Targeting allelic loss in colorectal cancer

Abstract: Targeted cancer therapy exploits molecular differences between tumor and normal cells to selectively kill cancer cells. Whereas targeting of activated oncogenes has proved clinically useful, few current therapies exploit loss-of-function mutations in tumor suppressor genes or in the genome at large. This thesis explores the consequences of allelic loss affecting tumor suppressor genes and passenger genes in colorectal cancer (CRC), aiming to identify vulnerabilities that can be exploited for therapy. In Paper I we used genome editing to model inactivation of PRDM2 and showed that PRDM2 loss impacts cell growth and invasiveness, potentially mediated by genes involved in epithelial-to-mesenchymal transition. We confirmed the role of PRDM2 as a tumor suppressor gene in CRC and proved that c.4467delA inactivating mutations constitute a driver event in CRC.In Paper II we investigated whether the reduced allelic diversity resulting from loss of heterozygosity (LOH) in cancers could be exploited for therapy. We identified target genes by mapping prevalent alleles frequently lost in cancer and investigated NAT2 loss in CRC. Drug discovery efforts identified a compound selectively toxic to tumor cells with reduced NAT2 activity, providing proof of concept for LOH targeting by small molecule drugs.In Paper III, we aimed to widen the cohort of CRC patients eligible for NAT2 allele-selective chemotherapy. We determined NAT2 slow acetylator frequencies and LOH events in two independent cohorts by next-generation sequencing and genomic arrays. Next, we demonstrated enhanced response to allele-selective chemotherapy of tumor cells encoding additional prevalent NAT2 slow acetylator alleles, and developed a method for detection of NAT2 allelic loss suitable for clinical use.In Paper IV, we extended the search of therapeutic target genes by mining loss-of-function (LoF) alleles retained in tumors after LOH. This effort identified a prevalent splice site disruption in CYP2D6 as a putative target and motivated the development of cell model systems to identify compounds targeting CYP2D6 loss in cancer cells.In Paper V we characterized a set of 56 microsatellite stable CRCs by whole-genome sequencing in an attempt to understand the genetic causes leading to genomic instability and colorectal tumorigenesis. We confirmed the mutation frequencies of known CRC genes and identified for the first time the contribution of an unknown mutational process in 10% of the analyzed tumors.