Previously, we reported that glycine N-methyltransferase (GNMT) knockout mice develop chronic hepatitis and hepatocellular carcinoma (HCC) spontaneously. For this study we used a phosphoenolpyruvate carboxykinase promoter to establish a GNMT transgenic (TG) mouse model. Animals were intraperitoneally inoculated with aflatoxin B(1) (AFB(1)) and monitored for 11 months, during which neither male nor female GNMT-TG mice developed HCC. In contrast, 4 of 6 (67%) male wild-type mice developed HCC. Immunofluorescent antibody test showed that GNMT was translocated into nuclei after AFB(1) treatment. Competitive enzyme immunoassays indicated that after AFB(1) treatment, the AFB(1)-DNA adducts formed in stable clones expressing GNMT reduced 51.4% compared to the vector control clones. Experiments using recombinant adenoviruses carrying GNMT cDNA (Ad-GNMT) further demonstrated that the GNMT-related inhibition of AFB(1)-DNA adducts formation is dose-dependent. HPLC analysis of the metabolites of AFB(1) in the cultural supernatants of cells exposed to AFB(1) showed that the AFM(1) level in the GNMT group was significantly higher than the control group, indicating the presence of GNMT can enhance the detoxification pathway of AFB(1). Cytotoxicity assay showed that the GNMT group had higher survival rate than the control group after they were treated with AFB(1). Automated docking experiments showed that AFB(1) binds to the S-adenosylmethionine binding domain of GNMT. Affinity sensor assay demonstrated that the dissociation constant for GNMT-AFB(1) interaction is 44.9 microM. Therefore, GNMT is a tumor suppressor for HCC and it exerts protective effects in hepatocytes via direct interaction with AFB(1), resulting in reduced AFB(1)-DNA adducts formation and cell death.