Objective

We have reported that glucose utilization regulates leptin expression and secretion from isolated rat adipocytes. In this study, we employed two antidiabetic agents that act to increase glucose uptake by peripheral tissues, metformin and vanadium, as pharmacological tools to examine the effects of altering glucose utilization on leptin secretion in primary cultures of rat adipocytes.

Research methods and procedures

Isolated adipocytes (100 microL of packed cells per well) were anchored in a defined matrix of basement membrane components (Matrigel) with media containing 5.5 mM glucose and incubated for 96 hours with metformin or vanadium. Leptin secretion, glucose utilization, and lactate production were assessed.

Results

Metformin (0.5 and 1.0 mM) increased glucose uptake in the presence of 0.16 nM insulin by 37 +/- 10% (p < 0.005) and 62 +/- 8% (p < 0.0001) over insulin alone, respectively. Metformin from 0.5 to 5.0 mM increased lactate production by 105 +/- 43% (p < 0.025) to 202 +/- 52% (p < 0.0025) and at 1.0 and 5.0 mM increased the proportional rate of glucose conversion to lactate by 78 +/- 18% (p < 0.005) and 166 +/- 41% (p < 0.0025), respectively. At concentrations less than 0.5 mM, metformin did not affect leptin secretion, but at 0.5 mM, the only concentration that significantly increased glucose utilization without increasing glucose conversion to lactate, leptin secretion was modestly stimulated (by 20 +/- 9%; p < 0.05). Concentrations from 1.0 to 25 mM inhibited leptin secretion by 25 +/- 8% (p < 0.005) to 89 +/- 4% (p < 0.0001). Across metformin doses, leptin secretion was inversely related to the percentage of glucose taken up and released as lactate (r = -0.74; p < 0.0001). Vanadium (5 to 20 microM) increased glucose uptake from 20 +/- 7% (p < 0.01) to 34 +/- 13% (p < 0.02) and increased lactate production at 5 microM by 17 +/- 8% (p < 0.025) and 10 microM by 61 +/- 20% (p < 0.02) but did not alter the conversion of glucose to lactate. Vanadium (5 to 50 microM) inhibited leptin secretion by 33 +/- 6% (p < 0.0025) to 61 +/- 8% (p < 0.0001).

Discussion

Both metformin and vanadium increase glucose uptake and inhibit leptin secretion from cultured adipocytes. The inhibition of leptin secretion by metformin is related to an increase in the metabolism of glucose to lactate. The inhibition by vanadium most likely involves direct effects on cellular phosphatases. We hypothesize that the effect of glucose utilization to stimulate leptin production involves the metabolism of glucose to a fate other than anaerobic lactate production, possibly oxidation or lipogenesis.

Adipose tissue secretes leptin, which interacts with receptors in the hypothalamus. In rodent models of obesity, leptin increases metabolism and decreases food intake, which helps to maintain normal body composition. Accurate and precise methods to quantitate circulating leptin concentrations are needed for physiological studies. We developed an RIA to measure leptin in rat plasma, serum, or adipocyte culture fluids. The working range of the assay, defined by the detection limit and the highest calibrator, was 0.5-50 micrograms/L. Recovery of 1.6-11.6 micrograms/L leptin added to serum was 92-103%. The rat leptin RIA correlated well with a previously developed mouse RIA when rat plasma was assayed with both methods (r = 0.94), but the mouse leptin assay underestimated rat leptin in plasma. Within- and between-run CVs were 2.4% to 5.7%. Plasma leptin concentrations correlated directly with percentage of body fat, and correlation improved when the results were separated by gender (r = 0.796, P < 0.001 for males; r = 0.710, P < 0.001 for females). Leptin concentrations were generally higher in male rats than in females; plasma leptin increased 0.60 microgram/L for each percentage of increase in body fat for males but only 0.22 microgram/L for females. We conclude that rat serum/plasma leptin concentrations are accurately and precisely measured with this new RIA.

Circulating leptin secreted from adipocytes is correlated with fat mass and plasma insulin concentrations in humans and rodents. Plasma leptin, insulin, and glucose decrease during fasting and increase after refeeding; however, the underlying mechanisms regulating the changes of leptin secretion are not known. To investigate the role of insulin-stimulated glucose metabolism in the regulation of leptin secretion, we examined the effects of insulin and inhibitors of glucose transport and metabolism on leptin secretion from rat adipocytes in primary culture. Insulin (0.16-16 nM) increased leptin secretion over 96 h; however, the increase in leptin was more closely related to the amount of glucose taken up by the adipocytes (r = 0.64; P < 0.0001) than to the insulin concentration per se (r = 0.20; P < 0.28), suggesting a role for glucose transport and/or metabolism in regulating leptin secretion. 2-Deoxy-D-glucose (2-DG), a competitive inhibitor of glucose transport and phosphorylation, caused a concentration-dependent (2-50 mg/dl) inhibition of leptin release in the presence of 1.6 nM insulin. The inhibitory effect of 2-DG was reversed by high concentrations of glucose. Two other inhibitors of glucose transport, phloretin (0.05-0.25 mM) and cytochalasin-B (0.5-50 microM), also inhibited leptin secretion. Inhibition of leptin secretion by these agents was proportional to the inhibition of glucose uptake (r = 0.60 to 0.86; all P < 0.01). Two inhibitors of glycolysis, iodoacetate (0.005-1.0 mM) and sodium fluoride (0.1-5 mM), produced concentration-dependent inhibition of leptin secretion in the presence of 1.6 nM insulin. In addition, both 2-DG and sodium fluoride markedly decreased the leptin (ob) messenger RNA content of cultured adipocytes, but did not affect 18S ribosomal RNA content. We conclude that glucose transport and metabolism are important factors in the regulation of leptin expression and secretion and that the effect of insulin to increase adipocyte glucose utilization is likely to contribute to insulin-stimulated leptin secretion. Thus, in vivo, decreased adipose glucose metabolism may be one mechanism by which fasting decreases circulating leptin, whereas increased adipose glucose metabolism would increase leptin after refeeding.