Differential genome-wide DNA methylation in prostate tumours from South African men

Abstract

Background: DNA methylation is an epigenetic mechanism known to aid the progression of cancer, including prostate cancer. It is part of a cluster of molecular processes that initiate tumorigenesis and drive its early evolution by altering other molecular processes. While studies have looked at DNA methylation in prostate cancer, most have been limited by targeted gene analysis, with further bias towards non-African cohorts. Considering the enhanced coverage of more recent genome-wide arrays, such as the Illumina Infinium HumanMethylationEPIC BeadChip, which measures DNA methylation over more than 850,000 CpG sites genome-wide, many studies that have employed a more global approach to DNA methylation analysis are further limited by frequently utilising lower-coverage arrays. Due to the bias against African cohorts, African-relevant bioinformatic tools for the processing of African DNA methylation data, particularly generated by the EPIC array, are scarce. As a result, the genomic mechanisms that underlie African prostate cancer as well as the contribution of DNA methylation alterations to African prostate cancer are poorly understood. Results: Working with EPIC DNA methylation data, I present a novel established African-relevant genome-wide bioinformatic pipeline for the processing and normalisation of African tumour-derived genome-wide DNA methylation data. Pilot application of this pipeline on prostate tissue identified differentially methylated CpG dinucleotides that may contribute to aggressive prostate cancer in a small cohort of men of South African ancestry. Additionally, I identified top genes in South African prostate cancer that are significantly enriched for differentially methylated CpG sites. Finally, patient-matched genomic-epigenomic data integration revealed preliminary evidence for interplay between these two systems in African prostate cancer, although the identification of DNA methylation signatures would prove more insightful. Conclusions: Ultimately, this work highlights the marginalization of Africans in scientific research. As a preliminary solution to this underrepresentation, this dissertation provides a novel toolset to appropriately handle African DNA methylation data with the ultimate goal of generating a deeper understanding of the genomic mechanisms harboured within African prostate cancer, a field with limited knowledge. Potential improvements to this tool, complications encountered when interpreting epigenome-wide results as well as the near future of cancer genomics is discussed.