Madhur K. Sinha

Serum immunoreactive-leptin concentrations in normal-weight and obese humans.

BACKGROUND Leptin, the product of the ob gene, is a hormone secreted by adipocytes. Animals with mutations in the ob gene are obese and lose weight when given leptin, but little is known about the physiologic actions of leptin in humans. METHODS Using a newly developed radioimmunoassay, wer measured serum concentrations of leptin in 136 normal-weight subjects and 139 obese subjects (body-mass index, > or = 27.3 for men and > or = 27.8 for women; the body-mass index was defined as the weight in kilograms divided by the square of the height in meters). The measurements were repeated in seven obese subjects after weight loss and during maintenance of the lower weight. The ob messenger RNA (mRNA) content of adipocytes was determined in 27 normal-weight and 27 obese subjects. RESULTS The mean (+/- SD) serum leptin concentrations were 31.3 +/- 24.1 ng per milliliter in the obese subjects and 7.5 +/- 9.3 ng per milliliter in the normal-weight subjects (P < 0.001). There was a strong positive correlation between serum leptin concentrations and the percentage of body fat (r = 0.85, P < 0.001). The ob mRNA content of adipocytes was about twice as high in the obese subjects as in the normal-weight subjects (P < 0.001) and was correlated with the percentage of body fat (r = 0.68, P < 0.001) in the 54 subjects in whom it was measured. In the seven obese subjects studied after weight loss, both serum leptin concentrations and ob mRNA content of adipocytes declined, but these measures increased again during the maintenance of the lower weight. CONCLUSIONS Serum leptin concentrations are correlated with the percentage of body fat, suggesting that most obese persons are insensitive to endogenous leptin production.

Nocturnal rise of leptin in lean, obese, and non-insulin-dependent diabetes mellitus subjects.

We studied 24-h profiles of circulating leptin levels using a sensitive and specific RIA in lean controls and obese subjects with or without non-insulin-dependent diabetes mellitus (NIDDM) during normal routine activity. Serum leptin levels were significantly higher in obese (41.7 +/- 9.0 ng/ml; n = 11) and obese NIDDM (30.8 +/- 6.7; n = 9) subjects compared with those in lean controls (12.0 +/- 4.4, n = 6). In all the three groups, serum leptin levels were highest between midnight and early morning hours and lowest around noon to midafternoon. The nocturnal rise in leptin levels was significant when data were analyzed by ANOVA (lean: F = 3.17, P < 0.0001, n = 4; obese: F = 2.02, P < 0.005, n = 11; and obese NIDDM: F = 4.9, P < 0.0001, n = 5). The average circadian amplitude between acrophase and nadir was 75.6% in lean, 51.7%, in obese and 60.7% in obese NIDDM groups, respectively. No significant correlations (P > 0.05) were observed between circulating levels of leptin and either insulin or glucose levels in any of the 20 subjects studied for 24-h profiles. The nocturnal rise in leptin observed in the present study resembles those reported for prolactin, thyroid-stimulating hormone, and free fatty acids. We speculate that the nocturnal rise in leptin could have an effect in suppressing appetite during the night while sleeping.

Evidence of free and bound leptin in human circulation. Studies in lean and obese subjects and during short-term fasting.

Little is known about leptins interaction with other circulating proteins which could be important for its biological effects. Sephadex G-100 gel filtration elution profiles of 125I-leptin-serum complex demonstrated 125I-leptin eluting in significant proportion associated with macromolecules. The 125I-leptin binding to circulating macromolecules was specific, reversible, and displaceable with unlabeled leptin (ED50: 0.73 +/- 0.09 nM, mean +/- SEM, n = 3). Several putative leptin binding proteins were detected by leptin-affinity chromatography of which either 80- or 100-kD proteins could be the soluble leptin receptor as approximately 10% of the bound 125I-leptin was immunoprecipitable with leptin receptor antibodies. Significantly higher (P < 0.001) proportions of total leptin circulate in the bound form in lean (46.5 +/- 6.6%) compared with obese (21.4 +/- 3.4%) subjects. In lean subjects with 21% or less body fat, 60-98% of the total leptin was in the bound form. Short-term fasting significantly decreased basal leptin levels in three lean (P < 0.0005) and three obese (P < 0.005) subjects while refeeding restored it to basal levels. The effects of fasting on free leptin levels were more pronounced in lean subjects (basal vs. 24-h fasting: 19.6 +/- 1.9 vs. 1.3 +/- 0.4 ng/ml) compared with those in obese subjects (28.3 +/- 9.8 vs. 14.7 +/- 5.3). No significant (P > 0.05) decrease was observed in bound leptin in either group. These studies suggest that in obese individuals the majority of leptin circulates in free form, presumably bioactive protein, and thus obese subjects are resistant to free leptin. In lean subjects with relatively low adipose tissue, the majority of circulating leptin is in the bound form and thus may not be available to brain receptors for its inhibitory effects on food intake both under normal and food deprivation states.

Insulin receptor kinase in human skeletal muscle from obese subjects with and without noninsulin dependent diabetes.

We have studied the structure and function of the insulin receptors in obese patients with and without noninsulin dependent diabetes mellitus (NIDDM) and in nonobese controls using partially purified receptors from muscle biopsies. Insulin binding was decreased in obesity due to reduced number of binding sites but no differences were observed in insulin binding between obese subjects with or without NIDDM. The structural characteristics of the receptors, as determined by affinity labeling methods and electrophoretic mobility of the beta-subunit, were not altered in obese or NIDDM compared to normal weight subjects. Furthermore, the ability of insulin to stimulate the autophosphorylation of the beta-subunit and the phosphoamino acid composition of the phosphorylated receptor were the same in all groups. However, insulin receptor kinase activity was decreased in obesity using Glu4:Tyr1 as exogenous phosphoacceptor without any appreciable additional defect when obesity was associated with NIDDM. Thus, our data are supportive of the hypothesis that in muscle of obese humans, insulin resistance is partially due to decreased insulin receptors and insulin receptor kinase activity. In NIDDM the defect(s) in muscle is probably distal to the insulin receptor kinase.

Insulin-like growth factor I binding in hepatocytes from human liver, human hepatoma, and normal, regenerating, and fetal rat liver.

Insulin-like growth factor-I (IGF-I) in human hepatoma cells (HEP-G2) has, in addition to its effect on cell growth, short-term metabolic effects acting through its own receptor. We have demonstrated that normal human hepatocytes, compared with HEP-G2 cells, have virtually no IGF-I binding sites. Because the rate of growth is the major difference between the hepatoma and the normal liver, we asked if normal liver might express IGF-I binding sites under physiologic growth conditions. Indeed, whereas adult rat hepatocytes have low IGF-I binding sites similar to those in human liver, hepatocytes from regenerating liver after 3 d subtotal hepatectomy have an approximately sixfold increase (P less than 0.005) and those from fetal rat liver a approximately 12-fold increase (P less than 0.005), to levels comparable to those in the HEP-G2 cells. The specificity of 125I IGF-I binding to its receptor was demonstrated by competition studies with monoclonal antibodies directed toward the IGF-I and the insulin receptors, with unlabeled IGF-I and insulin and by affinity labeling experiments. Thus, if IGF-I has any short-term metabolic functions in the adult human liver, it is not through interaction with its own receptor. Autocrine regulation by IGF-I of liver growth appears possible since IGF-I binding sites are expressed under pathological and physiological conditions of growth. The mechanism that couples these two phenomena remains to be elucidated.

Studies on the mechanism of insulin resistance in the liver from humans with noninsulin-dependent diabetes. Insulin action and binding in isolated hepatocytes, insulin receptor structure, and kinase activity.

We have developed a method to isolate insulin-responsive human hepatocytes from an intraoperative liver biopsy to study insulin action and resistance in man. Hepatocytes from obese patients with noninsulin-dependent diabetes were resistant to maximal insulin concentration, and those from obese controls to submaximal insulin concentration in comparison to nonobese controls. Insulin binding per cell number was similar in all groups. However, insulin binding per surface area was decreased in the two obese groups because their hepatocytes were larger. In addition, the pool of detergent-extractable receptor was further decreased in diabetics. Insulin receptors in all groups were unaltered as determined by affinity-labeling methods. However, insulin-stimulated insulin receptor kinase activity was decreased in diabetics. Thus, in obesity, decreased surface binding could explain resistance to submaximal insulin concentrations. In diabetes, diminished insulin-stimulated protein kinase activity and decreased intracellular pool of receptors could provide an explanation for postinsulin-binding defect(s) of insulin action in human liver.

Insulin-Receptor Kinase Activity of Adipose Tissue From Morbidly Obese Humans With and Without NIDDM

We have determined glucose transport, insulin binding, and insulin-receptor kinase activity in adipose tissue from morbidly obese patients with and without non-insulin-dependent diabetes mellitus (NIDDM). The insulin sensitivity and responsiveness of glucose transport in freshly isolated adipocytes were significantly reduced in NIDDM subjects compared with nondiabetics. This was due in part to decreased insulin binding in adipocytes. Reduced specific 125I-labeled insulin binding was also observed in crude detergent extracts and partially purified insulin receptors from adipose tissue. In addition, the basal and insulin-stimulated tyrosine-specific protein kinase activity per milligram of protein was significantly decreased in NIDDM patients compared with nondiabetics. The differences between maximally insulin-stimulated and basal kinase activities expressed by insulin-binding activity were also significantly reduced in NIDDM subjects. We conclude that insulin resistance in morbidly obese patients with NIDDM is due to both insulin-binding and postbinding defects. One of the postbinding defects in NIDDM appears to be impaired insulin-receptor kinase activity of fat tissue.

Adipose Tissue Glucose Transporters in NIDDM: Decreased Levels of Muscle/Fat Isoform

We investigated the mechanism of peripheral insulin resistance in the adipose tissue of obese and non-insulin-dependent diabetes mellitus (NIDDM) patients at the level of the glucose-transport effector system. Freshly isolated adipocytes from obese nondiabetic and obese NIDDM subjects had decreased insulin sensitivity and responsiveness for glucose-transport stimulation compared with control subjects, with more pronounced changes associated with obese NIDDM patients. The relative abundance of muscle/fat glucose-transporter isoform in the three groups of subjects was determined by Western-blot analysis of detergent-soluble adipose tissue extracts with monoclonal antibody 1F8. Obesity per se had no effect on adipose tissue muscle/fat glucose-transporter isoform (3150 ± 660 vs. 4495 ± 410 counts/min [cpm]/mg protein). Furthermore, decreased levels of muscle/fat isoform was observed in adipose tissue of NIDDM patients (1560 ± 600 cmp/mg protein.) Furthermore, decreased levels of muscle/fat isoform in adipose tissue of NIDDM patients were also reflected in isolated adipocytes. Our results demonstrate that insulin resistance in isolated adipocytes of NIDDM patients could at least partly be due to a significant depletion of adipose tissue muscle/fat glucose-transporter isoform.

Orally Administered Liposome-Entrapped Insulin in Diabetic Animals

We investigated the effects of liposome-entrapped insulin (LEI) administered orally one-half to one-tenth of previously reported doses, on plasma glucose and insulin in the spontaneously diabetic BB Wistar rat and in alloxan-induced diabetic rabbits. Incorporation of insulin within the liposome fraction ranged from 15 to 23%. Radioimmunoassay of Triton X-100 treated LEI yielded insulin values in high agreement (82 +/- 10%) with those predicted based on estimated incorporation. Whereas insulin alone, or liposomes devoid of insulin had no effect, LEI 5 U/kg significantly reduced glucose and raised insulin in 54% of rats (13 of 24) and 67% of the rabbits (6 of 9). Among the rats that responded, blood glucose fell from a basal of 318 +/- 21 mg/dl to a nadir of 186 +/- 22 mg/dl at 2 h (p less than 0.001); values at 1, 2 and 4 h were all significantly less (58-69%) than basal. Similarly, glucose declined significantly for 3 h post LEI in the rabbits while IRI rose from 30 +/- 7 micro U/ml to a peak of 399 +/- 75 micro U/ml at 1 h (p less than 0.001); values at 2 and 3 h remained significantly elevated. Some batches of LEI failed to reduce glucose despite apparently adequate incorporation, while even with effective batches some animals failed to respond. Thus, although orally administered LEI can be effective, their stability and effectiveness are not completely predictable.

Mechanism of IGF-I-Stimulated Glucose Transport in Human Adipocytes: Demonstration of Specific IGF-I Receptors Not Involved in Stimulation of Glucose Transport

We demonstrate the presence of specific insulinlike growth factor I (IGF-I) receptors in human adipocytes. Competition studies with 125I-labeled IGF-I and unlabeled IGF-I, IGF-II, and insulin showed the specificity of 125I-IGF-I binding to the IGF-I receptors in adipocytes, membranes, and partially purified detergent-solubilized extracts. The monoclonal antibody to the IGF-I receptor (α-IR3) inhibits 125I-IGF-I binding and immunoprecipitates the IGF-I receptor. In addition, the α-subunit of IGF-I receptor is ∼10,000 Mr larger than the α-subunit of insulin receptor, and IGF-I stimulates phosphorylation of the β-subunit of the IGF-I receptor. IGF-I stimulates basal glucose transport in human adipocytes, but the concentrations of IGF-I required for half-maximal and maximal stimulation of glucose transport are 800- and 1000-fold greater than that of insulin. The possibility of IGF-I stimulating glucose transport by interacting predominantly with insulin receptors is suggested by data showing that 1) IGF-I competes with insulin-binding sites, 2) there is a lack of an additive effect with IGF-I and insulin in stimulating glucose transport, 3) α-IR3, which specifically inhibits IGF-I binding, does not inhibit IGF-I or insulin-stimulated glucose transport, 4) insulinreceptor antibody MA-10 inhibits IGF-I and insulinstimulated glucose transport, and 5) IGF-I stimulates insulin-receptor autophosphorylation, although its effect is markedly decreased compared with insulin. In summary, human adipocytes possess specific IGF-I receptors. However, IGF-I stimulates glucose transport predominantly by interacting with the insulin receptor.