Identification of Biomarkers Related to Immune Cell Infiltration with Gene Coexpression Network in Myocardial Infarction

Abstract

Background: There is evidence that the immune system plays a key critical role in the pathogenesis of myocardial infarction (MI). However, the exact mechanisms associated with immunity have not been systematically uncovered.

Methods: This study used the weighted gene coexpression network analysis (WGCNA) and the CIBERSORT algorithm to analyze the MI expression data from the Gene Expression Omnibus database and then identify the module associated with immune cell infiltration. In addition, we built the coexpression network and protein-protein interactions network analysis to identify the hub genes. Furthermore, the relationship between hub genes and NK cell resting was validated by using another dataset GSE123342. Finally, receiver operating characteristic (ROC) curve analyses were used to assess the diagnostic value of verified hub genes.

Results: Monocytes and neutrophils were markedly increased, and T cell CD8, T cell CD4 naive, T cell CD4 memory resting, and NK cell resting were significantly decreased in MI groups compared with stable coronary artery disease (CAD) groups. The WGCNA results showed that the pink model had the highest correlation with the NK cell resting infiltration level. We identified 11 hub genes whose expression correlated to the NK cell resting infiltration level, among which, 7 hub genes (NKG7, TBX21, PRF1, CD247, KLRD1, FASLG, and EOMES) were successfully validated in GSE123342. And these 7 genes had diagnostic value to distinguish MI and stable CAD.

Conclusions: NKG7, TBX21, PRF1, CD247, KLRD1, FASLG, and EOMES may be a diagnostic biomarker and therapeutic target associated with NK cell resting infiltration in MI.

Figure 1

(a) The expression matrix from 199 samples in the training dataset. (b) The landscape of tumor-infiltrating immune cells. The difference of the proportions of tumor-infiltrating immune cells between MI and stable CAD sample.

Figure 2

WGCNA revealed gene coexpression networks. (a) Clustering dendrogram of 139 samples corresponding to clinical characteristics. (b) Analysis of the scale-free fit index and mean connectivity for various soft-thresholding powers (β).

Figure 3

Identification of gene modules associated with the immune cell infiltration of MI. (a) The horizontal line defines the threshold, so 15 distinct genes modules were identified. (b) The dendrogram of all genes is clustered based on a dissimilarity measure. (c) The heatmap shows the correlation between MEs and the immune cell infiltration of MI. (d) The scatter plot shows the correlation between gene significance for MI and module membership in pink module.

Figure 4

Key modules and identification of hub genes. (a) The first 20 enriched terms are shown as a bar chart on the left. The network diagram on the right is constructed with each enrichment term as a node and the similarity of the node as the edge. Nodes with the same cluster ID are the same color. (b) PPI network of genes from the pink module. The higher the number of connected nodes, the larger the size of the node. (c) Hub genes were selected based on overlap between PPI and coexpression networks.

Figure 5

Validation of hub genes. (a) NK cell resting infiltration level between MI and stable CAD. (b) Relationship between 11 hub genes expression and NK cell resting infiltration level. P < 0.05 is considered statistically significant. (c) Scatter plot of NKG7 expression and NK cell resting infiltration level. (d) Hierarchical clustering analysis of 7 verified genes.

Figure 6

ROC analysis of 7 verified genes. The AUC was analyzed to evaluate the performance of each hub genes. x-axis indicated 1-specificity, and y-axis indicated sensitivity.