Gut Microbiota Dysbiosis in Acute Ischemic Stroke Associated With 3-Month Unfavorable Outcome

Abstract

Background: Alterations in the gut microbiota after ischemic stroke have been demonstrated, whereas the effect on stroke outcome remains to be established.

Methods: A total of 132 consecutive patients with acute ischemic stroke were prospectively enrolled. Their gut microbiomes within 24 h of admission were profiled using 16S ribosomal RNA (rRNA) gene (V3-V4 region) sequencing. Microbiota comparisons were made between groups with good outcome (n = 105) and poor outcome (n = 27) based on 3-month modified Rankin Scale scores of 0-2 and 3-6. Propensity score-matching (PSM) analysis was conducted to assess the robustness of our findings. The functional potential was predicted using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt).

Results: Patients in the poor outcome group were characterized by a significant reduction in the alpha diversity (Shannon index, p = 0.025; Simpson index, p = 0.010), an increase in the pathogenic bacteria (e.g., Enterococcaceae and Enterococcus), and a decrease in the short-chain fatty acids (SCFAs)-producing bacteria (e.g., Bacteroidaceae, Ruminococcaceae, and Faecalibacterium) to those with good outcome group (all p < 0.05). Similar results of microbial composition were obtained after PSM. The PICRUSt revealed that the pathway for membrane transport was relatively dominant in patients with poor outcome (p < 0.05).

Conclusion: This study demonstrated that stroke patients with 3-month poor outcome had baseline gut microbiota dysbiosis featured by increased pathogenic bacteria and decreased SCFAs-producing bacteria.

Keywords: 16S rRNA; dysbiosis; functional outcome; gut microbiota; ischemic stroke.

Figure 1

Comparison of alpha- and beta-diversities of gut microbiota between the two groups. (A) Six indices were used to represent the alpha-diversity: observed species, Chao1, Abundance-based Coverage Estimator (ACE), Shannon, Simpson and Coverage. (B) Principal coordinates analysis (PCoA) of the weighted uniFrac distances for the mRS 0–2 and mRS 3–6 subjects. The two components explain 26.6 and 18.5% of the variance, respectively. A significant separation was found between the two groups (p < 0.001, permutation test with pseudo-F ratios). Blue, mRS 0–2 subjects (n = 105); red, mRS 3–6 subjects (n = 27). mRS, modified Rankin Scale.

Figure 2

Taxa with significantly different relative abundance between two groups at the genus level. *p < 0.05, **p < 0.01. mRS, modified Rankin Scale.

Figure 3

Significantly discriminant taxa between the good outcome and poor outcome patients determined using linear discriminate analysis size effect (LEfSe). LEfSe, linear discriminate analysis size effect; LDA, linear discriminate analysis; mRS, modified Rankin Scale.

Figure 4

Correlation between the gut microbiota and baseline characteristics. (A) Heatmap of Spearman's correlation analysis between bacteria genera and clinical characteristic parameters. (B) Heatmap of Spearman's correlation analysis between bacteria phyla and laboratory parameters. *p < 0.05, **p < 0.01. FBG, fasting blood glucose; TG, triglyceride; NIHSS, National Institute of Health Stroke Scale; DWI-ASPECTs, DWI-Alberta Stroke Program Early CT score; IL, interleukin; TMAO, trimethylamine N-Oxide; BDNF, brain-derived neurotropic factor.

Figure 5

The gut microbiota function differed between the two groups in KEGG pathways and was correlated with bacteria profile. (A) Significantly differed microbiota function in KEGG level 2 pathways. (B) Heatmap of Spearman's correlation analysis between bacteria genera and the KEGG level 2 pathways. KEGG, kyoto encyclopedia of genes and genomes; mRS, modified Rankin Scale.