NUTRITION AND SOMATOMEDIN .22. MOLECULAR REGULATION OF INSULIN-LIKE GROWTH FACTOR-I DURING FASTING AND REFEEDING IN RATS

The liver is thought to be the locus of nutritional/hormonal regulation of circulating insulin-like growth factor-I (IGF-I). To probe the basis of nutritional regulation, we examined changes in serum IGF-I, hepatic content of extractable IGF-I immunoreactivity (a high M(r) putative precursor) and hepatic IGF-I mRNA during fasting and refeeding in rats. Preliminary studies revealed that the hepatic level of IGF-I mRNA was consistently reduced only after food was withheld for 3 days, so the effects of refeeding were subsequently examined in such animals. After 3 days of fasting, animals lost 30% of their initial weight; weight regain was apparent within 3 h of refeeding ad libitum and, after 48 h, weight was comparable to initial fed levels. Fasting reduced levels of serum and extractable hepatic IGF-I to 19 and 26% of control (fed) values respectively (both P < 0.005 vs control). There was no change in levels of serum IGF-I over the first 3 h of refeeding, but IGF-I rose above fasted levels at both 9 and 48 h (both P < 0.005). Extractable hepatic IGF-I rose more slowly and was still below fasted levels after 9 h of refeeding, and modestly, but not significantly, greater than fasted levels after 48 h. The ratio of serum to hepatic IGF-I was decreased compared with control after 3 days of fasting, but increased after 3 and 9 h of refeeding (P < 0.02 at 9 h). Northern blot analysis of total hepatic RNA revealed four species of IGF-I mRNA 0.8-1.1, 2.0, 4.0 and 7.5 kb in size. Each mRNA species fell to 15-28% of control levels after 3 days of fasting (all P < 0.001). There was a prompt increase in each transcript after 3 h of refeeding, and all values were significantly (P < 0.05) greater than fasted levels at 9 h but, at 48 h, most species were still below control levels. Levels of mRNA for the cytoskeletal proteins beta-actin and cyclophilin also fell with fasting, but were restored more rapidly than IGF-I mRNA, to or above control levels after 3 h of refeeding. The observation that IGF-I expression was decreased at 3 h when beta-actin and cyclophilin were normalized suggests specificity of regulation. Despite the temporal incongruity between IGF-I mRNA and serum and hepatic IGF-I, there were highly significant correlations (all P < 0.002) between each pair of parameters. Since refeeding of fasted animals led to a rise in IGF-I mRNA which preceded rises in serum and hepatic IGF-I, our findings are consistent with the hypothesis that nutritional regulation of circulating IGF-I involves modulation at the level of hepatic IGF-I mRNA. The changes in the ratio of hepatic to serum IGF-I during fasting and refeeding indicate that there may also be regulation at the level of hepatic release.