. Characterization of the portal signal in a nonsteady hyperglycemic state in conscious dogs. Am J Physiol Endocrinol Metab 284: E148-E155, 2003. First published September 11, 2002 10.1152/ajpendo.00079.2002To characterize the "portal signal" in a nonsteady hyperglycemic state, the kinetic relationship between net hepatic glucose balance (NHGB) and either hepatic glucose load (HGL) or plasma insulin level was determined during glucose infusion using a catheter technique in 36 conscious dogs. Glucose was infused intraportally (Po group) and peripherally (Pe group) at 39, 56, and 83 mol ⅐ kg Ϫ1 ⅐ min Ϫ1 over 2 h. There was a linear relationship between mean NHGB and either mean HGL or plasma insulin levels at each rate in either delivery (HGL: Po r ϭ 0.99, Pe r ϭ 0.95; insulin: Po r ϭ 99, Pe r ϭ 0.79). The threshold levels for net hepatic glucose uptake were 3.8 and 11.7 mmol/l for plasma glucose and 65 and 392 pmol/l for plasma insulin, respectively. The slope of the regression line against the abscissa was four times larger in portal than in peripheral delivery (HGL: Po 0.20 vs. Pe 0.05, P Ͻ 0.05; insulin: Po 0.19 vs. Pe 0.04, P Ͻ 0.05). These results suggest that the portal signal overrules the threshold of glucose for hepatic uptake by increasing hepatic extraction rate in a nonsteady hyperglycemic state.glucose delivery route; hepatic glucose load; insulin dependency; hepatic glucose uptake NUMEROUS STUDIES HAVE DOCUMENTED that hepatic glucose uptake is dependent on hepatic glucose load (HGL), plasma insulin level, and route of glucose delivery. However, in most studies, either hyperinsulinemic or hyperglycemic clamps were used in the presence of constant glucose, insulin, glucagon, and/or somatostatin concentrations (1, 2, 6, 14-16). It has not been demonstrated yet whether the portal signal functions similarly in a nonsteady hyperglycemic state as in the case during a clamp state. Moore et al. (13) infused glucose at a gradually increasing rate via the portal and peripheral delivery routes in dogs, with insulin and glucagon being free to change. However, arterial and portal plasma glucose profiles were largely not different with time in either delivery. In many clamp studies, peripheral delivery was combined with portal delivery to keep hepatic load constant. However, the portal signal appears not to be associated with the magnitude of arterial-portal glucose gradients at high glucose load (16). Recently, it has been reported that somatostatin blocks activation of the glucose sensor (5). Furthermore, several studies have demonstrated that pulsatile insulin concentrations have greater effects on glucose uptake than a constant concentration (3,7,11,17), and the initial acute rise in plasma insulin provokes a decrease in postprandial hyperglycemia (4, 10). Against this background, the present studies were undertaken to characterize the portal signal in a nonsteady hyperglycemic state. To this end, we determined the kinetic relationship between hepatic glucose disposal and HGL, plasma insulin levels, arterial...
