OCancer occurs spontaneously in pet dogs, with an increased frequency particularly in large breeds. The enrichment of specific cancers in specific breeds suggests that inherited risk factors are involved. However, somatic tumor mutations are also important for cancer development. We have mapped inherited risk factors for several cancers including osteosarcoma and lymphoma as well as performed whole exome sequencing in several dog breeds and identified significantly mutated genes also found in the corresponding human cancers, as well as some novel genes.
Osteosarcoma is a relatively rare childhood cancer in humans, primarily affecting the long bones while being highly metastatic. In dogs the disease is more common, but clinically similar to humans, as well as highly enriched in several breeds, i.e. Rottweilers, greyhounds and Irish wolfhounds. Lymphomas consist of multiple subtypes with some more common in humans and some more common in dogs, but with a substantial overlap.
Systematic and genome-wide interrogations of human cancers have enabled characterization of their mutational landscapes and generated comprehensive pan-cancer multi-omics resources, intersecting mutational profiles with gene expression data. However, cancer driver discovery has mostly focused on coding mutations, with less attention paid to mutations in the non-coding parts, which make up >98% of the genome. Although most non-coding mutations are passenger events, mutations in regulatory regions should be elucidated for their contribution to disease initiation and progression. In a recent study of human glioblastoma, we combined with 241 Zoonomia mammalian constraint data set with whole genome sequencing data to identify genes enriched in non-coding constraint mutations (NCCMs). Mammalian evolutionary constraint is important based on the fact that bases that have a function will not mutate randomly across the mammalian tree. By looking for bases with high constraint scores, our analysis points to genes enriched for NCCMs and helps identify novel candidate cancer driver genes where the disease mutations are primarily regulatory in function.
Here we will focus on NCCM analysis of 90 osteosarcoma samples from multiple breeds and on 60 lymphoma samples with a comparison between different subtypes. NCCM discovery, gene discovery and pathway analysis will yield a better understanding of the molecular mechanisms underlying osteosarcoma and lymphoma. Our results will also be compared to similar human data sets from the ICGC. (The human data is already being analyzed on Bianca to follow regulations for human data). Integration/comparison will happen at the NCCM and gene level.