Gene-environment interaction is a seminal concept in the molecular epidemiology of human cancer. Our case-control (using hospital- and population-based controls) studies focus on lung cancer, a tobacco-related cancer, and colon cancer, one of the cancer types associated with chronic inflammation. These studies require the integration of an increasing understanding of genetic variation in cancer susceptibility, analysis of carcinogen exposure, rapidly developing technologies, bioinformatics, social-ethical concerns, and epidemiological study-design methods. We have discovered that a deficient G2/M cell cycle checkpoint that responded to DNA damage, a phenotypic trait, is associated with a higher lung cancer risk in African Americans than in Caucasians. We have developed a novel bioinformatic approach to identify SNP-SNP interactions and generate hypotheses. For example, the GSTT1-null polymorphism, the Asp72Pro TP53 polymorphism or the Asp302His CASP8 polymorphism was positively correlated with colon cancer risk. In collaboration with Alavanja, we have extended our previous molecular epidemiological studies of lung cancer in never-smoking women. In addition to the GSTM1-null polymorphism increasing the risk of secondhand smoke-induced lung cancer, GSTM1 null also increases the risk of environmental radon-induced lung cancer in these women. We are continuing our longstanding studies of human lung carcinogenesis. The molecular profile of adenocarcinoma identified smoking- versus nonsmoking-associated cancers and short-term versus long-term survivors. We have also discovered molecular profiles of microRNAs, nonprotein coding genes that identify lung cancer, its different histological types and prognosis. These findings are being extended to other cancer types including colon and esophageal cancer, and to animal models of human cancer.
Showing the most recent 10 out of 57 publications
