CLEC7A as a potential biomarker of atherosclerotic response to aspirin therapy: a multi-omics analysis

Figures

• 

  Figure 1. Removal of batch effects and identification of DEGs in the AR and ASCVD datasets. (A, B) DEG heatmap and volcano plot in the AR group. (C, D) PCA plots show the expression patterns in the two ASCVD datasets before and after batch-effect removal. (E, F) DEG heatmap and volcano plot in the ASCVD group. AR, Aspirin resistant; ASCVD, Atherosclerotic cardiovascular disease; DEG, differentially expressed gene; PCA, Principal-component analysis.

• 

  Figure 2. WGCNA of the AR and ASCVD datasets. (A) Determination of soft-threshold power for AR. (B) Cluster dendrogram of AR highly connected genes in key modules. (C) Relationships between modules and traits in AR. Correlations and P values are included in each cell. (D) Calculation of soft-threshold power for ASCVD. (E) Cluster dendrogram of ASCVD modules with highly connected genes. (F) Module–trait relationships in ASCVD. A correlation and P value are included in each cell. AR, Aspirin resistant; ASCVD, Atherosclerotic cardiovascular disease; WGCNA, Weighted gene co-expression network analysis.

• 

  Figure 3. Shared gene signatures and functional enrichment between AR and ASCVD. (A) The shared DEGs between AR and ASCVD are obtained by overlapping their DEGs of them. (B) The shared genes between the WGCNA modules for AR and ASCV were identified by overlapping the modules. (C, D) Shared genes were represented by bar plots displaying GO and KEGG enrichment. AR, Aspirin resistant; ASCVD, Atherosclerotic cardiovascular disease; DEG, differentially expressed gene; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; WGCNA, Weighted gene co-expression network analysis.

• 

  Figure 4. PPI network and Hub genes analysis. (A) The first module of the PPI network. (B–E) Four algorithms (MNC, MCC, Degree, EPC) from the cytoHubba plugin were used to identify the top 20 hub genes. (F) Seventeen common hub genes were identified by four algorithms in the cytoHubba plugin. (G) Gene networks were further constructed using GeneMANIA. MNC, Neighborhood Component Centrality; MCC, Maximum Clique Centrality; EPC, Edge Percolated Component; PPI, Protein–protein interaction.

• 

  Figure 5. Identification of the key genes of aspirin non-responsiveness. (A, B) Six key genes were identified from the hub genes by machine learning LASSO regression. (C, D) Ten key genes were also identified from the hub genes by the Random Forest (RF) algorithm. (E, F) Two key genes were also identified from the hub genes by the SVM-RFE algorithm. (G) Two key genes were identified by overlapping CLEC7A and RGS1. (H, I) ROC curves of CLEC7A and RGS1 in the training group for AR and ASCVD. (J, K) Differential expression of CLEC7A and RGS1 in the training group for AR and ASCVD. AR, Aspirin resistant; ASCVD, Atherosclerotic cardiovascular disease. Data were shown as mean ± SD (ns P > 0.05, ****P > 0.0001).

• 

  Figure 6. AR and ASCVD immune cell composition. (A) A violin diagram indicating that the AR group exhibited a significantly different immune cell composition. (B) Correlation between CLEC7A expression and immune cells in the AR group. (C) The violin diagram of the ASCVD group shows an evident difference in immune cell types. (D) Relationship between CLEC7A expression and immune cells in the ASCVD group. AR, Aspirin resistant; ASCVD, Atherosclerotic cardiovascular disease.

• 

  Figure 7. The single-gene GSEA and molecular regulatory networks of CLEC7A. (A, B) GSEA analysis of CLEC7A. (C) The alluvial diagram of the mRNA–miRNA–lncRNA network. (D) Circular regulatory signaling pathway. GSEA, Gene set enrichment analysis.

• 

  Figure 8. Validation of CLEC7A expression in mice. (A) Oily red staining of aortic tissue from mice, n = 3, bar = 100 μm. (B, C) Expression levels of CLEC7A in aortic tissue and serum from mice, n = 8. (D) Dual immunofluorescence staining of the T cell marker Foxp3 and neutrophil marker Ly6g with CLEC7A in mouse heart aortic valve tissue. (E) Immunofluorescence quantitative analysis, n = 3, bar = 100 μm. Data were shown as mean ± SD (nsP>0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

• 

  Figure 9. Virtual screening and CLEC7A-molecule interaction analysis. (A) Specific information on small molecules with DrugCLIP score >3 and Vina docking score <–5. (B–E) The docking conformations of the top four small molecules (Z394411138, Z28028664, Z28130510, Z28143312) with the CLEC7A protein.