Transcriptome-wide association study of prostate cancer in the Chinese population

Prostate cancer (PCa) is among the most prevalent malignancies in men, with a substantial genetic predisposition. Despite extensive genome-wide association studies (GWAS) identifying numerous risk loci, the causal target genes remain largely undetermined. To address this knowledge gap, we conducted the first transcriptome-wide association study (TWAS) of PCa in the Chinese population to systematically identify putative target genes.

This study included 5048 participants, comprising 2425 individuals from the Chinese Prostate Cancer Genetics Consortium cohort (1417 PCa cases and 1008 healthy controls) and 2623 from the Ruijin Biopsy cohort (1133 cases and 1490 controls). We employed both single-tissue and cross-tissue TWAS approaches to prioritize candidate genes. Causal inference was performed using two-sample Mendelian randomization, while differential mRNA expression and methylation analyses were conducted to investigate potential regulatory mechanisms. The prognostic significance of the identified genes was further assessed through survival analysis.

A total of 179 and 114 putative target genes were identified in the ChinaPCa and Ruijin Biopsy cohorts, respectively. The meta-analysis of the single-tissue TWAS revealed 23 genes achieving false discovery rate significance, while cross-tissue TWAS identified nine additional significant genes (one gene previously reported in the single-tissue TWAS was excluded). GRAM domain-containing protein 2 (GRAMD2) emerged as the most consistently associated gene across analyses. Furthermore, the Mendelian randomization analysis provided evidence supporting a causal role for trafficking protein particle complex subunit 6B (TRAPPC6B) and patatin-like phospholipase domain-containing 6 (PNPLA6) in PCa etiology.

This study represents the first comprehensive TWAS of PCa in the Chinese population, identifying 290 putative causal genes. These findings enhance our understanding of PCa pathogenesis and may inform future research on molecular mechanisms, therapeutic targets, and prognostic biomarkers.