Insulin resistance induced by sucrose feeding in rats is due to an impairment of the hepatic parasympathetic nerves.

AIMS/HYPOTHESIS: A considerable proportion of whole-body insulin-stimulated glucose uptake is dependent upon the hepatic insulin-sensitising substance (HISS) in a pathway mediated by the hepatic parasympathetic nerves (HPNs). We tested the hypothesis that a high-sucrose diet leads to the impairment of the HPN-dependent component of insulin action. METHODS: We quantified insulin sensitivity using the rapid insulin sensitivity test, a modified euglycaemic clamp. Quantification of the HPN-dependent component was achieved by administration of a muscarinic receptor antagonist (atropine, 3 mg/kg). RESULTS: Insulin sensitivity was higher in standard-fed than in sucrose-fed Wistar rats (305.6+/-34.1 vs 193.9+/-13.7 mg glucose/kg body weight; p<0.005) and Sprague-Dawley rats (196.4+/-5.9 vs 95.5+/-16.3 mg glucose/kg body weight; p<0.01). The HPN-independent component was similar in the two diet groups. Insulin resistance was entirely due to an impairment of the HPN-dependent component in both Wistar rats (164.3+/-28.1 [standard-fed] vs 26.5+/-7.5 [sucrose-fed] mg glucose/kg body weight; p<0.0001) and Sprague-Dawley rats (111.7+/-9.5 vs 35.3+/-21.4 mg glucose/kg body weight; p<0.01). Furthermore, HPN-dependent insulin resistance in Sprague-Dawley rats was already evident after 2 weeks of a high-sucrose diet (28.5+/-7.6 [2 weeks], 35.3+/-21.4 [6 weeks], 17.9+/-5.4 [9 weeks] mg glucose/kg body weight) and was independent of the nature of sucrose supplementation (12.3+/-4.7 [solid] and 17.9+/-5.4 [liquid] mg glucose/kg body weight). CONCLUSIONS/INTERPRETATION: Our results support the hypothesis that insulin resistance caused by sucrose feeding is due to an impairment of the HPN-dependent component of insulin action, leading to a dysfunction of the HISS pathway.

Hepatic parasympathetic (HISS) control of insulin sensitivity determined by feeding and fasting.

In response to insulin, a hormone [hepatic insulin sensitizing substance (HISS)] is released from the liver to stimulate glucose uptake in skeletal muscle but not liver or gut. The aim was to characterize dynamic control of HISS action in response to insulin and regulation of release by hepatic parasympathetic nerves. Insulin action was assessed by the rapid insulin sensitivity test, where the index is the glucose required (mg/kg) to maintain euglycemia after a bolus of insulin. Blocking HISS release by interruption of the hepatic parasympathetic nerves by surgical denervation, atropine, or blockade of hepatic nitric oxide synthase produced similar degrees of insulin resistance and revealed a similar dynamic pattern of hormone action that began 3--4 min after, and continued for 9--10 min beyond, insulin action (50 mU/kg). HISS action accounted for 56.5 +/- 3.5% of insulin action at insulin doses from 5 to 100 mU/kg (fed). We also tested the hypothesis that HISS release is controlled by the feed/fast status. Feeding resulted in maximal HISS action, which decreased progressively with the duration of fasting.

Rapid insulin sensitivity test (RIST).

A rapid insulin sensitivity test (RIST) was recently introduced to assess insulin action in vivo (H. Xie, L. Zhu, Y.L. Zhang, D.J. Legare, and W.W. Lautt. J. Pharmacol. Toxicol. Methods, 35: 77-82. 1996). This technical report describes the current recommended standard operating procedure for the use of the RIST in rats based upon additional experience with approximately 100 tests. We describe the manufacture and use of an arterial-venous shunt that allows rapid multiple arterial samples and intravenous administration of drugs. The RIST procedure involves determination of a stable arterial glucose baseline to define the ideal euglycemic level to be maintained following a 5-min infusion of insulin, with the RIST index being the amount of glucose required to be infused to maintain euglycemia over the test period. Insulin administration by a 5-min infusion is preferable to a 30-s bolus administration. No significant difference was determined between the use of Toronto pork-beef or human insulin. Four consecutive RISTs were carried out in the same animal over 4-5 h with no tendency for change with time. The RIST index is sufficiently sensitive and reproducible to permit establishment of insulin dose-response curves and interference of insulin action by elimination of hepatic parasympathetic nerves, using atropine. This technical report provides the current recommended standard operating procedure for the RIST.

Insulin sensitivity tested with a modified euglycemic technique in cats and rats.

A new insulin sensitivity test (IST) is described using a modified euglycemic clamp in cats and rats. The IST uses the amount of glucose required to be infused to maintain euglycemia over a 30-min period in rats and 60 min in cats following a bolus administration of insulin as the index of insulin sensitivity. Glucose levels are determined at short time intervals (2-5 min), and variable glucose infusion is used to hold glucose levels within a few percentage points of the basal pre-test glucose level. A new blood sampling procedure is described that allows each IST to be carried out using a total of only 0.5 mL of blood. The IST is sensitive and allows clear insulin dose effects to be demonstrated with 100 mU/kg requiring 355.0 +/- 14.3 mg/kg over 30 min and 50 mU/kg requiring 198.7 +/- 11.1 mg/kg. Five consecutive tests were reproducibly carried out (%CV = 3.0 +/- 0.5) over a 12-hr period in the cat with insulin, glucagon, and glucose levels remaining stable prior to each IST. Glucagon and norepinephrine plasma concentrations do not change significantly during the IST. The IST is sufficiently sensitive to allow demonstration of dose-response relationships for atropine-induced insulin resistance. The IST is thus sensitive, reproducible, and able to demonstrate acute insulin resistance in anesthetized cats and rats. The test is demonstrated in fed (rats) and fasted (cats) state.

Evaluation of vascular tone in portacaval shunts comparing the index of contractility and resistance in cats.

A model of prehepatic chronic portal hypertension in cats was used to determine portacaval shunt responses to infused norepinephrine and to possible transmitter overflow into portal blood from nerves supplying the gut. Responses are compared using a new index of contractility. Four weeks after application of a slowly constricting occluder, the portal vein was completely occluded and acute experiments were carried out under pentobarbital anesthesia. Portal pressure was elevated to 15.0 +/- 0.9 mmHg and all portal flow passed through the shunts. In response to intraportal norepinephrine (0.25, 0.5 and 1.25 micrograms.min-1.kg-1) shunt resistance rose by 6% +/- 3%, 19% +/- 4% and 26% +/- 5%, respectively, whereas the index of contractility rose (by 22% +/- 8%, 46% +/- 10% and 89% +/- 20%, respectively), the distending blood pressure also rose (5% +/- 1%, 7% +/- 1% and 14% +/- 3%, respectively). The difference in percentage increase of resistance and the index of contractility is a result of the passive dilator effect of the elevated distending pressure acting on the distensible shunt vessels. Stimulation of mesenteric nerves caused the mesenteric artery to constrict, but the shunt vessels showed no effect. In conclusion, the shunt vessels respond actively to norepinephrine and passively to altered distending pressure. However, transmitter overflow from nerves supplying the intestines is unlikely to play a role in determining resistance in the shunts. Vascular resistance is affected by both active and passive effects, so that the active contractile responses are best evaluated using the index of contractility, which is not altered passively.

Distensibility of portacaval shunts in portal hypertensive cats: index of contractility model.

Complete shunting of portal blood flow through portacaval shunts was obtained using a constrictor around the portal vein to gradually produce a total occlusion. After 4 weeks, acute experiments were conducted in anesthetized cats. Blood from the femoral artery was shunted through a pump to supply and control the entire portal blood flow. As shunted portal blood flow was varied over a wide range, the portal shunt resistance showed distensibility. Decreasing portal venous pressure from 15.0 +/- 0.9 to 11.1 +/- 0.6 mmHg (1 mmHg = 133.3 Pa) resulted in elevations of resistance of 58%. The relation between the resistance (R) and the distending pressure (Pd) was a constant, the index of contractility (IC), where IC = R.Pd3. In steady state, the IC was 485 +/- 55 mmHg4.mL-1.min.kg and did not change passively in response to changes in portal blood flow. In conclusion, portacaval shunts are passively distensible, and resistance is altered as a cubic function of the distending pressure. Because resistance is altered both actively and passively, the IC should prove useful to differentiate these alternatives for evaluation of changes in portal hypertensive therapy.

Development of portacaval shunts in portal-stenotic cats.

The goal of the present study was to investigate the formation of portacaval shunts in a new experimental model of chronic portal hypertension, portal vein stenosis in the cat. The procedure gradually occluded the portal vein by use of an Ameroid constrictor around the portal vein. After 4 weeks, the portal vein was completely occluded and portal venous pressure was elevated to 15.6 +/- 0.3 mmHg (1 mmHg = 133.3 Pa) (n = 8). The hemodynamic changes did not affect the functional capacity of the liver. Latex injection was used to study the shunts. This revealed the spontaneous development of porta-systemic collaterals in all hypertensive cats, mainly between the gastrosplenic and right gastroepiploic veins and the left renal vein. Fine small branches also drained directly into the cava. The left renal vein was markedly dilated in all cats. Collateral circulation also developed between the inferior vena cava and the inferior mesenteric vein through both left internal testicular and iliolumbar veins. Some branches of the inferior mesenteric vein were connected directly to the cava. Esophageal varices in the mucosa or submucosa were not demonstrated. However, the presence of latex in the pulmonary veins and the visualization of periesophageal collaterals suggest the opening of porta-pulmonary shunts. A constant feature in all cats was the presence of a dilated azygos vein, which drained collaterals retroperitoneally and from the abdominal wall. In conclusion, an experimental model of prehepatic portal hypertension of gradual onset has been developed in cats. The formation of the porta-systemic shunts mimics other animal models and the human form of the disease.(ABSTRACT TRUNCATED AT 250 WORDS)

Hepatic blood flow distribution: consideration of gravity, liver surface, and norepinephrine on regional heterogeneity.

Blood flow distribution within the livers of cats and dogs was assessed using 15-microns microspheres injected into the hepatic artery and portal vein. Representative vertical core samples (n = 11-18) were taken from the thickest part of each liver. Heterogeneity was assessed in several ways. The difference in total flow to different lobes was greater in dogs than in cats, and in dogs, those lobes with highest portal venous flow had lowest hepatic arterial flow. Overall flow variance was very high in both species, with adjacent surface samples in a single lobe showing variance of 15-22% for both vessels. The ratio of highest to lowest flow within core samples averaged 2.1-3.4 for both vessels in both species. The hepatic arterial flow was highest to the surface 2 mm of the liver. Portal flow most often (31% of all samples) showed a pattern of highest flow to the top, graduating down to lowest flow to the bottom (dorsal side) of the vertical cores. However, this pattern appeared much more frequently in the most ventral liver lobes and very seldom in the lobes lying beneath the liver mass. Norepinephrine reduced heterogeneity. Hepatic arterial occlusion for 10 min produced minor and inconsistent reduction of heterogeneity. Rotating cats from back to front and again to back disrupted patterns of distribution but not in a way that could be interpreted as due to effects of gravity. Flow patterns changed with time. The heterogeneity of perfusion appears to be under dynamic and multiple interacting forces.

Passive autoregulation of portal venous pressure: distensible hepatic resistance.

Hepatic resistance to portal blood flow is extremely low and both the pre- and postsinusoidal resistance sites are distensible. Both isolated in situ and in vivo vascular circuitry were used in cats to demonstrate the principle of distensible resistance as a mechanism for the observation that blood flow was able to be decreased from 50 to 20 ml.min-1 x kg-1 while intrahepatic pressure decreased by only 1.4 +/- 0.2 mmHg and portal pressure by 2.0 +/- 0.4 mmHg. Presinusoidal resistance increased by 226% and hepatic venous resistance by 57%, thus accounting for passive autoregulation of portal pressure. The relation between vascular resistance and the distending blood pressure that acts on the resistance is predictable from the relationship IC = R.Pd3, where IC is the index of contractility (does not change passively, but does change with active vascular tone changes), R is vascular resistance (changes actively and passively), and Pd is distending blood pressure (estimated as the average of pressure on either side of the resistance vessels). The relatively minor effect of portal flow on portal pressure is accounted for by a combination of factors including the low basal resistance, the distensible resistance, the hepatic arterial buffer response, and hepatic blood volume compliance. By calculation of IC, the venous distensibility can be quantified and the passive effect of flow changes on portal and intrahepatic pressure determined.

Evaluation of hepatic venous resistance responses using index of contractility.

Hepatic vascular responses to 1.25 micrograms.kg-1.min-1 norepinephrine, infused into the hepatic artery, and 8-Hz nerve stimulation were monitored in anesthetized cats using a recently introduced index of contractility (IC). IC was validated in that it did not change passively in response to passive changes in portal flow or distending blood pressure, whereas the distensible venous resistance sites showed dramatic changes in resistance. Resistance is altered by both active contractile responses and passive distensibility; IC is not altered passively but is affected by changes in vascular tone. Resistance was a less sensitive index of vasoconstriction because, although the constriction increased resistance, the subsequent elevation in portal and intrahepatic pressure counteracted the constriction; the extent of active neurogenic response using resistance as the index was grossly underestimated due to venous distensibility. IC showed that pre- and postsinusoidal constriction occurred to both norepinephrine and nerves; extensive vascular escape from neurogenic constriction occurred for the portal vein so that by 5 min almost all the rise in portal pressure was due to hepatic venous constriction.

Dilazep potentiation of adenosine-mediated superior mesenteric arterial vasodilation.

The present study investigated the effects of dilazep, an inhibitor of adenosine uptake, on adenosine-mediated vasodilation in vivo. Intravenous and intraportal venous infusions of exogenous adenosine (0.04-1.0 mg/kg/min) did not recirculate to cause increases in superior mesenteric arterial conductance (SMAC) or arterial plasma adenosine levels except at the higher doses tested (0.4-1.0 mg/kg/min). After administration of dilazep, however, even low doses (0.04-0.1 mg/kg/min) of exogenous adenosine significantly increased SMAC and elevated arterial plasma adenosine concentration. The increased adenosine levels were highly correlated with the increased percentage of change of SMAC and values for Rmax and EC50 were 193.4 +/- 27.3% change of SMAC and 2.8 +/- 1.3 microM, respectively. Administration of bolus doses of 8-phenyltheophylline abolished the ability of dilazep to potentiate vasodilation, but did not affect isoproterenol-induced relaxation. Together, these results suggest that potentiation of the vasodilating effect of exogenous adenosine by dilazep is mediated through inhibition of adenosine uptake in vivo which increases the availability of plasma adenosine to act on adenosine receptors.

Index of contractility: quantitative analysis of hepatic venous distensibility.

The low-pressure resistance vessels of the splanchnic circulation are passively distensible, and changes in regional blood pressures can lead to large changes in vascular resistance. The relationship between distending blood pressure (Pd) and vascular resistance (R) is described as a constant, the index of contractility (IC) where IC = R x Pd3. IC was derived in an isolated blood-perfused liver and was confirmed in vivo for both pre- and postsinusoidal resistance sites. IC does not change passively in response to wide changes in blood flow or hepatic outflow pressure. IC is dramatically altered in response to active vasoconstriction. In vivo, the presinusoidal IC rose from a control level of 12.2 +/- 4.2 to 92.7 +/- 20.6 IC units (mmHg4.ml-1.min.kg body wt) in response to 1.25 micrograms.kg-1.min-1 norepinephrine intraportal; the postsinusoidal IC rose from 20.4 +/- 2.3 to 59.6 +/- 14.2 IC units. IC reflects resistance changes secondary to active contractile responses independent of the passive consequences of the distensible nature of the resistance sites. We suggest that these concepts can be applied to any vascular bed with distensible resistance vessels.

Effect of adenosine and glucagon on hepatic blood volume responses to sympathetic nerves.

Hepatic blood volume responses were studied in cats using in vivo plethysmography. The maximal response (Rmax) to sympathetic nerve stimulation and to infusions of norepinephrine into the hepatic artery or portal vein was similar (12-14 mL expelled per liver in 2.9-kg cats; average liver weight, 76.8 +/- 6.8 g). The ED50 for norepinephrine intraportal (0.44 +/- 0.13) and intrahepatic arterial infusions (0.33 +/- 0.08 micrograms.kg-1.min-1) were similar indicating equal access of both blood supplies to the capacitance vessels. Adenosine (2.0 mg.kg-1.min-1) did not cause significant volume changes but produced a mild (27%) suppression of Rmax due to nerve stimulation with no change in the frequency (3.4 Hz) needed to produce 50% of Rmax. Rmax tended (not statistically significant) to decrease during glucagon (1.0 micrograms.kg-1.min-1) infusion but the nerve frequency needed to produce 50% of Rmax rose to 5.6 Hz. Thus both adenosine and glucagon produced modulation of sympathetic nerve-induced capacitance responses without having significant effects on basal blood volume. Adenosine, by virtue of its marked effects on arterial resistance vessels (at substantially lower doses than those used here) and the relative lack of effect on venous capacitance vessels, may be useful for producing clinical afterload reduction without venous pooling.

Dilazep-induced vasodilation is mediated through adenosine receptors.

Administration of dilazep, an inhibitor of adenosine uptake, significantly reduced systemic arterial blood pressure and increased superior mesenteric arterial conductance without affecting the plasma adenosine levels of femoral arterial or portal venous blood. Administration of a bolus dose of 8-phenyltheophylline (8-PT), an antagonist of adenosine receptors, blocked adenosine-mediated autoregulation of the superior mesenteric artery. After the blockade of adenosine receptors by 8-PT, dilazep did not produce vasodilation. These data suggest that dilazep has a vasodilating effect in vivo that is mediated by adenosine.

Evaluation of hepatic venous balloon occluder to estimate portal pressure.

The hepatic venous balloon occluder method of estimating portal venous pressure (PVP) was evaluated in cats and dogs, during basal state and active vasoconstriction, and during passive presinusoidal resistance elevation in cats. In the dog, the balloon catheter measured a pressure not different from PVP when the balloon was inflated both in basal state and during active vasoconstriction induced by hepatic nerve stimulation, intraportal infusion of histamine or norepinephrine, regardless of whether the balloon was distal or proximal to hepatic venous sphincters. In the cat, the inflated balloon measures a pressure not different from PVP in basal state but slightly overestimated PVP during nerve stimulation or norepinephrine infusion in some protocols. Blood clots were injected intraportally in cats to produce a pure, passive presinusoidal resistance as shown by unchanged intrahepatic pressure but elevated PVP. The balloon method accurately measured PVP in this condition and clearly cannot differentiate pre- from postsinusoidal resistance sites in cats or dogs. The balloon method and the classical wedged pressure method will represent PVP when resistance is primarily postsinusoidal; they provide different measurements when resistance is presinusoidal, the wedged method representing intrahepatic pressure but the balloon method reflecting portal pressure. These differences require confirmation in human presinusoidal cirrhosis.