Cross-Species Identification and Validation of Hub Genes and Potential Therapeutic Targets in Myocardial Infarction

Figures

• 

  Figure 1. Identification of differentially expressed genes (DEGs) in myocardial infarction models across species. (A–D) Volcano plots showing DEGs in four GEO datasets, including human cardiac organoid models (GSE115031 and GSE113871) and mouse myocardial infarction models (GSE46395 and GSE223208). The numbers of upregulated and downregulated genes in each dataset are indicated. The x-axis represents log2 fold change (log2FC), and the y-axis represents −log10 (p value). Dashed lines indicate the thresholds for differential expression. (E) Venn diagram showing the overlap of DEGs between the two human cardiac organoid datasets (GSE115031 and GSE113871), identifying 465 shared DEGs. (F) Venn diagram showing the overlap of DEGs between the two mouse myocardial infarction datasets (GSE223208 and GSE46395), identifying 1051 shared DEGs. DEGs were defined as the union of upregulated and downregulated genes identified in each dataset (|log2FC| ≥ 0.5, adjusted p value ≤ 0.05). Shared DEGs represent the intersection of DEGs within each species and were subsequently used for cross-species homolog mapping and downstream integrative analyses.

• 

  Figure 2. Cross-species integration and functional enrichment analysis of shared genes. (A) Venn diagram showing the overlap between human and mouse datasets. Mouse DEGs obtained from the intersection of two myocardial infarction models were first converted to their human homologs and subsequently intersected with the shared DEGs from two human cardiac organoid datasets, yielding 49 common genes. (B) Gene Ontology (GO) biological process enrichment analysis of the 49 shared genes. The top enriched terms include wound healing, muscle contraction, blood coagulation, response to decreased oxygen levels, and plasminogen activation. The x-axis represents gene ratio, and dot size indicates gene count, while color represents adjusted p value. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the 49 shared genes. Significantly enriched pathways include cytoskeleton in muscle cells, proteoglycans in cancer, complement and coagulation cascades, HIF-1 signaling pathway, apelin signaling pathway, and fluid shear stress and atherosclerosis. Dot size indicates gene count, and color represents adjusted p value.

• 

  Figure 3. Protein–protein interaction (PPI) network construction and hub gene identification. (A) PPI network of the shared genes. Among the 49 cross-species common genes, two housekeeping genes (ACTB and ACTG1) were excluded, and the remaining genes were subjected to PPI network construction using the STRING database. Nodes represent proteins, and edges represent protein–protein interactions. Node size reflects the degree of connectivity, with highly connected genes shown as larger nodes. Node color indicates connectivity or ranking based on network topology analysis. (B) Top 10 hub genes ranked by degree centrality in the PPI network. The degree values indicate the number of interactions for each gene, with IL6, SERPINE1, and MMP14 among the most highly connected genes.

• 

  Figure 4. Identification of hub genes using multiple topological algorithms. (A) UpSet plot showing the intersection of hub genes identified from the PPI network using nine topological analysis methods in the cytoHubba plugin of Cytoscape, including MCC, MNC, Degree, EPC, Closeness, Betweenness, BottleNeck, Stress, and Radiality. The top 10 genes from each algorithm were selected, and five overlapping genes (IL6, SERPINE1, MMP14, PLAUR, and ENO1) were identified as final hub genes. Asterisks indicate intersecting genes shared across multiple topological methods and do not represent statistical significance. Red bars highlight the final intersection gene set. (B) Ranking of the top 10 hub genes based on maximal clique centrality (MCC) scores. IL6, SERPINE1, and MMP14 were among the highest-ranked genes, indicating their central roles in the network.

• 

  Figure 5. Expression patterns of hub genes in human and murine myocardial infarction models. (A) Heatmap showing the expression patterns of the five hub genes (IL6, SERPINE1, MMP14, PLAUR, and ENO1) in human cardiac organoid samples from dataset GSE115031 under control and myocardial infarction (MI) conditions. (B) Violin plots illustrating the normalized expression levels of the five hub genes in human cardiac organoid samples (GSE115031), comparing control and MI groups. (C) Heatmap showing the expression patterns of the five hub genes in murine heart tissue samples from dataset GSE46395 under control and MI conditions. (D) Violin plots illustrating the normalized expression levels of the five hub genes in murine heart tissue samples (GSE46395), comparing control and MI groups. Statistical significance is indicated as * p < 0.05, ** p < 0.01, *** p < 0.001.

• 

  Figure 6. Independent validation of hub genes across multiple datasets and species. (A) Violin plots showing the normalized expression levels of the five hub genes (IL6, SERPINE1, MMP14, PLAUR, and ENO1) in human peripheral blood samples under control and myocardial infarction (MI) conditions. (B) Expression levels of the five hub genes in human engineered heart tissue models comparing control and MI groups. (C) Expression patterns of the five hub genes in porcine (scrofa) myocardial infarction models. (D) Expression patterns of the five hub genes in rat myocardial infarction models. (E) Expression patterns of homologous genes in zebrafish myocardial infarction models. (F) Expression patterns of homologous genes in medaka myocardial infarction models. Statistical significance is indicated as * p < 0.05, ** p < 0.01, *** p < 0.001; ns, not significant.

• 

  Figure 7. Integrated analysis of diagnostic performance and potential therapeutic targets of hub genes. (A) Drug–gene interaction network showing potential therapeutic compounds targeting the five hub genes (IL6, SERPINE1, MMP14, PLAUR, and ENO1), based on the DGIdb database. Nodes represent genes and drugs, and edges indicate known or predicted interactions. Drug approval status is indicated in the figure. (B) Receiver operating characteristic (ROC) curves of individual hub genes and the combined model in an independent peripheral blood dataset. The combined model demonstrated superior diagnostic performance (AUC = 0.904) compared with individual genes. The grey diagonal line represents the reference line (AUC = 0.5), and different colors indicate individual genes and the combined model.