Cancer genomes carry hundreds of mutations that reflect exposures to exogenous carcinogens or the effect of endogenous factors and the selection of events driving the cell growth and survival. Analyses of carcinogen-specific mutation patterns can thus reveal past exposures and inform hypotheses on causative agents for specific cancer types. While data on somatic mutations in human cancers accumulate at a high speed through high-throughput sequencing efforts, there is a lack of experimental data on genome-wide mutation patterns induced by specific carcinogens. Such experimental data would allow tracing the impact of specific carcinogens in human tumors, thus helping the interpretation of human mutation patterns and providing clues about possible links between a specific carcinogen and a specific tumor types. Here we propose to model in experimental systems exome-wide mutation patterns induced by suspected human carcinogens. Our approach is based on the use of an in vitro assay based on Hupki mouse embryonic fibroblasts (HUF) and on archived mouse tumors obtained in carcinogenicity assays. In the in vitro assay, HUF are exposed to carcinogens and DNA mutations are examined after a selective step relevant to tumor development, namely clonal immortalization. Mouse tumors of exposed mice will be retrieved from the archives of the NIEHS National Toxicology Program. We will perform exome sequencing of HUFs immortalized after exposure to candidate compounds and of mouse tumors, and apply statistical models to identify exposure-specific mutation signatures. Signatures obtained in these experimental models will be searched for in genomes from human primary cancers to identify fingerprints of exposures to the tested compounds. Data on human cancers available from public repositories, and new series of tumors occurring in patients with documented exposure to the tested carcinogens will be analyzed or mined. Our pilot work using the in vitro assay showed very good concordance between experimentally-induced mutation patterns and those observed in human cancers.
The aim of this proposal is to further investigate the capacity of our approach to discriminate between different types of carcinogens having similar genotoxic properties, and to identify suspected carcinogens in series of occupational cancers.
Somatic mutations that accumulate during cancer development reflect exposures to exogenous carcinogens. The analysis of tumor mutation patterns can thus provide hypotheses on causative agents for specific cancer types. While a large amount of data on somatic mutations in human tumors is accumulating through large-scale sequencing efforts, there is a lack of experimental data on genome- wide mutation patterns caused by exposures to specific genotoxic compounds that would allow their identification as potential carcinogens to humans. This project proposes to fill this gap by modeling mutational landscapes of human cancers using a cell immortalization in vitro assay. The data generated are expected to accelerate the interpretation of mutation patterns observed in human cancers and to provide mechanistic evidence for the classification of potential human carcinogens.
| Huang, Mi Ni; Yu, Willie; Teoh, Wei Wei et al. (2017) Genome-scale mutational signatures of aflatoxin in cells, mice, and human tumors. Genome Res 27:1475-1486 |
