Melanocortin-4-receptor (MC4R)-expressing neurons modulate food intake and preference in rodents but their role in human food preference is unknown. Here we show that compared with lean and weight-matched controls, MC4R deficient individuals exhibited a markedly increased preference for high fat, but a significantly reduced preference for high sucrose food. These effects mirror those in Mc4r null rodents and provide evidence for a central molecular circuit influencing human macronutrient preference.
Neural circuits involving the hypothalamus, brainstem and mesolimbic system play a pivotal role in the regulation of eating behaviour in response to changes in nutritional state and meal consumption1,2. Primary leptin-responsive neurons in the arcuate nucleus of the hypothalamus expressing the anorectic peptide pro-opiomelanocortin (POMC) or the orexigen agouti-related peptide (AgRP) project to the paraventricular nucleus of the hypothalamus where they synapse with neurons expressing the melanocortin 4 receptor (MC4R) to regulate food intake3,4. Targeted disruption of melanocortin signalling in Pomc, Agouti and Mc4r null mice leads to an increase in food consumption and more specifically to increased preference for a high fat diet in food choice experiments5,6,7. Studies in Mc4r−/− mice and rodents harbouring a point mutation in Mc4r, suggest that disruption of melanocortin signalling paradoxically reduces consumption of high sucrose solutions and food6,8, an effect that is not seen with the non-nutritive sweetener sucralose, and is not explained by conditioned taste aversion6. Interestingly, live optical recording of AgRP and POMC neurons in awake mice has revealed direct activation/inhibition of these neurons by a highly palatable diet offered in the energy-deficient state9. While there is considerable data regarding the contribution of central melanocortin signalling to food preference in rodent models, the relevance of these findings to food preference in humans has not been explored. Highly penetrant loss of function MC4R variants in humans are rare (1–5% of people with severe obesity), but as the phenotype of human MC4R deficiency closely parallels that seen in mice lacking Mc4r10, studies in genetically characterized individuals allow us to directly test whether disruption of melanocortin signalling alters food preference in humans. We tested whether MC4R deficient individuals have an altered preference for foods of varying fat and sucrose content. We found that compared with obese and lean controls, individuals with MC4R mutations show an increased preference for high fat food but a decreased preference for sucrose containing food.
Fat preference test
We tested whether MC4R deficient individuals have an altered preference for foods of varying fat and sucrose content, compared with obese and lean controls. We developed a three choice, ad libitum meal paradigm in which fat content was covertly manipulated to provide 20 (low), 40 (medium) and 60% (high) of the total caloric content (Table 1; Methods), critically without altering appearance, texture or taste. Twenty lean controls (BMI 23.1±0.3 kg m−2), 20 obese controls (BMI 34.1±1.1 kg m−2) and 14 adults with heterozygous loss of function MC4R variants (BMI 38.8±2.2 kg m−2; Table 2) participated; BMI did not differ significantly between the obese and MC4R groups. Results were analysed using either ANOVA analysis (with interaction terms for study group and study meal) with Tukey's HSD post-hoc comparisons for the fat preference study. Significance was set at P=0.05.
Participants were initially provided with ‘tasters’ (15 g) of the 3 meals, and liking/disliking ratings obtained using visual analogue scales (VAS) before and after meal consumption. Liking ratings for the low/medium/high fat meals were similar before and after meal consumption for all three groups (Fig. 1a,b). Although liking scores for the high-fat meal were comparable to those for the low and medium fat meals, MC4R deficient individuals consumed 95% more of the high fat meal than lean, and 65% more than obese controls (Fig. 1c,d; P=0.0222). There was no difference in total intake collapsed across fat level (mean±s.e.m.; g) between lean controls (488.5±43.1), obese controls (551.0±48.8) and MC4R deficient individuals (561.5±60.6); P=0.53 for the group comparison.
Sucrose preference test
In the second study, we investigated sucrose preference in a further 20 lean (BMI 22.5±0.4 kg m−2) and 20 obese controls (BMI 35.9±1.2 kg m−2) and 10 adults with heterozygous loss of function MC4R variants (Table 2, BMI 41.3±3.2 kg m−2). The test dessert provided 8 (low), 26 (medium) and 54% (high) sucrose; to avoid interference from fat preference in this paradigm, the energy content from fat was clamped at ∼30% (Table 3).
We found that lean and obese volunteers liked the high sucrose meal more than the low or medium sucrose meals in keeping with previous studies of sucrose preference in volunteers11 (Fig. 2a,b). In MC4R deficient individuals, although liking ratings for low and medium sucrose meals were comparable with the lean and obese groups, liking ratings for the high sucrose meal were significantly reduced (Fig. 2a,b; P=0.0252). Individuals with MC4R variants also consumed significantly less of all three sucrose meals compared with lean and obese controls (Fig. 2c,d; P=0.0064). In summary, our findings in MC4R deficient individuals indicate that central melanocortin circuits play a key role in modulating fat and sucrose preference in humans.
Sensory discrimination test
It is well recognized that oral perception can affect food preference11. To test whether there was a perceptible sensory difference between the low (20%; A) and high (60%; B) fat test meals, we employed a sensory discrimination test (Triangle Test). We found that 41 out of the 78 panellists we tested, correctly identified the odd sample in the test. According to international standards for this test (www.astm.org/Standards/E1885.htm), in a study with 78 panellists (α=0.001) at least 40 correct responses are required to reject the null hypothesis (that is: that there is no difference between the meals). Since we observed 41 correct responses, we cannot conclude that the two meal types are the same, although these results are suggestive. As we were not able to perform this test in a comparable number of people with MC4R deficiency due to the rarity of this disorder, we cannot formally exclude the possibility that attenuated perception of fat may have contributed to the effect that we observed. However, it is noteworthy that on debriefing, participants did not report differences between the foods and were unaware that we had manipulated fat content.
Although common genetic variants in the fatty acid translocase CD36 have been associated with liking of high-fat foods in African Americans12 and common obesity-associated variants have been associated with diary reports of food choices in a number of studies13,14, this is to our knowledge one of the first experimental studies to show a direct association between macronutrient preference (other than alcohol) and a specific genetic/molecular mechanism in humans. Additional studies using a forced choice paradigm which makes a direct comparison between fat and sucrose will be needed to test whether disruption of melanocortin signalling primarily alters sucrose or fat sensing or the reward from these stimuli.
The potential neural mechanisms involved in fat and sucrose preference have been explored in rodent studies. Fat preference may in part be mediated by MC4R-expressing neurons in the amygdala where injection of AgRP acutely increases15, whereas the agonist Melanotan-II decreases16, fat consumption in rats. The central nucleus of the amygdala receives and provides projections to the parabrachial nucleus, a region which responds to gustatory and visceral signals, is involved in sucrose preference and is connected to excitatory MC4R-expressing hypothalamic neurons17,18,19. Previous studies have shown that pre-prandial gut hormone levels (total ghrelin, PYY and GLP-1), as well as plasma insulin and glucose are comparable in MC4R deficiency and weight-matched controls20,21,22. Although postprandial total ghrelin suppression was attenuated in MC4R-deficient individuals compared with lean controls in a previous study22, as this difference emerges 30 min after meal initiation, it is unlikely to contribute to the effects on food preference seen in this study.
The additional experimental flexibility afforded by investigating humans in this study enabled us to dissociate consumption from subjective liking. Interestingly, the increased consumption of high fat food in MC4R deficient individuals did not appear to be related to increased subjective liking. One possible explanation is that the motivating effect of high fat content is implicit and non-conscious, consistent with a theoretical distinction between ‘liking’ and ‘wanting’23. However, a failure to observe an effect on rated (subjective) liking may be an issue of power and/or reflect the choice of scaling method used.
What is the potential physiological relevance of these findings? In free living (nutritionally replete) humans, food preference is complicated by the fact that sweet foods are often also high fat energy dense foods, thus the choice is rarely between a high fat food and a high sucrose food. We hypothesize that the preference for energy-dense high-fat foods in animals and humans may represent part of the adaptive response to nutritional deprivation. Fasting/starvation leads to a fall in leptin levels which triggers physiological and behavioural responses to restore energy homoeostasis24. In keeping with this hypothesis, leptin-deficient ob/ob mice provided with single energy-source diets self-select a higher proportion of energy from high-fat (94%) and a lower proportion from high-carbohydrate food (3%) (ref. 25). Critical to mediating the response to fasting/starvation is an increase in the endogenous MC4R antagonist AgRP, which potently stimulates food intake26, increases fat preference and reduces carbohydrate intake when administered centrally27. Thus, the preference for fat (which delivers twice as many calories/gram as carbohydrates or protein and can be readily stored in adipose tissue), at the expense of sugars/carbohydrates may represent an advantageous behaviour that is expressed in the face of nutritional depletion. Further studies of food preference in the weight reduced/partially-leptin deficient state will be needed to test this hypothesis. The detailed behavioural study of humans with naturally occurring mutations affecting specific central signalling pathways provides a powerful tool to link contemporary advances in the understanding of the molecular control of energy balance to the biology underpinning food preference in humans.
Recruitment and ethical committee approval
All studies were approved by the Cambridge Local Research Ethics Committee (03/103) and undertaken after informed consent. Lean and obese individuals were recruited from the local population by advertisement and MC4R deficient individuals were recruited from our existing cohort (the Genetics of Obesity Study20). Exclusion criteria were the use of any medication, any medical or psychiatric disease and food allergies/intolerances.
Fat preference test
We developed a three choice, ad libitum meal paradigm (Chicken Korma and Rice) in which fat content was covertly manipulated (by adding rapeseed oil) to provide 20 (low), 40 (medium) and 60% (high) of the total caloric content (Table 1), critically without altering appearance, texture or taste. The test was conducted 4 h after a standardized breakfast which provided 20% of individual energy requirements. Participants were initially provided with ‘tasters’ (15 g) of the three meals and liking/disliking ratings obtained using VAS before meal consumption. We then provided large dishes containing the low/medium/high fat chicken korma curry and rice (finely chopped and mixed to prevent selection of specific components) in excess (10 MJ; ∼2,390 kcal). Participants were encouraged to try the three dishes without explicitly stating that the three meals were different and instructed to eat until they were comfortably full. The amount consumed was covertly weighed. After the meal, liking ratings were measured as before.
Sucrose preference test
The test dessert consisted of Eton mess (sweet British pudding consisting of yoghurt, cream, meringues and strawberries) to provide 8% (low) sucrose (Table 3). We then added caster sugar (sucrose) to increase the energy content from sucrose to 26 (medium) and 54% (high). To avoid interference from fat preference in this paradigm, the energy content from fat was clamped at ∼30% (Table 3). Participants were admitted to the Clinical Research Facility at 3:00 pm after a standardized breakfast and lunch. Participants were initially provided with ‘tasters’ (15 g) of the three desserts and liking/disliking ratings obtained using VAS before and after consumption. Participants were invited to try all three desserts, provided in excess, and eat until they were comfortably full. The amount consumed was covertly weighed.
Sensory discrimination test
To test whether there was a perceptible sensory difference between the low (20%; A) and high (60%; B) test meals, we employed a Triangle Test using the American Society for Testing and Materials (ASTM) standard protocol28. Seventy-eight healthy volunteers received a triad of samples (15 g) coded with a three-digit random number. Each set had two identical samples and one odd sample, presented simultaneously in a predetermined counterbalanced order. Assessors were asked to record which they believed to be the ‘odd’ sample. If the participant was unable to determine which sample was different, they were asked to guess, since there was a possibility that they were not consciously aware that they could taste the difference. Samples were randomly presented to avoid positional bias since the middle sample is usually chosen as odd. Possible combinations of samples were: AAB, ABA, BAA, BBA, BAB and ABB. Between tasting samples, participants were asked to take a sip of water to avoid sensory fatigue.
Results were analysed using either ANOVA analysis (with interaction terms for study group and study meal) with Tukey's HSD post-hoc comparisons for the fat preference study or a linear mixed-effects model for the sucrose preference study. Significance was set at P=0.05. All statistical analyses were performed in R.
Data availability statement
All relevant data are available from the authors.
How to cite this article: van der Klaauw, A. A. et al. Divergent effects of central melanocortin signalling on fat and sucrose preference in humans. Nat. Commun. 7, 13055 doi: 10.1038/ncomms13055 (2016).
Gautron, L., Elmquist, J. K. & Williams, K. W. Neural control of energy balance: translating circuits to therapies. Cell 161, 133–145 (2015).
Waterson, M. J. & Horvath, T. L. Neuronal regulation of energy homeostasis: beyond the hypothalamus and feeding. Cell Metab. 22, 962–970 (2015).
Fan, W., Boston, B. A., Kesterson, R. A., Hruby, V. J. & Cone, R. D. Role of melanocortinergic neurons in feeding and the agouti obesity syndrome. Nature 385, 165–168 (1997).
Garfield, A. S. et al. A neural basis for melanocortin-4 receptor-regulated appetite. Nat. Neurosci. 18, 863–871 (2015).
Koegler, F. H., Schaffhauser, R. O., Mynatt, R. L., York, D. A. & Bray, G. A. Macronutrient diet intake of the lethal yellow agouti (Ay/a) mouse. Physiol. Behav. 67, 809–812 (1999).
Panaro, B. L. & Cone, R. D. Melanocortin-4 receptor mutations paradoxically reduce preference for palatable foods. Proc. Natl Acad. Sci. USA 110, 7050–7055 (2013).
Tung, Y. C., Rimmington, D., O'Rahilly, S. & Coll, A. P. Pro-opiomelanocortin modulates the thermogenic and physical activity responses to high-fat feeding and markedly influences dietary fat preference. Endocrinology 148, 5331–5338 (2007).
Mul, J. D. et al. Melanocortin receptor 4 deficiency affects body weight regulation, grooming behavior, and substrate preference in the rat. Obesity 20, 612–621 (2012).
Chen, Y., Lin, Y. C., Kuo, T. W. & Knight, Z. A. Sensory detection of food rapidly modulates arcuate feeding circuits. Cell 160, 829–841 (2015).
van der Klaauw, A. A. & Farooqi, I. S. The hunger genes: pathways to obesity. Cell 161, 119–132 (2015).
Drewnowski, A. Sensory preferences for fat and sugar in adolescence and adult life. Ann. N. Y. Acad. Sci. 561, 243–250 (1989).
Keller, K. L. et al. Common variants in the CD36 gene are associated with oral fat perception, fat preferences, and obesity in African Americans. Obesity 20, 1066–1073 (2012).
Cecil, J. E., Tavendale, R., Watt, P., Hetherington, M. M. & Palmer, C. N. An obesity-associated FTO gene variant and increased energy intake in children. N. Engl. J. Med. 359, 2558–2566 (2008).
Bauer, F. et al. Obesity genes identified in genome-wide association studies are associated with adiposity measures and potentially with nutrient-specific food preference. Am. J. Clin. Nutr. 90, 951–959 (2009).
Boghossian, S., Park, M. & York, D. A. Melanocortin activity in the amygdala controls appetite for dietary fat. Am. J. Physiol. Regul. Integr. Comp. Physiol. 298, R385–R393 (2010).
Samama, P., Rumennik, L. & Grippo, J. F. The melanocortin receptor MCR4 controls fat consumption. Regul. Pept. 113, 85–88 (2003).
Shah, B. P. et al. MC4R-expressing glutamatergic neurons in the paraventricular hypothalamus regulate feeding and are synaptically connected to the parabrachial nucleus. Proc. Natl Acad. Sci. USA 111, 13193–13198 (2014).
Yamamoto, T., Shimura, T., Sakai, N. & Ozaki, N. Representation of hedonics and quality of taste stimuli in the parabrachial nucleus of the rat. Physiol. Behav. 56, 1197–1202 (1994).
King, B. M. et al. Amygdaloid-lesion hyperphagia: impaired response to caloric challenges and altered macronutrient selection. Am. J. Physiol. 275, R485–R493 (1998).
Farooqi, I. S. et al. Clinical spectrum of obesity and mutations in the melanocortin 4 receptor gene. N. Engl. J. Med. 348, 1085–1095 (2003).
Greenfield, J. R. et al. Modulation of blood pressure by central melanocortinergic pathways. N. Engl. J. Med. 360, 44–52 (2009).
van der Klaauw, A. A. et al. Postprandial total ghrelin suppression is modulated by melanocortin signaling in humans. J. Clin. Endocrinol. Metab. 98, E288–E292 (2013).
Castro, D. C. & Berridge, K. C. Advances in the neurobiological bases for food 'liking' versus 'wanting'. Physiol. Behav. 136, 22–30 (2014).
Ahima, R. S. et al. Role of leptin in the neuroendocrine response to fasting. Nature 382, 250–252 (1996).
Romsos, D. R. & Ferguson, D. Self-selected intake of carbohydrate, fat, and protein by obese (ob/ob) and lean mice. Physiol. Behav. 28, 301–305 (1982).
Betley, J. N., Cao, Z. F., Ritola, K. D. & Sternson, S. M. Parallel, redundant circuit organization for homeostatic control of feeding behavior. Cell 155, 1337–1350 (2013).
Tracy, A. L., Clegg, D. J., Johnson, J. D., Davidson, T. L. & Benoit, S. C. The melanocortin antagonist AgRP (83-132) increases appetitive responding for a fat, but not a carbohydrate, reinforcer. Pharmacol. Biochem. Behav. 89, 263–271 (2008).
Drewnowski, A., Krahn, D. D., Demitrack, M. A., Nairn, K. & Gosnell, B. A. Naloxone, an opiate blocker, reduces the consumption of sweet high-fat foods in obese and lean female binge eaters. Am. J. Clin. Nutr. 61, 1206–1212 (1995).
Yeo, G. S. et al. Mutations in the human melanocortin-4 receptor gene associated with severe familial obesity disrupts receptor function through multiple molecular mechanisms. Hum. Mol. Genet. 12, 561–574 (2003).
Rene, P. et al. Pharmacological chaperones restore function to MC4R mutants responsible for severe early-onset obesity. J. Pharmacol. Exp. Ther. 335, 520–532 (2010).
This work was supported by the Wellcome Trust (to A.A.v.d.K., P.C.F., I.S.F.), the National Institute for Health Research Cambridge Biomedical Research Centre (to S.O’R., I.S.F.), the Bernard Wolfe Health Neuroscience Fund (to A.A.v.d.K., I.S.F., P.C.F.) and the European Research Council (I.S.F.). This work was supported by the NeuroFAST consortium which is funded by the European Union’s Seventh Framework Programme (FP7/2007–2013) under grant agreement no 245009.
The authors declare no competing financial interests.
About this article
Cite this article
van der Klaauw, A., Keogh, J., Henning, E. et al. Divergent effects of central melanocortin signalling on fat and sucrose preference in humans. Nat Commun 7, 13055 (2016). https://doi.org/10.1038/ncomms13055
This article is cited by
Adropin correlates with aging-related neuropathology in humans and improves cognitive function in aging mice
npj Aging and Mechanisms of Disease (2021)
Immunoglobulin G modulation of the melanocortin 4 receptor signaling in obesity and eating disorders
Translational Psychiatry (2019)
Neuronal SIRT1 regulates macronutrient-based diet selection through FGF21 and oxytocin signalling in mice
Nature Communications (2018)
Formalising recall by genotype as an efficient approach to detailed phenotyping and causal inference
Nature Communications (2018)