Search

The AMP-activated kinase (AMPK) is a serine threonine kinase that functions as a fuel sensor to regulate energy balance at both cellular and whole-body levels. Here we studied how hepatic AMPKalpha2 isoform affects hepatic glucose production and peripheral glucose uptake in vivo. We generated mice deleted for the AMPKalpha2 gene specifically in the liver (liveralpha2KO). Liveralpha2KO mice were glucose intolerant and hyperglycemic in the fasted state. Hyperglycemia was associated with a 50% higher endogenous glucose production than in controls as assessed in vivo. We then investigated whether this increased glucose production was sensitive to insulin. Insulin, when infused at a rate inducing physiological hyperinsulinemia, totally inhibited endogenous glucose production in liveralpha2KO mice, showing that they had normal insulin sensitivity. This was confirmed in vivo by normal insulin-induced phosphorylation of Akt and transcriptional regulation of the phosphoenolpyruvate carboxykinase, glucose-6 phosphatase, and pyruvate kinase in liver during the fasted/fed transition. Leptin and adiponectin regulate hepatic glucose production, so we then infused these adipokines into liveralpha2KO mice. Neither of these adipokines regulated hepatic glucose production in mice lacking hepatic AMPKalpha2, whereas both did so in control mice. In conclusion, we show that the hepatic AMPKalpha2 isoform is essential for suppressing hepatic glucose production and maintaining fasting blood glucose levels in the physiological range. We also demonstrate that regulation of hepatic glucose production by leptin and adiponectin, but not insulin, requires hepatic AMPKalpha2 activity.

The AMP-activated kinase has been proposed to be an important intracellular energy sensor because the enzyme controls lipid and glucose oxidation. In the corresponding knockout mice, insulin-stimulated muscle glycogen synthesis and glucose tolerance are reduced. In addition, these mice excrete catecholamines in excess, suggesting that the central and autonomic nervous systems are impaired. Indeed, in the brain, fuel sensor mechanisms have been described, and recently, evidence has shown that the AMP-activated kinase could control food intake. We show in this study that the intracerebroventricular infusion of 5-aminoimidazole-4-carboxamide-1-beta-D-ribofuranoside (AICAR), a pharmacological AMP-activated kinase activator, increased insulin-stimulated muscle glycogen synthesis and insulin sensitivity during a hyperinsulinemic clamp. Similarly, we infused AICAR in the brain of fasted mice, i.e. when insulinemia was low, and showed that muscle glycogen synthesis was also increased. We then studied the effect of a cerebral infusion of the peripheral signals, i.e. insulin and glucose, known to be detected by the brain. The cerebral infusion of insulin increased muscle glycogen synthesis. This effect was blunted by the coinfusion of glucose, which induced insulin resistance. Importantly, the cerebral injections of AICAR, insulin, and glucose were associated with variations in the phosphorylation state of the AMP-activated kinase in the hypothalamus. In conclusion, our data showed for the first time that 1) the brain is sensitive to insulin and glucose for the regulation of muscle glycogen synthesis; and 2) the cerebral infusion of AICAR enhances insulin sensitivity. Although the above mechanisms are correlated with the regulation of AMP-activated kinase, the direct involvement of the enzyme in the mechanism remains to be demonstrated.