Hypothalamic Melanin-Concentrating Hormone and Estrogen-Induced Weight Loss
Hypothalamic Melanin-Concentrating Hormone and Estrogen-Induced Weight Loss
Hypothalamic Melanin-Concentrating Hormone and Estrogen-Induced Weight Loss
Paul Mystkowski,1 Randy J. Seeley,5 Tina M. Hahn,1 Denis G. Baskin,1 Peter J. Havel,3 Alvin M. Matsumoto,4
Charles W. Wilkinson,2,4 Kimberly Peacock-Kinzig,5 Kathleen A. Blake,5 and Michael W. Schwartz1
Departments of 1Medicine and 2Psychiatry and Behavioral Sciences, University of Washington and Veterans Affairs (VA)
Puget Sound Health Care System, Seattle, Washington 98104, 3Department of Nutrition, University of California at Davis,
Davis, California 95616, 4Geriatric Research/Education and Clinical Center, VA Puget Sound Health Care System, Seattle,
Washington, and 5Department of Psychiatry, University of Cincinnati Medical Center, Cincinnati, Ohio 45267
Melanin-concentrating hormone (MCH) is an orexigenic neu- Whereas estrogen-induced weight loss engaged multiple sys-
ropeptide produced by neurons of the lateral hypothalamic area tems that normally favor recovery of lost weight, the expected
(LHA). Because genetic MCH deficiency induces hypophagia increase of MCH mRNA expression induced by energy restriction
and loss of body fat, we hypothesized that MCH neurons may was selectively and completely abolished. These findings identify
represent a specific LHA pathway that, when inhibited, contrib- MCH neurons as specific targets of estrogen action and suggest
utes to the pathogenesis of certain anorexia syndromes. To test that inhibition of these neurons may contribute to the hypophagic
this hypothesis, we measured behavioral, hormonal, and hypo- effect of estrogen.
thalamic neuropeptide responses in two models of hyperestro- Key words: MCH; estrogen; body weight; energy balance;
genemia in male rats, a highly reproducible anorexia paradigm. lateral hypothalamus; anorexia; pair-feeding
The lateral hypothalamus plays a critical role in the regulation of effects on “second order” neurons situated in adjacent hypotha-
energy intake and expenditure. Initial work performed ⬎50 years lamic areas, such as the LHA and the paraventricular nucleus
ago demonstrated that bilateral lesions of the lateral hypothalamic (PVN) (Broberger et al., 1998; Elias et al., 1998; Sawchenko, 1998).
area (LHA) result in profound hypophagia in rodents (Stellar, The abundance of ARC projections to MCH neurons in the LHA
1954), and recent progress has begun to identify candidate neuro- (Elias et al., 1998) supports the hypothesis that the MCH system
nal subpopulations that may mediate these LHA feeding effects. lies “downstream” of ARC neurons involved in the adaptive re-
Neurons producing melanin concentrating hormone (MCH) rep- sponse to weight loss. Therefore, under conditions in which MCH
resent one such subset. In rats and mice, intracerebroventricular signaling is selectively disrupted, anorexia might be expected to
administration of MCH induces hyperphagia, whereas MCH defi- persist despite intact upstream signaling events (e.g., low levels of
ciency induced by targeted gene deletion leads to a hypophagia plasma leptin and ARC POMC and high levels of plasma cortico-
syndrome and loss of body fat (Qu et al., 1996; Shimada et al., 1998; sterone and ARC NPY).
Tritos et al., 1998). Thus, selective inhibition of MCH neurons One anorexia paradigm of particular utility for investigating this
might be expected to induce an anorexic response. hypothesis is chronic hyperestrogenemia in male rats (Wade, 1972,
Converging evidence suggests that the hypothalamic arcuate 1986; Mordes and Rossini, 1981; Mordes et al., 1984; Bernstein et
nucleus (ARC), in addition to the LHA, plays a key role in the al., 1986). This model provides reliable and sustained decreases in
transduction of peripheral signals reflecting negative energy bal- food intake and body weight without overt toxicity, is easily used
ance into behavioral, autonomic, and metabolic responses that and reversible, and allows quantification of the anorexic stimulus.
favor the recovery of lost weight (Elmquist et al., 1998; Woods et Furthermore, estrogen was recently shown to inhibit hypothalamic
al., 1998; Schwartz et al. 2000). Accordingly, hormonal responses to MCH expression in ovariectomized female rats (Murray et al.,
negative energy balance (declining plasma levels of insulin and 2000). We therefore hypothesized that anorexia because of hypere-
leptin in concert with increasing levels of corticosterone) induce strogenemia in male rats is associated with selective inhibition of
the ARC to increase biosynthesis and release of the coexpressed MCH. To investigate this hypothesis, we measured changes in food
orexigenic molecules neuropeptide Y (NPY) and agouti-related intake, body weight, plasma hormone levels, and hypothalamic
protein (AgRP) and to decrease expression of pro- levels of MCH, NPY, AgRP, POMC, and corticotropin-releasing
opiomelanocortin (POMC), the precursor of the anorexic melano- hormone (CRH) mRNA in two rat models of hyperestrogenemia:
cortin ␣-MSH (Schwartz et al., 1997; Hahn et al., 1998; Havel et al. Leydig cell tumor implantation (whose anorexic effects are pro-
2000). Changes in the release of these neuropeptides from ARC posed to be primarily mediated by estrogen secretion) (Mordes and
neurons are hypothesized to influence energy homeostasis via Rossini, 1981; Mordes et al., 1984; Bernstein et al., 1986) and
chronic subcutaneous estrogen administration.
Received May 23, 2000; revised Aug. 21, 2000; accepted Sept. 5, 2000.
This work was supported by Grants HL07028 (P.M.), NS-32273, DK-12829, DK- MATERIALS AND METHODS
52989-01 (M.W.S.), DK-50129, DK-35747 (P.J.H.), DK-54080, and DK-54890 (R.J.S.)
from the National Institutes of Health, the American Diabetes Association, the United Study animals and protocols
States Department of Agriculture (P.J.H.), the Department of Veterans Affairs Animals. Adult male Wistar-Furth rats (Simonsen Laboratories, Gilroy,
(A.M.M. and C.W.W.), and by the Diabetes Endocrine Research Center and Clinical CA) weighing 200 –300 gm were housed individually and maintained on a
Nutrition Research Unit of the University of Washington. We recognize the excellent 12 hr light /dark cycle (lights on at 7:00 A.M.) with ad libitum access to
technical assistance of Hong Nguyen, Vicki Hoagland, Kimber Stanhope, Elizabeth drinking water and pelleted Purina lab chow (#5001), except as noted. All
Colasurdo, Robert Beckham III, and Dr. Supriya Kelkar, and we thank Linda Walters study protocols were approved by the University of Washington Animal
for assistance in manuscript preparation. C are Committee.
Correspondence should be addressed to Dr. Michael W. Schwartz, Division of E xperiment 1: effects of Leydig cell tumor implantation on food intak e,
Endocrinology/Metabolism, Harborview Medical Center, 325 Ninth Avenue, Box plasma hormones, and hypothalamic neuropeptide levels. Body weight and
359757, Seattle, WA 98104. E-mail: [email protected]. food intake was measured daily for each rodent. One day before tumor
Copyright © 2000 Society for Neuroscience 0270-6474/00/208637-06$15.00/0 implantation surgery, 30 animals were weight-matched and placed into one
8638 J. Neurosci., November 15, 2000, 20(22):8637–8642 Mystkowski et al. • MCH and Estrogen-Induced Weight Loss
of three groups (n ⫽ 10/group): tumor-bearing or sham-implanted controls Table 1. Physiologic and hormonal responses to Leydig cell tumor-
fed ad libitum, or sham-implanted controls that were pair-fed to the food mediated hyperestrogenemia from experiment 1
intake of the tumor bearing group. The pair-fed group received two daily
meals of equal size (given at lights out and 4 hr into the dark cycle). The
total amount of chow provided daily to pair-fed animals was equal to that Sham Tumor bearing Pair-fed
consumed on the previous day by a previously assigned partner in the b
Daily food intake, gm 20.3 ⫾ 0.4 13.5 ⫾ 0.7a 13.8 ⫾ 0.7a
tumor-bearing group. This feeding regimen continued for 22 d.
Tumor implantation procedure. A LTW(m) Leydig cell tumor (graciously Daily food intake, % vehicle 100 ⫾ 2.1b 66 ⫾ 3.6a 68 ⫾ 3.5a
provided by Dr. I lene Bernstein) was removed from a donor rat anesthe- Final body weight, gm 305 ⫾ 5b 235 ⫾ 5ab 261 ⫾ 4a
tized with ketamine and xylazine (6.5 ml /10 ml, 1 ml / kg, i.p. body weight Final body weight, % vehicle 100 ⫾ 1.5b 77 ⫾ 1.5ab 86 ⫾ 1.1a
for each drug), and the excised tumor was cut into 5–10 mg fragments and
placed in 0.9% sterile saline. Under sterile conditions, a tumor fragment Plasma values
was inserted into a small subcutaneous pocket created in the right flank of Estradiol, pg/ml 38 ⫾ 21 305 ⫾ 77ab 20 ⫾ 2
similarly anesthetized recipient rats. Ad libitum fed and pair-fed animals Leptin, ng/ml 4.5 ⫾ 0.3b 1.7 ⫾ 0.3ab 3.3 ⫾ 0.3a
underwent the same procedure, but tumor cells were not implanted.
Blood and tissue anal yses. On day 22 after implantation (1 hr after lights Insulin, pmol/l 555 ⫾ 39b 380 ⫾ 47ab 565 ⫾ 61a
on), tumor-bearing and sham-implanted rats were individually briefly Glucose, mg/dl 134 ⫾ 1.4 136 ⫾ 5.0 136 ⫾ 5.7
exposed to C O2 and decapitated, followed by rapid brain removal and Corticosterone, g/dl 6 ⫾ 4b 59 ⫾ 15a 75 ⫾ 21a
collection of trunk blood in cooled heparinized tubes. Brains and plasma
were frozen at ⫺80°C. T umors were weighed and stored at ⫺20°C. The a
p ⬍ 0.05 versus sham rodents; bp ⬍ 0.05 versus pair-fed rodents.
mean postmortem tumor weight was 0.82 ⫾ 0.46 gm, and in no case did
tumor weight exceed 1% of total body weight at the time of killing. Given
the 24 hr lag required for pair feeding, collection of blood and tissue from
the pair-fed group occurred 24 hr later. Plasma levels of leptin (Linco, St.
L ouis, MO) and corticosterone (IC N Biomedicals, Costa Mesa, CA) were RESULTS
measured by RIA. Plasma insulin was assayed by a modification of the Experiment 1: effects of Leydig cell tumor implantation
double-antibody method of Morgan and Lazarow (1963). Serum estradiol
was measured using a solid phase, competitive, time-resolved immunoflu- on food intake, plasma hormones, and hypothalamic
orometric assay (DEL FIA; Wallac Oy, Oulansalo, Finland) using neuropeptide levels
europium-labeled estradiol, polyclonal anti-estradiol antibodies derived To investigate the mechanisms underlying anorexia induced by
from rabbit, and a second antibody directed against rabbit IgG coated onto
96 well plates. The intra-assay coefficient of variation of this assay was subcutaneous implantation of estrogen-secreting Leydig cell tumor
3.8%, and the lower limit of sensitivity was 0.05 nmol / l (13.6 pg /ml). cells, we studied three groups of male Wistar rats for 22 d: Leydig
Plasma glucose levels were determined by the glucose oxidase method. cell tumor-bearing, sham-implanted controls that were pair-fed to
E xperiment 2: dose-effect of estrogen administration on food intak e, body the intake of tumor-bearing group, and sham-implanted controls
weight, and plasma hormone levels. To identif y a dose of estradiol that fed ad libitum. Relative to the ad libitum fed controls, mean daily
yields plasma levels comparable with those achieved in tumor-bearing rats
of experiment 1, rats were anesthetized with ketamine and xylazine and food intake and final body weight of the tumor-bearing rats were
implanted (subcutaneously) with a pellet containing either vehicle only reduced by 34 and 23% ( p ⬍ 0.05 for both), respectively (Table 1).
(n ⫽ 15) or 2.5 mg (n ⫽ 7), 7.5 mg (n ⫽ 8), or 15 mg (n ⫽ 7) of Whereas food consumption was designed to be comparable in
17-estradiol (Innovative Research of America, Sarasota, FL). Of the pair-fed and tumor-bearing groups, final body weights were signif-
vehicle-treated rats, eight were pair-fed, as in experiment 1, to a rat of
similar initial body weight in the group receiving the 15 mg estradiol dose. icantly lower in tumor-bearing rodents, suggesting an increase in
Food intake and body weights were measured daily for 29 d and plasma was energy expenditure relative to pair-fed animals. Mean plasma
obtained for analysis on day 29. levels of 17-estradiol were ⬃10-fold greater in the tumor-bearing
E xperiment 3: effect of chronic 17-estradiol administration on hypotha- rats than those measured in either sham-implanted group. As
lamic neuropeptide mR NA levels. Rats were implanted subcutaneously with expected, leptin levels declined in both tumor-bearing and pair-fed
a pellet containing either vehicle (n ⫽ 10) or 7.5 mg (n ⫽ 8) of 17-
estradiol, a dose selected on the basis of results from experiment 2 to groups relative to ad libitum fed controls (4.5 ⫾ 0.3 ng/ml), but
approximate plasma levels of tumor-bearing rats in experiment 1, and food were significantly lower in tumor-bearing (1.7 ⫾ 0.3 ng/ml; p ⬍ 0.05
intake and body weights were measured daily for 22 d. Plasma and brains vs ad libitum fed controls) than in pair-fed rats (3.3 ⫾ 0.3 ng/ml;
were obtained and analyzed as described for experiment 1. p ⬍ 0.05 vs ad libitum fed controls and tumor-bearing rodents).
Plasma insulin levels were also significantly decreased in tumor-
In situ hybridization bearing rodents, but not in pair-fed animals, whereas corticoste-
T issue preparation. Frozen brains were sectioned in a coronal plane at 14 rone levels were significantly elevated in both tumor-bearing and
m with a cryostat, mounted on RNase-free slides, and treated with 4% pair-fed animals. Nonfasting plasma glucose levels were similar
paraformaldehyde, acetic anhydride, ethanol, and chloroform. For each across groups (Table 1).
animal, six slides (12 brain sections) of hypothalamus in the region 2.0 –3.5
mm posterior to bregma were selected for hybridization for each ribo- The results of in situ hybridization performed on coronal brain
probe. Sections for quantitation of POMC mRNA were selected from the sections are shown in Figures 1 and 2. Relative to ad libitum fed
ARC rostral to the ventromedial nucleus (V M N), whereas those used for controls, NPY and AgRP mRNA expression were comparably
N PY and AgRP mRNA hybridization contained the V M N and corre- increased in the pair-fed and tumor-bearing groups (Figs. 1a, 2),
sponded to the midregion of the ARC. Sections for MCH hybridization whereas POMC and CRH levels were reduced in these groups (Fig.
were selected from an area just rostral to slides chosen for POMC.
Particular care was taken to anatomically match sections among animals 1b). As with body weight and plasma leptin levels, POMC mRNA
for each probe by an investigator blinded to study conditions. For each levels in the tumor-bearing group were reduced to a significantly
experiment, all brain slices were concurrently prepared for hybridization greater extent than was observed in the pair-fed group. The direc-
and used in the same assay. Riboprobes from cDNA templates for N PY, tion of changes of each of these neuropeptide mRNAs was similar
AgRP, POMC, and MCH mRNAs were transcribed in the presence of
33P-UTP. Unincorporated label was separated using a QIAquick Nucleo- in tumor-bearing and pair-fed groups and consistent with a com-
tide Removal K it (Qiagen, Santa C larita, CA). The hybridization signal on pensatory response to chronic energy restriction. In contrast, ex-
film autoradiograms in the ARC of each brain slice was quantified using an pression of MCH mRNA in the LHA differed markedly between
MCI D image analysis system (Imaging Research, St. C atherine’s, Ontario, tumor-bearing and pair-fed groups. Thus, whereas MCH gene
C anada), as described previously (Schwartz et al., 1997). Both the hybrid- expression was markedly increased by pair-feeding (585% ⫾ 117 of
ization density and image area were measured, the product of which was
calculated as an index of mRNA levels from the mean of six to eight ad libitum fed controls, p ⬍ 0.05 vs ad libitum fed controls and
sections per animal (Schwartz et al., 1997). tumor-bearing rodents), MCH levels in the tumor-bearing group
Statistical anal ysis. Group mean values were compared by one-way were nonsignificantly lower than in ad libitum fed rats (74% ⫾ 17
ANOVA for experiments 1 and 2, using Fisher’s test for multiple compar- of ad libitum fed rodents; p ⫽ 0.19 for ad libitum fed controls vs
isons of between-group differences. For E xperiment 3, group mean values
were compared with a two-tailed unpaired Student’s t test. The null tumor-bearing animals) (Figs. 1c, 2). MCH mRNA levels in the
hypothesis of equal means was rejected at the p ⫽ 0.05 level of significance. LHA of pair-fed rats were therefore eightfold higher than in the
Data are expressed as mean ⫾ SEM. tumor-bearing rats ( p ⬍ 0.05).
Mystkowski et al. • MCH and Estrogen-Induced Weight Loss J. Neurosci., November 15, 2000, 20(22):8637–8642 8639
DISCUSSION
A critical role for LHA neurons, particularly those producing
MCH, in energy homeostasis is suggested by the hypophagic and
Figure 2. Representative autoradiograms of hypothalamic neuropeptide Y lean phenotype arising from targeted gene deletion of this neu-
(NPY ) and melanin-concentrating hormone (MCH ) mRNA by in situ ropeptide (Shimada et al., 1998). We therefore hypothesized that
hybridization in Leydig cell tumor-bearing (TB), pair-fed (PF ), and sham-
implanted (Sham) rodents of experiment 1. NPY mRNA levels were visibly decreased MCH signaling contributes to the pathogenesis of cer-
increased in tumor-bearing and pair-fed controls relative to Sham animals, tain forms of anorexia. Here we report the complete abolition of
whereas MCH mRNA expression was reduced in tumor-bearing compared the effect of chronic negative energy balance to increase LHA
with pair-fed animals. Arc, Arcuate nucleus of the hypothalamus; ZI, zona MCH expression in two rat models of estrogen-mediated anorexia,
incerta; LHA, lateral hypothalamic area; 3v, third cerebral ventricle.
despite the preservation of key hormonal (decreased plasma leptin
and insulin levels with increased plasma levels of corticosterone)
and hypothalamic (increased expression of NPY and AgRP mRNA
Experiment 2: dose-effect of estrogen administration on along with decreased expression of CRH and POMC mRNA)
food intake, body weight, and plasma hormone levels responses to weight loss. These findings identify MCH neurons as
To identify a dose of estradiol that achieves plasma levels compa- specific targets of estrogen action that may contribute to its an-
rable with those detected in tumor-bearing rats in experiment 1 orexic effects.
and to assess its effects on food intake and body weight, rats were Hyperestrogenemic rats displayed similar or exaggerated hor-
monitored for 29 d after subcutaneous implantation with sustained- monal responses to weight loss as compared with pair-fed controls.
release pellets containing either vehicle or 17-estradiol in one of Plasma leptin and insulin concentrations were significantly below
three doses (2.5, 7.5, and 15 mg). Relative to vehicle-implanted, ad those of both ad libitum-fed and pair-fed controls in experiment 1,
libitum fed controls, mean daily food intake in each of the estrogen- whereas corticosterone elevations paralleled those seen in pair-fed
treated groups was reduced by ⬃25%, an effect detected on the first rodents. Thus, hyperestrogenemia had no apparent adverse affect
day of treatment (Table 2). The magnitude of the food intake and on the recruitment of adaptive hormonal responses to energy
body weight changes did not differ across the three estradiol groups deficit. Moreover, the ability of these hormonal cues to engage
and approximated that seen in the tumor-bearing group of exper- ARC and PVN neuronal responses that accompany energy deficit
8640 J. Neurosci., November 15, 2000, 20(22):8637–8642 Mystkowski et al. • MCH and Estrogen-Induced Weight Loss
Table 2. Dose effect of 17-estradiol administration on food intake, body weight, and plasma hormone levels from experiment 2
Table 3. Effect of chronic 17-estradiol administration on body weight sive weight loss, whereas important ARC and PVN responses were
and plasma hormone levels from experiment 3 unimpeded, is consistent with the model illustrated in Figure 4, in
which estrogen-sensitive MCH neurons in the LHA are “down-
Vehicle 7.5 mg estradiol stream” in a pathway used by ARC neurons to trigger feeding
Final body weight, gm 399 ⫾ 6 267 ⫾ 6a
responses (Elias et al., 1998; Elmquist et al., 1998; Sawchenko,
1998; Schwartz et al., 2000). According to this model, interventions
Final body weight, % vehicle 100 ⫾ 1.6 67 ⫾ 1.5a
that disrupt MCH signaling can cause sustained anorexia by inter-
Plasma values
rupting the pathway used by ARC neurons to control feeding
Estradiol, pg/ml 8.5 ⫾ 0.5 1046 ⫾ 105a behavior.
Leptin, ng/ml 5.8 ⫾ 0.7 3.3 ⫾ 0.4a This hypothesis is supported by the finding that MCH-deficient
Glucose, mg/dl 137 ⫾ 2 113 ⫾ 2a mice have reduced ARC POMC production, yet continue to un-
Corticosterone, g/dl 11 ⫾ 9 67 ⫾ 31ns dereat, and by growing evidence that ARC neurons can directly
a ns
influence LHA activity (Shimada et al., 1998). Recent work has
p ⬍ 0.05 versus vehicle; p ⫽ not significant.
demonstrated rich connections between NPY- and POMC-
containing neurons of the ARC and MCH neurons in the LHA
(Broberger et al., 1998; Elias et al., 1998) and that pharmacological
blockade of melanocortin-4 receptors, which are concentrated in
the LHA, increases MCH mRNA expression (Hanada et al., 2000).
Furthermore, leptin-deficient ob/ob mice display increased MCH
mRNA expression (Qu et al., 1996), despite relatively low expres-
sion of leptin receptor in the LHA (Schwartz et al., 1996). Thus,
leptin appears to influence MCH neurons via an indirect mecha-
nism. The recent demonstration that ARC-derived leptin-
responsive neurons project to the LHA provides a plausible mech-
anism, whereby leptin-deficiency increases MCH gene expression
(Elias et al., 1999).
Our data are in agreement with the recently demonstrated effect
of estrogen administration to decrease hypothalamic MCH mRNA
Figure 3. Hypothalamic mRNA levels as measured by in situ hybridization
in rodents bearing a subcutaneous pellet releasing 17-estradiol (E2) or levels in female rats rendered estrogen-deficient by ovariectomy
vehicle (Veh) (experiment 3). Results (mean ⫾ SEM) are presented as a (Murray et al., 2000). Thus, estrogen deficiency in female rats is
percentage of mRNA expression relative to Veh. The pellets released 7.5 associated with increased hypothalamic MCH gene expression that
mg of E2 over 21 d. *p ⬍ 0.05 versus Veh; #p ⫽ 0.19. may, in turn, contribute to the accompanying hyperphagia and
weight gain in this setting (Wade and Gray, 1979). We found in
appeared to be intact in the hyperestrogenemic groups. Tumor- male rats that, in contrast to the sixfold increase of MCH mRNA
bearing rodents of experiment 1 displayed 45–100% increases in expression observed in pair-fed controls, hyperestrogenemic ani-
NPY and AgRP mRNA relative to ad libitum fed control rodents, mals mounted MCH responses no different from ad libitum fed
an effect similar to that observed in the pair-fed group, whereas controls. Together, these observations establish MCH neurons as
POMC and CRH mRNA expression was inhibited to an even targets of estrogen action. Our data further suggest that supra-
greater extent in hyperestrogenemic versus pair-fed rodents. This physiological levels of estrogen can override the ability of weight
combination of hypothalamic neuropeptide responses was also loss and decreased leptin signaling to induce MCH expression.
observed in estradiol-treated rodents in experiment 3. Thus, an- These findings also suggest, but do not establish, a specific contri-
orexia persisted in hyperestrogenemic animals, despite the appar- bution of MCH signaling to the alterations of energy balance
ent preservation of a set of compensatory hormonal and medial observed in settings of estrogen excess and deficiency. Further
hypothalamic responses expected to potently stimulate food intake. work is warranted to test the hypotheses that MCH neurons me-
Although the PVN, like the LHA, contains second-order neu- diate the feeding effects observed with physiological, as well as
rons involved in energy homeostasis (such as those containing supraphysiological, alterations in estrogen status.
CRH), the finding that electrolytic lesions of the PVN do not The hypothesis that estrogen receptors mediate the effects of
attenuate estrogen-induced anorexia (Butera et al., 1992) suggests estrogen on energy balance is supported by the observation that
that key neurons that mediate the effects of estrogen on energy these receptors are expressed at high levels in medial and lateral
homeostasis must be located elsewhere. Our demonstration that hypothalamic areas implicated in energy homeostasis (Couse et al.,
hyperestrogenemia selectively and completely suppressed the ex- 1997). However, whereas Sar et al. (1990) reported the colocaliza-
pected increase of MCH mRNA expression in response to progres- tion of NPY and estrogen receptor in a subpopulation of ARC
Mystkowski et al. • MCH and Estrogen-Induced Weight Loss J. Neurosci., November 15, 2000, 20(22):8637–8642 8641
Hahn T, Breininger J, Baskin D, Schwartz M (1998) Coexpression of neuropeptide-Y immunoreactivity in estradiol-concentrating cells in the
AgRP and NPY in fasting-activated hypothalamic neurons. Nat Neurosci hypothalamus. Endocrinology 127:2752–2756.
1:271–272. Sawchenko P (1998) Toward a new neurobiology of energy balance, appe-
Hanada R, Nakazato M, Matsukura S, Murakami N, Yoshimatsu H, Sakata tite, and obesity: the anatomists weigh in. J Comp Neurol 402:435– 441.
T (2000) Differential regulation of melanin-concentrating hormone and Schwartz MW, Seeley RJ, Campfield LA, Burn P, Baskin DG (1996)
orexin genes in the agouti-related protein/melanocortin-4 receptor sys- Identification of targets of leptin action in rat hypothalamus. J Clin Invest
tem. Biochem Biophys Res Comm 268:88 –91. 98:1101–1106.
Havel P, Hahn T, Sindelar D, Baskin D, Dallman M, Weigle D, Schwartz Schwartz M, Seeley R, Woods S, Weigle D, Campfield L, Burn P, Baskin D
M (2000) Effects of streptozotocin-induced diabetes and insulin treat- (1997) Leptin increases hypothalamic pro-opiomelanocortin (POMC)
ment on the hypothalamic melanocortin system and muscle uncoupling mRNA expression in the rostral arcuate nucleus. Diabetes 46:2119 –2123.
protein 3 expression in rats. Diabetes 49:244 –252. Schwartz MW, Woods SC, Porte D, Seeley RJ, Baskin DG (2000) Central
Laudenslager ML, Wilkinson CW, Carlisle HJ, Hammel HT (1980) En- nervous system control of food intake. Nature 404:667– 671.
ergy balance in ovariectomized rats with and without estrogen replace- Shimada M, Tritos NA, Lowell BB, Flier JS, Maratos-Flier E (1998) Mice
ment. Am J Physiol 238:R400 –R405. lacking melanin-concentrating hormone are hypophagic and lean. Nature
Mordes JP, Rossini A (1981) Tumor-induced anorexia in the Wistar rat. 396:670 – 674.
Science 213:565–567. Stellar E (1954) The physiology of motivation. Psychol Rev 61:5–22.
Mordes JP, Longcope C, Flatt J, MacLean DB, Rossini A (1984) The rat Tritos NA, Vicent D, Gillette J, Ludwig DS, Flier ES, Maratos-Flier E
LTW(m) Leydig cell tumor: cancer anorexia due to estrogen. Endocri- (1998) Functional interactions between melanin-concentrating hor-
nology 115:167–173. mone, neuropeptide Y, and anorectic neuropeptides in the rat hypothal-
Morgan C, Lazarow (1963) Immunoassay of insulin: two antibody system: amus. Diabetes 47:1687–1692.
plasma insulin levels of normal, subdiabetic, and diabetic rats. Diabetes van Dijk G, Thiele TE, Seeley RJ, Woods SC, Bernstein IL (1997)
12:115–163. Glucagon-like peptide-1 and satiety [letter, comment]. Nature 385:214.
Murray J, Baker B, Levy A, Wilson C (2000) The influence of gonadal Wade GN (1972) Gonadal hormones and behavioral regulation of body
steroids on pre-pro melanin-concentrating hormone mRNA in female weight. Physiol Behav 8:523–534.
rats. J Neuroendocrinol 12:53–59. Wade GN (1986) Sex steroids and energy balance: sites and mechanisms
Qu D, Ludwig DS, Gammeltoft S, Piper M, Pelleymounter MA, Cullen of action. Ann NY Acad Sci 474:389 –399.
MJ, Mathes WF, Przypek J, Kanarek R, Maratos-Flier E (1996) A role Wade GN, Gray J (1979) Gonadal effects on food intake and adiposity: a
for melanin-concentrating hormone in the central regulation of feeding metabolic hypothesis. Physiol Behav 22:583–593.
behaviour. Nature 380:243–247. Woods S, Seeley R, Porte D, Schwartz M (1998) Signals that regulate food
Sar M, Sahu A, Crowley WR, Kalra SP (1990) Localization of intake and energy homeostasis. Science 280:1378 –1383.