Introduction: A keloid is a type of benign fibrotic disease with similar features to malignancies, including anti-apoptosis, over-proliferation, and invasion. Epithelial-mesenchymal transition (EMT) is a crucial mechanism that regulates the metastatic behavior of tumors. Thus, identifying EMT biomarkers is paramount in comprehensively understanding keloid pathogenesis.
Methods: To identify the differentially expressed genes (DEGs) GSE92566 dataset, with 3 normal skin and 4 keloid tissues, was downloaded from GEO databases to identify the differentially expressed genes (DEGs). Further, EMT-related genes were downloaded from dbEMT 2.0 databases and intersected with GSE92566 DEGs to identify EMT-related-DEGs (ERDEGs). Subsequently, the ERDEGs were used for GO, KEGG, gene set enrichment analysis (GSEA), protein-protein interaction (PPI), and miRNAs-mRNAs network analysis. To predict small molecules for EMT inhibition, the ERDEGs were imported to cMAP databases, whereas hub genes were imported to DGidb databases. Finally, we carried out qRT-PCR and in vitro experiments to validate our findings.
Results: A total of 122 ERDEGs were identified, including 59 upregulated and 63 down-regulated genes. Moreover, enrichment analysis revealed that focal adhesion, AMPK signal pathway, Wnt signal pathway, and EMT biological process were significantly enriched. STRING databases and Cytoscape software were used to construct the PPI network and EMT-related hub genes. Further, 3 modules were explored from the PPI network using the Molecular Complex Detection (MCODE) plugin. In the Cytohubba plugin, 10 hub genes were explored, including FN1, EGF, SOX9, CDH2, PROM1, EPCAM, KRT19, ITGB1, CD24, and KRT18. These genes were then enriched for the focal adhesion pathway. We constructed a microRNA (miRNA)-mRNA network, which predicted hsa-miR-155-5p (8 edges), hsa-miR-124-3p (7 edges), hsa-miR-145-5p (5 edges), hsa-miR-20a-5p (5 edges) and hsa-let-7b-5p (4 edges) as the most connected miRNAs regulating EMT. Based on the ERDEGs and 10 hub genes mentioned above, ribavirin demonstrated high drug-targeting relevance. Subsequently, qRT-PCR confirmed that the expression of FN1, ITGB1, CDH2, and EPCAM corroborated with previous findings. qRT-PCR also showed that the expression levels of hsa-miR-124-3p and hsa-miR-145-5p were significantly lower in keloids and hsa-miR-155-5p was upregulated in keloids. Finally, by treating human keloid fibroblasts (HKFs) with ribavirin in vitro, we confirmed that ribavirin could inhibit HKFs proliferation and EMT.
Conclusion: In summary, this work provides novel EMT biomarkers in keloids and predicts new small target molecules for keloid therapy. Our findings improve the understanding of keloid pathogenesis, providing new treatment options.
Keywords: Bioinformatic; Epithelial-mesenchymal transition; Fibrosis; Keloids.
Copyright © 2023. Published by Elsevier Ltd.