Vagal innervation limits brain injury by inhibiting gut-selective integrin-mediated intestinal immunocyte migration in intracerebral hemorrhage

Abstract

Rationale: The vagus nerve, which connects the brain and gastrointestinal tract, helps to maintain immune balance in the intestines. Gut-specific integrins, on the other hand, help to keep immune cells in the intestines. Since immune cells from outside the intestines can significantly affect the outcome of strokes, we investigated how immune cells from the intestines affect the immune response in the brain during intracerebral hemorrhage (ICH). Methods: We aimed to examine the impact of vagal innervation on intestinal immunocyte trafficking and its influence on ICH outcomes using Kikume Green-Red (KikGR) and wildtype (WT) mice, with or without prior subdiaphragmatic vagotomy (SDV). Furthermore, we sought to elucidate the regulatory effects of vagal innervation on intestinal immunocyte trafficking by activating α7 nicotinic acetylcholine receptors (α7nAChR) in WT mice that underwent ICH after SDV. Additionally, we explored the potential intermediary role of gut-selective integrins in cholinergic transmitters-mediated intestinal immunocyte trafficking. Our methodology encompassed in vivo fluorescence imaging, flow cytometry, Western blotting, immunofluorescence staining, histopathology, and behavioral assessments to evaluate the outcomes. Results: Our findings reveal that during the acute phase of ICH, intestinal immunocytes migrated to various anatomical locations, including the circulation, hemorrhagic brain, meninges, and deep cervical lymph nodes. Pertinently, SDV resulted in diminished expression of α4β7 and αEβ7 integrins on immunocytes, leading to heightened intestinal immunocyte trafficking and exacerbated ICH outcomes. Conversely, the administration of α7nAChR agonists countered the adverse effects of vagotomy on α4β7 and αEβ7 integrin expression, thereby constraining the migration of immune cells from the intestines after ICH. The implication of α4β7 and αEβ7 integrins in this setting was supported by the ineffective influence of α7nAChR agonists on the trafficking of intestinal immunocytes enhanced by administering beta-7 integrin antagonists, such as etrolizumab. It was further supported by the exacerbated ICH outcomes by administering beta-7 integrin antagonists like etrolizumab alone. Conclusion: The identification of vagus nerve-mediated modulation of α4β7 and αEβ7 integrin expression in the trafficking of immune cells within the intestinal tract holds significant implications. This discovery presents new opportunities for developing therapeutic interventions for ICH and stimulates further investigation in this area.

Keywords: Gut-selective integrin; Immunomodulatory therapy; Intestinal immunocyte trafficking; Intracerebral hemorrhage; Vagus nerve; α7nAChRs.

Figure 1

Dynamic fluorescent changes detected with the in vivo imaging system in KikGR mice experienced SDV, photoconversion, and ICH. A) Workflow for processing KikGR mice. Photoconversion of the MLNs was induced with ultraviolet light irradiation in KikGR mice 20 d after SDV. The ICH model was established 4 d after photoconversion. Fluorescent detection was conducted immediately before photoconversion, immediately before ICH, and 3 d after ICH. B) Representative green (KikG) and red (KikR) fluorescent images observed with the in vivo imaging system in the heads and bodies of WT mice 3 d after ICH and KikGR mice had previously undergone SDV immediately before photoconversion and 4 d (immediately before ICH) and 7 d (3 d after ICH) after photoconversion. The fluorescent intensity of KikG and KikR in WT mice was set as zero standard. C-H) Analysis for the average intensity of red fluorescence (KikR) in the heads and bodies of KikGR mice that had previously undergone a sham vagotomy or SDV at the above 3 time points. n = 5 mice per group. The two-tailed Mann-Whitney U test for the red fluorescence (KikR) was performed in the bodies 7 d after photoconversion and the two-tailed paired t-test for others. No difference was detected. I) Representative fluorescent green (KikG) and red (KikR) images detected with the in vivo imaging system in intact isolated brains, dCLNs, MLNs, and brain sections of KikGR mice that had previously undergone a sham vagotomy or SDV 7 d after photoconversion (3 d after ICH). Similar tissues, isolated from WT mice 3 d after ICH, were used as zero controls to detect green (KikG) and red (KikR) fluorescence. J-M) Analysis of the mean intensity of red fluorescence (KikR) in the immediately removed brain, dCLNs, MLNs, and brain sections of KikGR mice had previously undergone a sham vagotomy or SDV 7d after photoconversion (3 d post-ICH). n = 5 mice per group. The two-tailed paired t-test for red fluorescence (KikR) was performed in the intact brain and brain sections, and the Mann-Whitney U test for red fluorescence (KikR) was performed in dCLNs and MLNs. *P < 0.05, **P < 0.01.

Figure 2

SDV increases intestinal immunocyte trafficking in KikGR mice with ICH. A) Representative flow cytometric results for photoconverted immunocytes (CD45⁺KikR⁺ cells) in MLNs of mice 7 d after sham surgery or photoconversion. These mice did not experience SDV or ICH before and after the conversion surgery. B) Analysis of photoconverted immunocytes (CD45⁺KikR⁺ cells) in MLNs of mice mentioned above. n = 5 mice per group. The two-tailed Mann-Whitney U test was performed. **P < 0.01. C) The fluorescence in brain sections from non-photoconverted and photoconverted mice on day 3 after ICH. These mice did not experience SDV but underwent sham conversion surgery or photoconversion 4 d before ICH. The insets in the lower figures show a higher magnification of non-photoconverted (KikG⁺) and photoconverted (KikR⁺) cells. Scale bar of the upper figures = 500 μm. Scale bar of the lower figures = 50 μm. D, F, H, J, and L) Representative flow cytometric results for the detection of CD45⁺KikR⁺ cells in brain tissues (D), meninges (F), peripheral blood (H), dCLNs (J), and MLNs (L) 3 d after ICH. These mice experienced SDV 20 d before photoconversion and underwent ICH 4 d after photoconversion. E, G, I, K, and M) Analysis for the percentages of red fluorescent cells in CD45⁺ cells in brain tissues (E), meninges (G), peripheral blood (I), dCLNs (K), and MLNs (M) 3 d after ICH. n = 5 mice per group. The two-tailed paired t-test was performed. *P < 0.05, **P < 0.01.

Figure 3

SDV enhances the cellular and molecular inflammatory response in the hemorrhagic brain. A) Schematic diagram of the selected fields for the quantification of fluorescent-stained cells. B) Representative images of immunofluorescence staining for Iba-1, GFAP, MPO, and FJB in the perihematomal regions 3 d after ICH in WT mice had previously undergone sham vagotomy or SDV. The insets show a higher magnification of immunofluorescence-staining positive cells. Scale bar = 100 μm. C) Analysis for microglia/macrophage and astrocytic activation, neutrophil infiltration, and neural degeneration 3 d after ICH. n = 6 mice per group. The two-tailed Mann-Whitney U test for MPO and the two-tailed paired t-test for others were performed. *P < 0.05, **P < 0.01. D) Representative Western blot bands of HMGB1, IL-1β, IL-6, and TNF-α 1 day after sham operation and hemorrhagic brains of mice that had previously experienced sham vagotomy or SDV earlier. β-actin was used as the loading control. E) Quantitative analysis of the expression of pro-inflammatory factors HMGB1, IL-1β, IL-6, and TNF-α in hemorrhagic brains 1 d after ICH. n = 5 mice per group. The Kruskal-Wallis test for IL-1β and one-way analysis of variance (ANOVA) with Bonferroni correction for others were performed. *P < 0.05, **P < 0.01.

Figure 4

SDV exacerbates the histopathological and behavioral deficits of mice with ICH. A) Representative brain sections stained with LFB/CV 3 d after ICH in WT mice that had previously experienced SDV or sham vagotomy. The areas of the lesion lacking staining are circled with a black curve (indicated by a red arrow); scale bar = 500 μm. B) Analysis for brain injury volume 3 d after ICH in WT mice that had previously experienced sham vagotomy or SDV. n = 10 mice per group. The two-tailed paired t-test was performed, *P < 0.05. C) Analysis of brain swelling 3 d after ICH in WT mice that had previously experienced a sham vagotomy or SDV. n = 10 mice per group. The two-tailed paired t-test was performed. No difference was found. D) Changes in brain water content 3 days after ICH in sham-operated mice and ICH mice that had previously experienced sham vagotomy or SDV. n = 6 mice per group. A one-way ANOVA analysis with Bonferroni correction for multiple comparisons was performed, *P < 0.05, **P < 0.01. E) Representative residual lesions (indicated by red arrows) in brain sections 28 d after ICH in WT mice that had previously undergone SDV or sham vagotomy. Scale bar = 500 μm. F) Analysis of residual lesions in the hemorrhagic brain 28 d after ICH in WT mice that had previously undergone sham vagotomy or SDV. n = 10 mice per group. The two-tailed paired t-test was performed, *P < 0.05. G) Analysis of brain atrophy in the hemorrhagic brain 28 d after ICH in WT mice that had previously undergone a sham vagotomy or SDV. n = 10 mice per group. The Mann-Whitney U test was performed. No significant difference was found. H) A schematic diagram of the fields selected for quantifying myelin loss. I) Representative images in the perihematomal regions stained with LFB/CV 28 d after ICH in WT mice that had previously undergone sham vagotomy or SDV. Scale bar =100 μm. J) Quantitative analysis of white matter damage 28 d after ICH in WT mice that had previously undergone sham vagotomy or SDV. n = 10 mice per group. The two-tailed paired t-test was performed. No significant difference was found. K) Neurological deficits evaluated with neurologic deficit scores (NDS) throughout the 28-day research process in WT mice that had previously undergone a sham vagotomy or SDV. n = 10 mice per group. Generalized estimation equations (GEE) with a two-tailed Mann-Whitney U test were performed at multiple time points. Wald χ² = 61.68 and P < 0.001 for the total trend of SDV versus sham vagotomy; P < 0.001 on day 3, P = 0.036 on day 7, P = 0.004 on day 14, and P = 0.001 on day 28 for SDV versus sham vagotomy, respectively. L) Neurologic deficits were evaluated with the corner turn test (CTT) daily throughout the 28-day research process in WT mice that had previously undergone SDV or sham vagotomy. n = 10 mice per group. Generalized estimation equations (GEE) with a two-tailed Mann-Whitney U test were performed at multiple time points. Wald χ² = 11.243 and P = 0.010 for the total trend of SDV versus sham vagotomy; P = 0.03 on day 7 and P = 0.004 on day 28 for SDV versus sham vagotomy.

Figure 5

α7nAChR agonists reverse the homing and retention of intestinal immunocytes of ICH mice that previously underwent SDV. A) Analysis of the influence of previous SDV and α7nAChR agonist treatment on the infiltration of CD3⁺, CD4⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells quantified by flow cytometry in the brain lesion 3 d after ICH. n = 5-6 mice per group. One-way ANOVA followed by Bonferroni correction for CD3⁺ and CD3⁺CD8⁺ cells, while the Kruskal-Wallis test was performed for others. *P < 0.05, **P < 0.01. B) Proportional changes of α4β7 integrin-positive CD45^high, CD3⁺, CD4⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the hemorrhagic brain 3 d after ICH. n = 5-6 mice per group. The Kruskal-Wallis test was performed. *P < 0.05. C) Percentage changes of CD103-positive CD4⁺ and CD3⁺CD4⁺ cells in the hemorrhagic brain 3 d post-ICH. n = 5-6 mice per group. The Kruskal-Wallis test was performed. *P < 0.05. D) Comparison of the proportions of CD3⁺, CD4⁺, and CD3⁺CD4⁺ cells to CD45⁺ cells in the peripheral blood 3 d after ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction was performed for multiple comparisons. *P < 0.05. E) The ratios of α4β7 integrin-positive CD45⁺, CD3⁺, CD4⁺, CD8⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the bloodstream 3 d after ICH. n = 6 mice per group. One-way ANOVA followed by Bonferroni correction for α4β7 integrin-positive CD45⁺ and CD3⁺ cells, while the Kruskal-Wallis test was performed for others. *P < 0.05. F) Analysis of the proportions of CD3⁺, CD4⁺, CD8⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells to CD45⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test for CD3⁺, CD8⁺, and CD3⁺CD8⁺ cells, while one-way ANOVA followed by Bonferroni correction was performed for others. *P < 0.05, **P < 0.01. G) Percentage changes of α4β7 integrin-positive CD45⁺ and CD3⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction for multiple comparisons was performed. *P < 0.05, **P < 0.01. H) Proportional changes of CD103-positive CD4⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. One-way ANOVA followed by Bonferroni correction for CD103-positive CD4⁺ and CD3⁺CD8⁺ cells, while the Kruskal-Wallis test was performed for CD103-positive CD3⁺CD4⁺ cells. *P < 0.05. I) Quantification of the percentages of CD3⁺ and CD3⁺CD4⁺ cells to CD45⁺ cells in the Peyer's patch 3 d after ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction was performed for multiple comparisons. No significant difference was found. J) Proportional changes of α4β7 integrin-positive CD45⁺, CD3⁺, CD4⁺, and CD3⁺CD4⁺ cells in the Peyer's patch 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test for α4β7 integrin-positive CD45⁺ cells and one-way ANOVA followed by Bonferroni correction was performed for others. *P < 0.05, **P < 0.01. K) The percentages of CD103-positive CD45⁺, CD3⁺, CD8⁺, and CD3⁺CD8⁺ cells in the Peyer's patch 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test was performed. *P < 0.05.

Figure 6

α7nAChR agonists have no influence on intestinal immunocyte homing and retention enhanced by β7 integrin antagonist etrolizumab in ICH. A) Evaluation for the influence of β7 integrin antagonist or β7 integrin antagonist with α7nAChR agonist treatment on the infiltration of CD3⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the lesioned brain 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test for CD3⁺ cells and one-way ANOVA followed by Bonferroni correction for others were performed. *P < 0.05, **P < 0.01. B) Proportional changes of α4β7 integrin-positive CD3⁺, CD4⁺, and CD3⁺CD4⁺ cells in the hemorrhagic brain 3 d after ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction for multiple comparisons was performed. *P < 0.05. C) Percentage changes of CD103-positive CD45^high, CD3⁺, CD4⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the hemorrhagic brain 3 d post-ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction for multiple comparisons of CD45^high and CD3⁺ cells, while the Kruskal-Wallis test for others was performed. *P < 0.05, **P < 0.01. D) Analysis for the proportions of CD3⁺, CD4⁺, CD8⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells to CD45⁺ cells in the bloodstream 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test was performed. **P < 0.01. E) The ratio of α4β7 integrin-positive CD45⁺, CD3⁺, CD4⁺, CD8⁺, CD3⁺CD4⁺, and CD3⁺CD8⁺ cells in the periphery 3 d post-ICH. n = 6 mice per group. The Kruskal-Wallis test was performed. *P < 0.05. F) Percentage change of CD103-positive CD45⁺ cells in the peripheral blood 3 d after ICH. n = 6 mice per group. A one-way ANOVA with Bonferroni correction for multiple comparisons was performed. *P < 0.05. G) Comparison of the proportions of CD3⁺, CD8⁺, and CD3⁺CD8⁺ cells to CD45⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. One-way ANOVA followed by Bonferroni correction was performed. *P < 0.05, **P < 0.01. H) Percentage changes of α4β7 integrin-positive CD45⁺, CD3⁺, CD4⁺, and CD3⁺CD4⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction for the proportions of CD45⁺ and CD3⁺CD4⁺ cells, while the Kruskal-Wallis test for others was performed. *P < 0.05, **P < 0.01. I) Proportional changes of CD103-positive CD45⁺, CD3⁺, CD8⁺, and CD3⁺CD8⁺ cells in the MLNs 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test for CD103-positive CD45⁺ cells and one-way ANOVA followed by Bonferroni correction for others were performed. *P < 0.05, **P < 0.01. J) Quantification of the ratios of CD3⁺, CD4⁺, CD8⁺, and CD3⁺CD4⁺ cells to CD45⁺ cells in the Peyer's patches 3 d after ICH. n = 6 mice per group. The Kruskal-Wallis test for CD3⁺, CD4⁺, and CD3⁺CD4⁺ cell percentages and one-way ANOVA with Bonferroni correction for multiple comparisons of CD8⁺ cell percentage were performed. *P < 0.05. K) The proportions of α4β7 integrin-positive CD45⁺, CD4⁺, and CD3⁺CD4⁺ cells in the Peyer's patches 3 d after ICH. n = 6 mice per group. A one-way ANOVA with Bonferroni correction for multiple comparisons of CD45⁺ and CD4⁺ cells, while the Kruskal-Wallis test for CD3⁺CD4⁺ cells was performed. *P < 0.05. L) The percentages of CD103-positive CD45⁺, CD3⁺, and CD3⁺CD8⁺ cells in the Peyer's patches 3 d post-ICH. n = 6 mice per group. One-way ANOVA with Bonferroni correction for multiple comparisons. No significant differences were found.

Figure 7

The β7 integrin antagonist etrolizumab enhances cellular and molecular inflammatory reactions in the lesioned brain after ICH. A) Immunostaining for GFAP, Iba-1, MPO, and FJB in the perihematomal region 3 d after ICH in mice treated with vehicle or β7 integrin antagonists. The insets show a higher magnification of immunofluorescence-stained cells, Scale bar = 100 μm. B) Analysis of microglia/macrophage and astrocytic activation, neutrophil infiltration, and neural degeneration 3 d after ICH. n = 6 mice per group. The two-tailed paired t-tests were used for GFAP-, IBA-1- and FJB-positive cells, and the Mann-Whitney U test for MPO-positive cells. **P < 0.01. C) Representative Western blot bands of proinflammatory factors HMGB1, IL-1β, IL-6, and TNF-α 3 d after ICH, and β-actin was used as the loading control. D) Analysis of the expression of HMGB1, IL-1β, IL-6, and TNF-α in the hemorrhagic brain 1 day after ICH in sham-operated and ICH mice treated with vehicle or β7 integrin antagonist. n = 5 mice per group. One-way ANOVA with Bonferroni correction for HMGB1 and IL-1β, and the Kruskal-Wallis test for IL-6 and TNF-α. *P < 0.05, **P < 0.01.

Figure 8

The β7 integrin antagonist etrolizumab exacerbates brain injury and functional deficits in mice with ICH. A) Representative brain sections stained with LFB/CV 3 d after ICH. The areas of the lesions lacking staining are circled with a black curve (red arrow indicated); scale bar = 500 μm. B) Brain injury volume 3 d after ICH in mice that received vehicle or β7 integrin antagonists. n = 10 mice per group. The two-tailed paired t-test was performed. **P < 0.01. C) Brain swelling 3 d after ICH in mice that received vehicle or β7 integrin antagonists. n = 10 mice per group. The two-tailed paired t-test was performed. **P < 0.01. D) Brain water content 3 d after ICH in sham and ICH mice that received vehicle or α7nAChR agonists. n = 6 mice per group. One-way analysis of variance followed by Bonferroni correction for multiple comparisons. *P < 0.05, **P < 0.01. E) Representative images of brain sections stained with LFB/CV 28 d after ICH. The areas of the lesions are circled with a black curve (the red arrow indicated). Scale bar = 500 μm. F) Analysis of residual lesion volume and brain atrophy in ICH mice received vehicle or β7 integrin antagonist. n = 10 mice per group. The two-tailed paired t-tests were performed. **P < 0.01. G) LFB-stained myelin in the perihematomal region of brain sections 28 d after ICH. Scale bar = 100 μm. H) Analysis of white matter damage in the perihematomal regions 28 d after ICH. n = 10 mice per group. The two-tailed paired t-test was performed. No significant difference was detected. I) Dynamic changes in neurologic deficit scores (NDS) throughout the 28-day research process in WT mice received vehicle or β7 integrin antagonists. n = 10 mice per group. Generalized estimation equations (GEE) with a two-tailed paired t-test were performed at multiple time points. Wald χ2 = 33.824, P < 0.001 for the total trend of the vehicle versus β7 integrin antagonists; all P < 0.01 on days 3, 7, 14, and 28 for the vehicle versus β7 integrin antagonists. J) Corner turn test (CTT) throughout the 28-day research process in WT mice received vehicle or β7 integrin antagonists. n = 10 mice per group. Generalized estimation equations (GEE) with the two-tailed paired t-test were performed at multiple time points. Wald χ2 = 23.064, P < 0.001 for the total trend of the vehicle versus β7 integrin antagonists; all P < 0.01 on days 7, 14, and 28 for the vehicle versus β7 integrin antagonists.