Genetics, Endogenous and Exogenous Exposures, and Human Cancer

Abstract

Cancer is a complex disease influenced by inherited germline variations and acquired somatic mutations from a variety of endogenous and exogenous exposures. Over the past decades, cancer research has mainly focused on studying the role of each of these factors separately in cancer development. However, emerging evidence suggests that tumorigenesis is a consequence of complex interactions between these genetic factors and exposures, which not only jointly influence cancer initiation and progression, but are also relevant to therapeutic resistance and patients’ outcome. It is therefore critical for novel genetic epidemiological research to go beyond studying the link between genotypes and disease phenotypes and further characterize the interactive relationship between cancer risk factors to enhance our understanding of the underlying mechanisms of cancer development. In this dissertation, I present three studies that investigated the relationships between germline genetic variations, somatic mutational profiles, and endogenous and exogenous exposures in human cancer. In Chapter 1, we evaluated the associations between germline polygenic risk scores (PRS) of cancer and cancer-related traits and tumor somatic mutational profiles across 12 cancer types using data from The Cancer Genome Atlas. We identified 17 significant associations between germline PRS and single-base substitution (SBS) mutational signatures or the total somatic mutation counts, including positive associations between germline inflammatory bowel disease PRS and somatic mutations attributable to signature SBS1 in prostate cancer and APOBEC-related signatures in breast cancer, and age at menarche PRS and mutation counts of SBS1 in overall and estrogen receptor-positive (ER+) breast cancer; inverse associations were found between several cancer PRS and tumor mutation counts, and age at menarche PRS and mutation counts of SBS1 in prostate cancer, likely reflecting hormone-related mechanisms. These findings provide insights into the underlying mechanisms of somatic mutagenesis and cancer development. In Chapter 2, we conducted a genome-wide association study to assess the genetic basis of mammographic texture variation (V), an independent risk factor for breast cancer, in the Nurses’ Health Study (NHS), NHSII, and Mayo Mammography Health Study cohorts. Three genome-wide significant loci associated with V were identified, including 6q24.1 (ECT2L), 8q24.22 (LINC01591), and 12q22 (PGAM1P5). We identified additional shared susceptibility loci for V, mammographic density (MD), and breast cancer risk, which may indicate shared mechanisms. On genome-wide or single-nucleotide polymorphism-set scale, V was genetically correlated with MD phenotypes, ER+ breast cancer, adult body mass index, and childhood body fatness. These findings provide the first evidence of the genetic basis of V and the shared genetic components between V, MD, and breast cancer risk. In Chapter 3, we further characterized the germline-somatic relationship in cancer by assessing the associations between germline cancer gene expression quantitative trait loci (eQTL) and local and global tumor mutations across 11 cancer types in the Dana-Farber Profile cohort. We identified significant associations between germline eQTL for several well-established cancer genes, including APC, ATM, CBFB, EPHA5, FANCA, GLI1, GLI2, WRN, and TP53, and local mutation status of one cancer gene or the global mutational burden. These findings provide further evidence for the potential role of cancer gene expression regulation in germline-somatic associations.