Comprehensive analysis of immune-related biomarkers and pathways in intracerebral hemorrhage using weighted gene co-expression network analysis and competing endogenous ribonucleic acid

Abstract

The immune response is an important part of secondary brain injury following intracerebral hemorrhage (ICH), and is related to neurological deficits and prognosis. The mechanisms underlying the immune response and inflammation are of great significance for brain injury and potential functional restoration; however, the immune-related biomarkers and competing endogenous ribonucleic acid (RNA) (ceRNA) networks in the peripheral blood of ICH patients have not yet been constructed. We collected the peripheral blood from ICH patients and controls to assess their ceRNA profiles using LCHuman ceRNA microarray, and to verify their expression with qRT-PCR. Two-hundred-eleven DElncRNAs and one-hundred-one DEmRNAs were detected in the ceRNA microarray of ICH patients. The results of functional enrichment analysis showed that the immune response was an important part of the pathological process of ICH. Twelve lncRNAs, ten miRNAs, and seven mRNAs were present in our constructed immune-related ceRNA network, combining weighted gene co-expression network analysis (WGCNA). Our study was the first to establish the network of the immune-related ceRNAs derived from WGCNA, and to identify leukemia inhibitory factor (LIF) and B cell lymphoma 2-like 13 (BCL2L13) as pivotal immune-related biomarkers in the peripheral blood of ICH patients, which are likely associated with PI3K-Akt, the MAPK signaling pathway, and oxidative phosphorylation. The MOXD2P-miR-211-3p -LIF and LINC00299-miR-198-BCL2L13 axes were indicated to participate in the immune regulatory mechanism of ICH. The goal of our study was to offer innovative insights into the underlying immune regulatory mechanism and to identify possible immune intervention targets for ICH.

Keywords: ICH; WGCNA; ceRNA; immune-related; lncRNA.

FIGURE 1

The flow chart of this study. The peripheral blood from ICH patients and control subjects were collected for LCHuman ceRNA microarray and verification using qRT-PCR. The DElncRNAs and DEmRNAs were identified and GO, KEGG, PPI, and GSEA were performed. Then, we got the highest related lncRNAs to ICH by taking the intersection of strongest correlated lncRNAs module by WGCNA, DElncRNAs from our ceRNA microarray, and predicted lncRNAs. We took the intersection of DEmRNAs and IRGs to obtain immune-related DEmRNAs. TargetScan and miRanda were used to predict the interactions between three types of RNAs, and the immune-related ceRNA network was constructed. In addition, we performed GO and KEGG on the immune-related DEmRNAs, verified their expression by the GSE24265 and the GSE125512 from GEO datasets, and conducted GSEA to identify potential pathways of the immune-related DEmRNAs. Finally, we constructed the immune-related ceRNA sub-network.

FIGURE 2

The DElncRNAs and DEmRNAs in the ICH group compared with the control group. (A) The volcano map of DElncRNAs in the ICH group compared with the control group. (B) The volcano map of DEmRNAs in the ICH group compared with the control group. (C) The heat map of the top 100 DElncRNAs in the ICH group compared with the control group. (D) The heat map of DEmRNAs in the ICH group compared with the control group. Red represented up-regulation and high expression, while blue represented down-regulation and low expression.

FIGURE 3

Functional enrichment analysis and PPI network of DEmRNAs in ICH. (A) The top-level Gene Ontology biological processes of DEmRNAs in ICH. (B) PPI network of DEmRNAs in ICH group and MCODE component in gene table. (C) The main protein interaction domains extracted from the protein network.

FIGURE 4

The representative pathways with significant differences of mRNAs in ICH based on GSEA. (A) JAK-STAT signaling pathway. (B) Intestinal immune network pathway. (C) NOD-like receptor signaling pathway. (D) Toll-like receptor signaling pathway. (E) Wnt signaling pathway. (F) MAPK signaling pathway.

FIGURE 5

Identification and functional enrichment analysis of the immune-related DEmRNAs in ICH. (A) The venn diagram showed the intersection genes of DEmRNAs and IRGs in ICH. (B) The heatmap and density plot of the 17 immune-related DEmRNAs in ICH. (C) The top 10 Gene Ontology biological processes of immune-related DEmRNAs in ICH. (D) The top 10 KEGG enrichment analysis of immune-related DEmRNAs in ICH. (E) The heatmap of KEGG pathways of immune-related DEmRNAs in ICH.

FIGURE 6

The expression of ten DElncRNAs in peripheral blood of ICH patients compared with control subjects using qRT-PCR. (A) The qRT-PCR results of XIST. (B) The qRT-PCR results of FAM182B. (C) The qRT-PCR results of LINC00472. (D) The qRT-PCR results of LOC101927210. (E) The qRT-PCR results of LINC02731. (F) The qRT-PCR results of PCBP1-AS1. (G) The qRT-PCR results of LINC00174. (H) The qRT-PCR results of LOC100131626. (I) The qRT-PCR results of HOXB-AS3. (J) The qRT-PCR results of LINC00299. *P < 0.05, **P < 0.01, ***P < 0.001.

FIGURE 7

Analysis of lncRNAs modules using WGCNA and the construction of lncRNAs associated ceRNA network in ICH. (A) Module trait relationships between ICH patients and control subjects. Red represented positive correlation, blue represented negative correlation, and darker colors represented higher correlation. (B) The venn diagram showed the intersection of the green lncRNAs module by WGCNA, the DElncRNAs from our microarray, and the predicted lncRNAs. (C) The constructed ceRNA network of lncRNAs from the intersection, the predicted miRNAs, and the DEmRNAs from our microarray was visualized by Cytoscape. Red diamond represented lncRNAs, green triangle represented miRNAs, and blue ellipse represented mRNAs. (D) The Sankey diagram showed the immune-related lncRNA-miRNA-mRNA ceRNA network in ICH.

FIGURE 8

Identification of key biomarkers and pathways of the immune-related DEmRNAs in the peripheral blood and the brain tissue from ICH patients. (A) The venn diagram showed LIF was at the intersection of the IRGs from immune database, the DEmRNAs in peripheral blood from our microarray, and the DEmRNAs in brain tissue from the GSE24265 dataset of ICH patients. (B) The expression of LIF in the peripheral blood from our ceRNA microarray. (C) The expression of LIF in the brain tissue from the GSE24265 dataset of ICH patients. (D) The venn diagram showed BCL2L13 was at the intersection of the IRGs from immune database, the DEmRNAs in the peripheral blood from our microarray, and the DEmRNAs in the peripheral blood from the GSE125512 dataset of ICH patients. (E) The expression of BCL2L13 in the peripheral blood from our ceRNA microarray. (F) The expression of BCL2L13 in the peripheral blood from the GSE125512 dataset of ICH patients. (G) The five representative results of GSEA for LIF. (H) The five representative results of GSEA for BCL2L13. (I) The immune-related ceRNA sub-network of LIF and BCL2L13, including MOXD2P-miR-211-3p-LIF and LINC00299-miR-198-BCL2L13. *P < 0.05, **P < 0.01.