PA2G4 in CAFs promotes biochemical recurrence of prostate cancer via H3K18la

Cancer-associated fibroblasts (CAFs) exhibit heterogeneity and play diverse roles in prostate cancer (PCa) progression, yet their specific impact via lactate metabolism remains unexplored. This study identifies a CAFs lactate metabolism-associated transcriptomic signature (CLS) through multi-omics analysis and establishes a Lactate Metabolism-Related Clinical Prognostic Index (LMCAFCPI) using machine learning. Lactate metabolism genes from MsigDB, were used to derive CLS from single-cell RNA sequencing (scRNA-seq). Unsupervised clustering divided samples into two risk subgroups, and GSVA and ssGSEA assessed biological functions and immune features. The LMCAFCPI demonstrated superior prognostic performance. A nomogram incorporating Gleason grade, PSA, T-stage, and LMCAFCPI showed stable biochemical recurrence (BCR) prediction. The study also revealed positive correlations between LMCAFCPI and specific immune cells, and validated signature genes in cellular models. CUT-TAG sequence confirmed PA2G4 knockdown in CAFs suppresses PCa proliferation and metastasis by inhibit H3K18la. Overall, LMCAFCPI offers a novel prognostic tool for PCa management and highlights potential targets for future research, with PA2G4 emerging as a potential therapeutic target due to its inhibition of H3K18la in PCa.