Overproduction of insulin in the chromium-deficient rat.

The hypothesis that the insulin secretory hyperresponsiveness observed in rats with diet-induced insulin resistance may be a basic characteristic of dietary chromium (Cr) deficiency was evaluated. Two groups of weanling rats were fed ad libitum a purified diet containing 64% sucrose, 20% casein, 5% corn oil, and the recommended levels of vitamins and minerals without added Cr. Cr-deficient (-Cr) rats were provided with distilled drinking water only, while Cr-supplemented (+Cr) rats received water containing 5 ppm Cr as CrCl3. A third group of rats fed a commercial chow diet served as sucrose controls. Effects of Cr deficiency were assessed by comparing fasting levels of glucose, insulin, and plasma lipids in blood samples collected biweekly from the -Cr and +Cr groups over a 3-month period. Both groups of rats fed the low-Cr sucrose diet developed a transient hyperinsulinemia and hyperlipidemia relative to the chow-fed control rats. There were significant effects of Cr supplementation on plasma triglycerides during the initial 2 weeks of dietary adaptation. Effects of the low-Cr diet were evaluated after the 12-week period by comparing the insulin response area and glucose clearance during a 40-minute intravenous glucose tolerance test (IVGTT). The rates of glucose clearance (KG) in -Cr and +Cr rats were similar (4.2 +/- 1.0 and 4.3 +/- 0.8%/min, respectively) and were comparable to the K(G) in chow-fed rats (4.6 +/- 0.8). In contrast, insulin secretory responses in -Cr rats were exaggerated (area, 14,083 +/- 3,399 microU/mL x min), being twofold greater (P < .05) relative to the +Cr group (6,183 +/- 864). The insulin secretory response area in chow-fed rats (7,081 +/- 408 microU/mL x min) was similar to the value in the +Cr group. These observations provide support for the hypothesis that Cr deficiency can lead to elevated insulin secretory responses to glucose.

Role of angiotensin-converting enzyme inhibition in glucose metabolism and renal injury in diabetes.

The role of angiotensin-converting enzyme (ACE) inhibition in glucose metabolism and renal injury in diabetes has been extensively investigated in diabetic humans, as well as in animal models of diabetes. Accumulated data indicate that ACE inhibitors have either no adverse effect on glucose control or insulin sensitivity or may even improve them. ACE inhibitors also appear to have neutral or positive effects on lipid metabolism. The variability of results between studies may relate to differences in experimental design, the degree of glycemia or insulin resistance, potassium balance, and dose or duration of ACE inhibitor treatment, among others. In contrast, ACE inhibitors have proved effective in limiting proteinuria and retarding renal function loss in insulin-dependent diabetes mellitus (IDDM) or non-insulin-dependent diabetes mellitus (NIDDM) patients. In rats with experimental or spontaneous diabetes, ACE inhibitors also reduce proteinuria and limit glomerular as well as tubulointerstitial damage, independent of their effects on systemic arterial pressure. How ACE inhibitors limit renal injury in diabetes is not entirely clear, but hemodynamic and nonhemodynamic mechanisms may be involved. Increasing evidence suggests that the intrarenal renin-angiotensin system (RAS) may be altered or activated in the diabetic kidney. Such activation may be specifically inhibited by ACE inhibitors and may explain the superiority of this class of agents over other antihypertensive agents in reducing proteinuria and slowing the progression of diabetic nephropathy.

Long-term effects of perindopril on metabolic parameters and the heart in the spontaneously hypertensive/NIH-corpulent rat with non-insulin-dependent diabetes mellitus and hypertension.

The spontaneously hypertensive/NIH-corpulent (SHR/N-cp) rat is a genetic model that exhibits both non-insulin-dependent diabetes mellitus (NIDDM) and hypertension. To determine the impact of long-term treatment with the long-acting angiotensin-converting enzyme (ACE) inhibitor perindopril (PE) on the glucose metabolism, lipid levels, and heart in this model, studies were performed in three groups of SHR/N-cp rats maintained on a diet containing 54% carbohydrate with 18% sucrose and 36% starch. One group of obese rats received PE (0.5 to 1.0 mg/kg body weight/d) for 3 to 4 months, a second group of obese rats received no treatment, and a third group of lean rats were used as controls. The mean systolic blood pressure (SBP) increased gradually in both untreated obese and lean rats, with lean animals showing slightly higher levels compared with untreated obese rats. By contrast, SBP was reduced to normal levels in PE-treated obese rats throughout the treatment period. Compared with lean rats, obese rats showed significantly higher body weight and fasting serum levels of glucose, insulin, total cholesterol (TC), and triglyceride (TG). However, no significant differences were observed in these metabolic parameters between PE-treated and untreated obese rats. Plasma renin activity measured at the end of the treatment period was significantly higher in PE-treated rats compared with untreated obese and untreated lean rats. The mean heart weight and left ventricular weight, expressed in absolute terms or indexed to body weight, were significantly lower in PE-treated versus untreated obese and untreated lean rats. To further determine whether glucose metabolism is directly affected by PE treatment, in vitro glycogen synthesis was evaluated in isolated soleus muscles obtained from three additional groups of animals. The basal rate of muscle glycogen synthesis was significantly lower in obese compared with lean rats (P < .05), but did not differ between PE-treated and untreated obese rats. Maximal insulin-stimulated glycogen synthesis increased threefold in PE-treated obese rats, but this increase did not differ from the increases observed in untreated obese and lean rats. In conclusion, the present study shows that long-term PE treatment in obese SHR/N-cp rats with NIDDM and hypertension effectively controlled systemic arterial pressure and resulted in a significant reduction in left ventricular weight. However, these favorable effects of PE were not associated with significant improvement in glucose tolerance, hyperinsulinemia, and hyperlipidemia in this model. PE also had no direct stimulatory effects on either basal or insulin-mediated glycogen synthesis in the isolated soleus muscle of obese rats, perhaps because of the severe insulin-resistant state of the animals. Our results support the clinical observations that antihypertensive therapy with ACE inhibitors has neutral effects on glucose metabolism and insulin sensitivity in patients with combined hypertension and NIDDM.

Dietary chromium decreases insulin resistance in rats fed a high-fat, mineral-imbalanced diet.

The effects of chromium (Cr) supplementation on diet-induced insulin resistance produced by feeding a high-fat, low-Cr diet were studied in rats to ascertain the role of Cr in insulin resistance. Wistar male rats were maintained for 16 weeks after weaning on a basal diet containing 40% lard, 30% sucrose, and 25% casein by weight and adequate vitamins and minerals without added Cr (-Cr). Fasting levels of insulin, glucose, and triglycerides and the responses during an intravenous glucose tolerance test (IVGTT) were compared as indices of insulin resistance and the effectiveness of dietary Cr. IVGTTs and blood sampling for data analyses were performed over a 40-minute period after IV glucose injection (1.25 g/kg body weight) in overnight-fasted animals under pentobarbital anesthesia (40 mg/kg body weight). All animals were normoglycemic (-Cr, 109 +/- 3 mg/dL; +Cr, 119 +/- 5), with fasting insulin levels elevated in the -Cr group (65 +/- 10 microU/mL) versus the +Cr group (31 +/- 4 microU/mL). Increases in plasma triglycerides in the -Cr group were not significant. Following glucose injection, the rate of glucose clearance was lower in the -Cr group (1.74 +/- 0.22 v2.39 +/- 0.11%/min), and 40-minute glucose areas in the -Cr group tended to be higher than in the +Cr group. The insulin response to glucose injection was 20% higher in the -Cr group. Forty-minute plasma triglyceride areas were lower in +Cr rats (875 +/- 62 v 1,143 +/- 97 mg/dL.min in -Cr rats). These data demonstrate that the insulin resistance induced by feeding a high-fat, nutrient-stressed diet is improved by Cr.

Perindopril ameliorates glomerular and renal tubulointerstitial injury in the SHR/N-corpulent rat.

We compared the effects of long-term treatment with the angiotensin-converting enzyme inhibitor perindopril and triple therapy (hydrochlorothiazide, reserpine, and hydralazine) on the metabolic and renal features in the SHR/N-corpulent (cp) rat, a genetic model of non-insulin-dependent diabetes mellitus and hypertension. Obese male SHR/N-cp rats (4 to 6 weeks old) were fed a 54% carbohydrate diet containing 18% sucrose and 36% starch. After 2 months on the diet, rats were assigned to one of three groups: one group (n=8) received perindopril (PE); the second group (n=8) received triple therapy (TT); and the third group (n=8) did not receive therapy. Treatment was maintained for 3 to 4 months. Body weight, food intake, and fasting levels of serum glucose and insulin did not differ among the three groups. Control rats exhibited progressive proteinuria in parallel with the rise in systolic blood pressure (SBP). Both PE and TT equally lowered SBP to normal levels and reduced proteinuria in treated rats. However, the reduction of proteinuria was greater and more sustained with PE than with TT (P<.05), whereas the effect of TT on proteinuria was delayed. Plasma renin activity was increased in PE and TT rats compared with control rats (P<.02). Semiquantitative analysis of renal lesions showed that the percentage of glomeruli with mesangial expansion and sclerosis and the tubulointerstitial score (an index of severity of tubulointerstitial lesions, namely tubular atrophy, inflammatory cellular infiltrates, and interstitial fibrosis) was reduced in both PE and TT rats. However, the reduction of glomerulosclerosis and tubulointerstitial lesions was greater in PE than in TT rats (P<.01). The percentage of glomerular sclerosis was positively correlated with the severity score of tubulointerstitial lesions (r=.60, P<.01). We conclude that PE is more effective than TT in halting the progression of proteinuria in the SHR/N-cp rat with non-insulin-dependent diabetes mellitus and hypertension. The antiproteinuric effect of PE is associated with significant reduction in glomerulosclerosis and tubulointerstitial lesions, independent of the effect of treating hypertension.

Chromium improves insulin response to glucose in rats.

The effects of chromium (Cr) supplementation on insulin secretion and glucose clearance (KG) during intravenous glucose tolerance tests (IVGTTS) were assessed in rats with impaired glucose tolerance due to dietary Cr deficiency. Male Wistar rats were maintained after weaning on a basal low-Cr diet containing 55% sucrose, 15% lard, 25% casein. American Institute of Nutrition (AIN)-recommended levels of vitamins, no added Cr, and an altered mineral content as required to produce Cr deficiency and impaired glucose tolerance. The Cr-supplemented group ([+Cr] n = 6) were provided with 5 ppm Cr as CrCl3 in the drinking water, and the Cr-deficient group ([-Cr]n = 5) received purified drinking water. At 12 weeks on the diet, both groups of rats were hyperinsulinemic (+Cr, 103 +/- 13; -Cr, 59 +/- 12 microU/mL) and normoglycemic (+Cr, 127 +/- 7; -Cr, 130 +/- 4 mg/dL), indicating insulin resistance. After 24 weeks, insulin levels were normal (+Cr, 19 +/- 5; -Cr, 21 +/- 3 microU/mL) and all rats remained normoglycemic (+Cr, 124 +/- 8; -Cr, 131 +/- 6 mg/dL). KG values during IVGTTS were lower in -Cr rats (KG = 3.58%/min) than in +Cr rats (KG = 5.29%/min), correlating with significantly greater 40-minute glucose areas in the -Cr group (P < .01). Comparisons of 40-minute insulin areas indicated marked insulin hyperresponsiveness in the -Cr group, with insulin-secretory responses increased nearly twofold in -Cr animals (P < .05). Chromium deficiency also led to significant decreases in cyclic adenosine monophosphate (cAMP)-dependent phosphodiesterase (PDE) activity in spleen and testis (P < .01). In these studies, Cr deficiency was characterized by both beta-cell hypersecretion of insulin and tissue insulin resistance that were associated with decreased tissue levels of cAMP PDE activity.

Decreased hepatic insulin extraction precedes overt noninsulin dependent (Type II) diabetes in obese monkeys.

Many obese middle-aged rhesus monkeys (Macaca mulatta) spontaneously develop noninsulin dependent diabetes mellitus (NIDDM). Basal hyperinsulinemia and increased stimulated plasma insulin levels are associated with this obesity and precede the onset of overt diabetes. The present studies sought to determine the relative contributions of enhanced insulin secretion and of reduced insulin clearance to this early obesity-associated hyperinsulinemia. Direct simultaneous measurement of portal and jugular vein insulin levels in two normal monkeys showed a constant rate of hepatic insulin extraction of 56+/-3% over the range of peripheral insulin levels from 351+/-113 to 625+/-118 pmol/L. In 33 additional monkeys ranging from normal to diabetic, basal C-peptide levels were examined as an indicator of beta-cell secretion and the molar ratio of plasma C-peptide to insulin (C/I ratio) under basal steady state conditions calculated as an index of hepatic insulin extraction. Well in advance of overt diabetes, there was a progressive decline of 67% in the apparent hepatic insulin extraction rate in association with increased obesity and plasma insulin levels. Basal insulin levels and hepatic insulin extraction returned toward normal in monkeys with impaired glucose tolerance and in those with overt diabetes. We conclude that reduced insulin disposal, probably due to reduced hepatic extraction of insulin, in addition to increased beta-cell activity, contributes to the development of basal hyperinsulinemia in obese rhesus monkeys progressing toward NIDDM. In addition, in overt diabetes, normal hepatic insulin extraction in the presence of limited beta-cell secretion may exacerbate the hypoinsulinemic state.

Peripheral neuropathy in diabetic monkeys.

Peripheral neuropathy is a significant complication of human diabetes and a source of morbidity. Appropriate experimental models may aid in understanding its pathogenesis and in developing therapeutic strategies. We sought to determine whether spontaneously diabetic obese adult monkeys developed peripheral neuropathy and whether it occurred early or late in relation to the onset of hyperglycemia. We studied nerve conduction in both motor (peroneal, median, and ulnar) and sensory (median and ulnar) nerves in 13 adult male rhesus monkeys, 4 overtly diabetic and 9 nondiabetic (mean age 21 +/- 2 and 16 +/- 2 yr, respectively, NS; mean fasting plasma glucose 14.5 +/- 3.4 and 4.4 +/- 0.6 mM, P = .001). The diabetic animals had significantly reduced motor conduction velocities and prolonged F-wave latencies. Motor-evoked amplitudes did not differ. In the diabetic monkeys, nerve conduction times were increased in motor fibers, which could be identified as early as 2 yr after the onset of hyperglycemia. These abnormalities are similar to those seen in humans and suggest further study of these animals as a primate model of human diabetic neuropathy.

Insulin clearance and microsomal glutathione-insulin transhydrogenase in perfused livers of fed and fasted rats.

Clearance of exogenous insulin measured in perfused livers from rats fed ad lib or fasted X 24 or X 48 h was correlated with changes in activity and distribution of the insulin-degrading enzyme glutathione-insulin transhydrogenase measured in microsome fractions, post-perfusion. For comparison with endogenous insulin removal (Endocr. Res. Commun. 7: 231, 1980), a single-pass perfusion mode was used and clearance of insulin at levels (less than or equal to 15 ng/ml) typically observed in perfused rat liver-pancreases during glucose stimulation was studied. Similar to the endogenous data, exogenous insulin removal followed an ogival pattern during fasting. In the fed state, clearance was relatively low, corresponding to a hepatic extraction of approximately 29%. Insulin extraction increased nearly 2-fold after a 24 h fast to approximately 48% (p less than .01), declining to approximately 30% (p less than .025) when fasting was prolonged (X 48 h). At portal insulin concentrations greater than 8 ng/ml (approximately 200 microUnits/ml), clearance tended to decrease in all 3 nutritional states, with apparent saturation of the insulin capturing mechanism being strongest in the 24 h fasted state. In conjunction with these changes in whole organ insulin removal, GSH-insulin transhydrogenase nonlatency, viz., nonlatent activity in intact microsomes relative to total activity in disrupted microsomes, did not change during the first 24 h of fasting; whereas the proportion of nonlatent activity was significantly decreased (p less than .01) after 48 h. Homogenate activity remained essentially constant during the initial fasting period, and declined by approximately 16% (p less than .01) after 48 h of fasting.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of a short-term fast on hepatic microsomal glutathione-insulin transhydrogenase.

The effect of nutritional state on the hepatic insulin degrading enzyme glutathione-insulin transhydrogenase (GIT) was assessed by comparing the distribution of GIT activity between its nonlatent and latent forms in fractionated liver microsomes from ad lib fed (n = 11) and overnight fasted (n = 11) rats. In fed state microsomes, treatment with the membrane disrupting agent phospholipase-A2 (PLA2) over a range of PLA2 concentrations (less than or equal to 2.0 micrograms/ml) caused biphasic release of GIT with a peak activity of 651 +/- 58 U/mg microsomal protein (n = 11) occurring at PLA2 = 1.0 microgram/ml. In total liver microsomes from fasted animals, GIT release in response to PLA2 was sigmoidal over the entire range of PLA2 concentrations, with a plateau of activities (450 U/mg microsomal protein) occurring at PLA2 greater than or equal to 0.75 microgram/ml. Peak activities (478 +/- 88 U/mg prot., n = 11, PLA2 = 1.0 microgram/ml) were 30% lower as compared to the fed state (p less than .05). In untreated (intact) microsomes from fed rat liver nonlatent activity was 126 +/- 8 U/mg protein, representing 19.9 +/- 1.2% of the total GIT activity. In contrast, nonlatent activity measurable in suspensions of intact microsomes from fasted rat liver (110 +/- 6 U/mg) expressed as a % of total activity was significantly increased (p less than .05) being 23.3 +/- 1.1%. Similar fasting-induced changes were also apparent in isolated smooth microsomes but not in rough membrane preparations.(ABSTRACT TRUNCATED AT 250 WORDS)

Effects of glucagon on hepatic microsomal glucose-6-phosphatase in vivo.

The smooth endoplasmic reticulum (SER) has been implicated in glycogen deposition and depletion. It has been suggested that SER proliferation plays a role in elevated glucose release during rapid glycogenolysis (Striffler et al., Am. J. Anat. 160: 363, 1981). In these studies, the role of SER in glucagon-stimulated hepatic glucose release was examined by assessing changes in microsomal glucose-6-phosphatase (G-6-Pase) and membrane cholesterol to phospholipid ratios. In control fed rats given 6 i.p. injections of glucagon 350 micrograms/injection) at hourly intervals, percentage hepatic glycogen decreased nearly 30 fold, with liver homogenate G-6-Pase (U/mg protein) increasing 50% (p less than .02 relative to vehicle-injected controls) from .055 +/- .003 at 0h (n = 12) to .081 +/- .004 at 6h (n = 11). The increase in homogenate G-6-Pase was reflected in the isolated SER fraction by a 48% rise (p less than .01 relative to controls) in activity from a 0h value of .210 +/- .010 (n = 10) to a peak activity of .310 +/- .027 U/mg microsomal protein at 5 h (n = 12). G-6-Pase also increased in the rough endoplasmic reticulum (RER) reaching .242 +/- .023 U/mg protein at 4h (n = 14), but then declining to .209 +/- .019 U/mg protein at 6 h (n = 11). The changes in G-6-Pase were accompanied by alterations in the ratio of microsomal cholesterol to phospholipid, which decreased by 50% (p less than .002) in both RER and SER fractions signifying changes in membrane lipid environment. Ultrastructural analysis revealed a marked reduction in hepatic glycogen and conspicuous presence of elements of the SER in regions of the cytoplasm where glycogen was or presumably had been located.

Morphological and biochemical observations on hepatic glycogen metabolism in mice on a controlled feeding schedule. II. Streptozotocin-diabetic mice.

Morphological aspects of hepatic glycogen metabolism in streptozotocin (SZ) diabetic mice on a controlled feeding cycle (6 hr fed, 18 hr fasted) were studied and correlated with plasma glucose and insulin levels at various time intervals after feeding. Hepatic glycogen was low at initiation of feeding (approximately 0.05%) and increased to a maximum of 4.54 +/- 0.30% (N = 8) as compared to 6.90 +/- 0.20% (N = 12) in normal animals. Plasma glucose levels were similar to those of normal mice at initiation of feeding (80 +/- 5 mg/dl) but much higher during the feeding period (diabetic: 540 +/- 15 mg/dl, normal: 150 +/- 10 mg/dl). At the end of the feeding period, plasma glucose levels rapidly declined, reaching lower than normal levels. In contrast to insulin responses to feeding in normal animals, plasma insulin levels in SZ-diabetic mice remained very low, never exceeding 16 muU/ml. At maximum hepatic glycogen deposition, light microscopic studies showed atypical patterns of glycogen distribution with periportal cells having generally smaller-than-normal glycogen masses. Ultrastructural studies indicated that these cells contained more abundant quantities of smooth endoplasmic reticulum (SER) than is characteristically seen. The aberrant distribution patterns of glycogen observed in the diabetic mice provided morphological evidence for the proposal that the SER is involved in hepatic glycogenesis, with insulin deficiency resulting in abnormal functioning of the organelle.

Morphological and biochemical observations on hepatic glycogen metabolism in mice on a controlled feeding schedule. I. Normal mice.

Changes in hepatocyte morphology were correlated with chemically measured liver glycogen, blood glucose, and plasma insulin levels in control-fed mice (6 hr fed, 18 hr fasted) sacrificed at various time intervals after initiation of a 6-hr meal. At initiation of feeding hepatic glycogen was low(0.05%) but deposition proceeded rapidly, reaching a maximum of 6.99 +/- 0.13% by the sixth hour. Glycogen was depleted during the subsequent fasting period, reaching the prefeeding levels by 24 hr. A relative hyperglycemia (140-192 mg/100/ ml) predominated during all stages of glycogen deposition and depletion until the 21st hour. Plasma insulin levels were maximum during feeding (63 +/- 7 microU/ml, 3 hr) with mild hyperinsulinemia (insulin greater than 16 microU/ml) occurring during glycogen depletion (9-21 hr). Histochemical determinations (PAS) showed lobular patterns of hepatic glycogen which correlated with chemically measured glycogen levels. Six hours after initiation of feeding, periportal cells showed intensely stained masses of glycogen while centrilobular cells showed relatively diffuse staining. At 24 hours after initiation of feeding (18 hr of fasting), no significant staining was observed in the hepatocytes. Ultrastructurally, during all stages of glycogen deposition and depletion, centrilobular cells were characterized by the presence of dispersed glycogen particles with elements of smooth endoplasmic reticulum (SER) between the particles, while periportal cells showed dense glycogen deposits with SER restricted to the periphery of the glycogen masses.

Morphological and biochemical observations on hepatic glycogen metabolism in genetically diabetic (db/db) mice.

Changes in hepatocyte glycogen morphology, liver glycogen, and blood glucose and insulin levels were studied at various time points after initiation of feeding in 10 week old genetic diabetic (db/db) mice (n = 8) and their lean littermates adapted to a controlled feeding regimen (6 h fed - 18 h fast). In spite of hyperinsulinemic (insulin greater than or equal to 97.60 +/- 88 uU/ml) and hyperglycemic (glucose greater than or equal to 437.1 +/- 49.5 mg/dl) conditions favoring glycogen synthesis, feeding-induced hepatic glycogen deposition was less efficient in the diabetic mice. An apparent decreased ability to mobilize hepatic glycogen during fasting and maintenance of relatively high fasting liver glycogen concentrations were however attributable to the anti-glycogenolytic effects of high blood glucose and insulin concentrations. Histochemical determinations (PAS) on livers of db/db mice showed typical lobular patterns of glycogen distribution during deposition and depletion of the polysaccharide. At stages of maximum glycogen (6 h after initiation of feeding), periportal cells displayed intensely stained masses of glycogen with centrilobular cells staining more diffusely. Ultrastructural studies revealed smooth endoplasmic reticulum (SER) in close association with glycogen during its deposition and depletion. The preservation of normal glycogen morphology and SER-glycogen ultrastructure indicates that functioning of the SER in hepatic glycogen metabolism is normal in midly diabetic (db/db) mice.

Kinetics of insulin clearance by the liver in perfused liver-pancreas.

The relationship between hepatic insulin extraction velocity and portal plasma insulin concentration was assessed using in situ perfused liver-pancreas preparations from rats fasted 24 or 48 hours. During hepatic transit of the first phase of glucose-stimulated insulin secretion, hepatic extraction velocities followed first order kinetics for all insulin concentrations observed (less than 1 to 17 ng/ml). During the second phase of insulin release, saturation of the insulin removal process occurred at insulin concentrations between 12 and 14 ng/ml. At non-saturating insulin levels, livers from rats fasted 24 hours clear insulin at about 4.0 ml/min corresponding to an extraction ratio of 52%. In contrast, livers of liver-pancreases from 48 hour fasted rats clear insulin at 2.4 ml/min with an extraction ratio of 31%.

Effect of fasting on insulin removal by liver of perfused liver-pancreas.

Using in situ perfused rat liver-pancreas preparations, the hepatic clearance of biphasically released endogenous insulin was studied using livers of fed and fasted rats. Hepatic insulin extraction was determined by subtracting the mean endogenous insulin output found in the posthepatic vena cava effluent of liver pancreas preparations from the amount of insulin found in the prehepatic portal effluent of perfused pancreas preparations. All pancreas perfusions and liver-pancreas perfusions were subjected to the same plasma glucose stimulus (250 mg/dl) for 40 min. In the fed state, no significant hepatic extraction of insulin occurred. The biphasic pattern of insulin appearance in the posthepatic vena cava effluent, which ranged from less than 1 to 21 ng/ml, was unaffected by passage through the liver. In contrast, when rats were fasted for 24 h, approximately 50% of the insulin released into the portal vein was removed by their livers during one transhepatic passage. After a 48-h fast, insulin extraction capacity decreased relative to the capacity observed for livers of 24-h fasted rats.