Mouse models of insulin resistance and IGF-1 deficiency in colon carcinogenesis

2421 Insulin resistance is a risk factor for colon cancer, but it is not clear which of the metabolic consequences of insulin resistance are involved. Muscle-specific Insulin Receptor Knockout (MIRKO) mice exhibit elevated serum triglycerides (TG), free-fatty acids (FFA) and fat mass, but have normal body weight, circulating glucose and insulin and normal insulin sensitivity relative to their wild-type (WT) littermates. Liver-specific IGF-1 knockout (LID) mice have a significant reduction in circulating IGF-1 levels relative to their WT littermates. The objective of this study was to test in MIRKO mice the hypothesis that hyperlipidemia promotes colon cancer independent of changes in insulin, and in LID mice, that a reduction in IGF-1 protects against colon cancer. Seven-week old male MIRKO mice, 12-week old male and female LID mice and their WT littermates, were treated with AOM to induce aberrant crypt foci (ACF) or colon tumors. In MIRKO mice, ACF were assessed at 24 weeks of age and tumors at 40 weeks of age. Unexpectedly, the metabolic phenotype of the MIRKO mice changed significantly as the mice aged. Seven week-old MIRKO mice displayed hyperinsulinemia and at 16 weeks they displayed reduced insulin sensitivity. The previously reported MIRKO phenotype developed between 16 and 24 weeks of age. By 40 weeks of age, however, the serum metabolites as well as insulin were not different from WT, but the MIRKO mice were again insulin resistant. ACF development was not different in MIRKO and WT mice, but MIRKO mice had a reduced susceptibility to the development of tumors. Our results suggest that circulating TG and FFA are not promoters of colon carcinogenesis. Circulating adiponectin, however, was elevated in MIRKO mice by 24 weeks of age, suggesting that this adipokine may have inhibited formation of colon tumors. In LID mice, ACF were assessed at 16 and tumors at 40 weeks of age. Male and female LID mice had approximately 60% lower serum IGF-1 levels compared to WT littermates. Male LID mice were hyperinsulinemic relative to WT, though female LID mice had normal insulin levels. There was a significant inverse correlation between insulin and IGF-1 levels in males only, and there were no differences in total ACF between male LID and WT mice. Female LID mice, however, developed significantly fewer ACF compared to WT and there was a significant positive correlation between serum IGF-1 and total ACF. By 40 weeks of age, both male and female LID mice were hyperinsulinemic and there were no differences in tumor development compared to WT. There was a significant inverse correlation between insulin and IGF-1 in both males and females at this age suggesting that the reduced susceptibility to colon carcinogenesis caused by reductions in circulating IGF-1 may be masked by hyperinsulinemia. Taken together, these studies exclude circulating lipids as factors that increase risk of colon cancer development, but support the involvement of the insulin-IGF-1 axis.