A comparative genomics approach to identify novel inherited cancer risk variants in dogs and humans

Abstract

Dogs provide a very special and unique opportunity for novel discovery in inherited disease studies. Through breeding practices to aid in the development of specific breeds, dog breeds are a very homogenous population, which has resulted in an increase of inherited diseases in purebred dog populations. Early genetic studies of the dog often employed linkage maps within familial studies, which were largely helpful due to the high linkage disequilibrium that exists in the dog population from breeding. However, as sequencing technology has developed, more dog genetic studies are being carried out, and many of them utilizing next generation sequencing (NGS) technology. From this technology, whole genome sequencing (WGS) is a sequencing option that provides an unbiased survey of the entire genome. This complete genome of information has become more crucial in genetic studies as it is estimated that the vast majority of disease influencing mutations within dogs will be outside of the coding portion of the genome. Furthermore, there are many diseases that have genetic similarities in both dogs and humans, allowing the dog to benefit from previous human disease studies and also to serve as a model for human diseases. Dogs have been successful models for very heterogeneous human diseases. WGS has been an effective method for identifying mutations associated with inherited diseases through multiple different analyses methods, and identifying disease influencing risk genes in dogs can be easier due to the high homogeneity within breeds. This can then be translated to human disease studies, potentially as candidate gene approaches. This approach also translates well to cancer studies, as cancer is a genetic disease, and WGS can aid in identifying mutations in both species. Due to similar presentations and previously known similar genetic links between breast cancer and canine mammary tumors (CMT), a cohort of purebred CMT-affected dogs were investigated through pedigree analysis and WGS to identify risk variants within the cohort. This involved an initial analysis of mutations in orthologs of human breast cancer risk genes. Variants within both BRCA2 and STK11 were associated with CMT risk; breed-specific associations were identified. This initial analysis highlighted the effectiveness of WGS and elucidating CMT risk in small breed-specific cohorts. In search for novel risk variants, the WGS data of five Golden Retrievers were subsequently analyzed. Upon identifying and validating mutations shared amongst all five Golden Retrievers, the results were compared to human breast cancer cases to elucidate risk. Rare protein truncating variants (PTVs, nonsense, frameshifting and splice-site affecting mutations) were investigated in the Golden Retrievers WGS data and then genotyped in the remaining Golden Retriever cohort. From this a frameshifting mutation in CEACAM24 was identified in the CMT-affected Golden Retriever cohort, which translated to a significant association of rare PTVs in the CEACAM gene family in human breast cancer cases. This was the first time inherited mutations the CEACAM gene family were associated with inherited breast cancer risk. The CEACAM gene family has long been tied to colorectal cancer (CRC) development and progression; however, there is limited to no information on this gene family and inherited CRC risk. An analysis to investigate an association with CEACAM genes and inherited CRC risk was carried out. Rare PTVs and missense mutations were both investigated for influence, and no gene-based or –family associations were identified. However, certain individual mutations were associated, highlighting the need for further exploration. Ultimately, this work represents one of the first investigations of the CEACAM gene family and inherited CRC risk. This dissertation highlights the power of WGS of dogs and how such studies can benefit human health through comparative oncology.