Characterization of a novel peripheral pro-lipolytic mechanism in mice: role of VGF-derived peptide TLQP-21

Biochem. J. (2012) 441, 511–522 (Printed in Great Britain) doi:10.1042/BJ20111165 511

Characterization of a novel peripheral pro-lipolytic mechanism in mice: roleof VGF-derived peptide TLQP-21Roberta POSSENTI*†, Giampiero MUCCIOLI‡, Pamela PETROCCHI*†, Cheryl CERO§‖, Aderville CABASSI¶,Lucy VULCHANOVA**, Maureen S. RIEDL††, Monia MANIERI‡‡, Andrea FRONTINI‡‡, Antonio GIORDANO‡‡, Saverio CINTI‡‡,Paolo GOVONI§§, Gallia GRAIANI‖‖, Federico QUAINI‖‖, Corrado GHE‡, Elena BRESCIANI¶¶, Ilaria BULGARELLI¶¶,Antonio TORSELLO¶¶, Vittorio LOCATELLI¶¶, Valentina SANGHEZ‖, Bjarne D. LARSEN***, Jorgen S. PETERSEN***1,2,Paola PALANZA‖, Stefano PARMIGIANI‖, Anna MOLES*†††, Andrea LEVI* and Alessandro BARTOLOMUCCI§‖3

*Institute of Cell Biology and Neurobiology, National Research Council and Fondazione Santa Lucia, Rome, Italy, †Department of Neuroscience, University of Roma II – Tor Vergata,Rome, Italy, ‡Department of Drug Science and Technology, University of Turin, Turin, Italy, §Department of Integrative Biology and Physiology, University of Minnesota, Minneapolis,MN 55455, U.S.A., ‖Department of Evolutionary and Functional Biology, University of Parma, Parma, Italy, ¶Department of Internal Medicine, Nephrology and Health Sciences,University of Parma, Parma, Italy, **Department of Veterinary and Biomedical Sciences, University of Minnesota, St. Paul, MN 55108, U.S.A., ††Department of Neuroscience, Universityof Minnesota, Minneapolis, MN 55455, U.S.A., ‡‡Department of Molecular Pathology, Marche Polytechnic University, Ancona, Italy, §§Department of Experimental Medicine,University of Parma, Parma, Italy, ‖‖Department Internal Medicine, University of Parma, Parma, Italy, ¶¶Department of Experimental Medicine, University of Milano-Bicocca, Monza,Italy, ***Zealand Pharma A/S, Glostrup, Denmark, and †††Genomnia srl, Lainate, Italy

The peptides encoded by the VGF gene are gaining biomedicalinterest and are increasingly being scrutinized as biomarkersfor human disease. An endocrine/neuromodulatory role for VGFpeptides has been suggested but never demonstrated. Furthermore,no study has demonstrated so far the existence of a receptor-mediated mechanism for any VGF peptide. In the presentstudy, we provide a comprehensive in vitro, ex vivo and invivo identification of a novel pro-lipolytic pathway mediatedby the TLQP-21 peptide. We show for the first time thatVGF-immunoreactivity is present within sympathetic fibres inthe WAT (white adipose tissue) but not in the adipocytes.Furthermore, we identified a saturable receptor-binding activityfor the TLQP-21 peptide. The maximum binding capacity forTLQP-21 was higher in the WAT as compared with other

tissues, and selectively up-regulated in the adipose tissue ofobese mice. TLQP-21 increases lipolysis in murine adipocytesvia a mechanism encompassing the activation of noradrenaline/β-adrenergic receptors pathways and dose-dependently decreasesadipocytes diameters in two models of obesity. In conclusion, wedemonstrated a novel and previously uncharacterized peripherallipolytic pathway encompassing the VGF peptide TLQP-21.Targeting the sympathetic nerve–adipocytes interaction mightprove to be a novel approach for the treatment of obesity-associated metabolic complications.

Key words: adipocyte, β-adrenergic receptor, granin-derivedpeptide, obesity, sympathetic nervous system, VGF.

INTRODUCTION

Obesity-associated pathologies are increasing their incidenceworldwide to pandemic levels [1]. The obesity pandemic hasstimulated the identification of molecular mechanisms regulatingmetabolic functions and controlling BW (body weight) with thefinal goal to generate new pharmacotherapy to control obesityand/or its deleterious consequences [1,2]. The WAT (whiteadipose tissue) is a complex organ with multiple topographicallocations (e.g. visceral and subcutaneous), specific functionalcharacteristics, with direct sympathetic and sensory innervationsand capable of influencing global metabolic response throughtheir repertoire of secreted adipokines [3,4]. Lipolysis, theprocess ultimately required to degrade TAGs (triacylglycerols)and release non-esterified fatty acids, is a very complex processwith multiple pathways involved [5–7]. A key player in lipolysisis the activation of the sympathetic nervous system whichsends efferent projections to the adipose tissue and activatesβ-ARs (β-adrenergic receptors) in adipocyte membranes [4,5].

In agreement, mice deficient in the three known β-ARs showobesity due to a failure of diet-induced thermogenesis [8].β-AR-less mice also show a transdifferentiation of browninto white-like adipocytes [8]. Unfortunately, sympathomimeticdrugs and β-agonists are often associated with mild-to-severeside effects or with contrasting clinical outcome, which haslimited their development or use in clinical practice [2,7].Peptidergic candidates are increasingly considered in drugdiscovery programmes [2,9,10]. In this context, peptides encodedby the granin protein family member VGF are gaining increasingbiomedical interest (e.g. [11–13]). A potential link betweenVGF and obesity is demonstrated by the observation that VGFpeptides are up-regulated in rat hypothalamus after a high-fat dietor cold exposure [14]. VGF-deficient mice are hypermetabolic,lean and resist obesity [15,16], a complex phenotype probablyresulting from the deletion of several encoded bioactive peptidesas demonstrated by the observation that VGF-derived peptidesTLQP-21 and NERP-2 centrally modulate metabolic functionsand adiposity in opposite directions, with the former being

Abbreviations used: AMPK, AMP-activated protein kinase; β-AR, β-adrenergic receptor; BAT, brown adipose tissue; BW, body weight; CNS, centralnervous system; CSS, chronic subordination stress; DIO, diet-induced obesity; ERK, extracellular-signal-regulated kinase; HFD, high-fat diet; HSD, HonestlySignificant Difference; HSL, hormone-sensitive lipase; iBAT, interscapular BAT; ir, immunoreactivity; LV, left ventricular; PKA, protein kinase A; PKC, proteinkinase C; QPCR, quantitative PCR; RT, reverse transcription; RV, right ventricle; STD, standard chow; TAG, triacylglycerol; TH, tyrosine hydroxylase; TNFα,tumour necrosis factor α; WAT, white adipose tissue; pWAT, perigonadal WAT; rpWAT, retroperitoneal WAT; scWAT, subcutaneous inguinal WAT.

1 Present address: Merck Serono International S.A., Geneva, Switzerland.2 Jorgen Petersen owns shares in Zealand Pharma A/S.3 To whom correspondence should be addressed (email [email protected]).

c© The Authors Journal compilation c© 2012 Biochemical Society

www.biochemj.org

Bio

chem

ical

Jo

urn

al

512 R. Possenti and others

catabolic and the latter being anabolic [17,18]. The C-terminalinternal peptide termed TLQP-21 has emerged as a new promisinganti-obesity target [17]. Chronic intracerebroventricular injectionof TLQP-21 in mice increased energy expenditure and rectaltemperature and prevented obesity [17,19]. Besides having acentral role in nerve terminal- and in neuron-derived cellcultures secretory granules, VGF co-localizes with molecules asdiverse as substance P, proSAAS, noradrenaline (norepinephrine),adrenaline (epinephrine) and insulin [14,20–22]. However, noresults supporting a peripheral role for VGF peptides have beenpublished so far. In the present paper, we report for the first timethat VGF is present in sympathetic fibres in WAT, binds withhigh affinity to adipocytes membranes and increases lipolysis viaa mechanism that requires the activation of noradrenaline/β-ARpathways.

EXPERIMENTAL PROCEDURE

Animals and diet

Male CD1 mice (Charles River) were individually housed onarrival in a fully controlled environment (light on at 07:00 h, lightoff at 19:00 h, T = 21 +− 1 ◦C). Standard chow (STD) was the4RF21 (Mucedola; 10.6% kcal from fat and 3.9 kcal/g; 1 kcal= 4.184 kJ); the high-fat diet (HFD) was a special pelleted dietderived from the standard 4RF21 (Mucedola; 45 % kcal from fatand 5.2 kcal/g). All animal experimentation was conducted inaccordance with the European Communities Council Directive of24 November 1986 (86/EEC), and was approved by local ethicalcommittee and the Italian Institute of Health.

Identification of VGF in WAT

VGF immunohistochemistry

Perigonadal rat WAT was fixed by immersion or transcardial per-fusion using modified Zamboni’s fixative (4 % paraformaldehydeand 0.2% picric acid). The tissue was cryoprotected in 10%sucrose for a minimum of 24 h before freezing and sectioning.Slide-mounted cryostat sections (30 μm) were incubated inblocking buffer (PBS containing 0.03 % Triton X-100, 1%BSA, 1% normal donkey serum and 0.01% sodium azide)for 1 h at room temperature (20–22 ◦C), followed by overnightincubation at 4 ◦C in primary antisera, which were guinea pig anti-TLQP-21, 1:1000; rabbit anti-TLQP-21, 1:3000; guinea pig anti-AQEE-30, 1:3000 [23]; mouse anti-TH (tyrosine hydroxylase,1:1000; Immunostar); rabbit anti-TH, 1:1000 (Millipore). TLQP-21 antisera were generated using previously described methods[24]. Briefly, the peptides TLQP-21 and SSRR-14 (correspondingto the last 14 amino acids of TLQP-21) were conjugated tobovine thyroglobulin (Sigma) using glutaraldehyde. The peptideconjugates (1 mg/ml) were emulsified with an equal volume ofFreund’s adjuvant (Difco). TLQP-21 was injected into femaleNew Zealand rabbits (n = 3, Harlan) at 2-week intervals (1 mgof peptide for initial and 0.5 mg of peptide for subsequentimmunizations). SSRR-14 was injected in female guinea pigs(n = 4, Harlan) at 2-week intervals (0.5 mg of peptide for initialand 0.25 mg of peptide for subsequent immunizations). No cross-reactivity was observed with the C-terminal peptide AQEE-30which is adjacent to TLQP-21 (Supplementary Figure S1 availableat http://www.BiochemJ.org/bj/441/bj4410511add.htm). Thespecificity of the antisera was ascertained by dot blot analysis, inwhich the antisera bound preferentially to TLQP-21 rather thanthe extended fragment TLQP-62 (Supplementary Figure S1). Theslides were rinsed in PBS, incubated in secondary antisera [Cy2

(carbocyanine)- and Cy3 (indocarbocyanine)-conjugated donkeyanti-guinea pig or anti-rabbit respectively obtained from JacksonImmunoresearch Laboratories] for 1 h at room temperature andoverlaid with a coverslip. Images were collected using anOlympus FluoView 1000 BX2 Upright Confocal microscope andprocessed using ImageJ and Adobe Photoshop.

Receptor-binding assay and lipolysis

TLQP-21-binding assay

Binding of TLQP-21 to crude membranes (30000 g pellet)isolated from pWAT (perigonadal WAT), scWAT (subcutaneousinguinal WAT), rpWAT (retroperitoneal WAT), BAT (brownadipose tissue), adrenals, cerebellum, skeletal muscle and liverwas carried out using 125I-YA-TLQP-21 as a ligand (New EnglandBiolabs). The addition of tyrosine-alanine (YA) amino acidicresidues to the TLQP-21 molecule made possible the radio-iodination of the peptide and to obtain a highly specific radioligandwith the same in vitro biological activity as TLQP-21 (resultsnot shown). For the single-point binding assay, cell membranes(corresponding to 100 μg of membrane protein) were incubatedin triplicate at 23 ◦C for 4 h under constant shaking with 0.5 nM125I-YA-TLQP-21 in a final volume of 0.5 ml of assay buffer [25].Parallel incubations, in which 1 μM unlabelled YA-TLQP-21 wasalso present, were used to determine non-specific binding, whichwas subtracted from total binding to yield specific binding values.The binding reaction was terminated by the addition of ice-coldassay buffer, followed by rapid filtration through Whatman GF/Bfilters and the radioactivity bound to membranes was measuredby a Packard auto-γ -counter. Specific binding was expressed as apercentage of the total radioactivity added. For saturation-bindingstudies, cell membranes were incubated with increasing concen-trations of the radioligand (0.03–4 nM). Competition studies wereperformed by incubating cell membranes (150 μg/tube) with 1 nM125I-YA-TLQP-21 with or without increasing concentrations (10pM–0.1 μM) of unlabelled YA-TLQP-21 or TLQP-21. Bindingspecificity was also tested with a scrambled peptide made usingthe same amino acid residues of TLQP-21. Results were plottedand curves fit using the GraphPad Prism 4 software assuming thatthe binding was due to a single class of binding sites, thus allowingdetermination of the maximum binding capacity (Bmax), dissoci-ation constant (Kd), Hill slope and concentration of the competitorcausing 50% inhibition (IC50) of specific radioligand binding.

Glycerol assay

Approx. 2–3 g of the epididymal fat was surgically removedand adipose tissue suspension was incubated as described byMuccioli et al. [26] and Student et al. [27]. Fat cells wereincubated in the presence of TLQP-21 (0.1–100 nM) with orwithout isoprenaline (15 nM, also called isoproterenol). Glycerollevels were measured in triplicate using the Free Glycerol Reagent(Sigma) and expressed relative to the cellular protein content. EC50

for inhibition of isoprenaline-induced lipolysis was determinedfor each compound under study with GraphPad Prism 4 software.

Signal transduction pathway in 3T3-L1 adipocytes

Cells and differentiation

3T3-L1 cell line was originally obtained from the ATCC(Manassas, VA, U.S.A.). Cells were maintained according tostandard procedures. The cells were differentiated according to awell-established protocol described previously [28].

c© The Authors Journal compilation c© 2012 Biochemical Society

A pro-lipolytic role for TLQP-21 peptide 513

Western blotting 3T3-L1 cells

Equivalent amounts of cell extracts (∼5×105 cells) weremixed with SDS-reducing sample buffer according to NuPAGE®

(Invitrogen). For detailed procedures, see [29]. Incubationwith primary antibodies was performed overnight at 4 ◦C:rabbit polyclonal antibodies anti-phospho-p44/p42 MAPK(mitogen-activated protein kinase; Thr202/Tyr204), anti-phospho-AMPK (AMP-activated protein kinase; Thr172) anti-phospho-PKA (protein kinase A) substrate (serine/threonine) and anti-phospho-HSL (hormone-sensitive lipase; Ser660) (Cell SignalingTechnology). Anti-tubulin was a mAb (monoclonal antibody)from Sigma. Secondary antibodies were horseradish peroxidase-coupled donkey anti-rabbit or anti-mouse (GE Healthcare). Asemi-quantitative analysis of a Western blotting densitometryscan was performed using Scan Analysis software ImageJ. Five–six different experiments run in duplicate taking the time point15 min were analysed with non-parametric Kruskall–Wallis testfollowed by Mann–Whitney U test for pair-wise comparisons.

Physiological and biochemical effects of chronic TLQP-21 deliveryin vivo in animal model of obesity

TLQP-21 peptide and surgery

TLQP-21 peptide (Primm) 40 and 400 μg/day (referred to as theLOW and HIGH dose respectively) was suspended in sterile salineand delivered (subcutaneously) via Alzet osmotic mini-pumps(model 1002) for 14 days. Controls received saline only. Prefilledosmotic mini-pumps were surgically implanted subcutaneouslyin the dorsal area of the animals. Surgery was performedin sterile conditions in mice anaesthetized with a mixture ofketamin/xilazine (100–5 mg/kg).

Animal models of obesity: DIO (diet-induced obesity)

Male mice initially weighing 32–35 g were fed for 12 weekson an HFD. DIO mice had a final BW being at least higherthan the average mean plus 2 S.D. of the BW of mice fed withSTD for the same amount of time. DIO mice were randomlyallocated to saline or TLQP-21 LOW or HIGH dose avoidingany pre-experimental difference in BW or food intake. Anadditional group of mice fed with STD and treated with salinewas used as lean control. BW and food intake were determinedevery other day. Perigonadal fat pads were manually dissectedand weighed. Fat pads were split into two parts and one-half was snap-frozen in liquid nitrogen and stored at − 80 ◦Cfor later measurement of TH enzymatic activity, noradrenaline(norepinephrine) and TAG/DNA ratio and TNFα (tumour necrosisfactor α) mRNA (see below). The second half was immersed inan ice-cold solution of 4% paraformaldehyde and embeddedin paraffin for morphometric analyses and immunohistochemistry(see below). Plasma was analysed for adrenaline concentration.BW, food intake, fat pad weight, TH activity, noradrenaline,adipocyte diameter and TNFα were analysed with a two-way ANOVA for repeated measures. All other parameterswere analysed with a one-way between factor ANOVA. Meancomparisons were performed with Tukey’s HSD (HonestlySignificant Difference) post-hoc after ANOVA or Student’sunpaired t test. Correlations were evaluated with Pearson’s test.

CSS (chronic subordination stress)-induced obesity

The CSS-induced obesity procedure has been previouslydescribed in detail [29]. Briefly, subordinate male mice wereexposed to chronic sensory contact with a dominant male and

received a brief aggressive encounter daily. The stress phasefeeding regimen was scheduled as follows (see Figure 6A): STDduring baseline and first 3 weeks of stress; HFD during thefourth and fifth week of stress. Subordinate mice were surgicallyimplanted mini-pumps delivering 40 μg/day TLQP-21 or saline.For the subsequent 2 weeks mice were fed on the HFD. Thefollowing experimental groups were used: CSS/TLQP-21 (n =7), CSS/saline (n = 7). BW and food intake were determinedevery other day. pWAT and interscapular (i)BAT were manuallydissected and weighed. BW and food intake were analysed witha two-way ANOVA for repeated measures followed by Tukey’sHSD post-hoc. Fat pad weight, TH-activity, noradrenaline andadipocyte diameter were analysed with Student’s unpaired t test.

Home cage phenotyping

Body temperature was recorded using temperature-sensingsubcutaneous transponders (Bio Medic Data Systems). Sensorswere implanted at least 15 days before the beginning of theexperiment. The assessment of individual locomotor activity inthe home cage was carried out by means of an automated systemthat uses small passive infrared sensors positioned on the top ofeach cage (TechnoSmart), as described previously [29]. Activityand body temperature results were analysed with ANOVA forrepeated measures.

Tissue and plasma biochemical assays

Detailed methods of tissue and plasma biochemical assays can befound in [30]. Briefly, tissue biopsies homogenate was incubatedwith L-tyrosine and TH activity was determined by calculatingthe amount of L-dopa (L-3,4-dihydroxyphenylalanine) generatedfrom L-tyrosine per minute per mg of tissue measuring withHPLC–ECD (electron-capture-induced dissociation) system.Noradrenaline was measured in fat biopsies by HPLCusing electrochemical detection, as previously described[30]. TAG/DNA content was analysed in tissue biopsiesby measuring TAG content using a Sigma TAG assay kit and bynormalizing to the DNA content (DNeasy tissue kit; Qiagen).Adrenaline was determined in plasma by HPLC usingelectrochemical detection, as previously described [17].

Real-time RT–PCR (reverse transcription–PCR)

pWAT TNFα mRNA was analysed in triplicate by real-timeRT–PCR. Briefly, 400 ng of total digested RNA of eachsample was subjected to RT with revertAid H Minus FirstStrand cDNA Synthesis Kit (Fermentas). QPCR (quantitativePCR) for TNFα was performed with Assay-on-Demandtechnology (Applied Biosystems, Warrington, U.K.), witha Brilliant QPCR Master Mix (Stratagene, La Jolla, CA,U.S.A.). All QPCRs were performed on the ABI 7900 HTsystem (Applied Biosystems). Samples were normalized toGAPDH (glyceraldehyde-3-phosphate dehydrogenase; Assay-on-Demand, Applied Biosystems) and 2��Ct analysis applied.

WAT morphometric analysis

Paraffin-embedded specimens of perigonadal adipose tissues fromdifferent mice (DIO and CSS) were processed for morphometricanalysis. Sections 5 μm thick were stained with haematoxylinand eosin. Optical microscopy images (Nikon Microscope Eclipse80i) were digitally captured with NIS-Elements imaging softwareF 2.20, and the diameter of 200 adipocytes for each mouse wasmeasured with ImageJ software (NIH).

c© The Authors Journal compilation c© 2012 Biochemical Society

514 R. Possenti and others

WAT TH immunohistochemistry

Paraffin sections 3 μm thick of pWAT DIO mice were usedfor immunohistochemical localization of TH protein andMAC-2 according to the ABC (avidin–biotin-peroxidase)method [30]. The primary antibodies were polyclonal anti-rabbitTH (Chemicon; 1:600 dilution) and monoclonal anti-MAC-2(Cedarlane Laboratories; 1:1000 dilution).

Heart histomorphology and morphometric analysis

The abdominal cavity was opened and the aorta was cannulatedwith a PE-50 catheter connected to a perfusion apparatus. Thechest was opened and the heart was stopped in diastole byLV (left ventricular) intracavitary injection of cadmium chloride(100 nmol). The right atrium was then cut and the myocardialvasculature was shortly perfused at a physiological pressure witha heparinized PBS solution, followed by 10 min of perfusion with10% formalin. The heart was separated by surgical excision ofthe major thoracic vessels and fixated in 4% formalin. After24 h, the free walls of the RV (right ventricle) and the LVinclusive of interventricular S were separated and their weightswere recorded. The major cavity axis of the LV, from the aorticvalve to the apex, was measured under a stereomicroscope witha ruler calibrated exactly to 0.1 μm (2Biological Instruments).LV transversal diameters and wall thickness were analysed with astereomicroscope connected to a digital camera (Kodak, DC290ZOOM). Captured microphotographs were analysed by the useof Image-Pro Plus 4.0 software (Media Cybernetics). The cavityvolume was calculated using the Dodge equation [31]. The volumeof the LV myocardium was computed by dividing ventricularweight by the specific gravity of the tissue (1.06 g/ml). Thefollowing parameters were determined: HW (heart weight) toBW ratio, LV to BW, RV to BW, LV free wall thickness, andLV transversal diameters and volume, LV thickness to chamberradius and mass to chamber volume [32]. One section, stainedwith Masson’s trichrome, was analysed by optical microscopy(magnification ×250) in order to evaluate the volume fractionof myocytes, and perivascular and interstitial fibrosis in theLV myocardium. In accordance with a procedure previouslydescribed [33], for each section a quantitative evaluation of thevolume fraction of myocytes and fibrotic tissue was performed in60 adjacent fields with the aid of a grid defining a tissue area of0.160 mm2.

RESULTS

VGF co-localizes with TH in pWAT

VGF immunoreactivity (-ir) has been detected in the dorsalroot and sympathetic ganglia and in a variety of neuro-endocrine and endocrine cells [12] and in sympatheticneural pathways innervating iBAT [34]. Our result showsthat VGF is expressed in perigonadal (p)WAT as proteinbut not as mRNA (Supplementary Figure S2 availableat http://www.BiochemJ.org/bj/441/bj4410511add.htm), thussuggesting that peptides are sorted into the regulated secretorypathway granules [35,36]. To unambiguously identify VGF withinsympathetic nerve fibres, we examined the co-localization of VGFwith TH, a marker of catecholaminergic nerves and terminals[37]. We used antisera generated against TLQP-21 (anti-TLQP-21) that recognized preferentially the processed peptide ratherthan the unprocessed precursor TLQP-62 (Supplementary FigureS1), as well as antisera generated against a region located C-terminally to TLQP-21 (anti-AQEE-30) that recognized VGF C-terminal fragments of various lengths (Supplementary Figure S1).

No cross-reactivity was observed with the C-terminal peptideAQEE-30 which is adjacent to TLQP-21 (Supplementary FigureS1). Both anti-TLQP-21 and anti-AQEE-30 labelled fibres withinpWAT that were also TH-positive (Figure 1). Staining with the twoTLQP-21 antisera was blocked by pre-absorption of the antiserawith the cognate peptides (10 μg/ml). Moreover, TLQP-21 andAQEE-30 co-localized extensively in nerve fibres, consistentwith the notion that the antisera recognize fragments of thesame neuropeptide precursor. We observed some fibres that wereanti-AQEE-30-positive and anti-TLQP-21-negative (but not viceversa) (Figures 1E and 1I).

A saturable receptor binding site for TLQP-21 in adipocytemembranes

Having established that VGF is present in sympathetic fibres inWAT we aimed to characterize binding sites for TLQP-21 [26].We used 125I-YA-TLQP-21 (an analogue of TLQP-21 with similarbiological activity, results not shown) to show that the binding tocell membranes was saturable, with high Bmax in different WATfat pads as well as in the adrenals and BAT, whereas low-bindingactivity occurred in CNS (central nervous system), liver andskeletal muscle (Figure 2A). Representative saturation isothermsand Scatchard plots of 125I-YA-TLQP-21 binding to membranesfrom WAT and iBAT are shown in Figure 2(B). Experimentsusing increasing concentrations of radiotracer revealed evidenceof saturable specific binding in the scWAT, pWAT and rpWAT padsas well as in the iBAT. Scatchard analysis (Figure 2C) showed theexistence in all four tissues of a single class of high-affinity sitewith Kd values ranging from 0.54 to 0.65 nM. The specificity of125I-YA-TLQP-21 binding to tissue membranes was assessed bycompetitive binding studies with a scrambled TLQP-21 peptide(LRSP-21) and isoprenaline (Figure 2D). Unlabelled TLQP-21and YA-TLQP-21 completely displaced radiolabelled 125I-YA-TLQP-21 from binding sites. In contrast, scrambled peptide andisoprenaline were capable of displacing only 10–25 % of thespecifically bound radiolabelled 125I-YA-TLQP-21. Finally, obesemice (fed on an HFD for 4 weeks which induced a BW gain of∼25% when compared with baseline and controls; results notshown) showed increased density of 125I-YA-TLQP-21 bindingsite in WAT and BAT fat pads only, whereas Bmax in the othertissues remained unaffected (Figure 2A).

TLQP-21 potentiates β-AR-induced lipolysis and phosphorylationof HSL in vitro

Having established that TLQP-21 binds to adipocytes, we askedwhether TLQP-21 might affect spontaneous or β-AR-mediatedlipolysis in mouse mature adipocytes. Cultured mouse adipocyteswere incubated with TLQP-21, with and without the non-selectiveβ-AR agonist isoprenaline. TLQP-21 did not affect glycerolrelease (Figure 3A), but dose-dependently increased isoprenaline-induced glycerol release (Figure 3B). The maximal dose of TLQP-21 tested (100 nM) reduced lipolytic EC50 of isoprenaline from 16to 7.5 nM (Figure 3C). TLQP-21 did not show any binding affinityfor β-AR, thus ruling out a primary agonist effect (Figure 3D).Noradrenaline promotes lipolysis via β-AR induction of cAMP-dependent PKA-mediated phosphorylation of HSL [6,38]. Basedon TLQP-21-induced potentiation of isoprenaline-induced lipoly-sis (Figure 3B), we asked whether TLQP-21 might potentiateisoprenaline-induced phosphorylation of HSL and upstreamprotein kinases in 3T3-L1 adipocytes as a model. As expected[39,40], isoprenaline increased phosphorylation of AMPK,ERK (extracellular-signal-regulated kinase), PKA and HSL(Figure 4). In the absence of isoprenaline, equimolar TLQP-21increased phosphorylation of AMPK and ERK but not PKA

c© The Authors Journal compilation c© 2012 Biochemical Society

A pro-lipolytic role for TLQP-21 peptide 515

Figure 1 Localization of VGF-peptides in the adipose tissue

(A–E) Co-localization of TH-ir (red) and TLQP-21-ir (green) in mouse pWAT. (A and B) Volume of 3 μm (stack of four optical sections, 1 μm apart). (C–E) A single optical section of a portion ofthe nerve fibre seen in (A and B). Yellow puncta in (E) represent the co-localization of TH-ir (C, red) and TLQP-21-ir (D, green). TH-ir was visualized using rabbit anti-TH. TLQP-ir was visualizedusing guinea pig anti-TLQP-21. (F–I) Co-localization of TH-ir (red), TLQP-21-ir (green) and AQEE-30-ir (blue) in rat pWAT. The images represent a volume of 6 μm (stack of seven optical sections,1 μm apart). In (I), the merging of (F) and (G) shows the overlap of TH-ir, TLQP-21-ir and AQEE-30-ir. TH-ir was visualized using mouse anti-TH. TLQP-ir was visualized using rabbit anti-TLQP-21.AQEE-30-ir was visualized using guinea pig anti-AQEE-30.

or HSL (Figure 4). TLQP-21 also failed to phosphorylate Akt(Ser473), PKC (protein kinase C; pan Ser660), p38 (Thr180/Tyr182)and JNK (c-Jun N-terminal kinase; Thr183/Tyr185; results notshown), suggesting that TLQP-21 does not increases intracellularcAMP. In contrast, in the presence of isoprenaline, TLQP-21 potentiated and prolonged phosphorylation of AMPK andHSL (Figure 4), thus providing a mechanistic support to itspotentiation of isoprenaline-induced lipolysis (Figure 3). Overall,in vitro cellular models support a mechanism whereby TLQP-21 potentiates β-AR-induced lipolysis by up-regulating HSLactivity.

Chronic subcutaneous infusion of TLQP-21 decreases adipocytediameter in two mouse models of obesity

Having established that TLQP-21 is present in nerve terminals,binds to adipocytes and potentiates lipolysis in vitro, we setout to investigate whether chronic peripheral in vivo treatmentwith TLQP-21 might affect adipocyte function in two modelsof obesity. In DIO mice, TLQP-21 dose-dependently decreasedthe diameter of pWAT adipocytes [F(2,18) = 49.5, P < 0.0001;Figure 5A]. TLQP-21 also dose-dependently increased enzymaticactivity of TH in both pWAT [F(2,24) = 25.1, P < 0.0001]and scWAT [F(2,24) = 34.5, P < 0.0001] (Figure 5B) as wellas the concentration of noradrenaline in pWAT [F(2,24) =

24.3, P < 0.0001] and scWAT [F(2,24) = 14.2, P < 0.0001](Figure 5C). As expected, TH enzymatic activity and tissuenoradrenaline were positively correlated in both pWAT (r = 0.67,P < 0.001) and scWAT (r = 0.65, P < 0.001). In addition, thedensity of TH-positive nerve fibres was increased in pWAT ofmice receiving the HIGH dose of TLQP-21 (Figure 5D). Increasedsympathetic-related parameters and reduced adipocytes diametersuggest increased sympathetic-driven lipolysis in the adiposetissue [37]. Two lines of evidence suggest that the reduction indiameter and the increased sympathetic tone could be functionallyrelated. First, the high dose of TLQP-21 increased lipolysis in thepWAT expressed as TAG/DNA ratio [41] when compared with theHFD-saline group (Figure 5E; F(2,18) = 4.5. P < 0.05). Secondly,the diameter of pWAT adipocytes was negatively correlated withboth TH (r = − 0.69, P < 0.001) and noradrenaline content(r = − 0.65, P < 0.001), but positively correlated with TAG/DNAratio (r = 0.54, P < 0.05). The TAG/DNA ratio also negativelycorrelates with tissue noradrenaline (r = − 0.52, P < 0.05).

Peripheral TLQP-21 delivery did not affect plasma adrenalinelevels (STD-saline, 1819.8 +− 138.1; HFD-saline, 2419.2 +− 210.9;HFD-TLQP-21-LOW, 1885.3 +− 163.4; HFD-TLQP-21-HIGH,2016.3 +− 182.7 pg/ml), suggesting that the peripheral action ofthis peptide is distinct from its central effects [17].

Obesity has been associated with tissue inflammation [42,43].However, this is unlikely to be the mechanism of TLQP-21 for

c© The Authors Journal compilation c© 2012 Biochemical Society

516 R. Possenti and others

Figure 2 Pharmacological characterization and tissue distribution of TLQP-21-binding sites

(A) The highest 125I-YA-TLQP-21-binding activity was observed in WAT as well as in the adrenals, whereas much lower binding activity occurred in the BAT, CNS, liver and skeletal muscle. Bindingactivity was significantly increased by exposure to 4 weeks of HFD in the adipose tissue only (P < 0.001), whereas no change occurred in the other tissues. (B and C) Representative saturationisotherms and Scatchard plots of 125I-YA-TLQP-21 binding to membranes from pWAT, rpWAT, scWAT and iBAT. (D) The specificity of TLQP-21 peptide to adipocyte membranes was assessed bycompetitive binding studies with 125I-YA-TLQP-21. Scrambled peptide or isoprenaline was unable to significantly displace 125I-YA-TLQP-21 from tissue membranes. Pooled results are representedas means + S.E.M.

two reasons. First, TLQP-21 did not affect density of crown-like structures, a histological evidence of inflamed WAT ([44];Supplementary Figure S3 available at http://www.BiochemJ.org/bj/441/bj4410511add.htm). Secondly, as expected, TNFα mRNAwas increased in obese mice but was not affected byTLQP-21 treatment (STD-saline = 0.04 +− 0.007; HFD-saline= 0.08 +− 0.008; HFD-TLQP-21-LOW = 0.07 +− 0.008; HFD-TLQP-21-HIGH = 0.08 +− 0.009). Finally, TLQP-21 did notaffect BW, adipose fat pad weight which is likely to be dueto the limited infusion time, and the absence of increased foodintake and energy expenditure (body temperature and locomotoractivity) (Supplementary Table S1 available at http://www.BiochemJ.org/bj/441/bj4410511add.htm).

We also provided an independent validation of the in vivoeffects of TLQP-21 in a different model of obesity, CSS (chronicsubordination stress) [29]. In agreement with the results obtainedin DIO mice, chronic subcutaneous treatment of TLQP-21(40 mg/day for 14 days) in CSS mice significantly decreasedthe diameter of pWAT adipocytes (t12 = 2.912, P < 0.05),but increased TH-enzymatic activity (t12 = 2.2, P < 0.05)and noradrenaline content (t12 = 2.7, P < 0.05; Figure 6).BW gain, food intake, adipose fat pad weight and locomotoractivity were not affected (Supplementary Table S2 available athttp://www.BiochemJ.org/bj/441/bj4410511add.htm).

Finally, TLQP-21 did not affect cardiac anatomy and structureas indicated by a preserved mass-to-chamber volume ratio and the

c© The Authors Journal compilation c© 2012 Biochemical Society

A pro-lipolytic role for TLQP-21 peptide 517

Figure 3 Effects of TLQP-21 on lipolysis in mouse adipocytes

(A–C) TLQP-21 did not affect glycerol release in mouse adipocytes, whereas the same ineffective dose, up-regulated isoprenaline (ISO)-induced glycerol release. (D) ISO binding to β-AR is notdisplaced by increasing concentrations of TLQP-21. *P < 0.05, **P < 0.01. Scrambled peptide (LRPS-21). Pooled results are represented as means + S.E.M.

Figure 4 Effect of TLQP-21 on phosphorylation cascade in 3T3-L1 adipocytes

(A) An equimolar concentration (10 μM) of TLQP-21 in the presence or absence of isoprenaline induced a time- and substrate-dependent phosphorylation of AMPK, ERK and HSL which reached amaximum between 15 and 30 min which were used for quantitative experiments in (C–E). (B) cAMP-activated PKA substrates are not affected by TLQP-21. (C) TLQP-21 increased phosphorylationof AMPK and potentiated the ISO-induced effect (Mann–Whitney U test, **P < 0.01, *P < 0.05). (D) TLQP-21 increased phosphorylation of ERK but failed to further up-regulate ISO-inducedeffects (Mann–Whitney U test, **P < 0.01). (E) TLQP-21 did not increase phosphorylation of HSL in the absence of ISO, but potentiated the ISO-induced effect (Mann–Whitney U test, **P < 0.01,*P < 0.05). Pooled results are represented as means + S.E.M.

absence of myocardial fibrosis (Table S1) [32] in DIO mice (TableS1), either iBAT TH activity or noradrenaline concentration inCSS mice (Supplementary Table S2), thus ruling out a generalizedincrease in sympathetic tone to peripheral organs.

In conclusion, results from two independent models ofobesity show that chronic peripheral TLQP-21 delivery decreasedadipocyte size in obese mice and up-regulated sympatheticfunctions within the WAT.

c© The Authors Journal compilation c© 2012 Biochemical Society

518 R. Possenti and others

Figure 5 Chronic infusion of TLQP-21 in DIO mice

TLQP-21 determined a dose-dependent (40–400 μg/day subcutaneously for 14 days) decrease in perigonadal adipocyte diameter (A). TLQP-21 also dose-dependently increased pWAT and scWATTH enzymatic activity (B) and noradrenaline content (C). (D) Density of TH-ir showed a trend to be increased by the HIGH dose of TLQP-21 in pWAT parenchyma. (E) TLQP-21 increased lipolysisin pWAT as shown by a decrease in the TAGs to DNA content ratio in HIGH dose group. The broken line indicates the mean value for control mice fed with STD. Different letters indicate statisticaldifference (P < 0.01) between groups within each fat pad. *P < 0.05. Pooled results are represented as means + S.E.M.

DISCUSSION

In the present paper we report the original identification of anovel peripheral lipolytic sensitizer mechanism mediated by theVGF-derived peptide TLQP-21. Our results shows that TLQP-21-ir is present in sympathetic nerve terminals in the WAT, thatTLQP-21 binds to an as yet unidentified membrane receptor inadipocyte membranes and that ex vivo and in vivo it exerts a pro-lipolytic effect. Chronic peripheral TLQP-21 does not determinea generalized up-regulation of the sympathetic tone as shown bythe normal plasma adrenaline, BAT noradrenaline content andthe absence of ventricular remodelling and myocardial damage[45,46].

We unambiguously identified the presence of VGF peptidesin sympathetic nerve fibres within the WAT. Specifically,we observed the presence of both TLQP-21-ir and VGF C-terminal peptide-ir in sympathetic fibres innervating the visceralWAT. Anti-TLQP-21-ir and C-terminal peptide-ir co-localized

extensively in nerve fibres. However, we also observed fibres thatwere anti-AQEE-30-positive and anti-TLQP-21-negative (but notvice versa), which would suggest a discrete processing of C-terminal peptides [20,47,48] and that TLQP-21 can be cleavedfrom its precursor TLQP-62 within neurite secretory vesicles[17,47]. Furthermore, we have also demonstrated the existence ofa high-affinity binding site for TLQP-21 in adipocyte membranes.Taken together, these results suggest that TLQP-21 releasedlocally from synaptic nerves participates in neuromodulationof lipolysis (Figure 7). Diet-induced obesity led to an increasein TLQP-21 maximum binding capacity in the adipose tissuebut not in other tissues investigated (adrenals, CNS, muscle andliver). To establish an in vivo functional role for TLQP-21 wechronically delivered subcutaneously the peptide for 2 weeksin obese mice. TLQP-21 dose-dependently decreased adipocytediameter and increased TAG lipolysis in obese mice. Decreasedadipocyte diameter and increased lipolysis were paralleled byincreased sympathetic tone in the adipose fat pads as demonstrated

c© The Authors Journal compilation c© 2012 Biochemical Society

A pro-lipolytic role for TLQP-21 peptide 519

Figure 6 Chronic infusion of TLQP-21 in CSS-induced obesity in mice

(A) Schematic diagram of experimental procedure. (B) TLQP-21 (40 μg/day subcutaneously for 14 days) determined a decrease in perigonadal adipocyte diameter, an increase in pWAT TH enzymaticactivity (C) and noradrenaline content (D). *P < 0.05. Pooled results are represented as means + S.E.M.

Figure 7 Proposed mechanism of TLQP-21. TLQP-21 is released by nerve terminals, binds a putative receptor (TLQP-21R) in adipocyte membrane, potentiateβ-AR-induced phosphorylation of AMPK and HSL and increases lipolysis

We hypothesize that intracellular signalling downstream of a putative TLQP-21 receptor activation will encompass activation of a Gq protein and lead to intracellular calcium mobilization. ChronicTLQP-21 treatment in vivo increased WAT TH-activity/immunoreactivity and tissue noradrenaline (NE). We hypothesize two mechanisms of action. Hypothesis 1 (Hyp: 1): a presynaptic TLQP-21receptor would directly affect nerve functions. Hypothesis 2 (Hyp: 2): TLQP-21-receptor-mediated signalling in adipocytes will activate a retrograde signalling to nerve terminals.

by increased enzymatic activity/immunoreactivity of TH, therate-limiting enzyme for the biosynthesis of catecholamine, aswell as the neurotransmitter noradrenaline in both visceral andsubcutaneous fat pads. Taken together these observations establisha new and previously unrecognized role for TLQP-21 in obesityand adipocyte biology.

Three lines of evidence support a mechanism whereby TLQP-21 affects adipocyte diameter via modulation of noradrenalinesignalling. First, correlation analysis supports a connectionbetween TLQP-21-dependent changes in TH/noradrenaline,lipolysis and adipocyte diameter in vivo. Secondly, TLQP-21 dose-dependently up-regulated isoprenaline-induced glycerol

release in mouse adipocytes. Thirdly, TLQP-21 increasesisoprenaline-induced phosphorylation of HSL in 3T3-L1adipocytes. Lipolysis in adipocytes is facilitated via a seriesof regulatory phosphorylation events linking receptor-mediatedincreases in cAMP to the phosphorylation of several keyproteins, including perilipin and HSL, resulting in an increase inTAG hydrolysis. Lipolysis is under multidimensional regulationby hormones and neurotransmitters with noradrenaline beingconsidered the main regulator via β-AR activation [6,38]. Severalinvestigators have shown that PKA-dependent phosphorylationof HSL (Ser659 and Ser660) is required for both its translocationto the lipid droplet and increased hydrolytic activity (e.g. [49]).

c© The Authors Journal compilation c© 2012 Biochemical Society

520 R. Possenti and others

In addition, although the role of AMPK in adipocytes is stillcontroversial [50], AMPK-dependent phosphorylation of HSL(Ser565) also results in activation of the lipase and is necessaryfor translocation [51]. TLQP-21 did not affect lipolysis in theabsence of β-AR activation. However, when TLQP-21 was co-administered with a submaximal lipolytic dose of isoprenaline,glycerol release increased dose-dependently up to ∼80%. Atthe same time TLQP-21 increased phosphorylation of AMPK butnot HSL in the absence of isoprenaline, whereas it potentiatedthe isoprenaline-induced phosphorylation of the same kinase(Figure 7). The TLQP-21 receptor has not been identified so far(see below), but previous results demonstrated that the putativeTLQP-21 receptor would increase Ca2 + [52] and not cAMP(e.g. lack of PKA activation). On the other hand, although thephysiological relevance and role of AMPK in the regulationof lipolysis in adipocytes remain to be fully addressed andcontrasting results have been reported [40,53–54], TLQP-21potentiation of isoprenaline-induced lipolysis might result fromconvergent isoprenaline-induced cAMP-PKA and TQLP-21-induced Ca2 + -AMPK [52,56,57] signalling pathways leading toincreased HSL activity (Figure 7). Future studies should clarifythis issue.

Increased sympathetic innervation to fat pads has beendemonstrated after cold exposure, fasting and chronic angiotensinII treatment [37,58,59]. Local sympathetic denervation ofWAT leads to marked increases in WAT mass and fat cellnumber [60]. Conversely, electrical stimulation of WAT nervespromotes lipid mobilization [61]. In the present study, wedemonstrate that chronic TLQP-21 treatment increased THactivity/immunoreactivity and noradrenaline content in fat padswhile increasing lipolysis. Therefore we speculate that TLQP-21might directly (e.g. binding to a presynaptic receptor on nerveterminals, Hypothesis 1 in Figure 7) or indirectly (e.g. increasingneurotrophin signalling in adipocytes [62,63], Hypothesis 2 inFigure 7) increase sympathetic nerve activity, leading to increasednoradrenergic signalling, and stimulate nerve growth within theWAT. Following the observation that iBAT noradrenaline, bodytemperature and heart morphology are not affected by TLQP-21,we conclude that TLQP-21 selectively modulates sympatheticnoradrenergic pathways in WAT. Understanding the molecularmechanism of this selective modulation of sympathetic nervesrequire further studies.

Besides its functional role in the WAT, one of the major findingsof the present study is the identification of the first receptor-mediated mechanism for a VGF-derived peptide. Research ongranin-derived peptides has so far been constrained by the poorunderstanding of receptor-mediated mechanisms [64]. In thepresent study, we identified for the first time a selective andsaturable TLQP-21-receptor-binding site and showed that thehighest affinity and Bmax is present in white adipocytes. We havealso demonstrated that TLQP-21 binds to adrenals and BAT withan affinity comparable with that of WAT. In contrast, binding tothe muscle, liver and cerebellum was modest. Previous studies,in a neuronal model, established that TLQP-21 significantlyactivated ERK1/2 serine/threonine protein kinases as an effect ofits anti-apoptotic neurotrophic action [52]. In addition, TLQP-21 significantly increased intracellular calcium [as measuredby fura 2/AM (fura 2 acetoxymethyl ester)] in approx. 60%of the recorded neurons [52]. In an attempt to analyse theTLQP-21 signalling in adipocytes we have undertaken a broadapproach which took into account different signalling pathways.We confirmed the phosphorylation of ERK and most interestinglythe prolonged effect on activation of HSL and AMPK by TLQP-21 with β-AR agonist isoprenaline. On the other hand TLQP-21failed to activate the cAMP/PKA pathway since PKA substrates

were not substantially phosphorylated by TLQP-21 [38]. We canspeculate that, similarly to other peptides [65], TLQP-21 willbind a seven transmembrane domain receptor, which leads tothe activation of a Gq protein, stimulation of a phospholipaseB, increase in DAG and PI3K (phosphoinositide 3-kinase)and activation of PKC and intracellular calcium mobilization(Figure 7). However, lacking a molecular characterization ofTLQP-21 receptor, we cannot exclude other types of receptor-mediated effect, such as tyrosine receptor or cytokine receptor,but the mild and weak effect suggests a more neuromodulatoryrole typical of neuropeptides more than the strong effects typicalof grow factors or cytokines (e.g. [66]).

Conclusion

The VGF gene is highly conserved across evolution [13,64].Studies on VGF in humans are limited (see [11,64] for reviews)and to the best of our knowledge no study directly addresseda potential gene association of VGF with obesity or othermetabolic disorders in humans. We report in the present paperthe original characterization of a novel peripheral lipolyticmechanism encompassing the VGF peptide TLQP-21. Ourfindings are remarkable and novel for three different aspects. First,this is the first demonstration of a peripheral biological role fora VGF-derived peptide. Secondly, the present study establishesfor the first time a selective receptor-mediated mechanism for aVGF-derived and one of the first for the entire family of graninpeptides [64]. Thirdly, the present study identified a completelynovel pro-lipolytic mechanism which, in the absence of adversecardiovascular effects, may open promising drug discoveryperspective to counteract obesity. In conclusion, our resultsprovide a mechanistic rationale and translational potential for thedevelopment of TLQP-21 analogues for the treatment of obesity.

AUTHOR CONTRIBUTION

Roberta Possenti and Alessandro Bartolomucci led the design, conceived the study,performed experiments and analysed the results. Roberta Possenti, Bjarne D. Larsen,Jorgen S. Petersen, Anna Moles, Andrea Levi and Alessandro Bartolomucci conceivedan initial part of the study. Giampiero Muccioli, Pamela Petrocchi, Cheryl Cero, AdervilleCabassi, Lucy Vulchanova, Maureen S. Riedl, Monia Manieri, Andrea Frontini, AntonioGiordano, Saverio Cinti, Paolo Govoni, Gallia Graiani, Federico Quaini, Corrado Ghe, ElenaBresciani, Ilaria Bulgarelli, Antonio Torsello, Vittorio Locatelli, Valentina Sanghez, PaolaPalanza and Stefano Parmigiani performed experiments, and analysed and contributed tothe interpretation of results. Alessandro Bartolomucci wrote the paper with input from allother authors.

FUNDING

This work was supported by Zealand Pharma A/S (to A.B. and S.P.), University of MinnesotaMedical School (to A.B.), Regione Lazio (to A.L. and R.P.), University of Milano-Bicocca-Fondo di Ateneo per la Ricerca (to A.T. and V.L.), Cariparma-Credit Agricole Foundation(to A.C.) and Fondazione Enrico and Enrica Sovena (to P.P.).

REFERENCES

1 Friedman, J. M. (2004) Modern science versus the stigma of obesity. Nat. Med. 10,563–569

2 Cooke, D. and Bloom, S. (2006) The obesity pipeline: current strategies in thedevelopment of anti-obesity drugs. Nat. Rev. Drug Discovery 5, 919–931

3 Cinti, S. (2005) The adipose organ. Prostaglandins Leukotrienes Essent. Fatty Acids 73,9–15

4 Bartness, T. J. and Song, C. K. (2007) Sympathetic and sensory innervation of whiteadipose tissue. J. Lipid Res. 48, 1655–1672

5 Sethi, J. K. and Vidal-Puig, A. J. (2007) Adipose tissue function and plasticity orchestratenutritional adaptation. J. Lipid Res. 48, 1253–1262

c© The Authors Journal compilation c© 2012 Biochemical Society

A pro-lipolytic role for TLQP-21 peptide 521

6 Thompson, B. R., Lobo, S. and Bernlohr, D. A. (2010) Fatty acid flux in adipocytes: thein’s and out’s of fat cell lipid trafficking. Mol. Cell. Endocrinol. 318, 24–33

7 Langin, D. (2006) Adipose tissue lipolysis as a metabolic pathway to definepharmacological strategies against obesity and the metabolic syndrome. Pharmacol. Res.53, 482–491

8 Bachman, E. S., Dhillon, H., Zhang, C. Y., Cinti, S., Bianco, A. C., Kobilka, B. K. andLowell, B. B. (2002) βAR signalling required for diet-induced thermogenesis and obesityresistance. Science 297, 843–845

9 Crowley, V. E., Yeo, G. S. and O’Rahilly, S. (2002) Obesity therapy: altering the energyintake-and-expenditure balance sheet. Nat. Rev. Drug Discovery 1, 276–286

10 Leader, B., Baca, Q. J. and Golan, D. E. (2008) Protein therapeutics: a summary andpharmacological classification. Nat. Rev. Drug Discovery 7, 21–39

11 Bartolomucci, A., Pasinetti, G. M. and Salton, S. R. (2010) Granins asdisease-biomarkers: translational potential for psychiatric and neurological disorders.Neuroscience 170, 289–297

12 Levi, A., Ferri, G. L., Watson, E., Possenti, R. and Salton, S. R. (2004) Processing,distribution, and function of VGF, a neuronal and endocrine peptide precursor. Cell. Mol.Neurobiol. 24, 517–533

13 Salton, S. R., Ferri, G. L., Hahm, S., Snyder, S. E., Wilson, A. J., Possenti, R. and Levi, A.(2000) VGF: a novel role for this neuronal and neuroendocrine polypeptide in theregulation of energy balance. Front. Neuroendocrinol. 21, 199–219

14 Chakraborty, T. R., Tkalych, O., Nanno, D., Garcia, A. L., Devi, L. A. and Salton, S. R.(2006) Quantification of VGF- and pro-SAAS-derived peptides in endocrine tissues andthe brain, and their regulation by diet and cold stress. Brain Res. 1089, 21–32

15 Hahm, S., Fekete, C., Mizuno, T. M., Windsor, J., Yan, H., Boozer, C. N., Lee, C., Elmquist,J. K., Lechan, R. M., Mobbs, C. V. and Salton, S. R. (2002) VGF is required for obesityinduced by diet, gold thioglucose treatment, and agouti and is differentially regulated inpro-opiomelanocortin- and neuropeptide Y-containing arcuate neurons in response tofasting. J. Neurosci. 22, 6929–6938

16 Hahm, S., Mizuno, T. M., Wu, T. J., Wisor, J. P., Priest, V. A., Kozak, C. A., Boozer, C. N.,Peng, B., McEvoy, R. C., Good, P. et al. (1999) Targeted deletion of the Vgf gene indicatesthat the encoded secretory peptide precursor plays a novel role in the regulation of energybalance. Neuron 23, 537–548

17 Bartolomucci, A., La Corte, G., Possenti, R., Locatelli, V., Rigamonti, A. E., Torsello, A.,Bresciani, E., Bulgarelli, I., Rizzi, R., Pavone, F. et al. (2006) TLQP-21, a VGF-derivedpeptide, increases energy expenditure and prevents the early phase of diet-inducedobesity. Proc. Natl. Acad. Sci. U.S.A. 103, 14584–14589

18 Toshinai, K., Yamaguchi, H., Kageyama, H., Matsuo, T., Koshinaka, K., Sasaki, K., Shioda,S., Minamino, N. and Nakazato, M. (2010) Neuroendocrine regulatory peptide-2 regulatesfeeding behavior via the orexin system in the hypothalamus. Am. J. Physiol. Endocrinol.Metab. 299, E394–E401

19 Bartolomucci, A., Bresciani, E., Bulgarelli, I., Rigamonti, A. E., Pascucci, T., Levi, A.,Possenti, R., Torsello, A., Locatelli, V., Muller, E. E. and Moles, A. (2009) Chronicintracerebroventricular injection of TLQP-21 prevents high fat diet induced weight gain infast weight-gaining mice. Genes Nutr. 4, 49–57

20 D’Amato, F., Noli, B., Brancia, C., Cocco, C., Flore, G., Collu, M., Nicolussi, P. and Ferri,G. L. (2008) Differential distribution of VGF-derived peptides in the adrenal medulla andevidence for their selective modulation. J. Endocrinol. 197, 359–369

21 Rizzi, R., Bartolomucci, A., Moles, A., D’Amato, F., Sacerdote, P., Levi, A., La Corte, G.,Ciotti, M. T., Possenti, R. and Pavone, F. (2008) The VGF-derived peptide TLQP-21: a newmodulatory peptide for inflammatory pain. Neurosci. Lett. 441, 129–133

22 Matsuo, T., Yamaguchi, H., Kageyama, H., Sasaki, K., Shioda, S., Minamino, N. andNakazato, M. (2010) Localization of neuroendocrine regulatory peptide-1 and -2 inhuman tissues. Regul. Pept. 163, 43–48

23 Riedl, M. S., Braun, P. D., Kitto, K. F., Roiko, S. A., Anderson, L. B., Honda, C. N.,Fairbanks, C. A. and Vulchanova, L. (2009) Proteomic analysis uncovers novel actions ofthe neurosecretory protein VGF in nociceptive processing. J. Neurosci. 29, 13377–13388

24 Vulchanova, L., Riedl, M. S., Shuster, S. J., Buell, G., Surprenant, A., North, R. A. and Elde,R. (1997) Immunohistochemical study of the P2X2 and P2X3 receptor subunits in rat andmonkey sensory neurons and their central terminals. Neuropharmacology 36, 1229–1242

25 Muccioli, G., Ghe, C., Ghigo, M. C., Papotti, M., Arvat, E., Boghen, M. F., Nilsson, M. H.,Deghenghi, R., Ong, H. and Ghigo, E. (1998) Specific receptors for synthetic GHsecretagogues in the human brain and pituitary gland. J. Endocrinol. 157, 99–106

26 Muccioli, G., Pons, N., Ghe, C., Catapano, F., Granata, R. and Ghigo, E. (2004) Ghrelinand des-acyl ghrelin both inhibit isoproterenol-induced lipolysis in rat adipocytes via anon-type 1a growth hormone secretagogue receptor. Eur. J. Pharmacol. 498, 27–35

27 Student, A. K., Hsu, R. Y. and Lane, M. D. (1980) Induction of fatty acid synthetasesynthesis in differentiating 3T3-L1 preadipocytes. J. Biol. Chem. 255, 4745–4750

28 Petrocchi, P., Quaresima, S., Mongiardi, M. P., Severini, C. and Possenti, R. (2010)Activation of kinase phosphorylation by heat-shift and mild heat-shock. Cell. Biol. Int.Rep. 17, e00002

29 Bartolomucci, A., Cabassi, A., Govoni, P., Ceresini, G., Cero, C., Berra, D., Dadomo, H.,Franceschini, P., Dell’Omo, G., Parmigiani, S. and Palanza, P. (2009) Metabolicconsequences and vulnerability to diet-induced obesity in male mice under chronic socialstress. PLoS ONE 4, e4331

30 Cinti, S., Zingaretti, M. C., Cancello, R., Ceresi, E. and Ferrara, P. (2001) Morphologictechniques for the study of brown adipose tissue and white adipose tissue. Methods Mol.Biol. 155, 21–51

31 Emanueli, C., Maestri, R., Corradi, D., Marchione, R., Minasi, A., Tozzi, M. G., Salis, M.B., Straino, S., Capogrossi, M. C., Olivetti, G. and Madeddu, P (1999) Dilated and failingcardiomyopathy in bradykinin B2 receptor knockout mice. Circulation 100, 2359–2365

32 Graiani, G., Lagrasta, C., Migliaccio, E., Spillmann, F., Meloni, M., Madeddu, P., Quaini,F., Padura, I. M., Lanfrancone, L., Pelicci, P. and Emanueli, C. (2005) Genetic deletion ofthe p66Shc adaptor protein protects from angiotensin II-induced myocardial damage.Hypertension 46, 433–440

33 Costoli, T., Bartolomucci, A., Graiani, G., Stilli, D., Laviola, G. and Sgoifo, A. (2004)Effects of chronic psychosocial stress on cardiac autonomic responsiveness andmyocardial structure in mice. Am. J. Physiol. Heart. Circ. Physiol. 286, H2133–H2140

34 Watson, E., Fargali, S., Okamoto, H., Sadahiro, M., Gordon, R. E., Chakraborty, T.,Sleeman, M. W. and Salton, S. R. (2009) Analysis of knockout mice suggests a role forVGF in the control of fat storage and energy expenditure. BMC Physiol. 9, 19

35 Gentile, F., Calı, G., Zurzolo, C., Corteggio, A., Rosa, P., Calegari, F., Levi, A., Possenti, R.,Puri, C., Tacchetti, C. and Nitsch, L. (2004) The neuroendocrine protein VGF is sortedinto dense-core granules and is secreted apically by polarized rat thyroid epithelial cells.Exp. Cell Res. 295, 269–280

36 Possenti, R., Rinaldi, A. M., Ferri, G. L., Borboni, P., Trani, E. and Levi, A. (1999)Expression, processing, and secretion of the neuroendocrine VGF peptides by INS-1cells. Endocrinology 140, 3727–3735

37 Giordano, A., Frontini, A., Murano, I., Tonello, C., Marino, M. A., Carruba, M. O., Nisoli,E. and Cinti, S. (2005) Regional-dependent increase of sympathetic innervation in ratwhite adipose tissue during prolonged fasting. J. Histochem. Cytochem. 53, 679–687

38 Carmen, G. Y. and Victor, S. M. (2006) Signalling mechanisms regulating lipolysis. Cell.Signalling 18, 401–408

39 Omar, B., Zmuda-Trzebiatowska, E., Manganiello, V., Goransson, O. and Degerman, E.(2009) Regulation of AMP-activated protein kinase by cAMP in adipocytes: roles forphosphodiesterases, protein kinase B, protein kinase A, Epac and lipolysis. Cell Signal.21, 760–766

40 Yin, W., Mu, J. and Birnbaum, M. J. (2003) Role of AMP-activated protein kinase in cyclicAMP-dependent lipolysis in 3T3-L1 adipocytes. J. Biol. Chem. 278, 43074–43080

41 Cabassi, A., Coghi, P., Govoni, P., Barouhiel, E., Speroni, E., Cavazzini, S., Cantoni, A. M.,Scandroglio, R. and Fiaccadori, E. (2005) Sympathetic modulation by carvedilol andlosartan reduces angiotensin II-mediated lipolysis in subcutaneous and visceral fat.J. Clin. Endocrinol. Metab. 90, 2888–2897

42 Weisberg, S. P., McCann, D., Desai, M., Rosenbaum, M., Leibel, R. L. and Ferrante, Jr,A. W. (2003) Obesity is associated with macrophage accumulation in adipose tissue. J.Clin. Invest. 112, 1796–1808

43 Xu, H., Barnes, G. T., Yang, Q., Tan, G., Yang, D., Chou, C. J., Sole, J., Nichols, A., Ross,J. S., Tartaglia, L. A. and Chen, H. (2003) Chronic inflammation in fat plays a crucial rolein the development of obesity-related insulin resistance. J. Clin. Invest. 112, 1821–1830

44 Cinti, S., Mitchell, G., Barbatelli, G., Murano, I., Ceresi, E., Faloia, E., Wang, S., Fortier,M., Greenberg, A. S. and Obin, M. S. (2005) Adipocyte death defines macrophagelocalization and function in adipose tissue of obese mice and humans. J. Lipid Res. 46,2347–2355

45 Brum, P. C., Kosek, J., Patterson, A., Bernstein, D. and Kobilka, B. (2002) Abnormalcardiac function associated with sympathetic nervous system hyperactivity in mice. Am. J.Physiol. Heart. Circ. Physiol. 283, H1838–H1845

46 Swynghedauw, B. (1999) Molecular mechanisms of myocardial remodeling. Physiol. Rev.79, 215–262

47 Trani, E., Giorgi, A., Canu, N., Amadoro, G., Rinaldi, A. M., Halban, P. A., Ferri, G.-L.,Possenti, R., Schinina, M. E. and Levi, A. (2002) Isolation and characterization of VGFpeptides in rat brain. Role of PC1/3 and PC2 in the maturation of VGF precursor. J.Neurochem. 81, 565–574

48 Trani, E., Ciotti, T., Rinaldi, A. M., Canu, N., Ferri, G. L., Levi, A. and Possenti, R. (1995)Tissue-specific processing of the neuroendocrine protein VGF. J. Neurochem. 65,2441–2449

49 Su, C. L., Sztalryd, C., Contreras, J. A., Holm, C., Kimmel, A. R. and Londos, C. (2003)Mutational analysis of the hormone-sensitive lipase translocation reaction in adipocytes.J. Biol. Chem. 278, 43615–43619

50 Long, Y. C. and Zierath, J. R. (2006) AMP-activated protein kinase signaling in metabolicregulation. J. Clin. Invest. 116, 1776–1783

51 Holm, C., Østerlund, T., Laurell, H. and Contreras, J. A. (2000) Molecular mechanismsregulating hormone-sensitive lipase and lipolysis. Annu. Rev. Nutr. 20, 365–393

c© The Authors Journal compilation c© 2012 Biochemical Society

522 R. Possenti and others

52 Severini, C., Ciotti, M. T., Biondini, L., Quaresima, S., Rinaldi, A. M., Levi, A., Frank, C.and Possenti, R. (2008) TLQP-21, a neuroendocrine VGF-derived peptide, preventscerebellar granule cells death induced by serum and potassium deprivation. J.Neurochem. 104, 534–544

53 Daval, M., Diot-Dupuy, F., Bazin, R., Hainault, I., Viollet, B., Vaulont, S., Hajduch, E.,Ferre, P. and Foufelle, F. (2005) Anti-lipolytic action of AMP-activated protein kinase inrodent adipocytes. J. Biol. Chem. 280, 25250–25257

54 Greenberg, A. S., Shen, W. J., Muliro, K., Patel, S., Souza, S. C., Roth, R. A. and Kraemer,F. B. (2001) Stimulation of lipolysis and hormone-sensitive lipase via the extracellularsignal-regulated kinase pathway. J. Biol. Chem. 276, 45456–45461

55 Reference deleted56 Hawley, S. A., Pan, D. A., Mustard, K. J., Ross, L., Bain, J., Edelman, A. M., Frenguelli,

B. G. and Hardie, D. G. (2005) Calmodulin-dependent protein kinase kinase-β is analternative upstream kinase for AMP-activated protein kinase. Cell Metab. 2, 9–19

57 Gabler, N. K., Davis, J. E., Walker-Daniels, J. and Spurlock, M. E. (2008) Increasingintracellular calcium in adipocytes increases fatty acid oxidation via the activation ofAMP-activated protein kinase. FASEB J. 22, 147.2

58 Garofalo, M. A., Kettelhut, I. C., Roselino, J. E. and Migliorini, R. H. (1996) Effect of acutecold exposure on norepinephrine turnover rates in rat white adipose tissue. J. Auton.Nerv. Syst. 60, 206–208

59 Brito, N. A., Brito, M. N. and Bartness, T. J. (2008) Differential sympathetic drive toadipose tissues after food deprivation, cold exposure or glucoprivation. Am. J. Physiol.Regul. Integr. Comp. Physiol. 294, R1445–R1452

60 Bowers, R. R., Festuccia, W. T., Song, C. K., Shi, H., Migliorini, R. H. and Bartness, T. J.(2004) Sympathetic innervation of white adipose tissue and its regulation of fat cellnumber. Am. J. Physiol. Regul. Integr. Comp. Physiol. 286, R1167–R1175

61 Correll, J. W. (1963) Adipose tissue: ability to respond to nerve stimulation in vitro.Science 140, 387–388

62 Bullo, M., Peeraully, M. R. and Trayhurn, P. (2005) Stimulation of NGF expression andsecretion in 3T3-L1 adipocytes by prostaglandins PGD2, PGJ2, and �12-PGJ2. Am. J.Physiol. Endocrinol. Metab. 289, E62–E67

63 Peeraully, M. R., Jenkins, J. R. and Trayhurn, P. (2004) NGF gene expression andsecretion in white adipose tissue: regulation in 3T3-L1 adipocytes by hormonesand inflammatory cytokines. Am. J. Physiol. Endocrinol. Metab. 287, E331–E339

64 Bartolomucci, A., Possenti, R., Loh, Y. P., Mahata, S., Fisher-Corbie, R. and Salton, S.(2011) The extended granin family: structure, functions and biomedical implications.Endocr. Rev., doi:10.1210/er.2010-0027

65 Kristiansen, K. (2004) Molecular mechanisms of ligand binding, signaling, and regulationwithin the superfamily of G-protein-coupled receptors: molecular modeling andmutagenesis approaches to receptor structure and function. Pharmacol. Ther. 103,21–80

66 Fleenor, D., Arumugam, R. and Freemark, M. (2006) Growth hormone andprolactin receptors in adipogenesis: STAT-5 activation, suppressors of cytokinesignaling, and regulation of insulin-like growth factor I. Horm. Res. 66,101–110

Received 30 June 2011/25 August 2011; accepted 31 August 2011Published as BJ Immediate Publication 31 August 2011, doi:10.1042/BJ20111165

c© The Authors Journal compilation c© 2012 Biochemical Society

Biochem. J. (2012) 440, 511–522 (Printed in Great Britain) doi:10.1042/BJ20111165

SUPPLEMENTARY ONLINE DATACharacterization of a novel peripheral pro-lipolytic mechanism in mice: roleof VGF-derived peptide TLQP-21Roberta POSSENTI*†, Giampiero MUCCIOLI‡, Pamela PETROCCHI*†, Cheryl CERO§‖, Aderville CABASSI¶,Lucy VULCHANOVA**, Maureen S. RIEDL††, Monia MANIERI‡‡, Andrea FRONTINI‡‡, Antonio GIORDANO‡‡, Saverio CINTI‡‡,Paolo GOVONI§§, Gallia GRAIANI‖‖, Federico QUAINI‖‖, Corrado GHE‡, Elena BRESCIANI¶¶, Ilaria BULGARELLI¶¶,Antonio TORSELLO¶¶, Vittorio LOCATELLI¶¶, Valentina SANGHEZ‖, Bjarne D. LARSEN***, Jorgen S. PETERSEN***1,2,Paola PALANZA‖, Stefano PARMIGIANI‖, Anna MOLES*†††, Andrea LEVI* and Alessandro BARTOLOMUCCI§‖3

*Institute of Cell Biology and Neurobiology, National Research Council and Fondazione Santa Lucia, Rome, Italy, †Department of Neuroscience, University of Roma II – Tor Vergata,Rome, Italy, ‡Department of Drug Science and Technology, University of Turin, Turin, Italy, §Department of Integrative Biology and Physiology, University of Minnesota, Minneapolis,MN 55455, U.S.A., ‖Department of Evolutionary and Functional Biology, University of Parma, Parma, Italy, ¶Department of Internal Medicine, Nephrology and Health Sciences,University of Parma, Parma, Italy, **Department of Veterinary and Biomedical Sciences, University of Minnesota, St. Paul, MN 55108, U.S.A., ††Department of Neuroscience, Universityof Minnesota, Minneapolis, MN 55455, U.S.A., ‡‡Department of Molecular Pathology, Marche Polytechnic University, Ancona, Italy, §§Department of Experimental Medicine,University of Parma, Parma, Italy, ‖‖Department Internal Medicine, University of Parma, Parma, Italy, ¶¶Department of Experimental Medicine, University of Milano-Bicocca, Monza,Italy, ***Zealand Pharma A/S, Glostrup, Denmark, and †††Genomnia srl, Lainate, Italy

Figure S1 VGF expression

(A) VGF cannot be detected in WAT by RT–PCR. Total RNAs from mouse tissues were extractedusing commercially available extraction method (Invitrogen), according to the manufacturer’sinstructions. Briefly, 2 μg of total RNA was reverse-transcribed in RT buffer using anoligonucleotide-dT primer (Promega) and the First Strand cDNA Synthesis Kit. The preparationwas supplemented with 0.01 M DTT (dithiothreitol) and 1 mM of each dNTP and 200 units ofhuman placenta ribonuclease inhibitor (Code No. 2310, TaKaRa) in a final volume of 25 μl.The reaction mixture was incubated at 37◦C for 60 min. VGF and 18S mRNA expression wasanalysed by PCR, i-Cycler 35 cycles (Bio-Rad Laboratories) and run on gel. VGF ampliconapprox. 400 bp; control RNA 18 amplicon approx. 250 bp (C = brain cortex, W = WAT,B = BAT). (B) VGF probed with anti-TLQP-21 antibodies is present in brain cortex, pWAT andiBAT, but not in tissue extract from 3T3-L1 fibroblast and adipocytes or NGF-treated PC12 cells.3T3-L1 cells and NGF-treated PC12 cells washed once with ice-cold PBS and then solubilizedin lysis buffer [50 mM Tris/HCl, pH 7.5, 10 mM EDTA and 1 % Nonidet P40 containing proteaseinhibitor mixture (Sigma)]. Adipose tissue cell preparations were washed six times in PBSthen lysed in cold lysis buffer. Brain extract was lysed in cold lysis buffer. After removingintact nuclei at 5000 rev./min for 10 min, supernatant was analysed for protein content (Pierceprotein reagent). Approx. 2 μg of protein for tissues and 5 μg for cell extracts were mixed withsample reagent buffer (Invitrogen NuPAGE® protocol) spotted on to PVDF membranes and fixedfor 10 min with blotting buffer containing 15 % methanol. Primary antibodies raised againstTLQP-21 peptide (PRIMM) were used at 1:1000 overnight at 4◦C.

Figure S2 Dot blot

The antisera generated against TLQP-21 showed no cross-reactivity with the C-terminal peptideAQEE-30 which is adjacent to TLQP-21. The specificity of the antisera was ascertained by dotblot analysis, in which the antisera bound preferentially to TLQP-21 rather than the extendedfragment TLQP-62.

1 Present address: Merck Serono International S.A., Geneva, Switzerland.2 Jorgen Petersen owns shares in Zealand Pharma A/S.3 To whom correspondence should be addressed (email [email protected]).

c© The Authors Journal compilation c© 2012 Biochemical Society

R. Possenti and others

Figure S3 Crown-like structures

For each section, the density of CLSs (crown-like structures) was obtained by counting thenumber of CLSs in 20 randomly selected area compared with the number of adipocytes and wasexpressed as CLSs/100 adipocytes. Morphometric analysis of CLS density was determined asfollows: two sections from different levels of each pWAT depot were analysed. For each section,the density was obtained counting the number of CLSs in 20 randomly selected areas comparedwith the number of adipocytes and was expressed as CLSs/100 adipocytes.

Table S1 Diet-induced obese mice

HFD-saline HFD-TLQP-21 LOW HFD-TLQP-21 HIGH

Physiological parametersFinal BW (g) 59.06 +− 2 59.76 +− 1.6 59.08 +− 1.6Food intake (treatment phase, average kcal/day) 26.4 +− 1.1 27.8 +− 1.2 26.5 +− 1.0Epidydimal fat pad (g) 2.5 +− 0.16 2.8 +− 0.18 2.8 +− 0.26Body temperature (baseline) 35.5 +− 0.3 35.4 +− 0.3 35.7 +− 0.2Body temperature (treatment phase) 35.6 +− 0.27 35.2 +− 0.18 35.79 +− 0.19Locomotor activity dark phase (baseline, counts) 19503 +− 3260 22262 +− 1549 19744 +− 2415Locomotor activity dark phase (treatment, counts) 16164 +− 1805 19644 +− 1175 19563 +− 2389Locomotor activity light phase (baseline, counts) 6247 +− 1035 6826 +− 1539 6515 +− 1554Locomotor activity light phase (treatment, counts) 6725 +− 1700 7253 +− 1374 6812 +− 1353

Cardiac anatomy and morphometryDiffused fibrosis (volume fraction,%) 0.09 +− 0.02 0.12 +− 0.05 0.26 +− 0.07Perivascular fibrosis (volume fraction,%) 0.28 +− 0.06 0.23 +− 0.02 0.43 +− 0.05Cardiac mass/chamber volume 7.1 +− 0.8 6.9 +− 0.6 6.3 +− 0.3

Table S2 Chronic subordination stress mice

HFD-saline HFD-TLQP-21 LOW

Final BW (g) 43.27 +− 0.7 43.14 +− 1Food intake (baseline phase, average kcal/day) 20.4 +− 1 21.41 +− 1.3Food intake (treatment phase, average kcal/day) 29.5 +− 0.6 29.7 +− 1Epidydimal fat pad (g) 0.94 +− 0.08 0.96 +− 0.09Locomotor activity dark phase (baseline, counts) 20882 +− 4610 20448 +− 2877Locomotor activity dark phase (treatment, counts) 16911 +− 2128 18598 +− 1627Locomotor activity light phase (baseline, counts) 5480 +− 1821 3506 +− 686Locomotor activity light phase (treatment, counts) 7733 +− 1926 6307 +− 1306iBAT TH enzymatic activity (fmol/min per mg of tissue) 45 +− 6 60 +− 14iBAT noradrenaline (pg/mg of protein) 35 +− 5 45 +− 5

Received 30 June 2011/25 August 2011; accepted 31 August 2011Published as BJ Immediate Publication 31 August 2011, doi:10.1042/BJ20111165

c© The Authors Journal compilation c© 2012 Biochemical Society