Leptin is a key regulator of glucose metabolism in mammals, but the mechanisms of its action have remained elusive. We now show that signaling by extracellular signal–regulated kinase (ERK) and its upstream kinase MEK in the ventromedial hypothalamus (VMH) mediates the leptin-induced increase in glucose utilization as well as its insulin sensitivity in the whole body and in red-type skeletal muscle of mice through activation of the melanocortin receptor (MCR) in the VMH. In contrast, activation of signal transducer and activator of transcription 3 (STAT3), but not the MEK-ERK pathway, in the VMH by leptin enhances the insulin-induced suppression of endogenous glucose production in an MCR-independent manner, with this effect of leptin occurring only in the presence of an increased plasma concentration of insulin. Given that leptin requires 6 h to increase muscle glucose uptake, the transient activation of the MEK-ERK pathway in the VMH by leptin may play a role in the induction of synaptic plasticity in the VMH, resulting in the enhancement of MCR signaling in the nucleus and leading to an increase in insulin sensitivity in red-type muscle.
Leptin is an adipocyte-derived hormone that plays an important role in glucose metabolism in peripheral tissues as well as in overall energy metabolism in mammals (1,2). Treatment with leptin ameliorates diabetes in lipodystrophic mice and humans (3–5) as well as type 1 (6,7) and obesity-unrelated type 2 diabetes (8) in rodents. Although the antidiabetic effects of leptin are known to be mediated by the central nervous system (9–11), the mechanism by which leptin stimulates glucose utilization in muscle has remained unclear.
Neurons in the arcuate hypothalamic nucleus (ARC) and ventromedial hypothalamus (VMH) contribute to the effects of leptin on glucose metabolism. Restoration of expression of the Ob-Rb receptor for leptin in proopiomelanocortin (POMC) neurons of db/db mice (which lack Ob-Rb) normalizes blood glucose concentration (12,13). The hyperinsulinemia and insulin resistance characteristic of these animals remain unaffected, however, suggesting that other brain regions may also regulate glucose metabolism. We previously showed that injection of leptin into the VMH increases glucose uptake by skeletal muscle (mainly the red type), brown adipose tissue (BAT), and the heart, but not by white adipose tissue, through activation of the melanocortin receptor (MCR) in the VMH (14). These effects of leptin were manifest at 6 h after injection (14) and were abolished by attenuation of sympathetic nerve signaling through surgical denervation or through administration of either a blocker of sympathetic nerve activity (guanethidine) or the β-adrenergic antagonist propranolol (11,15). Furthermore, whereas leptin injection into the VMH increased glucose uptake in muscle, BAT, and the heart, injection into the ARC increased glucose uptake in BAT alone, and injection into the dorsomedial hypothalamus (DMH) or paraventricular hypothalamus (PVH) had no effect (14). The effect of leptin on muscle glucose uptake is thus dependent on Ob-Rb activation in the VMH, as well as on Ob-Rb activation in the ARC.
Activation of Ob-Rb stimulates intracellular signaling pathways, including those mediated by signal transducer and activator of transcription 3 (STAT3), phosphoinositide 3-kinase (PI3K), and extracellular signal–regulated kinase 1 or 2 (ERK1/2) (1,2,16). Leptin also downregulates the activity of AMP-activated protein kinase in the ARC and PVH, an effect that contributes to the anorexic action of leptin (17). With the use of a hyperinsulinemic-euglycemic clamp and measurement of 2-deoxyglucose (2DG) uptake, we have now examined the role of leptin signaling in the VMH in the acute effects of leptin injected into the periphery or the VMH on glucose metabolism in skeletal muscle of lean mice. Our results reveal that signaling by ERK and its upstream kinase MEK in the VMH mediates the leptin-induced increase in glucose utilization and its insulin sensitivity both in the whole body and in red-type skeletal muscle through activation of MCR in the VMH. In contrast, leptin in the VMH was found to enhance the insulin-induced suppression of endogenous glucose production (EGP), which largely reflects hepatic glucose production, through a STAT3-dependent, MCR-independent pathway in this nucleus.
RESEARCH DESIGN AND METHODS
Male FVB mice (CLEA Japan, Tokyo, Japan) were studied at 12–16 weeks of age. The animals were housed individually in plastic cages at 24 ± 1°C with lights on from 0600 to 1800 h, and they were maintained with free access to a laboratory diet (Oriental Yeast, Tokyo, Japan) and water. All animal experiments were approved by the ethics committee for animal experiments of the National Institute for Physiological Sciences.
A chronic double-walled stainless steel cannula was implanted stereotaxically and either unilaterally into the right side of the VMH or bilaterally into the VMH as described previously (14,18). Bilateral cannula placement was performed for examination of the effects of bilateral injection of the MEK inhibitor U0126 or a STAT3 inhibitor into the VMH in those on systemic injection of leptin. Unilateral cannula implantation was performed for all other studies. For the hyperinsulinemic-euglycemic clamp, polyethylene catheters were inserted into the right carotid artery and jugular vein of mice. Animals were handled repeatedly during the recovery period (2 weeks) after cannula implantation. Correct placement of the cannula tips was verified microscopically in brain sections, with >95% of animals manifesting correct placement; the few animals with incorrect cannula placement were excluded from analysis. Food was removed immediately before the administration of inhibitors, leptin, or the MCR agonist melanotan-II (MT-II).
Administration of leptin, MT-II, and inhibitors.
Leptin (5 ng) (PeproTech, Rocky Hill, NJ), the MCR agonist MT-II (10 pmol) (Phoenix Pharmaceuticals, Burlingame, CA), or the MCR antagonist SHU9119 (10 pmol) (Phoenix Pharmaceuticals) in 0.1 µL of physiological saline was injected with a Hamilton microsyringe into the right side of the VMH of freely moving mice through the unilateral cannula. The MEK inhibitor U0126 (10 μmol/L) (Cell Signaling Technology, Beverly, MA) or the PI3K inhibitor LY294002 (10 μmol/L) (Merck, Darmstadt, Germany) in 0.1 µL of 0.01% DMSO was injected into the same side of the VMH at 1 h before injection of leptin or MT-II. Cell-permeable SH2 domain–binding phosphopeptide (STAT3 inhibitor, Merck) (0.1 µL of a 250 μmol/L solution in saline) was injected into the VMH twice, at 1 h and 5 min before leptin injection. In some experiments, 0.1 µL of U0126 (10 μmol/L in 0.01% DMSO) or the STAT3 inhibitor (250 μmol/L in saline) was injected into the VMH bilaterally at 1 h or at both 1 h and 5 min, respectively, before intraperitoneal injection of leptin (5 mg/kg). Control mice were injected with the same volume of saline or 0.01% DMSO into the VMH or with intraperitoneal saline as appropriate.
Hyperinsulinemic-euglycemic clamp and measurement of associated 2-[14C]DG uptake.
Four hours after leptin or MT-II injection, the hyperinsulinemic-euglycemic clamp protocol was initiated in conscious and unrestrained mice. The protocol was modified slightly from that described on the website of the Mouse Metabolic Phenotyping Center at Vanderbilt University (http://www.mc.vanderbilt.edu/root/vumc.php?site=mmpc&doc=32773). The 120-min basal period (t = –120 to 0 min) was initiated at 1300 h and was followed by a 105-min clamp period (t = 0–105 min) beginning at 1500 h (Fig. 1A). A priming dose of [3-3H]glucose (5 μCi) (American Radiolabeled Chemicals, St. Louis, MO) was administered via the jugular vein catheter at t = –120 min and was followed by infusion of the tracer at a rate of 0.05 µCi/min for 2 h. The clamp period was initiated at t = 0 min by primed and continuous infusion of bovine insulin (bolus of 16 mU/kg followed by a rate of 5 mU ⋅ kg–1 ⋅ min–1) (Sigma-Aldrich Japan, Tokyo, Japan) through the jugular vein catheter. The rate of [3-3H]glucose infusion was increased to 0.1 µCi/min for the remainder of the experiment in order to minimize changes in specific activity relative to the equilibration period. Blood was collected every 5–10 min from the carotid artery catheter, and blood glucose was monitored (One Touch Ultra; Lifescan, Johnson & Johnson). Glucose (30%) was infused at a variable rate via the jugular vein catheter in order to maintain blood glucose levels at 130–150 mg/dL. Withdrawn erythrocytes were suspended in sterile 0.9% saline and returned to each animal.
Tissue 2DG uptake was measured as described previously (14). For assessment of 2DG uptake during the basal period, mice were infused with 2-[14C]DG (5 µCi) (American Radiolabeled Chemicals) at t = –45 min through the jugular vein catheter. At t = –40, –30, –20, –10, and 0 min, an arterial blood sample (50 µL) was collected for assessment both of the rate of blood glucose appearance (Ra), which reflects EGP, and of 2DG uptake. For measurement of 2DG uptake during the clamp period, another group of mice was infused with 2-[14C]DG (5 µCi) at t = 60 min, and blood samples (50 µL) were collected at t = 65, 75, 85, 95, and 105 min. Immediately after collection of the final blood sample (t = 0 or 105 min), mice were killed with an overdose of pentobarbital sodium, and the soleus, red (Gastro-R) and white (Gastro-W) portions of the gastrocnemius, epididymal white adipose tissue (epiWAT), and liver were rapidly dissected. Gastro-R was dissected from the inner surface of the gastrocnemius attached to the soleus, whereas Gastro-W was dissected from the outer surface of the muscle. The rate of disappearance of blood glucose (Rd), which reflects whole-body glucose utilization, as well as Ra, the rates of whole-body glycolysis and glycogen synthesis, and the rates of glycolysis and glycogen synthesis in muscle were determined as described previously (19,20). Rd is equal to Ra plus the glucose infusion rate (GIR) during the clamp period, whereas Rd is equal to Ra during the basal period. Plasma concentrations of insulin (Insulin ELISA; Shibayagi, Gunma, Japan) and glucagon (Glucagon EIA kit; Yanaihara Institute, Shizuoka, Japan) were measured with the use of kits. Plasma epinephrine and norepinephrine concentrations were measured by high-performance liquid chromatography as described previously (18). Glycogen phosphorylase a activity in liver was measured as described previously (21) and was expressed as the ratio of activity in the absence of AMP to that in the presence of 3 mmol/L AMP.
The right side of the ARC, VMH, or DMH sampled at 30 min after leptin injection into the VMH or at 1 h after intraperitoneal injection of leptin was dissected from a 1-mm-thick sagittal section prepared from the midline of the fresh brain and was subjected to immunoblot analysis as described previously (14) (Supplementary Fig. 1). The primary antibodies included those to Tyr705-phosphorylated STAT3 (pSTAT3), Thr202- and Tyr204-phosphorylated p44/42 MAPK (pERK1/2), and Ser473-phosphorylated Akt (pAkt) (all from Cell Signaling Technology); those to Ser549-phosphorylated synapsin (Thermo Scientific Pierce, Rockford, IL); and those to total forms of these various proteins (Cell Signaling Technology). All antibodies were used at a dilution of 1:1,000.
Immunofluorescence analysis of phosphorylated STAT3 and ERK.
At 1 h after intraperitoneal leptin injection, mice were anesthetized and perfused transcardially with 4% paraformaldehyde in 0.1 mol/L phosphate buffer. Brain tissue was removed, fixed again, and embedded in OCT compound (Sakura Finetechnical, Tokyo, Japan). Phosphorylated forms of STAT3 and ERK in the same cryosections (thickness of 7 μm) were detected by consecutive incubations with rabbit polyclonal antibodies to Tyr705-phosphorylated STAT3 (1:100 dilution) (Cell Signaling Technology) and Alexa Fluor 568–labeled goat antibodies to rabbit immunoglobulin G (1:400 dilution) (Life Technologies, Carlsbad, CA), and then with rabbit polyclonal antibodies to Thr202- and Tyr204-phosphorylated ERK1/2 (1:500 dilution) (Cell Signaling Technology) and Alexa Fluor 488–labeled goat antibodies to rabbit immunoglobulin G (1:400 dilution) (Life Technologies). Sections were examined with a fluorescence microscope (Olympus AX-70) and a laser confocal microscope (Digital Eclipse C1; Nikon). Staining was absent in control sections processed without primary antibodies.
Extraction of RNA and reverse transcription PCR analysis.
Reverse transcription and real-time PCR analysis with Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) were performed as described previously (14,18). The sequences of PCR primers (forward and reverse, respectively) were 5′-CATGGGCGCAGCAGGTGTATACT-3′ and 5′-CAAGGTAGATCCGGGACAGACAG-3′ for glucose-6-phosphatase (G6Pase), 5′-GGTGTTTACTGGGAAGGCATC-3′ and 5′-CAATAATGGGGCACTGGCTG-3′ for PEPCK, and 5′-AACTTTGGCATTGTGGAAGG-3′ and 5′-ACACATTGGGGGTAGGAACA-3′ for glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Data were normalized by the corresponding abundance of GAPDH mRNA.
Data are presented as means ± SEM. Statistical comparisons between two groups and among multiple groups were performed with Student t test and with ANOVA followed by Tukey HSD post hoc test, respectively. A P value of <0.05 was considered statistically significant.
Leptin injection into the VMH stimulates whole-body glucose metabolism.
We examined the effects of leptin injection into the VMH on glucose metabolism with the use of a hyperinsulinemic-euglycemic clamp (Fig. 1A). The dose of leptin was selected on the basis of the results of our previous study (14). We also selected the dose of insulin as 5 mU ⋅ kg–1 ⋅ min–1 for the clamp on the basis of the results of preliminary experiments showing that this dose increased the rate of glucose disappearance (Rd, reflecting whole-body glucose utilization) about twofold and suppressed the rate of glucose appearance (Ra, reflecting EGP) by about one-half. Immunoblot and immunohistochemical analyses revealed that leptin injection into the VMH increased the phosphorylation of STAT3, ERK1/2, and Akt (which functions downstream of PI3K) in this brain region but not in the ARC or DMH (Fig. 1B and Supplementary Fig. 2). Preliminary data revealed that the phosphorylation of ERK1/2 peaked at 30 min and returned to the control level at 6 h after leptin injection into the VMH (data not shown).
During the clamp period, the plasma insulin concentration increased 1.78- and 1.63-fold in mice injected with saline or leptin, respectively, with these values not differing significantly (Supplementary Table 1). The blood glucose concentration was maintained constant by glucose infusion (Fig. 1C). Leptin injection into the VMH necessitated an increase in GIR (Fig. 1D). The increase in GIR was associated with an increase in Rd (Fig. 1E) and a decrease in Ra (Fig. 1F). Leptin injection also significantly increased whole-body glycolysis and tended to increase glycogen synthesis (Fig. 1G). Leptin enhanced the insulin-induced increase in 2DG uptake in muscle, with this effect being more pronounced in red-type muscle (soleus and Gastro-R) than in white-type muscle (Gastro-W) (Fig. 1H). The increase in 2DG uptake in red-type muscle was accompanied by an increase in glycolysis and glycogen synthesis in the muscle tissue (Fig. 1I).
In the basal condition, leptin injection into the VMH increased Rd (Fig. 1E) without affecting the blood glucose level (Fig. 1C). This effect was accompanied by an increase in Ra (Fig. 1F), given that Ra is equal to Rd in the steady-state condition. Leptin increased 2DG uptake in soleus and to a lesser extent in Gastro-R, but not in Gastro-W, similar to its effects during the clamp period (Fig. 1H). Plasma insulin levels did not differ between control and leptin-injected groups during the basal period (Supplementary Table 1). Leptin injection into the VMH thus increased glucose turnover at the whole-body level during the basal period. These results suggested that leptin injection into the VMH increases whole-body and muscle glucose utilization under both basal and clamp conditions, whereas it suppresses hepatic EGP in a manner dependent on plasma insulin concentration.
Leptin in the VMH induces glucose utilization and enhances insulin-induced suppression of EGP differentially via MEK-ERK and STAT3 pathways.
We examined the effects of injection of a MEK inhibitor (U0126), a STAT3 inhibitor (cell-permeable SH2 domain–binding phosphopeptide), and a PI3K inhibitor (LY294002) into the VMH. Whereas determination of the acute effects of specific inhibitors minimizes the influence of adaptive responses of neuronal circuits in the brain, the specificity of such agents depends on their concentration. We therefore first determined the doses of inhibitors that preferentially attenuated the activation of their target molecules (Fig. 2A). Plasma insulin and blood glucose levels were not affected by injection of these inhibitors into the VMH, or by similar injection of leptin, the MCR agonist MT-II, or the MCR antagonist SHU9119 (Supplementary Fig. 3 and Supplementary Table 1). The MEK inhibitor U0126 and the STAT3 inhibitor, but not the PI3K inhibitor LY294002, suppressed the leptin-induced increase in GIR during the clamp period, with the effect of U0126 being greater than that of the STAT3 inhibitor (Fig. 2B). None of these three inhibitors affected GIR in response to insulin infusion alone. U0126 inhibited the effect of leptin on Rd but not on Ra during the clamp period, whereas the STAT3 inhibitor attenuated the effect of leptin on Ra but not on Rd (Fig. 2C and D). LY294002 had no effect on either Rd or Ra. The leptin-induced increase in whole-body glycolysis was suppressed by U0126 but not by the other inhibitors (Fig. 2E). Furthermore, the MEK inhibitor, but not the STAT3 inhibitor or PI3K inhibitor, attenuated leptin-induced 2DG uptake, glycolysis, and glycogen synthesis in soleus and Gastro-R, but not in Gastro-W, during the clamp period (Fig. 3A and C–E). Uptake of 2DG by epiWAT was not affected by leptin or by the inhibitors (Fig. 3B and F). U0126 also suppressed the increase in Rd (equal to that in Ra) induced by leptin during the basal period (data not shown).
Insulin downregulated the amounts of PEPCK and glucose-6-phosphatase (G6Pase) mRNAs in the liver during the clamp period, whereas leptin and the leptin signaling inhibitors did not affect the hepatic abundance of these mRNAs (Supplementary Fig. 4A). In contrast, leptin attenuated the activity of glycogen phosphorylase a in the liver during the clamp period, and this effect was blocked by the STAT3 inhibitor but not by the MEK inhibitor or the PI3K inhibitor (Fig. 3G). These results suggested that leptin in the VMH enhances the insulin-induced suppression of hepatic glucose production via STAT3 signaling in the VMH.
The plasma glucagon concentration did not differ between control and leptin-injected mice either with or without insulin infusion (Supplementary Fig. 4B). Injection of inhibitors of leptin signaling pathways into the VMH also had no effect on the plasma glucagon level. Furthermore, plasma norepinephrine and epinephrine concentrations had decreased to undetectable levels at the end of the clamp period in both control and leptin-injected mice (data not shown), probably as a result of adequate training and handling of the animals as well as the high plasma insulin concentration. Whereas humoral factors may contribute to the leptin-induced suppression of EGP and hepatic glycogen phosphorylase a activity, the effect of leptin on EGP may be mediated by the autonomic nervous system, including the vagus nerve, as described previously (8,22,23).
Systemic leptin increases glucose utilization via MEK-ERK signaling in the VMH.
The intraperitoneal injection of leptin (5 mg/kg) increased the amounts of pERK1/2 and pSTAT3, but not pAkt, in the VMH (Fig. 4A). Leptin also increased the amounts of pERK1/2 and pSTAT3 in the ARC as well as pSTAT3 in the DMH (Supplementary Fig. 5A). The intraperitoneal injection of leptin increased the number of cells positive for both pSTAT3 and pERK in the VMH (Fig. 4B) as well as in the ARC (Supplementary Fig. 5B). Systemic leptin thus activates ERK and STAT3 in the same neurons in the VMH as well as in the ARC, probably in a manner dependent on Ob-Rb. Prior bilateral injection of the MEK inhibitor U0126 into the VMH suppressed the increase in ERK phosphorylation in the VMH induced by intraperitoneal injection of leptin without affecting the phosphorylation of STAT3 or Akt in the VMH (Fig. 4A) or of any of the three signaling molecules in the ARC or DMH (Supplementary Fig. 5A).
Injection of U0126 into the VMH partially inhibited the increase in GIR induced by intraperitoneal injection of leptin (Fig. 4C). This effect of U0126 was associated with suppression of the leptin-induced increases in Rd (Fig. 4D) and 2DG uptake in the soleus (Fig. 4E). The MEK inhibitor did not affect the leptin-induced enhancement of the suppression of Ra by insulin (Fig. 4F). In the basal period, U0126 injection into the VMH inhibited the leptin-induced increases in Rd (Fig. 4D), 2DG uptake in the soleus (Fig. 4E), and Ra (Fig. 4F), the latter being equal to the increase in Rd (Fig. 4D). Blood glucose and plasma insulin levels during the basal or clamp periods were not affected by intraperitoneal injection of leptin with or without prior injection of U0126 into the VMH (Supplementary Fig. 3 and Supplementary Table 1). The enhancement by leptin of the insulin-induced suppression of hepatic EGP was only partially inhibited by injection of the STAT3 inhibitor into the VMH (data not shown). Hepatic EGP is thus regulated by leptin via other brain sites or peripheral tissues as well as via the VMH.
MCR in the VMH mediates leptin-induced glucose utilization (but not suppression of EGP) in a manner independent of MEK-ERK signaling.
Injection of the MCR antagonist SHU9119 into the VMH resulted in partial inhibition of the increase in GIR induced by leptin injection into the VMH (Fig. 5A). This inhibition of GIR by SHU9119 was associated with suppression of the leptin-induced increases in Rd (Fig. 5B) and 2DG uptake in soleus muscle (Fig. 5C). SHU9119 did not affect the enhancement by leptin of the insulin-induced suppression of Ra (Fig. 5D). During the basal period, SHU9119 injection into the VMH inhibited the leptin-induced increases in Rd (Fig. 5B), 2DG uptake in soleus (Fig. 5C), and Ra (Fig. 5D), with the latter being equal to the increase in Rd (Fig. 5B).
Injection of the MCR agonist MT-II into the VMH increased both GIR (Fig. 6A) and Rd (Fig. 6B) during the clamp period. Injection of MT-II into the VMH also enhanced insulin-induced 2DG uptake in soleus but not in Gastro-W or epiWAT (Fig. 6C). However, the MT-II–induced increases in GIR, Rd, and 2DG uptake in soleus were not inhibited by injection of the MEK inhibitor U0126 into the VMH (Fig. 6A–C). Moreover, MT-II did not enhance the suppression of Ra induced by insulin during the clamp period (Fig. 6D). Consistent with our previous results showing that MT-II increased 2DG uptake in red-type skeletal muscle (14), MT-II increased both Rd (Fig. 6B) and Ra (Fig. 6D) in the basal period, and these effects were not inhibited by U0126. Whereas MT-II did not increase glucose uptake in Gastro-W during the clamp period, we previously found that MT-II induced a small (1.3-fold) increase in 2DG uptake in white muscle in the absence of insulin infusion (14). The effect of MT-II as well as leptin on glucose uptake and its insulin sensitivity in white skeletal muscle is thus markedly smaller than in red muscle. These data thus suggested that the increases in whole-body glucose utilization as well as in glucose utilization by red-type muscle induced by leptin in the VMH, but not the enhancement by leptin of the suppression of EGP by insulin, are mediated by MCR in the VMH.
Leptin has previously been shown to increase the density of hippocampal synapses and of N-methyl-d-aspartate–sensitive glutamate receptors at these synapses in a manner dependent on ERK signaling (24). Finally, we examined the effects of leptin on the phosphorylation of synapsin, which contributes to synapse formation, in the VMH. Leptin injection into the VMH increased the phosphorylation of synapsin in the VMH in a manner dependent on MEK-ERK signaling (Fig. 6E).
Central or peripheral administration of leptin has beneficial effects on diabetes in lipodystrophic mice and humans (3–5) as well as on type 1 (6,7) and obesity-unrelated type 2 diabetes in rodents (8). We have now shown that MEK-ERK signaling in the VMH mediates the leptin-induced acute increase in glucose uptake in red-type muscle as well as the insulin sensitivity of this process, and that it thereby contributes to the leptin-induced increase in whole-body glucose utilization (Fig. 7). In contrast, leptin in the VMH enhances insulin-induced suppression of hepatic EGP through the action of STAT3 in the VMH and the inhibition of glycogen phosphorylase a activity in the liver (Fig. 7). Given that leptin stimulates hepatic EGP under basal conditions, it reciprocally regulates this process in a manner dependent on the plasma insulin concentration. MCR activation in the VMH contributes to the leptin-induced increase in whole-body and muscle glucose utilization but not to its suppression of EGP. Activation of the MEK-ERK pathway in the VMH appears to occur upstream of and to be necessary for the activation of MCR signaling in the VMH and is required for the increase in muscle glucose uptake.
We previously showed that injection of leptin into the VMH increased muscle glucose uptake in mice at 6 h but not at 3 h after the injection, whereas intracerebroventricular injection of the MCR agonist MT-II increased muscle glucose uptake within 3 h (14). Similarly, intravenous or intracerebroventricular injection of leptin was found to induce a slow but progressive increase in sympathetic nerve activity in peripheral tissues (25,26), whereas intracerebroventricular injection of α-melanocyte–stimulating hormone (α-MSH) resulted in immediate activation of sympathetic nerves (26). We previously showed that the increase in muscle glucose uptake induced by leptin injection into the VMH is mediated by the sympathetic nervous system and β-adrenergic receptors (β-ARs) (15). We also found that injection of the hypothalamic neuropeptide orexin into the VMH stimulates glucose uptake in the muscle of mice, similar to the effect of leptin, and that this action of orexin is mediated via sympathetic nerves and β2-ARs (18). The orexin-induced increase in glucose uptake was blunted in β-AR–deficient mice (β-less mice), whereas restoration of β2-AR expression in the muscle of these mice resulted in recovery of the orexin effect. Furthermore, preliminary data revealed that the phosphorylation of ERK1/2 peaked at 30 min and returned to the control level at 6 h after leptin injection into the VMH (data not shown). These observations suggest that leptin exerts its effects on glucose metabolism by altering neuronal plasticity in the VMH through the activation of MEK-ERK signaling in this region. Consistent with this notion, we found that leptin injection into the VMH increased the phosphorylation of synapsin, which contributes to synapse formation, in the VMH in a manner dependent on MEK-ERK signaling.
We previously showed that injection of leptin into the VMH induced expression of the transcription factor c-FOS in the ARC as well as in the VMH at 6 h after the injection (14). Intense stimulation of VMH neurons by a photoactivatable caged form of glutamate increased the electrical activity of POMC neurons in the ARC (27). Moreover, the ventrolateral region of the VMH and the VMH shell contain a high number of dendrites that harbor a substantial number of axons and boutons immunoreactive for α-MSH, and MT-II and SHU9119 reciprocally regulate the activity of VMH neurons (28). Brain-derived neurotrophic factor is expressed in the ventrolateral region of the VMH and acts downstream of MCR in the VMH (29). Brain-derived neurotrophic factor enhances neuronal plasticity through retrograde action. Early activation of the MEK-ERK signaling pathway in the VMH by leptin may thus induce synaptic plasticity in the VMH, resulting in the enhancement of MCR signaling in the VMH via POMC neurons in the ARC and leading to an increase in insulin sensitivity in red-type muscle (Fig. 7).
STAT3 and the leptin receptor Ob-Rb in VMH neurons are implicated in the homeostatic regulation of glucose metabolism by leptin. Ablation of leptin receptors in steroidogenic factor 1 (SF1)–expressing VMH neurons of mice induced insulin resistance before the onset of obesity (30). Conversely, SF1-specific ablation of suppressor of cytokine signaling-3 (SOCS-3), a feedback inhibitor of the leptin-induced JAK-STAT3 signaling pathway, improved glucose homeostasis in mice (31). Furthermore, intracerebroventricular injection of a STAT3 inhibitor suppressed leptin-induced enhancement of insulin sensitivity in the liver, and abolishment of Ob-Rb–dependent STAT3 signaling (in s/s mice) results in pronounced hepatic insulin resistance (32). Our data suggest that STAT3 in the VMH regulates liver insulin sensitivity. However, other brain regions, such as the ARC and brain stem, may also contribute to such regulation, given that the enhancement of insulin-induced suppression of hepatic EGP elicited by intraperitoneal injection of leptin was only partially attenuated by injection of the STAT3 inhibitor in the VMH.
We found that leptin suppressed EGP during the clamp period of the hyperinsulinemic-euglycemic clamp protocol, whereas it increased EGP during the basal period. The insulin-dependent reciprocal regulation of EGP by leptin may explain why leptin does not induce hypoglycemia in the postprandial state. Insulin inhibits hepatic EGP via the central nervous system (33) as well as through the inhibition of agouti-related peptide–containing neurons in the ARC (34). Insulin in the brain modulates leptin signaling in the hypothalamus (35). The increased plasma insulin levels during the prandial state may thus influence the effect of leptin on hepatic EGP in both the brain and liver.
PI3K signaling in the hypothalamus also regulates glucose and lipid, as well as energy, metabolism (36–38). Ablation of the p110α catalytic subunit of PI3K in POMC neurons was found to impair glucose metabolism (36). In contrast, ablation of this subunit in SF1-expressing neurons did not affect glucose homeostasis (37). These observations as well as our present data suggest that the PI3K pathway in Ob-Rb–expressing neurons in other brain regions, such as the ARC, rather than in the VMH, might contribute to the effects of leptin on glucose metabolism. Furthermore, the PI3K pathway in Ob-Rb–expressing neurons in the VMH might contribute to the chronic rather than the acute effects of leptin on glucose metabolism.
We previously showed that whereas leptin injection into the VMH increased glucose uptake in muscle, BAT, and the heart, injection into the ARC increased glucose uptake in BAT but not in muscle or heart, and injection into the DMH or PVH had no effect (14). Injection of MT-II either into the VMH or intracerebroventricularly increased glucose uptake in muscle, BAT, and the heart, whereas injection into the PVH increased glucose uptake in BAT alone, and injection into the DMH or ARC had no effect (14). Thus, whereas leptin-induced glucose uptake in BAT and muscle is mediated by Ob-Rb and MCR in the VMH, glucose uptake in BAT may also be mediated by Ob-Rb in POMC neurons in the ARC and then MCR in the PVH. PVH neurons project to the solitary nucleus and to the raphe nucleus, and thereby stimulate thermogenesis in BAT (39). Further investigation is necessary to explore the neuronal pathway responsible for leptin-induced glucose uptake in muscle from the VMH to the hindbrain and muscle tissue.
In summary, we have found that the MEK-ERK pathway in the VMH plays an important role in leptin-induced glucose uptake in red-type muscle as well as whole-body glucose utilization. MCR in the VMH also mediates these effects of leptin. In contrast, activation of STAT3 in the VMH by leptin enhances the insulin-induced suppression of hepatic EGP by inhibiting glycogen phosphorylase a activity. These results suggest that VMH neurons mediate the antidiabetic effects of leptin in the control of muscle as well as liver glucose metabolism. They thus provide important insight into the mechanism of the antidiabetic effects of leptin in humans as well as rodents.
This work was supported by Grants-in-Aid for Scientific Research (B) (21390067 and 24390058 to Y.M.), Grants-in-Aid for Young Scientists (B) (20790656 and 22790875 to S.O. and 23790282 to C.T.), and a Grant-in-Aid for Scientific Research on Innovative Areas (Research in a Proposed Research Area, “Molecular Basis and Disorders of Control of Appetite and Fat Accumulation;” 22126005 to Y.M.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan, as well as by the Specific Research Fund of the National Institutes for Natural Sciences (to Y.M.).
No potential conflicts of interest relevant to this article were reported.
C.T. researched data, contributed to discussion, and wrote and edited the manuscript. T.S., H.K., and S.S. researched data, contributed to discussion, and reviewed and edited the manuscript. S.O. contributed to discussion. E.A.C., T.S., Y.O.-O., S.Y., K.T., L.T., and K.S. researched data. Y.M. designed the study, contributed to discussion, and wrote and edited the manuscript. Y.M. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
The authors thank N. Kawai and K. Nagatani for laboratory management, the Center for Analytical Instruments at the National Institutes for Basic Biology (Okazaki, Japan) for biochemical analysis, and K.W. Brocklehurst, an independent scientific editorial consultant (Washington, DC), for editorial assistance.