Single-cell RNA sequencing, spatial transcriptomics, and genetic risk localization reveal dysregulated immune status and metabolic reprogramming in castration-resistant prostate cancer

Purpose As one of the most common cancers, prostate cancer is partially treated with androgen deprivation therapy. Following treatment failure, patients progress to castration-resistant prostate cancer (CRPC), which is associated with poor prognosis. This study aims to elucidate the dynamic changes in the tumor microenvironment, genetic risk architecture, and therapeutic targets across prostate cancer progression through omics analysis. Materials and methods Tissues were collected from patients in three groups: benign prostatic hyperplasia (BPH, n = 3), treatment-naive prostate cancer (PCa, n = 4), and CRPC ( n = 3) for scRNA-seq analysis. Publicly available spatial transcriptomics data (GSE278936, 10x Visium) from 10 representative samples spanning the benign-to-advanced prostate-disease spectrum (BPH, PCa, Neoadjuvant, CRPC, Metastatic) were analyzed. gsMap was used to map GWAS genetic risk signals onto spatial niches. Geneformer was fine-tuned for virtual perturbation of key CRPC genes (LAS1L, IL6ST, TNFAIP8). Results We report that inherited prostate-cancer genetic risk undergoes a systematic spatial redistribution as disease progresses. By integrating scRNA-seq (10 patient samples) with spatial transcriptomics and gsMap genetic-risk mapping (10 samples from GSE278936), we found that genetic-risk enrichment is concentrated in epithelial niches in localized disease but shifts to endothelial, fibroblast, and Treg/neutrophil niches in CRPC. This genetic redistribution was accompanied by a tumor-microenvironment shift from immune surveillance toward Treg-mediated suppression and by the emergence of a metabolically active CRPC2 tumor-cell state. Virtual perturbation of LAS1L altered the malignant transcriptional state, although it did not reverse it. This nominates LAS1L as a hypothesis-generating, spatially risk-associated candidate whose causal role requires experimental validation. Together, these multi-layered analyses indicate that the cell types carrying inherited disease risk are remodeled during progression to CRPC, providing a spatial-genetic framework for studying the advanced-disease microenvironment. Conclusion In conclusion, this study delineates the dynamic shifts in the immune microenvironment during prostate cancer progression, characterized by a transition from immune surveillance to immune suppression. This study highlights the redistribution of germline genetic risk across cellular compartments as a key feature of advanced prostate cancer, providing a multi‑omics spatial resource for understanding microenvironmental remodeling in CRPC. The predicted association between LAS1L and ETV7 may provide novel insights into the tumor-immune interface in advanced prostate cancer.