Inconsistencies in the hypophagic action of intracerebroventricular insulin in mice.

Insulin inhibits eating after its intracerebroventricular (ICV) administration in multiple species and under a variety of conditions. Nevertheless, the results across reports are inconsistent in that ICV insulin does not always reduce food intake. The reasons for this variability are largely unknown. Using mice as a model, we performed several crossover trials with insulin vs. vehicle when infused into the third cerebral ventricle (i3vt) to test the hypothesis that recent experience with the i3vt procedure contributes to the variability in the effect of ICV insulin on food intake. Using a cross-over design with two days between injections, we found that insulin (0.4 μU/mouse) significantly reduced food intake relative to vehicle in mice that received vehicle on the first and insulin on the second trial, whereas this effect was absent in mice that received insulin on the first and vehicle on the second trial. Higher doses (i3vt 4.0 and 40.0 μU/mouse) had no effect on food intake in this paradigm. When injections were spaced 7 days apart, insulin reduced food intake with no crossover effect. Mice that did not reduce food intake in response to higher doses of i3vt insulin did so in response to i3vt infusion of the melanocortin receptor agonist melanotan-II (MT-II), indicating that the function of the hypothalamic melanocortin system, which mediates the effect of insulin on eating, was not impaired by whatever interfered with the insulin effect, and that this interference occurred upstream of the melanocortin receptors. Overall, our findings suggest that associative effects based on previous experience with the experimental situation can compromise the eating inhibition elicited by i3vt administered insulin.