JAM3 is a potential risk biomarker for predicting prognosis and immune cell infiltration by regulating epithelial–mesenchymal transition in bladder cancer

To identify novel biomarkers for predicting prognosis and immune function in bladder cancer (BC) patients, we combined weighted correlation network analysis (WGCNA) and least absolute shrinkage and selection operator (LASSO) regression analysis by using data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases, and finally screened the junctional adhesion molecule 3 (JAM3) as an independent risk factor in BC. High levels of JAM3 were linked to adverse clinical parameters, such as higher T and N stages. Additionally, a JAM3-based nomogram model accurately predicted 1-, 3- and 5-year survival rates of BC patients, indicating potential clinical utility. Functional enrichment analysis revealed that high JAM3 expression activated calcium signaling pathway, the extracellular matrix (ECM)-receptor interaction and the PI3K-Akt signaling pathway, and was positively correlated with genes associated with epithelial–mesenchymal transition (EMT). Subsequently, we found that overexpression of JAM3 promoted the migration and invasion abilities in BC cells, regulating the expression levels of N-cadherin, matrix metallopeptidase 2 (MMP2) and Claudin-1 thereby promoting EMT levels. Additionally, we showed that JAM3 was negatively correlated with anti-tumor immune cells such as CD8+ T cells, while positively correlated with pro-tumor immune cells such as M2 macrophages, suggesting its involvement in immune cell infiltration. The immune checkpoint CD200 also showed a positive correlation with JAM3. Our findings revealed that elevated JAM3 levels are predictive of poor prognosis and immune cell infiltration in BC patients by regulating EMT process.