Distinct effects of different Bacteroides strains on depressive-like behavior via a gut-Th1/Th17 cells-brain axis

Abstract

Extensive evidence links gut microbiota to the pathogenesis of major depressive disorder (MDD), yet the specific microbial species involved remain unclear. Here, we identify distinct roles of three Bacteroides species—B. uniformis, B. vulgatus, and B. thetaiotaomicron—in depression. B. uniformis increases susceptibility to depression in mice, significantly enhances Th17 cell differentiation in vivo and in vitro, and upregulates hippocampal IL-17A level. However, treatment with SR1001, a Th17 cell inhibitor, alleviates B. uniformis-induced depressive-like behaviors. Conversely, B. thetaiotaomicron and B. vulgatus attenuate depressive behaviors in mice, significantly suppresse the differentiation of Th1 and Th17 cells in vivo, and reduce the levels of hippocampal cytokines, including IL-17A, IFN-γ, and TNF-α. Clinical analyses reveal increased Th1 and Th17 cells in MDD patients, correlating with depression severity. B. uniformis is enriched in MDD fecal samples and positively associated with Th17 levels, whereas B. thetaiotaomicron showes an inverse correlation. Mechanistically, targeted metabolomic shows that B. uniformis reduces butyric acid and cholesterol sulfate, whereas B. thetaiotaomicron increases butyric acid, propionic acid, and biotin, all of which are linked to Th1 and Th17 regulation. These findings highlight the role of Bacteroides species in depression via a gut-Th1/Th17 cells-brain axis, providing mechanistic insights and ideas for therapeutic strategies.

A study provides insights into how distinct Bacteroides strains oppositely regulate depressive-like behavior through a gut–Th1/Th17–brain immune axis, revealing microbiota-driven mechanisms underlying depression.

Introduction

Major depressive disorder (MDD) is a prevalent mental health condition affecting approximately 300 million individuals worldwide, characterized by symptoms such as persistent low mood, diminished motivation, and anhedonia^1,2. In 2018, the World Health Organization (WHO) ranked MDD as the third leading cause of global disease burden, with projections indicating it could rise to the top position by 2030³. However, the etiology and pathogenesis of MDD remain complex and poorly understood, presenting significant challenges to the development of effective diagnostic and therapeutic strategies. Therefore, it is essential to investigate novel perspectives to deepen our understanding of the underlying mechanisms of MDD. Recent studies have highlighted the critical role of gut microbiota in depression⁴, yet the specific key bacterial strains and their mechanistic contributions remain elusive.

Emerging evidence has revealed that the gut microbiota may contribute to the development of depression through the “gut-brain” axis, involving neurotransmitters, metabolites, endocrine, immune, and neuroactive pathways⁵. Although findings on gut microbiota alterations in depression have been inconsistent, Bacteroides species have demonstrated consistent variations across multiple studies^6–12. In our previous research, metagenomic analysis of fecal samples from MDD patients and healthy controls (HCs) also revealed significant differences in the abundance of several Bacteroides species¹³. Among these, B. uniformis (BU), B. vulgatus (BV), and B. thetaiotaomicron (BT) have been implicated in psychiatric disorders. For instance, in the context of depression, B. uniformis has been reported to increase the serotonin precursor 5-hydroxytryptophan (5-HTP) in the brain, thereby mitigating depressive-like behaviors induced by chronic unpredictable mild stress (CUMS) in mice¹⁴. Conversely, other studies have indicated that B. uniformis reduces serotonin (5-HT) levels in the hippocampus and induces depressive -like behaviors in mice¹⁵. Thus, the role of B. uniformis in depression remains inconclusive and warrants further investigation. Additionally, B. thetaiotaomicron has been shown to alleviate anxiety-like, repetitive, and communicative behaviors in autism spectrum disorder (ASD) mouse models¹⁶; however, its potential role in depression has not yet been explored. Furthermore, B. vulgatus and its metabolite, 4-hydroxyphenylacetic acid (4-HPAA), have been reported to alleviate depressive-like behaviors in mice with colitis¹⁷, suggesting a potential role for B. vulgatus in mitigating depressive symptoms.

The mechanisms by which Bacteroides species influence depression remain poorly understood. However, extensive research has demonstrated that Bacteroides species modulate disease pathogenesis through the regulation of T helper (Th) 1 and Th17 cell differentiation. For example, a consortium of nine bacterial strains, including B. uniformis, was shown to promote Th1 cell differentiation in germ-free mice¹⁷. Similarly, B. thetaiotaomicron alleviated colitis by suppressing Th1 and Th17 cell responses¹⁸, whereas B. vulgatus reduced Th1 polarization, thereby conferring protection against Escherichia coli-induced colitis¹⁹. Moreover, Th1 and Th17 cells have been strongly implicated in the pathogenesis of depression. For example, Th17 cells were significantly elevated in the hippocampus and prefrontal cortex of depressed mice, and their transfer into the hippocampus has been shown to induce depressive-like behaviors^20,21. Recent studies have further reported an increased proportion of peripheral Th17 cells and elevated serum IL-17A levels in patients with MDD^22–25. Additionally, Th1 cells were elevated in the brains of mice subjected to learned helplessness^20,21. Notably, knockout mice deficient in the Th1-specific transcription factor T-bet failed to exhibit depressive-like behaviors²⁶. These findings collectively suggest that Th1 and Th17 cells may play a promotive role in depressive-like behaviors. Based on this evidence, we aim to investigate whether B. uniformis, B. vulgatus, and B. thetaiotaomicron contribute to the development or alleviation of depression by regulating Th1 and Th17 cell differentiation.

In this study, we first investigated the regulatory effects of B. uniformis, B. vulgatus, and B. thetaiotaomicron on Th1 and Th17 cell differentiation in vitro. Based on these findings, we employed a 7-day subthreshold chronic restraint stress (CRS) mouse model and a 14-day CRS-induced depression mouse model to examine the roles of these Bacteroides species in depression, as well as their regulatory effects on Th1 and Th17 cell differentiation and associated cytokine levels in the hippocampus. Finally, we validated the variations in Bacteroides abundance, alterations in Th cell populations, and their correlations using fecal and peripheral blood samples from HCs and MDD patients. Collectively, this study aimed to elucidate the roles of B. uniformis, B. vulgatus, and B. thetaiotaomicron in depression and their involvement in the “gut-Th1/Th17 cells-brain” axis mechanisms.

Results

B. uniformis, B. vulgatus, and B. thetaiotaomicron regulated Th1 and Th17 cell differentiation in vitro

Previous studies have shown that Bacteroides species are involved in modulating Th1 and Th17 cell differentiation^18,19,27, and abnormalities in Th1 and Th17 cells have been identified in MDD patients^28–30. Accordingly, the modulatory effects of B. uniformis, B. vulgatus, and B. thetaiotaomicron on Th1 and Th17 cell differentiation were first examined using a Bacteroides-DC-T cell co-culture system. The results revealed that B. uniformis (Fig. 1A and Supplementary Fig. 1A, D) significantly promoted Th1 (vs. no infection, MOI = 1 and 5, P < 0.05) and Th17 (vs. no infection, MOI = 10, P < 0.05) cell differentiation in a dose-dependent manner. Conversely, B. thetaiotaomicron (Fig. 1C and Supplementary Fig. 1C) markedly suppressed Th1 (vs. no infection, MOI = 1, 5 and 10, P < 0.05) and Th17 (vs. no infection, MOI = 5, P < 0.05) cell differentiation in a concentration-dependent manner. However, B. vulgatus (Fig. 1B and Supplementary Fig. 1B) exhibited no significant effect on Th1 and Th17 cell differentiation (vs. no infection, P > 0.05). These results highlight the distinct regulatory roles of B. uniformis and B. thetaiotaomicron in Th1 and Th17 cells differentiation, with B. vulgatus showing minimal influence in vitro.

Fig. 1. Modulation of Th1 and Th17 cell differentiation by B. uniformis, B. vulgatus, and B. thetaiotaomicron in vitro.

DCs were infected with heat-killed B. uniformis, B. vulgatus, and B. thetaiotaomicron at 0, 1, 5, or 10 MOI or were treated with LPS (1 µg/mL) for 24 h. The DCs were then cocultured with naïve CD4⁺ T cells isolated from the spleen for 3 days in the presence of anti-CD3ε monoclonal antibodies. Flow cytometry results showing the regulatory effects on Th1 and Th17 cell differentiation by (A) B. uniformis, (B) B. vulgatus, and (C) B. thetaiotaomicron. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 6). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: P > 0.05. DCs dendritic cells, MOI multiplicity of infection, LPS lipopolysaccharide, from Escherichia coli 055: B5, as a positive control.

B. uniformis increased depression susceptibility

Evidence has indicated that Th1 and Th17 cells may contribute to depressive-like behaviors^20,21,31. The in vitro results demonstrated that B. uniformis promoted the differentiation of Th1 and Th17 cells. Based on these findings, we hypothesized that B. uniformis might enhance susceptibility to depression. Given that B. vulgatus exhibited no significant effect on Th1 and Th17 cell differentiation in vitro, a 7-day subthreshold CRS model was employed to assess whether B. uniformis and B. vulgatus influence depression susceptibility.

Pseudo-germ-free mice were established by administering antibiotics for 10 days¹⁵. Subsequently, mice were administered B. uniformis or B. vulgatus and exposed to a 7-day subthreshold CRS (Fig. 2A and Supplementary Fig. 3A). Compared to the ABX group, the ABX + CRS⁻⁷ group showed no significant behavioral changes (P > 0.05, Fig. 2B–G), while the ABX + CRS⁻⁷ + BU group significantly exhibited reduced center activity (time, distance, and entries) in the open-field test (OFT, P < 0.05, Fig. 2B–D), decreased sucrose preference in the sucrose preference test (SPT, P < 0.01, Fig. 2E), prolonged immobility in the forced swim test (FST, P < 0.001, Fig. 2F), and delayed feeding latency in the novelty-suppressed feeding test (NSFT, P < 0.001, Fig. 2G). Notably, administration of heat-killed B. uniformis also induced similar depressive-like phenotypes, including reduced center activity (time and distance) in the OFT, decreased sucrose preference, and prolonged immobility in the FST (P < 0.05, Supplementary Fig. 2A–G). Conversely, B. vulgatus treatment group demonstrated improved behavioral outcomes relative to the ABX + CRS^-7 group (Supplementary Fig. 3B–G). These findings suggest that B. uniformis increases susceptibility to depression, whereas B. vulgatus may alleviate stress-induced depressive behaviors following 7-day CRS.

Fig. 2. B. uniformis increased susceptibility to depression.

A Schematic illustration of the experimental design, including antibiotic treatment, bacterial gavage, and CRS procedures. ABX-treated male mice were orally administered PBS or live B. uniformis (10⁸ CFU/mL, 0.2 mL per mouse) every other day for 7 days, followed by 7 days of CRS (3–4 h daily). Behavioral tests were conducted 24 h later, including (B–D) time, distance, and number entries in center (%) in the OFT, E sucrose preference index in the SPT, F immobility time in the FST and G feeding latency in the NSFT. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 10). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns: P >0.05. ABX, antibiotic cocktail; CRS⁻⁷, chronic restraint stress for⁷ days; BU Bacteroides uniformis, OFT open-field test, SPT sucrose preference test, FST forced swim test, NSFT novelty-suppressed feeding test.

B. vulgatus and B. thetaiotaomicron alleviated depression

B. thetaiotaomicron was demonstrated to inhibit the differentiation of Th1 and Th17 cells in vitro, which are implicated in depression³⁰. Thus, we hypothesized that B. thetaiotaomicron might alleviate depressive-like behaviors. To investigate the role of B. thetaiotaomicron in depression and further confirm the depression-alleviating effects of B. vulgatus, we utilized a 14-day CRS model, followed by antibiotic treatment and oral administration of B. thetaiotaomicron or B. vulgatus (Fig. 3A).

Fig. 3. B. thetaiotaomicron and B. vulgatus ameliorated depressive-like behaviors in mice.

A Schematic illustration of the experimental design, including CRS, antibiotic treatment, and bacterial gavage procedures. Male mice were subjected to CRS of 3–4 h daily for 14 days, followed by 10 days of ABX treatment and a 3-day washout period. Mice were orally administered PBS or live B. thetaiotaomicron and B. vulgatus (10⁸ CFU/mL, 0.2 mL) every other day for 7 days. Behavioral assessments were conducted 24 h later, including (B–D) time, distance and number entries in center (%) in the OFT, (E) sucrose preference index in the SPT, (F) immobility time in the FST, and (G) feeding latency in the NSFT. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 10). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns: P > 0.05. ABX, antibiotic cocktail; CRS⁻¹⁴, chronic restraint stress for days; BT B. thetaiotaomicron, BV B. vulgatus, OFT open-field test, SPT sucrose preference test, FST forced swim test, NSFT novelty-suppressed feeding test.

Behavioral tests demonstrated that the ABX + CRS⁻¹⁴ group showed significant depressive-like behaviors compared to the ABX group (P < 0.05, Fig. 3B–G). However, treatment with B. thetaiotaomicron or B. vulgatus significantly alleviated these behaviors, including increased center time and distance (P < 0.05, Fig. 3B, C), enhanced sucrose preference (P < 0.05, Fig. 3E), and reduced immobility and feeding latency (P < 0.05, Fig. 3F, G). Moreover, administration of heat-killed B. thetaiotaomicron also attenuated depressive-like behaviors, including increased time in the OFT, enhanced sucrose preference, and reduced immobility in the FST (P < 0.05, Supplementary Fig. 4A–G). These findings suggest that B. thetaiotaomicron and B. vulgatus may ameliorate CRS-induced depressive-like behaviors.

B. uniformis, B. vulgatus, and B. thetaiotaomicron modulated Th1 and Th17 cell differentiation in depressed mice

To investigate the mechanisms of the “gut-Th1/Th17 cells-brain” axis in depression mediated by the three Bacteroides strains, the spleens, blood, and Peyer’s patches were collected for flow cytometry analysis. The results revealed that B. uniformis significantly promoted Th17 differentiation in the spleen (vs. ABX + CRS⁻⁷, P < 0.05), blood (vs. ABX, P < 0.05; vs. ABX + CRS⁻⁷, P < 0.05), and Peyer’s patches (vs. ABX + CRS⁻⁷, P < 0.05; vs. ABX, P < 0.05, Fig. 4A–C and Supplementary Fig. 5A–C and 6A–C), but had no significant effect on Th1 cell differentiation (P > 0.05, Fig. 4A–C and Supplementary Fig. 6A–C). B. thetaiotaomicron and B. vulgatus significantly suppressed Th17 differentiation in the spleen (vs. ABX + CRS⁻¹⁴, P < 0.05, Fig. 4D and Supplementary Fig. 7A) and blood (vs. ABX + CRS⁻¹⁴, P < 0.05, Fig. 4E and Supplementary Fig. 7B), as well as Th1 differentiation in Peyer’s patches (vs. ABX + CRS⁻¹⁴, P < 0.05, Fig. 4F and Supplementary Fig. 7C). These findings demonstrate that B. uniformis, B. vulgatus, and B. thetaiotaomicron regulate Th1 and Th17 cell differentiation in depressed mice.

Fig. 4. B. uniformis, B. vulgatus, and B. thetaiotaomicron modulated Th1 and Th17 cell differentiation in depressed mice.

The ratio of Th1 and Th17 cells in (A, D) spleen, (B, E) blood and (C, F) Peyer’s patch of all groups of mice were determined by flow cytometry. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 6 ~ 7). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: P > 0.05. ABX, antibiotic cocktail; CRS⁻⁷, chronic restraint stress for 7 days; BU Bacteroides uniformis, CRS^-14 chronic restraint stress for 14 days, BT B. thetaiotaomicron, BV B. vulgatus.

B. uniformis, B. vulgatus, and B. thetaiotaomicron modulated the levels of Th1/Th17-associated cytokines in the hippocampus of depressed mice

Cytokines have been reported to play a critical role in depression through the “gut-brain” axis³². To assess alterations in Th1/Th17-associated cytokine levels in the hippocampus, cytokine arrays were employed for each group. Compared to the ABX + CRS⁻⁷ group, B. uniformis significantly increased hippocampal levels of IL-12p70 and IL-13 in depressed mice (P < 0.05, Fig. 5A). Relative to the ABX group, B. uniformis also significantly elevated the levels of IL-12p70, IL-13, IL-17A, IL-21, and IFN-γ in the hippocampus (P < 0.05, Fig. 5A). Conversely, compared to the ABX + CRS^-14 group, B. thetaiotaomicron significantly reduced the levels of IL-12p70, IL-17A, IFN-γ, TGF-β1, and TNF-α in the hippocampus (P < 0.05, Fig. 5B). Similarly, compared to the ABX + CRS^-14 group, B. vulgatus significantly decreased the levels of IL-5, IL-12p70, IL-17A, IL-17F, IL-28, TGF-β1, and TNF-α in the hippocampus (P < 0.05, Fig. 5B).

Fig. 5. B. uniformis, B. vulgatus, and B. thetaiotaomicron influenced Th1/Th17- associated cytokine levels in the hippocampus of depressed mice.

The concentrations of IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12p70, IL-13, IL-17A, IL-17F, IL-21, IL-22, IL-23, IL-28, IFN-γ, MIP-3a, TGF-β1, and TNF-α (pg/ml) in the hippocampus of mice from each group (A, B) were measured using a cytokine array kit. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 6). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns: P > 0.05. ABX, antibiotic cocktail; CRS⁻⁷, chronic restraint stress for days; CRS⁻¹⁴ chronic restraint stress for days, BU Bacteroides uniformis, BT B. thetaiotaomicron, BV B. vulgatus.

Th17 cells mediated the pro-depressive effects of B. uniformis

The above findings indicated that B. uniformis increased depression susceptibility by promoting Th17 cell differentiation in the spleen, blood, and Peyer’s patches, while also elevating IL-17A levels in the hippocampus of depressed mice. To determine whether Th17 cells mediate the pro-depressive effects of B. uniformis, we employed the Th17 differentiation inhibitor SR1001 (Fig. 6A). The findings demonstrated that the B. uniformis-treated group exhibited depressive-like behaviors relative to the ABX + CRS⁻⁷ group (P < 0.05, Fig. 6B–G). However, administration of BU + SR1001 significantly alleviated depressive-like behaviors compared to the ABX + CRS⁻⁷ + BU group, including increased center time, distance, and the number of entries into the center in OFT (P < 0.05, Fig. 6B–D), enhanced sucrose preference (P < 0.05, Fig. 6E), and reduced immobility (P < 0.05, Fig. 6F). Furthermore, flow cytometry results showed that SR1001 significantly inhibited the differentiation of Th17 cells in the spleen, blood, and Peyer’s patches of mice (vs. ABX + CRS⁻⁷ + BU, P < 0.05, Supplementary Fig. 8A–C).

Fig. 6. The pro-depressive effects of B. uniformis were mediated by Th17 cells.

A Schematic representation of ABX treatment, bacterial gavage, SR1001and CRS protocols. ABX-treated male mice received either PBS or live B. uniformis (10⁸ CFU/mL, 0.2 mL) via gavage every other day for 7 days, followed by 7 days of CRS (3–4 h/day). To inhibit Th17 cell differentiation, mice were administered SR1001 (25 mg/kg, i.p., every other day) for 2 weeks following a 3-day washout. Behavioral assessments were conducted 24 h later, including measurements of (B–D) time spent, distance traveled, and number of entries into the center (%) in the OFT, E sucrose preference in the SPT, F immobility time in the FST, and G feeding latency in the NSFT. Statistical analysis was performed using one-way ANOVA (post hoc analysis: Tukey’s multiple comparison test, n = 10). Data indicate the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns:P > 0.05. ABX antibiotic cocktail, CRS⁻⁷ chronic restraint stress for 7 days, BU Bacteroides uniformis, OFT open-field test, SPT sucrose preference test, FST forced swim test, NSFT novelty-suppressed feeding test.

Comparison of Th cell subsets and Bacteroides abundances in HCs and MDD groups

To further elucidate the role of Th cells in depression, we performed flow cytometry analysis on PBMCs from HCs and MDD patients. There were no significant differences in age, sex, and BMI between the HCs and MDD patients (P > 0.05; details of all individuals are presented in Supplementary Table 1). MDD patients demonstrated a significant increase in the proportions of Th1 (P < 0.01) and Th17 (P < 0.01) cells and a significant reduction in regulatory T (Treg) Cell (P < 0.001) cell proportions compared to HCs (Supplementary Fig. 9A). Moreover, the percentages of Th17 cell were positively correlated with HAMD scores which reflected severity of depression (r = 0.9127, P < 0.0001, Supplementary Fig. 9B). Additionally, metagenomic sequencing was performed on fecal samples to assess alterations in Bacteroides species. B. uniformis levels in feces were significantly higher in MDD patients compared to HCs (P < 0.05, Supplementary Fig. 9C), positively correlating with both depression severity (r = 0.7409, P < 0.05, and Supplementary Fig. 9D) and peripheral Th17 cell percentages (r = 0.7016, P < 0.01, and Supplementary Fig. 9E). Conversely, B. thetaiotaomicron exhibited a decreasing trend (P > 0.05, Supplementary Fig. 9B), negatively correlated with depression severity (r = −0.7199, P < 0.01, Supplementary Fig. 9D) and Th17 cell percentages (r = −0.6800, P < 0.05, Supplementary Fig. 9E). B. vulgatus showed no significant changes or correlations (P > 0.05, Supplementary Fig. 9C–E). Furthermore, peripheral Th1 cell percentages did not exhibit significant correlations with B. uniformis, B. vulgatus, or B. thetaiotaomicron levels (P > 0.05, Supplementary Fig. 9F).

Significant metabolic alterations induced by B. uniformis and B. thetaiotaomicron supplementation

In addition to the in vitro and in vivo findings, clinical data also revealed that B. uniformis and B. thetaiotaomicron were significantly correlated with peripheral Th17 cell percentages. To further elucidate the underlying mechanisms by which B. uniformis and B. thetaiotaomicron modulate Th17 cells, high-throughput targeted quantitative metabolomic profiling of fecal samples was performed in mice. Compared to the ABX + CRS⁻⁷ group, supplementation with either live or heat-killed B. uniformis led to significant metabolic alterations (Supplementary Fig. 10A), with increased levels of propionylcarnitine, leucyl−glycine, pantothenic acid, symmetric dimethylarginine, ethanolamine, heptadecenoylcarnitine, pentadecanoic acid, 3beta−hydroxy−5−cholenoic acid, tridecanoic acid, isobutyric acid, ethylmethylacetic acid, hydroxyisovaleric acid, valeric acid, phenylacetic acid and iodide (P < 0.001), and decreased levels of malonylcarnitine, butyric acid, cholesterol sulfate and melatonine (P < 0.001, Supplementary Fig. 10B). Similarly, compared to the ABX + CRS^-14 group, supplementation with either live or heat-killed B. thetaiotaomicron significantly altered fecal metabolites (Supplementary Fig. 11A), leading to significant increases in the levels of o−Sulfotyrosine, gamma−aminobutyric acid, n−acetyltyrosine, propionylcarnitine, n−acetylphenylalanine, n−acetylleucine, leucyl−glycine, asparagine, pentadecanoic acid, omega−muricholic acid, 3−epideoxycholic acid, glycocholic acid 3−sulfate, butyric acid, propionic acid, ethylmethylacetic acid, isobutyric acid, valeric acid, d−galacturonic acid and biotin (P < 0.001), together with a significant decrease in phenylethylamine levels (P < 0.001, Supplementary Fig. 11B).

Discussion

Multiple studies have identified significant alterations in the Bacteroides species in depression^6,7,9–11. Based on these findings, our study further investigated the roles of different Bacteroides strains in depression. The results demonstrated that B. uniformis significantly increased susceptibility to depression in mice, whereas B. thetaiotaomicron and B. vulgatus alleviated depressive-like behaviors. Further investigation revealed B. uniformis promoted Th17 cell differentiation and elevated Th17-related cytokines (e.g., IL-17A, IL-21) in the hippocampus. In contrast, B. thetaiotaomicron and B. vulgatus suppressed Th1 and Th17 cell differentiation and reduce related cytokines (e.g., IL-17A, TNF-α, IFN-γ) in the hippocampus. Moreover, targeted metabolomics showed that B. uniformis suppressed metabolites like butyric acid and cholesterol sulfate, while B. thetaiotaomicron elevated butyric acid, propionic acid, and biotin, all associated with Th1 and Th17 regulation. These findings suggest that Bacteroides strains may influence depression through the “gut-Th1/Th17 cells-brain” axis, offering insights into the pathogenesis of depression (Fig. 7).

Fig. 7. The roles of B. uniformis, B. vulgatus, and B. thetaiotaomicron in depression and the underlying “gut-Th1/Th17 cells-brain axis” mechanism.

B. uniformis promoted Th17 cell differentiation in the small intestine, spleen, and blood, leading to elevated IL-17A levels and other pro-inflammatory cytokines in the brain. This inflammatory response may further impair the BBB and contribute to neuronal damage, ultimately increasing susceptibility to depression. CRS-induced proliferation of Th1 cells in the small intestine may impair BBB integrity by releasing TNF-α and IFN-γ into the peripheral circulation and brain. Together with IL-17A and other inflammatory factors, this process facilitated Th17 cell migration into the brain, exacerbating depressive pathology. B. vulgatus and B. thetaiotaomicron inhibited Th1 cell differentiation in the small intestine and Th17 cell differentiation in peripheral blood and the spleen, thereby reducing levels of TNF-α, IFN-γ, and IL-17A in both the periphery and the brain. This suppression may mitigate BBB damage, exert neuroprotective effects, and ultimately alleviate depressive symptoms. Additionally, both B. uniformis and B. thetaiotaomicron significantly altered fecal metabolite profiles, suggesting that these changes may affect Th1 and Th17 differentiation (By Figdraw).

Bacteroides are major colonizers of the gut microbiota. These Gram-negative obligate anaerobes not only dominate in abundance but also play multiple roles in human health and disease³³. For instance, studies have demonstrated that B.uniformis can regulate metabolism, inflammation, and immune responses, contributing to the potential treatment of colitis^34–36, obesity^37,38, steatohepatitis^39,40, and pancreatic injury⁴¹. However, extracellular vesicles derived from B. uniformis may exacerbate gut inflammation⁴², and one study suggested that it may be involved in colorectal tumorigenesis induced by high-soluble fiber intake⁴³. In psychiatric disorders, certain B. uniformis strains could alleviate ASD-like (B. uniformis D2-69)⁴⁴, anxiety-like⁴⁵, and depressive-like behaviors (B. uniformis CECT 7771)¹⁴, while B. uniformis ATCC 8492 has been found to reduce hippocampal 5-HT levels and induce depressive-like behaviors¹⁵. Our study further revealed that this strain increased depression susceptibility. Additionally, the levels of B. uniformis were elevated in the feces of MDD patients and positively correlated with MDD severity. Consistent with our results, previous studies have also reported an increased abundance of B. uniformis in patients with MDD^8,12. These findings suggest that different B. uniformis strains may exert varying effects on depression, highlighting the need for further strain-specific research.

B. thetaiotaomicron has been reported to alleviate colitis^18,46,47, allergic airway inflammation⁴⁸, and fatty liver disease^49,50, and B. vulgatus similarly exhibits anti-inflammatory properties, improving colitis^17,51, atherosclerosis⁵², and osteoporosis⁵³. In psychiatric disorders, B. thetaiotaomicron significantly improved anxiety, repetitive behaviors, and communication deficits in ASD models¹⁶. Similarly, B. vulgatus was reported to alleviate morphine addiction⁵⁴ and depression-like behavior in a DSS-induced colitis model¹⁷. These findings suggest that B. thetaiotaomicron and B. vulgatus may function as probiotics with anti-inflammatory, immunoregulatory, and psychiatric protective effects. Our study also demonstrated that B. thetaiotaomicron and B. vulgatus ameliorate depression-like behaviors in mice. Additionally, the abundance of B. thetaiotaomicron negatively correlated with depression severity in MDD patients, supporting its potential protective role in depression.

Recent studies have identified B. uniformis, B. thetaiotaomicron, and B. vulgatus as key regulators of Th1 and Th17 cell differentiation, which are closely linked to immune and inflammatory disorders^17–19. Given the critical role of Th1 and Th17 cells in MDD, investigating their involvement along the “gut–Th1/Th17 cells–brain” axis is of significant importance. In our study, B. uniformis significantly increased susceptibility to depression in mice and promoted Th17 differentiation in the spleen, blood, and Peyer’s patches. Inhibition of Th17 differentiation via SR1001 abolished B. uniformis-induced depressive-like behaviors. Clinical metagenomic analysis further revealed a higher abundance of B. uniformis in MDD patients, which positively correlated with peripheral Th17 cell levels. Indeed, Th17 cells have been implicated in depression through the “gut–brain” axis. Adoptive transfer of Th17 cells into wild-type or Rag2^⁻/⁻ mice (lacking T cells) induced depressive-like behaviors, with the transferred Th17 cells infiltrating the hippocampus^20,21. Our clinical data demonstrated a significant increase in peripheral Th17 cells in MDD patients, correlating with depression severity. These findings suggest that the elevation of peripheral Th17 cells in mice following B. uniformis treatment and their potential accumulation in the brain may represent a key mechanism underlying depression. However, further investigation is required to determine whether Th17 cells infiltrate the brain.

Regarding the neurobiological mechanisms of depression, our results showed that B. uniformis treatment elevated hippocampal IL-17A and IL-21 levels. IL-21 plays a pivotal role in Th17 cell differentiation⁵⁵, while IL-17A, a signature pro-inflammatory mediator secreted by Th17 cells, induced depressive-like behaviors in mice⁵⁶. Clinical studies have further demonstrated that elevated IL-17A levels in patients with psoriasis were associated with an increased risk of depression, while IL-17A neutralization has been shown to alleviate depressive symptoms in approximately 40% of cases^57,58. Therefore, in addition to Th17 cells, IL-17A may also play a crucial role in B. uniformis-induced depression. Studies have shown that peripheral Th17 cells migrating into the brain secrete IL-17A, leading to microglial activation, exacerbation of oxidative and nitrosative stress, and the release of pro-inflammatory cytokines. This inflammatory cascade facilitates the intracerebral differentiation of naïve CD4⁺ T cells into Th17 cells, amplifying neuroinflammation, inducing neuronal damage, and increasing depression susceptibility^28,59. Thus, B. uniformis may increase depression susceptibility by promoting Th17 cell differentiation in the small intestine, spleen, and blood, leading to elevated IL-17A levels in the brain, further exacerbating neuroinflammation and neuronal damage. However, whether the increased IL-17A in the brain originated from the direct migration of Th17 cells from the periphery or formed in situ IL-17A secretion by infiltrating Th17 cells remained to be further elucidated.

Studies have demonstrated that Th17 cells were recruited to the central nervous system (CNS) during early CRS exposure⁶⁰, and IL-17 has been shown to disrupt the blood-brain barrier (BBB), facilitating the infiltration of immune cells and cytokines into the CNS, thereby exacerbating depressive symptoms^61,62. Consistently, our study found that after 14 days of CRS, mice exhibited depressive behaviors, accompanied by an increased presence of Th17 cells in the spleen and blood, along with elevated IL-17A and IL-17F levels in the hippocampus. Furthermore, in a learned helplessness depression mouse model, both Th17 and Th1 cells were elevated in the brain; however, only the adoptive transfer of Th17 cells promoted the onset of depression, whereas Th1 cell transfer had no such effect²¹. Nevertheless, accumulating evidence have suggested that Th1 cells may also contribute to depression. For instance, chronic stress has been reported to activate Th1 cells⁶³, and polymorphisms in the Th1-associated gene T-bet have been linked to MDD²⁶. In our study, a significant increase in Th1 cells has been observed in the peripheral blood of MDD patients compared to HCs. Therefore, these findings suggest that while Th1 cells may contribute to depression, their effect likely depends on the presence of additional factors, as their action alone appears insufficient to induce depressive symptoms²¹.

Our study revealed increased small intestinal Th1 cells and elevated hippocampal IL-28, IFN-γ, and TNF-α in CRS-induced depressive mice, while no significant changes were observed in Th1 cells within the blood and spleen. Research has shown that peripheral TNF-α could damage the BBB and infiltrate the hippocampus⁶⁴, disrupt mitochondrial function, and reduce dendritic spine density, contributing to depressive-like behavior⁶⁵. In addition, IFN-γ has been implicated in cytokine-induced depression^66,67 and was known to increase BBB permeability, promoting CD4⁺ T cell migration into the CNS. This effect was significantly amplified in the presence of TNF-α⁶⁸. Clinically, TNF-α and IFN-γ were associated with depressive symptom severity and emotional-cognitive dysfunction⁶⁹. Additionally, chronic stress has been reported to drive peripheral cytokines into circulation, disrupting BBB integrity and triggering neuroinflammation, all of which contribute to depressive-like behaviors⁷⁰. These findings suggest that CRS-induced expansion of small intestinal Th1 cells may contribute to BBB disruption by secreting TNF-α and IFN-γ into the peripheral circulation and brain, in combination with IL-17A and other inflammatory mediators, thereby facilitating Th17 cell infiltration into the brain and promoting depressive pathology. Interestingly, our study found that treatment with B. thetaiotaomicron and B. vulgatus alleviated depressive behaviors, potentially by inhibiting Th1 differentiation in the small intestine and Th17 differentiation in the spleen and peripheral blood, while reducing TNF-α, IFN-γ, and IL-17A levels in the hippocampus. Furthermore, in MDD patients, the relative abundance of B. thetaiotaomicron was significantly negatively correlated with the proportion of Th17 cells in peripheral blood.

Interestingly, our in vitro experiments showed that even heat-killed B. uniformis and B. thetaiotaomicron could modulate Th1 and Th17 cell differentiation. Consistently, behavioral assays revealed that heat-killed B. uniformis could increase depressive susceptibility, whereas heat-killed B. thetaiotaomicron could alleviate depressive-like behaviors. Metabolomic analysis further demonstrated that both live and heat-killed forms of these two strains significantly altered fecal metabolites. Specifically, both live and heat-killed B. uniformis markedly reduced the levels of butyric acid, while live and heat-killed B. thetaiotaomicron significantly increased its abundance. Notably, butyric acid has been reported to suppress Th1 and Th17 cell differentiation^71–75. In addition, B. uniformis significantly decreased cholesterol sulfate, which was known to suppress Th17 cell differentiation⁷⁶, whereas B. thetaiotaomicron increased propionic acid and biotin levels, which also suppress Th1 and Th17 differentiation^77,78. Collectively, these findings suggest that the differential effects of B. uniformis and B. thetaiotaomicron on Th1 and Th17 are likely mediated through both structural components—such as LPS, lipoproteins, and capsular polysaccharides—and their distinct influences on host metabolic pathways^79,80.

Our study has several limitations: First, the mechanisms by which Bacteroides strains regulate Th1 and Th17 cell differentiation—specifically how bacterial structural components and metabolites participate in this process—remain to be elucidated. Further investigation is required to elucidate the underlying mechanisms. Second, it remains unclear whether the regulatory effects of B. thetaiotaomicron and B. vulgatus on depressive-like behaviors are mediated through Th1 and Th17 cells. Additionally, the precise mechanisms by which Th1 and Th17 cells influence brain function and contribute to depression pathophysiology are not fully understood, necessitating further investigation. Third, the limited clinical sample size necessitates further studies with larger cohorts to enhance the robustness and clinical applicability of our findings. Finally, all mice experiments were conducted in antibiotic-treated pseudo–germ-free mice, a commonly used model to deplete intestinal bacteria; however, potential interactions with residual microbes cannot be completely excluded.

In conclusion, B. uniformis was found to increase susceptibility to depression, whereas B. thetaiotaomicron and B. vulgatus exhibited protective effects. These effects may be mediated by bacterial structural components and metabolites through the regulation of Th1/Th17 differentiation and hippocampal inflammation. Our findings provide novel insights into the pathogenesis of depression and suggest potential microbial-targeted therapeutic strategies.

Methods

Bacterial strain and culture conditions

Bacteroides strains, including B. uniformis ATCC8492, B. thetaiotaomicron ATCC29148, and B. vulgatus ATCC8482, were obtained from Beina Biology (Beijing, China). The strains were cultured anaerobically on Columbia blood agar plates (Cat. No. CP0160, HuanKai Microbial) for 24–48 h. After incubation, bacteria were harvested and resuspended in sterile phosphate-buffered saline (PBS) at 10⁸ CFU/mL for oral administration. Each strain was administered to mice (0.2 mL per mouse) within one hour of preparation. Bacterial concentrations were confirmed by OD600 readings and viable count analysis.

Preparation of murine bone-marrow-derived dendritic cells (BMDCs)

Bone marrow-derived dendritic cells (BMDCs) were prepared following a modified protocol based on Lutz et al.⁸¹. Bone marrow was collected from the femurs and tibiae of female C57BL/6 J mice and flushed with RPMI-1640 medium (Cat. No. C11875500BT, Gibco) containing 10% heat-inactivated fetal bovine serum (FBS, heat-inactivated at 56 °C for 30 min, Cat. No. 10099141 C, Gibco), 50 mM 2-mercaptoethanol (Cat. No. 21985023, Gibco), 10 mM HEPES (Cat. No. 15630080, Gibco), and 100× penicillin–streptomycin mix (Cat. No. FG101-01, TransGene Biotech). The marrow suspension was gently filtered through a 40-μm strainer to remove debris and treated with RBC lysis buffer (Cat. No. R1010, Solarbio) to eliminate red blood cells. The bone marrow cells were seeded at 2 × 10⁶ cells/ml in Petri dishes (Cat. No. 430591, Corning) with 10 ml of growth medium containing 10 ng/ml granulocyte-macrophage colony-stimulating factor (GM-CSF, 10 ng/ml, Cat. No. 415-ML-020, R&D Systems). Cultures were incubated at 37 °C with 5% CO₂. On day 3, 10 ml of fresh GM-CSF-supplemented medium was added. Medium changes were performed on days 6 and 8 using a half-depletion method. On day 10, BMDCs were purified using mouse CD11c MicroBeads (Cat. No. 130-125-835, Miltenyi Biotec) according to the manufacturer’s protocol. Purified BMDCs were plated in 24-well plates and treated with B. uniformis, B. thetaiotaomicron, or B. vulgatus (heat-killed at 65 °C for 30 min, at MOI = 1, 5, or 10) or LPS (1 μg/ml, Cat. No. L2880, Sigma) for 24 h. After exposure, BMDCs were washed three times with phosphate-buffered saline (PBS).

Isolation of naive CD4+ T cells from spleen

Spleens from female C57BL/6 mice were collected and placed in cold PBS containing 10% FBS. Under sterile conditions, the spleens were minced and passed through a 70-μm cell strainer using a syringe plunger to obtain a single-cell suspension. The suspension was washed twice with PBS, centrifuged at 300 g for 5 min, and the supernatant was discarded. Red blood cells were lysed using RBC lysis buffer (Cat. No. R1010, Solarbio), followed by another centrifugation step at 300 g for 5 min, after which the supernatant was removed. The cell pellet was resuspended in PBS and stored on ice for magnetic bead-based separation. Naive CD4⁺ T cells were isolated using the Naive CD4⁺ T Cell Isolation Kit (Cat. No. 130-104-453, Miltenyi Biotec) according to the manufacturer’s instructions. These purified initial T cells were then cocultured with infected or uninfected BMDCs at a 5:1 ratio for 72 h in the presence of anti-CD3ε (Cat. No. 553057, BD Biosciences) in 24-well plates.

Flow cytometry analysis

After 3 days of co-culture, DCs and T cells were stimulated for 6 h at 37 °C and 5% CO₂ with a leukocyte activation cocktail (Cat. No. 550583, BD Biosciences). The cells were then harvested, stained with Fixable Viability Stain 780 (FVS780, 0.2 mg/ml, Cat. No. 565388, BD Biosciences), anti-CD3 (0.2 mg/ml, Cat. No. 552774, BD Biosciences), and anti-CD4 (0.2 mg/ml, Cat. No. 563727, BD Biosciences), and subsequently fixed and permeabilized. Intracellular cytokines were stained in Perm buffer using anti-IFN-γ (0.2 mg/ml, Cat. No. 557724, BD Biosciences) and anti-IL-17A (0.2 mg/ml, Cat. No. 563354, BD Biosciences). Flow cytometry analysis was performed on a BD LSRFortessa system, and results were processed using FlowJo software.

Antibiotic cocktails

After 1-week adaptive feeding, the male C57BL/6 J mice were fed with antibiotic cocktails (ABX, 0.5 g/L ampicillin, Cat. No. A9518; 0.5 g/L metronidazole, Cat. No. M1547; 0.5 g/L neomycin, Cat. No. N6386, sigma; 0.25 g/L vancomycin, Cat. No. 222-01303, wako) in drinking water for 10 days. The oral administration was performed after a wash-out period for 3 days.

Chronic restraint stress (CRS)

In the subthreshold CRS model, ABX-treated male mice were orally administered with PBS, live, or heat-killed B. uniformis and B. vulgatus (10⁸ CFU/mL in sterile PBS, 0.2 mL per mouse) every other day for 7 days. Mice were then subjected to 7 days of CRS, placed in restrainers (Cat. No. GEGD-FM25, Globalebio) for 3–4 h daily. Behavioral tests were conducted 24 h post-CRS (Fig. 2A and Supplementary Fig. 2A). In addition, mice were injected intraperitoneally with SR1001 (25 mg/kg, Cat. No. 1335106-03-0, MedChemExpress), an inhibitor of Th17 cell differentiation and function, every other day for 2 weeks following a 3-day washout period (Fig. 6A). The efficacy of Th17 cell suppression was then evaluated using flow cytometry.

In the standard CRS depressive model, male mice underwent daily restraint for 3–4 h over 14 days, followed by 10 days of antibiotic treatment and a 3-day washout. Mice were subsequently orally administered with PBS, live, or heat-killed B. thetaiotaomicron and B. vulgatus (10⁸ CFU/mL in sterile PBS, 0.2 ml per mouse) every other day for 7 days. Behavioral assessments were performed 24 h after the final oral administration (Fig. 3A).

Open-field test (OFT)

Under dim lighting, mice were placed at the center of an open field (40 cm × 40 cm × 40.5 cm) and allowed to explore for 10 min⁸². A video camera positioned overhead recorded their movements, which were analyzed using Smart 3.0 software (Ruiwode). The time spent and distance traveled in the center, as well as the total distance traveled, were measured.

Sucrose preference test (SPT)

In the SPT, mice were individually housed and provided with one water bottle and one bottle containing a 2% sucrose solution for 2 days, with bottle positions switched after the first day. Following 24 h of water deprivation, mice were re-exposed to both bottles for 2 h, with their positions switched after 1 h. Sucrose preference was calculated as the ratio of sucrose consumption to total fluid intake.

Forced swim test (FST)

Each mouse was individually placed in a cylindrical tank (12 cm in diameter, 30 cm in height) filled with water maintained at 23–25 °C and allowed to swim for 6 min under standard lighting conditions⁸³. The water depth was adjusted to prevent the mice from contacting the tank bottom with their tails or hind limbs. Side-view recordings of their behavior were analyzed offline by an observer blinded to the experimental treatments. Immobility duration, defined as the minimal movements required to stay afloat without shifting the trunk or center of gravity, was calculated for the final 4 min of the test.

Novelty-suppressed feeding test (NSFT)

Before testing, mice were food-deprived for 24 h, with water available ad libitum. During the experiment, each mouse was placed in a plastic box (35 cm × 35 cm)⁸⁴. An observer manually recorded the time taken for the mouse to initiate pellet biting (feeding latency). If no biting occurred within 5 min, the latency was recorded as 5 min.

Cytokine quantification

Protein lysates from the hippocampus were analyzed using a mouse cytokine array kit (QAM-TH17-1, RayBiotech) according to the manufacturer’s instructions. Arrays were scanned with an InnoScan 300 Microarray Scanner (Innopsys, France), and raw signals were processed using MapPix 6.0 software. Data analysis was performed with Q-Analyzer software, which generated calibration curves for each cytokine. To ensure accuracy, saturated fluorescence signals were excluded, and only calibration curves with a strong fit (R² > 0.98) were considered, using linear or log–log regression models. Cytokine concentrations (pg/mL) were interpolated from these curves.

Monocytes isolation and flow cytometry analysis

Spleen, Peyer’s patches, and blood were collected from mice to generate single-cell suspensions. To quantify Th1 and Th17 cell populations, cells were stimulated with a leukocyte activation cocktail (2 μL/mL, Cat. No. 550583, BD Biosciences) and incubated at 37 °C with 5% CO₂ for 6 h. After incubation, cells were washed twice with staining buffer (Cat. No. 554656, BD Biosciences) and analyzed by flow cytometry. Surface staining was performed using Fixable Viability Stain 780 (0.2 mg/ml, Cat. No. 565388, BD Biosciences), anti-CD45 (0.2 mg/ml, Cat. No. 103114, BioLegend), and anti-CD4 (0.2 mg/ml, Cat. No. 563727, BD Biosciences), followed by washing, fixation, and permeabilization with Perm/Fix buffer (Cat. No. 00-5523-00, Thermo Fisher). After two additional washes with Perm/Wash buffer, intracellular cytokine staining was performed using anti-IFN-γ (0.2 mg/ml, Cat. No. 557724, BD Biosciences) and anti-IL-17A (0.2 mg/ml, Cat. No. 563354, BD Biosciences). Following two final washes, cells were resuspended in staining buffer for flow cytometry analysis. Flow cytometry was conducted using a Cytek NL-CLC3000 spectral flow cytometer, with data analyzed using FlowJo software.

Subjects

All ethical regulations relevant to human research participants were followed. This study was approved by the Independent Ethics Committee (IEC) of Beijing Anding Hospital (approval no. 2020-106) and retrospectively registered in the Chinese Clinical Trial Registry (ChiCTR-OOC-17012566) on September 4, 2017⁸⁵. Written informed consent was obtained from all participants. Inclusion criteria were as follows: age between 18 and 65 years; diagnosis based on Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria, confirmed through a clinical interview with experienced psychiatrists; a score of ≥11 on the 16-item Quick Inventory of Depressive Symptomatology-Self-Report (QIDS-SR-16)⁸⁶ and ≥14 on the 17-item Hamilton Depression Rating Scale (HAMD-17). Exclusion criteria included a lifetime history of bipolar disorder, schizophrenia, schizoaffective disorder, or other psychiatric conditions; prior use of antidepressants for more than 7 days within 14 days before the current episode; substance abuse (drug or alcohol dependence); pregnancy or breastfeeding; high suicide risk (HAMD-17 item 3 score ≥3); and serious medical conditions (e.g., chronic inflammatory disorders, diabetes, cardiovascular disease, thyroid disease, or cancer). HCs were recruited through advertisements and were excluded if they had a lifetime history of any DSM-IV psychiatric disorder, significant medical conditions (e.g., chronic inflammatory disorders, diabetes, cardiovascular disease, thyroid disease, or cancer), or were pregnant or breastfeeding.

Abundance analysis of Bacteroides species by shotgun metagenomic sequencing

Fecal samples were collected from participants, immediately frozen, and stored at −80 °C until analysis. DNA extraction, library preparation, and metagenomic sequencing were conducted¹³. DNA was extracted using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek, USA), and its concentration, purity, and integrity were assessed with a TBS-380 fluorometer, NanoDrop 2000 spectrophotometer, and 1% agarose gel electrophoresis, respectively. DNA was sheared to ~300 bp with a Covaris M220 (Gene Company Limited, China) and used for paired-end library construction with the NEXTFLEX® Rapid DNA-Seq Kit (Bioo Scientific, USA). After adapter ligation, libraries were sequenced on the Illumina NovaSeq platform (Illumina Inc., USA) at Shanghai Majorbio Bio-Pharm Technology Co., Ltd., following the manufacturer’s protocols. For this study, sequencing reads were first aligned to the NCBI NR database. Taxon abundance was then quantified at the species level using RPKM (Reads Per Kilobase per Million mapped reads). Subsequently, three Bacteroides species (B. uniformis, B. thetaiotaomicron, and B. vulgatus) were selected for comparative analysis of abundance between the HC and MDD groups.

Peripheral blood mononuclear cells (PBMCs) and flow cytometric analyses

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized whole blood using Ficoll-Hipaque density gradient centrifugation. Antibodies for staining were obtained from BD Biosciences (CA, USA; used at 5 µL per 10⁶ cells in 100 µL staining volume). PBMCs were divided into two panels for lymphocyte subset analysis. For Th1/Th2 and intracellular FoxP3 detection, cells were stained with anti-CD3 (Cat. No. 563800), anti-CD4 (Cat. No. 566392), anti-CD8 (Cat. No. 563677), anti-CD19 (Cat. No. 562947), anti-CD56 (Cat. No. 564057), anti-CD16 (Cat. No. 562293), anti-CD25 (Cat. No. 564467), anti-CD127 (Cat. No. 612794), anti-CD45RA (Cat. No. 562885), anti-CCR6 (Cat. No. 562724), anti-CCR7 (Cat. No. 557648), anti-CXCR3 (Cat. No. 565223) and FVS780 (Cat. No. 565388). After centrifugation, cells were fixed and permeabilized and labeled with anti-FoxP3(Cat. No. 17-4777-42, eBioscience). For Th17 cell analysis, PBMCs were stimulated with a leukocyte activation cocktail and stained with anti-CD3 (Cat. No. 563800), anti-CD4 (Cat. No. 566392) and anti-CD8 (Cat. No. 563677) and FVS780 (0.2 mg/ml, Cat. No. 565388). After fixation and permeabilization, cells were labeled with anti-IL-17A. Following PBS washes, samples were analyzed using a BD FACSCanto II flow cytometer, and data were processed with BD FACSDiva software.

Metabolomics profiling and data analysis

Targeted metabolomic profiling was performed using the AccuQuanter-1800 platform, which combines four complementary Liquid Chromatography–Tandem Mass Spectrometry (LC–MS/MS) assays to quantify approximately 600 metabolites in biological matrices. Metabolites were classified into assay panels based on their chemical characteristics and physiological ranges. Each sample underwent four sequential extractions, and pooled quality control (QC) samples were prepared to monitor analytical performance. For each extraction, 20–50 μL of serum was mixed with pre-cooled methanol or methanol/acetonitrile containing isotope-labeled internal standards for protein precipitation. After vortexing and centrifugation, the supernatant was analyzed by LC–MS/MS. For Assay 3, metabolites containing carboxyl or aldehyde groups were derivatized with 3-nitrophenylhydrazine before analysis. Assay 1 quantified 133 polar endogenous metabolites, including amino acids, nucleotide bases, oligopeptides, vitamins, and choline/carnitine derivatives, separated via hydrophilic interaction chromatography (HILIC). Assay 2 detected 57 bile acids and 54 fatty acids by reversed-phase chromatography under negative ESI mode. Assay 3 profiled 86 highly polar intermediates from the TCA cycle, glycolysis, and other organic acids, following derivatization. Assay 4 targeted 108 microbial-associated metabolites, including polyamines and catabolites of tryptophan, tyrosine, and phenylalanine. Data from Panels 4 were processed with Analyst 1.7 and OS-MQ (AB SCIEX, Singapore), and data from Panels 1, 2, and 3 with TraceFinder (Thermo Scientific, USA). Quantification was performed using isotope-labeled internal calibration.

Statistics and reproducibility

Results are presented as mean ± SEM. Statistical analyses were conducted using one-way ANOVA followed by Tukey’s post hoc test or an unpaired Student’s t-test, as appropriate. All analyses were performed using GraphPad Prism 8.0.2 (San Diego, CA, USA), with statistical significance set at p < 0.05. Sample sizes varied depending on the experiment and ranged from n = 6 to n = 13 per group. Exact sample sizes and relevant statistical details are provided in the figure legends.

Mice and ethical statement

Seven-week-old male and female C57BL/6 J mice were purchased from Weitong Lihua Co., Ltd. (Beijing, China) and housed in the animal facility of Capital Medical University (Beijing, China). Mice were maintained under specific pathogen-free (SPF) conditions in an individually ventilated cages (IVC) barrier system with controlled temperature (22 ± 1 °C), humidity (52.5 ± 2.5%), a standard laboratory diet, and ad libitum access to water. The light/dark cycle was set from 08:00 a.m. to 08:00 p.m. We have complied with all relevant ethical regulations for animal use. The study was approved by the Animal Use and Care Committee of Capital Medical University (Permit No. AEEI-2023-155).

Reporting summary

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

Author contributions

Zhiyu Li conducted the experiments, analyzed the data, and drafted the manuscript. Peilin Qin, Zuoli Sun, Liangkang Li, and Peng Liang participated in experiments and contributed to data collection; Yimei Zhao and Siyu Ren assisted in data analysis; Gang Wang and Jian Yang were responsible for project design and manuscript revision. All authors have reviewed and approved the final version of the manuscript.

Peer review

Peer review information

Communications Biology thanks Hassan Zafar and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Ibrahim Javed, David Favero, and Tobias Goris. A peer review file is available.

Data availability

All source data have been deposited in Figshare Repository (10.6084/m9.figshare.30814370) (ref. ⁸⁷) and are also provided as Supplementary Data. The metagenomic sequencing data has been deposited in Genome Sequence Archive in National Genomics Data Center (GSA: CRA035718) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa (refs. ^88,89).

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-09525-x.