A secreted signature discriminates indolent from aggressive prostate tumors

No non-invasive biomarkers are clinically available to distinguish aggressive prostate cancer (PCa) from the indolent PCa, as thus may lead the misdiagnose and overtreatment to the patients. This study aims to identify secreted proteins and assess their potentials to discriminate two PCa subtypes.

Microarray assay on tissue specimens of the discovery cohort identified secreted genes associated with invasive PCa and macrophage polarization, and the results were validated in an independent cohort. As elevated expression of ΔC499 has been linked to metastatic PCa and poorer prognosis, the relationship of key secretory genes and ΔC499 was investigated. Their roles on macrophage polarization were studied in co-cultured cells, PCa mouse model and patient blood samples. Finally, a multi-gene secreted signature model was developed and tested in training and validation cohorts.

ITGB5 was upregulated and TIMP1/TMEM176B were downregulated at RNA level in aggressive Pca, with bioinformatics linking these changes to M2 macrophage polarization. Additionally, we found that ITGB5 expression was positively regulated while TIMP1 and TMEM176B were negatively regulated through distinct transcriptional pathways. Further, we revealed that ΔC499 contributes to ITGB5-driven M2 macrophage polarization, enhancing PCa invasion in vitro and in vivo. Lastly, a multi-gene model integrating ITGB5, TIMP1, and TMEM176B distinguished indolent from aggressive PCa in training cohort (AUC = 0.88) and validation cohort (AUC = 0.90).

ΔC499 induces M2 macrophage polarization and drives PCa invasiveness by modulating ITGB5, TIMP1, and TMEM176B. Three genes signature is differentially expressed between aggressive and indolent tumors, providing potential non-invasive biomarkers to discriminate aggressive from indolent PCa.