| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215
Correspondence: Address all correspondence and requests for reprints to: Eleftheria Maratos-Flier, Division of Endocrinology, Department of Medicine, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts 02215. E-mail: emaratos{at}bidmc.harvard.edu
 |
Abstract |
|---|
|
I. Historical Perspective |
|---|
In fish, MCH is expressed in two groups of neurons (recently reviewed by Kawauchi and Baker in Ref. 7). One population of large MCH neurons is located in the nucleus tuberis, the fish homolog of the arcuate nucleus (8), and sends projections mainly to the neurohypophysis (posterior pituitary), although some projections are also sent to the pars distalis (anterior pituitary) as well as to the forebrain. Another group of MCH-positive neurons is found in the dorsal hypothalamus. This dorsal population is more abundant in primitive fish and is less prominent in teleost fish; these neurons are smaller and project to brain and not to pituitary. It is believed that projections to the pars distalis may play a role in regulating pituitary function, whereas the projections to the neurohypophysis are important in the regulation of skin color.
Subsequent to its discovery in fish, MCH was identified in the mammalian hypothalamus as part of a peptide discovery effort aimed at characterizing constituents of rat hypothalamus (9). The mammalian form of MCH is a cyclic 19-amino- acid neuropeptide that is highly homologous to salmon MCH. The N terminus contains two additional amino acids and two amino acid substitutions; there are two additional conserved amino acid substitutions, one in the ring structure and one near the C terminus (Fig. 1
).
|
, A and B). This neuronal population has characteristics similar to the population of neurons in the lateral hypothalamus that expresses orexin, an important regulator of arousal. Both populations of neurons are unusual, in that they make monosynaptic connections throughout the neuraxis, including the prefrontal cortex, the dorsal and ventral striatum, the piriform cortex, as well as to nuclei in the hindbrain, such as the nucleus tractus solitarius and the parabrachial nucleus. However, the MCH and orexin populations are distinct, and there is no overlap of expression.
|
More recently, accumulating evidence from physiological, genetic, and pharmacological studies indicates that MCH plays a critical role in energy homeostasis and is involved in the regulation of both feeding behavior and energy expenditure.
|
II. MCH Genes and Pro-MCH Peptides |
|---|
In addition to several species of trout (20) and salmon (21), MCH cDNA has been cloned from several genera of fish, including flounder (22) and tilapia (23). As shown in Fig. 3, homology between salmon and other fish is not particularly high. At the peptide sequence of the preprohormone, homology between chinook and coho salmon is 90%, and homology between chinook salmon and trout is also 90%. In contrast, homology between chinook and tilapia is only 61%. Similarly, homology between chinook and flounder is 59%. However, the peptide sequence of MCH is identical among these species.
|
Identification of the MCH gene in fish made it possible to clone MCH in mammals. Thus, rat and human mRNAs were identified within 1 or 2 yr (25, 26, 27), followed by the mouse mRNA (28). Furthermore, mRNA analysis revealed that the transcript encodes for two other potential products, neuropeptide E-I (NEI) and neuropeptide G-E (NGE) (25). In humans, rats, and mice, the gene contains three exons and two introns (Fig. 4). It is localized on the long arm of the human chromosome 12, the long arm of chromosome 7 in the rat, and on chromosome 10 in the mouse. There are untranslated regions at both the 5' end and the 3' end.
|
). Two different sources have reported two different amino acids at position 113, either alanine or aspartic acid, within the sequence of peptide NGE. The 165-amino-acid preprohormone is present in all mammalian species examined, including mice, rats, dogs, and cows. Across these different species, the degree of homology is quite high (Fig. 5). There is 91% amino acid identity and 95% homology between dog and human prepro-MCH. Homology between human and cow is almost as high (89 and 91%), and there is 80% amino acid identity and 90% homology between rat and human prepro-MCH. Partial cDNA sequence is available for the hormone from hogs and sheep (Fig. 5). In both of these species, the sequences are less homologous to the human than to the dog sequence.
|
The role of the two other peptides encoded by pro-MCH remains somewhat obscure. Limited experiments suggest that NEI is present in the same nerve terminals as MCH and injections of this neuropeptide produce behavioral changes in rats (see Section VI.B). There is no experimental evidence that NGE is a functional peptide in the central nervous system. Injections of synthetic NEI and NGE into the rat lateral ventricle, alone or combined with MCH, did not produce any effects on feeding behavior (29). Recently, two additional products of the MCH gene have been described. Translation of an alternative frame produces two peptides named MCH gene overprinted polypeptide (MGOP)-14 and MGOP-27 (30). The second product is transcribed from the antisense strand partially overlapping with the MCH gene and is named antisense RNA overlapping MCH (AROM) gene. This gene produces multiple products through alternative splicing, but the coding sequence of the AROM peptide does not overlap with MCH cDNA (31). The function of these novel transcripts from the MCH gene locus is unknown at the present time.
In humans, the authentic pMCH gene is reduplicated into two additional MCH-like genes on chromosome 5, pMCHL1 at 5p14 and pMCHL2 at 5q13. These genes have 92–95% homology with the authentic gene, but lack exon I. pMCHL1 produces several sense and antisense RNAs that are expressed in developing human brain (32, 33). pMCHL2 does not appear to yield any transcripts. Although at this time, the precise functional nature of these MCH-like genes is unknown, the genes have been useful in tracing primate brain evolution (34).
|
III. MCH and Energy Balance |
|---|
In another study, MCH was injected into the fourth ventricle of rats to target the nucleus tractus solitarius in the hindbrain. No effect on feeding was seen. However, there was a significant decrease in core body temperature and a trend toward decreased physical activity, suggesting that MCH acts on the hindbrain in a manner consistent with energy conservation (39).
In rats, MCH also increases water intake independent of food intake (40). ICV injections may also increase both alcohol intake (41) and sucrose intake (41, 42). These effects of MCH may be related, at least in part, to its effects on reward systems.
B. Genetic studies
Additional data about the role of MCH are derived from genetic studies. Knockout of the MCH gene produces a lean mouse that has both decreased feeding and increased resting energy expenditure (43). Mice lacking MCH weigh 25–30% less than control littermates, secondary to a reduction in body fat, and have low circulating leptin levels. Because leptin acts to suppress expression of orexigenic peptides such as neuropeptide Y (NPY) and agouti-related polypeptide (AgRP) and to increase expression of proopiomelanocortin (POMC; the precursor of 
-MSH) (12, 44, 45, 46, 47), one might predict altered expression of these neuropeptides in the hypothalamus of MCH-knockout mice, in a direction consistent with lowered leptin levels. However, although expression of POMC is reduced, expression of both AgRP and NPY is similar to that seen in normal animals.
On the basis of the high levels of MCH seen in ob/ob animals, coupled with the pharmacological effect of MCH to increase food intake, one would expect that leptin-deficient ob/ob mice also lacking MCH might be lean and that leanness would result from a decrease in food consumption. This hypothesis was evaluated in a study where mice lacking MCH were backcrossed onto a C57BL/6 background for six generations and then crossed with leptin-deficient mice of the same genetic background. Evaluation of these "double null" animals revealed that while still obese, they were markedly leaner than ob/ob animals (48). However, the attenuation of the obese phenotype that was seen was secondary to increased energy expenditure rather than to decreased feeding. The increase in energy expenditure was a result of both a 3-fold increase in locomotor activity and a 25% increase in metabolic rate (48). In terms of feeding, ob/ob mice also deficient in MCH ate the same amount of food as ob/ob mice with an intact MCH gene.
Additional insights into the role of MCH in regulating energy expenditure are derived from studies of MCH ablation in two different background strains. Mice lacking MCH were crossed for multiple generations onto either a C57BL/6 background or a 129/SvEv background. These two strains were chosen because the C57BL/6 strain is known to be quite susceptible to rodent obesity caused by high-fat/high-sucrose diets, whereas the 129/SvEv strain is relatively resistant to diet-induced obesity (DIO). Both C57BL/6 and 129/SvEv animals lacking MCH are lean relative to control animals with an intact MCH gene. In neither case is hypophagia a feature; in fact, 129/SvEv mice lacking MCH eat more than wild-type littermates. In both strains, increased energy expenditure appears to account for leanness. This difference in energy expenditure becomes more pronounced when animals are placed on a high-fat diet, which results in further increases in activity and metabolic rate in mice lacking MCH compared with control mice. The magnitude of the effect is different in the two strains; 129/SvEv mice lacking MCH resist DIO entirely, whereas C57BL/6 animals show significantly reduced weight gain (125).
The lean phenotype of mice lacking MCH persists out to at least 19 months of age. Aged MCH–/– male mice are 30% leaner and females are 25% leaner than their wild-type counterparts, an effect due entirely to decreased adipose mass. Furthermore, aged mice lacking MCH have improved glucose tolerance secondary to being more insulin sensitive than control mice (29). Plasma insulin levels during glucose tolerance tests are significantly lower in aged MCH–/– mice compared with wild-type animals. In addition, aging-associated decreases in locomotor activity and resting energy expenditure are attenuated in MCH–/– mice compared with wild-type animals. These data suggest that hypothalamic leanness may lead to an amelioration of the processes associated with aging.
Recently, a mouse model in which MCH neurons are ablated was developed using a genetic construct that mediates expression of the toxin, ataxin-3, in target MCH neurons (49). Mice expressing this gene have a time-dependent loss of MCH neurons, which begins at 4 wk of age and leads to loss of about two thirds of all MCH neurons at 15 wk. These mice show a divergence in weight at 7 wk of age, when ataxin-3 MCH mice become leaner than wild-type animals. The phenotype of these mice, lacking most MCH neurons as adults, is remarkably similar to the phenotype of mice lacking only the peptide. Leanness in ataxin-3/MCH mice may be due to a slight reduction in food intake; however, energy expenditure also contributes significantly, because ataxin-3/MCH mice show increased oxygen consumption. Furthermore, crossing the ataxin-3/MCH mouse to ob/ob mice leads to improved obesity and glucose tolerance.
The finding that both ablation of the MCH gene from MCH neurons and loss of MCH neurons themselves lead to a remarkably similar phenotype is of significant interest. This result indicates that the peptide itself is the critical mediator of signals from the MCH neuronal population that regulates energy balance. This is in contrast to comparable experiments performed with two other orexigenic peptides from the arcuate, NPY and AgRP. These peptides are coexpressed in a population of arcuate neurons that is suppressed by leptin. Ablation of either peptide individually, or both peptides, is not associated with a distinct phenotype (50). In contrast, ablation of the neuronal population using a toxin-based approach leads to a profound hypophagia (51, 52). In the case of AgRP/NPY, the peptides do not appear to be as critically important as the neurons in regulating energy balance.
Although MCH ablation leads to leanness, eutopic overexpression of an MCH transgene (MCH-OE) leads to increased susceptibility to obesity (53). On an FVB background, mice homozygous for the transgene overexpressing MCH have enhanced susceptibility to developing obesity when on a high-fat diet, compared with control animals on the same diet. MCH-OE mice gain 3–4 g of excess weight compared with control animals. MCH-OE mice also show significant hyperinsulinemia and islet cell hypertrophy, which are out of proportion to the mild obesity. The phenotype of overexpression is quite similar to that of chronic infusion, which has been shown to lead to hyperphagia and obesity in both rats (54) and mice (55). Mice infused with MCH also gain weight when pair fed to control animals, indicating that MCH causes alterations in energy balance beyond feeding. However, despite becoming obese, insulin levels in mice infused with MCH are not substantially higher than those seen in control animals.
Ablation of the rodent MCH receptor, MCHR1, also produces a lean phenotype. This phenotype is similar, but not identical, to that seen with ablation of the pMCH gene. Mice lacking MCHR1 are lean, with decreased fat mass, but surprisingly are hyperphagic compared with control littermates, when fed chow. When placed on a high-fat diet, they eat as much as control animals but gain substantially less weight. Mice lacking MCHR1 also develop osteoporosis and exhibit evidence of increased sympathetic activity, although the precise pathways mediating these effects are unknown.
Interestingly, a recent report indicates that the phenotype of ob/ob mice lacking MCHR1 (56) is somewhat different from that reported in mice lacking MCH (48) or MCH neurons (49). For example, mice lacking leptin and MCHR1 (receptor double null) weigh the same as ob/ob mice and have no difference in overall energy expenditure. However, they have an increase in total activity as seen in mice lacking leptin and pMCH (ligand double null). Also similar to ligand double null animals, receptor double null animals have lower circulating corticosterone levels, higher core temperature, and improved cold tolerance. However, in contrast to the ligand double null animals, there is no decrease in hepatic stearoyl-CoA desaturase-1 expression and no increase in brown adipose tissue uncoupling protein-1 protein expression.
These differences might be explained by the fact that ligand double null animals and ob/ob animals with ablation of MCH neurons lack MCH and also lack NEI and NGE, the two other neuropeptides encoded by the pMCH gene. The phenotype of these two animals is quite similar. In contrast, mice lacking the MCH receptor lack only signals mediated by MCH. Further studies will be necessary to reconcile these differences.
C. MCH and energy balance in fish
A few studies have addressed the potential role of MCH on feeding and weight in fish. In one study in medaka fish, a transgenic construct was used to drive MCH overexpression using a cytomegalovirus promoter (57). MCH-OE fish could be visually distinguished by their light body color. Microscopic examination of the scales revealed concentrated melanosomes. An increase in circulating MCH was demonstrated using a fish scale-based bioassay. However, no changes in body weight, size, or reproductive capacity were seen. In another study, barfin flounder were raised in a white environment, which is known to increase MCH expression and MCH serum levels. These fish showed very light skin pigmentation and increased somatic growth at 5 months compared with fish reared in a light environment (22, 58). In another feeding study performed in goldfish, administration of MCH resulted in a substantial inhibition of food intake (59). Interpretation of these studies is limited by the fact that they are few in number, are performed in different species of fish, and use extremely different techniques. At this point, the role, if any, of MCH in fish energy balance is unclear.
D. Interactions with other systems
The interaction of MCH with other peptidergic systems was initially described in fish scales, where the action of MCH to lighten color is antagonistic to the action of - MSH, which acts to darken skin color. This mutually antagonistic action persists in the mammalian hypothalamus, where the effect is focused on feeding behavior. Thus, although MCH administration increases feeding, -MSH acts to decrease feeding. When the peptides are administered together, depending on the relative dose, one can antagonize the action of the other (15). For example, 10 µg of -MSH completely blocks the feeding effect induced by 5 µg of MCH; however, 10 µg of -MSH has no effect on feeding induced by 25 µg of MCH. Glucagon-like peptide-1 and neurotensin also inhibit the orexigenic effects of MCH (60).
Changes in MCH expression have also been reported in two genetic mouse models of altered energy homeostasis, suggesting that there is regulation of the MCH system by other neuropeptides. MCH expression is markedly decreased in mice without muscarinic 3 receptors, which are lean (61). In contrast, mice lacking the bombesin receptor, BRS3, are obese and show increased hypothalamic expression of MCH and the MCH receptor (62). However, the mechanisms mediating these changes and the relative contribution of MCH to the phenotype are unknown.
Increased MCH levels have also been reported in another obese model, the fat/fat mouse (63). This mouse lacks carboxypeptidase E, and as a result, a number of neuropeptide preprohormones are not processed appropriately. These mice have increased levels of both MCH and NEI but very reduced levels of NGE, and it is possible that MCH contributes to their obesity.
It has also been shown that two compounds affecting glucose and fatty acid oxidation, 2-deoxyglucose and mercaptoacetate, respectively, increase MCH mRNA expression (64, 65). The mechanisms responsible for this up-regulation remain poorly defined, but these studies implicate MCH in the hyperphagic response elicited by administration of these compounds.
|
IV. MCH Receptor and Receptor Signaling |
|---|
MCHR1 contains three consensus N-glycosylation sites and several potential phosphorylation sites in its intracellular loops (75). As illustrated in Fig. 6, biochemical analysis has shown that an aspartic acid residue (Asp123) in the third transmembrane domain is crucial for ligand binding (76) and that an asparagine residue (Asn23) in the extracellular N-terminal region is important for N-linked glycosylation of MCHR1 and for cell surface expression (77). Recent studies have also shown a critical role for an arginine residue (Arg155) in the second intracellular loop of MCHR1, because mutation of this residue leads to a complete loss of initiation of signal transduction (78). The C-terminal region of MCHR1 has also been shown to be integral for receptor function (79). The actin- and intermediate filament-binding protein, periplakin, appears to be coexpressed with MCHR1 in mouse brain and may interact with the intracellular C terminal of MCHR1 to impede MCHR1-initiated signal transduction (Fig. 6) (80). Calcium mobilization is inhibited by periplakin; however, ERK1/2 phosphorylation is induced normally. In addition, an MCHR1 interacting zinc-finger protein (MIZIP) that interacts with the C terminus of MCHR1 has been identified from human brain and is postulated to be a regulatory molecule in MCHR1 signaling (81). The physiological significance of interactions with both MIZIP and periplakin is unknown.
|
) (82, 83). The receptor is primarily expressed in the brain, where highest expression of MCHR1 is detected in the piriform cortex and olfactory tubercle. Significant expression also occurs in the hippocampal formation, the shell of the nucleus accumbens, and the amygdala (73, 84, 85). MCHR1 localization in these areas suggests a role for MCH in olfactory learning and reinforcement mechanisms, which are fundamental processes in the regulation of feeding. Moderate expression of MCHR1 is observed in the arcuate and the ventromedial nuclei of the hypothalamus, which are both implicated in the regulation of feeding behavior. This localization of the receptor indicates that MCHR1 may specifically mediate the effects of MCH on appetite. As will be discussed in more detail in Section VII, there are also some reports that the receptor is expressed peripherally at lower levels compared with those found in brain (73, 85, 86).
Subsequent to the identification of MCHR1, a second 340-amino-acid MCH receptor subtype designated MCHR2 was identified in humans (87, 88, 89, 90, 91, 92). Although the overall homology to MCHR1 is relatively low (
38% amino acid identity), MCHR2 expression is similar to that of MCHR1, with the highest expression occurring in the brain, in particular the frontal cortex, amygdala, and nucleus accumbens (87, 88, 91). MCHR2 is also reported to be present in peripheral tissues, including adipose, prostate, and intestine (87, 88). In general, MCHR1 and MCHR2 are coexpressed in the same tissues. However, MCHR1 appears to have a more abundant and wider distribution pattern than MCHR2 (93). Both receptor subtypes are also present in the rhesus monkey (94, 95) as well as in dogs and ferrets (95). Notably, unlike MCHR1, expression of the MCHR2 subtype has not been detected in rodents (95). In addition, whereas MCHR2 increases intracellular Ca2+ levels, unlike MCHR1, it does not suppress cAMP (87, 88, 91). Overall, the physiological role of MCHR2 remains unclear. Based on its expression in cortical regions of the brain and the digestive tract, it has been suggested that MCHR2 may potentially have roles in cognitive and emotional aspects of food intake and/or regulation of nutritional homeostasis by MCH (90, 92).
B. MCH receptor polymorphisms
Several single-nucleotide polymorphisms (SNPs) and infrequent variations have been identified in the coding and/or flanking sequences of MCHR1 and MCHR2 (96, 97, 98). One study has reported a trend toward association of several MCHR1 SNPs with an obese phenotype in independent study groups of obese German children and adolescents (96). However, the same communication reported no association of these SNPs with any obese phenotypes in another German sample group, as well as no association in Danish, Finnish, and American cohorts (96). A separate study has identified two SNPs in MCHR1 but reported no association of these SNPs with obesity-related phenotypes in a population of 541 Caucasians (98). Furthermore, a third study has reported no abnormalities either in receptor binding assays or in functional assays measuring changes in intracellular Ca2+ and cAMP in Chinese hamster ovary cells transfected with DNA constructs containing MCHR1 and MCHR2 SNPs (97). Associated phenotypic characteristics, if any, for other SNPs identified in the MCH receptors have yet to be analyzed (97). Additional studies will be required to clarify whether MCH receptor SNPs are indeed associated with any obesity phenotypes.
It is not particularly surprising that thus far, no SNPs associated with MCH receptors and obesity have been identified. Obesity associated with the MCH gene would require a gain of function rather than a loss of function. Typically, mutations in the coding sequence of peptides lead to loss of function. This is what has been reported in human syndromes associated with obesity, including leptin and leptin receptor deficiency, POMC deficiency, and melanocortin receptor mutations. A loss of function of the MCH system would lead to leanness and resistance to DIO. Individuals with these characteristics are likely to be identified as "normal," and genetic studies of such families have not yet been performed.
C. MCH receptor antagonists
Given the data implicating the MCH system in energy homeostasis, it is reasonable to hypothesize that molecules capable of antagonizing the MCH receptor could be potential treatments for obesity. Recently, several pharmaceutical companies have reported the synthesis of potent MCHR1 antagonists. The chemical nature and properties of these compounds are beyond the scope of this review, and interested readers are encouraged to consult a recent review by Handlon and Zhou on this topic (99). In vivo, administration of these compounds decreases food intake and weight gain in rodents (99, 100). Several other studies also describe anxiolytic and antidepressant properties of these compounds, in addition to effects on energy balance (101, 102). These effects are most likely mediated by MCH receptors outside the hypothalamus, such as the prefrontal cortex and the amygdala. These exciting results show the potential of MCHR1 antagonists in treating a diverse array of diseases. It should be noted however, that very few studies have examined the specificity of these compounds in mice lacking MCHR1 (103). Evaluation of MCH receptor antagonists is particularly difficult, because rodents lack MCHR2. Although most of the MCHR1 antagonists synthesized show sufficient selectivity for MCHR1, it is unclear what the effects of these antagonists may be in organisms having another functional MCH receptor. Thus, the potential role of these agents in human obesity remains to be explored.
|
V. Neuronal Responses |
|---|
Numerous studies have examined the properties of orexin and MCH neurons and have found that these two closely apposed populations of neurons respond in opposite ways to several neurotransmitters. For example, both norepinephrine and acetylcholine inhibit MCH neurons, but activate orexin neurons (106). Glucose also has opposite effects on these two populations of neurons (107). These results suggest that the activity of MCH and orexin neurons would be differentially affected by the noradrenergic and cholinergic arousal systems and therefore participate in different stages of the wake-sleep cycle.
B. Electrophysiological responses of neurons to MCH
The effects of MCH on the electrical properties of neurons have been examined in a few studies. Consistent with the Gi/o coupling of MCHR1, MCH generally causes an inhibition of neuronal firing by presynaptic and postsynaptic mechanisms. Inhibition of voltage-gated calcium channels has been observed in recording from lateral hypothalamic neurons. This inhibition is pertussis toxin-sensitive, indicating Gi coupling of MCHR1 (108). MCH-mediated inhibition of the voltage-gated calcium channels affects all classes of currents, L, N, and P/Q, with the N-current showing the greatest inhibition (108, 109). A paradigm of rat parental overnutrition (litters were restricted to small numbers of pups) has been used to examine the responses of hypothalamic neurons to neuropeptides involved in energy homeostasis. The results showed that MCH excites medial arcuate (presumably some NPY/AgRP) neurons and inhibits ventromedial hypothalamus and PVN neurons in overnourished rats. As these rats become hyperphagic and obese, the enhanced responsiveness of these neurons to MCH is consistent with its role in promoting feeding and energy conservation (110, 111).
In other brain regions examined, MCH may potentiate synaptic transmission in the dentate gyrus of the hippocampus (112), whereas no effects of MCH have been observed on ventral tegmental area dopaminergic and nondopaminergic neurons (113). Actions in other brain regions expressing high levels of MCH receptors, such as the dorsal and ventral striatum, have not been examined, but should provide interesting clues regarding the proposed effects of MCH on locomotion and goal-directed behaviors.
|
VI. Central Effects Not Related to Energy Balance |
|---|
B. Other behavioral effects
Given the widespread expression of MCHR1, it is not surprising that MCH affects several behaviors in rodents. One of the first reports addressing potential behavioral effects of MCH used a paradigm of auditory gating in anesthetized rats. During these studies, MCH and -MSH were injected alone, or in combination, into the lateral ventricle, and the evoked potential was recorded in the dorsal hippocampus. In this model, MCH antagonizes the -MSH increase in auditory gating (14). Direct injection of MCH into the ventromedial nucleus of the hypothalamus of female rats stimulates sexual behavior (120), whereas injection of MCH into the hippocampus reverses the amnesic effects of N(G)-nitro-L-arginine (112).
Other behaviors examined after injections of MCH include grooming, locomotor activity, and rearing. MCH injected ICV into rats antagonizes the increase in grooming behavior after -MSH injection. MCH also antagonizes the increase in locomotor activity, grooming, and rearing after injection of NEI (a peptide produced from the same prohormone like MCH) (121).
Genetic inactivation of MCHR1 in mice results in increased spontaneous locomotor activity (122, 123), as well as increased wheel running activity (124). Although mice without MCH show little, if any increase in spontaneous activity, breeding MCH-deficient to leptin-deficient mice not only decreases their body weight but also increases their locomotor activity (48). In addition, feeding MCH–/– mice a high-fat diet increases both spontaneous activity (125) and wheel running (124). The locomotor effects of MCH might be mediated by its action on the mesolimbic dopamine systems. MCHR1 is highly expressed throughout the striatum and especially in the medial shell of the nucleus accumbens. Injection of MCH directly into the medial shell of the nucleus accumbens increases feeding in sated rats, and injection of MCHR1 antagonist produces the opposite effect, suggesting that this nucleus is an important site of MCH actions (126). MCHR1 ablated mice have also been found to be supersensitive to psychostimulants, such as amphetamine, by measuring the locomotor activity of wild-type and mutant mice after amphetamine injection (127).
In contrast to the metabolic effects of MCH, which consistently show that blocking MCH-MCHR1 interaction by pharmacological or genetic means results in leanness and resistance to DIO, there are conflicting reports on behavioral effects of MCH and MCHR1 antagonists, especially on exploration and anxiety. On elevated plus maze, a test for anxiety, ICV injection of MCH in rats increases the number of entries into the open arms and increases exploration in open field test, both indicative of anxiolytic effects (128).
MCH also produces anxiolytic effects when injected ICV in rats in a test of anxiety called Vogel’s punished drinking test; however, in the same study, no effects on locomotor activity in habituated or nonhabituated rats were observed (129).
Another study utilizing MCH injections in the rat third ventricle reported increased nonspecific ingestive behavior as reflected by increased ethanol and sucrose consumption with no effects observed on elevated plus maze (41).
Paradoxically, MCH antagonists also mimic some of the effects seen with MCH. Administration of the MCHR1 antagonist, SNAP7941, decreases anxiety and depression in rats in a forced swim test, by decreasing the time of immobility in the water; significantly increases rat social interaction; and decreases anxiety in guinea pig pups in a test of maternal separation (101). Similarly to SNAP7941, two novel orally administered MCHR1 antagonists, ATC0065 and ATC0175, also induce anxiolytic and antidepressant effects in rats on elevated plus maze, decrease stress-induced hyperthermia, increase social interaction in unfamiliar rats, and reduce ultrasonic vocalization in guinea pig pups after maternal separation (102).
Antidepressant effects have also been observed after MCHR1 antagonist injection into the medial shell of the nucleus accumbens using a forced swim test; injection of MCH produces the opposite effect, suggesting that some of the behavioral effects of MCH might be mediated by the mesolimbic dopamine system (126). Finally, genetic inactivation of MCHR1 results in anxiolytic effects in mice, as assessed by open field test, elevated plus maze, social interaction, and also triggers stress-induced hyperthermia. A possible decrease in serotonergic transmission has been observed in the prefrontal cortex, indicating a potential region for these complex effects of MCH (130).
It is impossible at the present time to reconcile the differences between the studies outlined above. Several variables could affect the outcome, such as the different species used in these studies (rats vs. mice); various tests used to score the behaviors (open field test, elevated plus maze, social interaction, etc.); different routes of entry (ICV vs. oral administration); the use of small diffusible compounds as antagonists for MCHR1 vs. MCH; and the use of various genetic models of MCHR1 deficiency. Unless the various experimental models are tested side by side in controlled studies, it will be difficult to assign conclusively, the function of MCH in the regulation of anxiety and exploratory behaviors.
Although the pharmacology of MCHR1 antagonists requires further studies, the physiological role of MCH in the behaviors mentioned is even more obscure. Subpopulations of MCH neurons almost certainly exist, performing distinct functions. At the present time, we have very limited knowledge about the nature of inputs into these neurons or their detailed projection pattern, and we do not understand the mechanisms underlying conditions that activate MCH neurons. Therefore, a more detailed understanding of the circuitry of the lateral hypothalamus will be necessary, to assign specific behaviors to subpopulations of MCH neurons.
|
VII. Peripheral Effects of MCH |
|---|
Expression of MCH receptor has been reported in both a skin cell line (136) and human skin, which also may express MCH (133). In addition, in primary human melanocytes, MCH appears to antagonize the action of MSH to increase melanin production (133). One group has reported antibodies to MCHR1 in patients with vitiligo, a skin disease associated with loss of pigmentation (137). Thus, it is tempting to speculate that MCH might play a role in pigmentation in mammals. However, in knockout models, loss of either MCH or the MCH receptor is not associated with change in coat color, either for the C57BL/6 strain of mouse, which has a black coat color, or for 129/SvEv mice, which have a brown coat. This is in contrast to the effects on coat color caused by agouti protein or melanocortin mutations. Hence, it seems unlikely that MCH has a physiological role in mammalian pigmentation.
MCHR1 is also present on primary rat adipocytes, and MCH can act on these cells to stimulate production of leptin (86). A physiologically relevant regulation of leptin synthesis/secretion by MCH could contribute to overall energy homeostasis. It is known that adipocytes respond to MCH. Adipocytes exposed to MCH in vitro show a 2-fold increase in leptin secretion after 4 h of treatment, which occurs secondary to a rapid and transient increase in ob mRNA.
MCHR1 is also expressed in the murine 3T3-L1 adipocyte cell line, and treatment of these cells with MCH down-regulates MCHR1 expression (138). This effect is acute and persistent up to a 2-h assay endpoint, thus indicating a mechanism of ligand-induced down-regulation for MCHR1. MCH-induced internalization of MCHR1 has also been reported in HEK293 cells stably expressing MCHR1 (70). More recent studies have indicated that MCH-induced internalization of MCHR1 occurs through a protein kinase C-, ß-arrestin-2-, and dynamin1-dependent pathway and that the C-terminal tail of MCHR1 has an important role in the internalization process (139).
In adipocytes, MCH action is mediated by activation of intracellular kinases (140, 141, 142). MCH incubation has been shown to activate both ERKs and pp70 S6 kinase in 3T3-L1 cells. Induction of phosphorylation of ERK1 and ERK2 is 3-fold, whereas induction of pp70 S6 kinase is in the 5-fold range. These respective increases occur acutely (within 5 min) and gradually decline over a 1-h incubation (138). MCH also stimulates ERK phosphorylation in HEK293 cells stably expressing MCHRI as well as in ex vivo brain slices (83). Stimulation of ERK 1/2 phosphorylation has also been reported in a human melanoma cell line endogenously expressing MCHR1 (136). These pathways are potentially important in driving increased leptin expression, because MCH-stimulated leptin promoter-driven luciferase activity can be attenuated in the presence of inhibitors of ERK 1/2 and pp70 S6 kinase activation (138).
These findings suggest that adipocytes contain an intact MCH signaling pathway that is mediated by MCHR1 (Fig. 7) and suggest that there may be a functional role for MCHR1 in the periphery. Expression of MCH has been reported in intestine and in testis (131) and it is possible that MCH, or a MCH-like peptide circulates in rat plasma (86). A separate study has also reported the presence of MCH in plasma from both lean and obese Zucker rats, with increased circulating levels of MCH being present in obese Zucker rats compared with lean rats (143). Although another report suggests that MCH circulates in human plasma and correlates with body weight (144), an earlier study using column-extracted plasma detected no activity in either rat or human samples (145). Furthermore, another report has questioned the validity of the RIA used to measure MCH in the human study (146). Thus, if MCH does play a role in the periphery, it seems unlikely that it circulates. However, it may be present in nerve terminals or secreted as a paracrine factor.
|
|
VIII. Summary and Conclusion |
|---|
). These areas coincide with sites of expression of the receptor. The distribution of these projections led to early speculation that MCH would be involved in motivated behavior. Accumulating evidence confirms that MCH regulates feeding behavior and more broadly energy homeostasis. Furthermore, increasing evidence suggests that MCH has a role in modulating more complex behaviors such as anxiety and aggression. Studies using pharmacological agents indicate that the MCH system can be targeted by agonists and antagonists to induce changes in both eating behavior and more complex behaviors. In aggregate, these findings suggest that it may be possible to manipulate the MCH system in humans, with the aim of treating obesity and potentially disorders of cognitive and motor function.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online June 20, 2006
1 P.P. and R.L.B. contributed equally to this work. 
Abbreviations: AgRP, Agouti-related polypeptide; DIO, diet-induced obesity; ICV, intracerebroventricularly; MCH, melanin-concentrating hormone; MCH-OE, MCH overexpressing; MCHR1, MCH receptor type 1; MIZIP, MCHR1 interacting zinc-finger protein; MSH, melanocyte-stimulating hormone; NEI, neuropeptide E-I; NGE, neuropeptide G-E; NPY, neuropeptide Y; POMC, preproopiomelanocortin; PVN, paraventricular nucleus; SNP, single-nucleotide polymorphism; ZI, zona incerta.
|
References |
|---|
-MSH and MCH are functional antagonists in a CNS auditory gating paradigm. Peptides 14:431–440[CrossRef][Medline]
Top
Abstract
I. Historical Perspective
