Epigenetic mechanisms in prostate cancer metastasis to the bone

In Australia, cancer accounted for approximately 30% of registered deaths in 2016. Although the 5-year survival rate for cancer is more than 68%, once a primary solid tumour has metastasised the 5-year survival of the disease markedly declines. Prostate cancers primarily metastasise to the bone and five-year survival for those with secondary bone lesions is 3%. Globally, prostate cancer is the second most commonly diagnosed cancer and the fifth leading cause of cancer death among men, with more than 1.4 million new cases of prostate cancer and 375,304 associated deaths worldwide estimated in 2020. With limited effective therapies, current diagnostic tools are unable to distinguish indolent disease from that with a propensity to become metastatic. Recent evidence suggests that transcriptional reprogramming may drive the metastatic phenotype, and that this reprogramming is controlled at least partly by epigenetic changes to the DNA of tumour cells, including DNA methylation. These changes, referred to as ‚ÄövÑv¿epigenetic drivers‚ÄövÑvp, have previously been shown to be involved in cancer cell survival. As genetic mutations alone don‚ÄövÑv¥t explain cancer metastasis, it has been postulated that epigenetic drivers are integral to the development of metastatic disease. Here it was hypothesised that, although there is considerable interindividual variation, fewer DNA methylation differences between primary prostate tumour and bone metastasis will be shared between individuals and these are more likely to drive the metastatic phenotype. Key DNA methylation changes driving bone metastasis in prostate cancer were identified using a powerful genome-wide approach to analyse formalin-fixed paraffin embedded tissue of primary tumour and matched bone metastases. Differentially methylated regions (DMRs) that were shared between individuals were identified using the Infinium¬¨vÜ MethylationEPIC array. This analysis found that significant DNA methylation differences are evident between matched metastatic and primary lesions and that a subset of these are shared between individuals. A selection of top candidate DMRs, associated with the promoters of genes CD81, WNT5A, and an intergenic region upstream of TBX3, were further analysed in prostate cancer cell lines to identify any functional changes in gene expression associated with the observed methylation changes. This analysis revealed that, particularly for the regions associated with WNT5A and TBX3, differential methylation was associated with differential expression between cell lines representative of localised disease (22Rv1) and bone metastasis (PC-3). A genome-wide CRISPR/Cas9 negative screen in prostate cancer cell line, 22Rv1, was then used to identify genes associated with increased survival and proliferative abilities. The genes identified in this analysis were cross referenced with genes associated with DMRs found between the paired primary prostate tumour and bone metastases. Nine genes were identified to overlap between those identified by the CRISPR/Cas9 screen as required for survival/proliferative capability and those for which significant methylation changes were identified between bone metastasis and primary tumours. Two genes were selected for further analysis, GATA2 and MDC1. Detailed analysis of the two selected genes found that, for both, differential DNA methylation was observed in prostate cancer cell lines 22Rv1 and PC-3 across regions identified as DMRs by the EPIC array analysis. The changes in DNA methylation were shown to be associated with gene expression changes. Both initial DNA methylation and CRISPR/Cas9 analyses have identified several genes exhibiting differential DNA methylation and gene expression in prostate cancer cell lines, with the changes associated with WNT5A and GATA2 indicating that these DMRs have the potential to act as epi-drivers of metastasis in prostate cancer. Breast cancer also has a high propensity to metastasise to bone, typically presents initially as hormone-dependent and has defects in similar gene pathways, for example mutations in DNA repair genes contribute to risk of both breast and prostate cancer. Given these shared characteristics it was hypothesized that alterations characteristic of prostate cancer metastasis to the bone may also be apparent in breast cancer. As such, investigation of genes identified as potential epi-drivers of metastasis in prostate cancer were examined in breast cancer cell lines. The genes further examined, WNT5A and GATA2, were found to have differential DNA methylation and expression in breast cancer cell lines also, showing potential for acting as epi-drivers across both cancer types. Taken together, here an innovative combination of approaches was taken with a view to identifying putative epigenetic markers of metastatic capability in bone metastases. Outcomes of these experiments are consistent with the hypothesis that epigenetic alterations exist between primary and metastatic disease and that these are associated with gene expression changes. Together these key molecular changes may act as drivers of the metastatic phenotype. Given the dearth of effective therapies available for metastatic cancers, particularly those that have metastasized to bone, these genes present as potential putative targets for therapy.