Integrated multi-omics profiling uncovers the epigenetic, transcriptional, and metabolic landscape of prostate cancer progression

Abstract

A comprehensive understanding of the underlying molecular mechanisms of prostate cancer is essential for the development of precise diagnostic biomarkers. In this study, we applied unsupervised multi-omics factor analysis (MOFA) to integrate DNA methylation, gene expression, and metabolic profiles derived from the same individuals, aiming to characterize the biological landscape of normal, malignant, and aggressive prostate tissue. Our analysis identified distinct molecular pathways associated with aggressive disease, specifically those involved in zinc metabolism, cell cycle regulation, smooth muscle architecture, immune activation, and tissue morphology. Key metabolites within the TCA cycle, amino acid metabolism, and lipid pathways were central to these signatures. Furthermore, we observed a consistent co-enrichment of SP1 and CTCFL binding regions among factor-associated CpGs, suggesting a model of global epigenetic reprogramming. These findings indicate a novel interplay between Polycomb deregulation, CTCFL-mediated chromatin remodeling, and SP1-driven transcriptional activation in shaping the prostate cancer epigenome. Apart from immune activation, the identified molecular signatures were validated in the TCGA cohort and demonstrated significant predictive value for disease recurrence. Overall, these results underscore the power of multi-omics integration in providing a holistic understanding of prostate cancer biology and its potential for clinical translation into prognostic biomarkers.