Biochimica et Biophysica Acta 1831 (2013) 1267–1275

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bba l ip

α-MSH signalling via melanocortin 5 receptor promotes lipolysis andimpairs re-esterification in adipocytes

Adriana R. Rodrigues a,b, Henrique Almeida a,b, Alexandra M. Gouveia a,b,c,⁎a Department of Experimental Biology, Faculty of Medicine of Porto, Alameda Prof Hernâni Monteiro, 4200-319 Porto, Portugalb IBMC—Instituto de Biologia Molecular e Celular, Universidade do Porto, Rua do Campo Alegre, 823, 4150-180 Porto, Portugalc Faculty of Nutrition and Food Sciences of the University of Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal

Abbreviations: ACC, acetyl-CoA carboxylase; AMPK,activated protein kinase; ATGL, adipose triglyceridregulated protein kinase; HSL, hormone-sensitive lipprotein kinase;MC5R,melanocortin 5 receptor;MCRs,melnocyte-stimulating hormone; NEFA, nonesterified fatty acicarboxykinase; PKA, protein kinase A; PLIN1, perilipin 1siRNA, small interfering RNA; TG, triglyceride.⁎ Corresponding author at: Department of Experimen

of Porto, Alameda Prof Hernâni Monteiro, 4200-319 Por36 54; fax: +351 22 551 36 55.

E-mail address: agouveia@med.up.pt (A.M. Gouveia)

1388-1981/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.bbalip.2013.04.008

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 September 2012Received in revised form 5 April 2013Accepted 9 April 2013Available online 17 April 2013

Keywords:MelanocortinsMelanocortin 5 receptorAdipocyteLipolysisRe-esterification

The melanocortin system has a clear effect on the mobilisation of stored lipids in adipocytes. The aim of thecurrent study was to investigate the role of melanocortin 5 receptor (MC5R) on α-melanocyte-stimulatinghormone (α-MSH)-induced lipolysis in 3T3-L1 adipocytes. To this end, MC5R expression was decreased bysmall interfering RNA (siRNA), which significantly impaired the α-MSH stimulation of lipolysis, as deter-mined by glycerol and nonesterified fatty-acid (NEFA) quantification. The functional role of α-MSH/MC5Ron triglyceride (TG) hydrolysis was mediated by hormone-sensitive lipase (HSL), adipose triglyceride lipase(ATGL), perilipin 1 (PLIN1) and acetyl-CoA carboxylase (ACC). Immunofluorescence microscopy revealedthat phosphorylated HSL clearly surrounded lipid droplets in α-MSH-stimulated adipocytes, whereas PLIN1left the immediate periphery of lipids. These observations were lost when the expression of MC5R wassuppressed.In 3T3-L1 adipocytes, α-MSH-activated MC5R signals through the cAMP/PKA and MAPK/ERK1/2 pathways.PKA was fundamental for HSL and PLIN1 activation and lipolysis regulation. ERK1/2 inhibition stronglyinterfered with the release of NEFAs but not glycerol. In addition, the intracellular TG levels, which weredecreased after MC5R activation, were restored after ERK1/2 inhibition, indicating that these kinases areinvolved in NEFA re-esterification rather than lipolysis regulation. This notion is also supported by the obser-vation that the α-MSH-mediated activation of phosphoenolpyruvate carboxykinase (PEPCK) was abolishedin the presence of ERK1/2 inhibitors.Altogether, these results indicate that α-MSH-activated MC5R regulates two tightly coupled pathways in ad-ipocytes: lipolysis and re-esterification. The global effect is a decrease in adipocyte fat mass, which is impor-tant for strategies to ameliorate obesity.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Melanocortins are a group of neuropeptides derived from pro-opiomelanocortin (POMC) and include adrenocorticotropic hormone(ACTH) and α, β and γ-melanocyte-stimulating hormone (MSH)(see Ref. [1] for a review). The melanocortin system also comprisesa family of five melanocortin receptors (MCRs, MC1R to MC5R) thatregulate a variety of cellular functions, such as skin pigmentation,

adenosine 5′-monophosphate-e lipase; ERK, extracellular-ase; MAPK, mitogen-activatedanocortin receptors;MSH,mela-d; PEPCK, phosphoenolpyruvate; POMC, pro-opiomelanocortin;

tal Biology, Faculty of Medicineto, Portugal. Tel.: +351 22 551

.

rights reserved.

steroidogenesis and energy homeostasis. In fact, in the central ner-vous system, melanocortins activate MC3R and MC4R, which haveimportant anorexigenic roles in the control of food intake and bodyweight (see Refs. [2–4] for a review). Accordingly, MC4R-knockoutmice exhibit a phenotype of obesity [5], and the most frequent mono-genic obesity syndrome in humans, responsible for up to 6% of mor-bidly obese patients, results from MC4R mutations [6].

Melanocortins also interfere with energy homeostasis by acting onperipheral tissues. A significant increase in plasma α-MSH levelswere reported after successful weight reduction in obese adolescents[7,8]. Moreover, there is evidence that the peripheral administrationof a melanocortin analogue effectively reduced weight in diet-inducedobese rodents [9,10] and in POMC-null mice [11]. In contrast withthese findings, Nogueiras et al. (2007) [12] did not observe detectablechanges in rat body weight, body fat, thermogenesis or lipid metabo-lism after the peripheral infusion of melanocortin analogues.

Melanocortins are known to exert a lipolytic effect on white adi-pose tissue (WAT) [13], and compelling evidence has revealed thatsympathetic nervous system (SNS) innervation of WAT constitutes

1268 A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

an important mediator of lipid mobilisation [12,14,15]. Beyond thisaction through SNS, melanocortins have an additional and directrole on adipocytes. In fact, melanocortins bind to specific receptorson adipocytes, with MCRs expression being described for several spe-cies, including humans [16]. Real-time PCR analyses in subcutaneousand abdominal human adipose tissues revealed high levels of MC1R,low levels of MC4R and MC5R and undetectable levels of MC2R andMC3R [17]. Although some studies have reported MC5R and MC1R[18] and also MC4R [19] in human fat, Smith et al. (2003) foundonly MC1R and MC2R [20].

The direct lipolysis-promoting effect of melanocortins on adipo-cytes has also been reported in vitro using different adipocyte cellmodels [21–25]. In murine 3T3-L1 adipocytes, cells in which MC2Rand MC5R expression was observed, melanocortins are known to ac-tivate the PKA and ERK1/2 signalling pathways and several proteinsdirectly linked to the control of lipolysis [24,26–28]. However, al-though MC5R has been indicated as the potential target, the specificmelanocortin receptor that mediates the α-MSH effect was neverclarified.

Here, for the first time, we provide evidence for the role of MC5Rin α-MSH-regulated lipolysis and re-esterification in 3T3-L1 adipo-cytes. Furthermore, we show that lipolysis involves the activation ofHSL, ATGL and PLIN1 and that re-esterification is under the controlof the MAPK/ERK1/2 signalling pathway.

2. Materials and methods

2.1. Reagents

DMEM/Ham's F-12, Hank's balanced salt solution (HBSS), foetalbovine serum, glutamine, penicillin and streptomycin were pur-chased from Gibco®. DMEM, newborn calf serum and insulin wereobtained from Biochrom AG. The MEK1/2 inhibitors U0126 andPD98059 and the PKA inhibitor H89 were obtained from Cell Signal-ing and malate dehydrogenase from Calbiochem. All other chemicalswere purchased from Sigma, unless otherwise indicated.

2.2. Cell differentiation and treatment

Mouse 3T3-L1 preadipocytes (Zenbio) were grown and differenti-ated as described previously [28]. On day 8, the differentiated cellswere subjected to an over-night serum starvation and subsequentlytreated with α-MSH (1 μM) for the time periods indicated in thefigures. For ERK1/2 signalling inhibition, the cells were treated withUO126 (0.1, 1 and 10 μM) or PD98059 (50 μM) 30 minutes prior toα-MSHstimulation. PKA inhibition involved incubation of the adipocyteswith H89 (1 and 10 μM) under the same conditions as for ERK1/2. Con-trol cells were mock-treated in parallel.

2.3. Small interfering RNA

Silencing ofMC5R andATGLwas performedusing appropriatemousesiRNA oligonucleotides from Ambion. Adipocytes were transfected with10 nM siRNA oligonucleotides using Lipofectamine 2000 (Invitrogen),according to the manufacturer's protocol. After 48 hours, the cells weresplit into 24-well dishes or 25-cm2

flasks for protein or RNA prepara-tions, respectively. All assays were performed 72 hours after siRNAtransfection. A non-targeting siRNA oligonucleotide (Sigma) was usedas the control. The knockdown levels of MC5R and ATGL were assessedusing real-time PCR and western blotting, respectively.

2.4. Real-time PCR

The extraction of total RNA was performed using the RNeasy®Plus kit (Qiagen). The RNA concentration and purity were assessedusing a NanoDrop spectrophotometer. cDNA was generated using

RevertAid™ H Minus (Fermentas) with 2 μg of total RNA. Thereal-time PCR was performed using iTaq™ SYBR® Green Supermix(Bio-Rad) with the 7000 Real-Time PCR System (Applied Biosystems).β-Actin was used as a control to normalise gene expression. The fol-lowing primer sequences were used: MC5R forward, 5′-CTGGTCTCCCTTCTTTCTTCA-3′, and reverse, 5′-GGTCCTCCGCATCTCTTGG-3′,and β-actin forward, 5′-TCATGAAGTGTGACGTTGACATCC-3′, and re-verse, 5′-GTAAAACGCAGCTCAGTAACAGTC-3′.

2.5. Glycerol, NEFA and TG quantification

Differentiated adipocytes were grown in 24-well plates and stim-ulated with 1 μM α-MSH in HBSS supplemented with 2% BSA for4 hours. The cell medium was collected and used for the glyceroland NEFA quantification with the commercially available kits FreeGlycerol Reagent (Sigma) and NEFA-HR(2) (Wako Chemicals), re-spectively. For TG quantification, the adipocytes were lysed with100 μl of 5% Triton X-100 for 30 minutes at 37 °C. The cells were col-lected, diluted to 500 μl (1% Triton X-100 final concentration) andsonicated. The protein content was determined by the Bradfordassay, and 20 μg of protein was used for the quantification of the in-tracellular initial glycerol level, as described above. The TG reagent(Sigma) was then added for 5 minutes at 37 °C, and the final absor-bance at 540 nm was recorded. The total TG levels were determinedby subtracting the initial glycerol level from the final level.

2.6. cAMP assay

After serum starvation, the 3T3-L1 adipocytes were treated for15 minutes with 1 μMα-MSH in the presence of IBMX (a phosphodies-terase inhibitor). Themediumwas removed, the cells were immediate-ly lysed, and the intracellular cAMP concentration was quantifiedusing the cAMP-Screen®System (Applied Biosystems) following themanufacturer's instructions.

2.7. Western blotting analysis

After treatment, the adipocytes were washed with ice-coldphosphate-buffered saline (PBS) and solubilised in lysis buffer(50 mM Tris–HCl [pH 7.6], 10 mM NaCl, 5 mM EDTA, 1 mMβ-glycerophosphate, phosphatase and protease inhibitor cocktails)containing 0.25% Triton X-100. The lysates were sonicated, and theprotein concentration was determined using the Bradford proteinassay (Bio-Rad). A 20- μg sample of cell lysate was used for immu-noblotting, as previously described [29]. Specific primary antibod-ies were used for the detection of the phosphorylated forms ofERK1/2 (Thr202/Tyr204), HSL (Ser563, Ser565 and Ser660), ACC (Ser79),AMPKα (Thr172) and AMPKβ1 (Ser108) (1:1000, Cell Signaling) andthe expression levels of PLIN1 (1:1000 Cell Signaling) and α-tubulin(1:10000, Sigma). Chemiluminescent detection was performed usingthe SuperSignalWest Pico reagent (Pierce). After the detection of phos-phorylated ERK1/2, ACC, HSL, AMPKα and β1, the membranes werestrippedwith 10% SDS for 30 minutes at room temperature and then in-cubated with antibodies that recognise total HSL, ERK1/2, AMPKβ1/2and ACC (1:1000, Cell Signaling).

2.8. Fluorescence microscopy

Afterα-MSH stimulation, the cells were washed with ice-cold PBS,fixed with 4% paraformaldehyde and permeabilised for 5 minuteswith 1% Triton X-100. After blocking with 10% bovine serum albuminfor 1 hour, the immunodetection of PLIN1 and phosphorylated HSLon Ser660 was performed by incubating the cells with specific anti-bodies (1:400 and 1:800, respectively, Cell Signaling) overnight at4 °C and a secondary antibody (Alexa®488, Molecular Probes) for1 hour at room temperature. The nuclei were counterstained with

Fig. 1. MC5R promotes lipolysis after α-MSH stimulation. A, 3T3-L1 adipocytes weretransfectedwith a non-targeting siRNA oligonucleotide (siCont) orwith an siRNAdirectedagainst MC5R (siMC5R). The MC5R mRNA levels were determined by real-time PCRand are expressed relative to the MC5R levels in the siCont-transfected adipocytes.B, Transfected cells (siCont and siMC5R) were mock-treated (non-stimulated) or stimu-lated with 1 μM α-MSH (α-MSH-stimulated) or 10 μM isoproterenol (Iso-stimulated)for 4 hours. The effect on glycerol and NEFA release into the cell medium was evaluatedand is expressed relative to the non-stimulated siCont group. All the experiments wererepeated at least three times (*P b 0.05; **P b 0.01; ns— no statistical significance).

1269A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

4′,6-diamidino-2-phenylindole (DAPI), and the images were capturedusing an ApoTome fluorescence microscope (Zeiss).

2.9. PEPCK activity

For the PEPCK assay, fully differentiated adipocytes were plated in12-well plates, serum starved overnight and stimulated with α-MSHfor 4 hours. To clarify the role of ERK1/2 activity in glyceroneogenesis,the cells were pre-treated with U0126 or PD98059 30 minutes priorto α-MSH stimuli. The cells were collected in 50 mM Tris–HCl(pH 7.4), lysed by sonication and centrifuged at 15,000 ×g for 5 mi-nutes at 4 °C. The protein content of the supernatant was quantifiedby the Bradford assay (Bio-Rad), and 50 μg of protein was used forthe determination of the PEPCK activity, as described elsewhere[30]. Briefly, the samples were incubated with PEPCK assay buffercontaining 50 mM Tris–HCl (pH 7.4), 50 mM NaHCO3, 1 mM MnCl2,1 mM phosphoenolpyruvate, 2 U malate dehydrogenase and 0.25 mMreduced nicotinamide adenine dinucleotide (NADH) at 30 °C for3 minutes. The reaction was initiated by the addition of 0.15 mM2′-deoxyguanosine 5′-diphosphate (dGDP). The reduction in absor-bance was measured at 340 nm for 5 minutes at one-minute inter-vals. One unit of PEPCK activity was expressed as the number ofnanomoles of NADH oxidised per minute.

2.10. Data analysis

Digital images of the western blots were analysed by densitometryusing Scion Image. Each result represents the mean ± standard devi-ation of at least three independent experiments. All the data wereanalysed using the two-tailed Student t-test.

3. Results

3.1. Small interfering RNA knockdown of MC5R decreases the lipolysisrate in α-MSH-stimulated 3T3-L1 adipocytes

The expression of MC5R was abolished by siRNA to clearlydefine the role of this receptor in the regulation of adipocyte lipid me-tabolism. 3T3-L1 cells were transfected with MC5R-specific siRNAoligonucleotide (siMC5R) and also with a non-targeting control oligo-nucleotide (siControl), which does not interfere with the expressionof any gene in mouse cells. Real-time PCR analysis revealed that theMC5R mRNA levels were decreased 85% in the MC5R siRNA-silencedcells compared to the MC5R levels in the control siRNA-transfectedcells (Fig. 1A). However, because there are no specific antibodiesagainst MC5R available, it was not possible to evaluate the MC5Rknockdown at the protein level.

α-MSH-treated adipocytes with suppressed MC5R expressionwere used to specifically study the role of this receptor in the regula-tion of lipolysis. As shown in Fig. 1B, α-MSH activates lipolysis by in-ducing the release of glycerol and NEFAs into the cell medium(approximately 10- and 12-fold above the control, respectively).MC5R silencing significantly reduced the lipolytic effect of α-MSH:the glycerol and NEFA levels were decreased by 47 and 40%, respec-tively (Fig. 1B). Isoproterenol was used as a positive control for lipol-ysis stimulation by a mechanism that is MC5R independent. Asexpected, the silencing of MC5R expression did not influence the li-polysis rate of isoproterenol, indicating that MC5R activation in thissystemwas exclusively mediated by the addition of α-MSH. These re-sults provide the first compelling evidence that MC5R has a role in li-polysis regulation in adipocytes.

3.2. MC5R action is mediated by HSL, ACC, PLIN1 and ATGL

TG hydrolysis involves twomain lipases: ATGL and HSL. ATGL con-verts TG onto diacylglycerol; although HSL hydrolyses both TG and

diacylglycerol, the conversion of diacylglycerol to monoacylglycerolis more efficient [31]. HSL activity is regulated by PKA phosphoryla-tion at Ser660 and Ser563; the phosphorylation of HSL at Ser565

prevents the PKA activation of Ser563 and, consequently, inhibitslipolysis. To gain insight into the molecular mechanisms governingα-MSH-induced lipolysis in adipocytes, we next investigated theeffect of α-MSH on HSL activation. Western blotting experiments re-vealed that α-MSH treatment (for 5 to 360 minutes) did not alter thetotal HSL content but significantly increased HSL Ser660 and Ser563

phosphorylation (Fig. 2A). HSL phosphorylation at these residueswas rapid, achieving maximum within 5 minutes and remainingabove basal levels for at least 2 hours. Analysis of HSL Ser565 phos-phorylation revealed that this residue was not affected by α-MSHtreatment (Fig. 2A). Altogether, these results indicate that α-MSH

1270 A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

1271A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

leads to specific HSL phosphorylation at residues that promote itsactivation, thus inducing lipolysis.

ATGL expression was also not altered by α-MSH (Fig. 2A) or byMC5R silencing (data not shown). However, because there are noavailable antibodies against the activated form of ATGL, it was notpossible to assess ATGL activation by western blotting. To evaluatethe involvement of ATGL in MC5R-induced lipolysis, the expressionof this lipase was reduced by siRNA (siATGL). The quantification ofthe ATGL protein level by western blotting revealed a decrease of75% after siRNA treatment (Fig. 2B, top graph); as expected forthese conditions, the lipolysis rate was also significantly decreasedin isoproterenol-treated cells (Fig. 2B, middle and bottom graphs).This result is in agreement with the requirement of ATGL in lipolysisactivation by a broad spectrum of lipolytic agents. α-MSH is notan exception, as glycerol and NEFA release was also significantlyreduced in the stimulated cells with ATGL expression interference.These results unequivocally demonstrate the involvement of ATGLin α-MSH-mediated lipolysis.

Lipid storage and lipolysis are also regulated by PLIN1. This proteinassociates with lipid droplets under basal conditions, whereas it is ac-tivated by phosphorylation upon lipolytic stimulation and dissociatesfrom lipid droplets towards the cytosol [32]. A western blottinganalysis of adipocyte samples stimulated with α-MSH revealedPLIN1 at a slightly higher molecular weight as a consequence of phos-phorylation, which was observed by a slower SDS–PAGE migration(Fig. 2A). To confirm PLIN1 phosphorylation after α-MSH treatment,the cells were additionally subjected to alkaline phosphatase diges-tion. The PLIN1 upward shift in mobility upon SDS–PAGE analysisdisappeared in the phosphatase-treated samples (SupplementaryFig. 1).

The biosynthesis of fatty acids is mediated by ACC, which becomesinactivated by phosphorylation. As shown in Fig. 2A, ACC is also rap-idly phosphorylated in α-MSH-treated adipocytes, resulting in a de-crease in its activity. However, the melanocortin α-MSH did notinfluence AMPKα and β1 activation, fuel-sensing enzymes presentin most mammalian tissues (Fig. 2A). The total ACC, AMPKα andβ1/2 protein levels were also not affected by α-MSH treatment.

The role of MC5R in the activation of HSL, PLIN1 and ACC wasdetermined in adipocytes with silenced expression of the receptor.A western blotting analysis revealed that a low level of MC5R(siMC5R) results in the complete inhibition of the phosphorylationof HSL Ser660 and Ser563, ACC and PLIN1 after α-MSH stimulation(Fig. 2C). The HSL565 residue was not subjected to activation byα-MSH and, accordingly, was not affected by MC5R silencing. The in-volvement of MC5R in the location/activation of HSL Ser660 and PLIN1was also evaluated by fluorescent microscopy (Fig. 2D). The HSLSer660 phosphorylation levels were very low in the cells transfectedwith control siRNA but increased substantially after α-MSH treat-ment, in accordance with the western blotting results. It was also pos-sible to observe the intracellular localisation of the phosphorylatedform of HSL, which was more concentrated at the periphery of thelipid droplets. In contrast, α-MSH lost the ability to promote signifi-cant HSL phosphorylation at Ser660 in adipocytes with a suppressedMC5R level. PLIN1 formed a protective coat surrounding the lipiddroplets under non-stimulating conditions (Fig. 2D), whereas it wasphosphorylated and dissociated from the lipid droplets after α-MSH

Fig. 2. MC5R action is mediated by HSL, ACC, PLIN1 and ATGL. A, Fully differentiated adipoctime periods (5, 10, 15, 30, 60, 120, 180 and 360 minutes). The phosphorylation levels of(pAMPKβ1) and the expression levels of HSL, PLIN1, ATGL, ACC, AMPKα, AMPKβ1/2 andwith a non-targeting siRNA (siCont) or with an siRNA directed against ATGL (siATGL) to tedetermined by western blotting and are presented relative to the ATGL levels in the siCoNEFA output into the medium (middle and bottom graphs, respectively) was evaluatC, MC5R-silenced adipocytes were mock-treated (0 minutes) or incubated with 1 μM α-MSer565) and ACC were determined by western blotting. The total levels of HSL, PLIN1, ACCwere treated with 1 μM α-MSH for 30 minutes, and the location of phospho-HSL (Ser660)stained with DAPI (blue). All the blots and immunofluorescence images are representative

treatment. This effect was substantially attenuated in the cellstransfected with MC5R siRNA (Fig. 2D).

3.3. α-MSH-activated MC5R signals through the ERK1/2 and cAMP/PKApathways

The characterisation of the signalling pathways induced byα-MSH in cell systems over-expressing MC5R revealed the involve-ment of the MAPK/ERK1/2 and Gs/cAMP/PKA pathways [28]. α-MSHwas also shown to increase cAMP and pERK1/2 levels in 3T3-L1 adi-pocytes, though the specific MCRs involved remain unknown[24,27,28,33]. The direct role of MC5R in the activation of these sig-nalling pathways was tested by siRNA experiments. As shown inFig. 3A, in the presence of a non-targeting siRNA oligonucleotide(siCont), α-MSH-stimulated adipocytes exhibited a marked increasein cAMP synthesis. More important, reducing MC5R expression withsiRNA (siMC5R) decreased the high cAMP levels achieved afterα-MSH treatment (a reduction of approximately 75%). The westernblotting analysis of adipocyte samples stimulated with α-MSH forseveral time periods (5, 15 and 30 minutes) revealed a markeddecrease in phosphorylated ERK1/2 with MC5R silencing (Fig. 3B),even though ERK1/2 expression was not affected, as shown bywestern blotting analysis of total ERK1/2.

3.4. MC5R-mediated lipolysis depends on the cAMP/PKA pathway but notthe ERK1/2 pathway

The increase in cAMP levels after α-MSH treatment and the factthat several proteins related to lipolysis, such as HSL and PLIN1, areclassically activated by PKA phosphorylation [31] implicate this path-way in the regulation of lipolysis. To assess this possibility, adipocyteswere pre-incubated with the PKA inhibitor H89 (1 and 10 μM) 30 mi-nutes prior to α-MSH treatment. A western blotting analysis demon-strated that the ability of α-MSH to induce HSL (Ser660 and Ser563)and PLIN1 phosphorylation was hampered in the presence of H89(Fig. 4A). In agreement with the results shown in Fig. 2, HSL Ser565

was not phosphorylated after α-MSH treatment.Additionally, blocking cAMP/PKA signalling with H89 in α-MSH-

stimulated cells led to a significant decrease in glycerol and NEFAlevels released into the cell medium (41% and 53%, respectively)when compared to the control cells (without the addition of PKA in-hibitor) (Fig. 4B).

We further evaluated whether the ERK1/2 signalling pathway wasinvolved in lipolysis regulation. We initially analysed the HSL (Ser660,Ser563 and Ser565), ACC and PLIN1 phosphorylation levels in α-MSHstimulated cells that were pre-treated with two specific MEK1/2 in-hibitors, UO126 and PD98059. The western blotting results revealedthat inhibiting ERK1/2 activation did not influence the α-MSH-mediated phosphorylation of these enzymes at the specific epitopesrecognised by the antibodies (Fig. 4C). The detection of ERK1/2 phos-phorylated forms (pERK1/2) tested the efficiency of both inhibitors,and total ERK1/2, HSL and α-tubulin analysis was used to normaliseprotein loading (Fig. 4C).

The NEFA and glycerol levels in the cell medium were also quanti-fied after α-MSH stimulation and under ERK1/2 inhibition. As shownin Fig. 4D, the glycerol levels remained unaltered when the cells were

ytes were mock-treated (0 minutes) or incubated with 1 μM α-MSH for the indicatedHSL (pHSL Ser660, Ser563 and Ser565), ACC (pACC), AMPKα (pAMPKα) and AMPKβ1

α-tubulin were evaluated by western blotting. B, 3T3-L1 adipocytes were transfectedst the involvement of ATGL in MC5R-mediated lipolysis. The ATGL protein levels werent-transfected adipocytes (top graph). The effect of ATGL depletion on glycerol anded and is expressed as described in the legend for Fig. 1 (*P b 0.05; **P b 0.01).SH for 5, 15 or 30 minutes, and the phosphorylated levels of HSL (Ser660, Ser563 andand α-tubulin were also analysed. D, Adipocytes transfected with siCont or siMC5R

and PLIN1 (green) was evaluated by immunofluorescence microscopy. The nuclei areof three independent experiments.

Fig. 3. MC5R signals through the cAMP/PKA and ERK1/2 pathways. 3T3-L1 adipocyteswere transfected with a non-targeting (siCont) or with an MC5R-specific (siMC5R)siRNA oligonucleotide. A, Cells were mock-treated (non-stimulated) or incubated for15 minutes with 1 μM α-MSH in the presence of IBMX. The cells were immediatelylysed, and the intracellular cAMP concentration was quantified. The data are expressedrelative to the non-stimulated siCont group. (**P b 0.001). B, Cells were mock-treated(0 minutes) or incubated with 1 μM α-MSH for 5, 15 or 30 minutes. The levels of totaland phosphorylated ERK1/2 were determined by western blotting. All the experimentswere conducted four times independently.

1272 A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

incubated with α-MSH in the presence of UO126 and PD98059. Sur-prisingly, ERK1/2 inhibition led to a decrease in the NEFA concentra-tion in the α-MSH-treated cells (Fig. 4D). It has been reportedthat lipolysis is more reliably monitored by the release of glycerolinto the incubation medium than the NEFA output, which stronglydepends on the regulation of its intracellular recycling in re-esterification mechanisms [34]. The diminished NEFA release andthe unaltered glycerol level suggest the involvement of MC5R andthe ERK1/2 pathway in the down-regulation of lipid re-esterification.

3.5. The ERK1/2 pathway regulates fatty acid re-esterification

To evaluate the role of α-MSH/MC5R and ERK1/2 in lipid re-esterification, we first quantified the intracellular TG content. In theα-MSH-treated adipocytes with normal MC5R expression (siCont)the TG levels were clearly diminished, in accordance with the actionof this melanocortin on the stimulation of lipolysis (Fig. 5A), andMC5R silencing (siMC5R) clearly abolished this effect. More impor-tantly, we quantified the TG levels in the presence of two ERK1/2inhibitors, UO126 and PD98059 (Fig. 5B). ERK1/2 inhibition in theα-MSH-stimulated cells (blank bars) led to a significant increase inTG levels (Fig. 5B). These data and results shown in Fig. 4D indicatethat ERK1/2 inhibition interferes with the effect ofMC5R on the impair-ment of re-esterification, as TG levels increase as a consequence of NEFAre-esterification, even though the release of glycerol remained high.

Re-esterification occurs through the synthesis of glycerol3-phosphate, which is mainly produced via glyceroneogenesis inadipocytes (by the conversion of such non-carbohydrate sub-strates as lactate or pyruvate), a mechanism that requires cytosolicphosphoenolpyruvate carboxykinase (PEPCK) activity [34,35]. To ex-plore the role of PEPCK, we evaluated the PEPCK activity in adipocytesafter stimulation with α-MSH and verified that it decreased incomparison to mock-treated cells (Fig. 5C). This result confirms that

α-MSH/MC5R has an inhibitory effect on re-esterification. Moreover,the α-MSH effect on PEPCK activity was abolished in the presence ofthe ERK1/2 inhibitors UO126 and PD98059 (Fig. 5C).

Altogether, these results indicate that MC5R signalling throughERK1/2 is not implicated in the regulation of lipolysis but rather me-diates fatty acid re-esterification through glyceroneogenesis.

4. Discussion

Lipid mobilisation from adipocytes is regulated by several agents,such as catecholamines, thyroid-stimulating hormone, insulin,glucagon, cortisol, leptin, growth hormone [13] and melanocortins[21,24,26,27,33]. Melanocortins appear to regulate lipolysis in adipo-cytes by two distinct pathways: sympathetic nervous system innervationor directly by circulating melanocortins. However, the physiologicalrelevance ofmelanocortin receptors in adipose tissue and their role in pe-ripheral tissues have not been fully clarified. Melanocortins are stress-related peptides and, although highly speculative, different types and in-tensities of stress could influence the engagement of a specific pathwayto stimulate adipose tissue lipolysis.

In attempts to elucidate the direct role of melanocortins in adiposetissue, previous studies have characterised the specific MCR subtypesin adipocytes from different species. Differentiated 3T3-L1 adipocytesmainly express MC2R and MC5R [24,27,33,36], and, because MC2Rbinds to ACTH with a higher affinity, the effect of α-MSH in thosecells has been attributed to MC5R interactions. However, no direct ev-idence for the role of MC5R in adipocytes had been published to date.Herein, the implication of MC5R in the α-MSH-mediated regulationof lipolysis is characterised for the first time using an siRNA-based ap-proach. We verified that MC5R induces lipolysis, a process that is me-diated by PKA activation of HSL, ATGL and PLIN1, and also regulatesERK1/2 phosphorylation-dependent re-esterification.

The mobilisation of fat via lipolysis is catalysed by two major en-zymes: HSL and ATGL. Our results clearly indicate that MC5R activa-tion upon α-MSH stimulation leads to HSL phosphorylation and itsconcentration around lipid droplets. In addition, lipid mobilisation isstrongly reduced upon ATGL silencing. Although the involvement ofHSL in α-MSH-induced lipolysis was previously suggested [27], theassessment of HSL activity was questionable as it was based on HSLesterase activity and the potential contribution of other esteraseswas not considered.

Lipolysis is also mediated by PLIN1, which surrounds the lipiddroplets in adipocytes and modulates the access of several lipases toTG. We verified that PLIN1 was activated after α-MSH treatment, inagreement with the data of Harmer et al. (2008) [26], and this effectwas lost when MC5R siRNA was employed. These results werecomplemented by immunofluorescence microscopy, which revealed,for the first time, the intracellular displacement of these proteinswhen subjected to a melanocortin stimulus. We also verified thatACC, an enzyme involved in TG synthesis, is phosphorylated by anMC5R-mediated mechanism to inhibit ACC function. Therefore, boththe activation of lipolytic enzymes and the inhibition of TG synthesissupport the role of MC5R in the inhibition of lipogenesis.

Surprisingly, we did not observe the phosphorylation of AMPKαand AMPKβ1, a result that, despite some controversy [27], confirmsthe data of Moller et al. (2011) [24], who did not detect AMPK stim-ulation with MCR agonists. Our result is also in accordance with theinability of α-MSH to phosphorylate HSL on Ser565, as this residue iscommonly activated by AMPK [31].

In adipocytes, cAMP/PKA and MAPK/ERK1/2 are the main signallingpathways activated by α-MSH [24,27,28,33]. The involvement of PKAin lipolysis regulation has been widely and consensually described,namely with regard to HSL and PLIN1 activation [13]. In agreement,our results show that PKAmediates the activation of PLIN1 and the phos-phorylation of HSL on Ser660 and Ser563 residues to regulate lipolysis. Incontrast, no effect was observed on Ser565, which is understandable

Fig. 4. Role of cAMP/PKA and ERK1/2 on MC5R-mediated lipolysis. A and C, Differentiated 3T3-L1 adipocytes were preincubated with different concentrations of the PKA inhibitorH89 (1 and 10 μM), the MEK inhibitors U0126 (0.1, 1 and 10 μM) and PD98059 (50 μM) or with vehicle (0) for 30 minutes and then treated without (−) or with (+) 1 μM α-MSHfor 30 minutes. The cell extracts were immunoblotted for the detection of pHSL (Ser660, Ser563 and Ser565), pACC, pERK1/2, PLIN1, HSL, ERK1/2 and α-tubulin, as described in“Materials and methods.” The images are representative of at least three experiments. B and D, Adipocytes were pre-incubated with 10 μM H89, 10 μM U0126, 50 μM PD98059or vehicle (control) for 30 minutes and then stimulated or not withα-MSH for 4 hours. The effect on glycerol and NEFA release into the cell mediumwas evaluated and is expressedrelative to the non-stimulated control group (*P b 0.05; **P b 0.01; ns — not significant).

1273A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

because HSL Ser660 and Ser563 are themajor sites regulating HSL activity,whereas Ser565 phosphorylation inhibits PKA action through Ser563 [31].With regard toMAPK/ERK1/2, we provide evidence that the inhibition of

this pathway does not interfere with α-MSH-MC5R-mediated lipolysis.In fact,α-MSH-mediated glycerol release and the phosphorylation levelsof the enzymes implicated in lipolysis did not change after ERK1/2

Fig. 5. ERK1/2 regulates MC5R inhibition of re-esterification. A, 3T3-L1 adipocytes weretransfected with a control (siCont) or with an siRNA oligonucleotide directed againstMC5R (siMC5R) and then mock-treated (non-stimulated) or stimulated with 1 μMα-MSH (α-MSH-stimulated) for 4 hours. The effect on the intracellular triglyceridelevels was evaluated and is expressed relative to the non-stimulated siCont group.All the experiments were repeated four times (**P b 0.01). B, Adipocytes werepre-incubated with 10 μM U0126, 50 μM PD98059 or vehicle (mock) for 30 minutesand then stimulated or not with α-MSH for 4 hours. The effect on the TG contentwas evaluated and is expressed relative to the non-stimulated group (*P b 0.05;**P b 0.01). C, To clarify ERK1/2 activity in glyceroneogenesis, the cells werepre-treated with U0126 or PD98059 for 30 minutes prior to 4 hours of α-MSH stimu-lation or mock-treatment. Cell lysates were used for the determination of the PEPCKactivity, as described in “Materials and methods.” The PEPCK activity was expressedin nanomoles of NADH oxidised per minute, and the results were plotted relative tomock-treated cells. The data represent the results from at least 7 independent experi-ments (**P b 0.01; ns — non-significant, compared to mock-treated cells). The statisti-cal significance of the effects of U0126 or PD98059 on the α-MSH-stimulated cells areindicated with ##P b 0.01.

1274 A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

inhibition. These data are in apparent conflict with the results of Cho etal. (2005) [27], reporting a dependence of ERK1/2 on α-MSH-mediatedglycerol release. However, our 3T3-L1 adipocytes behaved differentlyfrom the cultures of Cho et al. because the glycerol release rate upon

stimulation was not comparable. Indeed, we observed an increase of ap-proximately 11-fold above the controls, contrasting with the 1.1-fold in-crease reported by these authors [27]. These divergent resultsmost likelyare a consequence of distinct lipid accumulation during adipocyte differ-entiation. Another recent report [24] stated that ERK1/2 inhibition didnot decrease NEFA release from melanocortin-treated 3T3-L1 cells, indisagreement with our results. However, instead of using α-MSH totest the inhibitor, these authors employed the melanocortin analoguesPG-901 and NDP-α-MSH, which could explain the differences reported.

Although the MAPK/ERK1/2 pathway is not involved in α-MSH/MC5R-mediated lipolysis, we demonstrated that it does regulatere-esterification by inhibiting the recycling of NEFAs into TG. Re-esterification occurs with significant rates in mammalian adipose tis-sue during periods of active lipolysis [34]. In fact, in humans, NEFArecycling back to TG in adipose tissue was estimated to be as highas 40% [37]. The re-esterification of glycerol requires the synthesisof glycerol-3-phosphate, which can be produced from glucose duringglycolysis, from glycerol after glycerol kinase phosphorylation orthrough the conversion of pyruvate by PEPCK via glyceroneogenesis.Adipose tissue is known to present an insignificant glycerol kinaseactivity, which disfavours the phosphorylation of the glycerol re-leased during lipolysis [34,35]. In addition, there is evidence thatglycolysis is not the major source of glycerol-3-phosphate for triglyc-eride synthesis; in fact, sources other than glycerol or glucose werepredominantly observed in mammals, both during starvation andafter the ingestion of a diet high in carbohydrates [35]. Therefore,we consider glyceroneogenesis as the main metabolic pathway forTG turnover in adipocytes, with PEPCK as the rate-limiting enzyme.Accordingly, we did observe that, in parallel with lipolysis, α-MSHdecreased glyceroneogenesis by reducing PEPCK activity through amechanism mediated by ERK1/2 signalling.

This work provides strong evidence in favour of a role for MC5R inmurine adipocytes as a lipolysis stimulator and re-esterification re-pressor. The function of this receptor has been controversial due toits broad tissue expression and apparent diversity of effects upon ac-tivation. In mammals, MC5R has been implicated in the regulation ofsebogenesis [38], fatty acid oxidation in skeletal muscle [39], ocularimmunity [40] and adipose tissue inflammation [41]. MC5R-knockoutmice display a severe defect in water repulsion and thermoregulationdue to the decreased production of sebaceous lipids, generally implicat-ingMC5R in the regulation of exocrine glands [42]. Although the precisefunction of MC5R may vary according to the site of expression, all thedata converge in favour of its involvement in lipid metabolism.

In conclusion, we present evidence for a direct role of MC5Rin promoting lipolysis and repressing re-esterification after α-MSHstimulation. Strategies that aim to increase lipolysis and subsequentfatty acid utilisation might be useful in ameliorating or preventingobesity. In fact, further insight into the modulation of the down-stream targets of MC5R would increase our knowledge of adipocytebiology and allow the development of new anti-obesity drugs.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.bbalip.2013.04.008.

Acknowledgements

Rodrigues A.R. was supported by POCI 2010, FSE and “Fundaçãopara a Ciência e Tecnologia” (SFRH/BD/41024/2007). We gratefullythank Sociedade Portuguesa de Endocrinologia, Diabetes eMetabolismo,ABBOTT and Tanita HealthyWeight Community Trust for financial sup-port of this work.

References

[1] S.N. Cooray, A.J. Clark, Melanocortin receptors and their accessory proteins, Mol.Cell. Endocrinol. 331 (2011) 215–221.

[2] O.J. Marston, A.S. Garfield, L.K. Heisler, Role of central serotonin and melanocortinsystems in the control of energy balance, Eur. J. Pharmacol. 660 (2011) 70–79.

1275A.R. Rodrigues et al. / Biochimica et Biophysica Acta 1831 (2013) 1267–1275

[3] R. Pandit, J.W. de Jong, L.J. Vanderschuren, R.A. Adan, Neurobiology of overeatingand obesity: the role of melanocortins and beyond, Eur. J. Pharmacol. 660 (2011)28–42.

[4] B.C. De Jonghe, M.R. Hayes, K.K. Bence, Melanocortin control of energy balance:evidence from rodent models, Cell. Mol. Life Sci. 68 (2011) 2569–2588.

[5] D. Huszar, C.A. Lynch, V. Fairchild-Huntress, J.H. Dunmore, Q. Fang, L.R. Berkemeier,W. Gu, R.A. Kesterson, B.A. Boston, R.D. Cone, F.J. Smith, L.A. Campfield, P. Burn, F.Lee, Targeted disruption of the melanocortin-4 receptor results in obesity in mice,Cell 88 (1997) 131–141.

[6] C. Vaisse, K. Clement, E. Durand, S. Hercberg, B. Guy-Grand, P. Froguel,Melanocortin-4 receptor mutations are a frequent and heterogeneous cause ofmorbid obesity, J. Clin. Invest. 106 (2000) 253–262.

[7] W.L. Prado, L.M. Oyama, M.C. Lofrano-Prado, A. de Piano, S.G. Stella, C.M.Nascimento, J. Carnier, D.A. Caranti, L. Tock, S. Tufik, M.T. de Mello, A.R. Damaso,Alterations in downstream mediators involved in central control of eating behav-ior in obese adolescents submitted to a multidisciplinary therapy, J. Adolesc.Health 49 (2011) 300–305.

[8] C.L. Roth, P.J. Enriori, U. Gebhardt, A. Hinney, H.L. Muller, J. Hebebrand, T. Reinehr,M.A. Cowley, Changes of peripheral alpha-melanocyte-stimulating hormone inchildhood obesity, Metabolism 59 (2010) 186–194.

[9] D.D. Pierroz, M. Ziotopoulou, L. Ungsunan, S. Moschos, J.S. Flier, C.S. Mantzoros,Effects of acute and chronic administration of the melanocortin agonist MTII inmice with diet-induced obesity, Diabetes 51 (2002) 1337–1345.

[10] A.D. Strader, H. Shi, R. Ogawa, R.J. Seeley, O. Reizes, The effects of themelanocortin agonist (MT-II) on subcutaneous and visceral adipose tissue inrodents, J. Pharmacol. Exp. Ther. 322 (2007) 1153–1161.

[11] L. Yaswen, N. Diehl, M.B. Brennan, U. Hochgeschwender, Obesity in the mousemodel of pro-opiomelanocortin deficiency responds to peripheral melanocortin,Nat. Med. 5 (1999) 1066–1070.

[12] R. Nogueiras, P. Wiedmer, D. Perez-Tilve, C. Veyrat-Durebex, J.M. Keogh, G.M.Sutton, P.T. Pfluger, T.R. Castaneda, S. Neschen, S.M. Hofmann, P.N. Howles, D.A.Morgan, S.C. Benoit, I. Szanto, B. Schrott, A. Schurmann, H.G. Joost, C. Hammond,D.Y. Hui, S.C. Woods, K. Rahmouni, A.A. Butler, I.S. Farooqi, S. O'Rahilly, F.Rohner-Jeanrenaud, M.H. Tschop, The central melanocortin system directly con-trols peripheral lipid metabolism, J. Clin. Invest. 117 (2007) 3475–3488.

[13] V.E. Chaves, D. Frasson, N.H. Kawashita, Several agents and pathways regulatelipolysis in adipocytes, Biochimie 93 (2011) 1631–1640.

[14] T.J. Bartness, Y.B. Shrestha, C.H. Vaughan, G.J. Schwartz, C.K. Song, Sensory andsympathetic nervous system control of white adipose tissue lipolysis, Mol. Cell.Endocrinol. 318 (2010) 34–43.

[15] C.K. Song, R.M. Jackson, R.B. Harris, D. Richard, T.J. Bartness, Melanocortin-4 re-ceptor mRNA is expressed in sympathetic nervous system outflow neurons towhite adipose tissue, Am. J. Physiol. Regul. Integr. Comp. Physiol. 289 (2005)R1467–R1476.

[16] K.G. Mountjoy, Functions for pro-opiomelanocortin-derived peptides in obesityand diabetes, Biochem. J. 428 (2010) 305–324.

[17] M. Hoch, A.N. Eberle, U. Wagner, C. Bussmann, T. Peters, R. Peterli, Expression andlocalization of melanocortin-1 receptor in human adipose tissues of severelyobese patients, Obesity (Silver Spring) 15 (2007) 40–49.

[18] V. Chhajlani, Distribution of cDNA for melanocortin receptor subtypes in humantissues, Biochem. Mol. Biol. Int. 38 (1996) 73–80.

[19] Y.C. Chagnon, W.J. Chen, L. Perusse, M. Chagnon, A. Nadeau, W.O. Wilkison, C.Bouchard, Linkage and association studies between the melanocortin receptors4 and 5 genes and obesity-related phenotypes in the Quebec Family Study, Mol.Med. 3 (1997) 663–673.

[20] S.R. Smith, B. Gawronska-Kozak, L. Janderova, T. Nguyen, A. Murrell, J.M.Stephens, R.L. Mynatt, Agouti expression in human adipose tissue: functionalconsequences and increased expression in type 2 diabetes, Diabetes 52 (2003)2914–2922.

[21] R.L. Bradley, J.P. Mansfield, E. Maratos-Flier, Neuropeptides, including neuropep-tide Y and melanocortins, mediate lipolysis in murine adipocytes, Obes. Res. 13(2005) 653–661.

[22] N. Hoggard, L. Hunter, J.S. Duncan, D.V. Rayner, Regulation of adipose tissue leptinsecretion by alpha-melanocyte-stimulating hormone and agouti-related protein:further evidence of an interaction between leptin and the melanocortin signallingsystem, J. Mol. Endocrinol. 32 (2004) 145–153.

[23] B.A. Boston, The role of melanocortins in adipocyte function, Ann. N. Y. Acad. Sci.885 (1999) 75–84.

[24] C.L. Moller, K. Raun, M.L. Jacobsen, T.A. Pedersen, B. Holst, K.W. Conde-Frieboes,B.S. Wulff, Characterization of murine melanocortin receptors mediating adipo-cyte lipolysis and examination of signalling pathways involved, Mol. Cell.Endocrinol. 341 (1-2) (2011) 9–17.

[25] M.J. Betz, N. Hatiboglu, B. Mauracher, D. Hadaschik, A. Sauter, H. Demmelmair, B.Koletzko, F. Beuschlein, M. Slawik, Mc2 receptor knockdownmodulates differentiationand lipid composition in adipocytes, Horm. Metab. Res. 44 (9) (2012) 670–675.

[26] S.C. Harmer, D.J. Pepper, K. Cooke, H.P. Bennett, A.B. Bicknell, Evidence of a possiblerole for Lys-gamma3-MSH in the regulation of adipocyte function, J. Endocrinol. 196(2008) 149–158.

[27] K.J. Cho, J.H. Shim, M.C. Cho, Y.K. Choe, J.T. Hong, D.C. Moon, J.W. Kim, D.Y.Yoon, Signaling pathways implicated in alpha-melanocyte stimulatinghormone-induced lipolysis in 3T3-L1 adipocytes, J. Cell. Biochem. 96(2005) 869–878.

[28] A.R. Rodrigues, H. Almeida, A.M. Gouveia, Melanocortin 5 receptor signaling andinternalization: role of MAPK/ERK pathway and β-arrestins 1/2, Mol. Cell.Endocrinol. 361 (2012) 69–79.

[29] A.R. Rodrigues, D. Pignatelli, H. Almeida, A.M. Gouveia, Melanocortin 5 receptoractivates ERK1/2 through a PI3K-regulated signalingmechanism,Mol. Cell. Endocrinol.303 (2009) 74–81.

[30] Y.Y. Chia, S.Y. Liong, S.H. Ton, K.B. Kadir, Amelioration of glucose homeostasis byglycyrrhizic acid through gluconeogenesis rate-limiting enzymes, Eur. J. Pharmacol.677 (2012) 197–202.

[31] A.D. Lampidonis, E. Rogdakis, G.E. Voutsinas, D.J. Stravopodis, The resurgence ofhormone-sensitive lipase (HSL) in mammalian lipolysis, Gene 477 (2011) 1–11.

[32] D.L. Brasaemle, Thematic review series: adipocyte biology. The perilipin family ofstructural lipid droplet proteins: stabilization of lipid droplets and control oflipolysis, J. Lipid Res. 48 (2007) 2547–2559.

[33] D. Norman, A.M. Isidori, V. Frajese, M. Caprio, S.L. Chew, A.B. Grossman, A.J. Clark,G. Michael Besser, A. Fabbri, ACTH and alpha-MSH inhibit leptin expression andsecretion in 3T3-L1 adipocytes: model for a central-peripheral melanocortin-leptin pathway, Mol. Cell. Endocrinol. 200 (2003) 99–109.

[34] L. Reshef, Y. Olswang, H. Cassuto, B. Blum, C.M. Croniger, S.C. Kalhan, S.M.Tilghman, R.W. Hanson, Glyceroneogenesis and the triglyceride/fatty acid cycle,J. Biol. Chem. 278 (2003) 30413–30416.

[35] C. Nye, J. Kim, S.C. Kalhan, R.W. Hanson, Reassessing triglyceride synthesis in ad-ipose tissue, Trends Endocrinol. Metab. 19 (2008) 356–361.

[36] B.A. Boston, R.D. Cone, Characterization of melanocortin receptor subtype expres-sion in murine adipose tissues and in the 3T3-L1 cell line, Endocrinology 137(1996) 2043–2050.

[37] M.D. Jensen, K. Ekberg, B.R. Landau, Lipid metabolism during fasting, Am. J. Physiol.Endocrinol. Metab. 281 (2001) E789–E793.

[38] L. Zhang, W.H. Li, M. Anthonavage, M. Eisinger, Melanocortin-5 receptor: a mark-er of human sebocyte differentiation, Peptides 27 (2006) 413–420.

[39] J.J. An, Y. Rhee, S.H. Kim, D.M. Kim, D.H. Han, J.H. Hwang, Y.J. Jin, B.S. Cha, J.H. Baik,W.T. Lee, S.K. Lim, Peripheral effect of alpha-melanocyte-stimulating hormone onfatty acid oxidation in skeletal muscle, J. Biol. Chem. 282 (2007) 2862–2870.

[40] A.W. Taylor, N. Kitaichi, D. Biros, Melanocortin 5 receptor and ocular immunity,Cell. Mol. Biol. (Noisy-le-grand) 52 (2006) 53–59.

[41] D.J. Jun, K.Y. Na, W. Kim, D. Kwak, E.J. Kwon, J.H. Yoon, K. Yea, H. Lee, J. Kim, P.G.Suh, S.H. Ryu, K.T. Kim, Melanocortins induce interleukin 6 gene expression andsecretion through melanocortin receptors 2 and 5 in 3T3-L1 adipocytes, J. Mol.Endocrinol. 44 (2010) 225–236.

[42] W. Chen, M.A. Kelly, X. Opitz-Araya, R.E. Thomas, M.J. Low, R.D. Cone, Exocrinegland dysfunction in MC5-R-deficient mice: evidence for coordinated regulationof exocrine gland function by melanocortin peptides, Cell 91 (1997) 789–798.