Gut microbiota depletion accelerates hematoma resolution and neurological recovery after intracerebral hemorrhage via p-coumaric acid-promoted Treg differentiation

Abstract

Hematoma volume significantly influences the prognosis of patients with intracerebral hemorrhage (ICH). Effective resolution of hematoma through enhanced clearance mechanisms and reduced hematoma lysis is essential for neurological recovery following ICH. Regulatory T cells (Tregs), known for their anti-inflammatory properties, exert neuroprotective effects in various central nervous system disorders. Additionally, gut microbiota profoundly impacts Treg development through multiple regulatory pathways. Nonetheless, the precise roles of Tregs and gut microbiota in facilitating hematoma resolution after ICH remain unclear. This study, therefore, aimed to investigate the contributions of Tregs and gut microbiota to hematoma resolution post-ICH, as well as the underlying mechanisms. Methods: The impact of gut microbiota depletion on neurological deficits and hematoma resolution, including erythrophagocytosis and erythrocyte lysis, was assessed using antibiotic cocktail (ABX) gavage administered prior to ICH induction in mice. Flow cytometry analysis and targeted cell depletion techniques were employed to identify peripheral immune cell populations mediating the beneficial effects of gut microbiota depletion on neurological recovery and hematoma resolution. The functional roles of Tregs in erythrophagocytosis, erythrocyte lysis, and associated downstream molecular signaling pathways were investigated through adoptive Treg transfer experiments. The mechanisms underlying Treg population expansion post-microbiota depletion in ICH mice were explored using multi-omics analysis of serum and fecal metabolites via mass spectrometry and fecal microbial composition using 16S rRNA sequencing. Additionally, the effects of p-coumaric acid (PCA) gavage and clindamycin-mediated depletion of PCA-metabolizing gut microbiota on Treg abundance, hematoma resolution, and neurological recovery post-ICH were assessed. Results: Gut microbiota depletion by ABX gavage increased brain Treg populations, thereby enhancing erythrophagocytosis, suppressing erythrocyte lysis, and ultimately promoting hematoma resolution and neurological recovery. Adoptive Treg transfer experiments further established that Tregs facilitate scavenger pathway-mediated erythrophagocytosis and suppress complement-mediated erythrocyte lysis. These effects occurred via upregulation of efferocytosis receptors (MERTK and AXL), ligands (Gas6 and C1q), and the hemoglobin scavenger receptor CD163, alongside downregulation of complement C3 expression and reduced formation of membrane attack complexes (MACs). Multi-omics analysis demonstrated that ABX gavage eliminated PCA-metabolizing microbiota, thereby increasing PCA concentrations in serum and feces. Elevated PCA levels promoted peripheral Treg differentiation by inhibiting the PKCθ-AKT-FoxO1/3a signaling pathway, leading to higher brain Treg numbers. PCA gavage and clindamycin treatment similarly enhanced brain Treg populations, accelerated hematoma resolution, and improved neurological recovery following ICH. Conclusion: Gut microbiota depletion facilitates hematoma resolution and neurological recovery through PCA-mediated induction of Treg differentiation.

Keywords: Regulatory T cells; gut microbiota; hematoma resolution.; intracerebral hemorrhage; p-coumaric acid.

Figure 1

Gut microbiota depletion promotes hematoma resolution and neurological recovery by enhancing erythrophagocytosis and inhibiting hematoma lysis on day 3 post-ICH. (A) Schematic diagram of the experimental workflow illustrating the effects of antibiotic cocktail (ABX) gavage on gut microbiota composition, hematoma resolution, and neurological recovery after ICH in mice. Mice received ABX by oral gavage for 7 days, followed by a 2-day washout period, and subsequently underwent injection to induce ICH. Fecal samples were collected on day 3 post-ICH for 16S rRNA sequencing. Neurological function and residual hematoma volume were assessed on days 3 and 7 post-ICH. (B) Chao1 α-diversity indices of fecal microbiota in SPF and ABX-treated mice. Data are presented as median ± IQR (n = 5 per group). (C) Neurological function assessments in SPF and ABX-treated mice using the cylinder, corner, and forelimb placement tests. Data are presented as median ± IQR (n = 9 per group). (D) Representative coronal brain sections showing residual hematoma on days 3 and 7 post-ICH. Scale bar = 5 mm. Quantification of residual hematoma volume in SPF and ABX groups on days 3 and 7 post-ICH. Data are presented as mean ± SD (n = 9 per group). (E) Schematic diagram illustrating the effects of fecal microbiota transplantation (FMT) on hematoma resolution, neurological recovery, and erythrophagocytosis in ABX-treated mice post-ICH. (F) Representative coronal brain sections showing residual hematoma in SPF, ABX, and FMT groups on day 3 post-ICH. Scale bar = 5 mm. Quantification of residual hematoma volume in each group. Data are presented as median ± IQR (n = 9 per group). (G) Neurological function assessments in SPF, ABX, and FMT groups using the cylinder, corner, and forelimb placement tests on day 3 post-ICH. Data are presented as median ± IQR (n = 9 per group). (H, I) Schematic diagram illustrating the workflow for in vivo erythrophagocytosis analysis. Red blood cells (RBCs) were isolated via Ficoll gradient centrifugation from whole blood collected by cardiac puncture, labeled with the fluorescent dye DiD, and resuspended in autologous plasma (1:4 ratio). A 30-μL mixture was injected into the striatum to induce ICH. Microglia/macrophages engaging in erythrophagocytosis were identified by flow cytometry using the gating strategy for LIVE/DEAD^-CD45^int/hiCD11b⁺-DiD⁺ cells. (J) Percentages of erythrophagocytic microglia/macrophages in SPF, ABX, and FMT groups on day 3 post-ICH (n = 6 per group). Data are presented as mean ± SD. (K) Schematic diagram illustrating the experimental workflow used to evaluate the effects of FMT on hematoma lysis and ventricular compression in ABX-treated mice after ICH. Mice in each group were anesthetized on day 3 post-ICH, and T2- and T2*-weighted magnetic resonance imaging (MRI) was performed to assess hematoma lysis and ventricular compression. (L) Representative T2- and T2*-weighted MRI scans of SPF, ABX, and FMT groups on day 3 post-ICH. Dashed lines indicate hematoma location in the brain. (M) Quantification of T2* hyperintense or isointense lesion volume, and ipsilateral ventricular compression as an indicator of brain swelling (n = 6 per group). Data are presented as mean ± SD. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001; ns = not significant.

Figure 2

Gut microbiota depletion increases the brain Treg population to promote hematoma resolution and neurological recovery after ICH. (A) FACS gating strategy used to isolate neutrophils (LIVE/DEAD^-CD45^int/hiCD11b⁺ly6G⁺CD4^-), monocytes/macrophages (LIVE/DEAD^-CD45^int/hiCD11b⁺ly6G^-CD4^-), Tregs (LIVE/DEAD^-CD45^int/hly6G^-CD4⁺CD25⁺Foxp3⁺), and other immune cells (LIVE/DEAD^- CD45^int/hily6G^-CD4⁺CD25^-Foxp3^-) among peripheral infiltrating immune cells in the brains of SPF and ABX-treated mice on day 3 post-ICH. Quantification of immune cell populations was performed via flow cytometry. Data are presented as mean ± SD (n = 6 per group). (B) FACS gating strategy used to quantify Tregs (LIVE/DEAD^-CD45^int/hiCD4⁺CD25⁺Foxp3⁺) among CD4⁺ T helper cells (LIVE/DEAD^-CD45^int/hiCD4⁺) in the brains of SPF, ABX, and FMT mice on day 3 post-ICH. Data are presented as mean ± SD (n = 9 per group). (C) Schematic diagram illustrating the experimental workflow for evaluating the effects of Treg depletion on hematoma resolution, neurological recovery, microglial/macrophage erythrophagocytosis, and hematoma lysis after ICH in ABX-treated mice. Mice received an intraperitoneal injection of anti-CD25 antibody or IgG control 2 days before ICH induction. (D) Representative coronal brain sections from SPF + IgG, ABX + IgG, and ABX + anti-CD25 antibody-treated mice on day 3 post-ICH, showing residual hematomas. Scale bar = 5 mm. Quantification of residual hematoma volume on day 3 post-ICH. Data are presented as mean ± SD (n = 9 per group). (E) Neurological function assessed on day 3 post-ICH in SPF + IgG, ABX + IgG, and ABX + anti-CD25 antibody-treated mice using the cylinder, corner, and forelimb placement tests. Data are presented as median ± IQR (n = 9 per group). (F) Representative T2- and T2*-weighted MRI scans of SPF + IgG, ABX + IgG, and ABX + anti-CD25 antibody-treated groups on day 3 post-ICH. Dashed lines indicate the location of hematoma in the brain. Quantification of T2* hyperintense or isointense lesion volumes and ipsilateral ventricular compression as an indicator of brain swelling. Data are presented as mean ± SD (n = 6 per group). (G) Flow cytometry analysis of erythrophagocytic microglia/macrophages in SPF + IgG, ABX + IgG, and ABX + anti-CD25 antibody-treated groups on day 3 post-ICH. Data are presented as mean ± SD (n = 6 per group). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

Tregs promote scavenger pathway-mediated erythrophagocytosis and suppress complement-mediated erythrolysis to facilitate hematoma resolution and neurological recovery. (A) Schematic diagram illustrating the experimental workflow to assess the effects of adoptive Treg transfer on hematoma resolution, neurological recovery, and hematoma lysis after ICH. Tregs were magnetically isolated from splenocytes of donor mice and administered via tail vein injection to recipient mice immediately after ICH induction. (B) Neurological function assessed in ICH + Vehicle and ICH + Treg mice using the cylinder, corner, and forelimb placement tests on day 3 post-ICH. Data are presented as median ± IQR (n = 9 per group). (C) Representative coronal brain sections showing residual hematomas in ICH + Vehicle and ICH + Treg groups on day 3 post-ICH. Scale bar = 5 mm. Quantification of residual hematoma volume in each group on day 3 post-ICH. Data are presented as median ± IQR (n = 9 per group). (D) Representative T2- and T2*-weighted MRI scans from ICH + Vehicle and ICH + Treg groups on day 3 post-ICH. T2* hyperintense or isointense lesion volumes were quantified, and ipsilateral ventricular compression was measured as an indicator of brain swelling. Data are presented as mean ± SD (n = 9 per group). (E) Schematic diagram illustrating the workflow to evaluate the effects of adoptive Treg transfer on microglial/macrophage erythrophagocytosis and associated downstream molecular markers. Tregs were transferred into recipient mice via tail vein injection after ICH. (F) Percentages numbers of erythrophagocytic microglia/macrophages assessed by flow cytometry in ICH + Vehicle and ICH + Treg groups on day 3 post-ICH. Data are presented as mean ± SD (n = 6 per group). (G) Membrane attack complex (MAC) content measured in perihematomal brain tissue of Sham + Vehicle, ICH + Vehicle, and ICH + Treg groups on day 3 post-ICH using ELISA. Data are presented as median ± IQR (n = 6 per group). (H) Western blot analysis of C1q, C3, AXL, MERTK, Gas6, CD163, and β-actin protein levels in perihematomal brain tissue from Sham + Vehicle, ICH + Vehicle, and ICH + Treg groups on day 3 post-ICH. (I) Densitometric quantification of C1q, C3, AXL, Mertk, Gas6, and CD163 protein levels based on Western blot results. Data are presented as mean ± SD (n = 6 per group). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Gut microbiota depletion triggers Treg production in the gut and spleen. (A) Percentages of Tregs in the lamina propria of large intestine, small intestine, and spleen in SPF, ABX, and FMT groups, determined by flow cytometry on day 3 post-ICH. Representative contour plots represent Treg populations in the lamina propria of the large intestine (top), small intestine (middle), and spleen (bottom). Data are presented as mean ± SD (n = 6 per group). (B) Schematic diagram illustrating the experimental workflow for assessing the effects of serum-derived factors from different experimental groups on Naïve CD4⁺ T cells differentiation into Tregs. Naïve CD4⁺ T cells were magnetically sorted from splenocytes and cultured under suboptimal Treg-inducing conditions (0.1 ng/mL TGF-β1, 10 ng/mL IL-2, CD3-T cell receptor [TCR] and anti-CD28 signaling) with serum from the following groups: SPF + Sham, SPF + ICH, ABX + Sham, and ABX + ICH. Treg differentiation was analyzed by flow cytometry on day 3. (C) Percentages of Tregs generated in vitro under the influence of serum from SPF + Sham, SPF + ICH, ABX + Sham, and ABX + ICH groups as determined by flow cytometry. Representative contour plots represent Treg populations in each group. Data are presented as mean ± SD (n = 6 per group).

Figure 5

Gut microbiota depletion enhances PCA levels in serum and feces, promoting Treg differentiation. (A) Schematic diagram illustrating the experimental workflow for serum metabolomic profiling. Blood samples were collected via cardiac puncture from ABX-treated mice on day 3 post-ICH and incubated at room temperature for 1 hour. Serum was isolated by centrifugation and subjected to targeted metabolomic analysis. (B, C) Heatmaps and volcano plots showing differential serum metabolite profiles among SPF + Sham, SPF + ICH, ABX + Sham, and ABX + ICH groups on day 3 post-ICH. The top three upregulated metabolites in ABX-treated mice compared with SPF mice were TCA, Tα/βMCA, and PCA (n = 8 per group). (D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of differentially expressed serum metabolites between SPF and ABX groups on day 3 post-ICH (n = 8 per group). Statistical analysis: ANOVA followed by Bonferroni post hoc tests in (B); and Student's t-test in (C-D). Fold-change scale adjusted to 4; Uni p < 0.05 considered statistically significant (B-D). (E) Naïve CD4⁺ T cells were cultured with Vehicle, TCA, TβMCA, or PCA under Treg-inducing conditions. Percentages of Tregs generated in each treatment group were determined by flow cytometry. Representative contour plots represent percentages of generated Tregs in each group. Data are presented as mean ± SD (n = 6 per group). (F) PCA concentrations in feces and serum of SPF + Sham, SPF + ICH, and ABX + ICH groups on day 3 post-ICH. Data are presented as median ± IQR (n = 8 per group).

Figure 6

p-Coumaric acid (PCA) triggers Treg differentiation via the PKCθ-AKT-FoxO1/3a pathway. Oral gavage of PCA or clindamycin to deplete PCA-metabolizing microbiota increases the brain Treg population, facilitates hematoma resolution, and accelerates neurological recovery after ICH. (A) Western blot analysis of naïve CD4⁺ T cells treated with Vehicle or PCA for 0, 36, and 72 hours. Target proteins included p-PKCθ, total PKCθ, p-AKT, total AKT, p-FoxO1, total FoxO1, p-FoxO3a, total FoxO3a, and β-actin. (B) Quantitative analysis of p-PKCθ to total PKCθ, p-AKT to total AKT, p-FoxO1 to total FoxO1, and p-FoxO3a to total FoxO3a ratios based on Western blot analysis. Data are presented as mean ± SD (n = 6 per group). (C) Flow cytometry analysis of Treg differentiation in naïve CD4⁺ T cells treated with Vehicle, PCA, or PCA + PMA. Percentages of generated Tregs in each treatment group are shown with representative contour plots. Data are presented as mean ± SD (n = 6 per group). (D) Schematic diagram illustrating the experimental workflow to assess the effects of PCA oral gavage on brain Treg population, hematoma resolution and neurological recovery after ICH. (E) Percentages of Tregs in the brains of ICH + Vehicle and ICH + PCA groups, determined by flow cytometry. Representative contour plots represent percentages of Tregs in the brains of ICH + Vehicle and ICH + PCA groups. Data are presented as mean ± SD (n = 6 per group). (F) Representative coronal brain sections showing residual hematomas in ICH + Vehicle and ICH + PCA groups on day 3 post-ICH. Scale bar = 5 mm. Quantification of residual hematoma volume in ICH + Vehicle and ICH + PCA groups on day 3 post-ICH. Data are presented as mean ± SD (n = 9 per group). (G) Neurological function assessed using the cylinder, corner, and forelimb placement tests in ICH + Vehicle and ICH + PCA groups. Data are presented as median ± IQR (n = 12 per group). (H) Proposed metabolic pathway of PCA metabolism via phenolic acid decarboxylase, an enzyme expressed by specific gut bacterial species. (I) Line graph illustrating reduced relative abundance of four PCA-metabolizing bacterial species in ABX-treated mice compared with SPF mice on day 3 post-ICH (p < 0.05). Red, blue, purple, and green lines represent Lactobacillus johnsonii, Rikenella microfusus, Clostridiales bacterium, and Clostridium dlviii, respectively. (J) Percentages of Tregs in the brains of ICH + Vehicle and ICH + Clindamycin groups on day 3 post-ICH, determined by flow cytometry. Representative contour plots represent percentages of Tregs in the brains of ICH + Vehicle and ICH + Clindamycin mice. Data are presented as mean ± SD (n = 6 per group). (K) Representative coronal brain sections from ICH + Vehicle and ICH + Clindamycin groups on day 3 post-ICH showing residual hematomas. Scale bar = 5 mm. Quantification of residual hematoma volume on day 3 post-ICH. Data are presented as mean ± SD (n = 10 per group). (L) Neurological function assessed by the cylinder, corner, and forelimb placement tests in ICH + Vehicle and ICH + Clindamycin groups. Data are presented as median ± IQR (n = 12 per group). Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001.