Influence of uremia and hemodialysis on the turnover and metabolic effects of glucagon

Glycaemic control in diabetic nephropathy

To investigate the quality of glycaemic control that is achievable in diabetic patients with persistent proteinuria and asymptomatic but declining renal function three matched groups of patients were studied. The first comprised diabetics with proteinuria receiving continuous subcutaneous insulin infusion; the second, diabetics without proteinuria receiving continuous subcutaneous insulin infusion; and the third, diabetics with proteinuria receiving conventional insulin treatment. Glycaemic control in patients receiving continuous subcutaneous insulin infusion was shown to be appreciably worse during the daytime in diabetics with proteinuria than in diabetics without proteinuria, although greatly superior to that in diabetics with proteinuria receiving conventional insulin treatment. The loss of glycaemic control in patients with proteinuria receiving continuous subcutaneous insulin infusion probably occurred as a response to daytime hypoglycaemia and a consequent reduction in the proportion of the total insulin dose given prandially. Difficulty in controlling blood glucose concentrations may be a factor in the failure of intensified insulin regimens to influence the progression of diabetic renal disease.

Effects of glucagon on plasma amino acids

The effects of glucagon deficiency and excess on plasma concentrations of 21 amino acids were studied in six normal human subjects for 8 h. During glucagon deficiency, produced by intravenous infusion of somatostatin (0.5 mg/h) and insulin (5 mU/kg per h), amino acid concentration (sum of 21 amino acids) rose from 2,607 +/- 76 to 2,922 +/- 133 microM after 4 h (P less than 0.025). The largest increases occurred in lysine (+26%), glycine (+24%), alanine (+23%), and arginine (+23%) concentrations. During glucagon excess produced by intravenous infusion of somatostatin (0.5 mg/h), insulin (5 mU/kg per h), and glucagon (60 ng/kg per h), amino acid concentration decreased from 2,774 +/- 166 to 2,388 +/- 102 microM at 8 h (P less than 0.01). The largest decreases occurred in citrulline (-37%), proline (-32%), ornithine (-30%), tyrosine (-23%), glycine (-20%), threonine (-21%), and alanine (18%) concentrations. Urinary urea nitrogen and total nitrogen excretions were lower during glucagon deficiency than during glucagon excess (3.1 +/- 0.2 vs. 6.3 +/- 2.3 g/8 h, P less than 0.05 and 4.8 +/- 1.0 vs 7.0 +/- 2.6 g/8 h, respectively, P less than 0.05). Biostator-controlled euglycemic glucagon deficiency was produced in four normal subjects for 4 h to eliminate possible effects of changes in glucose concentration on amino acids. Amino acid concentration (sum of 18 amino acids) increases occurred in arginine (+42%), alanine (+28%), glutamine (+25%), and glycine (+16%) concentrations. The data show that small changes (-66 pg/ml and +50 pg/ml) in basal glucagon concentrations cause plasma amino acid concentrations to change in opposite directions. The finding that urinary excretion of nitrogen and urea nitrogen was greater during glucagon excess than during glucagon deficiency suggested alterations in the rate of gluconeogenesis from amino acids as one mechanism by which glucagon controls blood amino acid levels.

Induction of insulin resistance in normal adipose tissue by uremic human serum

An incubation of uremic human serum with normal rat adipose tissue will make the subsequently isolated adipocytes less responsive to insulin. To examine the extent of insulin resistance, we obtained sera from nondiabetic, uremic patients, who had not undergone dialysis therapy. The sera were then dialyzed (3500 molecular-weight cutoff) for 18 hr against a defined culture medium to eliminate possible in vitro effects of altered levels of end-product metabolites, electrolytes, and metabolic substrates. After an incubation of epididymal fat tissue from normal rats, for 3 hr with the dialyzed sera (50% vol/vol), cells were isolated and washed. The insulin stimulation of 14C-glucose (0.2 mM) incorporation to 14CO2 and total lipids was significantly reduced in the adipocytes pretreated with sera from 19 of the 29 uremic patients. Although elevated in the uremic patients, the sera levels of insulin, and parathyroid and growth hormones were not correlated to insulin resistant activity. Furthermore, incubation of adipose tissue for 3 hr with insulin, glucagon, or PTH did not produce resistance. The uremic sera reduced glucose utilization equally at 0.2 and 50 mM glucose, suggesting that the insulin resistance was induced additionally at a site distal to the glucose transport system. However, the concentration of insulin (22 microunits/ml) required for half-maximal stimulation of glucose metabolism was not altered by pretreatment with uremic serum. Also, neither the isoproterenol-stimulated lipolysis nor the inhibition of this cellular event was influenced by pretreatment with uremic sera.(ABSTRACT TRUNCATED AT 250 WORDS)

The effects of chronic uremia on glucagon binding and action in isolated rat hepatocytes

Increased liver sensitivity to glucagon has been proposed to play a role in the complex metabolic state of chronic uremia. In order to assess this possibility at the cellular level, we studied basal and glucagon-stimulated alpha-aminoisobutyric acid (AIB) uptake, glucagon binding, and glucagon degradation in isolated hepatocytes from chronic uremic and pair-fed and ad libitum-fed control rats. The uremic rats were euglycemic and hyperglucagonemic when compared with their controls. The basal rate of AIB uptake was enhanced in hepatocytes from both the uremic and pair-fed control rats. Hepatocytes from ad libitum-fed control animals responded significantly to glucagon at concentrations from 1 X 10(-11) to 1 X 10(-7) mol/L, and those from pair-fed control animals at concentrations from 1 X 10(-8) to 1 X 10(-7) mol/L. Hepatocytes from the uremic rats were unresponsive to glucagon with regard to AIB uptake. 125I-labeled glucagon binding was increased in the uremic rats. This increase of glucagon binding appears to be the results of an increase in the number of binding sites rather than a consequence of a change in binding affinity or decreased glucagon degradation. In conclusion, our data are not supportive of the hypothesis that there exists in uremia an increased sensitivity to glucagon in the liver. The uremic liver is resistant to glucagon with regard to AIB uptake. Despite the high level of circulating immunoreactive glucagon, hepatocytes from uremic rats did not show the expected "down regulation" of their 125I-labeled glucagon binding sites. These studies emphasize the primary role of post-binding events in the regulation of glucagon action and binding.

Effects of glucose-containing peritoneal-dialysis solutions on rates of lipogenesis in vivo in the liver, brown and white adipose tissue of chronic uraemic rats

Chronic uraemic rats had decreased food intake, and this was accompanied by decreased weight of the epididymal fat-pads and interscapular brown adipose tissue. Normal rats whose food intake was restricted to an amount similar to that of the uraemic rats showed similar decreases in weight of the adipose-tissue depots. In addition, the food-restricted rats had decreased liver weight compared with normal or uraemic rats. The basal rate of lipogenesis was decreased in liver and epididymal fat-pads of food-restricted and uraemic rats and in interscapular brown adipose tissue of uraemic rats. Administration of a low-glucose-containing (1.36%) peritoneal-dialysis solution slightly increased lipogenesis in liver of uraemic rats, but had no significant effect in epididymal fat-pads. For brown fat, the rate of lipogenesis was increased in normal, food-restricted and uraemic groups, but the values for the last group were 4-5-fold lower than for the food-restricted or control groups. A high-glucose-containing (3.86%) peritoneal-dialysis solution gave similar rates of lipogenesis in liver, epididymal fat-pads and brown fat of all three groups, but for brown fat moderately uraemic rats showed a considerably lower rate of lipogenesis than did mildly uraemic rats. The basal plasma insulin concentration was lower in the food-restricted (50%) and uraemic (70%) groups than in the control group. The low-glucose peritoneal-dialysis solution increased plasma insulin to control values in the food-restricted rats, but had no significant effect on plasma insulin in the uraemic rats, despite a significant increase in blood glucose in this group. It is concluded that there is an impairment of the lipogenic response to intraperitoneal glucose loads in interscapular brown adipose tissue of uraemic rats, and that this is not due to the accompanying decrease in food intake. The hypoinsulinaemia may be an important factor. The possible relevance of this finding to the obesity observed in some uraemic patients treated by peritoneal dialysis with glucose-containing solutions is discussed.

Streptozotocin diabetes results in increased responsiveness of adipocyte lipolysis to glucagon

Adipocytes from streptozotocin-diabetic rats are approximately 50-times more sensitive to the lipolytic action of glucagon. This change is only perceived in the presence of a small quantity of adenosine deaminase which itself has little effect on basal lipolysis. Insulin treatment restores glucagon sensitivity to normal.

Hyperglucagonemia following cisplatin treatment

These studies were initiated to determine (1) if cisplatin (cis-DDP)-induced hyperglucagonemia is related to decreased hormone degradation, (2) the relationship between impaired kidney function associated with cis-DDP nephrotoxicity and hyperglucagonemia, and (3) the contribution of cis-DDP-induced hyperglucagonemia to disturbances in glucose metabolism in male F-344 rats. Administration of 5 or 7.5, but not 2.5, mg/kg cis-DDP iv increased fasting plasma immunoreactive glucagon (IRG) concentrations. Hyperglucagonemia following cis-DDP treatment was characterized by an increase in the biologically active or true pancreatic form of IRG as well as an increase in an extrapancreatic component. cis-DDP treatment (5 mg/kg) resulted in a prolonged half-life and a reduced rate of plasma disappearance of exogenous glucagon. Reducing cis-DDP nephrotoxicity, via mannitol pretreatment, resulted in a significant reduction in total, true pancreatic, and extrapancreatic plasma IRG. Other nephrotoxicants, such as glycerol or gentamicin, also resulted in hyperglucagonemia, indicating that the effects of cis-DDP on glucagon metabolism are also characteristic of other nephrotoxicants and, therefore, may be secondary to kidney toxicity. Despite marked hyperglucagonemia following cis-DDP treatment, neither severe fasting hyperglycemia nor increased hepatic and renal gluconeogenic enzyme activity was apparent in treated animals. This apparent discrepancy cannot be attributed to glucagon resistance at the target tissue level since cis-DDP-treated animals responded appropriately to exogenous glucagon. These results indicate that hyperglucagonemia following cis-DDP treatment (1) may be related to decreased glucagon degradation associated with impaired renal function and (2) does not markedly disrupt glucose homeostasis.

Role of the kidneys in elimination of glucagon, insulin, secretin and somatostatin in the rat

Immunoreactive plasma glucagon and secretin in the rat was elevated 48 hours after nephrectomy and ureteral ligation. Since kidneys obstructed by ureteral ligation were unable to remove glucagon and secretin from the blood, renal handling of glucagon and secretin must include glomerular filtration. Insulin and somatostatin levels were significantly elevated 48 hours after nephrectomy, but not after ureteral ligation, indicating partial uptake from peritubular capillaries.

Gastric inhibitory polypeptide (GIP) and insulin release in response to oral and intravenous glucose in uremic patients

Eight patients with advanced renal failure of long duration were studied 1 day after hemodialysis. A 50 g oral glucose load (OGTT) and an intravenous glucose infusion (IVGI), giving the same plasma glucose profile as the OGTT, were carried our in order to study the relation between Gastric Inhibitory Polypeptide (GIP) plasma levels after oral glucose and the insulin release during OGTT and IVGI. The plasma GIP increase during OGTT was significantly elevated compared to a group of eight healthy volunteers. The insulin potentiation during OGTT in relation to GIP was significantly depressed in the uremic patients. It is proposed that a factor of intestinal origin is released during intake of carbohydrates, which blocks the B-cell response to the combined glucose-GIP stimulus. Alternatively, the concentrations of plasma GIp measured have included GIp fragments without insulin releasing capability.

Role of glucagon as a contributor to glucose intolerance in acute and chronic uremia

It has been suggested that increased sensitivity to glucagon may contribute to glucose intolerance in uremia. In order to evaluate this possibility systematically, we have assessed the effect of glucagon on hepatic glucose outflow, formation of cAMP, and activation of adenylate cyclase by livers obtained from acutely and chronically uremic rats and their respective sham operated controls. Glucagon infused at rates of 6 ng/min/kg rat resulted in minimal and equivalent increases in hepatic glucose outflow and cAMP accumulation when livers from acutely uremic and control rats were perfused for 30 min. However, at glucagon infusion rates of 18 ng/min/kg, glucose efflux from perfused livers of acutely uremic rats was significantly reduced (p less than 0.001) compared to perfused livers of control rats (4.64 +/- .9 vs 12.7 +/- 2.4 mumol/g liver) and cAMP accumulation was also significantly lower (p less than 0.01) (1352 +/- 222 vs 3100 +/- 348 pmol/g liver). Basal adenylate cyclase activity of hepatic membranes obtained from uremic and control rats was similar, and was stimulated by glucagon concentrations ranging from 10(-8) to 10(-6) at equivalent rates in both groups. In livers from chronically uremic rats, glucagon infused at rates of 6 ng/kg/min significantly increased hepatic glucose outflow (32.5 +/- 6.9 mumol/g liver). However this was not greater than that of control animals (37.6 +/- 9.2). Furthermore, cAMP accumulation was significantly lower (p less than .02) in chronically uremic rats than in controls, and activation of adenylate cyclase by glucagon was similar in both groups. These findings indicate that glucagon does not increase glucose efflux, cAMP accumulation or enhance activation of adenylate cyclase by isolated perfused livers from either acutely or chronically uremic rats. Thus, glucose intolerance in uremic rats does not appear to be due to increase hepatic glucose output resulting from increased sensitivity to glucagon.

Studies on the role of the liver and splanchnic tissues in the production of carbohydrate intolerance in uremia

The potential contribution of the splanchnic tissues to the carbohydrate intolerance of uremia was studied in fasted, partially nephrectomized rats. The livers of sham operated (C) and partially nephrectomized (Nx) rats were perfused with physiologic concentrations of potential gluconeogenic substrates using a nonrecirculating perfusion apparatus. Glucose release was slightly greater in the livers of Nx rats as compared to C rats. The portal vein concentrations of the potential gluconeogenic precursors were not different in the two groups. Moreover, there were no differences in the net hepatic extraction of alanine, glutamine or glutamate between the two groups of rats. There was also no difference in the production of glucose from U14C alanine. The livers of Nx rats, however, demonstrated less net extraction of lactate and released greater concentrations of betahydroxybutyrate. The increased release of glucose by livers of Nx rats may be at least partially due to their greater hepatic glycogen content.

The acute metabolic effects of glucagon and its interactions with insulin in forearm tissue

The acute effects of glucagon (mol. wt. 3500) and its interactions with insulin were studied in the forearm during eight studies in seven normal, post-absorptive males. The protocol consisted of a 2 h baseline, 1 h glucagon perfusion (mean glucagon increment, 691 +/- 50 pg/ml), 1 h perfusion of both insulin and glucagon (mean insulin increment of 105 insulin and glucagon (mean insulin increment of 105 /- 13 mU/l) and a 30 min recovery period. Simultaneous arterial (A), deep venous (DV), and superficial venous (SV) blood samples were obtained at 30 min intervals. Perfusion of glucagon resulted in a decrease in (A-DV) non-esterified fatty acids of -0.128 +/- 0.057 mmol/l (n = 7, p less than 0.05) and (A-SV) non-esterified fatty acids of -0.081 +/- 0.36 mol/l (n = 7, p less than 0.05), as well as a change in deep compartment uptake of glycerol after 60 min of -0.044 +/- 0.019 mumol/min/100 ml of forearm tissue (n=6, p less than 0.05), indicating increased lipolysis. There was also a decrease in net glucose uptake as reflected by a change in (A-Dids of -0.081 +/- 0.36 mol/l (n = 7, p less than 0.05), as well as a change in deep compartment uptake of glycerol after 60 min of -0.044 +/- 0.019 mumol/min/100 ml of forearm tissue (n=6, p less than 0.05), indicating increased lipolysis. There was also a decrease in net glucose uptake as reflected by a change in (A-Dids of -0.081 +/- 0.36 mol/l (n = 7, p less than 0.05), as well as a change in deep compartment uptake of glycerol after 60 min of -0.044 +/- 0.019 mumol/min/100 ml of forearm tissue (n=6, p less than 0.05), indicating increased lipolysis. There was also a decrease in net glucose uptake as reflected by a change in (A-DV) of -0.24 +/- 0.09 mmol/l (n = 7, p less than 0.025) and (A-SV) of 0.10 +/- 0.05 mmol/l (n = 7, p less than 0.05). There was also a net decrease in deep arteriovenous differences of potassium in six of seven subjects. Insulin levels, similar to those found after a meal, rapidly reversed the effects of glucagon on non-esterified fatty acid, glucose and potassium. These effects persisted throughout the recovery period.

Hepatic metabolism of glucagon in the dog: contribution of the liver to overall metabolic disposal of glucagon

The hepatic extraction (HE) of glucagon (G) and insulin (I) was measured in 27 dogs, using peripheral infusion of the hormones following elimination of endogenous secretion by pancreatectomy (Px) or somatostatin (S) infusion. HE(G) was 22.5 +/- 1.7%, and HE(I) was 45.1 +/- 3%. HE(G) in seven Px dogs was 27.9 +/- 4.2%, not significantly different from the value of 20.6 +/- 1.6% in 20 S-infused dogs, with corresponding values for HE(I) being 44.9 +/- 6 and 46.0 +/- 3.6%, respectively, suggesting that S does not affect HE of either hormone. HE of endogenous G (22.1 +/- 2.8%) was similar to that of exogenously infused G (19.1 +/- 1.9). HE(G) was nonsaturable in the physiologic and pathophysiologic range of plasma G levels, but there was evidence of saturability in the pharmacologic range. Comparison of simultaneously measured parameters of I and G metabolism indicated independence of the metabolic processes of these two islet hormones, despite distinct similarities in their overall patterns of metabolic disposal. Metabolic clearance rates (MCR) for G and I were 12.6 +/- 0.8 and 19.5 +/- 1.0 ml . kg-1 . min-1, while simultaneously measured hepatic HE rates were 4.2 +/- 0.3 and 8.1 +/- 0.6 ml . kg-1 . min-1, respectively. MCR(G) was independent of arterial G levels. Half-life of infused G and I was 5.5 +/- 0.5 and 4.1 +/- 0.3 min, respectively. The liver accounted for 34.7 +/- 2.4% of the MCR(G) and 42.0 +/- 2.9% of MCR(I). The liver is thus an important site for G removal. However, HE(G) varies widely in different animals, and it is therefore not possible to predict portal vein G concentrations or G secretion rates from G levels in peripheral vessels.

Insulin resistance in uremia

Tissue sensitivity to insulin was examined with the euglycemic insulin clamp technique in 17 chronically uremic and 36 control subjects. The plasma insulin concentration was raised by approximately 100 microU/ml and the plasma glucose concentration was maintained at the basal level with a variable glucose infusion. Under these steady-state conditions of euglycemia, the glucose infusion rate is a measure of the amount of glucose taken up by the entire body. In uremic subjects insulin-mediated glucose metabolism was reduced by 47% compared with controls (3.71 +/- 0.20 vs. 7.38 +/- 0.26 mg/kg . min; P less than 0.001). Basal hepatic glucose production (measured with [3H]-3-glucose) was normal in uremic subjects (2.17 +/- 0.04 mg/kg . min) and suppressed normally by 94 +/- 2% following insulin administration. In six uremic and six control subjects, net splanchnic glucose balance was also measured directly by the hepatic venous catheterization technique. In the postabsorptive state splanchnic glucose production was similar in uremics (1.57 +/- 0.03 mg/kg . min) and controls (1.79 +/- 0.20 mg/kg . min). After 90 min of sustained hyperinsulinemia, splanchnic glucose balance reverted to a net uptake which was similar in uremics (0.42 +/- 0.11 mg/kg . min) and controls (0.53 +/- 0.12 mg/kg . min). In contrast, glucose uptake by the leg was reduced by 60% in the uremic group (21 +/- 1 vs. 52 +/- 8 mumol/min . kg of leg wt; P less than 0.005) and this decrease closely paralleled the decrease in total glucose metabolism by the entire body. These results indicate that: (a) suppression of hepatic glucose production by physiologic hyperinsulinemia is not impaired by uremia, (b) insulin-mediated glucose uptake by the liver is normal in uremic subjects, and (c) tissue insensitivity to insulin is the primary cause of insulin resistance in uremia.

Removal of IR-GIP by the kidneys in man, and the effect of acute nephrectomy on plasma GIP in rats

In a group of 10 patients undergoing routine cardiac catheterization, mean plasma gastric inhibitory polypeptide (GIP) was significantly lower (p = 0.002) in blood drawn from the renal vein than in blood drawn from the hepatic vein, the right atrium, the femoral vein, and the femoral artery, demonstrating the removal of immunoreactive GIP by the kidneys. In a group of eight nephrectomized rats, mean plasma GIP was significantly higher 30 and 60 min after start of intraduodenal glucose infusion (p = 0.03 and p = 0.002, respectively) as compared with eight control rats, which probably reflects reduced removal of GIP in the nephrectomized group.

Carbohydrate metabolism and uraemia-mechanisms for glycogenolysis and gluconeogenesis

Disturbances of carbohydrate metabolism during acute uraemia are characterized by the degradation of liver and muscle glycogen with a simultaneous activation of hepatic gluconeogenesis. After binephrectomy, the substitution of essential amino acids and keto analogues stimulate liver, but not skeletal muscle glycogen synthesis. Serine proves to be an optimal substrate for liver gluconeogenesis and muscle glycogen generation under acute uraemic conditions. Propranolol does not influence glycogenolysis of skeletal muscle in acutely uraemic rats. During starvation, acute uraemia leads to an increase of total carbohydrate content as well as of glycogen and glucose concentrations in heart muscle Alterations in carbohydrate contents are not observed in the kidney after ureter ligation. Enhanced glycogenolysis of skeletal muscle and liver during acute uraemia may be due to activation of phosphorylase kinase caused by the increased serum concentrations of various hormones (glucagon, catecholamines, parathormone) as well as free proteolytic activity, an increase of intracellular Ca2+-concentration and finally by alterations in the structure of contractile proteins.

Glucagon deficiency and hyperaminoacidemia after total pancreatectomy

The first goal of this study was to investigate whether totally pancreatectomized patients are glucagon deficient and if so, to what degree. Immunoreactive glucagon (IRG) concentrations in peripheral plasma of nine pancreatectomized patients were not significantly different from those of 10 normal controls as measured by two antisera (30-K and RCS-5) both detecting the COOH-terminal portion of the molecule and one (RCS-5) postulated to be specific for pancreatic glucagon. Plasma from six of nine pancreatectomized patients were fractionated over Sephadex G-50 and IRG was measured with both antisera in the column eluates. Using 30-K, 80.8 +/- 9% of the IRG eluted within the void volume. This material was rechromatographed on Sephadex G-200 and found to have an apparent mol wt of approximately 200,000. Only 18.3 +/- 9% eluted in the IRG3500 region. IRG3500 was significantly reduced in pancreatectomized patients as compared to normal controls (49 +/- 9 vs. 18 +/- 9 pg/ml, P less than 0.05). Using RCS-5, all IRG (corresponding to 20 +/- 6 pg/ml of plasma) eluted in the IRG3500 region. The second goal of this study was to investigate the effects of chronic glucagon deficiency on plasma amino acids. In the nine pancreatectomized patients studied, postabsorptive plasma concentrations of serine, alanine, arginine, glycine, threonine, citrulline, alpha-aminobutyrate, and tryosine were significantly elevated compared to values obtained from 20 normal controls. Physiological glucagon increments produced in two pancreatectomized patients by infusion of glucagon (6.25 and 8.0 microgram/h, respectively) resulted in normalization of the hyperaminoacidemia within 22 h. We conclude (a) that pancreatectomized patients are partially glucagon deficient because of diminished basal as well as diminished stimulated glucagon secretion; (b) that fasting concentrations of certain glucogenic amino acids are elevated in pancreatectomized patients probably as result of reduce; hepatic gluconeogenesis; and (c) that the RCS-5 antiserum is not "pancreatic glucagon" specific.

Effect of prolonged uremia on insulin metabolism by isolated liver and muscle

As the prolonged metabolic clearance rate of insulin in chronic uremia cannot be entirely explained by impaired removal and degradation of insulin by the kidney, we set out to determine whether prolonged uremia depresses other major sites of insulin degradation. The study was conducted with livers and skeletal muscle obtained from normal control rats and uremic rats 4 weeks after 80% nephrectomy. Despite a significant difference between renal function in the control and uremic rats (BUN, 18 vs. 46 mg/dl), there was no significant difference in the clearance of insulin by isolated uremic or control livers perfused with a bloodless medium. Similarly, the 125I-insulin degrading activity of liver homogenates was not depressed by uremia. In contrast, binding and degradation by uremic liver cell membranes was significantly reduced to 58% and 85% of the controls, respectively. Degradation by homogenates of skeletal muscle and by intact epitrochlaris muscle was significantly less in uremics than in controls. These results indicate that chronic uremia depresses skeletal muscle insulin degradation but not hepatic insulin removal or degradation despite a decrease in insulin binding and degradation by liver plasma membranes. It thus appears that depression of insulin degradation by muscle may contribute to the prolonged insulin metabolic clearance rate seen in chronic uremia. Furthermore, it is possible that the impaired binding of insulin to liver membranes may play a role in the insulin resistance of uremia.

Glucagon metabolism in normal subjects and in cirrhotic patients before and after portasystemic venous shunt surgery

The effect of portasystemic shunt surgery on basal immunoreactive glucagon (IRG) levels, metabolic clearance rate (MCR) and t 1/2 for glucagon decay, and basal systemic delivery rate (BSDR) of glucagon was investigated in paired studies in ten cirrhotic subjects. The degree of hepatocellular dysfunction and extent of portasystemic venous shunting was also recorded. Basal IRG levels were highest in the post-shunt (mean +/- SEM, 382 +/- 73 pg/ml) as compared to the pre-shunt (213 +/- 27 pg/ml; P less than 0.05) cirrhotic and control (53 +/- 13 pg/ml; P less than 0.005) groups. The MCR of glucagon was similar in control (13.0 +/- 1.3 ml/kg/min) and pre-shunt cirrhotic patients (13.3 +/- 1.7 ml/kg/min) but was significantly (P less than 0.02) decreased in the post-shunt cirrhotics (7.6 +/- 1.3 ml/kg/min). The t 1/2 for glucagon decay was similar in the control and cirrhotic groups. The BSDR, an estimate of pancreatic A cell secretion, was increased four-fold (P less than 0.01) in the pre-shunt (3042 +/- 454 pg/kg/min) and post-shunt (2518 +/- 535 pg/kg/min) cirrhotic groups, as compared to controls (750 +/- 244 pg/kg/min). It is concluded that (a) in the presence of cirrhosis, the magnitude of portasystemic shunting is important in determining the degree of hyperglucagonaemia; (b) in preshunt cirrhotics raised basal IRG levels are principally due to A cell hypersecretion of glucagon whereas in post-shunt cirrhotics riased IRG levels reflect both A cell hypersecretion and delayed clearance of glucagon; and (c) acute shunting of splanchnic venous blood away from the liver reduces the clearance of glucagon, suggesting that, in man, the liver contributes to the clearance of circulating glucagon.

Impact of hemodialysis on the abnormal glucose and alanine kinetics of chronic azotemia

Rates of alanine and glucose turnover and precursor-product interrelationships were determined in patients on chronic hemodialysis and in matched controls using simultaneous primed injection-continuous infusions of [U-14C] alanine and [2-3h] glucose. In eight chronically dialyzed patients studied before their first dialysis of the week, glucose turnover was 866 +/- 120 micromole/min (mean +/- SE); after their last dialysis of the week, glucose turnover was 880 +/- 63 micromole/min. These rates were 35% (p less than 0.05) and 37% (p less than 0.01) greater than rates observed in ten normal volunteers (642 +/- 28.3 micromole/min). Fasting glucose and insulin levels in dialyzing patients were unchanged from normal. Alanine turnover was increased predialysis (318 +/- 55.2 micromoles/min; p less than 0.01) and postdialysis (248 +/- 32.4 micromole/min; p less than 0.01) as compared to normal (168 +/- 14.3 micromole/min). In patients pre- and postdialysis, gluconeogenesis from alanine was increased to 34.6 +/- 10.9 micromole/min (p less than 0.05) and 39.0 +/- 6.33 micromole/min (p less than 0.05) compared to 20.9 +/- 1.63 micromole/min in normal subjects. We conclude that neither acute nor chronic hemodialysis corrects the increased glucose and alanine production and utilization and gluconeogenesis observed in chronic renal failure.

Pattern of growth hormone response to insulin, arginine and haemodialysis in uraemic children

Plasma growth hormone (GH) concentrations after insulin and arginine stimulation were estimated in 11 dialyzed and 6 non-dialyzed children with chronic renal failure. Twenty healthy children served as controls. Plasma GH peak concentration and estimation of the total area under the plasma GH concentration-time curve by the trapezoidal rule were used to evaluate results. Elevated basal GH levels and an exaggerated response to the stimuli were seen in several of the patients. The causes of the abnormal GH secretion and the role of high GH levels in carbohydrate intolerance are discussed. No consistent pattern was seen in GH secretion during haemodialysis without glucose in the dialysate. In children undergoing haemodialysis with a fluid containing glucose, plasma GH fell considerably.

Diabetic management by insulin infusion during major surgery

In five insulin-requiring, uremic diabetic patients undergoing renal transplantation, we infused insulin intravenously at a low rate to maintain plasma glucose levels between 100 and 200 mg/100 ml. In those patients receiving 100 mg or more of prednisone per day and 5 per cent dextrose solution, the hourly infusion rate was determined from tthe following equation: insulin (U) = plasma glucose value divided by 100. When prednisone was not given or when the patient was thin, the ratio became: plasma glucose value divided by 150. Results were compared with those of nineteen similar transplant patients treated with conventional subcutaneous insulin therapy during surgery, and significantly better glucose control was achieved with the low dosage, intravenous infusion.

Dietary carbohydrate and metabolism of ingested protein

Six normal people were fed a lean beef meal while on a normal diet and after seven days on a very low carbohydrate (greater than 25 g/day) 2000 kcal/day diet. After carbohydrate restriction, the protein-induced rise in branched chain aminoacids was 40-50% greater than the rise after the control diet. Intravenous leucine also produced a 40% greater rise in plasma-leucine after carbohydrate restriction. Three days of fasting exaggerated protein-induced increases in plasma branched-chain aminoacids by 55-77%. Hypocaloric, pure carbohydrate refeeding restored the branched-chain aminoacid responses to normal. Severe carbohydrate restriction thus leads to increased accumulation of plasma branched-chain aminoacids after protein feeding which is at least in part due to reduced utilisation of these aminoacids.

Pathogenesis of glucose intolerance in uremia

The pathogenesis of glucose intolerance in uremia was examined with the glucose clamp technique. Hyperglycemic clamp (n = 8): The plasma glucose concentration is acutely raised and maintained at 125 mg/dl above basal levels. Under these steady state conditions the glucose infusion rate, M, equals the amount of glucose metabolized: Predialysis M averaged 4.23 +/- 0.36 mg/kg/min and increased to 7.71 +/- 0.43 postdialysis (p less than 0.001). The plasma insulin response predialysis was 90 +/- 20 microU/ml and decreased to 80 +/- 23 microU/ml following dialysis. Consequently the M/l ratio, a measure of tissue sensitivity to insulin, increased by 80% +/- 25% (p less than 0.001) but still remained less than controls (p less than 0.01). Euglycemic insulin clamp (n = 10): The plasma insulin concentration is acutely raised by 100 microU/ml and the plasma glucose concentration is held constant at the basal level. Predialysis both M (3.37 +/- 0.36 mg/kg/min) and M/l (3.56 +/- 0.33 mg/kg/min per microU/ml X 100) were significantly less than controls (p less than 0.01). Postdialysis both M and M/l increased significantly (p less than 0.01) to a mean that was not significantly different from controls. Basal hepatic glucose production (n = 6), 2.15 +/- 0.09 mg/kg/min, was similar to controls and fell (87% +/- 4%) normally during the insulin clamp. In five uremic subjects in wom insulin binding to monocytes was measured, there was no correlation with tissue sensitivity to insulin (M/l). Significant abnormalities in both growth hormone and glucagon physiology were present in uremic individuals, but no correlation with either the presence or degree of glucose intolerance was demonstrable. In conclusion, glucose intolerance is universally present in uremic subjects and results primarily from peripheral tissue insensitivity to insulin. Insulin secretion is usually enhanced in an attempt to compensate for this insulin resistance but in occasional subjects uremia also inhibits beta cell sensitivity to glucose. Hepatic glucose production is unaffected by uremia. The lack of correlation between insulin binding and tissue sensitivity to insulin suggests that the cellular mechanism accounting for the insulin resistance is probably the result of a defect in intracellular metabolism or in the glucose transport system.

Metabolic and nutritional factors in children with renal insufficiency

Uremia is associated with a decrease in muscle and adipose tissue mass and a low weight-for-height ratio. These findings are related to dietary deficiencies in uremia--particularly energy deficiency and to metabolic disorders characteristic of uremia. These latter have features of an exaggerated catabolic state which may be modified by other stresses, e.g. short starvation or high-protein diets. Recommendations for diet therapy for children with uremia are of limited value because of the lack of definitive studies. At present, diet should be adequate in energy to improve nitrogen balance and weight gain commensurate with age. There may be advantages to using a protein:energy ratio in the diet that is lower than the ratio used in conventional diets.

Abnormal amino acid and protein metabolism in uremia

The many alterations in amino acid and protein metabolism in renal failure are often poorly defined, and the available data concerning them are usually descriptive. Nonetheless, certain factors play an important role in the altered amino acid and protein metabolism of uremia. These include malnutrition caused by poor nutrient intake, loss of nutrients during dialysis, and abnormal metabolism of nutrients. Other factors include uremic toxins, superimposed catabolic illnesses, endocrine disorders, and the reduced capacity of the failing kidney to synthesize or degrade certain hormones, amino acids, peptides, and small proteins. These aberrations have complex interrelationships which sometimes potentiate each other. It is possible that the administration of sufficient quantities of energy, vitamins, and minerals, as well as the dietary manipulation of protein, amino acid and ketoacid intake may improve the metabolism of amino acids and proteins. Vitamin B6 and zinc have special requirements that may affect protein or amino acid metabolism.

Glucose intolerance in uremia. Quantification of pancreatic beta cell sensitivity to glucose and tissue sensitivity to insulin

The relative contributions of impaired insulin secretion and of tissue insensitivity to insulin to the carbohydrate intolerance of uremia were investigated in 10 chronically uremic subjects. Two types of glucose-clamp experiments were performed in each patient before and after 10 wk of thrice weekly hemodialysis. In both types the blood glucose concentration was maintained at a constant level by the periodic adjustment of a variable glucose infusion with a negative feedback formula.Hyperglycemic clamp. The blood glucose concentration was acutely raised and maintained 125 mg/dl above basal levels for 2 h. Since the glucose concentration was held constant, the glucose infusion rate is an index of glucose metabolism (M). After dialysis M increased in all patients from an average of 4.23 to 6.30 mg/kg body wt per min (P < 0.001). The plasma insulin responses (I) both pre- and postdialysis were biphasic with an early burst within the first 2-5 min, followed by a phase of gradually increasing insulin concentration. After dialysis the plasma insulin response diminished slightly. Consequently, the M/I ratio, an index of tissue sensitivity to endogenous insulin, increased postdialysis in all subjects by an average of 92% (P < 0.01). Euglycemic clamp. The plasma insulin concentration was acutely raised and maintained by a primecontinuous insulin infusion. The blood glucose concentration was held constant at the basal level by a variable glucose infusion as above. M/I again is a measure of tissue sensitivity to insulin (exogenous) and increased in all patients postdialysis by an average of 57% (P < 0.01). In two patients hepatic glucose production was measured with tritiated glucose during the euglycemic clamp and declined by 84% predialysis. A similar decrease (82%) was observed postdialysis. Thus, both the hyperglycemic and euglycemic clamp techniques demonstrated tissue insensitivity to insulin to be the dominant carbohydrate defect in uremia. The surprising apparent lack of consistency in the change in beta cell response postdialysis is explained by the strong inverse correlation between beta cell sensitivity to glucose and tissue sensitivity to insulin (r = -0.920; P < 0.001). Those individuals who showed the most striking improvement in tissue sensitivity to insulin actually decreased their serum insulin response to hyperglycemia; those whose improvement in tissue sensitivity was more modest showed increases in beta cell responses.

Hyperlipoproteinemia in experimental chronic renal insufficiency in the rat

Lipid metabolism was studied in experimental uremia. Uremic (U) rats were compared with sham-operated, pair-fed (PF) controls and with ad-lib-fed (AL) controls. In U animals, fasting glucose concentrations were normal, immunoreactive serum insulin (IRI) levels were decreased, and immunoreactive glucagon levels were increased. A significant increase in the serum concentration of all lipid classes was observed: triglycerides were elevated 10-fold above the values in PF and AL controls; phospholipids, twofold; total cholesterol, threefold; and free cholesterol, sixfold. Cholesterol concentration was increased in beta- and pre-beta-lipoproteins and even more so in alpha- and pre-alpha-lipoproteins. There was an increase in the ratio of free cholesterol/total cholesterol. The fatty acid composition of serum lipoproteins was unchanged. Concomitantly, in liver tissue, there was no change in lipid content (triglyceride, cholesterol) and fatty acid composition. These findings argue against glucose- or insulin-mediated changes in hepatic de novo fatty acid synthesis, chain elongation, or poly-desaturation. In U animals, the HMG-CoA-reductase activity of liver microsomes was slightly, but not significantly, reduced as was tritiated water incorporation into cholesterol in isolated perfused liver preparations. In adipose tissue, there was a decrease in triglyceride content. The results provide evidence against insulin-mediated hepatic overproduction as a major cause of hyperlipoproteinemia in this model of experimental renal insufficiency and point to peripheral under-utilization of lipoproteins.

Abnormal carbohydrate metabolism in chronic renal failure. The potential role of accelerated glucose production, increased gluconeogenesis, and impaired glucose disposal

To delineate the potential role of disordered glucose and glucose-precursor kinetics in the abnormal carbohydrate metabolism of chronic renal failure, alanine and glucose production and utilization and gluconeogenesis from alanine were studied in patients with chronic compensated renal insufficiency and in normal volunteers. With simultaneous primed injection-continuous infusions of radiolabeled alanine and glucose, rates of metabolite turnover and precursor-product interrelationships were calculated from the plateau portion of the appropriate specific activity curves. All subjects were studied in the postabsorption state. In 13 patients with chronic renal failure (creatinine = 10.7+/-1.2 mg/100 ml; mean+/-SEM), glucose turnover was found to be 1,035+/-99.3 mumol/min. This rate was increased 56% (P = 0.003) over that observed in control subjects (664+/-33.5 mumol/min). Alanine turnover was 474+/-96.0 mumol/min in azotemic patients. This rate was 191% greater (P = 0.007) than the rate determined in control subjects (163+/-19.4 mumol/min). Gluconeogenesis from alanine and the percent of glucose production contributed by gluconeogenesis from alanine were increased in patients with chronic renal failure (192% and 169%, respectively) as compared to controls (P < 0.05 for each). Alanine utilization for gluconeogenesis was increased from 40.2+/-3.86 mumol/min in control subjects to 143+/-39.0 mumol/min in azotemic patients (P < 0.05). The percent of alanine utilization accounted for by gluconeogenesis was not altered in chronic renal insufficiency. In nondiabetic azotemic subjects, mean fasting glucose and immunoreactive insulin levels were increased 24.3% (P = 0.005) and 130% (P = 0.046), respectively.These results in patients with chronic renal failure demonstrate (a) increased glucose production and utilization, (b) increased gluconeogenesis from alanine, (c) increased alanine production and utilization, and (d) a relative impairment to glucose disposal. We conclude that chronic azotemia is characterized by increased rates of glucose and glucose precursor flux and by a relative impairment to glucose disposal. These findings may suggest an underlying hepatic and peripheral insensitivity to the metabolic action of insulin in patients with chronic renal insufficiency.

Glucagon metabolism in the rat

The renal handling of the biologically active glucagon component (the 3,500-mol wt fraction of immunoreactive glucagon [IRG]) and the contribution of the kidney to its overall peripheral metabolism were studied in normal and uremic rats. The metabolic clearance rate of glucagon was 31.8 +/- 1.2 ml/min per kg in normal animals and was diminished by approximately one-third in each of three groups of rats with compromized renal function: 22.3+/-1.6 ml/min per kg in partially (70%) nephrectomized; 22.9+/-3.3 ml/min per kg in bilaterally ureteral ligated; and 23.2+/-1.2 ml/min per kg in bilaterally nephrectomized animals. In normal rats the kidney contributed 30% to the overall metabolic clearance of the hormone and the renal extraction of endogenous and exogenous glucagon was similar, averaging 22.9+/-1.6% and was independent of plasma IRG levels over a wide range of arterial concentrations. The remnant kidney of partially (70%) nephrectomized animals continued to extract substantial amounts (16.6+/-4.2%) of the hormone, but accounted for only 8% of the total peripheral catabolism of IRG. In the two groups of animals with filtering kidneys, renal glucagon uptake was linearly related to its filtered load and could be accounted for by glomerular filtration and tubular reabsorption. However, the kidneys of animals with both ureters ligated (renal extraction of inulin = 3.2+/-1.8%) and hence virtual absence of glomerular filtration, continued to extract 11.5+/-1.9% of the renal arterial glucagon, contributing by 9% to its overall metabolic clearance, indicating that IRG uptake occurs also from the post glomerular capillaries.

Norepinephrine: hormone and neurotransmitter in man

To determine whether norepinephrine could subserve a hormonal as well as a neurotransmitter function, norepinephrine was infused for 60 min into each of five normal young men in doses of 0.1, 0.5, 1.0, 2.5, and 5.0 microgram/min. After infusion, the plasma norepinephrine concentration fell with a mean (+/-SD) half-time of 2.4 +/- 0.7 min. The mean (+/-SD) norepinephrine metabolic clearance rate was 3,070 +/- 200 ml/min. The calculated basal plasma norepinephrine production rate was 0.7 microgram/min. The blood pressure and circulating glycerol, acetoacetate, beta-hydroxybutyrate, and glucose (increased) and the heart rate and circulating insulin, lactate, pyruvate, and alanine (decreased) exhibited highly significant parabolic relationships with the steady-state plasma norepinephrine concentrations. However, norepinephrine levels in excess of 1,800 pg/ml were required to produce hemodynamic and/or metabolic effects. Thus, under usual conditions, the biologic actions of norepinephrine can be attributed only to its sympathetic neurotransmitter function. Plasma norepinephrine concentrations do at times exceed 1,800 pg/ml during exercise and during major acute illness. Thus, under conditions of stress, norepinephrine may subserve a hormonal, as well as a neurotransmitter, function.

Glucagon binding and adenylate cyclase activity in liver membranes from untreated and insulin-treated diabetic rats

To investigate the role of hepatic glucagon receptors in the hypersensitivity to glucagon observed in insulin-deprived diabetics, liver plasma membranes were prepared from control rats and from streptozotocin-induced diabetic rats some of whom were treated with high-dose and low-dose insulin. The untreated diabetic animals exhibited hyperglycemia, weight loss, hypoinsulinemia, and hyperglucagonemia. High-dose insulin treatment (2 U Protamine-zinc-insulin/100 g per day) resulted in normoglycemia, normal weight gain, mild hyperinsulinemia, and return of glucagon levels toward base line. The low-dose (1 U protamine-zinc-insulin/100 g per day) insulin-treated diabetic group demonstrated chemical changes intermediate between the untreated and the high-dose insulin-treated animals. In liver plasma membranes from the untreated diabetic rats, specific binding of (125)I-glucagon was increased by 95%. Analysis of binding data suggested that the changes in glucagon binding were a consequence of alterations in binding capacity rather than changes in binding affinity. Furthermore, in the untreated diabetic rats, both basal and glucagon (2 muM)-stimulated adenylate cyclase activity were twofold higher than in controls. In the high-dose insulin-treated diabetic rats, glucagon binding and basal and glucagon-stimulated adenylate cyclase activity were normalized to control values, whereas low-dose insulin treatment resulted in changes intermediate between control and untreated diabetic rats. In contrast to glucagon-stimulated adenylate cyclase activity, fluoride-stimulated adenylate cyclase activity was similar in all groups of rats. Liver plasma membranes from untreated and insulin-treated diabetic animals degraded (125)I-glucagon to the same extent as control rats. The specific binding of (125)I-insulin in the untreated diabetic animals was 40% higher than in control rats. In low-dose insulin-treated diabetic rats, insulin binding was not significantly different from that of control rats, whereas in the high-dose insulin-treated group in whom plasma insulin was 70% above control levels, insulin binding was 30% lower than in control rats. These findings suggest that alterations in glucagon receptors may contribute to the augmented glycemic and ketonemic response to glucagon observed in insulin-deprived diabetics.

The role of the liver in glucagon metabolism

Total plasma immunoreactive pancreatic glucagon (IRG) was measured in samples taken simultaneously from the proximal portal vein and superior vena cava of 26 healthy rats. The portal-peripheral ratio of IRG was 2.80+/-0.25, the portal-peripheral difference (Delta) 124+/-15 pg/ml, and percentage extraction 58+/-3. Gel filtration of paired portal and peripheral vein samples showed that reduction in the 3,500-dalton IRG component (glucagon) in peripheral samples accounted for almost all the differences, there being minimal and inconsistent changes in the high molecular weight (>40,000) fraction. The portal-peripheral ratio of the 3,500-dalton glucagon was 5.24+/-1.10, the portal-peripheral difference 130+/-33 pg/ml, and the percentage extraction 81+/-5. To study the transhepatic differences in the 9,000-dalton "proglucagon-like" material, the experiment was repeated in nine rats 24 h after bilateral nephrectomy, a procedure which increases plasma levels of this fraction. The portal-peripheral ratio for plasma IRG in these rats was 1.48+/-0.12, the portal-peripheral difference 140+/-29 pg/ml, and percentage extraction 28+/-5. Gel filtration revealed no consistent differences between portal and peripheral concentrations of the 9,000- and >40,000-dalton components, which comprised 40 and 13%, respectively, of the mean IRG level of 492+/-35 pg/ml. In contrast, there were marked differences between portal and peripheral levels of the 3,500-dalton component the ratio being 3.42+/-0.63, the portal-peripheral difference 182+/-32 pg/ml, and percentage extraction 64+/-5. Similar studies in a healthy dog, in which species there are significant circulating levels of the 9,000-dalton IRG component, confirmed the selective hepatic extraction of the 3,500-dalton fraction. We conclude that the various IRG fractions are metabolized differently by the liver, and that portal-peripheral ratios based on direct assay of plasma IRG will vary depending on the percentage glucagon immunoreactivity in each fraction; the greater the combined contribution of fractions other than the 3,500-dalton component to total plasma IRG, the lower will be the ratio. Because of the heterogeneity of circulating IRG and significant differences in the metabolism of its various components, gel filtration of plasma samples is necessary for precise quantitation of the hepatic uptake of each particular fraction.

Glucagon and insulin binding to liver membranes in a partially nephrectomized uremic rat model

To investigate the role of glucagon and insulin receptor binding in the glucagon hypersensitivity and insulin resistance which characterize the glucose intolerance of uremia, liver plasma membranes were prepared from control rats (blood urea nitrogen [BUN] 15+/-1 mg/100 ml, creatinine 0.7+/-0.2 mg/100 ml), and from 70% nephrectomized rats (BUN 30+/-2 mg/100 ml, creatinine 2.2+/-0.2 mg/100 ml), and from 90% nephrectomized rats (BUN 46+/-3 mg/100 ml, creatinine 4.20+/-0.7 mg/100 ml), 4 wk after surgery. As compared to controls, the 90% nephrectomized rats had significantly higher levels of plasma glucose (95+/-4 vs. 125+/-11 mg/100 ml), plasma insulin (28+/-9 vs. 52+/-11 muU/ml), and plasma glucagon (28+/-5 vs. 215+/-18 pg/ml). Similar, but less marked, elevations were observed in the 70% nephrectomized animals. In liver plasma membranes from nephrectomized rats, specific binding of (125)I-glucagon was increased by 80-120%. Furthermore, glucagon (2 muM)-stimulated adenylate cyclase activity in nephrectomized rats was twofold higher than in controls. In contrast, fluoridestimulated adenylate cyclase activity was similar in both groups of rats. In marked contrast to glucagon binding, specific binding of (125)I-insulin to liver membranes from nephrectomized rats was reduced by 40-50% as compared to controls. Data analysis suggested that the changes in both glucagon and insulin binding are a consequence of alterations in binding capacity rather than changes in affinity. Liver plasma membranes from nephrectomized rats degraded (125)I-glucagon and (125)I-insulin to the same extent as control rats. THESE RESULTS DEMONSTRATE THAT: (a) the 70 and 90% nephrectomized rats simulate the hyperglycemia, hyperinsulinemia, and hyperglucagonemia observed in clinical uremia; (b) in these animals specific binding of glucagon to liver membranes is increased and is accompanied by higher glucagon-stimulated adenylate cyclase activity; and (c) specific binding of insulin is markedly decreased. These findings thus provide evidence of oppositely directed, simultaneous changes in glucagon and insulin receptor binding in partially nephrectomized rats. Such changes may account for the hypersensitivity to glucagon and may contribute to resistance to insulin observed in the glucose intolerance of uremia.

Renal effects on serum gastric inhibitory polypeptide (GIP)

Fasting and meal-stimulated serum immunoreactive gastric inhibitory polypeptide (GIP) concentrations were measured in normal subjects and in uremic patients undergoing chronic hemodialysis. Mean fasting GIP was higher in the uremic patients (1006 +/- 145 (SE) pg/ml) than in the normal control subjects (132 +/- 31 pg/ml, p less than 0.001). Also, postcibal absolute and incremental serum GIP concentrations between 15 and 180 min were greater (p less than 0.05) in the uremic patients than in the control subjects; in the former they failed to return to fasting levels 180 min after the meal. In a second study, using anesthetized normal dogs, simultaneous renal arterial and venous serum GIP concentrations were measured during an intraduodenal perfusion of glucose. The renal arterial-venous (A-V) GIP gradient became greater as serum arterial GIP concentrations increased. The correlation between renal A-V GIP gradient and renal arterial GIP concentration was quite good (r = 0.85), with a 39% maximum mean A-V reduction in serum GIP concentrations observed across the kidney. This large renal A-V GIP gradient observed under nonsteady conditions suggests that the kidney may be an important site for the removal of GIP from the circulation. Thus, the higher than normal fasting and stimulated serum GIP concentrations observed in uremic patients can be attributed, at least in part, to a loss of the renal extraction mechanism for GIP.