Research ArticleMetabolism Open Access | 10.1172/JCI178250

Naima S. Dahir,1,2 Yijun Gui,1,3 Yanan Wu,1,3 Patrick R. Sweeney,4 Alix A.J. Rouault,1,2 Savannah Y. Williams,1 Luis E. Gimenez,1 Tomi K. Sawyer,5 Stephen T. Joy,1 Anna K. Mapp,1,6 and Roger D. Cone1,2,3

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Dahir, N. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Gui, Y. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Wu, Y. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Sweeney, P. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Rouault, A. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Williams, S. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Gimenez, L. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Sawyer, T. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Joy, S. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Mapp, A. in: JCI | PubMed | Google Scholar

1Life Sciences Institute,

2Department of Molecular and Integrative Physiology, and

3Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan, USA.

4Department of Molecular and Integrative Physiology, University of Illinois, Urbana-Champaign, Illinois, USA.

5Courage Therapeutics, Newton, Massachusetts, USA.

6Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan, Ann Arbor, Michigan, USA.

Address correspondence to: Roger D. Cone, Life Sciences Institute, University of Michigan, 210 Washtenaw Avenue, Ann Arbor, Michigan 48109-2216, USA. Phone: 503.804.2370; Email: rcone@umich.edu.

Find articles by Cone, R. in: JCI | PubMed | Google Scholar

Published July 15, 2024 - More info

Published in Volume 134, Issue 14 on July 15, 2024
J Clin Invest. 2024;134(14):e178250. https://doi.org/10.1172/JCI178250.
© 2024 Dahir et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Published July 15, 2024 - Version history
Received: December 26, 2023; Accepted: May 15, 2024 View PDF Abstract

The melanocortin-3 receptor (MC3R) regulates GABA release from agouti-related protein (AgRP) nerve terminals and thus tonically suppresses multiple circuits involved in feeding behavior and energy homeostasis. Here, we examined the role of the MC3R and the melanocortin system in regulating the response to various anorexigenic agents. The genetic deletion or pharmacological inhibition of the MC3R, or subthreshold doses of an MC4R agonist, improved the dose responsiveness to glucagon-like peptide 1 (GLP1) agonists, as assayed by inhibition of food intake and weight loss. An enhanced anorectic response to the acute satiety factors peptide YY (PYY3-36) and cholecystokinin (CCK) and the long-term adipostatic factor leptin demonstrated that increased sensitivity to anorectic agents was a generalized result of MC3R antagonism. We observed enhanced neuronal activation in multiple hypothalamic nuclei using Fos IHC following low-dose liraglutide in MC3R-KO mice (Mc3r–/–), supporting the hypothesis that the MC3R is a negative regulator of circuits that control multiple aspects of feeding behavior. The enhanced anorectic response in Mc3r–/– mice after administration of GLP1 analogs was also independent of the incretin effects and malaise induced by GLP1 receptor (GLP1R) analogs, suggesting that MC3R antagonists or MC4R agonists may have value in enhancing the dose-response range of obesity therapeutics.

Introduction

The central melanocortin system is fundamental in regulating feeding behavior and energy homeostasis. Mutations in the melanocortin-4 receptor (MC4R) or genetic deletion of Mc4r results in hyperphagia, hyperinsulinemia, and obesity (1–3). In contrast, Mc3r deletion does not produce hyperphagia or hyperinsulinemia in ad libitum–fed animals (4, 5) However, the MC3R acts as a negative regulator of the MC4R system, controlling responses of the melanocortin system to both internal and external challenges (6–10), controlling the upper and lower boundaries of energy homeostasis (9), and providing sexually dimorphic inputs to the system (8, 10, 11). For example, MC3R is essential in promoting normal compensatory refeeding and neuroendocrine responses to fasting (7, 8, 10, 12). The MC3R accomplishes this by regulating GABA release onto MC4R neurons from presynaptic MC3R expressed on agouti-related protein (AgRP) neurons (9, 12) and controlling the activation of AgRP neurons (13). Furthermore, the MC3R is requisite for normal growth (4, 5) and reproductive function, with mutations leading to reduced linear growth, lean mass, and late onset of puberty in mice and humans (2).

We have previously examined the effect of the MC3R on MC4R-expressing paraventricular hypothalamic (PVH) neurons and demonstrated that MC3R blockade enhances the magnitude and duration of the inhibition of food intake in response to a single dose of the MC4R agonist LY2112688 (9). The MC3R is expressed in nearly all AgRP neurons, and the AgRP neurons are known to broadly affect multiple aspects of feeding behavior and suppress competing behavioral states, including fear, alarm, and thirst. For example, we have demonstrated that deletion of the MC3R increases sensitivity to anorexia induced by restraint and social isolation (10). Thus, MC3R activity may be found to broadly regulate responsiveness to anorectic stimuli. Here, we examined the ability of the MC3R, and the downstream effects of the MC4R, to regulate responsiveness to a variety of anorexigenic agents, from acute satiety factors like CCK and PYY3–36 to the long-term adipostatic factor leptin, as well as to leading glucagon-like peptide-1–based (GLP1-based) therapeutic agents like liraglutide and tirzepatide.

Results

Disruption of MC3R signaling increases sensitivity to GLP1R agonists. Previously, we reported that deletion of the MC3R (Mc3r–/–) in male and female mice results in enhanced acute inhibition of food intake in response to a single dose of the glucagon-like peptide-1 receptor (GLP1R) agonist liraglutide (10). We tested the effect of various concentrations of liraglutide in mice lacking Mc3r and their WT counterparts. In this testing paradigm, we administered concentrations of 0.05–0.4 mg/kg liraglutide at least 30 minutes before the onset of the dark cycle. We manually measured a single time point (24-hour) for food intake (spillage included) and body weight change in male and female mice. As expected, liraglutide reduced food intake and body weight in WT male (Figure 1, A and B) and female (Figure 1, C and D) mice in a concentration-dependent manner. In the Mc3r–/– mice, we found an increased dose sensitivity to liraglutide-induced suppression of food intake and weight loss 24 hours after administration. Unlike WT mice, Mc3r–/– mice exhibited reduced food intake and weight loss in response to low concentrations of liraglutide (0.05 mg/kg), indicating hypersensitivity to the compound. Next, we assessed the sensitivity of Mc3r–/– mice to other GLP1R agonists. We tested the effect of semaglutide, a secondary GLP1 drug shown to have a longer halflife than liraglutide, thereby increasing both the incretin and anorectic duration of drug action (14). We tested male mice and found that, like liraglutide, deletion of MC3R led to increased responsiveness to semaglutide; Mc3r–/– mice responded more to both the anorectic effects and weight loss following semaglutide treatment compared with the WT mice, illustrating that this hypersensitivity was in response to multiple GLP1R agonists (Supplemental Figure 1, A and B; supplemental material available online with this article; https://doi.org/10.1172/JCI178250DS1).

Figure 1

MC3R loss increases responsiveness to GLP1 drugs. Liraglutide (0.05–0.4 mg/kg) administration resulted in more significant inhibition of (A) food intake and (B) weight loss in Mc3r–/– male mice compared with Mc3r+/+ mice in a dose-dependent manner after 24 hours (n = 7–8/group). (C) Liraglutide-induced feeding and (D) body weight changes of Mc3r+/+ and Mc3r–/– female mice (n = 7–8/group; 0.05–0.4 mg/kg). Tirzepatide (TZ) (1–4 nmol/kg) and coadministration of tirzepatide and C11-induced (E) feeding responses and (F) body weight changes of WT male mice (vehicle, n = 10; all other groups, n = 8) at 24 hours after injection. (G) Twenty-four-hour feeding and (H) body weight changes after chronic injections of tirzepatide (2 nmol/kg), C11 (0.5 nmol), tirzepatide plus C11, and vehicle. Data are expressed as the mean ± SEM, and statistical tests were performed by 2-way ANOVA with Tukey’s test for post hoc analysis (G–H). For all the dose-response curve data, a repeated-measures 2-way ANOVA was corrected for multiple comparisons using the Tukey-Kramer method for each time point, and data were fitted with 4 parameters: nonlinear fit, **P < 0.01.

We have previously shown that the MC3R peptide antagonist compound 11 (C11) (15) can reliably inhibit the activity of the MC3R and that coadministration of C11 and liraglutide can further lower the body weight of WT mice (10). We repeated the same experiment and additionally used Mc3r–/– mice to illustrate the specificity of C11. We found that pharmacological inhibition of the MC3R (C11; 1 nmol/1 μL, i.c.v.) and liraglutide, together, resulted in profound weight loss and decreased food intake over 8-hour, 12-hour, and 24-hour periods (Supplemental Figure 2). Furthermore, no effect was seen in mice with MC3R deletion, validating the pharmacological specificity of C11.

Next, we examined the weight loss and anorectic effects of MC3R pharmacological inhibition in response to increasing doses of tirzepatide. This diabetes drug acts as a dual agonist for GLP1R and the glucose-dependent insulinotropic polypeptide (GIP) in humans (16, 17) but acts as a potent GLP1R agonist in rodents (18). Administration of tirzepatide (1–5 nmol/kg, s.c.) to WT mice decreased food intake and promoted weight loss in male mice (Figure 1, E and F), measured 24 hours after administration. Furthermore, coadministration of C11 (1 nmol/1 μL, i.c.v.) and varying dosages of tirzepatide profoundly decreased 24-hour food intake and 24-hour body weight after a single injection of both compounds (Figure 1, E and F). To determine whether chronic administration of tirzepatide and C11 further promotes the hypersensitivity effect, 24-hour food intake and the body weight of male mice administered C11 (0.5 nmol, i.c.v.), tirzepatide ( 2 nmol/kg, s.c.), C11 and tirzepatide, or vehicle (DMSO, i.c.v. and saline, s.c.) were measured. The results for 4 days of vehicle treatment, followed by treatment for 3 days, and the recorded recovery for an additional 3 days are shown in Figure 1, G and H. Administration of C11 and tirzepatide alone showed comparable effects on weight loss and decreased food intake compared with mice in the vehicle-treated groups. However, administration of both compounds led to a greater inhibition of food intake and weight loss than did either compound alone, demonstrating a higher efficacy of tirzepatide with MC3R antagonism in a chronic treatment model.

MC3R deletion does not increase the incretin effects or malaise induced by liraglutide. GLP1 analogs have been historically associated with their incretin effects, stimulating insulin secretion and suppressing glucagon secretion in hyperglycemic or euglycemic states (19). GLP1R analogs approved by the FDA are reliably used to treat type 2 diabetes and regulate glucose homeostasis (20). We tested whether Mc3r–/– mice also exhibited increased sensitivity to the incretin effects of GLP1 analogs by measuring blood glucose in a glucose tolerance test (GTT) (Figure 2A). WT and Mc3r–/– mice were fasted and injected with varying doses of liraglutide (0.01–0.2 mg/kg, s.c.) or vehicle (PBS, s.c.). After 6 hours of fasting, the glucose responses were measured following a 1 g/kg oral glucose bolus. However, we found no effect of genotype; both Mc3r+/+ and Mc3r–/– mice responded similarly to the oral GTT (Figure 2, B–F). Furthermore, we subjected mice to varying glucose boluses (0.5–2 g/kg, oral) in response to liraglutide (0.2 mg/kg, s.c.) and found no effect of genotype; both Mc3r+/+ and Mc3r–/– mice responded similarly to the oral GTT (Supplemental Figure 3, A–C). Glucose disposal with liraglutide was comparably better than with vehicle in both genotype groups (Supplemental Figure 3, A–C).

Figure 2

Mc3r deletion has no effect on the incretin activity or malaise associated with liraglutide. (A–F) Glucose levels and AUC before and after oral administration of glucose (1 g/kg) after liraglutide (lira) (0.01–0.2 mg/kg) or vehicle (veh) treatment in Mc3r+/+ and Mc3r–/– male mice (n = 6–7/group). CTA test day 1 (G) and test day 2 (H) after liraglutide administration in Mc3r+/+ and Mc3r–/– male mice. (I) Representative images from the AP showing Fos IHC after saline or liraglutide injection from Mc3r+/+ and Mc3r–/– male mice. CC, central canal. Scale bars: 100 μm. (J) Quantification of cells expressing Fos after saline or liraglutide injection in Mc3r+/+ and Mc3r–/– male mice. (K) Representative images from the hypothalamus showing Fos IHC after liraglutide treatment in Mc3r+/+ and Mc3r–/– mice. Scale bars: 100 μm. (L) Quantification of cells expressing Fos in the ARH, VMH, and DMH after liraglutide treatment. Statistical analysis was done by 2-tailed Student’s t test, *P < 0.05, and ****P < 0.0001 (J, G, H, and L).

The most reported adverse effects of GLP1R agonists are gastrointestinal; patients report increased emetic responses (21–24), as well as exacerbation of gastroparesis (25). To test whether the observed hypersensitive anorectic responses to GLP1 analogs were accompanied by increased malaise, we tested whether liraglutide triggered increased conditioned taste aversion in Mc3r–/– mice, as previously described (23, 26). Low concentrations of liraglutide that resulted in decreased food intake and promoted weight loss (0.05 mg/kg) in Mc3r–/– mice were paired with the non-nutritive sweetener saccharin (0.1%). A second group of mice received saccharin paired with LiCl (150 mM) to induce gastric malaise. A third group of mice received saccharin paired with saline (150 mM NaCl). Mice underwent training for 2 days, conditioning for 2 days, and testing for 2 days. Mc3r+/+ and Mc3r–/– mice showed low avoidance for saccharin when paired with saline and high avoidance when paired with either the liraglutide or LiCl. We found that genetic deletion of Mc3r did not increase the conditioned taste aversion associated with administering liraglutide (Figure 2, G and H). GLP1 agonists are also known to activate neurons in the area postrema (AP) at doses that induce conditioned taste aversion (CTA) (27). Deletion of the MC3R was not observed to increase activation of AP neurons in response to 0.1 mg/kg liraglutide (Figure 2, I and J). In contrast, a profound increase in neuronal activation following liraglutide treatment was seen in hypothalamic feeding circuits in the ventromedial hypothalamus (VMH), dorsomedial hypothalamus (DMH), and arcuate nucleus of the hypothalamus (ARH) of the Mc3r–/– mice relative to WT mice (Figure 2, K and L). Collectively, these data suggest that the observed anorectic hypersensitization of GLP1 analogs in Mc3r–/– mice was independent of emetic responses.

MC3R deletion produces increased sensitivity to diverse anorectic hormones. We next sought to determine whether the role of the MC3R in sensitivity to GLP1 agonists is unique to this family of hormones or more generalizable. We first tested the effect of the long-term adipostatic hormone leptin, a hormone produced by the adipose tissue in proportion to fat stores (28). Given that Mc3r–/– mice display late-onset weight gain and hyperleptinemia (7, 13), these mice should theoretically exhibit leptin resistance. In contrast, Mc3r–/– mice also exhibited increased anorectic sensitivity to leptin; low doses of leptin, which otherwise did not affect food intake in WT mice, showed robust anorectic activity in male Mc3r–/– mice (Figure 3A). The effect was not observed in female mice (Figure 3D), although the estrous cycle, an important determinant of daily food intake and leptin levels (29, 30), was not synchronized in this study. We next tested gut hormones known to act acutely as satiety factors, including a form of the peptide YY (PYY3–36) and cholecystokinin (CCK), both secreted by enteroendocrine cells in the small intestine. We found that male and female Mc3r–/– mice had increased responses to PYY3–36 in a dose-dependent manner over a 2-hour nocturnal feeding period (Figure 3, B and E). Similarly, administration of CCK produced an increased anorectic response in male and female Mc3r–/– mice compared with the WT mice in a dose-dependent manner (Figure 3, C and F).

Figure 3

Deletion of Mc3r results in generalized enhanced sensitivity to anorectic hormones. Nocturnal feeding in response to leptin (0.1–1 mg/kg) 12 hours after injection (vehicle, n = 11, n = 6/group) in Mc3r+/+ and Mc3r–/– (A) male and (D) female mice. Acute dark-phase feeding after administration of PYY3–36 (n = 7/group) in Mc3r+/+ and Mc3r–/– (B ) male and (E) female mice. Acute dark-phase feeding after administration of CCK (n = 7–8/group) in Mc3r+/+ and Mc3r–/– (C ) male and (F) female mice. Data represent the mean ± SEM. Statistical analysis was done by 2-way, repeated-measures ANOVA with Tukey’s post hoc test (A–F).

Relative contributions of the MC3R and MC4R to liraglutide sensitivity. Earlier research showed that the MC3R presynaptically releases GABA from AgRP neurons that synapse onto downstream MC4R neurons, highlighting that the MC3R is a negative regulator of MC4R-expressing neurons (9). Furthermore, it has been reported that coadministration of the GLP1R analog liraglutide and the MC4R agonist setmelanotide has an additive effect on weight loss (29). Hence, we tested the hypothesis that the administration of MC4R agonists might also increase sensitivity to liraglutide-induced inhibition of food intake and weight loss (Figure 4, A–H). Peripheral administration of the MC4R peptide agonist CTX-1211 also increased the sensitivity to liraglutide-induced inhibition of food intake and weight loss in WT male mice 24 hours after treatment (Figure 4, A and B). Sensitization of feeding behavior was less evident in female mice (Figure 4C), but liraglutide-induced weight loss was clearly sensitized by CTX-1211 (Figure 4D). Setmelanotide (Imcivree) (32), an MC4R agonist marketed to treat certain forms of genetic obesity, resulted in similar responses with the combination of liraglutide (Figure 4, E–H), illustrating the idea that multiple MC4R agonists can increase the effect of liraglutide in inhibiting food intake and weight loss. Next, we tested whether MC3R deletion would further enhance sensitization of liraglutide action by CTX 1211 and setmelanotide. Vehicle, setmelanotide, or CTX 1211 (1.5 mg/kg, i.p.), liraglutide (0.1 mg/kg, i.p.), or both an MC4R agonist and liraglutide were administered to WT and Mc3r–/– mice, and 24-hour food intake and changes in body weight were measured (Figure 4, I–L). MC4R agonists and liraglutide decreased food intake and promoted weight loss in WT mice, the magnitude of which was increased for both agents in Mc3r–/– mice. Notably, the anorectic and weight loss responses to the coadministration of MC4R agonists and liraglutide were yet further increased in Mc3r–/– mice compared with WT mice (Figure 4, I–L). Next, we tested whether the heightened responsiveness observed with concurrent administration of the MC4R agonist and liraglutide was additive or synergistic. To test this, we treated WT mice with a low dose of CTX 1211 (1 mg/kg), a concentration known not to affect food intake or body weight, in combination with an optimal dose of liraglutide (0.1 mg/kg). Both male and female mice that received the cotreatment of liraglutide and CTX 1211 exhibited significant reductions in food intake and weight loss compared with those receiving liraglutide alone (Figure 5, A–D).

Figure 4

Mc3r deletion enhances the ability of an MC4R agonist to increase sensitivity to liraglutide. (A, C, E, and G) Twenty-four-hour food intake and (B, D, F, and H) body weight changes in response to CTX 1211 or setmelanotide (2 mg/kg, i.p., n = 8/group) or to liraglutide alone (0.05–0.4 mg/kg, s.c., n = 8), as indicated in male and female mice. (I and K) Twenty-four-hour food intake and (J and L) body weight changes in response to CTX 1211 (1.5 mg/kg, i.p., n = 5–6) or setmelanotide (1.5 mg/kg, i.p., n = 9) alone, liraglutide alone (0.2 mg/kg, s.c.), liraglutide plus CTX 21 or setmelanotide, and vehicle in Mc3r+/+ and Mc3r–/– mice. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, by 2-way, repeated-measures ANOVA.

Figure 5

MC4R agonist synergistically increases the sensitivity of liraglutide. (A and C). Twenty-four-hour food intake and (B and D) body weight changes in response to vehicle (n = 10/group), liraglutide (0.1 mg/kg, n = 10/group), CTX 1211 (1 mg/kg, n = 10/group), and the combination of liraglutide plus CTX 1211 in male (A and B) and female(C and D) mice. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, by 2-way ANOVA.

Circuit mechanisms underlying anorectic hypersensitivity mediated by MC3R inhibition. Previous studies have demonstrated that much of the physiological function of the MC3R derives from the expression of the receptor in AgRP neurons. The MC3R is expressed in nearly all AgRP neurons (10), which project widely throughout the CNS (11). For example, the defective fasting-induced activation of AgRP neurons essential for the stimulation of food intake, observed in the global Mc3r–/– mouse, is also seen in mice with deletion of Mc3r exclusively from AgRP neurons (13). Thus, the Agrp-Cre Mc3rfl/fl mouse was used to study the role of MC3R expression in AgRP neurons in the hypersensitivity to anorectic hormones described here. Male Agrp-Cre Mc3rfl/fl mice were more responsive to the anorectic effects of a single 1 mg/kg dose (i.p.) of leptin than were control Agrp-Cre mice (Figure 6, A and B). Male (Figure 6, C and D) and female (Figure 6, E and F) Agrp-Cre Mc3rfl/fl mice exhibited greater inhibition of feeding and greater weight loss, measured 24 hours following a 0.1 mg/kg (s.c.) dose of liraglutide than did control Agrp-Cre mice. These data suggest that hypersensitization to leptin and liraglutide can be recapitulated by MC3R loss from AgRP neurons alone. Given the impaired activation of AgRP neurons observed in both the global deletion of Mc3r and the selective deletion of Mc3r in AgRP neurons (13), we used a chemogenetics approach involving designer receptors exclusively activated by designer drugs (DREADDs) to assess the functional outcomes of specifically activating AgRP prior to liraglutide treatment. We hypothesized that in the presence of AgRP activation, we could partially rescue the hypersensitivity seen in response to liraglutide in mice with Mc3r deletion. We found that liraglutide-mediated food inhibition and weight loss persisted in mice following DREADD-mediated AgRP neuronal activation (Figure 7, A–D). However, mice with selective deletion of Mc3r in AgRP neurons and DREADD-mediated AgRP neuronal activation showed reduced responsiveness to liraglutide action on feeding and weight loss compared with similarly treated Agrp-Cre (control) mice (Figure 7, C–F). Finally, we examined the role of the Pomc gene, the source of the MC3R and MC4R agonist, in liraglutide and leptin action. Leptin and liraglutide directly activate pro-opiomelanocortin (POMC) neurons and also enhance presynaptic inputs to POMC neurons (12, 30). Leptin- and liraglutide-mediated inhibition of food intake and weight loss were significantly blunted in male mice with a neural-specific deletion of the Pomc gene; no effect was seen in female mice (Figure 8, A–G). While these data do not examine the role of POMC in the hypersensitization to liraglutide by MC3R ablation/antagonism, they do implicate an important role for melanocortin signaling in liraglutide action.

Figure 6

Specific deletion of Mc3r in AgRP neurons increases the responsiveness to liraglutide and leptin. (A) Time-course feeding for Agrp-Cre and (B) Agrp-Cre Mc3rfl/fl mice in response to leptin (1 mg/kg, i.p., n = 8/group). Twenty-four-hour food intake and change in body weight in (C and D) male and (E and F) female mice in response to liraglutide (0.1 mg/kg, s.c., n = 8/group/males, n = 4–6/group/females). Data represent the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, by 2-way, repeated-measures ANOVA with Tukey’s post hoc test (A and B) and 2-tailed Student’s t test (C–F).

Figure 7

The reduction in food intake and weight loss induced by liraglutide persists with DCZ-mediated activation of hM3Dq-DREADD in AgRP neurons. (A) Experimental timeline for DREADD experiments. (B) AAV-mediated functional expression of hM3Dq-mCherry in AgRP neurons after DCZ (0.3 mg/kg) or saline treatment. Representative confocal images are shown after staining for c-Fos. Scale bars: 50 μm. (C and E) Twenty-four-hour food intake and (D and F) changes in body weight after liraglutide (0.1 mg/kg, s.c.) or vehicle treatment in the presence of DCZ (0.3 mg/kg, i.p.) or saline in Agrp-Cre and Agrp-Cre Mc3rfl/fl mice. Data are from both male and female mice and represent the mean ± SEM. *P < 0.05, **P < 0.01, and ****P < 0.0001, by 2-way, repeated-measures ANOVA with Tukey’s post hoc test (A and B) and 2-tailed Student’s t test (C–F).

Figure 8

Genetic deletion of the neuronal Pomc gene abolishes the response to liraglutide and leptin in male mice. (A) Schematic of the FED3 feeding device in free-feeding mode. Twenty-four-hour food intake and changes in body weight 24 hours after injection of liraglutide in (B and C) male mice (0.2 mg/kg, s.c., n = 7/group) and (D and E) female mice (0.2 mg/kg, s.c., n = 4/group). (F and G) Twenty-four-hour food intake and body weight changes in response to leptin (3 mg/kg, i.p.) in male mice (n = 5/group). Data are expressed as the mean ± SEM. **P < 0.01, ***P < 0.001, and ****P < 0.0001, by repeated-measures 2-way ANOVA with Tukey’s post hoc analysis.

Discussion

Previous studies have shown that the Mc3r–/– mouse exhibits modest, late-onset obesity (4), yet increased anorexia and weight loss in response to various challenges relative to WT littermates. For example, the Mc3r–/– mouse was demonstrated to exhibit greater anorexia and weight loss in response to LPS treatment or IL-1β treatment and greater loss of both fat mass and lean mass in response to implantation of Lewis lung carcinoma cells (6). More recently, we demonstrated that the Mc3r–/– mouse exhibit greater anorexia and weight loss than do WT mice in response to behavioral challenges, including restraint and social isolation (10). Furthermore, we demonstrated that the Mc3r–/– mouse exhibits a greater extent and duration of anorexia in response to a single dose of the MC4R agonist LY2112688 (9). Here, we sought to examine this phenomenon further and determine its potential utility in the pharmacotherapy of dietary obesity. First, careful dose-response curves demonstrated that the Mc3r–/– mouse had increased sensitivity to the anorectic and weight loss activities of the GLP1 agonists. In striking contrast, the Mc3r–/– mouse did not show increased sensitivity to the liraglutide-induced incretin effect or malaise, as indicated by a conditioned taste aversion assay. We have demonstrated that assays of Fos IHC, used as an indicator of neuronal activation, identified a liraglutide-induced increase in Fos signal in multiple hypothalamic feeding centers, but no increase in a region mediating emesis, the AP, in Mc3r–/– versus WT mice. Since mice do not show measurable emesis, it will be important to assess whether MC3R antagonism can blunt the emesis induced by the GLP1 drugs in a susceptible species or if comparable weight loss can be obtained at lower GLP1 drug concentrations by MC3R antagonist treatment to avoid pharmacotherapy-induced side effects.

Importantly, the hypersensitization to GLP1 drugs could also be demonstrated using the MC3R antagonist C11 in acute (24-hour) and chronic (3-day) treatment models. Thus, we expect that most of the anorectic hypersensitization phenomena do not result from developmental consequences of MC3R deletion and may ultimately be replicated pharmacologically. Unfortunately, the C11 molecule is difficult to synthesize and has limited brain penetrance, thus requiring i.c.v. administration. To our knowledge, a potent, receptor subtype–specific MC3R antagonist with brain penetrance has yet to be described.

To determine whether the hypersensitivity phenomenon was restricted to GLP1 analogs, we next tested the responsiveness of the Mc3r–/– mouse to both the slow-acting adipostatic hormone leptin and the rapid, acute-acting satiety factors PYY3–36 and CCK. The Mc3r–/– mouse is obese and hyperleptinemic and thus should exhibit leptin resistance. Remarkably, the Mc3r–/– mouse showed increased sensitivity to leptin relative to WT littermates, as determined by 24-hour food intake and weight-loss monitoring. Moreover, pharmacological inhibition of the MC3R also increased leptin action by decreasing food intake and promoting weight loss in males and females (data not shown). Additional work will be needed to understand how MC3R antagonism disrupts the well-characterized phenomenon of leptin resistance. Interestingly, the Mc3r–/– mouse was also hypersensitive to the acute inhibition of feeding by the satiety factors PYY3–36 and CCK.

We have demonstrated that MC3R acts postsynaptically on AgRP neurons to negatively regulate MC4R neurons in PVH (9) and POMC neurons in the arcuate nucleus (12). However, these same AgRP and POMC neurons project througho