[Molecular cancer epidemiology--the present status and future possibilities]

Molecular cancer epidemiology is a relatively new strategy for malignancies. This strategy has made it possible to diagnose the predisposition to cancer. An individual is said to have a predisposition to cancer when a tumor-suppressor gene is inactivated in germ-line cells. Mainly the inactivation of the tumor-suppressor gene is caused by mutations at the coding region of the gene. However, we clarified that point mutations or hypermethylations of the retinoblastoma tumor-suppressor gene (RB) also cause inactivation of the gene, resulting in retinoblastoma. On the other hand, studies to improve the diagnosed predisposition to cancer have not been performed. We therefore started a basic study for this purpose. As the first example of limiting the predisposition to cancer, the cases with a point mutation at the RB promoter region might be good candidates. In these carriers, only the RB promoter region is inactivated in spite of a lack of abnormalities in the coding region. Therefore, if the RB promoter activity is recovered by drugs, predisposition to retinoblastoma should be limited. As the second example, Li-Fraumeni syndrome in which the p53 gene is hereditarily mutated might be a good candidate. Recently p53 has been reported to stimulate the WAF1 gene, and the WAF1 protein to inhibit cdk activity, which inactivates the RB gene product by phosphorylation. In addition, we found that p53 up-regulates the promoter activity of the RB gene. These findings suggest that p53 directly or indirectly activates RB at the transcriptional or post-transcriptional level. Therefore, reactivation of the WAF1 or RB gene by certain drugs might compensate for the loss of function of p53 in Li-Fraumeni syndrome. We then suggest that it might be possible to prevent cancer by enhancing some intact target genes of the genetically inactivated tumor-suppressor gene. We term this new strategy "gene-regulating chemoprevention." To test this hypothesis it is important to clarify the structure of the RB promoter. In summary we found that RBF-1 and ATF sites are the core promoter regions, that the neighboring E2F site is a silencer site, and that E4TF1 preferentially binds to the RBF-1 site. We then speculate that drugs interfering with the binding of the E2F complex might become good candidates enhancing RB promoter activity. To find drugs regulating the promoters of these genes, it is reasonable to try G1 arresting drugs, because WAF1 and RB are thought to arrest cells at the G1 phase. Actually we found several drugs causing G1 arrest. They are several flavonoids which we ingest daily from vegetables and fruits, or prostaglandin D2 and its metabolite. In summary, we propose that "gene-regulating chemoprevention" will be a useful method for molecular cancer epidemiology.