Abstract Clinical genomic sequencing has revealed that prostate cancer (PCa) progression is associated with changes to chromosome number and structure, termed chromosomal instability (or CIN). While mutations in common driver genes are rare in primary PCa, CIN is detected in 90% of cases, suggesting that CIN may be a driver of the disease. However, the underlying mechanisms that cause CIN in PCa are poorly understood. Cells are susceptible to CIN during mitosis, when the duplicated genome must be equally segregated into two daughter cells. Normally, two centrosomes (cytoplasmic, non-membranous organelles) nucleate and organize microtubules to guide assembly of the mitotic spindle, thus ensuring the fidelity of cell division. However, we recently discovered that cells within primary PCa lack centrosomes, the first cancer type to show this phenomenon. Furthermore, we showed that centrosome elimination in non-tumorigenic cells generates CIN which can transform these cells, causing them to form xenograft tumors in mice. Therefore, we sought to understand the underlying mechanisms that drive the centrosome loss observed in PCa. The prostate is a naturally hypoxic organ, averaging around 4% oxygen tension in the healthy organ which declines as men age. Additionally, tumor hypoxia is strongly correlated with high CIN, particularly in prostate cancers. However, there are few mechanistic links between hypoxia and CIN. Therefore, we hypothesized that hypoxia could cause centrosome loss and, thus, predispose cells to CIN. Indeed, we found that immortalized prostate RWPE-1 cells progressively lost centrosomes when exposed to 1% O2 for as short as 6 hours in vitro. Using ultrastructure expansion microscopy (U-ExM), we find centrosomes are replaced by non-functional microtubule husks devoid of centrosome proteins which we call ‘remnants’. These phenomena also occur in PC-3 and LNCaP prostate cancer lines. Furthermore, we observe centrosome loss specifically in the hypoxic regions of xenograft tumors. Using a combination of RNAseq and chemical inhibitors, we determined that centrosome elimination sits at the crossroads of 3 transcriptional programs: (1) centrosome elimination occurs in basal cells and basal-like cancers that express p63; (2) cells must be confluent with active Hippo pathway; and (3) hypoxia induces a HIF-independent transcriptional response. When these conditions are met, cells transcriptionally downregulate 4 key centrosome proteins (Cep63, Cep152, Cep192, and Cep215), resulting in the depletion of a protective shell of pericentriolar material (PCM) and subsequent centrosome elimination. By overexpressing the mitotic kinase Polo-like kinase 1 (PLK1), we can strengthen the PCM and prevent elimination. Finally, we identify PCa cells containing remnants within prostatectomy tissue samples. This work is significant as it identifies the first examples and mechanisms of centrosome elimination in human somatic and cancer cells, provides a mechanistic link between hypoxia and CIN, and identifies remnants as a cellular indicator for CIN predisposition. Citation Format: John M. Ryniawec, Natalya K. Seppanen, Gregory C. Rogers, Anne E. Cress. Hypoxia-induced centrosome elimination as a driver of chromosomal instability in prostate cancer abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Innovations in Prostate Cancer Research and Treatment; 2026 Jan 20-22; Philadelphia PA. Philadelphia (PA): AACR; Cancer Res 2026;86 (2Suppl): Abstract nr B065.
Ryniawec et al. (Tue,) studied this question.