Hypothalamic Nesfatin-1/NUCB2 Knockdown Augments Hepatic Gluconeogenesis That Is Correlated With Inhibition of mTOR-STAT3 Signaling Pathway in Rats

The hypothalamus, where numerous neuropeptides and transmitters are released to control essential body functions, has been considered a key integrator in the central control of food intake and energy balance (1,2). Oh-I et al. (3) discovered nesfatin-1, an 82–amino acid polypeptide, while searching for new appetite-controlling signals. Nesfatin-1 is the cleavage product of the calcium and DNA–binding protein nucleobindin2 (NUCB2), a 396–amino acid peptide that is highly conserved in rat, mouse, and humans. NUCB2 mRNA is detected in brain nuclei, including the arcuate nucleus (ARC), paraventricular nucleus (PVN), supraoptic nucleus (SON), and nucleus of the solitary tract (3), and in peripheral tissues, such as rat stomach, pancreas, pituitary gland, and testes (4).

Nesfatin-1 has been identified as an anorectic molecule. Thus, central injection of nesfatin-1 in rats decreases food intake and body weight in a dose-dependent manner (3), and several other studies have shown that nesfatin-1, when injected into the lateral (5,6) or the fourth brain ventricle (5), the PVN (7), or the abdominal cavity (8), induces a sustained suppression of dark-phase feeding in rodents. Furthermore, increasing evidence suggests that nesfatin-1 may play an important role in the regulation of glucose metabolism because elevated plasma nesfatin-1 concentrations have been found in subjects with impaired glucose tolerance and type 2 diabetes and to be associated with insulin resistance (9).

Signal transducer and activator of transcription 3 (STAT3), a member of the STAT family of cytoplasmic transcription factors, was found to be activated by glycoprotein 130–coupled cytokines, such as interleukin-6 (10,11). Studies have reported that liver-specific STAT3 knockout mice have increased expression of gluconeogenic genes (12). Furthermore, STAT3 has been found to directly target the regulatory regions of glucose-6-phosphatase (G6Pase) and PEPCK and regulate their expression (11) and has been shown to be involved in the pathogenesis of interleukin-6–induced insulin resistance in the liver (13).

We have previously reported that hypothalamic nesfatin-1 is involved in the regulation of hepatic glucose production (HGP) through the hepatic AKT kinase (AKT)/AMP-dependent protein kinase (AMPK)/mammalian target of rapamycin (mTOR) pathway (14), suggesting a connection between hypothalamic nesfatin-1 and the STAT signaling pathway. To our knowledge, such a connection has not been reported, and it has been suggested that this connection be confirmed by a loss-of-function study (15). In the current study, therefore, we determined whether attenuation of the central nesfatin-1/NUCB2 signal leads to a change in peripheral glucose metabolism. To this end, we created a loss-of-function animal model by knocking down the hypothalamic nesfatin-1/NUCB2, and this has allowed us to determine the molecular mechanisms by which the central nesfatin-1/NUCB2 knockdown can lead to alteration in insulin signaling.

To construct vectors expressing short hairpin RNA against nesfatin-1/NUCB2, we designed several oligonucleotides and complementary strands to target specifically rat nesfatin-1/NUCB2. A recombinant vector Ad-shNUCB2 was generated by transfection in 293 cells with the AdEasy system. The Ad-shNUCB2 (sequence 5′-GATCCCCG GTGGAAAGTGCAAGGATATTCAAGAGATATCCTTGCACTTTCCACCTTTTTGGAAA-3′) was the most effective one and, therefore, was selected for subsequent experiments. A control recombinant vector Ad-shGFP encoding enhanced green fluorescence protein (Clontech, Mountain View, CA) was used as a control for viral infection. Large-scale amplification and purification of recombinant adenoviruses were performed with the ViraBind Adenovirus Purification Kit according to the manufacturer’s instructions (Cell Biolabs, Inc., San Diego, CA).

One hundred two 14-week-old male Sprague-Dawley rats (Experimental Animal Center of Chongqing Medical University, Chongqing, China) were studied. Rats were housed in a controlled environment, subjected to a standard light cycle (6:00 a.m.–6:00 p.m.) and randomly divided into two groups fed with either a normal chow diet (NCD) or a high-fat diet (HFD) for 10 weeks. The NCD (3.49 kcal/g) provided 60% calories from carbohydrates, 21% from protein, and 19% from fat. The HFD (4.72 kcal/g) provided 33% calories from carbohydrates, 13% from protein, and 54% from fat (lard) (Medicience Ltd., Jiangsu, China). All experimental procedures were approved by the Animal Experimentation Ethics Committee (Chongqing Medical University) and were in accordance with the National Health and Medical Research Council of China Guidelines on Animal Experimentation.

Fourteen days before the in vivo studies, rats were equipped with chronic catheters in the third cerebral ventricle. After being anesthetized with ketamine 87 mg/kg i.p., rats were fixed in a stereotaxic apparatus with ear bars and a nosepiece set at +5.00 mm. A stainless steel guide cannula was implanted into the third ventricle. After full recovery (7 days), rats were equipped with indwelling catheters placed in the right internal jugular vein and the left carotid artery (Fig. 3A) and randomly divided into six groups: 1) NCD artificial cerebrospinal fluid (aCSF) (NCA) (n = 10), 2) NCD-Ad-shGFP (NCG) (n = 10), 3) NCD-Ad-shNUCB2 (NCN) (n = 10), 4) HFD-aCSF (HFA) (n = 10), 5) HFD-Ad-shGFP (HFG) (n = 10), and 6) HFD-Ad-shNUCB2 (HFN) (n = 10). Three days before the clamp studies, rats received an injection of aCSF (10 μL/rat), Ad-shGFP (10⁹ plaque-forming unit [pfu]/rat), or Ad-shNUCB2 (10⁹ pfu/rat). Only rats recovering completely from the surgery were studied.

In different experiments, NCD- and HFD-fed rats infused with aCSF 10 μL, Ad-shGFP 10⁹ pfu, or Ad-shNCUB2 10⁹ pfu into the third ventricle were individually housed. The infusion was initiated at 5:00 p.m. and continued for 1 h at a rate of 10 μL/h. Food intake and body weight were monitored every 24 h for 7 days. All injections were performed once in ad libitum–fed rats maintained in their familiar housing cages. At the end of the 7th day, rats were anesthetized, and liver samples were freeze-clamped in situ and stored at −80°C for glycogen analysis.

Twelve hours before the clamp studies, food was removed from every cage. The clamps were performed as previously described (14). Briefly, a primed continuous infusion of high-performance liquid chromatography–purified [3-H³] glucose (Amersham, Los Angeles, CA; 6 μCi bolus, 0.2 μCi/min) was initiated at 0 min and maintained throughout the study to assess glucose kinetics. A hyperinsulinemic-euglycemic clamp was performed during the final 2 h (120–240 min) of the study. Insulin (6 mU/kg/min) was continuously infused, and a variable infusion of 25% glucose was started and adjusted every 5 min to maintain the plasma glucose concentration at ∼6 mmol (Fig. 3B). Blood samples (100 μL) were obtained from the jugular vein catheter at 0, 120, 200, 220, 230, and 240 min for determination of insulin, nonesterified fatty acid (NEFA), and glucose-specific activity. In another study cohort (n = 4 for each group), 2-deoxy-d-[H³] glucose (2-DG) (Amersham; 30 μCi bolus) was administered 45 min before the end of the clamp studies to determine insulin-mediated glucose uptake in individual tissues. Extra blood samples (50 μL) were taken at 2, 5, 10, 15, 20, 30, and 45 min after the injection to determine tracer disappearance. At the end of the clamp, the rats were anesthetized and tissue samples freeze-clamped in situ with aluminum tongs precooled in liquid nitrogen and stored at −80°C for subsequent analysis.

Plasma glucose was measured with the glucose oxidase method, plasma insulin was measured with a commercial insulin ELISA kit (Diagnostic Products, Los Angeles, CA). NEFA was determined spectrophotometrically with an acyl-CoA oxidase-based colorimetric kit (Wako Pure Chemical Industries, Osaka, Japan). Triglyceride (TG), total cholesterol (TC), HDL cholesterol (HDL-C), and LDL cholesterol (LDL-C) concentrations were measured with enzymatic colorimetric kits. Hepatic glycogen levels were measured with glycogen assay kits (BioVision, Mountain View, CA) (14). Plasma [3-H³] glucose–specific activity was measured in duplicates in the supernatants of Ba(OH)₂ and ZnSO₄ precipitates (Somogyi procedure) of plasma samples after evaporation to dryness to eliminated tritiated water. Under steady-state conditions, the rate of glucose disappearance (GRd) equals the rate of glucose appearance. The latter was calculated as the ratio of the rate of [3-H³] glucose infusion (disintegrations per minute/minute) and the steady-state plasma [H³] glucose–specific activity (disintegrations per minute/mg). When exogenous glucose was given, the rate of HGP was calculated as the difference between the rate of glucose appearance and the glucose infusion rate (GIR) (16). Equation 1 shows the rate of 2-DG uptake (R_i) was calculated as described by Ferré et al. (17):

The lumped constant (LC), which is a correction factor for the discrimination against 2-DG in the glucose transport and phosphorylation pathway, is determined in vitro by comparing the glucose and 2-DG fractional extraction by the different tissues. C*_B is the blood 2-DG expressed in terms of radioactivity, and C_B is the blood glucose concentration.

Total RNA was isolated from frozen tissue with TRIzol reagent (Invitrogen, Grand Island, NY) according to the manufacturer’s instructions. Quantitative real-time RT-PCR was performed with a SYBR Green PCR kit (Takara Bio, Otsu, Japan) and a Corbett Rotor-Gene 6000 real-time PCR system (Corbett Research, Sydney, Australia) according to the manufacturer’s instructions. Gene expressions were analyzed with the comparative threshold cycle method and normalized with β-actin. The following sequences of the primers were used: 5′-CCCTGAACCCTAAGGCCAACCGTGAA AA-3′ and 5′-TCTCCGGAGTCCATCACAATGCCTGTG-3′ for β-actin, 5′-CAC CTTGACAC TACACCCTT-3′ and 5′-GTGGCTGTGAACACCTCT-3′ for G6Pase, and 5′-AGTCACCATCACTTCCTGGAAGA-3′ and 5′-GGTGCAGAATCGCGA GTT-3′ for PEPCK.

After being injected with Ad-shNUCB2 or aCSF for 72 h, the rats were anesthetized by sodium pentobarbital 1 mg/kg i.p. and perfused first with saline containing heparin 20 units/mL for 3 min and then with 4% paraformaldehyde in 0.1 mol/L PBS for 20 min. The brains were excised, transferred to 4% paraformaldehyde, and fixed at 4°C for 24 h. After being dehydrated with ethanol xylene, the specimens were embedded in paraffin, and coronal sections of the hypothalamus were obtained. Immunohistochemistry for nesfatin-1/NUCB2 protein (rabbit antic-NUCB2, 1:200; Abcam, Cambridge, MA) was performed as described previously (18).

For analysis of nesfatin-1/NUCB2 protein expression in the brain, the hypothalamus was dissected out by a horizontal cut 2 mm in depth with the following limits: 1 mm anteriorly from the optic chiasm, the posterior border of mammillary bodies, and the hypothalamic fissures (19). Hypothalamus, liver, and muscle tissues were homogenized, and protein concentration was measured with a BCA quantification kit (Pierce Biotechnology, Rockford, IL). Protein lysates were subjected to 8% SDS-PAGE and transferred to polyvinylidene difluoride membranes. Membranes were probed at 4°C in the presence of 1:1,000 dilutions of primary antibodies against the insulin receptor (InsR), phospho-InsR, insulin receptor substrate 1 (IRS-1), phospho-IRS-1, AKT, phospho-AKT, mTOR, phospho-mTOR, STAT3, phospho-STAT3, and suppressor of cytokine signaling 3 (SOCS3) (all from Cell Signaling Technology, Boston, MA); G6Pase and PEPCK (Santa Cruz Biotechnology, Dallas, TX); and β-actin (Research Diagnostics, Santa Clara, CA). After being washed three times with Tris-buffered saline containing 0.1% Tween-20, the membranes were incubated with horseradish peroxidase–labeled sheep anti-rabbit antibody for 1 h. Next, the blots were visualized with enhanced chemiluminescence. The intensity of the bands was quantified with a Bio Imaging System Densitometer (Bio-Rad, Hercules, CA), and quantification of antigen–antibody complexes was performed with Quantity One analysis software (Bio-Rad).

All results are presented as mean ± SEM. A two-way ANOVA with a least significant difference post hoc test was used to compare mean values between multiple groups, and a two-sample unpaired Student t test was used for two-group comparisons. P < 0.05 was considered significant.

To inhibit hypothalamic nesfatin-1/NUCB2 signaling, we constructed a short hairpin RNA expression vector Ad-shNUCB2 specific to rat nesfatin-1/NUCB2. Three days after injection of Ad-shNUCB2 into the third ventricle, hypothalamic nesfatin-1/NUCB2 protein levels were decreased by 66% (P < 0.05) compared with nesfatin-1/NUCB2 protein levels of rats who received aCSF or Ad-shGFP, whereas no reductions of nesfatin-1/NUCB2 protein abundance were seen in adipose, liver, or muscle tissues (Fig. 1A).

To determine the effect of intracerebroventricular (ICV) Ad-shNUCB2 on nesfatin-1/NUCB2–expressing neurons, we performed immunohistochemistry in rat hypothalamic tissue. As shown in Fig. 1B and C, ICV Ad-shNUCB2 led to a significant decrease in the number of nesfatin-1/NUCB2 immunoreactive cells in the ARC, PVN, SON, and supraoptic retrochiasmatic nucleus (SOR) (P < 0.01), which have been implicated in glucose homeostasis.

ICV injection of Ad-shNUCB2 (10⁹ pfu/rat) caused a significant increase in food intake in NCD- and HFD-fed rats between days 2 and 7 (Fig. 2A–C). Because the greatest increase in food intake induced by ICV Ad-shNUCB2 was on day 3, all subsequent experiments were performed on the third day after ICV Ad-shNUCB2 injection. We also examined the change in body weight on day 7, and no significant differences were found among the rats that received Ad-shNUCB2, Ad-shGFP, or aCSF injection (Fig. 2D). Although HFD-fed rats had significantly decreased hepatic glycogen stores compared with NCD-fed rats (57.04 ± 2.05 vs. 73.36 ± 2.64 mg/g, P < 0.01), hypothalamic nesfatin-1/NUCB2 inhibition failed to change glycogen levels in both NCD- and HFD-fed rats (Fig. 2E).

To examine the effect of ICV administration of Ad-shNUCB2 on glucose kinetics, both NCD- and HFD-fed rats received an injection of Ad-shNUCB2, Ad-shGFP, or aCSF 3 days before the hyperinsulinemic-euglycemic clamp studies. After 12 h of fasting, the plasma insulin, glucose, TC, TG, LDL-C, HDL-C, NEFA levels were similar in NCD-fed rats that received aCSF, Ad-shNUCB2, or Ad-shGFP injections. However, HFD-fed rats had significantly increased plasma insulin, TC, TG, LDL-C, and NEFA levels and decreased HDL-C levels compared with NCD-fed rats (P < 0.01) (Supplementary Table 1). During the clamp, NEFA and TG levels were significantly suppressed in all of groups, although they were higher in HFD-fed rats than in NCD-fed rats (P < 0.05) (Table 1).

GIR, GRd, and HGP were determined during the final 30 min of the clamp studies when steady-state conditions were achieved for plasma glucose and insulin concentrations and GIR. As expected, HFD-feeding decreased GIR (Fig. 3C), GRd (Fig. 3D), and suppression of endogenous glucose production by insulin (Fig. 3E) and increased HGP (Fig. 3F). Hypothalamic nesfatin-1/NUCB2 inhibition significantly decreased GIR in both the NCD-fed (P < 0.01) and the HFD-fed (P < 0.01) groups (Fig. 3C). In addition, ICV Ad-shNUCB2 administration significantly increased HGP in both NCD- and HFD-fed rats compared with the aCSF control rats (P < 0.05 or P < 0.01) (Fig. 3F), Finally, Ad-shNUCB2 injection significantly decreased insulin-induced suppression of endogenous glucose production (Fig. 3E). Taken together, the data show that hypothalamic nesfatin-1/NUCB2 knockdown significantly increased peripheral and hepatic insulin resistance.

To further determine whether central nesfatin-1/NUCB2 knockdown decreases insulin-induced peripheral glucose utilization, 2-DG was injected through the intracarotid catheter during the last 45 min of the hyperinsulinemic-euglycemic clamp studies (Fig. 4A). We found that the 2-DG utilization rate was significantly decreased in gastrocnemius and soleus muscle (Fig. 4B), interscapular brown adipose tissue (Fig. 4C), and white adipose tissue (Fig. 4D) by the central nesfatin-1/NUCB2 knockdown in the NCD and HFD groups. The results indicate that central nesfatin-1/NUCB2 knockdown attenuated peripheral insulin action.

Because central nesfatin-1/NUCB2 knockdown significantly attenuated hepatic insulin action (Fig. 3), we examined whether expressions of PEPCK and G6Pase, two gluconeogenic enzymes, were altered by this knockdown. HFD feeding for 10 weeks increased hepatic PEPCK and G6Pase protein levels (P < 0.05). Both NCD- and HFD-fed rats that received ICV Ad-shNUCB2 had an intriguingly further increased expression of PEPCK and G6Pase proteins and mRNAs (Fig. 5) compared with the aCSF control rats. These results indicate that central nesfatin-1/NUCB2 knockdown blocked, at least in part, the inhibitory effects of insulin on PEPCK and G6Pase in the liver and led to increased HGP.

To determine the underlying mechanisms by which central nesfatin-1/NUCB2 knockdown inhibited insulin signaling, we examined hepatic phosphorylation levels of InsR, IRS-1, and AKT by Western blots of liver and muscle tissue. As shown in Fig. 6, 10 weeks of HFD feeding induced a marked decrease in InsR, IRS-1, and AKT phosphorylation in both liver and muscle tissue (P < 0.05). Upon central Ad-shNUCB2 treatment, the phosphorylated InsR (Tyr1105) (Fig. 6A and B), IRS-1 (Tyr612) (Fig. 6C and D), and AKT (Ser473) (Fig. 6E and F) (P < 0.05) were dramatically reduced in liver and muscle tissues of both NCD and HFD rats. These data confirm that insulin signaling was attenuated in liver and muscle by the central Ad-shNUCB2 knockdown.

We previously reported that hypothalamic nesfatin-1 is involved in the regulation of HGP through hepatic AKT/AMPK/mTOR signaling (14). mTOR has been shown to activate STAT3 by phosphorylation of Ser727 (10,20–22). To further characterize the signaling mechanism associated with the action of nesfatin-1/NUCB2, we examined the effects of central nesfatin-1/NUCB2 knockdown on STAT3 signaling in the liver. As shown in Fig. 7A and B, total STAT3 protein levels were similar in all groups. However, in both NCD- and HFD-fed rats, ICV nesfatin-1/NUCB2 knockdown induced a dramatic decrease in STAT3 at Tyr705 (Fig. 7A) and Ser727 (Fig. 7B) (P < 0.05). Consistent with our previous reports, 10 weeks of HFD inhibited phosphorylation of mTOR (Ser2448) (Fig. 7C). Furthermore, ICV nesfatin-1/NUCB2 knockdown led to a significant decrease in SOCS3 (Fig. 7D), a transcriptional target of STAT3.

Nesfatin-1 is a hypothalamic signal peptide involved in feeding behaviors and body weight control (3). Although several lines of evidence have suggested that nesfatin-1 may function as a regulator for energy homeostasis (14), the mechanisms associated with the action of nesfatin-1 remain to be characterized. In this study, we have established an in vivo hypothalamic nesfatin-1 knockdown model using RNA interference of nesfatin-1/NUCB2 expression in the hypothalamus of rats. Through this model, we observed that central nesfatin-1/NUCB2 knockdown did not significantly change the body weight of the mice, although food intake was increased ∼10%, with a maximal effect seen on day 3. The insignificant body weight change could be associated with the relatively short treatment period (7 days) (3) or the relative small increase in food intake (∼10%). There are several possibilities for such a small effect on food intake. First, one injection of Ad-shNUCB2 may not be enough for significantly knocking down central nesfatin-1/NUCB2. Because adenovirus-mediated gene expression was considered to be dose and time dependent in a linear fashion (23,24), multiple injections to increase vector dose may be required for a large effect on gene expression and, consequently, on the food intake. Second, there was insufficient blockade (50%) of NUCB2 translation at key hypothalamic nuclei related to food intake. Finally, central nesfatin-1 knockdown may lead to an increase of an anoretic hormone, such as leptin (because nesfatin-1 and leptin expression are not parallel) (25). Of note, a recent study reported that knockdown of hypothalamic NUCB2 had no effects on food intake or body weight in pubertal female rats (19), suggesting a sex difference in nesfatin-1 signaling (i.e., a more prominent role in the control of feeding in male rodents).

ICV nesfatin-1/NUCB2 knockdown decreased GIR and glucose uptake in peripheral tissues and increased HGP during hyperinsulinemic-euglycemic clamp. These changes were accompanied by a significant increase in the hepatic expression of the enzymes G6Pase and PEPCK. In addition, hypothalamic nesfatin-1/NUCB2 knockdown markedly decreased InsR, IRS-1, and AKT phosphorylation in both liver and muscle. Because the insulin signaling pathway is generally believed to proceed through receptor-mediated tyrosine phosphorylation of IRS-1, which leads to activation of phosphatidylinositol 3-kinase (PI3K) and AKT. In blocking gluconeogenesis, insulin reduces transcription of several crucial genes in glucose production, including PEPCK and G6Pase (26). Consistent with our previous report (14), the current results further confirmed that nesfatin-1 plays an important role in hepatic insulin signaling through regulating hepatic glucose fluxes and G6Pase and PEPCK expression.

How does central nesfatin-1/NUCB2 signaling regulate hepatic G6Pase and PEPCK expression? It is well-known that STAT3 signaling is involved in hypothalamic regulation of food intake and hepatic glucose fluxes (27). STAT3 is activated through phosphorylation by a wide range of cytokines and growth factors in response to various stimuli (28–31). We found that central nesfatin-1/NUCB2 knockdown suppressed phosphorylation of STAT3 at Ser727 and Tyr705 in the liver. It is known that Tyr705 phosphorylation is required for activation, whereas phosphorylation at Ser727 is essential for maximal activation of STAT3 (32). After Tyr705 phosphorylation, STAT3 can bind to the upstream regions of the G6Pase and PEPCK promoters to suppress their expression (11). Decreased Tyr705 phosphorylation of STAT3 by knocking down nesfatin-1/NUCB2 signaling may release this suppression, resulting in increased G6Pase and PEPCK expression. Therefore, the findings provide evidence that nesfatin-1 regulated HGP in association with STAT3-induced suppression of gluconeogenetic genes.

SOCS3 is one of the transcriptional targets of STAT3 and plays a role in cytokine-induced insulin resistance by reducing IRS levels (33–35). The results show decreased SOCS3 protein abundance in the liver in hypothalamic nesfatin-1/NUCB2 knockdown rats, further supporting decreased STAT3 activity. Of note, although SOCS3 was downregulated in liver and total IRS levels were slightly increased (Fig. 6B), phosphorylation of both IRS and AKT was decreased, suggesting that decreased phosphorylation of IRS and AKT plays a major role in the increased hepatic insulin resistance induced by hypothalamic nesfatin-1/NUCB2 knockdown.

We have previously reported that hypothalamic nesfatin-1 is involved in the regulation of HGP through hepatic AKT/AMPK/mTOR signaling (14). As a member of the phosphatidylinositol kinase-like Ser/Thr kinase family, mTOR appears to activate STAT3 by phosphorylation of Ser727 (10,20–22). We found that central nesfatin-1/NUCB2 inhibition decreased the activity of mTOR, particularly in HFD-fed rats, leading to decreased phosphorylation of STAT3. Overactivation of the mTOR/S6K1 pathway has been proposed to contribute to defective activation of PI3K by directly phosphorylating IRS on serine residues (36,37). Ser2448 in mTOR has been identified as a phosphorylation site by AKT in vivo, and this phosphorylation depends on PI3K (38). Therefore, it is likely that hypothalamic nesfatin-1/NUCB2 knockdown decreased mTOR Ser2448 phosphorylation through inhibition of insulin signaling. Although it is not clear whether these signal transduction changes are direct effects of hypothalamic nesfatin-1 knockdown or indirect effects associated with increased food intake, our previous study demonstrated that central nesfatin-1 infusion could independently and immediately affect hepatic insulin action, supporting a direct effect (14). However, some limitations of our study need to be considered, such as a lack of pair-feeding controls, a lack of access to insulin levels in the portal vein, and an experimental approach to induce a selective inhibition in these signaling pathways. Therefore, further studies are needed to address these issues.

The neural pathways involved in the effects of nesfatin-1/NUCB2 on glucose homeostasis remain to be defined. A study reported that nesfatin-1 has either hyperpolarizing or depolarizing effects on most neurons in the PVN (39). In addition, glucose and insulin may directly activate nesfatin-1 neurons in PVN (40). Our previous study found that ICV nesfatin-1 activated c-fos-positive neurons in the PVN (14). It is known that the hypothalamus, especially the PVN, through the modulation of the sympathetic-parasympathetic balance, takes part in the control of whole-body glucose metabolism (41–46). Therefore, it is possible that like leptin, nesfatin-1/NUCB2 signaling is relayed to the liver, skeletal muscle, or adipocytes through the autonomous nervous system.

In summary, we demonstrate that an adenoviral-mediated RNA interference can induce a loss of function in hypothalamic nesfatin-1/NUCB2. With this tool, we detected an important role of hypothalamic nesfatin-1 knockdown in the regulation of liver glucose fluxes and insulin signaling, which was associated with inhibition of the mTOR-STAT3 pathway. Further studies are required to validate the cross-talk between the mTOR-STAT3 pathway and nesfatin-1/NUCB2 signaling.

Funding. This work was supported by research grants from the National Natural Science Foundation of China (30871199, 81270913, 81070640, 30971388, and 30771037), Doctoral Fund of Ministry of Education of China (20105503110002, 20125503110003), Natural Science Foundation Key Project of CQCSTC (cstc2012jj B10022), and American Diabetes Association (1-10-CT-06).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. D.W. researched data and wrote and gave final approval to the submission of the manuscript. M.Y., Y.C., and Y.J. researched data and gave final approval to the submission of the manuscript. Z.A.M. and G.B. reviewed, edited, and gave final approval to the submission of the manuscript. L.L. and G.Y. contributed to the discussion and reviewed, edited, and gave final approval to the submission of the manuscript. L.L. and G.Y. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.