Molecular Functions and Drug Prediction of MMP1 and TGFBR2 in Nasopharyngeal Carcinoma

Nasopharyngeal carcinoma (NPC) is partially driven by epithelial-mesenchymal transition (EMT). We identified potential EMT-related biomarkers and targeted drugs for NPC via computational and machine learning approaches.

Based on the NPC patient dataset derived from the public database, differentially expressed genes (DEGs) were screened using the limma package and intersected with the EMT gene set from the database of Epithelial-Mesenchymal Transition-associated genes (dbEMT). Subsequently, GO and KEGG enrichment analyses were performed, and a protein-protein interaction (PPI) network was constructed via the STRING database. Core genes were identified through LASSO regression and SVM-RFE algorithms. Immune cell infiltration was assessed using CIBERSORT; pathway enrichment was analyzed with GSEA software; targeted drugs were predicted via Enrichr; and molecular docking validation was performed using AutoDock. Finally, the selected genes were functionally validated using NPC cell lines.

Among 131 EMT genes, 18 were filtered via PPI. Machine learning algorithms combined with external dataset validation further pinpointed MMP1 and TGFBR2 as hub genes. MMP1 was correlated with M0/M1 macrophages and activated NK cells, whereas TGFBR2 was correlated with resting NK cells and memory B cells. Both genes were linked to EMT, angiogenesis, hypoxia, and MYC targets. Dexamethasone showed strong binding affinity to MMP1 (-7.72 kcal/mol) and TGFBR2 (-7.27 kcal/mol). Cellular assays revealed upregulated MMP1 and downregulated TGFBR2 in NPC cells; MMP1 knockdown suppressed NPC malignant phenotypes.

MMP1 and TGFBR2 were identified as EMT-related biomarkers that mediated immune evasion and metastasis in NPC. Dexamethasone may target these genes, offering a novel therapeutic strategy for NPC.