Search

In our ongoing effort to identify genes influencing the biological pathways that underlie the metabolic disturbances associated with obesity, we performed genome-wide scanning in 2,209 individuals distributed over 507 Caucasian families to localize quantitative trait loci (QTLs), which affect variation of plasma lipids. Pedigree-based analysis using a quantitative trait variance component linkage method that localized a QTL on chromosome 7q35-q36, which linked to variation in levels of plasma triglyceride [TG, logarithm of odds (LOD) score = 3.7] and was suggestive of linkage to LDL-cholesterol (LDL-C, LOD = 2.2). Covariates of the TG linkage included waist circumference, fasting insulin, and insulin:glucose, but not body mass index or hip circumference. Plasma HDL-cholesterol (HDL-C) levels were suggestively linked to a second QTL on chromosome 12p12.3 (LOD = 2.6). Five other QTLs with lower LOD scores were identified for plasma levels of LDL-C, HDL-C, and total cholesterol. These newly identified loci likely harbor genetic elements that influence traits underlying lipid adversities associated with obesity.

Pathways leading from obesity to the manifestations of metabolic syndrome involve a number of metabolic risk factors, as well as adipokines, mediators of inflammatory response, thrombogenic and thrombolytic parameters, and vascular endothelial reactivity. Increased adipose tissue mass contributes to augmented secretion of proinflammatory adipokines, particularly tumor necrosis factor-alpha (TNF alpha), along with diminished secretion of the "protective" adiponectin. In our view, TNF alpha and adiponectin are antagonistic in stimulating nuclear transcription factor-kappa B (NF-kappa B) activation. Through this activation, TNF alpha induces oxidative stress, which exacerbates pathological processes leading to oxidized low-density lipoprotein and dyslipidemia, glucose intolerance, insulin resistance, hypertension, endothelial dysfunction, and atherogenesis. NF-kappa B activation further stimulates the formation of additional inflammatory cytokines, along with adhesion molecules which promote endothelial dysfunction. Elevated free fatty acid, glucose, and insulin levels enhance this NF-kappa B activation and further downstream modulate specific clinical manifestations of metabolic syndrome.

Despite numerous studies, the in vivo regulation of plasma leptin levels in response to nutritional factors continues to remain unclear. We investigated temporal and dose-response relationships of plasma leptin in response to physiological changes in insulin/glucose. After an overnight fast of 10 h, lean, healthy subjects were investigated for an additional 16 h of either extended fasting or one of three levels of glycemia/insulinemia induced by stepwise increasing iv glucose infusions. During extended fasting, plasma leptin values declined steadily and significantly. Plasma leptin levels remained constant at glucose concentrations between 5.8-6.5 mmol/liter, which maintained normoinsulinemia at 41.5-45.4 pmol/liter and FFA at 106-123 mg/liter, but leptin concentrations were increased at higher rates of glucose infusion (with plasma glucose rising to 8.7 mmol/liter). Concentrations of serum leptin were inversely related to FFA levels during extended fasting and at all levels of glycemia. Our data indicate that in lean healthy subjects, physiological changes in glycemia and insulinemia significantly alter plasma FFA and leptin concentrations. The increases in leptin concentrations demonstrate dose-dependent relationships that appear to relate to changes in FFA levels as well as to changes in glycemia/insulinemia.

Recent research has emphasized the importance of the metabolic cluster, which includes glucose intolerance, dyslipidemia, and high blood pressure, as a strong predictor of the obesity-related morbidities and premature mortality. Fundamental to this association, commonly referred to as the metabolic syndrome, is the close interaction between abdominal fat patterning, total body adiposity, and insulin resistance. As the initial step in identifying major genetic loci influencing these phenotypes, we performed a genomewide scan by using a 10-centiMorgan map in 2,209 individuals distributed over 507 nuclear Caucasian families. Pedigree-based analysis using a variance components linkage model demonstrated a quantitative trait locus (QTL) on chromosome 3 (3q27) strongly linked to six traits representing these fundamental phenotypes [logarithm of odds (lod) scores ranged from 2.4 to 3.5]. This QTL exhibited possible epistatic interaction with a second QTL on chromosome 17 (17p12) strongly linked to plasma leptin levels (lod = 5.0). Situated at these epistatic QTLs are candidate genes likely to influence two biologic precursor pathways of the metabolic syndrome.

Obesity is a complex disease with multiple features that has confounded efforts to unravel its pathophysiology. As a means of distinguishing primary from secondary characteristics, we compared levels of fasting plasma leptin and insulin in a cohort of weight-reduced obese women who have attained and maintained a normal BMI for more than 1 year with the levels in cohorts of never-obese and currently obese women. Weight-reduced obese women showed decreased plasma concentrations of leptin and insulin compared with obese women, but these levels remained significantly higher than those of never-obese women. Plasma leptin levels were highly correlated with plasma insulin levels (r = 0.60, P < 0.001). To further explore relationships with body composition, total body fat was determined by dual-energy X-ray absorptiometry and body fat distribution by computed tomography in subsets of these groups. Weight-reduced obese women had a significantly greater percent body fat and subcutaneous abdominal fat mass than did the never-obese women, and these were highly correlated with plasma leptin (r = 0.90, P < 0.001, and r = 0.52, P < 0.001, respectively). In these weight-reduced obese women, visceral fat mass was similar to that of the never-obese. The insulin sensitivity index and first-phase insulin response were also comparable. These results demonstrate that higher leptin levels in weight-reduced obese women are related to the higher total fat and particularly the subcutaneous fat masses. Normalization of visceral fat mass in the weight-reduced obese was accompanied by normalization of insulin sensitivity index and first-phase insulin response. This study suggests that increases in plasma leptin and insulin in obesity are secondary features of the obese state.

We have used the partially pancreatectomized infusion model in order to examine individual and combined effects of glucose and insulin on insulin resistance in rat skeletal muscles. Infusing glucose or insulin can produce animals which are hyperglycemic, hyperinsulinemic, or both. Individual and combined effects of chronic hyperglycemia and hyperinsulinemia on basal and insulin-mediated glucose utilization indices in glycolytic and oxidative muscle fibers were examined by 2-deoxyglucose uptake. Hyperglycemia reduced the basal glucose utilization index by 49% and hyperinsulinemia by 55%, while combined hyperglycemia + hyperinsulinemia diminished 2-deoxyglucose uptake by 69%. Maximally insulin-stimulated utilization was diminished only 28% under hyperglycemia but by 81% in the hyperinsulinemic state. In order to assess utilization in individual muscle fibers, uptake was examined in three tissues of differing fiber composition. The slow-twitch oxidative soleus muscle demonstrated greater basal uptake than the fast-twitch gastrocnemius (glycolytic) and quadriceps (oxidative) muscles. In addition basal (though not maximally insulin-stimulated) glucose utilization in the fast-twitch fibers was affected by chronic glucose and insulin to a greater extent than the slow-twitch soleus muscle, indicating that chronic hyperglycemia is more likely to precipitate insulin resistance in fast-twitch muscles. Significant differences in glucose metabolism among muscle fiber types suggests that results from insulin resistance studies in mixed muscles may be skewed according to their fiber composition., (Copyright 1998 Academic Press.)

Defects in glucose uptake are among the primary defects associated with peripheral insulin resistance, but fundamental mechanisms leading to this state are poorly understood. In order to elucidate mechanisms leading toward defects in glucose transport, we have used a partially pancreatectomized infusion (PxI) animal model with infusions of saline, glucose, or insulin to examine individual and combined effects of hyperglycemia and hyperinsulinemia on skeletal muscle glucose utilization. Moderate hyperglycemia induced by pancreatectomy reduced basal hindlimb muscle glucose utilization by 57% without affecting maximal insulin-stimulated glucose utilization; insulin administered in an amount sufficient to correct this hyperglycemia did not alter basal glucose utilization, but maximal insulin-stimulated glucose utilization was sharply diminished (75%); hyperglycemia with hyperinsulinemia similarly reduced basal and maximal insulin-stimulated glucose utilization. In order to establish the role of the glucose transporter protein in these insulin-resistant states, we quantified GLUT 4 content by immunoblotting and GLUT 4 mRNA by solution hybridization/RNAse protection assays. Hyperglycemia (2 weeks) reduced total muscle GLUT 4 protein content (53%) and mRNA (46%), while subsequent hyperinsulinemia (72 h) with either normo- or hyperglycemia partially restored both total GLUT 4 protein and mRNA levels. As insulin-stimulated GLUT 4 content in plasma membranes was not diminished by combined hyperglycemia/hyperinsulinemia, these results indicate functional GLUT 4 translocation in this model and suggest suppression of GLUT 4 transporter activity.

Regulation of leptin production by the hormonal and metabolic milieu is poorly understood. Because abdominal obesity is commonly associated with elevated plasma free fatty acid (FFA) flux, we examined the effects of augmenting FFA on plasma leptin levels in women who were lean and of suppressing FFA in women with abdominal obesity. In study 1, nine subjects who were lean, after a 12-hour overnight fast, received either intravenous saline or Intralipid plus heparin to increase the plasma FFA concentration to approximately 1000 mumol/ L. After 3 hours of additional fasting, subjects underwent 3-hour hyperglycemic clamps. In study 2, seven subjects with abdominal obesity were evaluated by a similar protocol, but lipolysis and plasma FFA flux were instead maximally suppressed by acipimox. In the individuals who were lean, leptin levels were unchanged during clamping. Increasing plasma FFA reduced plasma leptin from 7.66 +/- 0.66 to 7.05 +/- 0.66 (p = 0.03), but 3 hours of hyperglycemia plus hyperinsulinemia had no additional effect on leptin levels (7.15 +/- 0.71). Basal leptin levels, 4-fold higher in the subjects with obesity, were reduced from 34.6 +/- 2.4 micrograms/L to 32.3 +/- 1.1 micrograms/L (p = 0.004) during the clamp period. When plasma FFA flux was suppressed, however, plasma leptin levels after clamped hyperglycemia/hyperinsulinemia were increased to 38.9 +/- 1.2 micrograms/L (p = 0.014 vs. time 0 and 0.001 vs. saline protocol). Changes in leptin concentrations are not correlated with changes in FFA. These results suggest that plasma FFA concentration does not regulate plasma leptin levels in basal, extended fasting, or hyperglycemic/hyperinsulinemic states.

The recently cloned adipose tissue hormone leptin has been proposed to be involved in the neuroendocrine regulation of adiposity and its metabolic sequelae. Visceral fat is known to predict reduced insulin sensitivity and associated adverse metabolic profiles. In this study, we report the first evaluation of the relationships between leptin levels and total body fat, visceral fat, and insulin sensitivity in a cohort of premenopausal African-American women. Thirty-four subjects were analyzed for total fat mass and visceral fat by dual-energy X-ray absorptiometry and computerized axial tomography, respectively. Insulin sensitivity (SI) was assessed using Bergman's minimal model. Results showed that fasting leptin levels strongly correlated with total body fat mass (r = 0.797, P < 0.001). Correlations of leptin with visceral fat (r = 0.54, P < 0.001) and SI (r = -0.419, P = 0.02) were dependent on total body fat. In conclusion, leptin levels reflect total body fat mass, and although visceral fat is known to predict reduced insulin sensitivity independently, leptin did not. Our data thus suggest that diverse mechanisms are responsible for the regulation of total body versus visceral fat distribution, with its metabolic and health risks.

We examined the effects of overexposure of testosterone (T) on fat cell morphology and adipocyte precursor pools in inguinal and retroperitoneal fat depots of ovariectomized rats. In both tissues peripubertal T decreased weights without affecting adipocyte mean cell size or the size distribution profiles, but adipocyte number was decreased by 65% in the inguinal and by 38% in the retroperitoneal depots. Immunofluorescent flow cytometry utilizing a specific antibody to rat adipose tissue lipoprotein lipase was used to quantify regional precursor cell populations. T sharply reduced the percentages of differentiated and undifferentiated preadipocytes in the inguinal depot, from 43.2 +/- 5.3 to 23.5 +/- 2.1% and from 57.7 +/- 4.0 to 43.6 +/- 5.3%, respectively, with a concomitant increase in fibroblasts from 1.6 to 32.9%. On the other hand, T had no effect on retroperitoneal preadipocyte pools. Perinatal androgenization exacerbated the decline in the inguinal weight (1.4 +/- 0.1 vs. 2.2 +/- 0.1g) but otherwise did not influence the actions of peripubertal T. Androgens may thus act in a tissue-specific manner to regulate fat cell growth potential in the femoral region in the female.

The effects of female sex hormones on insulin binding and receptor-mediated insulin degradation were investigated in hepatocytes from ovariectomized rats. The influences of perinatal and peripubertal androgenization on these events were examined. Estradiol treatment increased insulin binding and receptor-mediated insulin degradation by increasing cell surface insulin receptor number. Progesterone also increased both binding and degradation, but the increase in degradation exceeded the increase in binding. Perinatal exposure to testosterone blunted the estradiol-induced increase in insulin binding and decreased degradation, whereas the progesterone-mediated increases were completely suppressed. Peripubertal testosterone decreased binding, with a much greater reduction in insulin degradation. Perinatal androgenization did not influence the peripubertal testosterone effects. Thus peripubertal female sex hormones exert regulatory influences on both hepatic cell surface insulin receptor number and postreceptor events mediating insulin degradation. These events are modulated by perinatal and peripubertal exposure to androgens. Abnormalities in sex hormone levels and/or hepatic androgenization could therefore contribute to altered insulin metabolism and hyperinsulinemia in some hyperandrogenized women with abdominal obesity and increased androgenic activity.

Insulin binding and receptor-mediated insulin processing were investigated in isolated hepatocytes from sexually maturing female rats and from age-matched animals that had undergone prepubertal ovariectomy or whose sexual dimorphism had been disrupted by neonatal androgen treatment. Equilibrium insulin binding was determined after 18 h of incubation with 125I-TyrA14-monoiodoinsulin at 4 degrees C. Receptor-mediated insulin processing was studied after overnight binding with 100 pM insulin at 4 degrees C and subsequent incubation at 37 degrees C. Insulin binding at tracer concentrations increased 50% from sexual immaturity at 3-4 wk through pubescence at 6-8 wk and was further increased into young adulthood at 10-12 wk. Scatchard analysis indicated that the altered binding was due primarily to an increase in receptor number. Increased binding with sexual maturation resulted in correspondingly higher levels of insulin in each of the four compartments of processing studied (cell surface bound, internalized, degraded, and released). A corresponding increase in receptor-mediated insulin degradation after 10 min at 37 degrees C was observed, as changes in insulin degradation were proportional to the increase in insulin receptor binding. The age-related increase in insulin binding and subsequent increase in degradation were abolished by prepubertal ovariectomy. Furthermore, disrupting sexual dimorphism by perinatal androgen treatment resulted in a reduction of the age-related increase in insulin binding, but the percentage of insulin degraded was significantly greater than that accounted for by the increase in receptor number. As a result, insulin degradation per unit of receptor-bound hormone was increased.(ABSTRACT TRUNCATED AT 250 WORDS)

Acetyl-CoA carboxylase from rat epididymal fat tissue is activated by incubation at 30 C in the absence of citrate or metal ions. This activation is accompanied by a corresponding loss of 32P from the labeled enzyme, and it is not blocked by the heat-stable phosphorylase phosphatase inhibitor proteins from rabbit muscle. We have succeeded in separating an activity which activates and dephosphorylates acetyl-CoA carboxylase from the carboxylase using polyethylene glycol-6000. These results suggest that the temperature-dependent activation of acetyl-CoA carboxylase in crude or partially purified preparations results from dephosphorylation of the carboxylase by bound phosphatase.

A flow cytometric immunofluorescence procedure utilizing a specific antibody to rat adipose tissue lipoprotein lipase (LPL) was developed to quantify differentiated and undifferentiated preadipocytes present in the adipose tissue vascular stroma. This method is highly sensitive and specific for cells capable of synthesizing LPL in significant quantities. Pubescence in female rats was associated with an increase in differentiated preadipocytes and in fat cell number with enlargement of the fat depots in the perirenal, parametrial, and the subcutaneous dorsal and femoral regions. A concomitant decline in the percentage of undifferentiated preadipocytes occurred in all but the femoral depot. Ovariectomy reduced pubertal adipose growth in the femoral and parametrial but not the dorsal or perirenal regions. Furthermore, the femoral undifferentiated preadipocyte pool was not preserved in the ovariectomized animals. Thus, ovarian factors influence the pubescence-associated regional preadipocyte differentiation and conversion to adipocytes. The femoral depot contains an ovarian-dependent infinite pool of fat cell precursors. These features could account for the association between ovarian hormones and body fat topography.

Insulin promotes an association between acetyl CoA carboxylase and acetyl CoA carboxylase phosphatase. The association between rat epididymal fat tissue carboxylase and the phosphatase occurs in both a tissue culture system and in vivo and is accompanied by an increase in acetyl CoA carboxylase activity.

Acetyl-CoA carboxylase phosphatase has been purified from the rat epididymal fat pad. The phosphatase occurs in a complex with the carboxylase. In the purification of the phosphatase, the high molecular weight complex was initially separated by sucrose gradient centrifugation, and the phosphatase was isolated from the complex by adjusting to 80% saturation with ethanol and by chromatography on Sephadex G-75. The molecular weight of the phosphatase is 71,000 as determined by sodium dodecyl sulfate gel electrophoresis and gel chromatography on Sephacryl-200 in the presence of 6 M urea. The Km for acetyl-CoA carboxylase and glycogen phosphorylase a are 1.5 microM and 37 microM, respectively. The phosphatase has a broad substrate specificity, being active toward glycogen synthase, 3-hydroxy-3-methylglutaryl-CoA reductase, phosphorylase a, phosphoprotamine, and p-nitrophenyl phosphate, in addition to acetyl-CoA carboxylase from fat tissue and liver. Acetyl-CoA carboxylase inhibits the dephosphorylation of phosphoprotamine, indicating that the same activity is responsible for dephosphorylating both substrates. The phosphatase requires no metal ion for activity and is not inhibited by the rat liver phosphorylase phosphatase inhibitor protein. The significance of these findings is discussed in relation to the regulation of acetyl-CoA carboxylase, and the phosphatase is compared to other phosphoprotein phosphatases.

Rats given a single i.p. injection of 10 mg/kg of methotrexate rapidly accumulate drug in several tissues, including liver, kidneys, testes, brain and red cells. Using a tissue extraction procedure which minimizes the breakdown of methotrexate polyglutamate derivatives and molecular sieve high-performance liquid chromatography, we examined the formation of methotrexate polyglutamates in these tissues throughout a period of 1 week after a single dose of drug. Higher polyglutamates (n greater than 3) were found in the brain, liver, kidneys and testes as early as 3 hr. The percentage of methotrexate polyglutamates increases slowly in the liver, kidney, testis and brain during the course of the week, concomitant with a decline of unconjugated tissue methotrexate and a low plasma concentration of drug. This is in contrast to the blood cells, which have little polyglutamate formation and a rapid half-life for methotrexate. The consistently high levels of methotrexate polyglutamate levels found in the liver, kidneys, testes and brain are in agreement with data showing little or no disappearance of methotrexate from these tissues for up to 26 days.

The metabolism of the antifol methotrexate in rat erythrocytes in order to determine the basis for its persistence in this tissue was investigated. By using molecular sieve high-performance liquid chromatography and a methotrexate radioassay, we have detected methotrexate polyglutamates through the heptaglutamate in red cells of rats given a single i.p. injection of 10 mg/kg of methotrexate. Higher concentrations of methotrexate and its polyglutamates were found in blood from animals made reticulocytic with phenylhydrazine than in normal erythrocytes, lending credence to the hypothesis that red cell methotrexate polyglutamates arise from developing erythrocytes in the marrow. This was borne out in discrete populations of red cells and reticulocytes separated by density centrifugation through Percoll. Reticulocytes also were shown to have no significant gamma-glutamylcarboxypeptidase (conjugase) activity. We present the reticulocytic rat as an in vivo model for studying polyglutamate metabolism.