Genomic instability has been implicated as an important component in tumor progression. Evaluation of mutant frequencies (MFs) in tumors of transgenic mice containing nontranscribed marker genes should be useful for quantitating mutation rates in tumors as the physiologically inactive transgene provides neither a positive nor a negative selective pressure on the tumor. We have conducted long-term carcinogenicity studies in lambda/cII transgenic B6C3F1 mice using a variety of genotoxic and nongenotoxic test agents and have evaluated the mutant frequencies in both tumors and normal tissues from these animals. Mice were administered diethylnitrosamine (DEN) as three intraperitoneal injections of 15 mg/kg; phenobarbital (PB) or oxazepam (OXP) provided ad libitum at 0.1% or 0.25% in the diet, respectively; DEN initiation plus PB in the diet; or urethane (UTH) provided ad libitum at 0.2% in the drinking water. Normal tissues and tumors were isolated at various times over a 2-year period and half of each tissue/tumor was evaluated histopathologically and the other half was evaluated for MF in the cII transgene. Approximately 20 mutants from each of 166 individual tissues (tumor and nontumor) were sequenced to determine whether increases in MF represented unique mutations or were due to clonal expansion. UTH produced significant increases in MF in normal liver and lung. DEN either with or without PB promotion produced significant increases in MF in liver and correction of MF for clonality produced little change in the overall MF in these groups. PB produced a twofold increase in liver MF over controls after 27 weeks of treatment, but a similar increase was not observed with longer dosing times; at later time points, the MF in the PB groups was lower than that of the control group, suggesting that PB is not producing direct DNA damage in the liver. OXP failed to produce an increase in MF over controls, even after 78 weeks of treatment. Selected cases of genomic instability were observed in tumors from all treatments except OXP, with individual liver tumors showing very high MF values even after clonal correction. One rare and interesting finding was noted in a single mouse treated with UTH, where a mammary metastasis had an MF approximately 10-fold greater than the parent tumor, with 75% of the mutations independent, providing strong evidence of genomic instability. There was no clear correlation between tumor phenotype and MF except that pulmonary adenomas generally had higher MFs than normal lung in both genotoxic and nongenotoxic treatment groups. Likewise, there was no correlation between tumor size and MF after correction for clonality. The results presented here demonstrate that individual tumors can show significant genomic instability, with very significant increases in MF that are not attributed to clonal expansion of a single mutant cell.
2004 Wiley-Liss, Inc.