Genetics and environment contribute to the development of obesity, in both humans and rodents. However, the potential interaction between genes important in energy balance, strain background, and dietary environment has been only minimally explored. We investigated the effects of genetic ablation of melanin-concentrating hormone (MCH), a neuropeptide with a key role in energy balance, with chow and a high-fat diet (HFD) in two different mouse strains, one obesity-prone (C57BL/6) and the other obesity-resistant (129). Substantial differences were seen in wild-type (WT) animals of different strains. 129 animals had significantly lower levels of spontaneous locomotor activity than C57BL/6; however, 129 mice gained less weight on both chow and HFD. In both strains, deletion of MCH led to attenuated weight gain compared with WT counterparts, an effect secondary to increased energy expenditure. In both strains, feeding a HFD led to further increases in energy expenditure in both WT and MCH-KO mice; however, this increase was more pronounced in 129 mice. In addition, mice lacking MCH have a phenotype of increased locomotor activity, an effect also seen in both strains. The relative increase in activity in MCH−/− mice is modest in animals fed chow but increases substantially when animals are placed on HFD. These studies reinforce the important role of MCH in energy homeostasis and indicate that MCH is a plausible target for antiobesity therapy.
- melanin-concentrating hormone
- diet-induced obesity
- locomotor activity
- energy expenditure
obesity results from an interaction of genes and environment that leads to an imbalance of energy intake over energy expenditure (7, 14, 15, 27). A number of peptides synthesized both in the periphery (2, 9, 25) and in the brain (10, 13, 35) play an important role in regulating these processes. Among these peptides, the hypothalamic peptide, melanin-concentrating hormone (MCH), has emerged as a key regulator of energy balance (14, 20, 28, 35). When administered via intracerebroventricular injection, MCH causes an acute robust increase in feeding behavior (31, 32). Chronic infusion leads to excess weight gain and a shift of the animal to a “lipogenic” state associated with decreased energy expenditure and increased energy storage (18). Overexpression of MCH gene leads to mild obesity and increases susceptibility to high-fat diet (HFD)-induced obesity (21). Ablation of both the MCH gene (MCH−/−) (35) and the rodent receptor for MCH, designated MCHR-1, leads to a lean phenotype (6, 23). Furthermore, combined MCH deficiency and leptin deficiency, created by crossing ob/ob to MCH−/− mice, leads to a significant attenuation of the obese phenotype in ob/ob mice (34). Such “double null” mice remain hyperphagic, and weight loss results from increased energy expenditure secondary to both increased locomotor activity and increased resting metabolic rate.
Strain background plays an important role in determining the observed phenotype of specific mouse mutations (1, 24). Strain background also plays an important role in the development of diet-induced obesity (DIO) and susceptibility to type 2 diabetes (17, 33). To further evaluate the role of MCH in regulating energy homeostasis, we inbred mice onto both C57BL/6 and 129SvJ (129) genetic backgrounds. C57BL/6 mice are susceptible to DIO and develop glucose intolerance when fed HFD, while 129 mice are relatively resistant to weight gain, although there is less data available for this strain (3, 36, 37). Mice of mixed background lacking the MCH gene were backcrossed onto the strain of interest for 10 generations. Cohorts of male mice were fed chow or HFD for 12 wk. Food intake, locomotor activity, energy expenditure, assessed both as oxygen consumption (V̇o2) and heat production were compared. Comparison of wild-type (WT) strains revealed a substantial difference in weight gain on both diets; compared with C57BL/6, 129 animals gained less weight on either diet despite the fact that 129 mice had less than one-third the level of locomotor activity seen in C57BL/6 mice. Additional differences emerged when mice lacking MCH were evaluated. MCH−/− mice fed chow were leaner and gained less weight compared with WT controls. In 129 animals, this was seen in the context of increased food consumption in mice lacking MCH. When placed on HFD, an attenuation of weight gain was seen in C57BL/6 mice, as MCH−/− of this strain gained half as much weight as did control animals. On the 129 strain, MCH−/− mice gained no excess weight. Increased energy expenditure contributed to a lesser weight gain in MCH−/− mice; the effect was more apparent in the 129 strain.
These results further highlight the importance of MCH in regulating energy homeostasis and demonstrate the importance of genetic background and environmental factors such as mouse strains and diet in determining body weight.
MATERIALS AND METHODS
The generation of mice lacking the gene for prepro-MCH has been previously described (35). For the current studies, MCH−/− mice on a mixed 129 and C57BL/6 background were backcrossed onto 129 or C57BL/6 strains for at least 10 generations. The absence of MCH was confirmed by PCR amplification of genomic DNA. Mice were housed in the Joslin Animal Facility, maintained at 22°C, under an alternating 12:12-h light-dark cycle. The standard chow used in these studies was Purina Formulab diet 5008 (3.5 kcal/g), which contains 6.5% fat by weight and provides 16.7% total calories from fat. The HFD that we used was Research Diets D12451 (4.7 kcal/g), which contains 24% fat by weight and provides 45% calories from fat (Research Diets, New Brunswick, NJ). Weight of the animals and food intake were measured weekly. Blood for plasma leptin and insulin was collected at 3:00 PM (fed state) by tail bleeding. A total of three independent cohorts of C57BL/6 male mice was studied. Each cohort consisted of four groups of 4–6 animals each: WT on chow (WC; see figures), WT on HFD (WF), MCH−/− on chow (KC), and MCH−/− on HFD (KF). The same results with regard to weight gain and metabolic and physiological changes were observed in all three cohorts. For the 129 strain, only one cohort of WT and MCH−/− male animals was studied. In a separate study, the effect of HFD on leanness was reevaluated in female cohort, and the same results were seen as those observed in male animals. All procedures were approved by Joslin Diabetes Center Institutional Animal Care and Use Committee.
Glucose tolerance tests.
Intraperitoneal glucose tolerance test (IPGTT) was performed at the age of 18 wk. Mice were fasted overnight (1700–0800) and were subsequently injected with glucose (2 g/kg body wt ip). Tail blood was collected at −5, 15, 30, 60, and 120 min. A total volume of blood taken during IPGTT is less than 50 μl. Blood glucose concentrations were measured using a glucometer (Elite, Bayer, Mishawaka, IN).
V̇o2 was measured using, an eight-chamber open-circuit Oxymax system that is a component of the Comprehensive Laboratory Animal Monitoring System (CLAMS; Columbus Instruments, Columbus, OH). Mice were housed individually in specially built Plexiglass cages (5” × 4.5” × 8.5”, height × width × depth) maintained at ∼22°C under an alternating 12:12-h light-dark cycle (light period 0800–2000). Sample air was sequentially passed through oxygen (O2) and carbon dioxide (CO2) sensors (Columbus Instruments) for determination of O2 and CO2 content. Mice were acclimatized to monitoring cages for 24 h before data collection. Mice were weighed before each trial. Heat production on per animal basis was calculated from the following equation: (3.82 + 1.23 × RER) × V̇o2, where RER is the respiratory exchange ratio (volume of CO2 produced/volume of O2 consumed per hour) (5).
Resting energy expenditure.
Resting energy expenditure (REE) was determined by defining periods of inactivity and measuring V̇o2 during those periods. In C57BL/6 animals, intervals during which animals showed less than 50 beam breaks were considered inactive intervals. In 129 mice, intervals with less than 40 beam breaks were considered inactive.
Ambulatory activity of individually housed mice was evaluated on a relative, not absolute, basis using an eight-cage rack OPTO-M3 Sensor system (Columbus Instruments, Columbus, OH). Consecutive photobeam breaks occurring in adjacent photobeams were scored as an ambulatory movement. Cumulative ambulatory activity counts were recorded every 30 min throughout the light and dark cycles.
Body fat measurement.
Fat and lean body mass were assessed using a Dual Energy X-ray Absorptiometry (DEXA; Lunar PIXImus2 mouse densitometer, GE Medical Systems, Madison, WI), as described by the manufacturer and validated by others (4). Mice were anesthetized by intraperitoneal injection of a (1:1) mixture of tribromoethanol and tert-amyl alcohol (0.015 ml/g body wt) and scanned; total body fat and lean body mass were then determined using the analysis program provided by the manufacturer.
Values are reported as group means ± SE. Interactions between genotype (KO vs. WT) and diet (chow vs. HFD) were analyzed by two-way repeated-measures ANOVA when appropriate. One-way ANOVA and independent t-test were also used as appropriate. A P value of less than <0.05 was considered statistically significant. Statistical comparisons were made using Statview (Abacus Concepts, Berkeley, CA) software.
Weight and body composition in C57BL/6.
As previously reported (35), in mice with a mixed genetic background, C57BL/6 mice lacking the MCH gene were leaner than WT animals (Fig. 1A ; Table 1). MCH−/− mice gained less weight and accumulated less body fat when placed on HFD than WT animals (Fig. 1C). WT animals on chow gained 5.6 ± 0.3 g, whereas MCH−/− gained only 2.2 ± 0.7 g (P = 0.0003). WT mice fed HFD gained significantly more weight compared with MCH−/− mice on the same diet (13 ± 1.7 vs. 5.9 ± 1.2 g, P = 0.0007). Attenuated weight gain of HFD-fed MCH−/− animals was associated with a reduction in fat mass, which was 41.3% lower (8.4 ± 1.2 vs. 14.3 ± 1.9 grams, P = 0.018) (Table 1). The study was repeated in two additional cohorts, and similar results were obtained with each cohort.
Weight and body composition in 129 mice.
MCH−/− mice on 129 background were leaner than WT animals (Fig. 1). MCH−/− mice gained only 0.4 ± 0.7 g on chow, whereas WT animals on the same diet gained 1.4 ± 1.2 g (P = 0.019) (Fig. 1B; Table 1). WT mice on HFD gained more than four times as much weight as MCH−/− mice on the same diet over this time interval (4.9 ± 0.9 vs. 1.4 ± 1.2, P = 0.01) (Fig. 1D; Table 1). WT mice fed HFD had an almost 50% higher fat mass compared with MCH−/− mice on the same diet (10.1 ± 1.3 vs. 5.6 ± 0.1 g, P = 0.01).
Leptin and insulin levels.
There was no difference in leptin levels between WT and MCH−/− C57BL/6 mice on chow (Table 1). On HFD, WT mice had almost five times higher leptin levels compared with MCH−/− mice (38.4 ± 14.5 vs. 8.6 ± 2). Although 129 mice gained substantially less weight on HFD diet than C57BL/6, a significant increase in leptin levels was still seen. Thus leptin levels in 129 WT mice on HFD were significantly higher than MCH−/− on the same diet (18.5 ± 6.4 vs. 5.2 ± 3.2 ng/ml, P = 0.042). Insulin levels in WT C57BL/6 mice on HFD were significantly higher than all other groups. In 129 mice, there was no significant difference in insulin levels between groups.
In C57BL/6 mice, there was no difference in food intake (Kcal/day) between MCH−/− and WT in either diet. On a 129 background, MCH−/− mice on chow diet ate more than WT on the same diet (19.9 ± 1.1 vs. 15.1 ± 0.4). When they were fed a HFD, they ate the same amount (Table 1).
In C57BL/6 mice fed chow, MCH−/− mice had a small but significant increase in locomotor activity both during the dark cycle [1,333 ± 99 vs. 1,035 ± 90 average beam breaks (ABB) per 30 min, P < 0.001] and the 24-h period (1,085 ± 74 vs. 951 ± 72 ABB/30 min, P = 0.026) compared with WT mice on the same diet (Fig. 2A; Table 2). When animals were placed on HFD, activity of the C57BL/6 MCH−/− animals increased; MCH−/− mice had almost twofold higher locomotor activity compared with WT animals on HFD both during the dark cycle (1,402 ± 96 vs. 824 ± 72 ABB/30 min, P < 0.0001) and the 24-h period (1,268 ± 74 vs. 654 ± 52 ABB/30 min, P < 0.0001) (Fig. 2B; Tables 2 and 3).
Activity levels of 129 WT animals on chow were remarkably lower than those seen in C57BL/6 WT animals on chow (951 ± 72 vs. 167 ± 17 ABB/30 min over 24 h, P < 0.0001) (Fig. 2C; Tables 2, 3). Mice lacking MCH had twofold higher locomotor activity compared with WT counterpart (357 ± 25 vs. 167 ± 17 ABB/30 min over 24 h, P < 0.0001) when fed chow. On HFD, both WT and MCH−/− mice showed further increased activity, although the relative differences remained the same (537 ± 55 vs. 255 ± 23 ABB/30 min over 24 h, P < 0.0001) (Fig. 2D; Tables 2, 3).
C57BL/6 MCH−/− fed a chow diet had significantly higher V̇o2 than WT animals (3,556 ± 60 vs. 3,081 ± 63 mg·kg−1·h−1, P < 0.0001) (Fig. 3A; Tables 2, 3). When animals were fed HFD, only MCH−/− mice further increased their V̇o2 (from 3,556 ± 60 to 4,115 ± 84 ml·kg−1·h−1). WT C57BL/6 mice on HFD diet fail to increase V̇o2 compared with WT mice on chow (Fig. 3B; Tables 2, 3).
In the 129 strain that were fed chow, MCH−/− mice had 34% higher V̇o2 compared with WT animals (3,551 ± 56 vs. 2,647 ± 55 ml·kg−1·h−1, P < 0.0001) (Fig. 3C; Tables 2, 3). When placed on HFD, a similar pattern observed in C57BL/6 was seen in 129 mice. MCH−/− mice had a 39.2% higher V̇o2 (4,187 ± 64 vs. 3,008 ± 32 ml·kg−1·h−1, P < 0.0001) during a 24-h period than that of WT counterparts on the same diet (Fig. 3D; Tables 2, 3). REE accounts for the bulk of energy expenditure, that is, 75–80%. In all cases changes in average total V̇o2 were paralleled by similar changes in average REE.
Energy expenditure was also analyzed by examining Kcal produced per hour per animal. Chow fed MCH−/− and WT mice on a C57BL/6 background had similar heat production per animal. (Fig. 4A; Tables 2, 3). In both groups, heat production increased when animals were placed on HFD (Fig. 4B; Tables 2, 3).
In 129 animals fed chow, MCH−/− mice had a 50% higher heat production (0.52 ± .008 vs. 0.35 ± .007 Kcal/h, P < 0.0001) compared with WT counterparts (Fig. 4C; Tables 2, 3). This difference persisted when animals were placed on HFD. MCH−/− animals showed significantly higher heat production compared with WT mice on the same diet (0.59 ± 0.009 vs. .46 ± .005 Kcal/h, P < 0.0001) (Fig. 4D; Tables 2, 3).
Glucose tolerance tests.
In C57BL/6 animals, fasting glucose levels were the same in all cohorts. MCH−/− animals on chow trended toward lower glucose levels throughout 120 min compared with WT animals, but this difference did not reach statistical significance AUC: 24,436 ± 1,980 vs. 26,883 ± 2,530 mg/dl for 120 min, P = 0.11] (Fig. 5, A and C). Glucose tolerance worsened in both WT and MCH−/− animals fed HFD. There was no significant difference in glucose tolerance between MCH−/− and WT mice on HFD diet (36,322 ± 3,910 vs. 40,448 ± 3,350 mg/dl for 120 min, P = 0.20).
In 129 animals, there was no difference in glucose AUC between MCH−/− and WT animals on chow (14,521 ± 1,771 vs. 16,329 ± 2,797 mg/dl for 120 min, P = 0.53) (Fig. 5, B and D). When 129 mice were fed HFD, glucose tolerance worsened in both WT and MCH−/− mice. However, MCH−/− that were fed HFD had significantly better glucose tolerance than WT mice on HFD (21,435 ± 1,448 vs. 28,020 ± 1,342 mg/dl for 120 min, P = 0.027) (Fig. 5, B and D).
The results reported in this study illustrate the interaction of multiple factors in determining body weight. Thus the phenotype of mice lacking MCH is partially dependent on the strain of mice examined (Table 3). Initially, on a mixed background, leanness was secondary to hypophagia combined with a small increase in energy expenditure (35). Upon backcross to the 129 and C57BL/6 strains, the lean phenotype persists and is maintained through increased energy expenditure, as hypophagia is no longer a feature of the phenotype. Leanness associated with the absence of the MCH gene is more apparent in C57BL/6 animals than in 129 animals. This is in part due to basic differences between strains. Upon reaching maturity, C57BL/6 animals fed chow continue to gain weight, while 129 animals gain very little weight. Over the 12-wk observation period C57BL/6 fed chow gained an average of almost 6 g, while 129 animals gained only 1.4 g. Against a background of higher weight gain of C57BL/6, it is easier to observe the MCH effect; C57BL/6 MCH−/− mice gain only 2.2 g (compared to 6 g in WT). Although 129 MCH−/− mice gain 0.43 compared with the 1.43-g weight gain of WT mice, the small differences were not statistically significant.
The strain differences among WT mice in weight gain and energy expenditure are intrinsically noteworthy. Although 129 animals have slightly higher body weights at 8 wk of age, their subsequent weight gain is significantly less than that seen in C57BL/6 animals. This occurs in a context of relative hypoactivity of 129 mice, which over a 24-h period of observation had 20% of the locomotor activity seen in C57BL/6 mice. It is interesting that this difference in locomotor activity had little effect on V̇o2; VO2 of 129 mice on chow was only 15% lower than that of C57BL/6. Additional and substantial differences between strains emerge when animals are placed on HFD. The initial weight of 129 and C57BL/6 animals differs by only 1 g. After 12 wk on HFD, C57BL/6 animals gained 13 g while 129 mice gained only 5 g. The difference in weight gain appears to be attributable to compensations in energy expenditure made by 129 animals. WT 129 animals show a 50% increase in locomotor activity, as well as a 14% increase V̇o2 when fed HFD. In contrast, in C57BL/6 mice fed HFD, locomotor activity decreased significantly over a 24-h period, and there was no change in V̇o2.
These differences between mouse strains highlight the issue of strain choices in examining the effects of a single gene mutation in determining susceptibility to obesity and type 2 diabetes. The importance of genetic background in investigating the effects of single-gene mutation has been recently been reviewed and discussed (1, 24). For example, leptin-deficient mice backcrossed to the BALB/cJ genetic background have reduced adiposity, enhanced fertility, normal temperature, as well as improved insulin sensitivity, compared with leptin-deficient mice backcrossed to the C57BL/6 mice (30). In another recent report, leptin-deficient mice backcrossed onto FVB/N background had hyperglycemia and more severe insulin resistance compared with leptin-deficient mice on C57BL/6 background (16). These studies suggest the presence of genetic interactions between a mutation and the genetic background in developing obesity and type 2 diabetes. At present, there is no consensus on what might be ideal strains for metabolic studies. Given differences in DIO susceptibility, maintaining mutation on both 129 and C57BL/6 congenic strains is a valid choice.
Thus we further investigated the effects of MCH ablation on DIO on both strains and found that MCH has an additional effect on the physiological response to increased fat in the diet. In the absence of MCH, HFD leads to further increases in energy expenditure in both strains. MCH−/− C57BL/6 animals fed HFD have a 15.7% increase in V̇o2 compared with MCH−/− C57BL/6 fed chow. Although overall activity of chow fed MCH−/− mice is only slightly increased compared with WT, the difference is accentuated when animals are fed HFD; MCH−/− mice become twice as active as WT C57BL/6. In 129 mice, and similar differences in V̇o2 are seen. MCH−/− animals fed HFD have higher VO2 than WT animals on the same diet. As with C57BL/6, 129 MCH−/− mice also show relative increase in activity when fed HFD; however, WT 129 mice also increased activity when fed HFD.
When calculating V̇o2, we corrected values to the weight of the animals. This is a common correction (11, 22, 26) but has been criticized by some (12) as giving undue weight to adipose tissue and thereby creating artifactual differences in energy expenditure. We therefore also examined heat production, the calculation of which is independent of total body weight. Heat production is proportional to body mass, and animals with the same mass and same metabolic rate should have the same heat production. Thus in C57BL/6 animals, the finding of the same levels of heat production in leaner MCH−/− mice compared with WT animals confirms that MCH−/− animals are more metabolically active. In 129 mice, the difference in heat production was more pronounced and MCH−/− animals had ∼17% higher levels of heat production, despite their smaller size.
Interestingly, the changes in energy expenditure are sufficient to attenuate weight gain in both strains. C57BL/6 mice lacking MCH ate approximately the same number of calories compared with WT mice, while 129 mice ate more than WT when fed chow. The molecular basis underlying the higher energy expenditure among MCH−/− mice in either strain is undefined. We have previously postulated that MCH may negatively regulate sympathetic outflow, based on our findings in MCH and ob/ob double-null model (34). In ob/ob mice, known to have decreased sympathetic activity (38), MCH ablation corrects a number of defects, including core hypothermia and intolerance to cold exposure associated with the leptin-deficient state (34). In the absence of MCH, the resulting increase in energy expenditure and resistance to DIO reported in the current study would be consistent with increased sympathetic activity.
Increased sympathetic activity would not, however, explain the increased locomotor activity observed in mice lacking MCH, particularly, those fed HFD. Locomotion is typically regulated by the dopamine system, particularly the nucleus accumbens. MCH neurons are known to project from the lateral hypothalamus to both accumbens and caudate putamen, and both of these areas are rich in expression of MCHR-1 (19, 29). It is possible that MCH plays a role in regulating dopaminergic systems and that mice without MCH have altered dopaminergic tone. Such differences may also explain the increased activity in mice lacking MCHR-1, as well as in increased locomotor activity in ob/ob MCH-deficient mice.
Other metabolic parameters evaluated in these mice generally correlated with the degree of weight gain and body fat (8). In WT C57BL/6 animals fed HFD, leptin levels rose to almost 10-fold higher levels than animals fed chow. In WT 129 animals, HFD feeding was associated with a fourfold rise in leptin levels. Leptin levels in MCH−/− mice fed chow trended lower than leptin levels of WT animals fed chow; however, this difference was not statistically significant. Although leptin levels rose in MCH−/− animals of both strains fed HFD, the rise was blunted, reflecting attenuated weight gain. When leptin levels were expressed per gram body fat, MCH−/− mice on HFD in both C57BL/6 and 129 backgrounds have significantly lower leptin per gram body fat compared with WT controls, possibly suggesting increased sensitivity to leptin in MCH−/− mice compared with WT mice.
These results confirm the importance of MCH as an important regulator of energy homeostasis. Thus ablation of MCH is associated with leanness and relative resistance to DIO, and this result was confirmed in two different strains of mice. Despite the effect of exogenously administered MCH to stimulate appetite, leanness secondary to genetic ablation occurs as a result of increased energy expenditure, rather than decreased feeding. Furthermore, these results underscore basic and potentially profound physiological strain differences that may contribute to the observed phenotype resulting from both genetic and environmental manipulations.
This work was supported by grants from the National Institutes of Health Grant DK-56116 (Program Director Jeffrey Flier and Principal Investigator E. Maratos-Flier), DK-56113 and DK-53978 to E. Maratos-Flier. In addition, a portion of the work was supported by Grant from Yamanouchi Foundation. Justin Y. Jeon was supported by Natural Science and Engineering Council of Canada Postdoctoral Fellowship Award. CLAMS analysis and DEXA scanning were supported by Animal Physiology Core of the Joslin Diabetes Center and Endocrinology Research Center Grant DK-36836–16 (Directly to E. Maratos-Flier).
↵* E. Kokkotou and J. Y. Jeon contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society