(190aa) Exploring the Metabolic Shift Associated with Cancer Hypermutation

Somatic mutations are a key characteristic and driving force of malignant transformation, and their specific patterns facilitate diagnosis and prognosis across a wide variety of cancer types. In some cases, a tumor will exhibit an exceedingly high mutational burden, which can shroud driver mutations from passengers, and may alter its response to certain treatment types. Many of the causes of this “hypermutation” phenotype are known, and include environmental (e.g., tobacco smoke and UV light) and endogenous (e.g., dysfunctional DNA repair proteins) sources. Beyond genomic instability, however, the effects of hypermutation are less clear, especially with regards to resulting changes in the transcriptome, and much less the metabolic network. Given the importance of aberrant metabolism in cancer, we sought to investigate the link between DNA hypermutation and the metabolic network in tumors.

Using whole-exome sequencing data and RNA sequencing data from ~9,000 patients spanning over 30 different cancer types retrieved from The Cancer Genome Atlas (TCGA), we systematically explored the relationship between hypermutation status and metabolic gene expression. Samples were classified as hypermutated if they possessed at least 10 mutations per Mb, and metabolic genes were defined as those present in HMR2, a human genome-scale metabolic model. Although many cancer types exhibited a negligible connection between hypermutation status and metabolic transcriptome, four carcinomas (colon, endometrial, gastric, and lung) displayed a substantial shift in expression. These select carcinomas were explored in further detail, using differential gene expression analyses and integration with the HMR2 network to identify coordinated changes in metabolic subnetworks. The results of this study shed light on a largely unexplored link between two key characteristics of cancer cells—hypermutation and altered metabolism—and reveal specific metabolic processes that may contribute to observed differences in treatment response associated with hypermutation status.