Regulation of Glucose Tolerance and Sympathetic Activity by MC4R Signaling in the Lateral Hypothalamus

The incidence and prevalence of obesity and associated disorders, such as diabetes and hypertension, are increasing worldwide (1). Although there is a general consensus that the central nervous system (CNS) plays a key role in these processes (2,3), the underlying neural substrates remain to be fully defined. The central melanocortin signaling pathway, mainly through action on the melanocortin 4 receptors (MC4Rs), is a key relay of arcuate proopiomelanocortin and agouti-related peptide–expressing neurons to coordinate long-term energy balance, glucose homeostasis, and sympathetic nerve activity (SNA) (4). Loss of function of MC4R in both humans and rodents leads to the development of obesity, insulin resistance, and diabetes (5,6). MC4Rs are widely expressed in the CNS, including certain hypothalamic nuclei implicated in the regulation of energy balance, glucose metabolism, and sympathetic outflow, including the paraventricular nucleus, dorsomedial nucleus, and lateral hypothalamic area (LHA) (7). Previous studies identified distinct roles of MC4R signaling in specific neural circuits (8–11), offering the possibility that discrete functions of MC4R signaling in specific CNS sites will allow the development of selective treatments for obesity and associated disorders. The LHA, which receives direct input from leptin-responsive arcuate proopiomelanocortin neurons (12) and expresses MC4Rs (7), regulates multiple physiological processes, including food intake, reward-related behaviors, and autonomic function (13,14). However, the physiological role of MC4R signaling in the LHA is incompletely understood. We have previously shown that MC4R^LHA neurons coexpress the anoretic peptide neurotensin and functionally active form of the leptin receptor and are responsive to leptin administration (15). These observations suggested a potential role for MC4R^LHA neurons in metabolic regulation. Using a Cre-dependent reactivatable MC4R-null mouse line (8), we investigated the physiological function of MC4R^LHA signaling, with a specific focus on energy balance, glucose metabolism, and autonomic activity.

Mice with a transcriptional blocking (TB) cassette flanked by loxP sites placed in front of the MC4R gene (MC4R-TB) and backcrossed at least six generations onto the C57BL/6 line were previously reported (8,16). The tdTomato reporter mice were purchased from The Jackson Laboratory (stock number 007909). Mice were housed in the University of Iowa vivarium in a temperature-controlled environment (lights on 0600–1800 h) with ad libitum access to water and food. To avoid the potentially confounding effects of the female estrus cycle on metabolic profiling, only male mice were used. All animal procedures were performed in accordance with University of Iowa Institutional Animal Care and Use Committee guidelines.

Stereotactic surgery was performed as previously described (15,17). Briefly, 6–7-week-old male mice were anesthetized with ketamine/xylazine (100:10 mg/kg i.p.) and placed on a stereotaxic apparatus (David Kopf Instruments, Tujunga, CA). Following standard disinfection, an ∼1.0-cm incision was made to expose the skull, and a small hole was drilled into the skull bilaterally at defined positions (anteroposterior −1.5 mm; medial lateral +1.5 mm; dorsal ventral −4.9 mm with a 6° angle of injection arm), and a Hamilton microsyringe with small hub removable needle filled with adeno-associated virus (AAV) was slowly inserted. Injection was made by pressing the plunger at a rate of 0.05 μL/min. After 10 min of waiting to ensure a full penetration of AAV into the targeted area, the needle was removed and the incision closed by wound clips. Mice were then kept on a warming pad until awake. At the end of the study, brains were extracted and postfixed in 4% paraformaldehyde, and the injection sites were confirmed for every case by observing the green fluorescence protein (GFP) signal blindly to physiological outcomes. Two of the 59 MC4R-TB mice receiving AAV-Cre-GFP were excluded from the final data analysis due to complete mistargeting of the LHA.

Male mice aged 13–15 weeks were tested for locomotor activity using LABORAS (Metris, Hoofddorp, the Netherlands), an advanced system for automated recognition of small laboratory animal behaviors, such as locomotion, grooming, eating, and drinking. Data were collected uninterrupted in the home cage over a 23-h period and then analyzed with LABORAS software as previously reported (18).

After 1 week of full recovery from AAV microinfusion, mice were assigned ad libitum access to either a normal chow diet (#7913; Harlan Laboratories, Madison, WI) or a high-fat diet (HFD) (42% kcal from anhydrous milk fat, TD.88137; Harlan Laboratories). Body weight and food intake were measured periodically in both HFD and regular chow conditions up to 12 weeks after stereotaxic delivery of AAV into the LHA. For blood biochemistry, a trunk blood sample was collected from 18-week-old HFD-fed male mice. Sera were sent to the Vanderbilt University Mouse Metabolic Phenotyping Center for the measurement of blood hormone and cholesterol levels.

Seventeen- to 19-week-old male mice were used for glucose tolerance tests (GTTs) and pyruvate tolerance tests (PTTs). Mice were handled in the procedure room for 5 days before testing to allow habituation to the environment and to reduce handling stress. On the testing day, the mice were weighed, and food was removed at 9:00 a.m. for 6-h fasting with free access to water. At 3:00 p.m., blood glucose was measured at baseline (0′) followed by immediate intraperitoneal injection of d-glucose (1 g/kg) (G8270; Sigma-Aldrich) or sodium pyruvate (2 g/kg) (P2256; Sigma-Aldrich). Blood glucose level was then monitored at 15, 30, 60, and 120 min for each mouse.

Under continuous anesthesia, a small hole was drilled into the skull at a defined point (anteroposterior −0.2 mm, medial lateral +1.0 mm), and an intracerebroventricular (ICV) cannula with a specific length (0.23 mm below the plastic pedestal; Plastic One Inc.) was slowly inserted. The cannula was held in place with the appropriate amount of a glass ionomer luting cement anchored with a small machine screw in the right-front side of the cannula, allowing 3–5 min for stiffening. Open skin wounds were adhered onto mounted glass ionomer luting cement by applying Vetbond Tissue Adhesive, and the mice were kept on a warmed surgical table until fully awake.

Whole-body fluorodeoxyglucose (FDG) scanning was performed in the Small Animal Imaging Core at the University of Iowa as previously reported (19). Seventeen- to 19-week-old, chow-fed male mice were used. Fasted animals (∼14 h) were anesthetized with 1.5% isoflurane and their blood glucose measured before receiving a tail vein injection of FDG (7.5 ± 11.3 MBq). After a 60-min awake uptake period, the mice were imaged on an Inveon positron emission tomography (PET)/computed tomography (CT) scanner (Siemens, Knoxville, TN). A 15-min PET acquisition followed in the same workflow by a CT scan for attenuation correction was acquired. Three days after the initial scan, the same animals underwent PET imaging again. The same imaging protocol was followed except 30 min before FDG injection, an intraperitoneal injection of melanotan II (MTII 2 mg/kg) (H-3902; Bachem) was given. All images were reconstructed using a three-dimensional OP-MAP algorithm. Image analysis was completed using PMOD version 3.2 (PMOD Technologies, Zurich, Switzerland). Volumes of interest were drawn for interscapular brown adipose tissue (iBAT), muscle, brain, kidney, and heart; organ-specific uptakes of FDG were calculated from the scanned images; and normalized change from baseline FDG uptake was compared between groups.

Regional SNA was measured by direct multifiber recording of sympathetic nerve branches subserving the iBAT and hindlimb muscles as previously described (20). Seventeen- to 19-week-old chow-fed male mice were used for SNA recording. Under anesthesia, catheterization of the carotid arteries and jugular vein was performed for hemodynamic recording and maintenance of anesthesia with α-chloralose (25 mg/kg/h), respectively. A nerve fascicle to iBAT or hindlimb was carefully isolated under a dissecting microscope. A bipolar platinum-iridium electrode (Cooner Wire) was suspended under the nerve and secured with silicone gel (Kwik-Cast, World Precision Instruments). The electrode was attached to a high-impedance probe (HIP-511; Grass Technologies), and the nerve signal was amplified 105 times with a Grass P5 AC preamplifier. After amplification, the nerve signal was filtered at a 100- and 1,000-Hz cutoff with a nerve traffic analysis system (model 706C; Department of Bioengineering, University of Iowa). Subsequently, the amplified and filtered nerve signal was routed to an oscilloscope (model 54501A; Hewlett-Packard) for monitoring the quality of the sympathetic nerve recording and to a resetting voltage integrator (model B600c; University of Iowa Bioengineering) that summed the total voltage output to a unit of 1 V × s/min before resetting to 0.

After a 10-min recording of baseline activity, mice received ICV infusion of MTII (2 μg). SNA was continuously monitored for 4 h, and change from baseline activity was calculated and compared between groups. At the end of study, mice received ICV infusion of 2 μL blue dye, and brains were extracted to confirm correct ICV cannulation and AAV injection site.

Bilateral iBAT sympathetic denervation was performed as previously reported (21), with slight modification. Thirteen- to 15-week-old chow-fed male mice were used for this study. On the day of surgery, each mouse was weighed and anesthetized with isoflurane. When the proper state of anesthesia was reached, the mouse was shaved of hair on the nape region and secured on a warm surgical table with a rectal probe to measure and maintain body temperature at 37.5°C. After a standard skin disinfection procedure, a lateral incision was made just below the two shoulder blades to expose the intrascapular fat pads. A blunt forceps was then used to clamp the caudal edge of the intrascapular fat pad and retracted cranially over the head of the mouse. With this exposure, the bilateral sympathetic chain that innervates both the right- and the left-side BAT hemispheres was identified. All five branches of intercostal sympathetic nerves subserving the left- and right-side BAT fat pad were identified and carefully isolated, and the nerve branches between the root and the BAT were sectioned out (Fig. 4A–C). After completion of bilateral BAT denervation, the intrascapular fat pad was then returned to its original position and secured in place to the surrounding thoracic tissue and muscle with absorbable 6.0 VICRIL suture. A thin line of tissue adhesive (Vetbond) was applied along the edges of both skin flaps, and the two edges of skin were held in place until the entire length of the incision in the nape area was completely closed. Mice were monitored for 7 days postoperatively before subjected to further experimental procedures.

Total RNA was isolated from snap-frozen iBAT and inguinal white adipose tissue (WAT) of 16–18-week-old chow-fed male mice, and cDNA was generated by reverse transcription as previously described (22). Quantitative real-time PCR was performed with SYBR green method using a standard protocol as previously described (22). The following primer sets were used to determine relative abundance and expression of UCP1, Cidea, PGC1a, β3-AR, Glut1, and Glut4: 5′-AAGCTGTGCGATGTCCATGT-3′ and 5′-AAGCCACAAACCCTTTGAAAA-3′ for UCP1, 5′-AGACAAATGTGCTTCCAAAAAGAA3′ and 5′-GAAGAGATAAAGTTGTTGGTTTGGC-3′ for PGC1a, 5′-GGTTCAAGGCCGTGTTAAGG-3′ and 5′-CGTCATCTGTGCAGCATAGG-3′ for Cidea, 5′-TCCTTCTACCTTCCCCTCCTT-3′ and 5′-CGGCTTAGCCACAACGAACAC-3′ for β3-AR, 5′-CGCCCCCCAGAAGGTTAT-3′ and 5′-CGATGCGGTGGTTCCAT-3′ for Glut1, and 5′-TTGGCTCCCTTCAGTTTGG-3′ and 5′-CTACCCAGCCACGTTGCAT-3′ for Glut4. The expression of 36B4 (5′-CACTGGTCTAGGACCCGAGAAG-3′ and 5′-GGTGCCTCTGAAGATTTTCG-3′) was used as an internal control gene.

Nineteen- to 21-week-old overnight-fasted male mice were anesthetized with ketamine/xylazine (100:10 mg/kg) and received retro-orbital bolus insulin (5 units). iBAT, soleus muscle, and liver were then harvested at 10 min and snap frozen in liquid nitrogen. Small pieces of frozen tissue were homogenized in cold radioimmunoprecipitation assay buffer (#89901; Thermo Scientific) containing cocktails of proteinase inhibitors (11836170001; Roche Diagnostics) and phosphatase inhibitors (04906845001; Roche Diagnostics) and kept on ice for 30 min. Samples were then centrifuged (13,000 rpm) at 4°C for 20 min, and the supernatant was taken as total protein extraction. Protein concentration was determined by BCA protein assay method (#23228; Thermo Scientific), and 20 μg of protein were separated by SDS-PAGE gel, transferred onto a polyvinylidene fluoride membrane, and immunoblotted with phosphorylation of v-akt murine thymoma viral oncogene homolog 1 (AKT) at Ser⁴⁷³ site (pAKT) (1/2,000) (#4060; Cell Signaling Technology), AKT (1/1,000) (#9272; Cell Signaling Technology), phosphorylation of extracellular-signal-regulated kinases (ERK) at Thr²⁰² and Tyr²⁰⁴ sites (pERK) (1/1,000) (#9101S; Cell Signaling Technology), and ERK (1/1,000) (#9102; Cell Signaling Technology) as previously reported (23). Membranes were then incubated with horseradish peroxidase–conjugated secondary antibody (Jackson ImmunoResearch), and signals were detected by chemiluminescence. Signal intensity was measured and analyzed by a BioSpectrum 810 Imaging System (UVP, Upland, CA). For each experiment, signal intensity of targeted proteins was normalized to total AKT or ERK signal intensity and then compared between groups.

GraphPad Prism 5 software (GraphPad Software Inc., San Diego, CA) was used to perform all statistical analyses. Comparisons between groups were made by two-way or one-way ANOVA with Bonferroni post hoc analysis or by Student t test as indicated. P < 0.05 was considered statistically significant. Data are presented as mean ± SEM.

To directly study the physiological function of MC4R^LHA signaling, we used AAV to specifically deliver Cre-recombinase-GFP (AAV-Cre-GFP) or GFP-alone (AAV-GFP) into the LHA of reactivatable MC4R-TB mice to remove a TB cassette and restore endogenous expression of MC4Rs. To optimize the delivery volume of AAV-Cre-GFP specifically into the LHA, we first performed microinfusion of AAV-Cre-GFP into the LHA of tdTomato reporter mouse in which a loxP-flanked TB cassette prevents transcription of a ubiquitous promoter-driven tdTomato, which can be re-expressed upon removal of the TB cassette by Cre-mediated recombination. Two weeks after AAV-Cre-GFP microinfusion, a large number of tdTomato⁺ and GFP⁺ cells were observed in the LHA, and nearly all of GFP⁺ cells were tdTomato⁺ (Fig. 1C), indicating that AAV-Cre-GFP properly induced Cre/loxP recombination in the LHA neurons with limited infection to the surrounding structures (Fig. 1B and C). The volume and titer of AAV-Cre-GFP were optimized by repeated injections in tdTomato reporter mice (data not shown). Using 0.3 μL/side at a titer of 1.18 × 10¹² genome copies/mL, we were able to achieve a maximum coverage of dorsomedial LHA, where MC4R-positive neurons are mainly located (Fig. 1A), with minimized infection of adjacent hypothalamic nuclei. After validation and optimization of the AAV microinfusion protocol, AAV-Cre-GFP or AAV-GFP was bilaterally infused into the LHA of MC4R-TB mice and their wild-type (WT) littermates. In contrast to our hypothesis, restoration of MC4R^LHA expression had no measurable effects on body weight (Fig. 1D), food intake (Fig. 1E), locomotion (Fig. 1F), fasting glucose (Fig. 1G) and insulin (Fig. 1H) levels, body composition (Fig. 1L), or regional fat mass (Fig. 1M–O) when fed a chow diet. Subsequent assessment for glucose homeostasis, however, revealed that although both MC4R-TB groups were similarly obese (Fig. 1D), MC4R^LHA mice displayed significantly improved glucose tolerance compared with MC4R-TB (AAV-GFP) mice, especially at the 15′ and 30′ time points (Fig. 1I and J). To further test whether improved glucose tolerance is associated with altered insulin levels, we measured insulin levels at the 0′, 15′, and 30′ time points after intraperitoneal injection of glucose (1 g/kg) and did not observe significant difference between the two MC4R-TB groups (Fig. 1K). Because none of these parameters was significantly different between the WT AAV-GFP (n = 5) and WT AAV-Cre-GFP (n = 5) groups, both groups of WT mice were combined and are referred to as control to increase statistical power.

Because MC4R-null mice are hypersensitive to a palatable HFD with marked hyperphagia and weight gain compared with the chow-fed condition (24,25), we generated another cohort of MC4R-TB mice by MC4R-TB homo × MC4R-TB het breeding paradigm and divided it into two groups to receive either AAV-GFP or AAV-Cre-GFP microinfusion. After 1 week of recovery from surgery, mice were fed an HFD to assess for a diet-specific effect of MC4R^LHA signaling. Restoration of MC4R^LHA signaling did not affect weight gain (Fig. 2A) or food intake (Fig. 2B) in the HFD condition. Although both groups of mice were similarly obese, fasting glucose level was significantly lower in MC4R^LHA mice (Fig. 2D) without a significant change in circulating insulin levels (Fig. 2C). A subsequent GTT again confirmed a significant improvement in glucose tolerance in MC4R^LHA mice (Fig. 1E and F). No differences were seen in fasting serum glucagon, leptin, free fatty acid, triglyceride, total cholesterol, or direct HDL levels (Fig. 2G–L). Collectively, these data demonstrate a role for MC4R^LHA signaling in the regulation of glucose tolerance without altering body energy balance or the levels of critical glucoregulatory hormones, such as insulin and glucagon.

Improved glucose tolerance in MC4R^LHA mice without a significant change in body weight or circulating insulin or glucagon levels led us to ask whether glucose disposal was increased in any tissue of MC4R^LHA mice. To this end, we performed whole-body PET/CT scanning for FDG uptake by which biodistribution of FDG can be quantified in a spatially resolved manner. Although both groups of MC4R-TB mice were similarly obese (Fig. 3B), FDG uptake was significantly increased in iBAT of MC4R^LHA mice (Fig. 3A and C) after MTII administration compared with control MC4R-TB (AAV-GFP) mice. A similar trend was noted in skeletal muscle but did not reach statistical significance (Fig. 3D). No difference in FDG uptake was found in brain (Fig. 3E), kidney, or heart (data not shown).

To further test whether MC4R^LHA signaling affects hepatic glucose production, a PTT was performed as an indirect measurement of hepatic gluconeogenesis. Intraperitoneal injection of sodium pyruvate (2 g/kg) in 6-h fasted animals sharply increased blood glucose levels within 15 min, which gradually returned to baseline levels by 2 h in control mice (Fig. 4B and C). In contrast, glucose levels continuously increased in both groups of MC4R-TB mice, and no significant difference was noted between the groups during the 2-h test (Fig. 4B and C), suggesting that restoring MC4R^LHA signaling does not likely affect hepatic gluconeogenesis. Again, no difference of body weight was noted between the groups of MC4R-TB mice when performing GTT (Fig. 4A).

Additionally, to test the possibility that restoration of MC4R^LHA may affect insulin signaling, insulin-induced pAKT and ERK were also measured as markers of acute activation of the insulin receptor signaling cascade. After confirming a significant induction of pAKT and pERK by retro-orbital insulin bolus (5 units) in iBAT of WT mice (Fig. 4D and H), the same procedure was used to test insulin signaling in iBAT, soleus muscle, and liver. No significant difference was seen between the two MC4R-TB groups in any of the tissues tested (Fig. 4E–G and I–K), although there was a trend toward increased pERK/ERK levels in soleus muscle that did not reach statistical difference (Fig. 4J).

The observation that MC4R^LHA mice display improved glucose tolerance and increased FDG uptake in iBAT without affecting circulating insulin or peripheral insulin sensitivity suggests the possibility that MC4R^LHA signaling may facilitate glucose uptake in iBAT by increasing SNA. Given the known role of the LHA in the regulation of sympathetic outflow and direct multisynaptic innervation of LHA MC4R-positive neurons to iBAT (26,27), we tested the possibility that MC4R^LHA signaling mediates activation of SNA to iBAT induced by the melanocortin agonist MTII. Although direct multifiber recording demonstrated that control mice showed significantly increased iBAT SNA following infusion of MTII (Fig. 5A and B), the effect was completely abolished in MC4R-TB (AAV-GFP) mice (Fig. 5A and B). In contrast, MTII-induced iBAT SNA in MC4R^LHA mice was restored to control levels (Fig. 5A and B), indicating that reactivation of MC4R^LHA signaling is sufficient to rescue blunted sympathetic traffic to iBAT in obese MC4R-null mice. Because we noted a trend toward increased FDG uptake in skeletal muscle, we also recorded lumbar nerve SNA to hindlimb skeletal muscle. As expected, the control group demonstrated a robust increase in lumbar SNA following ICV injection of MTII (Fig. 5C). MC4R-TB mice receiving AAV-GFP in the LHA displayed a significantly delayed and blunted response to MTII-induced lumbar SNA compared with WT mice, with a modest increase in lumbar SNA observed after 180 min of recording. This blunted response of MTII-induced lumbar SNA is restored to WT levels in MC4R^LHA mice (Fig. 5C). These results indicate that MC4R^LHA signaling mediates MTII-induced sympathetic activation to iBAT consistent with the increased glucose uptake observed in this tissue in MC4R^LHA mice. To further test whether altered iBAT SNA affects the body’s thermoregulation, we measured core body temperature in a separate cohort of conscious mice by quick-inserting oiled rectal probe (Kent Scientific, Torrington, CT). In contrast to an earlier report (28), we found that obese MC4R-TB (AAV-GFP) mice have a slightly reduced core body temperature (34.81 ± 0.21°C) compared with WT control mice (35.59 ± 0.22°C), and a similar reduction was observed in MC4R-TB (AAV-Cre-GFP) mice (34.80 ± 0.23°C), suggesting that LHA MC4R signaling has no significant effect on core body temperature.

Activation of mitochondrial energy dissipation pathways in iBAT has previously been linked with improved glucose tolerance (29). SNA is known to regulate the expression of genes involved in mitochondrial energy dissipation, including UCP1, Cidea, PGC1a, β3-AR, Glut1, and Glut4 (30). Therefore, we measured mRNA levels of these genes in iBAT to correlate changes of gene expression with SNA and FDG uptake. Gene expression profiling in iBAT revealed that significantly decreased expression of UCP1 and Cidea (Fig. 6A and B) but not PGC1a and β3-AR (Fig. 6C and D) in MC4R-TB mice was normalized in MC4R^LHA mice. Of note, although Glut1 expression was similarly increased in both MC4R-TB groups (Fig. 6E), a twofold increase in Glut4 expression was only observed in MC4R^LHA mice (Fig. 6F). None of these transcriptional changes were observed in inguinal WAT (Fig. 7A–F), indicating an iBAT-specific effect of SNA by MC4R^LHA signaling. Significantly increased expression of Glut4 in iBAT may at least partly explain the facilitated glucose uptake in iBAT and the ameliorated hyperglycemia observed after restoration of MC4R^LHA signaling.

To directly demonstrate that SNA to iBAT is required for the improvement in glucose tolerance, another cohort of MC4R^LHA mice was generated and confirmed to have improved glucose tolerance without an effect on body weight (data not shown). One week after this GTT, mice underwent bilateral iBAT denervation by removal of five intercostal sympathetic nerve branches subserving iBAT (Fig. 8A–C) as previously described (21), which resulted in a significantly decreased UCP1 mRNA level compared with a sham control group in WT male mice, confirming effective iBAT denervation (Fig. 8D). One week after recovery from surgery, all three groups of mice were subjected to GTT again. Although there was no significant change in body weight during the 1-week recovery period (Fig. 8E), the improved glucose tolerance seen in MC4R^LHA mice was lost following iBAT sympathetic denervation (Fig. 8F and G), confirming that iBAT SNA driven by MC4R^LHA signaling is required to improve glucose intolerance in obese MC4R-TB mice.

In the present study, we report a previously unappreciated role of MC4R^LHA signaling in the regulation of glucose metabolism through the regulation of sympathetic outflow to iBAT. This effect may be mediated in part by upregulated expression of Glut4 in iBAT. Of note, although genes that are known to promote energy dissipation during excess feeding, such as UCP1 and Cidea, were upregulated in iBAT of MC4R^LHA mice, no body weight effect was observed in either chow- or HFD-fed mice. This result is consistent, however, with a previous report that UCP1-deficient mice are cold sensitive but do not develop obesity under normal housing conditions (31). The present findings suggest that the energy balance and glucoregulatory roles of iBAT can be dissociated. Consistent with this possibility, a thyroid hormone β-selective agonist and a β3-AR agonist have been shown to increase SNA and thermogenesis in iBAT without increasing glucose uptake (32). These observations led us to speculate that MC4R^LHA signaling-mediated sympathetic outflow may differentially innervate a portion of iBAT that affects glucose utilization without altering energy expenditure or blunting the severe obesity seen in MC4R-null mice.

The primary effect we detected for MC4R^LHA signaling was on iBAT, and we also observed a potential role for skeletal muscle. A trend toward increased FDG uptake was noted in MC4R^LHA mice, and restoration of MC4R^LHA signaling showed a delayed response to MTII in lumbar SNA. Although these observed effects of MC4R^LHA signaling on muscle was less profound compared with iBAT, a potentially important contribution of muscle for the glucoregulatory effect of MC4R^LHA signaling should be considered because there is a significantly larger volume of muscle compared with iBAT in humans. Another possibility is that our ability to detect significant uptake of FDG by PET/CT imaging in muscle was limited by the challenging spatial resolution of a diffuse organ such as muscle. The mechanism of MTII-mediated increase in lumbar SNA is unclear. One possibility is that because MTII can also act on the melanocortin 3 receptor, which may account for the residual lumbar SNA response to MTII. Further work is needed to test this possibility.

Although we show that restoration of MC4R^LHA signaling does not affect circulating insulin levels as well as bolus insulin–activated signaling cascades in iBAT and liver, the role of insulin in the glucoregulatory effect of MC4R^LHA signaling should be considered because Glut4-mediated transport of glucose is insulin dependent (33) and there was a trend toward increased pERK activation in muscle. Additionally, it is important to note that most of these measures in the current study were performed in chow-fed mice, and insulin resistance in muscle of MC4R-null mice is markedly worse under HFD feeding conditions (34,35). A protective effect of MC4R^LHA activity on insulin signaling in muscle may only emerge under the severe insulin resistance that occurs during HFD feeding. Nevertheless, future work with hyperinsulinemic-euglycemic clamp in HFD fed mice may be needed to clarify the role of insulin in the glucoregulatory effect of MC4R^LHA signaling. Somewhat surprising to see was that restoration of MC4R^LHA signaling reduces glucose excursion during GTT and fasting glucose levels in HFD-fed mice but without a concomitant reduction in insulin levels. In fact, there was a trend toward higher insulin levels during GTT in the MC4R^LHA group when the lower glycemic levels would predict lower insulin levels. This finding may indicate a degree of protection against obesity-related β-cell dysfunction conferred by MC4R^LHA activity. Whether such protection might be mediated by the demonstrated improvement of glucose disposal in other tissues, by modulating factors such as GLP-1 or circulating nutrients (36), or by more direct action of MC4R^LHA activity on islets is unclear.

Although the current study uses a combination of Cre-lox technology and advanced physiological measures to assess the function of MC4R^LHA signaling, several limitations should be noted. We used the reactivation model in which MC4R expression is restored in the LHA of adult mice. This method is well suited to determine biological sufficiency but cannot determine whether redundancy in MC4R function exists. Indeed, a role for MC4R signaling in the regulation of autonomic activity has been reported for several other brain sites (37,38). Additionally, we attempted to control for off-target effects of Cre-recombinase expression by infusing both AAV-GFP and AAV-Cre-GFP WT groups. However, we cannot exclude the possibility that Cre expression affects FDG uptake because we did not include WT mice as controls due to cost. Many of the methods and assessments used here are technically challenging, labor intensive, and expensive, which limits the statistical power of certain assessments. Although the primary outcome measures are of large effect size, we cannot exclude the possibility that we were unable to discern subtle differences in secondary outcome measures. Finally, the relevance to human studies should be noted. Many studies have proposed increasing activity of BAT to treat metabolic disorders, such as obesity and diabetes (39). The current study is consistent with this approach, although it suggests that the effects of BAT activity on body weight and glucose tolerance may be dissociable, depending on the method used. However, any approach targeting SNA or BAT activity in humans may have limited therapeutic potential due to side effects. Even so, understanding the neural circuitry that improves glucose tolerance will be important to designing optimal treatments.

In summary, the results expand the understanding of how MC4R signaling differentially regulates a diverse set of physiological functions. Isolating the function of this neural circuit within the LHA also improves our understanding of how specific hypothalamic circuits control glucose tolerance. The results represent a novel mechanism by which brain MC4R signaling regulates glucose homeostasis independent of energy balance and may help in the development of novel strategies for the treatment of diabetes.

Acknowledgments. We thank Joel Elmquist (University of Texas Southwestern Medical Center, Dallas, TX) and Bradford Lowell (Beth Israel Deaconess Medical Center, Boston, MA) for use of the MC4R-TB mouse line.

Funding. This work was funded by the following grants: National Institutes of Healthhttp://dx.doi.org/10.13039/100000002 (R01-DK-097820 to A.W.N., MH-084058-01A1 to M.L., HL-084207 to K.R.), the University of Iowa Fraternal Order of Eagles Diabetes Center Pilot Grant to M.L., the American Heart Associationhttp://dx.doi.org/10.13039/100000968 (Established Investigator Award 14EIA18860041 to K.R., postdoctoral fellowship 12POST9120037 and Scientist Developmental grant 14SDG20140054 to H.C.), and the Brain & Behavior Research Foundationhttp://dx.doi.org/10.13039/100000874 (formerly NARSAD) Young Investigator Award to H.C.

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

Author Contributions. D.A.M. contributed to performing most critical experiments and to the data analysis and review and editing of the manuscript. L.N.M., T.Y., M.K., and J.J. contributed to the stereotaxic surgery, mouse breeding and genotyping, and Western blotting. M.R.A. and S.A.W. contributed to the PET/CT scanning for FDG uptake. L.L.B.P. contributed to the FDG uptake data analysis. A.W.N. contributed to the review and editing of the manuscript. M.L. and K.R. contributed to the experimental design and review and editing of the manuscript. H.C. contributed to the experimental design and to performing most critical experiments. H.C. 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.