Melanocortin receptor 4 deficiency affects body weight regulation, grooming behavior, and substrate preference in the rat.

Obesity is caused by an imbalance between energy intake and expenditure and has become a major health-care problem in western society. The central melanocortin system plays a crucial role in the regulation of feeding and energy expenditure, and functional loss of melanocortin receptor 4 (MC4R) is the most common genetic cause of human obesity. In this study, we present the first functional Mc4r knockout model in the rat, resulting from an N-ethyl-N-nitrosourea mutagenesis-induced point mutation. In vitro observations revealed impaired membrane-binding and subsequent nonfunctionality of the receptor, whereas in vivo observations showed that functional loss of MC4R increased body weight, food intake, white adipose mass, and changed substrate preference. In addition, intracerebroventricular (ICV) administration of Agouti-Related Protein(79-129) (AgRP(79-129)), an MC4R inverse agonist, or Melanotan-II (MTII), an MC4R agonist, did affect feeding behavior in wild-type rats but not in homozygous mutant rats, confirming complete loss of MC4R function in vivo. Finally, ICV administration of MTII induced excessive grooming behavior in wild-type rats, whereas this effect was absent in homozygous mutant rats, indicating that MTII-induced grooming behavior is exclusively regulated via MC4R pathways. Taken together, we expect that the MC4R rat model described here will be a valuable tool for studying monogenic obesity in humans. More specifically, the relative big size and increased cognitive capacity of rats as compared to mice will facilitate complex behavioral studies and detailed mechanistic studies regarding central function of MC4R, both of which ultimately may help to further understand the specific mechanisms that induce obesity during loss of MC4R function.

Introduction

Melanocortin receptor 4 (MC4R) is a key element in hypothalamic control of short- and long-term energy homeostasis by integrating signals provided by α-melanocyte-stimulating hormone (α-MSH), an MC4R agonist, and Agouti-Related Peptide (AgRP), an MC4R inverse agonist (1,2,3,4). The central melanocortin system contains several neuronal circuits including leptin-sensitive neurons expressing proopiomelanocortin (Pomc) or Agrp in the arcuate nucleus, brainstem neurons expressing Pomc, and downstream targets of these neurons expressing melanocortin receptor 3 (Mc3r) and Mc4r (5). In short, a positive energy state activates anorexigenic α-MSH signaling and represses expression of orexigenic Agrp, thus increasing energy expenditure and lowering caloric intake through increased downstream MC3R and MC4R signaling, whereas a negative energy state triggers opposite events.

In humans, haploinsufficiency of Mc4r is the most common monogenic cause of severe obesity, accounting up to ~3% of all cases (6,7,8,9). Human genome-wide association studies have demonstrated that single-nucleotide polymorphisms in MC4R or genetic variances near MC4R are highly associated with increased BMI, adipose mass, early-onset obesity, and risk to develop obesity (10,11,12,13). Moreover, about 150 MC4R mutations, predominantly with a negative effect on MC4R function, have been reported to date in humans (reviewed in ref. 5). In addition, targeted disruption of Mc4r in mice resulted in increased food intake, body weight, body length, serum hormone levels, lean mass, and white adipose tissue (WAT) levels (14). Finally, these Mc4r mice show an identical clinical syndromal picture as compared to human patients with mc4r haploinsufficiency (14,15,16).

In this study, we present the first functional knockout rat model for Mc4r, which was identified in a recent N-ethyl-N-nitrosourea (ENU)-mutagenesis screen in our lab (17). We observed an ENU-induced premature stop codon in helix 8 of the G-protein-coupled receptor (Mc4r), and we hypothesized that this mutation would result in loss of function as the mutated MC4RK314X misses, amongst other important amino acids, two C-terminal isoleucines that are essential for correct receptor function (18). Therefore, we tested the functionality of the mutated receptor in vitro, and performed a basic characterization of rats heterozygous (Mc4r) and homozygous (Mc4r) for the mutation. Furthermore, we tested the effect of Agouti-Related Protein79–129 (AgRP79–129), an MC4R inverse agonist, on feeding behavior and the effect of Melanotan-II (MTII; ref. 19), an MC4R agonist, on feeding behavior and grooming behavior in Mc4r rats. Finally, we investigated whether Mc4r rats demonstrated a changed substrate preference when offered a high-fat/high-sucrose (HFHS) choice diet.

Methods and Procedures

Animals

The Animal Care Committee of the Royal Dutch Academy of Science approved all experiments according to the Dutch legal ethical guidelines. The Mc4r mutant rat line (Mc4r) was generated by target-selected ENU-driven mutagenesis (see ref. 17), and high-throughput resequencing of genomic target sequences in progeny from mutagenized rats (Wistar/Crl background) revealed an ENU-induced premature stop codon in helix 8 (K314X) of Mc4r. The heterozygous mutant rat was backcrossed to wild-type Wistar background for six generations to eliminate confounding effects from background mutations induced by ENU as described before (20). To further control for possible contributions of confounding mutations, we repeated several measurements in different outcross generations and could replicate previous findings in each generation. Additionally, we always compared littermates that were generated by crossing Mc4r rats. Experimental rats were obtained at the expected Mendelian frequency. Mc4r rats were viable into adulthood and appeared phenotypically normal despite their increased body weight. Two rats were housed together, unless noted otherwise, under controlled experimental conditions (12-h light/dark cycle, light period 0600–1800 hours, 21 ± 1 °C, ~60% relative humidity). Standard fed chow diet (semi high-protein chow: RM3, 27% crude protein, and 12% fat, 3.33 kcal/g AFE; SDS, Witham, UK) and water was provided ad libitum unless noted otherwise. All rats had access to home-cage enrichment (red rat retreat; Plexx, Elst, the Netherlands and aspen gnaw brick; Technilab-BMI, Someren, the Netherlands), unless noted otherwise. Only male rats were used in this study.

Rat genotyping

DNA isolation and genotyping were performed as described previously (21). In brief, a fragment of Mc4r, containing the ENU-induced mutation was amplified using gene-specific primers (forward (F):CCCAA CTTCT ACAGG CAGAC; reverse (R): TGGTA ATGAG GCAGA TGATG) and a touchdown PCR cycling program (92 °C for 60 s; 12 cycles of 92 °C for 20 s, 65 °C for 20 s with a decrement of 0.6 °C per cycle, 72 °C for 30 s, followed by 20 cycles of 92 °C for 20 s, 58 °C for 20 s, and 72 °C for 30 s; 72 °C for 180 s; GeneAmp9700; Applied Biosystems, Foster City, CA). The PCR reactions were diluted with 25 µl water, and 1 µl was used as template for the dideoxy sequencing reactions, which were preformed according to the manufacturer instructions (Applied Biosystems). Sequencing products were purified using Sephadex G50 (superfine, coarse; Sigma, Zwijndrecht, the Netherlands) mini-columns and analyzed on a 96-capillary 3730XL DNA analyzer (Applied Biosystems). Sequences were analyzed for polymorphisms using polyphred (22) and manual inspection of the mutated position. All pups were genotyped around postnatal day (PND) 21. Genotypes were confirmed when experimental procedures were completed.

In vitro membrane expression and cAMP response

N-terminally hemagglutinin (HA)-tagged MC4RWT and MC4RK314X constructs were cloned into the pcDNA3.1 expression vector (Invitrogen, Carlsbad, CA) and membrane expression of the HA-tagged receptors was tested in HeLa cells as described before (17). In vitro membrane expression data shown are general representation of observed results. MC4RK314X functionality was tested using a LacZ reporter-gene assay as previously described (23). Briefly, HEK293 cells co-transfected with either Cre::LacZ and pcDNA3.1-HA-Mc4rWT or with Cre::LacZ and pcDNA3.1-HA-Mc4rK314X were allowed to grow in a 96-well plate for 48 h. Subsequently, cells were incubated with the appropriate concentration α-MSH or forskolin (intrinsic activation of cyclic AMP (cAMP), used as positive control) in serum free Dulbecco's modified Eagle's medium. Dulbecco's modified Eagle's medium was discarded after 5 h of incubation, lysis buffer was added, and plates were stored directly at −20 °C. After defrosting the suspension, the substrate mix was added (1.6 g/liter o-nitrophenyl β--galactopyranoside with β-mercaptoethanol (67.5 mmol/l), and MgCl2 (1.5 mmol/l) in phosphate-buffered saline). Light absorbance was measured at 405 nm. The experiment was done in triplo per genotype and data is shown as average.

A receptor-binding competition assay using 125I-NDP-α-MSH

In vitro MC4R binding was measured by co-transfecting HEK293 cells with 7 µg of pcDNA-HA-MC4RWT or pcDNA-HA-MC4RK314X in combination with 400 ng TK-Renilla luciferase DNA. After 24 h of incubation, cells were split in two aliquots and one aliquot was treated with the indicated concentration of competing cold [Nle4, D-Phe7]-α-MSH (NDP-α-MSH) in combination with a stable concentration of the radio ligand 125I-NDP-α-MSH (7,000 counts per minute per well; both peptides diluted in Ham's F10 medium (Gibco) with 2.5 mmol/l CaCl2, 0.25% bovine serum albumin, and 200 KIU/ml aprotinin supplement) supplemented with Tris-buffered saline containing 2.5 mmol/l CaCl2 for 30 min at room temperature. The second aliquot was used to measure Renilla activity for transfection efficiency calculation and raw data corrections. The data are expressed in counts per minutes (cpm) and was measured in a lysed (using 1 M sodium hydroxide) cell suspension using a γ-counter. The experiment was done in triplo per genotype and data are shown as average.

Body composition

Starting at PND 33, rat body weight was measured weekly. For relative body weight, average body weight of wild-type rats was set at 100% and average body weight of Mc4r rats was expressed as a percentage of average wild-type body weight. At PND 96, left and right subcutaneous and perirenal WAT fat pads were isolated and weighted. Subsequently, perirenal WAT pads were fixed in 4% formaldehyde, rotated overnight at room temperature, washed with phosphate-buffered saline, and rinsed in 70% ethanol for at least 1 h. WAT pads were processed to paraffin (96% ethanol for 2 h, 2× 100% ethanol for 1.5 h, xylene for 2 h and embedded in paraffin overnight), cut to 14 µm sections, and stained with eosin. Adipocyte size of the perirenal WAT was calculated using NIH Image J software (Bethesda, MD). Per genotype, three rats were analyzed (10 pictures per rat, each picture taken at a different morphological site).

Plasma hormone levels

At PND 96, blood samples were collected after decapitation. Plasma insulin and leptin levels were measured as described before (20).

Basal food intake measurements

Starting at postnatal day (PND) 49, rats were housed individually and food intake, water intake, and body weight were measured during PND 56–63.

Bomb calorimetry

During basal food intake measurements, fecal samples were collected between PND 59 and 63, freeze-dried, and analyzed for gross energy content using adiabatic bomb calorimetry (IKA Calorimeter System C4000; IKA, Heitersheim, Germany). The feed conversion efficiency was calculated as the body weight gain (g) divided by the effective energy intake (kcal; kcal ingested minus kcal lost in feces) per day.

Ambulatory activity

Starting at PND 70, rats were allowed to habituate to the experimental room for 24 h before placement in a Phenotyper home-cage monitoring system (Noldus Information Technology, Wageningen, the Netherlands). After 20 h of acclimatization to the Phenotyper cages, basal ambulatory activity was measured during 24 h. Data were analyzed using Ethovision software (Noldus Information Technology) in combination with in-house developed software.

Hypothalamic mRNA expression

Ad libitum-fed rats were killed at PND 182 during the early afternoon and the hypothalamus was rapidly dissected and snap-frozen in liquid nitrogen. Hypothalamic mRNA expression was measured exactly as described before (20), with the addition that DNA was removed from the RNA samples using DNAse (Promega) as described by the manufacturer (Mc3r is a single exon gene). Primers were designed using SciTools PrimerQuest (IDT): Cyclophilin (F: ACTT CATG ATCC AGGG TGGA GACT; R: AAGT TCTC ATCT GGGA AGCG CTCA), Pomc (F: TCCA TAGA CGTG TGGA GCTG GT; R: TTCA TCTC CGTT GCCT GGAA ACAC), Cartpt (F: TGGA CATC TACT CTGC CGTG GAT; R: TTCC TGCA GCGC TTCA ATCT GCAA), Npy (F: AGAG GACA TGGC CAGA TACT ACTC; R: AATC AGTG TCTC AGGG CTGG ATC), and Mc3r (F: CATG TACT TCTT CCTG TGCA GCCT; R: AAGG TCAG GGAG TCGC TGTT GATA) Average Mc4r rat gene expression from the two experiments is expressed as a percentage of average wild-type rat gene expression.

ICV cannula implantation

Starting at PND 89, rats were allowed to habituate to the experimental room and housed individually. The next day, rats were anesthetized by subcutaneous injection of Hypnorm (1.0 ml/kg; 10 mg/ml fluanisone, 0.315 mg/ml fentanyl citrate; Janssen Animal Health, Beerse, Belgium) and pain was reduced by Carprofen (Pfizer Animal Health, New York, NY) intramuscularly administered (0.1 ml/100 g body weight). A 16-mm stainless gauge was implanted in the lateral ventricle (stereotaxic coordinates: 1.0 mm posterior and 2.0 mm lateral of the bregma) as described earlier (24). After surgery, rats received two intraperitoneal injections of 3 ml saline to aid recovery. Rats were allowed to recover for 9 days, during which rats were handled frequently.

ICV administration of MTII and AgRP79–129

At PND 99, body weight was measured and rats were placed on a caloric restriction for 6 h (12 till 6 ). Directly after start of the dark phase (6 ), rats were injected with 3 µl saline. Subsequently, food and water intake was measured for 22 h after which the body weight was measured again at the end of the experiment. This paradigm was repeated with the administration of 1 nmol/l MTII (PND 103; Tocris Bioscience, Bristol, UK) and 1 nmol/l AgRP79–129 (PND 106; Phoenix Pharmaceuticals, Burlingame, CA). Experimental peptides were dissolved in 3 µl saline.

Grooming assay

The grooming assay was done as described before (25). Briefly, at PND 109 rats were transported from the in-house animal facility, to an observation room 2 h before start of the behavioral test, during which rats had no cage enrichment, or access to food or water. Grooming was induced by ICV administration of 1 nmol/l MTII (dissolved in 3 µl saline). In our experiments, the observation cage consisted of a Phenotyper home-cage monitoring system (Noldus Information Technology) in which rats were placed immediately after the ICV injection. Observation started 15 min after the injection and continued for 50 min. Grooming was scored every 15 s during 50 min resulting in maximal 200 grooming scores. The grooming elements vibrating, face washing, body and genital grooming, body licking, scratching, and paw licking were scored. The grooming tests were performed during the early afternoon. After experimental handling, the rats were killed and correct implantation of the ICV cannula was checked using a Methylene Blue injection. Only data from rats with a correct cannula were used.

HFHS-choice diet

Rats were individually housed at PND 69 and received ad libitum access to standard fed semi high-protein diet and water. Body weight and food intake was monitored every 2 days for 10 days (between PND 69 and 79). At PND 79, rats received access to a food dispenser with standard semi high-protein diet, a food dispenser with HF diet (45%-AFE, 20% crude protein, 45% fat, and 35% carbohydrates; 4.54 kcal/g AFE; SDS), one bottle of water, and one bottle of 30% sugar water (1.2 kcal/g). After 6 days of acclimation to the novel diets, body weight and food intake was monitored every 2 days for 10 days (between PND 85 and 95). Dispensers and bottles were swapped every 2 days to control for side-preference effects.

Data analysis

All data are shown as mean ± SEM. All data were analyzed using a commercially available statistical program (SPSS for Macintosh, version 16.0) and were controlled for normality and homogeneity. , 4b, and 8a,b were analyzed using repeated measures analysis, followed by Bonferroni post hoc analyses if significant overall interactions were observed. was analyzed using one-way ANOVA, followed by Bonferroni post hoc analyses if significant overall interactions were observed. Figure 6a,b was analyzed using two-way ANOVA, followed by Bonferroni post hoc analyses if significant overall interactions were observed. All other data were analyzed using a Student's t-test. The null hypothesis was rejected at the 0.05 level.

Results

The Mc4r mutation results in loss of function

We have recently described a mutation in the Mc4r gene in the rat (Mc4r), which generates a premature stop codon and an 18-amino acid truncation in helix 8 of the receptor, thereby removing the C-terminus, which is essential for receptor cell surface localization (; refs. 17,18). Indeed, whereas HA-tagged MC4RWT was clearly detectable at the plasma membrane of HeLa cells, HA-tagged MC4RK314X could not be detected at the plasma membrane (), confirming earlier observations (17). Permeabilization of transfected cells reconfirmed the presence of approximately equal expression levels of both wild-type and mutant HA-MC4R fusion proteins (), indicating that mutant MC4R is probably synthesized but fails to be transported to the plasma membrane where it normally performs its function.

Using an in vitro cAMP assay, we show that HEK293 cells transfected with HA-tagged Mc4rK314X were unable to activate cAMP after addition of α-MSH, whereas cells transfected with HA-tagged Mc4rWT show a dose-dependent response (). The statistical analysis for cAMP readout after α-MSH treatment revealed a significant effect of dilution (F(11,99) = 22; P < 0.001), of genotype (F(2,9) = 418; P < 0.001), and a dilution × genotype interaction (F(22,99) = 34; P < 0.001; ). To control whether the HA tag could have an effect on the assay we measured α-MSH response of HEK293 cells transfected with non-HA-tagged Mc4rWT, which showed an equal cAMP readout as compared to HA-tagged Mc4rWT transfected cells when treated with the same dilution series of α-MSH or forskolin (data not shown). This demonstrates that the HA tag did not influence wild-type receptor activity. Furthermore, treatment with forskolin shows the same degree of cAMP activation in all tested conditions, demonstrating that an equal number of cells were transfected with the different expression vectors (). The statistical analysis for cAMP readout after forskolin treatment revealed a significant effect of dilution (F(7,63) = 272; P < 0.001), but no effect of genotype (F(2,9) = 3; P = 0.11), and no dilution × genotype interaction (F(14,63) = 1; P = 0.44; ).

Finally, we demonstrated that HEK293 cells transfected with HA-Mc4rK314X were unable to bind radioactive 125I-NDP-α-MSH during a receptor-binding competition assay whereas a significant signal was detected for cells transfected with HA-Mc4rWT (). The statistical analysis for 125I-NDP-α-MSH binding after cold NDP-α-MSH treatment revealed a significant effect of dilution (F(4,17) = 40; P < 0.001), of genotype (F(1,4) = 627; P < 0.001), and a dilution × genotype interaction (F(4,17) = 40; P < 0.001; ).

The Mc4r mutation results in obesity

Body weight analysis of rats during development revealed that Mc4r rats developed early-onset obesity whereas Mc4r rats developed late-onset obesity as compared to wild-type siblings (). The statistical analysis for body weight revealed a significant effect of time (F(9,261) = 2262; P < 0.001), an effect of genotype (F(2,29) = 20; P < 0.001), and a time × genotype interaction (F(18,261) = 38; P < 0.001; ). At PND 96, average body weight of Mc4r and Mc4r rats was 493 ± 8 g (116 ± 2%) and 602 ± 12 g (141 ± 3%), respectively, as compared to wild-type siblings (426 ± 15 g; ). Mc4r rats displayed increased subcutaneous and perirenal WAT mass as compared to wild-type siblings (405 ± 21% and 415 ± 18%, respectively; ), both of which remained increased when normalized for body weight (291 ± 15% and 299 ± 15%, respectively; ). Moreover, Mc4r rats demonstrated increased perirenal adipocyte cell size (), increased plasma leptin levels (553 ± 47%; ), and increased plasma insulin levels (372 ± 36%; ) as compared to wild-type siblings. Finally, an increase in body weight in Mc4r and Mc4r rats at PND 96 was accompanied by an increase in body length and waist circumference ().

The Mc4r mutation results in hyperphagia

During postnatal week 9 (PND 56–63), the average body weight of Mc4r rats was increased as compared to wild-type siblings (125 ± 3% as compared to wild-type rats), whereas body weights did not differ significantly between Mc4r and wild-type rats (98 ± 1% as compared to wild-type rats; ). Both Mc4r and Mc4r rats demonstrated increased food intake as compared to wild-type siblings, although the difference between Mc4r and wild-type rats was not significant (). Moreover, when food intake was normalized for body weight, both Mc4r and Mc4r rats were hyperphagic as compared to wild-type siblings (). Mc4r rats also consumed more water as compared to wild-type siblings (see online).

During PND 56–59, fecal output and fecal energy loss of Mc4r rats were increased as compared to Mc4r and wild-type siblings, whereas fecal output did not differ significantly between Mc4r and wild-type rats (). During these 3 days, the feed conversion efficiency did not differ significantly between wild-type, Mc4r, and Mc4r rats ().

The Mc4r mutation decreases ambulatory activity

During postnatal week 10, total ambulatory activity of Mc4r rats was decreased during the light phase, the dark phase, and during 24 h as compared to wild-type siblings (). The statistical analysis for 24 h ambulatory activity revealed a significant effect of time (F(16,217) = 7; P < 0.001), of genotype (F(1,14) = 13; P < 0.005), but no time × genotype interaction (F(16,217) = 1; P = 0.61; ).

The Mc4r mutation affects hypothalamic gene expression

Because the hypothalamus is an important brain region regulating energy balance, gene expression of Pomc, cocaine-, and amphetamine-regulated transcript (Cartpt), neuropeptide-Y (Npy), and Mc3r was investigated in hypothalamus samples of ad libitum-fed adult rats. At PND 182, relative hypothalamic gene expression of Pomc and Cartpt, both food intake-suppressing genes, was upregulated, whereas relative expression of Npy, a food intake-stimulating gene, was unchanged in Mc4r rats as compared to wild-type rats (). In addition, relative hypothalamic gene expression of Mc3r was increased in Mc4r rats as compared to wild-type rats ().

Pharmacological manipulation of the melanocortin system in vivo

Activation of MC4R decreases feeding, whereas blockade of MC4R increases feeding (26,27,28). Therefore, we tested whether Mc4r rats change their feeding behavior in response to ICV administration of AgRP79–129, a nonselective MC3R and MC4R inverse agonist, or MTII, a nonselective MC3R and MC4R agonist (19). ICV administration of 1 nmol/l AgRP79–129 increased the 22-h cumulative caloric intake in Mc4r rats as compared to saline-treated Mc4r rats, but had no effect on feeding in Mc4r rats (). Furthermore, ICV administration of 1 nmol/l MTII lowered caloric intake in Mc4r rats as compared to saline-treated Mc4r rats, but again had no effect on feeding in Mc4r rats (). Finally, body weight growth of Mc4r rats was positive after AgRP79–129 administration and negative after MTII administration, whereas no robust effects were observed in Mc4r rats ().

The MC4R mutation abolishes MTII-induced grooming behavior

ICV administration of MTII induces grooming behavior in rats (29). Therefore, we tested whether Mc4r rats increased their grooming behavior in response to ICV MTII administration. ICV administration of 1 nmol/l MTII resulted in excessive grooming behavior during a rat-grooming assay (87 ± 2% of experimental time spent on grooming behavior) in Mc4r rats, whereas this excessive behavior was absent in Mc4r rats (8 ± 3% of experimental time spent on grooming behavior; ). This demonstrates that melanocortin-induced grooming is exclusively mediated by MC4R.

The MC4R mutation increases preference for HF substrate

Loss or blockade of MC4R signaling increases HF substrate consumption, changes preference towards an HF diet, and increases appetitive responding for a fat, but not a carbohydrate, reinforcer (30,31,32). Moreover, Mc4r mice increase their body weight markedly as compared to wild-type siblings when fed a moderate-fat diet (33). Here, we gave Mc4r and Mc4r rats, raised on a standard chow diet, access to a HFHS-choice diet and measured caloric intake and changes in body weight on both diets. Mc4r rats increased their body weight slightly faster, albeit not significantly, during 10 days on a chow diet as compared to wild-type rats. The statistical analysis for body weight growth on a chow diet revealed a significant effect of time (F(4,28) = 111; P < 0.001), but not of genotype (F(1,7) = 1; P = 0.42) and no time × genotype interaction (F(4,28) = 2; P = 0.14; ). However, on an HFHS-choice diet, Mc4r rats increased their body weight faster as compared to wild-type rats. The statistical analysis for body weight growth on an HFHS-choice diet revealed a significant effect of time (F(2,12) = 386; P < 0.001) but not of genotype (F(1,7) = 1; P = 0.39), and revealed a time × genotype interaction (F(2,12) = 6; P < 0.05; ). Mc4r rats ingested more calories on a standard chow diet as well as on an HFHS-choice diet as compared to wild-type siblings, although this difference was not significant when fed an HFHS-choice diet (). Both genotypes increased their total caloric intake, albeit not significantly, on an HFHS-choice diet as compared to a standard chow diet (Mc4r: 120.4 ± 3.9%, Mc4r: 114.4 ± 4.9%; ). Finally, when fed an HFHS-choice diet, Mc4r rats ingested equal calories from the standard chow substrate (Mc4r: 35.5 ± 1.0%, Mc4r: 38.6 ± 6.2%), significantly more calories from the HF substrate (Mc4r: 42.8 ± 1.6%, Mc4r: 53.7 ± 4.3%), and significantly fewer calories from the HS substrate as compared to wild-type rats (Mc4r: 21.7 ± 1.8%, Mc4r: 7.7 ± 2.1%; ). This demonstrates that loss of MC4R function changes substrate preference toward HF energy-rich substrates.

Discussion

Taken together, our in vitro and in vivo data show that the Mc4r mutation results in loss of function. We observed that the Mc4r mutation abolishes correct membrane localization in vitro, and that Mc4r rats demonstrated increased body weight, adipose mass, plasma hormone levels, body length, and basal food intake in combination with decreased ambulatory activity. ICV administration of AgRP79–129 or MTII had no robust effect on feeding behavior in Mc4r rats. Furthermore, ICV MTII administration induced excessive grooming behavior in wild-type rats, whereas it had no robust effect on grooming behavior in Mc4r rats. Finally, when offered an HFHS-choice diet, Mc4r rats demonstrated increased preference for an HF substrate in combination with decreased preference for an HS substrate as compared to wild-type rats.

The in vitro data obtained with our rat model confirms earlier observations that the C-terminus is essential for correct MC4R cell surface localization, and that subsequent loss of cell surface localization abolishes correct MC4R signaling (18). Moreover, mutation of one of the two C-terminal isoleucines (Ile-317-Thr) in a human proband results in severe overweight (34), indicating that our rat model is a bonafide animal model resembling loss of MC4R function in humans.

Functional loss of MC4R in the rat results in a gene dosage effect on body weight that showed strong overlap with data observed after Mc4r disruption in mice (14). Moreover, Mc4r rats demonstrated increased body length, increased food intake, increased plasma insulin and leptin levels, and decreased 24 h ambulatory activity that are also in line with earlier observations in Mc4r-null mice (14,35). Finally, ICV administration of AgRP79–129 affected feeding behavior and ICV administration of MTII affected feeding and grooming behavior in wild-type rats, both as expected (26,28,29,36), but these effects were absent in Mc4r rats. In sum, our in vivo data confirm earlier observations in other rodent models and indicate that the Mc4r mutation results in functional loss of MC4R.

Hypothalamic gene expression analysis revealed no significant changes in Npy expression, but an increased expression of Pomc, Cartpt, and Mc3r in adult Mc4r rats as compared to wild-type siblings. Increased expression of Pomc was not observed in any hypothalamic brain region of Mc4r-null mice, whereas increased expression of Npy was observed in the dorsal medial hypothalamic nucleus of Mc4r-null mice (14). Thus, more specific analysis of gene expression in individual brain regions involved in the regulation of body weight might provide important information on differences between murine and rat models, and the processes underlying MC4R-deficiency-induced obesity.

Mc3r mice on a standard chow diet demonstrate normal body weights, albeit adiposity levels were increased in combination with a lower lean mass as well as a decreased food intake (37). Furthermore, Mc3r and Mc4r double knockout mice demonstrated heavier body weights than single gene knockouts (37,38). Finally, recently it was shown that MC3R function has an effect, although subtle, on food intake (39). In sum, MC3R, which is generally considered an autoreceptor (40), affects energy homeostasis. Thus, the relative hypothalamic upregulation of Mc3r observed in Mc4r rats might result in compensatory effects.

Administration of MTII induces excessive grooming behavior in wild-type rats (29), whereas Mc4r rats do not display excessive MTII-induced grooming behavior. Moreover, the time spent grooming by MTII-injected Mc4r rats did not significantly differ from time spent grooming by saline-injected wild-type rats during a comparable assay (data not shown; ref. 29). This indicates that MTII-induced grooming behavior is activated exclusively through MC4R-specific pathways, and not via MC3R-specific pathways.

Loss or blockade of MC4R signaling increases HF substrate consumption, changes preference towards an HF diet, and increases appetitive responding for a fat, but not a carbohydrate, reinforcer in rodents (30,31,32). Here, we show that functional loss of MC4R in the rat increases preference toward an HF substrate and away from an HS substrate, when offered an HFHS-choice diet. As the mechanism of these motivational changes is not yet fully understood, additional behavioral and molecular experiments with Mc4r rats on HF diets or on HFHS-choice diets will be of interest. Finally, Mc4r mice increased their body weight markedly as compared to wild-type mice when fed a moderate HF diet for 7 days, whereas body weight growth of Mc4r rats on an HFHS-choice diet for 10 days did not differ markedly from wild-type rats. An explanation for this observation might lie in species-specific differences, or in the fact that Mc4r rats had a substrate choice whereas Mc4r mice had not. Therefore, a detailed analysis of the homeostatic response of Mc4r rats to a (moderate) HF diet, without access to other substrates, will provide valuable information.

Given the amount of outcrossing to a wild-type background in our experimental Mc4r rats and the strong phenotypic overlap between Mc4r rats and Mc4r-null mice, we are confident that the observed phenotypes in Mc4r rats solely result from functional loss of MC4R. Also, an alternative to the conclusion that functional loss of Mc4r directly influences energy metabolism is that loss of Mc4r could result in a defect in hypothalamic brain development resulting in hyperphagia and obesity. This argument is formally difficult to exclude; however, no gross neuroanatomical defects were observed in brain sections from Mc4r rats. Furthermore, direct evidence for a pharmacological etiology is provided by the modulation of feeding behavior and body weight regulation by acute or chronic administration of MC4R agonists and antagonists, as described above.

We believe that the Mc4r rat model will be of great value to study the cause of human monogenic obesity by adding the rat to the field of comparative genomics. Moreover, due to the increased size and cognitive performance of the rat as compared to mice, allowing for more complex surgeries and behavioral experiments, we believe that our rat model can help further understand the specific mechanisms that induce obesity during loss of MC4R function.