Funding agencies: This work was funded by a studentship from the University of Aberdeen and AstraZeneca.
Disclosure: The authors declare no conflicts of interests.
The melanin-concentrating hormone (MCH) is a centrally acting peptide implicated in the regulation of energy homeostasis and body weight, although its role in glucose homeostasis is uncertain. Our objective was to determine effects of MCHR1 antagonism on energy budgets and glucose homeostasis in mice.
Effects of chronic oral administration of a specific MCHR1 antagonist (GW803430) on energy budgets and glucose homeostasis in diet-induced obese (DIO) C57BL/6J mice were examined.
Oral administration of GW803430 for 30 days reduced food intake, body weight, and body fat. Circulating leptin and triglycerides were reduced but insulin and nonesterified fatty acids were unaffected. Despite weight loss there was no improvement in glucose homeostasis (insulin levels and intraperitoneal glucose tolerance tests). On day 4-6, mice receiving MCHR1 antagonist exhibited decreased metabolisable energy intake and increased daily energy expenditure. However these effects had disappeared by day 22-24. Physical activity during the dark phase was increased by MCHR1 antagonist treatment throughout the 30-day treatment.
GW803430 produced a persistent anti-obesity effect due to both a decrease in energy intake and an increase in energy expenditure via physical activity but did not improve glucose homeostasis.
Melanin-concentrating hormone (MCH) is a 19-animo acid cyclic peptide predominantly expressed in the lateral hypothalamus and zona incerta, with projections to the dorsal and ventral striatum, prefrontal cortex, nucleus of the solitary tract, and the parabrachial nucleus (). MCH acts via two G protein-coupled receptors, MCHR1 and MCHR2 () expressed in humans, rhesus monkeys (Macaca mulatta), dogs (Canis familiaris), and ferrets (Mustela putorius furo) with similar distribution patterns (). However, MCHR2 is not expressed in rodents. MCHR1 is widely distributed in the mouse and rat brain ([5, 6]) including areas associated with energy homeostasis such as the arcuate, ventro-medial and dorso-medial nuclei in the hypothalamus ([5, 7]).
The MCH/MCHR1 pathway plays a key role in the regulation of feeding behavior and energy balance. Acute intra-cerebro-ventricular (ICV) injection of MCH stimulated food and water intake in rats ([8, 9]). Chronic ICV infusion of MCH in rats () and diet-induced obese (DIO) mice () induced hyperphagia and body weight gain, especially on a high-fat diet. In contrast, MCH−/− mice were hypophagic and lean compared to wild-type littermates (), and genetic ablation of the MCHR1 gene resulted in a lean phenotype accompanied by hypophagia and increased energy expenditure (). Furthermore, acute or chronic ICV injections of MCHR1 agonists in rodents resulted in the same phenotype observed with MCH treatment () whereas central and peripheral administration of a MCHR1 antagonist led to reduced food intake and body weight (). In addition to effects on energy balance, MCH has also been implicated in glucose homeostasis, but the relevant data are confused. MCHR1 knockout mice had lower insulin levels compared with wild-type mice, indicative of increased insulin sensitivity (). Similarly, MCHR1 deficiency in ob/ob mice caused a lower blood glucose response and markedly lower insulin levels despite no differences in body weight (). In contrast, central MCH injection to rats induced insulin resistance ().
Brain penetration is required for MCHR1 antagonists to inhibit food intake and to reduce body weight (). Recently a number of small-molecule antagonists for MCHR1 have been developed for the treatment of obesity. However, relatively few studies have addressed the mechanisms underlying the antiobesity effects of these MCHR1 antagonists. Huang et al. showed that rats treated with a MCHR1 antagonist (1-(4-Amino-phenyl)-pyrrolidin-3-yl-amine and 6-(3-aminopyrrolidin-1-yl)-pyridin-3-yl-amine) lost significantly more weight than pair-fed matches (). Acute administration of a highly selective and potent MCHR1 antagonist () that has high brain penetrability and oral bioavailability increased metabolic rate in DIO mice (). Moreover, mice chronically treated with this antagonist exhibited a significantly higher body temperature compared to pair-fed controls (). However, neither body weight nor fat mass differed between DIO mice treated with the MCHR1 antagonist SCH-A ((±)-N-[trans-5-(4-cyanophenyl)bicyclo[3.1.0]hex-2-yl]-N'-[4-fluoro-3-(trifluoromethyl)phenyl]-N-[3-(4-methyl-1-piperazinyl)propyl]urea), and the pair-fed group during a 5-day treatment (), which, together with unchanged energy expenditure (), suggests that the effect of SCH-A on body weight was only due to suppression of feeding. GW803430 is a potent and selective non-peptide MCHR1 antagonist () that has been suggested to have weight-reducing effects (). To examine exactly how GW803430 affected energy balance and glucose homeostasis, we evaluated energy budgets at early (day 4-6) and late (day 22-24) stages of a 30-day oral treatment, compared to a control group receiving only vehicle. Physical activity (PA) and body temperature (Tb) were monitored throughout the experiment, and resting metabolic rate (RMR) was measured on day 19 of treatment. We measured glucose homeostasis using i.p. glucose tolerance tests and levels of circulating insulin in fasted animals. This protocol permitted investigation of the effects of chronic MCHR1 antagonist administration on the different components of energy expenditure and glucose regulation.
Male C57BL/6 mice were fed on a high-fat diet (HFD) for 16 week to establish diet-induced obesity and then were randomly divided into MCHR1 antagonist-treated group and vehicle-treated group. GW803430, a specific MCHR1 antagonist, and its vehicle were orally administered to two groups of mice on a daily basis for 30 days (for more details see Supporting Information uploaded separately).
Body weight and food intake was measured every day. Body fatness was evaluated using dual energy X-ray absorptiometry (DXA) at baseline and on day 26 of dosing. PA and Tb was constantly recorded via Mini-Mitter and VitalView™ system throughout the experiment. All mice received an i.p. glucose tolerance test (GTT) on day 27 (see Supporting Information uploaded separately). At the end of 30-day treatment, mice were sacrificed and blood samples were collected for the determination of metabolite levels (leptin, insulin, triglyceride, and nonesterfied fatty acid). See Zhang et al. () for details.
On days 4-6 and day 22-24, energy budget profile was obtained by measuring energy intake, assimilation efficiency and energy expenditure. We measured daily energy expenditure (DEE) using the doubly labeled water (DLW) technique. Feces accumulated during DEE evaluation were collected and dried to a constant mass. The gross energy (GE) content of the samples was measured using bomb calorimetry. The apparent energy absorption efficiency (AEAE) and metabolizable energy intake (MEI) was calculated as detailed by Król et al. (). RMR was measured at baseline (week 15 of HFD) and on day 19 of treatment using an open-flow respiratory system.
All data were expressed as means ± SD. General linear modeling (GLM) with repeated measures was used to compare body mass, food intake, PA, Tb between MCHR1 antagonist and vehicle groups throughout the experiment. Independent t test with Bonferroni correction was performed to compare differences between groups at a given time point. Analysis of covariance (ANCOVA) was performed to examine differences in MEI, DEE, and RMR using body mass as a covariate (). MEI, DEE, and RMR were corrected for body mass using the residual method in linear regression and corrected data were then compared by independent t tests. Body fat mass and corrected RMR data were analyzed using repeated measures to examine differences between baseline and treatment. Paired t test was performed to compare body fat mass before and after dosing. Group differences in plasma leptin, insulin, TG, NEFA levels, and fat depots were examined with independent t test. P values < 0.05 were considered statistically significant. All data were analyzed using SPSS 17.0 statistical package for Windows (SPSS, Chicago, IL).
Body mass and fatness
Prior to the treatment, there was no difference in body mass between vehicle-treated control and MCHR1 antagonist-treated group (Vehicle: 42.94 ± 6.40 g; MCHR1 antagonist: 42.71 ± 5.52 g. Mice receiving MCHR1 antagonist showed a significant loss of body weight throughout the experiment (GLM repeated measures: F(1,19) = 43.55, P < 0.001, Figure 1A), and the relative weight change was significantly greater in MCHR1 antagonist-treated mice than vehicle-treated control (GLM repeated measures: F(1,19) = 38.90, P < 0.001, Figure 1B). At the end of the study, the mean body mass of mice treated with MCHR1 antagonist was 15.7% lower than for the control animals (t test on final body weight: t = 2.19, df = 19, P = 0.041).
In line with body weight changes, there was a significant effect of time (baseline and day 26 of treatment) on fat mass (F(1,19) = 8.54, P = 0.009) and a strong interaction between treatment group and time (F(1,19) = 32.51, P < 0.001) though the between-group effects on fat mass did not reach a statistically significant level (F(1,19) = 0.99, P = 0.331) again showing that fat mass had declined in the treatment group over time but not in the controls (Figure 1C). After 26 days of MCHR1 antagonist treatment there was a significant decrease in fat mass (Paired t test: t = 5.43, df = 9, P < 0.001) whereas vehicle-treated mice showed no significant change in fat mass (Paired t test: t = -2.23, df = 10, P = 0.05). At baseline, no difference was detected in body fat mass between groups. Following 26 days of MCHR1 antagonist treatment DIO mice had 35.5% less fat mass than vehicle-treated controls (Independent t test: t = 2.33, df = 19, P = 0.031). Moreover, on day 30 when all animals were dissected, MCHR1 antagonist-treated mice exhibited lower total fat mass (t = 2.26, df = 19, P = 0.036) accompanied by less perirenal fat (t = 3.21, df = 19, P = 0.005) and less mesenteric fat (t = 2.47, df = 19, P = 0.023) whereas the weights of gonadal fat and subcutaneous fat did not differ significantly, although both effects were on the borderline of significance (gonadal fat: t = 2.03, df = 19, P = 0.057; subcutaneous fat: t = 1.98, df = 19, P = 0.062; Figure 1D). No differences were detected in lean mass either at baseline or on day 26 of dosing (baseline: t = 0.23, df = 19, P = 0.819; day 26 of dosing; t = 0.30, df = 19, P = 0.770; Figure 1E).
Glucose tolerance and plasma metabolites
On day 27 of dosing, mice from both groups exhibited similar fasted blood glucose levels (Vehicle: 9.44 ± 0.55 (μg mL−1); MCHR1 antagonist: 8.60 ± 0.59 (μg mL−1)). There were no significant differences between vehicle-treated and MCHR1 antagonist-treated groups in glucose tolerance (F(1,19) = 0.36, P = 0.558, Figure 2A). However, age-matched non-DIO mice that were fed with a regular diet exhibited a better tolerance to glucose, suggesting that HFD impaired glucose metabolism (Figure 2A).
Plasma leptin levels were highly associated with body fat mass (R2 = 0.94, F(1,19) = 290.09, P < 0.001, Figure 2B). Despite an absolute difference in leptin (Independent t test: t = 2.48, df = 19, P = 0.022; Figure 2C), chronic MCHR1 antagonist treatment had no significant effect on circulating levels of leptin with the effect of fat mass removed (GLM: using fat mass as a covariate: F(1,18) = 0.96, P = 0.341; using square of fat mass as a covariate: F(1,18) = 0.88, P = 0.36). Circulating TG levels were significantly reduced by chronic MCHR1 antagonist dosing ((t = 2.49, df = 19, P = 0.022, Figure 2D). In contrast, no differences were observed in circulating insulin (t = 1.38, df = 19, P = 0.187; Figure 2E) and NEFA (t = 0.95, df = 19, P = 0.353; Figure 2F).
Food intake and energy budget
Chronic MCHR1 antagonist dosing markedly inhibited daily average food intake in DIO mice throughout the treatment (GLM repeated measures: group, F(1,19) = 5.31, P = 0.033; days of treatment, F(1,19) = 12.72, P < 0.001; days of treatment × group, F(1,19) = 2.49, P = 0.008; Figure 3A). Cumulative food intake was significantly reduced by chronic MCHR1 antagonist administration (GLM repeated measures: group: F = 7.121, P = 0.015, days of treatment, F(1,19) = 1638.26, P < 0.001, days of treatment × group, F(1,19) = 4.559, P < 0.001; Figure 3B). Over the 30-day treatment, accumulated food intake of DIO mice receiving the MCHR1 antagonist was 9.1% less than that of vehicle-treated counterparts (t test: t = 2.14, df = 19, P = 0.046). On day 26 when body fatness was evaluated, the difference in accumulated gross food intake between groups was 6.62 g, equivalent to 138.34 kJ (dry food intake × energy content of dry food). Body fat has an energy content of about 39 kJ g−1, so the difference in mean fat mass on day 26 (5.25 g) was equivalent to 204.75 kJ. The fact that reduced gross food intake did not fully account for fat loss suggests an increase in energy expenditure was also marked.
Indeed, chronic MCHR1 antagonist dosing altered energy budget on day 4-6 of treatment, when mice treated with MCHR1 antagonist showed an increase in DEE (ANCOVA, F(1,18) = 7.52, P = 0.013, Figure 4A) and a significant reduction in MEI (ANCOVA, F(1,18) = 4.90, P = 0.040, Figure 4B). As the treatment progressed, these effects of MCHR1 antagonist on energy budget disappeared on day 22-24 (DEE, F(1,18) = 0.02, P = 0.899; MEI, F(1,18) = 0.14, P = 0.709; Figure 4C,D). When both MEI and DEE were corrected for body mass, contrast tests revealed a 16.4% decrease in MEI (t = 2.13, df = 19, P = 0.027) and a 10.8% increase in DEE (t = -2.77, df = 19, P = 0.012) during day 4-6 of treatment whereas no changes were observed on day 22-24 (MEI: t = 0.61, df = 19, P = 0.548; DEE: t = 0.41, df = 19, P = 0.688) (Figure 4E,F). MCHR1 antagonist treatment had no effects on assimilation efficiency at both time points (day 4-6: t = 0.24, df = 19, P = 0.813; day 22-24: t = -0.31, df = 19, P = 0.759; Figure 4G). No changes in RMR were observed following 19 days of MCHR1 antagonist dosing (GLM repeated measures on corrected RMR: group: F(1,19) = 0.02, P = 0.901; time (baseline and day 19 of treatment), F(1,19) = 0.02, P = 0.902, time × group, F(1,19) = 0.01, P = 0.923; Figure 4H), indicating the enhancement on DEE was attributed to changes in physical activity. Notably, normalizing the DEE data with lean mass instead of total body mass at day 22-24 did not alter the results.
At baseline, mice treated with MCHR1 antagonist and with vehicle showed a circadian rhythm in body temperature with higher temperature during the dark phase and lower temperature during the light phase, and no differences between groups. Chronic MCHR1 antagonist treatment did not alter body temperature throughout the experiment (see Supporting Information S8 for detailed data).
Prior to the treatment, animals showed a circadian rhythm in daily physical activity, with most activity confined to the dark phase and there were no differences in daily activity patterns between groups (GLM repeated measures: group, F = 0.399, P = 0.535; time, F = 22.860, P < 0.001 time × group F = 0.585, P = 0.938; Figure 5A). During the first 4 days of dosing, no changes in activity pattern were observed (GLM repeated measures: group, F = 0.062, P = 0.807; time, F = 23.791, P < 0.001; time × group, F = 1.119, P = 0.320; Figure 5B). In contrast, DIO mice receiving the MCHR1 antagonist showed significantly elevated physical activity levels on day 18-22 of dosing (group, F = 9.531, P = 0.006; time, F = 26.623, P < 0.001; time × group, F = 2.167, P = 0.002; Figure 5C). Furthermore, chronic MCHR1 antagonist administration significantly increased physical activity levels during the dark phase (F = 8.586, P = 0.009; Figure 5D) despite no effects on activity during the light phase (F = 0.119, P = 0.734; Figure 5E) or on averaged daily activity (F = 3.134, P = 0.095; Figure 5F).
We have demonstrated here that chronic oral administration of GW803430, reduced body mass and adiposity in DIO mice fed a high fat diet, but with minimal effects on their glucose homeostasis. Previous work has shown GW803040 inhibited MCHR1 binding in different areas of the brain and exhibits substantial occupancy of the MCHR1 at doses of 1 mg kg−1 or greater (). These authors () further evaluated the effects of the same compound at a dose of 10 mg kg−1 in MCHR1+/+ and MCHR1−/− mice, respectively. The fact that MCHR1−/− mice did not show a decrease in body weight, but MCHR1+/+ mice did, suggested that GW803040 produced significant reductions in body mass, fatness and food intake via MCHR1 antagonism (). Our findings are consistent with previous studies reporting reductions in body fat mass following administration of other MCHR1 antagonists ([15, 21, 27, 28]). The loss of body mass and body fat mainly stemmed from a negative energy balance induced by the MCHR1 antagonist. The MCHR1 antagonist also caused a reduction of circulating leptin levels, but this was secondary to the loss of body fat since the effect of the MCHR1 antagonist on leptin disappeared after statistically accounting for body fat mass. In addition, plasma triglyceride levels were significantly reduced after chronic MCHR1 antagonist dosing. This observation is in agreement with the fact that MCHR1 antagonism is associated with the reduction of hepatic TG accumulation in ovariectomized mice ().
Paradoxically, no improvement in glucose metabolism was observed despite a significant reduction in body weight and fat mass. MCHR1 antagonist treatment recovered the weight gain induced by HFD but had no effect on impaired glucose tolerance. Moreover, chronic MCHR1 antagonist administration did not affect circulating insulin or NEFA. Consistent with this finding, Kowalski et al. also demonstrated that 28 days of MCHR1 antagonist treatment had no effect on circulating insulin levels (). Notably, previous studies showed that the effects of either MCHR1 antagonism or MCH administration on glucose homeostasis were independent of body weight change ([17, 29]). The absence of an effect on glucose homeostasis and lipid profiles however may indicate that while MCHR1 antagonism may improve obesity its impact on the metabolic sequalae of high body fatness may be less profound.
In agreement with the changes in body mass and fatness, the MCHR1 antagonist exerted a persistent anorexic effect on food intake in DIO mice on a high-fat diet. Mashiko et al. reported that a peptidic MCHR1 antagonist suppressed MCH-induced food intake in satiated Sprague-Dawley rat (). Furthermore, chronic 4-week infusion of the antagonist reduced food intake, prevented body weight gain; and also further reduced body weight in these obese mice (). Another study showed that the peptide antagonist blocked the increase in palatable food intake after icv administration of MCH (). The antagonist was more effective when rats were fed a highly palatable diet, suggesting that the reward system is involved in the feeding suppression effect of MCHR1 antagonism. In support of this notion, studies on MCH−/− mice showed that expression of dopamine transporter was significantly elevated in the nucleus accumbens and that evoked dopamine release was significantly increased in the nucleus accumbens shell (). Hypothalamic MCH also modulates nucleus accumbens activity ().
To further understand the mechanisms underlying the weight-reducing effect of MCHR1 antagonist, we evaluated energy budgets of DIO mice at different stages of the treatment. Hu et al. demonstrated that a brain permeable MCH1R antagonist resulted in a significant reduction in body weight and fat mass in DIO mice and that this effect appeared to be as partially driven by the decrease in food intake, which indicated an impact of blockade of MCHR1 signaling on energy expenditure (). Indeed, in our study we found that on day 4-6 of dosing, MCHR1 antagonism significantly inhibited metabolizable energy intake and enhanced daily energy expenditure. Consistent with these observations, Shearman et al. observed that antagonism of MCHR1 in rats for 14 days resulted in a decrease in caloric efficiency (expressed as mg body weight/kcal food intake) indicating an increase in energy expenditure. However, these effects on components from both sides of energy balance disappeared when the treatment progressed to day 22-24. At this stage, the modest surplus of energy expenditure over energy intake might contribute to the maintenance of weight loss. The animals could have adopted compensatory adaptations in the energy budget in response to a sustained weight loss. With regard to energy expenditure, there has been inconsistent evidence showing that acute (4 days) MCHR1 antagonism failed to increase energy expenditure in DIO mice (). It should be noted that in the aforementioned study energy expenditure was measured for 23 h using indirect calorimetry (Oxymax-CLAMS system), whereas daily energy expenditure in the current study was evaluated using the doubly labeled water method. DLW measures are made in the animals' home cage and therefore tend to involve fewer artifacts than 24-h indirect calorimetry measures. The discrepancy in results might be attributed to the different methods.
Because energy expenditure consists of resting metabolic rate (RMR), diet-induced thermogenesis, thermoregulation and physical activity, we specifically measured the effects of MCHR1 antagonism on each component of energy expenditure. RMR was not affected by MCHR1 antagonism after 19 days of treatment and nor was the respiratory quotient, which resembled the findings from DIO mice treated with MCHR1 antagonist for 4 consecutive days (). This implies that MCHR1 antagonism has no effects on RMR regardless of the duration of treatment. In our study, chronic MCHR1 antagonist administration had no effects on body temperature, which is in agreement with findings from Kowalski et al. who failed to detect any change in body temperature during the 6 h following acute MCHR1 antagonist administration (). In contrast, Ito et al. recently demonstrated that MCHR1 antagonist treated mice showed a significantly higher rectal temperature during the month of drug administration (), but these previous data, based on rectal probing to measure body temperature, are more prone to handling artifacts than the present study which employed completely noninvasive measurements from implanted transmitters. On the other hand, there has been evidence in support of a positive effect of MCH antagonism on body temperature. For example, mice centrally infused with MCH for 14 days significantly decreased body temperature (). In addition, Handlon et al. suggested that the effects of MCH antagonists on obese rodents may be mediated via brown adipose tissue (BAT) (), since hypothalamic MCH neurons project to thermogenic brown adipose tissue in the rat (), and there is evidence that MCH neural pathways regulate thermogenesis in BAT (). These currently inconsistent results indicate that further work is required to elucidate the effects of MCHR1 antagonism on body temperature.
It has been suggested that reduced locomotor activity may be a main contributor to diet-induced obesity in mice (). Kowalski et al. reported that the MCHR1 antagonist, SCH-A did not affect locomotor activity in DIO mice during the study period of 6 h after acute dosing (), which was consistent with findings showing that a peptidal MCHR1 antagonist failed to alter motor activity in lean mice () and rats (). Surprisingly, MCHR1−/− mice are hyperphagic and their leanness is a consequence of hyperactivity and altered metabolism. Our physical activity measurements showed that MCHR1 antagonism elevated activity levels during the dark phase throughout the experiment, despite no changes in physical activity during the light phase. The mechanisms underlying the effects of blockade of MCHR1 signaling on activity remain to be determined. It has been suggested that many peripheral and hypothalamic factors regulating appetite also appear to affect locomotor activity (). Later studies revealed that dopamine influences feeding and reward as well as motor activity (), and the MCH1 receptor antagonist SNAP 94847 induces sensitivity to dopamine D2/D3 receptor agonists in rats and mice (), indicating MCHR1 antagonist may play an important role in the regulation of physical activity via the dopaminergic system.
In conclusion, we have demonstrated that chronic administration of an MCHR1 antagonist caused a progressive reduction in body weight and adiposity, where the initiation of weight loss stemmed from both a decrease in energy intake and an increase in energy expenditure. While the weight-reducing effect of MCHR1 antagonist was sustained, the effects on energy budgets diminished as the treatment progressed. In addition, the chronic blockade of MCHR1 markedly elevated physical activity levels during the dark phase in DIO mice, although total daily activity was not significantly increased. The differences in physical activity persisted throughout the 30 day treatment and were perhaps sustained by the weight loss. Despite the effects on body weight and adiposity there was no improvement in blood NEFA profiles or in glucose homeostasis as measured by circulating insulin levels and glucose tolerance tests.
The authors would like to thank Peter Thomson and Paula Redman for their technical help with isotope analysis for DLW. Thanks to Lambertus Benthem and Caroline Wingoff for the formulation and delivery of the MCHR1 antagonist. We thank Yuko Gamo for assistance with surgeries, Lobke Vaan Holt for her advice on statistics and Cathy Wyse for constructive comments on the manuscript. Special thanks to Marianne Älholm Larsen Gr⊘nning, former senior director of Diabetes Research China, Novo Nordisk, for her generous support on the completion of this manuscript.