Investigation of selective constraints in colorectal cancer

Abstract

Next generation sequencing (NGS) has significantly advanced our understanding of colorectal cancer (CRC) and highlighted the inaccuracies of the traditional, linear model of adenoma-carcinoma evolution. NGS has provided evidence for intra-tumour heterogeneity (ITH) and colorectal carcinomas that lack a full complement of driver mutations. Alternatively, the branched model of adenoma-carcinoma evolution proposes that genetically differing clones can coexist and evolve in spatially discrete areas. Further refinements such as the just-right model and mini-driver models are encompassed by the branched model, providing additional insights.

By examining mutations in multi-regional, colorectal adenomatous polyp samples, I have provided supporting evidence for the branched model of CRC evolution. Following an assessment of the presence of parallel evolution in adenomas, I looked specifically at mutations in Wnt pathway genes to determine whether there is evidence for a “just-right” level of pathway activation by mutations. Canonically, the Wnt pathway is central to epithelial cell growth and regulation in the intestine. When mutationally activated, this pathway can lead to uncontrolled neoplastic initiation and progression. Previous evidence from familial adenomatous polyposis (FAP) suggests that a just-right number of APC 20-amino acid repeats results in an optimal level of Wnt pathway activation. The applicability of this model to the entire Wnt pathway has been suggested. The mini-driver model proposes that low-frequency mutations of minor selective advantage, compared with major-driver mutations, can finely adjust Wnt activation levels to provide a just-right level of Wnt signalling. Using a large, CRC dataset (100KGP), I explored mutations in both major-driver and mini-driver genes, seeking patterns of mutual exclusivity and co-occurrence. The results supported the just-right model, revealing mutual exclusivity between mutations in major-driver genes and co-occurrence between mutations in major-driver and mini-driver genes. These results suggested that combinations of Wnt pathway mutations are selected to provide a just-right amount of Wnt signalling for a tumour to evolve. In addition, this study also found potential discrepancies between particular combinations of mutations in different tumour genomic instability subtypes (MSS/MSI) and in different anatomical locations, which was suggestive of differing just-right signalling levels depending on the genomic instability/location context of the tumour. Furthermore, the proposed role of TCF7L2 as a mini-driver, and CRC predisposition gene, was investigated through genome-wide association studies (GWAS) and mouse model studies, showing some supportive results.

This research contributed valuable insights into the complex dynamics of CRC evolution, shedding light on the role of the just-right model involving the Wnt pathway and supporting the branched model of adenoma-carcinoma evolution.