Cbi

JD

R

taaahaArtatipnIbcmbm

sc

eAmsKihot

EoU

Biochemical and Biophysical Research Communications 258, 574–577 (1999)

Article ID bbrc.1999.0675, available online at http://www.idealibrary.com on

0CA

hanges in Hypothalamic Agouti-Related Protein (AGRP),ut not a-MSH or Pro-Opiomelanocortin Concentrations

n Dietary-Obese and Food-Restricted Rats

oanne A. Harrold,1 Gareth Williams, and Peter S. Widdowsoniabetes and Endocrinology Research Group, Department of Medicine, University of Liverpool, Liverpool, United Kingdom

eceived March 24, 1999

oa

itaptcMtSftm

wrMhrmest

ubortntpbv8tsw

Melanocortin-4 receptor (MC4-R) density is thoughto be regulated by synaptic availability of endogenousgonist, a-melanocyte-stimulating hormone (a-MSH),nd also by agouti-related protein (AGRP), which actss a competitive antagonist. As hypothalamic MC4-Rave been implicated in the regulation of energy bal-nce, we examined concentrations of a-MSH andGRP in hypothalami of dietary-obese and food-estricted rats. In dietary-obese rats, AGRP concentra-ions were significantly increased by 43% (p < 0.01)bove lean controls, whereas a 91% (p < 0.01) reduc-ion was observed in food-restricted rats. Surpris-ngly, hypothalamic concentrations of a-MSH and itsrecursor peptide, pro-opiomelanocortin (POMC), didot differ significantly from controls in either model.n conclusion, we suggest that MC4-R activity may note regulated by changes in agonist (a-MSH) but byhanges in the antagonist (AGRP) availability, whichay modulate background activation of the receptor

y tonic a-MSH release. AGRP may be an importantodulator of feeding behaviour. © 1999 Academic Press

Key Words: agouti-related protein; a-melanocytetimulating hormone; obesity; energy balance; melano-ortin-4 receptors; hypothalamus.

It is now clear that melanocortins play a key role innergy balance in mammals [1]. Obesity in the mutantgouti mouse results from blockade of hypothalamicelanocortin-4 receptors (MC4-R) by ectopic expres-

ion of an endogenous antagonist, agouti protein.nockout mice lacking MC4-R are phenotypically sim-

lar to agouti mice, in that they show hyperphagia,yperinsulinemia, hyperglycemia and severe late-nset obesity [2]. Furthermore, the recent identifica-ion of a mutation in the MC4-R gene in a morbidly

1 To whom correspondence should be addressed at Diabetes andndocrinology Research Group, Department of Medicine, Universityf Liverpool, Duncan Building, Daulby Street, Liverpool L69 3GA,K. Fax: (44)-(151)-706-5797. E-mail: harrold@liverpool.ac.uk.

574006-291X/99 $30.00opyright © 1999 by Academic Pressll rights of reproduction in any form reserved.

bese human demonstrates that the MC4-R system islso important in human energy balance [3].Activation of hypothalamic MC4-R in rodents inhib-

ts feeding. Food intake is reduced by intracerebroven-ricular (ICV) or intrahypothalamic injections of-MSH, the proposed endogenous agonist and cleavageroduct of the pro-opiomelancortin (POMC) precursorhat is expressed by neurons of the hypothalamic ar-uate nucleus (ARC), or the stable a-MSH analogue,T-II [1]. By contrast, MC4-R receptor blockade with

he moderately-selective cyclic peptide antagonist,HU9119, increases food intake and induces obesity

ollowing repeated ICV injections [1]. This suggestshat MC4-R are tonically activated, which might nor-ally help to restrain overeating.The complex regulation of melanocortin receptorsas recently highlighted by the discovery of Agouti-

elated peptide (AGRP). This potent antagonist of bothC3-R and MC4-R is synthesized in neurones of the

ypothalamic arcuate (ARC) nucleus and is up-egulated in obese (ob/ob) and diabetic (db/db) mutantice [4]. A role for the peptide in the regulation of

nergy balance is further supported by the recent ob-ervation of dense AGRP expression in the paraven-ricular nucleus and DMH [5].

We previously investigated the distribution and reg-lation of MC4-R in two rat models of altered energyalance, dietary-obesity and food-restriction. Dietary-bese rats exhibit marked regionally selective down-egulation of MC4-R in the hypothalamus, notably inhe ARC, dorsomedial (DMH) and ventromedial (VMH)uclei and in the median eminence (ME) [6]. By con-rast, food-restricted rats display up-regulation in hy-othalamic MC4-R density in the same areas [6]. It haseen shown that G-protein-linked receptors are in-ersely regulated by synaptic availability of agonist [7,]. We therefore interpreted these findings as evidencehat the hypothalamic MC4-R system operates homeo-tatically to try to maintain optimal energy stores,ith a-MSH release being stimulated in dietary-obese

rlvfilsAwLt

dtgippePfbtenWrmhe

M

soter

wFfc

hoUcLtpa

sathMhtt(

uoamtatmca

R

octPdH

TABLE 1

IFGPGPP

Vol. 258, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ats in order to curb their increased caloric intake,eading to down-regulation of MC4-R density. Con-ersely, we suggest that a-MSH release is inhibited inood-restricted rats, where satiety signals would benappropriate, leading to up-regulation in hypotha-amic MC4-R. We also suggested that leptin is respon-ible for signaling changes in energy balance to theRC pro-opiomelanocortin (POMC) neurons, some ofhich express the functional OB-Rb leptin receptor [9].eptin apparently stimulates these neurons and couldhus elicit the appropriate a-MSH response.

Our previous interpretation of alterations in MC4-Rensity infers changes in ligand availability. The aim ofhis study was to examine the relationship between re-ional MC4-R levels and hypothalamic concentrations ofts key ligands. We measured hypothalamic levels of theroposed endogenous receptor agonist, a-MSH, and itseptide precursor, POMC, together with putative endog-nous antagonist, agouti related peptide (AGRP). BothOMC and AGRP neurones lie within the ARC, but they

orm distinct populations which terminate independentlyut parallel to each other suggesting a postsynaptic in-eraction of signals from the two pathways [10]. AGRP isxpressed by nearly all (95%) ARC neurones that expresseuropeptide Y (NPY) [10], a major orexigenic peptide.e measured AGRP levels using an antibody which

ecognises the pharmacologically active C-terminal do-ain of AGRP (residues 83-132), ICV injection of whichas been shown to increase feeding and antagonise theffects of a-MSH [11].

ATERIALS AND METHODS

Male Wistar rats obtained from Liverpool University breedingtock had free access to tap water and were housed in groups of twor three under a 12:12 light:dark cycle (lights on 07:00h) and at aemperature of 19-22°C and 30-40% humidity. In dietary-obesityxperiments, nine Wistar rats were fed a highly-palatable energy-ich diet [12] for either two or eight weeks, while matched controls

Body Weights, Fat Pad and Gastrocnemius Muscle MasseRats and Rats Fed a Highly-Palatable Diet for E

2 Weeks

Control Diet-fed

nitial body weight (g) 216 6 3 220 6 7inal body weight (g) 336 6 6 348 6 9onadal fat pad (g) 2.7 6 0.1 2.9 6 0.1erirenal fat pad (g) 2.5 6 0.1 2.5 6 0.1astrocnemius muscle (g) 1.6 6 0.1 1.6 6 0.1lasma insulin (mU/ml) 13.8 6 0.5 14.3 6 0.4lasma leptin (ng/ml) 2.4 6 0.2 4.1 6 0.2 **

Note. Data are shown as means 6 SEM for groups of 8 or 9 rats.* p , 0.01, ** p , 0.001 compared with respective controls.

575

ere fed normal pelleted chow (CRM; Biosure, Cambridge, UK).ood-restricted Wistar rats (n58) were allowed 60% of their habitual

ood intake for 10 days to reduce weight by approximately 20%, whileontrols ate normal pelleted chow ad libitum.Rats were killed by CO2 inhalation and blood collected into cold

eparinized tubes, centrifuged and the plasma stored for the later assayf glucose, using a glucose oxidase-based kit (Boehringer MannheimK; Lewes, Sussex, UK); and leptin and insulin concentrations, using

ommercially available radioimmunoassay (RIA) kits, respectively frominco (Biogenesis, Poole, Dorset, UK) and Pharmacia Upjohn Diagnos-ics (Milton Keynes, Bucks, UK). Epididymal fat pads, perirenal fatads and one gastrocnemius muscle were also dissected and weighed,s representative depots of adipose tissue and muscle.Brains were quickly removed and the hypothalamus rapidly dis-

ected using consistent landmarks, being bordered by the optic chi-sma, mammillary bodies and hypothalamic sulcus, which enclosehe medial (ARC, VMH, DMH) and lateral nuclei and ME [13]. Theypothalamic blocks were weighed before being placed in 0.25 ml 0.1

acetic acid, boiled and sonicated to disrupt the tissue. The tissueomogenates were neutralised with 100 mM Tris (pH 7.0) and cen-rifuged (14,000 rpm) for 10 min. AGRP, POMC and a-MSH concen-rations were measured using commercially available RIA kitsPhoenix Pharmaceuticals Inc., Mountain View, CA, USA).

In a parallel experiment 500 mm hypothalamic sections were cutsing a vibratome, from two groups of nine rats following two weeksf palatable diet or standard laboratory diet feeding. The ARC, VMHnd DMH were punched out of the slices under a binocular dissectingicroscope [14], placed in 0.1 M acetic acid and treated in an iden-

ical manner to hypothalamic blocks. a-MSH was readily detectednd its concentrations were measured as described above. The pro-ein concentration in the pellet remaining after centrifugation waseasured using the BCA method [15]. Neither AGRP nor POMC

oncentrations could be measured in individual hypothalamic nucleis they fell below the sensitivity limit of their respective RIAs.

ESULTS

Rats fed a highly-palatable diet for eight weeks devel-ped obesity, with varying degrees of adiposity, but in-reased body weight and adiposity were not evident afterwo weeks diet feeding, compared with controls (Table 1).lasma leptin concentrations were increased in palatableiet fed rats at both two and eight weeks (Table 1).ypothalamic AGRP concentrations were significantly

nd Plasma Insulin and Leptin Concentrations of Controler 2 or 8 Weeks or a Restricted Diet for 10 Days

Palatable diet feeding

8 Weeks Food restriction

Control Diet-fed Control Restricted

159 6 7 170 6 6 289 6 7 293 6 10488 6 15 551 6 12** 350 6 11 281 6 8**5.3 6 0.3 11.0 6 0.6** 3.5 6 0.1 1.7 6 0.1**3.0 6 0.2 7.8 6 0.2** 3.2 6 0.1 1.1 6 0.1**2.8 6 0.1 2.8 6 0.1 2.0 6 0.1 2.0 6 0.12.8 6 1.3 34.2 6 2.0** 14.0 6 0.6 8.1 6 0.2**4.7 6 0.2 6.3 6 0.2** 2.6 6 0.2 1.4 6 0.1**

s, aith

2

int(mtc

etcrAHn

D

t

aiwtsrbtccMrwtiw

ciaappawAdpisslhtPmfmir

cfo

1dreD

Vol. 258, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ncreased in diet-fed rats at both time points, whereaseither a-MSH nor POMC concentrations in whole hypo-halamic blocks differed significantly from controlsFig. 1). Furthermore, concentrations of a-MSH inicropunched DMH, VMH and ARC nuclei from rats fed

he highly-palatable diet for two weeks were not signifi-antly different from controls (Fig. 2).

Food-restricted rats showed reduced body weight,pididymal and perirenal fat pad masses, plasma lep-in and insulin concentrations (Table 1), with no de-rease in gastrocnemius muscle mass. In the food-estricted group, there was almost complete loss ofGRP in the hypothalamus, versus freely-fed controls.owever, concentrations of a-MSH and POMC wereot significantly different (Fig. 1).

ISCUSSION

We have previously demonstrated specific changes inhe density of hypothalamic MC4-R in dietary-obese

FIG. 1. Whole hypothalamic concentrations of (A) AGRP (83-32), (B) a-MSH, and (C) POMC after 2 or 8 weeks highly-palatableiet feeding or 10 days food restriction. Open columns are controlats, whereas filled columns represent the rodent models of alterednergy balance. * p , 0.05, ** p , 0.01, compared to control animals.ata are shown as means 6 SEM for groups of 8 or 9 rats.

576

nd food-restricted rats. MC4-R were down-regulatedn the ARC, VMH, ME and DMH in dietary-obesity,hereas food-restricted rats showed up-regulation of

he same receptor in the same regions [6]. We haveuggested that these alterations point to a homeostaticole of the melanocortin system to maintain energyalance, consistent with the fact that agonist bindingo G-protein-linked receptors down-regulates the re-eptor. In dietary-obese rats, overconsumption of ex-ess calories enhances signalling through the leptin/C4-R system and this signals satiety, leading to

eceptor down-regulation in order to restore bodyeight towards normal levels. In contrast, reduced sa-

iety signalling through through the MC4-R pathwayn food-restricted rats, in order to stimulate hunger,ould result in receptor up-regulation.There are a number of possible permutations which

ould explain the increased MC4-R functional activityn dietary-obesity, namely increased synaptic avail-bility of the endogenous agonist (presumably a-MSH)nd/or reduced antagonist (AGRP) availability. In theresent study we did not detect changes in whole hy-othalamic a-MSH concentrations in rats fed the pal-table diet for either two or eight weeks and neitherere there changes in a-MSH concentrations in theRC, VMH or DMH, areas which exhibit markedown-regulation in MC4-R density at two weeks (un-ublished data). This unexpected finding suggests thatncreased activity at the MC4-R is not due to increasedynaptic availability of a-MSH. This suggestion istrengthened by our finding of unchanged hypotha-amic levels of the a-MSH precursor, POMC, in theypothalamus of dietary-obese rats. There is some con-roversy over the relationship between hypothalamicOMC expression and energy balance. Reduced POMCRNA levels have been reported in food-deprived and

ood-restricted rodents [16] and in ob/ob and db/dbice [17], while increased levels were observed follow-

ng ICV leptin injection [18]. However, a recent studyeports that POMC mRNA levels do not change in

FIG. 2. a-MSH concentrations in specific hypothalamic nuclei ofontrol rats (open columns) and animals fed a highly-palatable dietor 2 weeks (filled columns). Data shown as mean 6 SEM for groupsf 9 rats.

l y

(

scAedsHrhsmrscbaarbate

wcispfimrecatcmlpTtrpAfssl

paht

dia

A

MB

R

1

1

1

1

1

1

1

1

1

1

2

2

Vol. 258, No. 3, 1999 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ethal yellow (A ), MC4-R knockout or leptin-deficientob/ob) mice [19].

Hypothalamic AGRP concentrations were, however,ignificantly higher in dietary-obese rats than in leanontrols. This is apparently paradoxical, as increasedGRP (antagonist) availability at the MC4-R would bexpected to up-regulate the receptor, in contrast to theown-regulation which we have previously observed inpecific hypothalamic regions of dietary-obese rats.owever, increased peptide levels do not necessarily

eflect increased neuronal activity. Indeed, increases inypothalamic NPY concentrations have also been ob-erved in animals fed palatable diet [20]. Here, theost likely explanation given for the elevation was

educed synaptic release of NPY leading to increasedtorage in nerve terminals. As AGRP and NPY areo-localised our supposition is that peptide release islocked from the terminals of the NPY/AGRP neuronsnd that elevated AGRP levels may also representccumulation in nerve terminals. The overall effect ofeduced AGRP release, in the presence of a sustainedackground activation of MC4-R by tonic release of-MSH would have the effect of increasing the recep-or’s activity, through reduced blockade by the endog-nous antagonist.Opposite changes were found in food-restricted rats,hich exhibited reduced energy stores with signifi-

antly reduced fat-pad masses and a concomitant falln plasma leptin concentrations. Under these circum-tances, reduced satiety signalling would be antici-ated to stimulate hunger and searching for food. Ourndings of MC4-R up-regulation in food-restricted ratsay therefore reflect reduced MC4-R activity. In these

ats, as with dietary-obese animals, we could find novidence for changes in hypothalamic POMC or a-MSHoncentrations, suggesting that reduced agonist avail-bility does not contribute to the reduced MC4-R ac-ivity. However, the marked reductions in AGRP con-entration in the hypothalamus of food-restricted ratsay reflect dramatically increased level of peptide re-

ease such that presynaptic stores are severely de-leted resulting in increased blockade of the MC4-R.his would have the effect of reducing the overall ac-ivity of the MC4-R. To compensate for this elevatedelease, increased AGRP synthesis would be antici-ated. Interestingly, it has recently been reported thatGRP mRNA increases more than 13 fold in mice

asted for 2 days [21]. As other factors such as mRNAtability may also alter, this increased mRNA synthe-is will not necessarily correspond to increased peptideevels.

In conclusion, we have demonstrated changes in hy-othalamic AGRP(83-132)-immunoreactivity duringltered energy balance, in the absence of changes inypothalamic a-MSH concentrations. We suggest thathe changes in MC4-R activity in dietary-obese and

577

ietary-restricted rats are mediated solely by changesn the synaptic availability of the endogenous MC4-Rntagonist, namely AGRP.

CKNOWLEDGMENTS

We are very grateful for financial support from Astra Hassle AB,olndal, Sweden and to Drs. Gunnar Skogmann and Thomaserglindh for their valuable contributions.

EFERENCES

1. Fan, W., Boston, B. A., Kesterson, R. A., Hruby, W. J., and Cone,R. D. (1997) Nature 385, 165–168.

2. Huszar, D., Lynch, C. A., Fairchild-Huntress, V., Dunmore,J. H., Fang, Q., Berkemeier, L. R., Gu, W., Kesterson, R. A.,Boston, B. A., Cone, R. D., Smith, F. J., Campfield LA, Burn, P.,and Lee, F. (1997) Cell 68, 131–141.

3. Giles, S. H., Yeo, I., Farooqi, S., Aminian, S., Halsall, D. J.,Stanhope, R. G., and O’Rahilly, S. (1998) Nature Genet. 20,111–112.

4. Shutter, J. R., Graham, M., Kinsey, A.C., Scully, S., Luthy, R.,and Stark, K. L. (1997) Genes Dev. 11, 593–602.

5. Haskell-Luevano, C., Chen, P., Li, C., Chang, K., Smith, M. S.,Cameron, J. L., and Cone, R. D. (1999) Endocrinol. 140, 1408–1415.

6. Harrold, J. A., Widdowson, P. S., and Williams, G. (1999) Dia-betes 48, 267–271.

7. Shockley, M. S., Burford, N. T., Sadee, W., and Lameh, J. (1997)Neurochem. 68, 601–609.

8. Blake, A. D., Bot, G., Li, S., Freeman, J. C., and Reisine, T. (1997)J. Neurochem. 68, 1846–1852, 1997.

9. Cheung, C. C., Clifton, D. K., and Steiner, R. A. (1997) Endo-crinology 138, 4489–4492.

0. Broberger, C., Johansen, J., Johansson, C., Schalling, M., andHokfelt, T. (1998) Proc. Natl. Acad. Sci. USA 95, 15043–15048.

1. Rossi, M., Kim, M. S., Morgan, D. G., Small, C. J., Edwards,C. M., Sunter, D., Abusnana, S., Goldstone, A. P., Russell, S. H.,Stanley, S. A., Smith, D. M., Yagaloff, K., Ghatei, M. A., andBloom, S. R. (1998) Endocrinol. 139, 4428–4431.

2. Widdowson, P. S., Upton, R., Henderson, L., Buckingham, R.,Wilson, S., and Williams, G. (1997) Brain Res. 774, 1–10.

3. Pickavance, L., Dryden, S., Hopkins, D., Bing, C., Frankish, H.,Wang, Q., Vernon, R. G., and Williams, G. (1996) Peptides 17,577–582.

4. Corrin, S. E., McCarthy, H. D., McKibbin, P. E., and Williams, G.(1991) Peptides 12, 425–430.

5. Redinbaugh, M. G., and Turley, R. B. (1986) Anal. Biochem. 153,267–271.

6. Brady, L. S., Smith, M. A., Gold, P. W., and Herkenham, M.(1990) Neuroendocrinol. 52, 441–447.

7. Mizuno, T. M., Kleopoulos, S. P., Bergen, H. T., Roberts, J. L.,Priest, C. A., and Mobbs, C. V. (1998) Diabtetes 47, 294–297.

8. Schwartz, M. W., Seeley, R. J., Woods, S. C., Weigle, D. S.,Campfield, A., Burn, P., and Baskin, D. G. (1997) Diabetes 46,2119–2122.

9. Kesterson, R. A., Huszar, D., Lynch, C. A., Simerly, R. B., andCone, R. D. (1997) Mol. Endocrinol. 11, 630–637.

0. Wilding, J. P. H., Gilbey, S. G., Manna, M., Aslam, N., andGhatei, M.A. (1992) J. Endocrinol. 132, 299–304.

1. Mizuno T. M., and Mobbs, C. V. (1999) Endocrinol. 140, 814–817.