembly, potentially creating a more favorable niche for mucin-degrading specialists like Muribaculaceae [50, 51]. Additionally, VNS may alter intestinal nutrient availability through effects on gastrointestinal motility or by influencing intestinal immune cell activity through vagal sensory neuron signaling [48], thereby indirectly shaping microbial community structure. Furthermore, VNS may modulate the gut microbiota through several potential mechanisms: (1) altering intestinal nutrient availability, (2) impacting intestinal mucosal barrier function, or (3) via vagus nerve sensory neuron signaling to influence intestinal immune cell activity, thereby regulating microbial communities [52].

Muribaculaceae, a genus within the phylum Bacteroidetes, predominantly colonizes the intestinal tract and contributes to host health maintenance. Previous studies indicate Muribaculaceae encodes enzymes for O-glucan degradation and sialic acid inactivation, enabling the catabolism of complex polysaccharides into SCFAs [24, 53]. These metabolites regulate intestinal barrier integrity and immune responses, establishing Muribaculaceae as a candidate next-generation probiotic [54–56]. Based on the increased abundance of g_Muribaculaceae_unclassified in VNS-treated mice, we selected its representative strain Muribaculum intestinale S24-7 for mechanistic investigation. Notably, S24-7 gavage ameliorated S. aureus-induced mastitis in mice. To delineate the active components and mechanisms, we administered S24-7 supernatant and heat-inactivated preparations to mice. Both interventions partially attenuated mastitis, albeit less effectively than viable S24-7, suggesting a multi-component mechanism of action involving both structural components of the bacterial cell and secreted metabolites. The partial efficacy of heat-inactivated S24-7 aligns with established literature on par probiotics, wherein non-viable microbial cells or their components exert immunomodulatory effects through interactions with pattern recognition receptors [57]. Similarly, the protective effects of S24-7 supernatant indicate the involvement of soluble factors, potentially including SCFAs or other microbial metabolites known to influence host immunity [24, 58]. That neither intervention fully recapitulated the efficacy of viable S24-7 suggests possible synergistic interactions between structural components and bacterial metabolites, or the potential importance of ongoing bacterial metabolism and host-microbe crosstalk for optimal therapeutic effects.

Transcriptomic analysis of mammary tissue provided crucial insight into the molecular mechanisms underlying S24-7-mediated protection. While S. aureus perfusion robustly activated pro-inflammatory pathways, S24-7 supplementation specifically enhanced PPAR signaling pathway expression. PPAR, nuclear receptor superfamily transcription factors, comprise three primary isoforms (PPARα, PPARβ/δ, PPARγ) [59, 60]. Beyond their established roles in lipid metabolism, PPARs modulate inflammatory responses by inhibiting pro-inflammatory signaling, regulating mediator production, and influencing immune cell function [59]. Our findings that both S24-7 and VNS treatment significantly upregulated PPARγ expression in mammary tissue, and that PPARγ activation subsequently suppressed the NF-κB/NLRP3 signaling axis. This mechanism aligns with previous research demonstrating that VNS can upregulate PPARγ expression through α7nAChR activation in cerebral ischemia/reperfusion models [61, 62]. Our study extends this paradigm by revealing an additional, microbiota-dependent pathway for vagally-mediated PPARγ upregulation. Specifically, we demonstrate that VNS can increase PPARγ expression indirectly through enrichment of S24-7 in the intestinal tract. This dual-pathway model—comprising both direct neural signaling and indirect microbiota-mediated effects—significantly enhances our understanding of how neuromodulation can exert systemic anti-inflammatory effects.

In conclusion, our work identifies a novel “vagus nerve-Muribaculaceae-PPARγ” axis that protects against S. aureus-induced mastitis. The primary novelty of our findings lies in: (1) establishing the gut microbiota as an essential mediator of VNS effects in mastitis; (2) identifying Muribaculaceae as a key taxon linking VNS to mammary gland protection; and (3) elucidating PPARγ activation with subsequent suppression of NF-κB/NLRP3 signaling as the downstream molecular mechanism. Our study elucidates a vagally mediated gut-mammary axis and suggests that therapeutic strategies targeting VNS or augmenting Muribaculaceae abundance may prove effective against mastitis.

Limitations of the study

An additional limitation of our study concerns the bidirectional nature of the gut-mammary axis. While we focused on how gut-derived signals influence mammary health, we did not investigate whether S. aureus mastitis conversely affects intestinal homeostasis. Previous research has demonstrated that peripheral inflammation can alter gut microbiota composition and barrier function, suggesting the possibility of bidirectional communication along the gut-mammary axis. Future studies examining intestinal responses to mammary infection would provide a more comprehensive understanding of this physiological network. Future research should focus on several key areas: First, the translational potential of VNS should be evaluated in ruminant models, with careful optimization of stimulation parameters for agricultural applications. Second, the specific microbial metabolites and structural components responsible for S24-7 protective effects warrant identification and characterization. Third, the potential bidirectional communication along the gut-mammary axis merits investigation, particularly regarding how mammary inflammation affects gut homeostasis. Finally, the interactions between direct neural signaling and indirect microbiota-mediated effects in VNS-induced protection require further dissection to fully understand the integrated neuro-microbial immune network governing mammary gland health.

Authors’ contributions

Yuhong He: Conceptualization, Data curation, Investigation, Formal Analysis, Writing - Original DraftYue Zhang: Conceptualization, Funding acquisition, Supervision, Writing - Review & EditingLihua Zhao: Conceptualization, Data curation, Investigation, Project administrationZeming Zhou: Data curation, Formal analysis, Methodology, Project administrationNier Su: Software, Validation, Visualization, Writing – review & editingCan Zhang: Methodology, Project administration, ResourcesKeyi Wang: Methodology, VisualizationLei Jin: Methodology, Project administrationXiaoyu Hu: Funding acquisition, Supervision, Writing - Review & EditingYunhe Fu: Funding acquisition, Conceptualization, Conceptualization, Resources, Writing - Review & Editing.

Funding

The study is supported by the National Natural Science Foundation of China (32422086), National Natural Science Foundation of China (32301247) and Jilin Province Department of Science and Technology (212558JC010286222).

Data availability

The authors confirm that all data included in the study and its supplementary information are available in this article and that the relevant 16 S rRNA macro genome sequencing data have been deposited in the NCBI Sequence Read Archive (SRA) repository under accession number PRJNAxxxxxx.

Declarations

Ethics approval and consent to participate

The Institutional Animal Care and Use Committee (IACUC) of Jilin University approved all experiments involving animals. The IACUC ethics committee reviewed the full proposal and granted the animal care and use permit license. We followed the US National Institutes of Health’s manual of animal laboratory care and use during all experiments.

Competing interests

The authors declare no competing interests.