Identification of pivotal genes and regulatory networks associated with SAH based on multi-omics analysis and machine learning

Abstract

Subarachnoid hemorrhage (SAH) is a disease with high mortality and morbidity, and its pathophysiology is complex but poorly understood. To investigate the potential therapeutic targets post-SAH, the SAH-related feature genes were screened by the combined analysis of transcriptomics and metabolomics of rat cortical tissues following SAH and proteomics of cerebrospinal fluid from SAH patients, as well as WGCNA and machine learning. The competitive endogenous RNAs (ceRNAs) and transcription factors (TFs) regulatory networks of the feature genes were constructed and further validated by molecular biology experiments. A total of 1336 differentially expressed proteins were identified, including 729 proteins downregulated and 607 proteins upregulated. The immune microenvironment changed after SAH and the changement persisted at SAH 7d. Through multi-omics and bioinformatics techniques, five SAH-related feature genes (A2M, GFAP, GLIPR2, GPNMB, and LCN2) were identified, closely related to the immune microenvironment. In addition, ceRNAs and TFs regulatory networks of the feature genes were constructed. The increased expression levels of A2M and GLIPR2 following SAH were verified, and co-localization of A2M with intravascular microthrombus was demonstrated. Multiomics and bioinformatics tools were used to predict the SAH associated feature genes confirmed further through the ceRNAs and TFs regulatory network development. These molecules might play a key role in SAH and may serve as potential biological markers and provide clues for exploring therapeutic options.

Keywords: A2M; Microthrombus; Multi-omics; Subarachnoid hemorrhage; WGCNA.

Fig. 1

Construction of WGCNA. (A) The clustering dendrogram of transcriptomics, each leaf corresponds to a different gene module. (B) Heatmap of correlations between modules and traits, each cell containing the corresponding correlation and p value. (C) Heatmap of metabolite module-gene module correlation, with gene module on the right and metabolite module on the bottom. (D) Venn map is used to identify the intersection between DEGs and WGCNA related genes.

Fig. 2

Functional enrichment of 73 DEGs. (A) Significant enriched GO terms for DEGs. (B) The significantly enriched KEGG pathways. (C) DO analysis of DEGs. (D) Five active pathways of sham group in GSEA analysis. (F) Five active pathways of SAH 1d group in GSEA analysis.

Fig. 3

Machine learning screening for relevant genes. (A) Lasso regression analysis cross-validation curve. When 8 genes are used in the analysis, Lasso fits best and cross-validation error is minimized. (B) LASSO coefficient path diagram, each curve represents one gene. (B)Expression of 8 related genes in transcriptomics.

Fig. 4

Correlation between related genes and DEGs. (A) Venn map is used to identify the intersection between DEMs and WGCNA related genes. (B, D) Kegg enrichment analysis of 267 DEMs. (C, E) Correlation of Lasso-related genes with DEMs. (*p < 0.05, **p < 0.01, ***p < 0.001, n = 6)

Fig. 5

Proteomic analysis. (A) 3D PCA distribution of all samples. (B) Venn diagram showing proteins identified in different groups. (C) Differential histogram of protein quantification in various groups. (D) Heatmap of clustering analysis of differentially expressed proteins in different groups. (E) Venn diagram demonstrating the intersection of proteins with Lasso-associated genes. (F) Characterized genes in proteomics expression.

Fig. 6

Infiltration analysis of immune cells. (A) Relative abundance of 22 infiltrating immune cells between sham and SAH 1d samples. (B) Violin plot of all 22 immune cell differentially infiltrated fractions between sham and SAH 1d samples. (C) Relative abundance of 22 infiltrating immune cells between SAH 1d and SAH 7d samples. (D) Violin plot of all 22 immune cell differentially infiltrated fractions between SAH 1d and SAH 7d samples. (*p < 0.05, **p < 0.01, ***p < 0.001, n = 6)

Fig. 7

Characterized genes and immune cells correlation analysis between sham and SAH 1d samples. Lollipop charts of the correlation of LCN2 (A), A2M (B), GLIPR2 (C), GPNMB (D), GFAP (E), respectively, with 22 immune cell types.

Fig. 8

CeRNA of characterized genes. The ceRNA regulatory networks of A2M (A), LCN2 (B), GLIPR2 (C), GPNMB (D) and GFAP (E), separately. Red represents characterized genes, green represents miRNAs, and blue represents lncRNAs.

Fig. 9

TFs regulatory networks of characterized genes. TFs regulatory networks of A2M (A), LCN2 (B), GLIPR2 (C), GPNMB (D) and GFAP (E), respectively.

Fig. 10

A2M and GLIPR2 are highly expressed after SAH. (A, C ) Western blot images and quantitative analyses of A2M and GLIPR2 in rat cortical tissues. (B) Western blot images of A2M and GLIPR2 in the cerebrospinal fluid of patients. (*p < 0.05, **p < 0.01, ***p < 0.001)

Fig. 11

Relationship between A2M and microthrombosis. (A, B) Typical micrographs showing immunofluorescence staining of A2M with blood vessels and microthrombi in different experimental groups. n = 4 per group. Scale bar = 50 μm. (C) Quantitative analysis of A2M in different experimental groups. (D) Number of microthrombi in different experimental groups. (E, F, G) Immunodouble-labelling co-localisation curve analysis and Pearson’s R-value for A2M with microvessels and microthrombi. (*p < 0.05, **p < 0.01, ***p < 0.001)