Omics Approaches to Uncover Germline and Somatic Variation Underlying Inherited Sarcomagenesis

Sarcomas are rare mesenchymal tumors, making up 15% of all childhood and 1% of all adult tumors. They account for a disproportionate share of mortality in young adults, and if left untreated, are highly likely to metastasize. However, sarcoma etiology is poorly understood, and having numerous histological subtypes has complicated elucidation. To better understand factors underlying sarcomagenesis, we leveraged two rare inherited cancer predisposition syndromes, Li-Fraumeni Syndrome (LFS), and LFS-like (LFSL), both with a high incidence of sarcomas. LFS is caused by mutations in the tumor suppressor gene TP53 (p53), but has variable and incomplete penetrance, suggesting additional acquired somatic mutations are necessary for tumorigenesis. In contrast, LFSL has no known cause, although a 10-Mb region in 1q23 has been mapped by linkage analysis as a putative LFSL locus. Therefore, to better identify genetic variation underlying LFS and LFSL we utilized a 2-pronged approach. First, we evaluated LFSL families for rare, co-segregating, germline mutations, which identified a mutation in ARHGAP30 that was present in four LFSL families. Moreover, this mutation impacted both proliferation and migration when overexpressed in vitro. Subsequent analysis of publicly available data indicates a potential role for ARHGAP30 in sporadic cancers. Secondly, we endeavored to identify somatically acquired drivers of sarcomagenesis. In cancer, passenger events are acquired concomitantly with driver mutations, and distinguishing them remains a key challenge. To best address this, we used a comparative genomics approach, leveraging a “humanized” mouse model of LFS with a hotspot mutation, Trp53R172H, analogous to TP53R175H in humans. Hypothesizing that sarcoma etiology is similar in humans and mice, we then catalogued recurrent changes in the genome, transcriptome, and methlyome. We found little overlap in any of the omics approaches across the human tumors, which came from diverse p53 mutations and sarcoma types, but found strong overlap in the mouse tumors (fibrosarcomas and osteosarcomas). Recurrent data discovered in the mouse was mirrored in some human sporadic mesenchymal tumors, including novel genes like MROH2A, and MIR219A2. Our results emphasize the utility of a model disorder and comparative omics to uncover genes with relevance for both inherited and sporadic tumors