Patient-derived primary prostate cancer epithelial cultures for studying prostate cancer cell evolution under hypoxia.

Abstract

397 Background: Genetic instability is a hallmark of cancer and a driving force of tumourigenesis. An emerging concept is that hypoxia can potentiate genetic instability and tumour evolution leading to a poor clinical prognosis in localised prostate cancer. Mechanistic studies that explicitly test the ability for hypoxia to drive an increased DNA mutation burden and subsequent chromosomal instability(CIN) are lacking in sporadic and hereditary prostate cancer. We hypothesised that constant hypoxia selection pressure would increase clonal and sub-clonal selection for mutator phenotypes and CIN, beyond that achieved by aberrant cancer driver gene expression alone. Methods: We created isogenic single-hit and multi-hit prostate cell models for TP53, ATM, BRCA2 and RB loss and/or c-MYC overexpression using human telomerase ( hTERT )-immortalized prostate epithelial cell (PrEC) cultures. PrEC cultures were derived from morphologically-normal areas of radical prostatectectomy specimens. Isogenic cells were grown ex vivo under oxia (21% O 2 ), acute on chronic hypoxia (72h x 1% O 2 ) or constant long term, hypoxia exposure (20 generations x constant 1% O 2 ). Resulting DNA mutation burden and chromosomal instability were assayed using a combination of whole genome sequencing, RNAseq and immunofluorescent assessment of mitotic aberrations (antibodies against centrin-1, H3 phospho-serine10 and DAPI-DNA counterstain). Results: Under oxia, de novo percent genome alteration and micronuclei frequency was highest in the combined c-MYC/TP53 -/- clones compared to single gene c-MYC, BRCA2 -/- , ATM -/- or TP53 -/- clones. When growing the cells under acute on chronic hypoxic conditions (72 h, 1% O 2 ), there was an increase in micronuclei and centrosome aberrations across all genetic clone sub-types, with the highest incidence in c-MYC/TP53 -/- clones. Continued passaging of these cells led to increasing aneuploidy over time. Our CIN results were even more striking when using the PrEC cells with inactivated TP53 and RB (due to SV40 large T antigen transduction) following 20 generations of constant hypoxia. The observed increased percent genome alteration (45% increasing to 84%) was associated with increased micronuclei, PGCCs, erroneous anaphases and whole-genome duplication. In situ validation of these in vitro mechanistic studies is underway using spatial transcriptomic and whole genome profiling of 30 primary high-risk sporadic and hereditary prostatectomy specimens stained for hypoxic biomarkers (HYPROGEN trial: NCT05702619). Conclusions: Our results suggest that hypoxia augments genetic instability and CIN in cells secondary to cancer driver gene aberrations giving rise to clonal populations with selective advantage. These findings support the testing of therapeutic strategies that address both hypoxia and cancer driver mutations as a means to improve outcomes.