Exercise Improves Insulin and Leptin Sensitivity in Hypothalamus of Wistar Rats

The reagents for SDS-PAGE and immunoblotting were from Bio-Rad. Tris, aprotinin, ATP, dithiothreitol, phenylmethylsulfonyl fluoride, Triton X-100, Tween 20, glycerol, and BSA (fraction V) were from Sigma (St. Louis, MO). Protein A-Sepharose 6MB, ¹²⁵I-protein A, and nitrocellulose paper (Hybond ECL, 0.45 μm) were from Amersham Pharmacia Biotech (Buckinghamshire, U.K.). Sodium amobarbital (Amytal) and human recombinant insulin (Humulin R) were from Eli Lilly (Indianapolis, IN). Leptin was from Calbiochem (San Diego, CA). Ketamine hydrochloride was from Cristália (Itapira, Brazil). Antibodies to insulin receptor, IRS-1, IRS-2, Akt, JAK2, ObR, SOCS3 (suppressor of cytokine signaling 3), PTP1b (protein-tyrosine phosphatase 1b), STAT3, and IL-6 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The Akt phosphoserine 473–specific and phosphothreonine 308–specific antibodies and the STAT3 phosphotyrosine 705–specific antibody were from New England Biolabs (Beverly, MA), and the antibody to the p85 subunit of PI 3-kinase was from Upstate Biotechnology (Lake Placid, NY). Routine reagents were purchased from Sigma, unless otherwise specified.

Male Wistar rats (200–250 g) from the University of Campinas Central Animal Breeding Center were used in the experiments. All experiments involving animals were in accordance with the guidelines of the Brazilian College for Animal Experimentation (COBEA) and were approved by the ethics committee at the University of Campinas. Rats were maintained on a 12-h light/dark cycle and were provided free access to water and standard rodent chow before the exercise; they were randomly assigned to one of two groups: those exercised for 6 h or control rats. After an overnight fast, the rats were anesthetized with ketamine hydrochloride plus diazepan and positioned on a Stoelting stereotaxic apparatus. At 10 days before the exercise protocol, the catheter was implanted into the third ventricle as previously described (10). After a 1-week recovery period, catheter placement was confirmed by a positive drinking response after administration of angiotensin II (40 ng per 2 μl), and animals that did not drink 5 ml of water within 15 min after treatment were not included in the experiment.

Rats were acclimated to swimming for 10 min per day for 2 days. The swimming protocol was performed as previously described (31). The rats swam in groups of three in plastic barrels 45 cm in diameter that were filled to a depth of 50 cm, and the water temperature was maintained at 34–35°C. Animals performed two 3-h exercise bouts separated by one 45-min rest period.

After the last bout of exercise, animals were injected (2-μl bolus injection i.c.v.) with either vehicle, insulin (Eli Lilly), or leptin (rat leptin from National Institutes of Health) at the doses indicated. Thereafter, standard chow was given, and food intake was determined by measuring the difference between the weight of chow given and the weight of chow at the end of a 12-h period. Similar studies were carried out after the last bout of exercise in rats that were initially intracerebroventricularly injected with anti–IL-6 (rabbit anti–IL-6 at the doses indicated; Santa Cruz Biotechnology) or vehicle and then, after 30 min, with insulin or leptin. In preliminary experiments, we determined plasma glucose levels in animals that received intracerebroventricular insulin infusion. Plasma glucose was not altered by third ventricle insulin or saline microinjection.

Immediately after the last exercise bout, animals were treated with vehicle, insulin, or leptin, according to the protocols described in the preceding section, and then they were decapitated, and the hypothalami were removed at the time points indicated. The hypothalami were minced coarsely and homogenized immediately in solubilization buffer containing 100 mmol/l Tris (pH 7.6), 1% Triton X-100, 150 mmol/l NaCl, 0.1 mg aprotinin, 35 mg/ml phenylmethylsulfonyl fluoride, 10 mmol/l Na₃VO₄, 100 mmol/l NaF, 10 mmol/l Na₄P₂O₇, and 4 mmol/l EDTA, using a polytron PTA 20S generator (Model PT 10/35; Brinkmann Instruments, Westbury, NY) operated at maximum speed for 30 s and clarified by centrifugation. Equal amounts of protein were used for immunoprecipitation followed by Western blot analysis with the indicated antibodies and ¹²⁵I-protein A. ¹²⁵I-protein A bound to anti-peptide antibodies was detected by autoradiography, using preflashed Kodak XAR film (Eastman Kodak, Rochester, NY) with Cronex Lightning Plus intensifying screens (DuPont, Wilmington, DE) at −80°C for 12–48 h. Band intensities were quantitated by optical densitometry (Scion Image software; ScionCorp, Frederick, MD) of the developed autoradiographs.

Aliquots of supernatants containing equal amounts of protein were incubated overnight at 4°C, using antibodies against IRS-1 or -2, and the immunocomplexes were precipitated with a 50% solution of protein A-Sepharose 6MB. In vitro PI 3-kinase assays were performed as previously described (17). The ³²P-labeled 3-P-PI was quantitated using Scion Image software.

Where appropriate, the results are expressed as the means ± SE accompanied by the indicated number of rats used in experiments. Comparisons among groups were performed using parametric two-way ANOVA, where F ratios were significant; further comparisons were performed using the Bonferroni test.

The plasma glucose level was lower in the exercised group compared with the control group (3.39 vs. 4.41 mmol/l, respectively) (Fig. 1A), and the insulin levels were also lower (93.6 vs. 187.2 pmol/l) (Fig. 1B). Exercise did not, however, reduce plasma leptin (2.1 vs. 2.4 ng/ml) (Fig. 1C). As shown in Figs. 1B and C, insulinemia and leptinemia were not altered by third ventricle microinjection of insulin or leptin.

The effect of leptin (5, 10, and 20 μg), or its vehicle, on the control of food intake was studied by measuring the total food intake for 12 h after a session of exercise and a single intracerebroventricular injection of leptin or its vehicle. Leptin induced reductions in the 12-h food intake in both exercised and control rats in a concentration-dependent manner. In the exercised rats, leptin (5, 10, and 20 μg) reduced food intake by 38.5, 44.6, and 46.4%, respectively, whereas in the control group, these doses induced reductions of 15.7, 23.2, and 31.2%, indicating that leptin was much more effective in exercised rats (Fig. 2A).

To determine the effects of exercise on the early steps of the leptin signaling pathway, a dose of 10 μg was administered and the ObR and JAK2 tyrosine phosphorylation was assessed in the hypothalamus of exercised and control rats. Immunoprecipitation and Western blotting of hypothalamic extracts were performed using anti-ObR, anti-JAK2, and anti-phosphotyrosine antibodies. Leptin induced increases in ObR and JAK2 tyrosine phosphorylation levels in hypothalami from both control and exercised rats. In the exercised animals, leptin increased ObR and JAK2 tyrosine phosphorylation by 6.0- and 4.7-fold, respectively, compared with 2.1- and 2.7-fold increases in the hypothalami from control rats, representing increases in ObR tyrosine phosphorylation of 4.7- and 2.2-fold, respectively (Fig. 2B and C, upper panels). The same membranes used to detect tyrosine phosphorylation of ObR and JAK2 were reblotted with ObR and JAK2 antibodies, and, as expected, there were no changes in ObR and JAK2 protein expression (Fig. 2B and C, lower panels). Hypothalamic extracts from exercised and control rats that were stimulated with leptin (10 μg) were lysed and the proteins separated by SDS-PAGE gel and blotted with pSTAT3 antibodies. In the hypothalami from exercised animals, leptin increased STAT3 tyrosine phosphorylation by 5.3-fold compared with 2.4-fold increases in the hypothalami from control rats, representing 3.2-fold increases in STAT3 tyrosine phosphorylation (Fig. 2D, upper panel). No changes were observed in STAT3 protein expression (Fig. 2D, lower panel).

Immunoprecipitation and Western blotting of hypothalamic extracts were performed using anti–IRS-1, anti–IRS-2, and anti-phosphotyrosine antibodies. Leptin (10 μg) induced increases in IRS-1/2 tyrosine phosphorylation levels in hypothalami from both control and exercised rats. In the exercised animals, leptin increased IRS-1 and -2 tyrosine phosphorylation by 5.8- and 5.6-fold, respectively, compared with 3.4- and 3.8-fold increases in the hypothalami from control rats, representing increases in IRS-1 and -2 tyrosine phosphorylation of 2.0- and 1.6-fold, respectively (Fig. 3A and C, upper panels).

The same membranes used to detect tyrosine phosphorylation of IRS-1 and -2 were reblotted with antibodies against the p85 subunit of PI 3-kinase. The PI 3-kinase association with IRS-1 and -2 paralleled the changes in the phosphorylation of these proteins (Fig. 3A and C, middle panels). There were no changes in IRS-1 and -2 protein expressions (Fig. 3A and C, lower panels). To determine whether there was PI 3-kinase activity in IRS-1 and -2 immunoprecipitates, hypothalami were prepared and immunoprecipitated with anti–IRS-1 or anti–IRS-2 antibodies from both control and exercised rats. After treatment with leptin, there was an increase in PI 3-kinase activity associated with IRS-1 and -2. In the exercised animals, leptin increased PI 3-kinase activity associated with IRS-1 and -2 by 5.5- and 5.2-fold, respectively, compared with 2.4- and 3.8-fold increases in the hypothalami from control rats, representing increases in PI 3-kinase activity associated with IRS-1 and -2 of 2.0- and 1.6- fold, respectively (Fig. 3B and D).

The effect of insulin, or its vehicle, on the control of food intake was studied by measuring total food intake for 12 h after a session of exercise and a single intracerebroventricular injection of insulin or its vehicle. Insulin induced reductions in 12-h food intake in both exercised and control rats in a concentration-dependent manner. In the exercised animals, insulin (5, 10, and 20 mU) reduced food intake by 46.4, 49.1, and 63.3%, respectively, whereas in the control group, these doses induced reductions of 29.2, 27.7, and 47.9%, indicating that insulin was much more effective in exercised rats (Fig. 4A).

To determine the effects of exercise on the early steps of the insulin signaling pathway, insulin receptor, IRS-1, and IRS-2 tyrosine phosphorylation was assessed in the hypothalamus of trained and control rats. Immunoprecipitation and Western blotting of hypothalamic extracts were performed using anti–insulin receptor, anti–IRS-1, anti–IRS-2, and anti-phosphotyrosine antibodies. Insulin (10 mU) induced increases in insulin receptor, IRS-1, and IRS-2 tyrosine phosphorylation levels in the hypothalamus of both control and exercised rats. In the exercised animals, insulin increased insulin receptor, IRS-1, and IRS-2 tyrosine phosphorylation by 6.0-, 5.5-, and 5.8-fold, respectively, compared with 2.7-, 2.3-, and 2.5-fold increases in the hypothalami from control rats, representing increases in insulin receptor and IRS-1/2 tyrosine phosphorylation of 3.0-, 3.6-, and 3.2-fold, respectively (Fig. 4B, C, and D, upper panels). The same membranes used to detect tyrosine phosphorylation of IRS-1/2 were reblotted with antibodies against the p85 subunit of PI 3-kinase. The PI 3-kinase association to IRS-1 and -2 paralleled the changes in phosphorylation of these proteins (Fig. 4C and D, middle panels).

To determine whether there was PI 3-kinase activity in the IRS-1/2 immunoprecipitates, hypothalami from control and exercised rats treated with insulin (10 mU), after a session of exercise, were prepared and immunoprecipitated with anti–IRS-1 or anti–IRS-2 antibodies. After treatment with insulin, there was an increase in PI 3-kinase activity associated with IRS-1 and -2. In the exercised animals, insulin increased PI 3-kinase activity associated with IRS-1 and -2 by 5.8- and 6.0-fold, respectively, compared with 2.5- and 2.6-fold increases in the hypothalami from control rats, representing increases in PI 3-kinase activity associated with IRS-1 and -2 of 2.0- and 1.6-fold, respectively (Fig. 5A and B).

Hypothalamic extracts from exercised and control rats that were stimulated with insulin (10 mU) were lysed and the proteins separated by SDS-PAGE on gel and blotted with pAkt antibodies. In the hypothalami from exercised animals, insulin increased Akt serine 473 and Akt threonine 308 phosphorylation by 5.7- and 6.0-fold, respectively, compared with 2.8- and 3.7-fold increases in the hypothalami from control rats, representing increases in Akt serine phosphorylation of 2.7-fold (Fig. 5C, upper panel). No changes were observed in Akt protein expression (Fig. 5C, lower panel).

IL-6 expression was detected in control animals; however, a 2.8-fold increase was observed in exercised animals (Fig. 6A). We tested whether the inhibitory effects of leptin and insulin on food intake depend on IL-6 by intracerebroventricular infusion of anti–IL-6 into exercised rats. Treatment with leptin or insulin markedly reduced 12-h food intake in exercised rats pretreated with vehicle, although pretreatment with anti–IL-6 blocked exercise-induced leptin (Fig. 6B) and insulin (Fig. 6C) sensitivity in a concentration-dependent manner, respectively (Fig. 6B and C). Consistent with the increase in leptin sensitivity, JAK2 (Fig. 6D) and STAT3 (Fig. 6E) phosphorylation were induced by exercise and reversed by anti–IL-6 in accordance with the control of food intake. Insulin induced a significant increase in insulin receptor (Fig. 6F), Akt serine 473 (Fig. 6G), and threonine (Fig. 6H) phosphorylation in the hypothalamus of exercised rats pretreated with vehicle. In animals pretreated with anti–IL-6, the effect of exercise on insulin signaling was reversed in a concentration-dependent manner.

Exercise training has multiple effects on metabolism and gene expression (31). However, little is known about the mechanisms by which exercise leads to reduced appetite. Here, we provide evidence for a molecular mechanism to account for increased insulin and leptin action in hypothalamus after exercise. Intracerebroventricular insulin or leptin infusion in doses that did not alter insulinemia or leptinemia reduced food intake in exercised rats to a greater extent than that observed in control animals. Exercise was associated with a marked increase in the phosphorylation/activity of several proteins involved in leptin and insulin signal transduction in hypothalamus. In addition, we investigated the regulatory role of IL-6 in mediating the increase in leptin and insulin sensitivity in hypothalamus. Treatment with insulin or leptin markedly reduced food intake in exercised rats that were pretreated with vehicle, although no increase in sensitivity to leptin- and insulin-induced anorexia was detected after pretreatment with anti–IL-6 antibody. Increased leptin and insulin action in the brain may thus contribute to the modulation of energy homeostasis in exercised rats.

Despite a recent publication showing that the 12-week wheel exercise protocol reduced the expression of ObRb mRNA in the arcuate nucleus (32), our data demonstrate that after a session of exercise, there were no changes in the expression of hypothalamic proteins involved in leptin and insulin signal transduction. However, the phosphorylation status of these proteins was deeply modified. Exercise led to an increase in leptin- and insulin-stimulated ObR/JAK2 and insulin receptor tyrosine phosphorylation, respectively. The next step in leptin and insulin signaling may involve the tyrosine phosphorylation of IRS-1 and -2. As shown above, the amounts of IRS-1 and -2 were unchanged in the hypothalamus of exercised rats. In contrast, the phosphorylation of IRS-1 and -2 after stimulation with leptin or insulin increased in those rats compared with control animals. Because IRS-1 and -2 are the main molecules linking leptin and insulin signaling to PI 3-kinase activity, we examined the leptin- and insulin-induced association of IRS-1 and -2 with the p85 subunit of PI 3-kinase and found it to be increased in the hypothalamus of exercised rats. After IRS-1–or IRS-2–PI 3-kinase association, PI 3-kinase is activated and may in turn activate Akt, a serine kinase with pleiotropic actions in several tissues (33). The activation of Akt-1/protein kinase B is accompanied by an increase in its serine and threonine phosphorylation (22). Thus, the increase in the association between IRS-1 or IRS-2 and PI 3-kinase, and the increase in PI 3-kinase activity after leptin or insulin infusion in the hypothalamus of exercised rats, may play a role in the increased responsiveness to leptin and insulin in these animals.

Selective impairment of leptin and insulin signaling through the PI 3-kinase pathway in the hypothalamus could be pathophysiologically important in the development of obesity. Recent studies have shown that activation of the PI 3-kinase pathway could be involved in the anorexigenic effect of insulin or leptin (14,34,35). Our findings, in a model of exercise training, are relevant because insulin-induced tyrosine phosphorylation of IRSs and PI 3-kinase activity are reduced in the hypothalamus of different animal models of obesity (14,36,37). Thus, exercise training may be one therapeutic strategy to restore impaired leptin and insulin signal transduction in the hypothalamus of obese individuals.

In addition to the increased insulin and leptin sensitivity observed in the PI 3-kinase pathway, our data also provide evidence that there is an increase in leptin sensitivity in the JAK2/STAT3 pathway. Leptin activation of STAT3 requires the leptin receptor, which associates with and activates JAK2 in a ligand-dependent manner (8,9,12,38). One potential mediator of increased STAT3 activation in the hypothalamus of exercised rats is expression of SOCS-3, a suppressor of cytokine signaling. Forced expression of SOCS-3 in mammalian cells antagonizes leptin signaling, probably by binding and antagonizing JAK activity (39). Using Western blotting, we examined the expression of SOCS-3 in hypothalami of exercised rats. No significant differences were found between the two groups (data not shown). In addition, we found no significant difference in hypothalamic protein-tyrosine phosphatase 1b expression and its association with JAK2 and ObR between the two groups (data not shown). Thus, a molecular basis for the observed increase in leptin’s ability to activate STAT3 signaling after 1 day of exercise remains to be determined.

Perhaps the most striking finding was the reversal of exercise-induced increased hypothalamic insulin and leptin sensitivity by blocking the action of IL-6 action. These data are in accordance with earlier studies demonstrating that IL-6 treatment enhances energy expenditure in both rodents and humans (30,40–42). It has been previously shown that IL-6 treatment stimulates energy expenditure at the level of the brain in rodents (30,41,43), and it might be assumed that endogenous IL-6 also acts on the brain during exercise. The IL-6 exerting this effect during exercise could be produced by the brain itself, which has been shown to have increased IL-6 production during exercise (44). Alternatively, the large quantities of endocrine IL-6 produced from working skeletal muscle (45) might reach appropriate sites in the brain (23,46,47).

Numerous biological responses of different cell types are induced by IL-6, which activates STAT3 and Ras–extracellular signal-regulated kinase-1/2 via JAKS, and the balance of activation of both pathways is considered to direct the cell fate in response to IL-6 (48). The cross talk of signals mediated by a cytokine and growth factor has been previously reported in the case of the phosphorylation of tyrosine kinase receptors by the growth hormone–activated JAK2 (49). This suggests that the IL-6–induced activation of JAK2 is involved in the activation of insulin and leptin receptor–mediated signals in rat hypothalamus. Conversely, it has been reported that the PI 3-kinase and Akt pathways may be activated via gp130 (the glycoprotein of 130 kDa) recruitment of adaptor molecules to create binding sites to the SH2 domain of the p85 subunit of PI 3-kinase (50).

In the current study, using an in vivo approach, we saw a synergistic effect of IL-6 on the insulin-stimulated tyrosine phosphorylation of IRS-1 and on the serine phosphorylation of protein kinase B/Akt in rat hypothalamus. These results are clearly different from the findings of Senn et al. (51) in HepG2 cells and may indicate that the induction of SOCS-3 by IL-6 either follows a different time course in hepatocytes or that the major effect of IL-6 is exerted through other mechanisms, such as the transcriptional regulation identified in the current work. However, the recent finding (52) that a high IL-6 infusion for 2 h in rats did not reduce the insulin effect during a euglycemic clamp clearly supports the theory that any acute inhibitory effects of IL-6, mediated through a transient activation of SOCS-3, are of less importance for whole-body insulin sensitivity. Similar results have recently been reported in humans (53). Because IL-6 has been shown both to interfere with insulin-signaling pathways in the liver and adipocytes in an inhibitory manner and to reduce insulin-stimulated glycogen synthesis in hepatocytes, the cross talk of IL-6 with insulin-signaling pathways appears to be tissue specific.

The current study provides direct measurements of leptin and insulin signaling in the hypothalamus, and it documents increased sensitivity to these hormones in the hypothalamus of exercised rats in an IL-6–dependent manner. These findings provide support for the hypothesis that exercise could have appetite-suppressive actions mediated by the hypothalamus.

These studies were supported by grants from Fundacão de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).

We are thank Dr. Nicola Conran for English language editing.