Important etiological clues may arise from comparison of spectra and frequency of mutations in oncogenes/tumor suppressor genes in human and animal tumors. Systematic evaluation of a large series of these for such gene mutations, needed for statistical reliability in these studies, has been hampered by cumbersome, artifact-plagued techniques. Our group has continued to make important and recognized advances in technical development in this area. These have included, this year, improvements in sensitive use of single strand conformation polymorphism (SSCP) analysis with silver staining of bands, a method for distinguishing polymerase chain reaction (PCR) artifact from real mutation using a base-mismatch primer approach, and systematic study of conditions for extraction of DNA for PCR from formalin-fixed tumors. This year also we report interesting findings with both humans and animals. Collaboratively with the Genetic Epidemiology Branch, we focused on mutations in the p53 tumor suppressor gene in human gastric carcinomas from a high-risk area of Italy. We found that, in a subset of gastric tumors in a high-incidence area around Florence, G:C ~ A:T changes at CpG site were less common than in other parts of Italy. With other data, the results show that European gastric cancers had significantly more G:C ~ A:T and fewer A:T ~ G:C transitions in p53 than those in Orientals. These findings may provide insight on etiological factors for gastric cancer. In a new phase, we have increased the number of cases and included several categories of gastric tumors. Thus far, differences in the mutation spectrum and in frequency of p53 protein accumulation suggest that further important clues about etiology specific to certain histopathological types may be forthcoming. A human exposure-related rat study coming to completion involves nitroglycerin, used to treat angina pectoris and reported earlier to cause rat liver cancer. We have confirmed this observation, and found that K-ras oncogene mutations occur in a significant proportion of these tumors. Also in rats, we investigated possible involvement of the retinoblastoma (Rb) tumor suppressor gene in tumorigenesis. Since targeted disruption of this gene causes pituitary tumors, we studied the possible role of the Rb gene in similar pituitary tumors induced by nickel acetate. While levels of Rb mRNA, measured by the ribonuclease protection assay, were not consistently altered, Western blot analysis of the protein showed a uniform change in all tumors associated with complete absence of the hyperphosphorylated (115 kD) form of the protein. This finding suggests disruption of the normal phosphorylation/ dephosphorylation equilibrium for the Rb protein, believed to be part of cell cycle control. Detection of this abnormality in all 9 tumors is consistent with Rb involvement in the etiology of these tumors.
