Liporegulation in diet-induced obesity. The antisteatotic role of hyperleptinemia

To test the hypothesis that the physiologic liporegulatory role of hyperleptinemia is to prevent steatosis during caloric excess, we induced obesity by feeding normal Harlan Sprague-Dawley rats a 60% fat diet. Hyperleptinemia began within 24 h and increased progressively to 26 ng/ml after 10 weeks, correlating with an approximately 150-fold increase in body fat (r = 0.91, p < 0.0001). During this time, the triacylglycerol (TG) content of nonadipose tissues rose only 1-2.7-fold implying antisteatotic activity. In rodents without leptin action (fa/fa rats and ob/ob and db/db mice) receiving a 6% fat diet, nonadipose tissue TG was 4-100 times normal. In normal rats on a 60% fat diet, peroxisome proliferator-activated receptor alpha protein and liver-carnitine palmitoyltransferase-1 (l-CPT-1) mRNA increased in liver. In their pancreatic islets, fatty-acid oxidation increased 30% without detectable increase in the expression of peroxisome proliferator-activated receptor-alpha or oxidative enzymes, whereas lipogenesis from [14C]glucose was slightly below that of the 4% fat-fed rats (p < 0.05). Tissue-specific overexpression of wild-type leptin receptors in the livers of fa/fa rats, in which marked steatosis is uniformly present, reduced TG accumulation in liver but nowhere else. We conclude that a physiologic role of the hyperleptinemia of caloric excess is to protect nonadipocytes from steatosis and lipotoxicity by preventing the up-regulation of lipogenesis and increasing fatty-acid oxidation.

Slow recovery of body fat lost during adenovirus-induced hyperleptinemia

In normal rats, adenovirus-induced hyperleptinemia causes disappearance of visible body fat, downregulation of lipogenic enzymes, and upregulation of oxidative enzymes and thermogenic proteins. In addition, preadipocyte markers replace mature adipocyte markers, suggesting dedifferentiation. In weight loss induced by caloric restriction, by contrast, the lipogenic machinery is essentially intact. To determine if the radical changes induced by leptin would slow the reappearance of body fat, we compared normal lean rats made hyperleptinemic by infusing an adenovirus-leptin construct with diet-matched littermates. Initially, in plasma leptin the hyperleptinemic rats averaged approximately 50x the controls and, although it declined progressively, it was still slightly elevated at 150 days (P < 0.05). In the hyperleptinemics, body fat mass, quantified by magnetic resonance spectroscopy, remained below the pretreatment value for 60 days, while in diet-matched controls it exceeded the pretreatment value. Epididymal fat pad weight in hyperleptinemics was still 28% below paired controls at 150 days posttreatment. Histologic examination revealed adipocytes of hyperleptinemic animals to be smaller 60 days after treatment. At 60 days, adipose tissue UCP-2 gene expression in hyperleptinemics was still above controls, but expression of other lipogenic and oxidative enzymes had returned to baseline expression levels. We conclude that in normal rats recovery of body fat following adenovirus-induced hyperleptinemia is much slower than after caloric restriction, possibly because of persistent upregulation of adipocyte UCP-2.

Amylin-insulin relationships in insulin resistance with and without diabetic hyperglycemia

To determine if increased secretion of amylin can be implicated in the pathogenesis of non-insulin-dependent diabetes mellitus (NIDDM) in vitro and in vivo, we studied its relationships to insulin in insulin-resistant rats with and without NIDDM. In obesity-associated and dexamethasone-induced insulin resistance without diabetes, basal and stimulated secretion of amylin and insulin by isolated pancreata were proportionately elevated, leaving the amylin-to-insulin ratio (A/I) unchanged. By contrast, whenever diabetes occurred in dexamethasone-treated rats or in spontaneously diabetic obese insulin-resistant ZDF-drt male rats, a doubling of A/I was invariably observed due to an increase in amylin without a proportional increase in insulin secretion. Correction of dexamethasone-induced hyperglycemia with the glucocorticord receptor antagonist RU-486 was accompanied by a decline in A/I. Longitudinal in vivo studies demonstrated in both spontaneous and dexamethasone-induced models of NIDDM an increase in plasma A/I at the onset of hyperglycemia. In dexamethasone-induced diabetes, the increased A/I was associated with a high proamylin mRNA relative to proinsulin mRNA. We conclude that amylin and insulin expression and secretion rise in concert in compensated insulin-resistant states, but when hyperglycemia is present the increase in amylin exceeds that of insulin. Although a role of an increased A/I in the pathogenesis of NIDDM has not been established directly, these studies indicate that such a role could be possible.

Inhibition of glucose transport into rat islet cells by immunoglobulins from patients with new-onset insulin-dependent diabetes mellitus

Because glucose-stimulated insulin secretion is selectively impaired during the development of insulin-dependent diabetes mellitus (IDDM), we tested the possibility that the glucose transporter of pancreatic islet beta cells is a target of the autoimmune process in patients with IDDM. We measured the uptake of 3-O-methyl-beta-D-glucose by dispersed islet cells from rats after a 15-minute incubation with purified IgG from 27 patients with newly diagnosed IDDM, 28 normal subjects, and 5 patients with non-insulin-dependent diabetes mellitus (NIDDM). The IgG fractions from 26 of the 27 patients with IDDM (96 percent), but from none of the 5 patients with NIDDM, reduced the initial rates of 3-O-methyl-beta-D-glucose uptake to at least 1 SD below the mean of the rates observed in the presence of IgG fractions from normal subjects (P less than 0.001). In contrast, the uptake of L-leucine by islet cells was not affected by any of the IgG fractions. The inhibitory activity of IgG from the patients with IDDM was abolished by preincubation with islet cells and membranes from hepatocytes, which contain the same glucose transporter as beta cells, but not with erythrocytes, which do not contain this transporter. We conclude that IgG from patients with IDDM of recent onset, but not from those with NIDDM, inhibits glucose uptake by rat islet cells. The results are consistent with the presence of an antibody against a protein involved in glucose transport by beta cells that would thereby impair glucose-stimulated insulin secretion.

Molecular and cellular responses of islets during perturbations of glucose homeostasis determined by in situ hybridization histochemistry

We have evaluated in situ hybridization histochemistry as a means of estimating simultaneously the level of prohormone mRNA and the dimensions of rat pancreatic islets. Localization of the 27-mer 32P-labeled oligonucleotide probes for rat proinsulin I, glucagon, and prosomatostatin I corresponded with localization of antibodies to the three hormones. In normal rats subjected to chronic hyperglycemic clamping, the density of the proinsulin mRNA signal increased 54%, islet size and number increased approximately 100%, while proglucagon mRNA signal was reduced 81%. Resection of 50% of the pancreas increased proinsulin mRNA 36% and proglucagon mRNA 500%; islet area doubled and islet number increased 50%. In 150-day-old diabetic ob/ob mice, there was an 18-fold expansion in islet area, a 4-fold increase in islet number, but no increase in insulin gene expression. In insulin-dependent streptozotocin-treated diabetic rats, islet area and number were profoundly reduced; insulin deprivation failed to raise proinsulin mRNA in surviving beta cells above control levels. Proglucagon mRNA was high despite the hyperglycemia but was reduced by insulin within 1 hr, suggesting that insulin regulates glucagon gene expression or is required for its regulation by glucose. In situ hybridization of rat islets provides a valid semiquantitative index of insulin and glucagon biosynthesis and of islet dimensions and reveals that normal but not diabetic islets meet increased insulin demand by increasing both number and biosynthetic activity of beta cells.

Loss of insulin response to glucose but not arginine during the development of autoimmune diabetes in BB/W rats: relationships to islet volume and glucose transport rate

The insulin and glucagon responses to 10 mM glucose and 10 mM arginine were studied in pancreata isolated from nondiabetic diabetes-prone and diabetes-resistant BB/W rats at 60, 80, and 140 days of age and in diabetic BB/W rats on the 1st and 14th days of their diabetes. In the former group the insulin response to glucose declined progressively with age (r = -0.575; P less than 0.01) and at 140 days was significantly below age-matched diabetes-resistant controls (P less than 0.05). The insulin response to arginine did not decline with age in either group. For diabetic rats, on the first day of the diabetes, the insulin response to glucose was absent but the response to arginine did not differ from nondiabetic controls. On day 14 responses to glucose and arginine were both absent. The glucagon response to arginine showed no trend despite a decline in baseline glucagon secretion. Endocrine tissue in nondiabetic diabetes-prone rats made up 0.8 +/- 0.2% of the pancreas at 60 days of age and 0.52 +/- 0.22% at 140 days of age; the latter was significantly less than in 140-day-old diabetes-resistant controls (P less than 0.05). In diabetic rats on the 1st and 14th days of diabetes endocrine tissue was 0.2 +/- 0.1% and 0.07 +/- 0.02%, respectively. The glucose transport rate in islets isolated on the first day of diabetes was profoundly reduced compared to age-matched nondiabetic diabetes-prone controls. Thus, a population of arginine-responsive, glucose-unresponsive islets with low glucose transport rates is present at the onset of overt diabetes in BB/W rats.

Effect of gastric inhibitory polypeptide on plasma levels of chylomicron triglycerides in dogs

To determine whether gastric inhibitory polypeptide (GIP) promotes the clearance of chylomicron triglycerides (TG) from the circulation in dogs, chyle collected from donor dogs via a thoracic duct fistula was infused at a rate of 2 ml/min i.v. into normal recipient dogs during an infusion of either porcine GIP (1 microgram/kg per h) or saline as a control. In the GIP-infused dogs the rise in plasma TG was significantly below that of the control animals [mean peak of 36 +/- 4 mg/dl vs. 82 +/- 18 mg/dl (P less than 0.05)]. It is concluded that GIP exerts an effect upon the removal of chylomicron TG from the blood. The results suggest that GIP may play a physiologic role in the disposition of ingested fat.

High plasma free fatty acid levels contribute to the hypersomatostatinemia of insulin deficiency

High plasma levels of free fatty acids (FFA) stimulate the secretion of splanchnic somatostatin, and both are elevated in insulin deficiency. To determine if the hypersomatostatinemia of insulin deficiency is secondary to high FFA levels, plasma somatostatin-like immunoreactivity (SLI) was measured in a group of insulin-deprived alloxan-diabetic dogs during nicotinic acid-induced lowering of their elevated plasma FFA to normal, and in a group of nondiabetic dogs during nicotinic acid-induced lowering of their FFA to subnormal values. In insulin-deprived diabetic dogs, nicotinic acid reduced plasma FFA from 1.07 +/- 0.2 (M +/- SE) mmol/L to 0.6 +/- 0.1 mmol/L (P less than 0.02), approximately the basal FFA level in normal dogs. This was accompanied by a significant decline in plasma SLI levels from a mean baseline of 247 +/- 15 pg/ml to a mean nadir of 199 +/- 10 pg/ml (P less than 0.005). The latter was, nevertheless, significantly above the basal SLI level of the nondiabetic dogs. In contrast, in normal dogs, nicotinic acid-induced reduction in FFA from 0.54 +/- 0.02 mmol/L to 0.24 +/- 0.03 mmol/L (P less than 0.001) was associated with only a small and inconsistent decrease in SLI. These findings suggest that the hypersomatostatinemia of insulin-deficient alloxan-diabetic dogs is, in part, secondary to high plasma FFA levels.

Binding and biologic activity of glucagon in liver cell membranes of chronically hyperglucagonemic rats

Glucagon binding by liver cell membranes was examined in rats with chronically elevated plasma levels of immunoreactive glucagon (IRG) resulting from insulin deficiency, starvation, or twice daily glucagon injections. The concentration of specific glucagon binding sites was significantly reduced in the three chronically hyperglucagonemic (IRG greater than 125 pg/ml) groups as compared with nondiabetic controls and insulin-treated diabetic control rats with only mild hyperglucagonemia. A reduction in glucagon binding sites did not occur with hyperglucagonemia of 12 h or less. Despite the reduced binding of glucagon in the three chronically hyperglucagonemic groups, the ability of glucagon to stimulate cAMP production was not reduced. It is concluded that while decreased glucagon binding occures in the forms of chronic hyperglucagonemia studied, it is not associated with a reduction in the ability of glucagon to stimulate cAMP production.

Properties of immunoreactive glucagon fractions of canine stomach and pancreas

The present study was designed to identify the physicochemical, immunologic, and biologic properties of the immunoreactive glucagon (IRG) moieties of canine gastric fundus and to compare them with those of the canine pancreas. Acid-alcohol extracts of the gastric fundus and pancreas of dogs were subjected to Bio-Gel P-10 chromatography, The elution profiles of extracts of both organs revealed IRG peaks in the Mr = 2,000 3,500, and 9,000 zones; in the gastric extracts, a void volume peak was also present. On the basis of Sephadex G-150 rechromatography and sucrose density gradient ultracentrifugation the latter IRG was estimated to have a Mr = 65,000. Incubation of fundic IRG65,000 in 8 M urea failed to alter its elution position. Its pI was 6.4, while fundic IRG3,500 had a pI of 6.15 and pancreatic glucagon 6.25. Fundic IRG9,000 had a pI of 4.5 and pancreatic IRG9,000 4.65. Dilution curves of these three fundic and two pancreatic IRGs were parallel to crystalline beef-pork glucagon. The glycogenolytic activity of fundic IRG3,500 and IRG65,000, measured in the isolated rat liver system, was not different from that of immunoequivalent amounts of dog pancreatic glucagon or crystalline beef-pork glucagon. Both fundic and pancreatic IRG9,000 were devoid of glycogenolytic activity and lacker adenylate cyclase stimulating activity and 125I-glucagon displacing activity when tested on partially purified rat liver membranes. Fundic IRG65,000, however, stimulated adenylate cyclase and displaced 125I-glucagon to the same degree as immunoequivalent amounts of pancreatic glucagon. Fundic IRG3,500 was more active than pancreatic glucagon in stimulating adenylate cyclase activity. This was not clearly attributable to differences in binding to liver cell membranes.