Identification of pyroptosis-related immune signature and drugs for ischemic stroke

Abstract

Background: Ischemic stroke (IS) is a common and serious neurological disease, and multiple pathways of cell apoptosis are implicated in its pathogenesis. Recently, extensive studies have indicated that pyroptosis is involved in various diseases, especially cerebrovascular diseases. However, the exact mechanism of interaction between pyroptosis and IS is scarcely understood. Thus, we aimed to investigate the impact of pyroptosis on IS-mediated systemic inflammation. Methods: First, the RNA regulation patterns mediated by 33 pyroptosis-related genes identified in 20 IS samples and 20 matched-control samples were systematically evaluated. Second, a series of bioinformatics algorithms were used to investigate the contribution of PRGs to IS pathogenesis. We determined three composition classifiers of PRGs which potentially distinguished healthy samples from IS samples according to the risk score using single-variable logistic regression, LASSO-Cox regression, and multivariable logistic regression analyses. Third, 20 IS patients were classified by unsupervised consistent cluster analysis in relation to pyroptosis. The association between pyroptosis and systemic inflammation characteristics was explored, which was inclusive of immune reaction gene sets, infiltrating immunocytes and human leukocyte antigen genes. Results: We identified that AIM2, SCAF11, and TNF can regulate immuno-inflammatory responses after strokes via the production of inflammatory factors and activation of the immune cells. Meanwhile, we identified distinct expression patterns mediated by pyroptosis and revealed their immune characteristics, differentially expressed genes, signaling pathways, and target drugs. Conclusion: Our findings lay a foundation for further research on pyroptosis and IS systemic inflammation, to improve IS prognosis and its responses to immunotherapy.

Keywords: LASSO-cox regression; RNA modification; immunity; ischemic stroke; pyroptosis; systemic inflammation.

FIGURE 1

Identification of DEGs between healthy and IS samples. (A) Distribution of 33 PRGs on chromosomes. (B) Thirty-three PRG protein–protein interactions are presented. (C) Correlations among the expression of pyroptosis genes in all healthy and IS samples. (D) Volcano plot shows the summary of expression-changing information of 33 PRGs between healthy and IS samples. (E) Heatmap plot demonstrates the transcriptome expression status of 33 PRGs between healthy and IS samples. Blue: low expression level; red: high expression level. (F) Box plot demonstrates the transcriptome expression status of 33 PRGs between healthy and IS samples.

FIGURE 2

Pyroptosis genes can well distinguish healthy and IS samples. (A) Univariate logistic regression investigated the relationship between PRGs and IS. (B) Least absolute shrinkage and selection operator (LASSO) coefficient profiles of 33 PRGs. (C) Ten- fold cross-validation for tuning parameter selection in the LASSO regression. The partial likelihood of deviance is plotted against log(λ), where λ is the tuning parameter. Partial likelihood deviance values are shown, with error bars representing SE. The dotted vertical lines are drawn at the optimal values by minimum criteria and 1-SE criteria. (D) Distinguishing signature with three PRGs was developed by multivariate logistic regression, and the risk scores for IS were calculated. (E) Risk distribution between healthy and IS samples, where IS samples has a much higher risk score than healthy samples. (F) Discrimination ability for healthy and IS samples by PRGs was analyzed using the ROC curve and evaluated by the AUC value.

FIGURE 3

Correlation between infiltrating immunocytes and pyroptosis genes. (A) Difference in the abundance of each infiltrating immunocyte between healthy and IS samples. (B) Dot-plot demonstrated the correlations between each infiltrating immunocyte type and each PRG.

FIGURE 4

Correlation between immune reaction gene sets and pyroptosis genes. (A) Difference in the activity of each immune reaction gene set between healthy and IS samples. (B) Dot-plot demonstrated the correlations between each dysregulated immune reaction gene set and each PRG.

FIGURE 5

Identification of different pyroptosis expression patterns. (A) Heatmap of the matrix of co-occurrence proportions for IS samples. Blue: low-expression level; red: high-expression level. (B) Principal component analysis for the transcriptome profiles of three pyroptosis subtypes, showing a remarkable difference in the transcriptome between different modification patterns. (C) Expression status of 33 PRGs in the three pyroptosis subtypes. (D) Unsupervised clustering of 33 PRGs in the three patterns (E) Comparison of age, sex, smoking, alcohol consumption, hypertension, hypercholesterolemia, and diabetes and IS type among three pyroptosis regulation patterns. The heatmap illustrates the association of different clinical characteristics with the three subtypes.

FIGURE 6

Diversity of underlying biofunctional characteristics among three ischemic stroke subtypes. (A), (B), and (C) Upregulated gene enrichment among the three subtypes, pairwise comparisons were made for GO-biological processes, and potential drug targets predicted by DGIdb, respectively.

FIGURE 7

Biological characteristics and potential drugs of three PRG modification patterns. (A) Forty-two common genes were identified as genes associated with the pyroptosis phenotype. (B) Barplot graph for GO enrichment (the longer bar means the more genes enriched, and the increasing depth of red means the differences were more obvious). (C) Bubble graph for KEGG pathways (the bigger bubble means the more genes enriched, and the increasing depth of red means the differences were more obvious; q-value: adjusted p-value). (D) Sankey diagram showing 42 genes predicted by DGIdb for potential drugs.