Pancreatic cancer remains one of the most lethal malignancies worldwide, underscoring the urgent need for reliable biomarkers to predict prognosis and guide therapy. This study aimed to identify and characterize such biomarkers using an integrated bioinformatics and experimental approach. The GEO database was used to screen genes differentially expressed in pancreatic cancer. Next, the DAVID database was used for GO functional enrichment and KEGG pathway analyses. STRING database and Cytoscape software were utilized to generate PPI networks and identify key genes. The prognostic value was then evaluated by utilizing the GEPIA database and Cox regression analysis. The relationship between key genes and clinicopathological parameters was analyzed using the UALCAN database. The TISCH2, TIMER, and TISDB databases were used for immune infiltration analysis. The functional role of the key gene was ultimately confirmed through in vitro experiments. Through multiple rounds of strict screening, 10 key genes were obtained. After further verification, it was determined that MMP1 was significantly upregulated in pancreatic cancer tissue and correlated with poor prognosis in patients. Immunological analysis revealed correlations between MMP1 expression and the infiltration of specific immune cells (B cells, M1 macrophages, neutrophils, and monocytes) as well as the expression of immune checkpoint genes (VTCN1, LGALS9, TGFBR1, and IL10RB). In vitro functional assays confirmed that MMP1 knockdown in pancreatic cancer cells suppressed proliferation, migration, and invasion, while promoting apoptosis. Furthermore, a co-culture model demonstrated that MMP1 facilitated the recruitment of M1-polarized macrophages. MMP1 is an independent prognostic biomarker that drives pancreatic cancer progression through direct oncogenic effects and modulation of the tumor immune microenvironment. Thus, it emerges as an attractive, multifunctional therapeutic target, which calls for future research to confirm its translational potential and delineate the underlying molecular pathways.
Wang et al. (Thu,) studied this question.