Role of basal glucagon levels in the regulation of splanchnic glucose output and ketogenesis in insulin-deficient humans

Summary. The aim of the present study was to investigate the influence of hepatic glycogen depletion and increased lipolysis on the response of splanchnic glucose output and ketogenesis to combined glucagon and insulin deficiency in normal man. Healthy subjects were studied after a 60-h fast and compared with a control group studied after an overnight fast, Net splanchnic exchange of glucose, gluconeogenic precursors, free fatty acids (FFA) and ketone acids were measured in the basal state and during intravenous infusion of somatostatin (9 μg/min) for 90–140 min (overnight fasted subjects) or for 5 h (60-h fasted subjects). During the infusion of somatostatin, euglycemia was maintained by a variable intravenous infusion of glucose. Prior to somatostatin infusion, after an overnight (12–14 h) fast, splanchnic uptake of glucose precursors (alanine, lactate, pyruvate, glycerol) could account for 26% of splanchnic glucose output (SGO) indicating primarily glycogenolysis. Somatostatin infusion resulted in a 50% reduction in both insulin and glucagon concentrations and a transient decline in SGO which returned to baseline values by 86±ll min at which point the glucose infusion was no longer necessary to maintain euglycemia. Arterial concentrations of FFA and β-OH-butyrate and splanchnic β-OH-butyrate production rose 2.5-fold, 6-fold and 7.5-fold, respectively, in response to somatostatin infusion. In the 60-h fasted state, basal SGO (0.29±0.03 mmoymin) was 60% lower than after an overnight fast and basal splanchnic uptake of glucose precursors could account for 85% of SGO, indicating primarily gluconeogenesis. Somatostatin administration suppressed the arterial glucagon and insulin concentrations to values comparable to those observed during the infusion in the overnight fasted state. SGO fell promptly in response to the somatostatin infusion and in contrast to the overnight fasted state, remained inhibited by 50–100% for 5 h. Infusion of glucose was consequently necessary to maintain euglycemia throughout the 5-h infusion of somatostatin. Splanchnic uptake of gluconeogenic precursors was unchanged during somatostatin despite the sustained suppression of SGO. Basal arterial concentration and splanchnic exchange of β-OH-butyrate were respectively 22-fold and 6- to 7-fold elevated and basal FFA concentration was 70% increased as compared to the corresponding values in the overnight fasted state. Somatostatin infusion resulted in a rise in arterial FFA concentration (25–50% in all subjects) while the arterial concentrations and splanchnic release of ketone acids (acetoacetate +β-OH-butyrate) showed a variable response, rising in three subjects and declining in two. Nevertheless, splanchnic ketone acid production in the basal state and during the somatostatin infusion correlated directly with splanchnic inflow of FFA (arterial FFA concentration × hepatic plasma flow). The variable responses in ketogenesis could thus be ascribed to variable reductions in splanchnic blood flow induced by somatostatin and as a consequence, its varying effects on splanchnic inflow of FFA. These data thus demonstrate that combined hypoglucagonemia and hypoinsulinemia induced in humans by somatostatin (a) causes a persistent rather than transient inhibition of splanchnic glucose output when liver glycogen stores have been depleted by 60-h fasting and hepatic glucose production is dependent primarily on gluconeogenesis; and (b) fails to interfere with hepatic ketogenesis so long as FFA delivery to the splanchnic bed is maintained. These findings indicate that in the face of insulin deficiency, basal glucagon levels may not be necessary to maintain hepatic glycogenolysis or ketogenesis but may be essential to maintain gluconeogenesis.