To characterize the "portal signal" during physiological glucose delivery, liver glycogen was measured in unrestrained rats during portal (Po) and peripheral (Pe) constant-rate infusion, with minimal differences in hepatic glucose load (HGL) and portal insulin between the delivery routes. Hepatic blood flows were measured by Doppler flowmetry during open surgery. Changes in hepatic glucose, portal insulin, glucagon, lactate, and free fatty acid concentrations were generally similar in either delivery except for glucagon at 4 h. Hepatic glycogen, however, increased continuously in Po and was higher than Pe at 8 and 24 h, although it decreased to the level of Pe upon the removal of Po at 8 h. There was a near-linear relationship between hepatic glycogen and HGL in either delivery, with the slope being twice as high in Po and the intercepts converging to basal HGL. The hepatic response to Po did not alter upon 80% replacement by Pe. These results suggest that negative arterial-portal glucose gradients increase the rate of hepatic glycogen synthesis against the incremental HGL in an all-or-nothing mode.hepatic glycogen synthesis; hepatic glucose load; negative arterialportal glucose gradients; nonsteady state IT HAS BEEN WELL DOCUMENTED that portal glucose delivery (Po) creates an important signal to enhance net hepatic glucose uptake (NHGU) or hepatic glycogen deposition in rodents (6, 9, 33), dogs (1-3, 14, 24, 25, 29 -31), and humans (8). Evidence has shown that NHGU is dependent on hepatic glucose load (HGL), plasma insulin concentration, and negative arterial-portal (A-P) glucose gradients. Adkins et al. (1) and have shown that a negative A-P gradient might serve as the portal signal, whereas Pagliassotti et al. (31) found that the magnitude of NHGU is dependent on the magnitude of the A-P gradients. On the other hand, Myers and colleagues (24,25) demonstrated that NHGU is linearly related to HGL, whereas it is curvilinearly related to the plasma insulin concentration, and that the hepatic response to Po results in a leftward shift in the dose-response relationship. These results have been obtained under hyperglycemic hyperinsulinemic and/or pancreatic clamp conditions, where the concentration of plasma glucose, that of insulin, or the negative A-P gradients were varied while the other two conditions were held fixed. Therefore, the relationship within these parameters remains unsettled in the nonsteady hyperglycemic state. Moore et al. (22) showed that, even in the absence of somatostatin and fixed hormone concentrations, the portal signal acts to stimulate NHGU during glucose infusion at a constant rate in dogs. In a similar experiment, we (28) also observed that there was a linear relationship between NHGU and HGL in either method of delivery, with the slope being four times steeper in Po and the HGL intercepts converging to near-basal values. We hypothesized from these results that the portal signal might increase the slope for NHGU against HGL.Therefore, the aim of the present study was to prove the hypo...