Cannabinoid Type 1 Receptors Are Upregulated DuringAcute Activation of Brown Adipose TissueMinna Lahesmaa,1,2 Olof Eriksson,3,4 Thorsten Gnad,5 Vesa Oikonen,1 Marco Bucci,1 Jussi Hirvonen,1,6

Kalle Koskensalo,1,2 Jarmo Teuho,2 Tarja Niemi,7 Markku Taittonen,8 Salla Lahdenpohja,1 Mueez U Din,1

Merja Haaparanta-Solin,1,9 Alexander Pfeifer,5 Kirsi A. Virtanen,1,2 and Pirjo Nuutila1,10

Diabetes 2018;67:1226–1236 | https://doi.org/10.2337/db17-1366

Activating brown adipose tissue (BAT) could provide a po-tential approach for the treatment of obesity and metabolicdisease in humans. Obesity is associated with upregulationof the endocannabinoid system, and blocking the canna-binoid type 1 receptor (CB1R) has been shown to causeweight loss and to decrease cardiometabolic risk factors.These effects may be mediated partly via increased BATmetabolism, since there is evidence thatCB1Rantagonismactivates BAT in rodents. To investigate the significance ofCB1R in BAT function, we quantified the density of CB1Rin human and rodent BAT using the positron emissiontomography radioligand [18F]FMPEP-d2 and measuredBAT activation in parallel with the glucose analog [18F]-fluorodeoxyglucose. Activation by cold exposuremarkedlyincreased CB1R density and glucose uptake in the BAT oflean men. Similarly, b3-receptor agonism increased CB1Rdensity in the BAT of rats. In contrast, overweightmenwithreduced BAT activity exhibited decreased CB1R in BAT,reflecting impaired endocannabinoid regulation. Image-guided biopsies confirmed CB1R mRNA expression inhuman BAT. Furthermore, CB1R blockade increased glu-cose uptake and lipolysis of brown adipocytes. Our resultshighlight thatCB1Rs are significant for humanBATactivity,and the CB1Rs provide a novel therapeutic target forBAT activation in humans.

Brown adipose tissue (BAT) has emerged as a potentialtarget to combat obesity and metabolic disease in humans.

Activation of BAT is beneficial for human metabolism ata systemic level because it increases the metabolic rate (1)and is associated with increased lipid and glucose disposal(2–5). BAT can be activated by stimulating the sympatheticnervous system (SNS) with cold exposure or b3-adrenergicreceptor (b3-AR) agonists (1,6,7). However, BAT function isalso controlled by a number of other factors (8–10), whichcurrently are poorly understood and largely unexplored.

The endocannabinoid system (ECS) and specifically thecannabinoid type 1 (CB1) receptors (CB1Rs) control lipidand glucose metabolism (11). The ECS consists of a networkof receptors, their endogenous lipid ligands, and enzymes,which are significant in the brain and many peripheraltissues for modulating complex processes including metab-olism. Activation of CB1R promotes the conservation ofenergy by increasing food intake and inhibiting energyexpenditure and thermogenesis, leading to fat mass expan-sion (12). Conversely, the blockade of CB1R has been foundto decrease body weight and fat mass, improve glucosehomeostasis and insulin sensitivity, and decrease cardiome-tabolic risk factors, making CB1R antagonists potentialdrugs against obesity and diabetes (13). CB1R antagonistrimonabant (SR141716) was previously in clinical use withstrong efficacy for weight loss, but was withdrawn becauseof serious psychiatric adverse effects (14). Recently, novelCB1R antagonists, which act strictly peripherally, have beenfound to activate BAT in rodents, inducing lipolysis and lipid

1Turku PET Centre, University of Turku, Turku, Finland2Turku PET Centre, Turku University Hospital, Turku, Finland3Turku PET Centre, Åbo Akademi, Turku, Finland4Department of Medicinal Chemistry, Uppsala University, Uppsala, Sweden5Institute of Pharmacology and Toxicology, University of Bonn, Bonn, Germany6Department of Radiology, University of Turku, Turku, Finland7Department of Plastic and General Surgery, Turku University Hospital, Turku,Finland8Department of Anesthesiology, Turku University Hospital, Turku, Finland9MediCity Research Laboratories, University of Turku, Turku, Finland10Department of Endocrinology, Turku University Hospital, Turku, Finland

Corresponding author: Pirjo Nuutila, pirjo.nuutila@utu.fi.

Received 10 November 2017 and accepted 2 April 2018.

Clinical trial reg. no. NCT02941172, clinicaltrials.gov.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db17-1366/-/DC1.

© 2018 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, and thework is not altered. More information is available at http://www.diabetesjournals.org/content/license.

1226 Diabetes Volume 67, July 2018

METABOLISM

oxidation, thus improving metabolism (15,16). Activation ofBAT via CB1R antagonism without harmful centrally medi-ated adverse effects could be one way to improve metabolicdisease and combat obesity in humans (17).

CB1R physiology and pathology can be studied in vivousing positron emission tomography (PET). An inverseagonist radioligand for CB1Rs, [18F]FMPEP-d2, has previ-ously been used to quantify the density of CB1R in thehuman brain (18–20). So far, this radioligand has not beenused to study CB1R of other tissues in humans. Recently,we have shown in rodents that [18F]FMPEP-d2 binds toBAT in vivo, indicating high CB1R density in BAT (21). Thereis also preclinical evidence that CB1R and endocannabinoidsare upregulated in BAT after cold or b3-AR activation (22).Here we designed a clinical study aiming to investigate CB1Rdensity in human brown fat using [18F]FMPEP-d2-PETimaging in baseline and cold conditions. We also evaluatedthe effect of obesity on CB1R density in BAT and othertissues including the brain. We found that CB1Rs are upre-gulated when BAT is metabolically activated by cold, but thisresponse is blunted in overweight subjects. CB1R mRNAexpression in human BAT was confirmed from image-guidedbiopsies. In preclinical pharmacological experiments, CB1Rantagonism was shown to increase glucose uptake and lipol-ysis of brown adipocytes (BAs).

RESEARCH DESIGN AND METHODS

Human Studies

Study SubjectsThe clinical PET study included 18 healthy males, who weredivided into a lean group (n = 9) and an overweight group(n = 9), based on the combination of BMI (lean,25 kg/m2),waist circumference (lean ,100 cm), and body fat percent-age (lean,20%). The subjects were all Caucasian men, withan average age of 33 years (range 21–54 years). Subjects weredetermined to be healthy by means of clinical examination,blood tests, and anthropometric measurements (Tables 1and 2). The study protocol was reviewed and approved bythe Ethics Committee of the Hospital District of SouthwestFinland and was conducted according to the principles of the

Declaration of Helsinki. All study subjects provided writteninformed consent (Clinical trial reg. no. NCT02941172).

Study DesignEach subject was studied on 3 separate days after an over-night fast of 8–10 h (Fig. 1A–D). To measure CB1R density,dynamic PET/computed tomography (CT) examinationswere performed using the CB1R inverse agonist radioligand[18F]FMPEP-d2, once in room temperature (RT) conditionsand once during controlled cold exposure. To determinewhether the subject had metabolically active BAT, a PET/magnetic resonance (MR) study using the glucose analog[18F]fluorodeoxyglucose (FDG) was performed during con-trolled cold exposure. The radioligands [18F]FMPEP-d2 and[18F]FDG were synthesized according to standard operatingprocedures of the Turku PET Centre (Turku, Finland) aspreviously described (23,24).

[18F]FDG PET/MR Scanning Protocol and Image AnalysisGlucose uptake of BAT in the neck area was measured duringcold exposure after a standardized 2-h cooling protocol withcooling blankets (25) (SupplementaryData). [18F]FDG 148613 MBq was injected i.v. and a 40-min dynamic PET scanof the cervical region was performed using a 3 T PhilipsIngenuity TF PET/MR scanner (Philips Health Care, Amster-dam, the Netherlands). Eight consecutive modified two-point Dixon sequences (mDIXON scan, Philips Health Care)were used to provide anatomical reference in the whole bodyarea and for calculating signal fat fraction maps (see Supple-mentary Data). Image analysis was performed using Carimas2.9 software (Turku PET Centre). BAT regions of interest(ROIs) were identified in supraclavicular depots of adiposetissue, confirmed with MR signal fat fraction maps, andglucose uptake was quantified using the Patlak linearizationmodel (26).

[18F]FMPEP-d2 PET/CT Scanning Protocoland Image Analysis

To measure CB1R density in different tissues, areas of theneck, abdomen, and brain were scanned using a PET/CTscanner (GE Discovery STE16; General Electric Medical Sys-tems, Milwaukee, WI) once at RT and once during standard-ized cold exposure. [18F]FMPEP-d2 152 6 12 MBq wasinjected i.v., and dynamic scans of the cervical region(60 min), abdominal area (9 min), and the brain (9 min)were conducted. CT scans of each region were performed forphoton attenuation and anatomical reference. Carimas 2.9software was used for image analysis of BAT, white adiposetissue (WAT), and muscle. ROIs of adipose tissue weremanually drawn on the fused PET/CT images, includingonly voxels with CT Hounsfield units within the adiposetissue range (250 to2250Hounsfield units) (27). BAT ROIswere drawn bilaterally in supraclavicular adipose tissuedepots, subcutaneous and intraperitoneal WAT ROIs inabdominal regions, andmuscle ROIs in the trapezius muscle.Regional time-activity curves were calculated from the dy-namic images. Details about data acquisition, PET image

Table 1—Characteristics of study subjects

Anthropometric characteristics Lean Overweight

Number of male participants 9 9

Age (years) 32 6 9 34 6 11

Weight (kg) 77.6 6 8.2 106.5 6 11.1***

BMI (kg/m2) 24.9 6 1.7 32.9 6 4.6***

Waist circumference (cm) 83.2 6 7.1 113.6 6 10.7***

Waist-to-hip ratio 0.9 6 0.04 1.0 6 0.05***

Body fat percentage (%) 19.4 6 3.2 27.6 6 3.7***

Blood pressure systolic (mmHg) 125 6 8 134 6 15

Blood pressure diastolic (mmHg) 76 6 11 80 6 9

Data are the mean 6 SD unless stated otherwise. ***P , 0.001,independent t test.

diabetes.diabetesjournals.org Lahesmaa and Associates 1227

Table2—

Fastingplas

mabioch

emistryva

lues

ofstud

ysu

bjec

tsat

thesc

reen

ingvisitan

dbe

fore

andafterco

ldex

pos

ure

Fastingbioc

hemistry

Normal

rang

e

Lean

Overw

eigh

t

Scree

ning

Beforeco

oling

Afte

rco

oling

Scree

ning

Beforeco

oling

Afte

rco

oling

Gluco

se(m

mol/L)

4–6

5.06

0.6

5.26

0.5

5.26

0.5

5.56

0.5

5.66

0.4

5.36

0.3

Insu

lin(m

U/L)

6.16

2.4

5.76

1.7

4.26

1.8

12.9

65.8#

#13

.46

9.8

11.3

66.1

HbA

1c,%

(mmol/m

ol)

4–6(20–

42)

5.06

0.3(31.16

3.0)

5.16

0.3(31.76

2.6)

Cho

lesterol

(mmol/L)

,5.0

4.26

0.7

4.76

1.2

HDL(m

mol/L)

.1.0

1.56

0.4

1.26

0.2#

LDL(m

mol/L)

,3.0

2.36

0.7

3.16

1.0

Thyroid-stim

ulatingho

rmon

e(m

U/L)

0.3–

4.2

1.86

0.8

1.66

0.9

1.46

0.7

1.86

0.4

1.56

0.6

1.26

0.5**

Triiodo

thyron

ine(pmol/L)

3.1–

6.8

4.86

0.4

4.96

0.6

5.26

0.6

5.46

0.6#

5.26

0.8

5.16

0.7

Thyrox

ine(pmol/L)

11–22

16.1

62.3

15.9

62.3

16.4

62.1*

15.2

61.7

15.1

61.8

15.6

61.5

Triglyce

rides

(mmol/L)

0.45

–2.6

0.76

0.2

0.76

0.3

0.86

0.3**

1.06

0.6

1.06

0.6

1.16

0.6*

Free

fattyac

ids(m

mol/L)

0.46

0.1

0.66

0.1**

0.56

0.2

0.56

0.2

Lactate(m

mol/L)

0.6–

2.4

0.86

0.1

1.16

0.5*

0.96

0.2

1.16

0.5

NA(nmol/L)

0.59

–3.55

2.46

1.0

6.46

2.6***

2.66

0.7

5.16

1.9**

Ene

rgyexpe

nditu

re(kca

l/day)

1,66

46

195

1,90

76

293*

2,10

76

202

2,30

56

207*

Dataarethemea

n6

SD,u

nlessothe

rwiseindica

ted.

#P,

0.05

;##P

,0.01

,ind

epen

dent

ttestc

ompa

ringscreen

ingvalues

oflean

andov

erweigh

tsub

jects.*P

,0.05

;**P

,0.01

;***P=0.00

1,pa

iredttest

compa

ringvalues

with

ingrou

pbe

fore

andafterco

ldex

posu

re.

1228 CB1Rs and Brown Adipose Tissue Diabetes Volume 67, July 2018

Figure 1—A: Clinical PET study design. Each subject (lean n = 9, overweight n = 9) participated in imaging with [18F]FMPEP-d2 PET/CT in RT (B)and cold (C) conditions and with [18F]FDG PET/MRI in cold conditions (D). Coronal PET images from one lean study subject, arrows depictsupraclavicular BAT. E: Cold exposure increased BAT FUR of [18F]FMPEP-d2 in cold. Overweight (OW) subjects had a blunted cold response inBAT. F: BAT glucose uptake correlatedwith BAT [18F]FMPEP-d2 FUR in cold conditions.G: Image-guided humanBAT biopsy samples (lean n =5,overweight n = 4) confirmed cannabinoid receptor mRNA expression in BAT. CB1 mRNA expression was higher than CB2 mRNA expression. Hand I: b3-AR andUCP1mRNA expression in humanBAT. J and K: Pearson correlation between UCP1mRNA expression in BAT and [18F]FMPEP-d2 FUR in BAT (n = 9) (J) and glucose uptake in BAT (n = 8) (K). Data pooled from lean (white triangle) and overweight (black circle) subjects. *P,0.05; **P , 0.01.

diabetes.diabetesjournals.org Lahesmaa and Associates 1229

analysis, determination of plasma input, metabolite correc-tions, modeling of the data, and brain image analysis (usingSPM12 software; Wellcome Trust Centre for Neuroimaging,London, U.K.) are available in the Supplementary Data.CB1R density of a tissue can be determined by calculat-ing the volume distribution (VT) or the fractional uptakerate (FUR) of [18F]FMPEP-d2 from the dynamic PET images.Briefly, VT was calculated from the supraclavicular BATregions by applying the reversible one-tissue compartmentalmodel, using the metabolite-corrected plasma time-activitycurve as amodel input (21). FURwas also calculated for BAT,WAT, muscle, and brain in order to quantitatively compareCB1R density between the tissues.

Animal StudiesSprague Dawley rats (n = 39, male, 8–10 weeks old, weight2286 28 g) were bred at the animal facility of the Universityof Turku. All animal experiments were approved by theRegional State Administrative Agency for Southern Finland(ESAVI/3899/04.10.07/2013), and animal care compliedwith the principles of laboratory animal care and withguidelines of the International Council of Laboratory AnimalScience. The animals were housed at 21 6 3°C, in anatmosphere with 556 15% humidity and with a light periodfrom 6:00 A.M. to 6:00 P.M. All animals had free access to RM1(E) chow (801002; Special Diets Service, Witham, U.K.) andtap water.

Pretreatment of Animals Prior to PET ScanningThe rats were divided into three groups (n = 13), with eachgroup being administered radioligand alone, radioligandafter 10 min of i.v. preadministration of 2 mg/kg b3-ARagonist CL 316243 (Sigma-Aldrich), or 2 mg/kg CB1R an-tagonist rimonabant (Sigma-Aldrich). Briefly, CL 316243was dissolved in 0.9% NaCl prior to injection. Rimona-bant was dissolved in EtOH (150 mL) and then dilutedin biocompatible polar solvent Kleptose (b-cyclodextrin;Apoteket Pharmacy, Uppsala University Hospital, Uppsala,Sweden) to a final EtOH concentration of 20% beforeinjection.

[18F]FMPEP-d2 and [18F]FDG PET Scanning Proceduresand Image AnalysisDetailed imaging procedures, data acquisition, and analysisare described in the Supplementary Data. Briefly, each ratwas sedated, and anesthesia was maintained throughoutthe imaging studies. Animals were positioned in the PET/CT scanner with the BAT in the center of the field of view.[18F]FMPEP-d2 9.7 6 2.0 MBq (corresponding in all casesto,0.1 mg/kg) was administered i.v. in the tail vein (n = 24,or n = 8 from each treatment group). Each rat was thenexamined by PET/CT scanning over 120 min. VT of[18F]FMPEP-d2 in BAT in each rat was calculated inthe PMOD kinetic modeling module (PKIN; PMOD Tech-nologies, Zurich, Switzerland) using the one-tissue compart-ment model as previously described (21).

[18F]FDG 20.16 1.6MBqwas administered i.v. in the tailvein in another group of rats (n = 15, or n = 5 from each

treatment group). Each animal was then examined byPET/CT scanning over 60 min. The glucose utilization (inmicromoles per 100 g per minute) in BAT was estimated byfitting the PET data to a [18F]FDG two-tissue compartmentmodel, using a lumped constant of 1.3.

Ex Vivo Organ Distribution StudiesAfter the PET scanning, the animals were sacrificed (120minafter [18F]FMPEP-d2 injection [n = 24] or 60 min after[18F]FDG injection [n = 15]). Tissues were excised and mea-sured in aWizard2 Automatic GammaCounter (PerkinElmer,Turku, Finland). Measured radioactivity was corrected fordecay, weight of the organ, and background, and it wasexpressed as percentage of the injected dose/gram of tissue.

In Vitro Studies

Human BAT BiopsiesNine of the 18 subjects who underwent PET imaging gaveadditional written consent for acquiring BAT biopsy samplesfrom the supraclavicular neck region (five lean subjects, fouroverweight subjects). In sterile operating room conditionswithan anesthesiologist monitoring the procedure, biopsy samplesof BATwere taken by a plastic surgeon through one small skinincision, using local anesthesia. The anatomical location ofBAT was predetermined with [18F]FDG PET/MR images.After removal, samples were immediately snap frozen intoliquid nitrogen and stored at 270°C until analysis.

Human CellsHuman multipotent adipose-derived stem cells (hMADSs)were obtained from C. Dani (Université Côte d’Azur, CNRS,Inserm, iBV, Faculté de Médecine, Nice, France) and differ-entiated into BAs as described previously (28). Human whiteadipocytes (WAs) were obtained from PromoCell and dif-ferentiated according to manufacturer instructions.

Cannabinoid Receptor Expression AnalysisTotal RNA was isolated using TRIzol (human cells) orNucleoSpin RNA XS (human BAT samples; MACHERY-NAGEL). cDNA synthesis was performed with ProtoScriptFirst Strand cDNA Synthesis Kit (New England BioLabs)according to manufacturer instructions. Human cannabi-noid receptor mRNA expression was analyzed using an ABI7900HT Fast Real-Time PCR System with SybrGreen(Roche), and expression was calculated as 2^-dCt relativeto the TATA-box binding protein. PCR details are describedin the Supplementary Data. mRNA of the CB1R was mea-sured at baseline and after incubation with 1 mmol/L nor-adrenaline (NA) for 16 h.

Lipolysis and Glucose Uptake AssaysTo measure lipolysis, differentiated hMADS were washedtwice with lipolysis medium (DMEM21603; Life Technolo-gies) supplemented with 2% w/v fatty acid–free BSA (Sigma-Aldrich) followed by incubation with lipolysis medium con-taining 100 nmol/L each antagonist/inverse agonist (CB1SR141716A; CB2 SR144528), each agonist (CB1 ACEA; CB2JWH133), and/or 1 mmol/L NA at 37°C and 5% CO2 for 4 h.

1230 CB1Rs and Brown Adipose Tissue Diabetes Volume 67, July 2018

Antagonists were added 20 min prior to NA treatment.All substances were purchased from Tocris Bioscience.Fifty microliters of cell culture media was collected per welland incubated with 50 mL of Free Glycerol Reagent (Sigma-Aldrich) for 5 min at 37°C. Absorption was measured at540 nm. Glycerol release was calculated with glycerol stan-dard (Sigma-Aldrich) and normalized to protein content. Tomeasure glucose uptake, hMADSs were differentiated in12-well plates, starved for 16 h, treated with indicatedsubstances (see above) for 20 min, and glucose uptake assay(catalog #136955; Abcam) was performed according to man-ufacturer instructions. Antagonists were added 20 min priorto NA treatment. Experiments were performed with fourindependent cell cultures.

Statistical MethodsResults are presented as the mean6 SD. A paired two-tailedStudent t test (a = 0.95) was used for assessing differencesbetween baseline and cold conditions within a group. Un-paired two-tailed student t test (a = 0.95) was used forassessing differences between human lean and overweightgroups, and differences between treatment groups of ratsand for cell experiments. Pearson’s correlations were usedto study associations between [18F]FMPEP-d2 FUR in hu-man brain and BAT, BAT glucose uptake, and uncouplingprotein-1 (UCP1) expression. Statistical analyses were per-formed using IBM SPSS 23.0 software.

RESULTS

Cold Increases CB1R Density in Supraclavicular BATof Lean and Overweight MenCompared with baseline conditions, acute cold exposureincreased the FUR of the CB1R radioligand in supraclavicularBAT by threefold in lean subjects (P = 0.006) (Fig. 1E),indicating higher CB1R density. Interestingly, overweightsubjects exhibited low CB1R density at baseline, which in-creased during cold exposure (P = 0.026) (Fig. 1E), but onlyreached the baseline levels of lean subjects. Importantly,the uptake of [18F]FMPEP-d2 in BAT correlated strongly withthe degree of functional BAT activity, as measured by itsglucose uptake in cold (R = 0.89, P , 0.001) (Fig. 1F).Additionally, cold markedly increased energy expenditure inboth lean and overweight subjects (Table 2). The VT and FURof [18F]FMPEP-d2 are both indices of CB1R density in tissue.The BAT VT correlated with the BAT FUR under RT con-ditions (R = 0.80, P, 0.001) and cold conditions (R = 0.73,P = 0.001); hence, the FUR is a suitable index for CB1Rdensity when VT is unavailable (for details, see Supplemen-tary Data).

CB1Rs Are Expressed in Human BATTo verify our imaging findings, we analyzed image-guidedBAT biopsy samples from nine study subjects, specificallystudying BAT markers UCP1, b3-ARs, CB1R, and CB2R. Wefound that mRNA expression of CB1R was higher comparedwith that of CB2R (P = 0.039), but no significant differencewas found between lean and overweight subjects (Fig. 1G).

mRNA expression of UCP1 was significantly higher in leanthan in overweight subjects (P = 0.02) (Fig. 1I). When biopsydata from lean (n = 5) and overweight (n = 4) subjectswere pooled, UCP1 mRNA expression correlated with BAT[18F]FMPEP-d2 uptake (R = 0.73, P = 0.027) (Fig. 1J) as wellas glucose uptake under cold conditions (R = 0.76, P = 0.028)(Fig. 1K).

Overweight and Cold Exposure Cause Changes inHuman Brain CB1R DensityThe central nervous system is essential in mediating can-nabinoid signaling and BAT activation. When we quantifiedthe CB1R density of the brain at RT, CB1R density was 23%lower in overweight than in lean subjects (Fig. 2A and C).Cooling increased CB1R density in the pooled group, (P =0.033), specifically in the areas of the midbrain, pons, andparietal lobe. Furthermore, a positive correlation was foundbetween [18F]FMPEP-d2 uptake in BAT and brain graymatter in cold conditions (P , 0.0005) but not at baseline(Fig. 2D and E, midbrain region).

Overweight Decreases CB1R Density in WATIn overweight subjects, the uptake of [18F]FMPEP-d2 inWATboth subcutaneously and intraperitoneally was significantlylower comparedwith that in lean subjects, whereas uptake inmuscle was similar in both groups (Fig. 2A). Interestingly, inlean subjects, cold exposure induced a tendency of higherCB1R density in abdominal intraperitoneal fat (P = 0.07)(Fig. 2A).

Pharmacological Rodent Studies Confirmed CB1R-Specific Binding of [18F]FMPEP-d2To better understand CB1R physiology and the suitability ofthis tracer for BAT imaging, we conducted further PETstudies in rats during pharmacological activation of BATand blockade of CB1R. Similarly to humans, [18F]FMPEP-d2uptake markedly increased in interscapular BAT after i.v.administration of the b3-AR agonist CL 316243 (Fig. 3A).This clear increase persisted also when taking possiblealterations of [18F]FMPEP-d2 metabolism into account bykinetic modeling (Fig. 3B). The in vivo activation of BAT byCL 316243 was confirmed by increased glucose utiliza-tion (Fig. 3C), and BAT perfusion (Fig. 3D). Preadministra-tion of the CB1R antagonist rimonabant inhibited the[18F]FMPEP-d2 uptake in BAT (Fig. 3B) and brain at basalconditions (P , 0.0001), indicating that the uptake wasreceptor mediated rather than nonspecific. However, CB1Rantagonism with an i.v. dose of 2 mg/kg did not significantlyalter interscapular BAT glucose uptake or modulate perfu-sion (Fig. 3C and D).

Pharmacological Characterization of Human BAsReceptor mRNA expression analyzed in a human BA cell line(hMADS) and primary human WAs revealed significantlyhigher CB1R mRNA expression in BAs compared with WAs.NA stimulation significantly increased the mRNA expressionof CB1R in BAs, but significantly decreased it in WAs. Nochanges were observed in CB2R mRNA expression with NA(Fig. 3E and F).

diabetes.diabetesjournals.org Lahesmaa and Associates 1231

Figure 2—A: [18F]FMPEP-d2 uptake at baseline conditions and during cold exposure of lean and overweight subjects in BAT, subcutaneousWAT(SCWAT), intraperitoneal WAT (IP WAT), muscle, and brain gray matter. *Indicates paired t test between baseline and cold conditions: *P, 0.05;**P, 0.01; (*)P = 0.07. ¤Indicates independent t test between lean and overweight groups: ¤P, 0.05; ¤¤P, 0.01.B: Pearson correlation betweenBMI and [18F]FMPEP-d2 uptake in BAT in cold conditions of lean (triangle) and overweight (circle) subjects.C: PET brain images depicting FUR of[18F]FMPEP-d2 of one lean and one overweight subject at baseline and in cold conditions. D and E: Pearson correlation between [18F]FMPEP-d2FUR values of BAT and the midbrain ROI in cold and baseline conditions in pooled lean and overweight subjects.

1232 CB1Rs and Brown Adipose Tissue Diabetes Volume 67, July 2018

Figure 3—A–D: Pharmacological intervention in rodent BAT. A: Dynamic uptake of [18F]FMPEP-d2 into BAT over time. %ID/g, percentage of theinjected dose/gram of tissue. B: VT values of [18F]FMPEP-d2 uptake and retention in BAT calculated from dynamic PET scan data, withimage-derived arterial, metabolite-corrected input. C: Effect on glucose uptake into BAT. D: Modulation of perfusion based on the [18F]FDGextraction parameter K1, derived from kineticmodeling of [18F]FDGPETdata. b3-AR agonist CL 316243 (blue), CB1R antagonist rimonabant (red),and basal physiology (black). mRNA expression of CB1Rs and CB2Rs in human BAs (E) andWAs (F). NA (1mmol/L) stimulation for 16 h increasesCB1mRNAexpression in BAs but not inWAs. Glucose uptake (G) and glycerol release (H) in a humanBAcell line (hMADS) during pharmacologicalintervention with CB1R antagonist, CB1R agonist, CB2R antagonist, and CB2R agonist, combined with NA (1 mmol/L). *P , 0.05; **P , 0.01;***P , 0.001; ****P , 0.0001. Ago, agonist; Antago, antagonist.

diabetes.diabetesjournals.org Lahesmaa and Associates 1233

We further studied the effect of cannabinoid receptoragonism and antagonism on the function of human BAin vitro. Pharmacological blockade of CB1R increased glucoseuptake, whereas CB1R stimulation, CB2R stimulation, andCB2R blockade had no effect (Fig. 3G). Activation of humanBA with NA increased glucose uptake, and CB1R antagonismfurther enhanced this increase (Fig. 3G). Again, CB1R orCB2R agonism and CB2R antagonism combined with NAhad no effect (data not shown). Lipolysis is another hallmarkof BA activation, so we analyzed whether CB receptors canmodulate glycerol release from human BAs. Blocking theCB1R significantly increased lipolysis, whereas CB1R stim-ulation had no effect (Fig. 3H). CB1R blockade aftermaximalstimulation of BA with NA increased lipolysis, albeit not sig-nificantly. Taken together, these data show that CB receptorsare expressed in human BAs and that CB1R blockade canpositively modulate BA function cell autonomously.

DISCUSSION

To our knowledge, this is the first study investigating CB1Rexpression and function in human BAT in vivo. Our imagingresults show that acute cold exposure markedly increases[18F]FMPEP-d2 binding in supraclavicular BAT depots of leanhealthymen, indicating increased CB1R density when BAT isactivated. CB1R mRNA expression in human BAT was alsoconfirmed from samples obtained by image-guided biopsies.A physiological increase in receptor density suggests that theECS plays a role in the activation of human BAT.

Our results are consistent with those of a previous studyin mice showing that stimulation of BAT by cold and b3-ARagonism increased endocannabinoid levels in BAT (22). More-over, the activation of primary BAs induced transcription ofCnr1, the gene encoding the CB1R (22). CB1R agonismpromotes a positive energy balance (12), hence with theirfindings Krott et al. (22) hypothesized a negative feedbackmechanism, where endocannabinoids, their enzymes, andreceptors are upregulated during BAT activation as a po-tential autoregulatory loop to inhibit thermogenesis. Inline with this, another recent study (29) measured in-creased plasma endocannabinoid levels of healthy leanmen aftermild acute cold exposure. Kantae et al. (29) did notfind correlations between human BAT activity and plasmacannabinoids. However in this study, with our dynamic andhighly sensitive PET imaging methods, we show a positiveassociation between CB1R density and glucose uptake inhuman BAT during cold exposure. UCP1 mRNA expressionmeasured from BAT biopsy samples was also positivelyassociated with CB1R density and glucose uptake in BAT(n = 9, pooled lean and overweight subjects). An acute coldstress may stimulate the ECS to provide more CB1Rs forendocannabinoid binding in BAT in order to inhibit excessenergy expenditure and returnhomeostasis toward apositiveenergy balance.

Endocannabinoids are produced on demand, acting pri-marily in the brain, but exerting important regulatory effectson metabolism in adipose tissue (30). During cold exposure,CB1Rs were also upregulated in the brain, specifically in the

areas of the midbrain, pons, and parietal lobe, which areclosely related to the sympathetic control of BAT function.Furthermore, CB1R density in the midbrain correlated pos-itively with BAT CB1R density in cold, but not warm con-ditions. These results indicate a relationship among the ECS,the SNS, and BAT. The midbrain region includes the hypo-thalamus, which is one key site for controlling homeostasisand energy expenditure and is where endocannabinoids playa major regulatory role (11). The parietal lobe receives andprocesses sensory input, including temperature, whereasthe pons is a significant signaling route controlling auto-nomic functions (31). Temperature is sensed in peripheraltissues, and information is received and processed in theseareas of the central nervous system, after which efferentsympathetic outflow in the form of NA is increased to BAT,resulting in increased thermogenesis (6). Endocannabinoidsignaling in the brain and in BAT seems to be upregulatedacutely in cold to modulate a suitable thermogenic response.

In overweight subjects, CB1R density in BAT was lowand the increase in cold was blunted, merely reaching thebaseline values of lean subjects, possibly reflecting the gen-erally reduced activity of BAT in overweight subjects. How-ever, CB1R density was also significantly lower in abdominalWAT depots and in the brain compared with lean subjects.This suggests a broader downregulation of the CB1Rs,signifying the impaired regulation of the ECS in overweightsubjects, which is in line with the results of previous studies.Excessive activation of the ECS is associated with obesity(12), and a negative association between CB1R density in thebrain and BMI has been reported previously (19,20). Obesesubjects have increased circulating endogenous cannabinoidlevels, whereas mRNA expression of CB1Rs is lower in theWAT of obese subjects compared with lean subjects (32,33).Moreover, higher plasma endocannabinoid levels are relatedto increased abdominal adiposity and cardiometabolic riskfactors (34,35). The CB1R antagonist rimonabant resulted inweight loss in obese patients (13), demonstrating that block-ing the overactive ECS could improve metabolism. Thesefindings exhibit the negative feedback loop of the ECS; chron-ically high amounts of circulating endocannabinoids in obesityare associated with fewer CB1Rs in brain and adipose tissue.

In addition to studying the physiological effects of cold onCB1R signaling in human BAT, we conducted pharmacologicalstudies targeting the CB receptors in rodents and in a humanBA system. Previous preclinical evidence suggests that CB1Rblockade enhances BAT function. In mice, CB1R antagonistsblocked the inhibition of b3-AR, leading to increasedactivation in BAT, lipolysis, and the uptake of fatty acidsfrom plasma (36). Moreover, adipocyte-specific CB1R de-letion results in the browning of WAT, the promotion of athermogenic program, and an increase in alternativelyactivated macrophages, which increase local NA levels(37). In human BA, we showed that CB1R antagonismincreased glucose uptake and lipolysis, but CB1R agonism,CB2R antagonism, or CB2R agonism did not have any effect.During increased NA availability, we alsomeasured increasedCB1R mRNA expression in BA, but not in WA. Our results

1234 CB1Rs and Brown Adipose Tissue Diabetes Volume 67, July 2018

add to the existing evidence that CB1Rs are significant inregulating BA function in an SNS-dependent manner.

Interestingly, in rats we did not measure any appreciableincrease in glucose utilization in BAT up to 60 min afterCB1R antagonist administration, which is in disagreementwith previous findings (36,38,39). This discrepancy may beexplained by differences in methodology, such as the acutei.v. dose given here. Longer lasting plasma exposure to theCB1R antagonist might be required for inducing a recordableincrease in glucose utilization in BAT in rat. Unfortunately,the effect of CB1R antagonism in humans could not beinvestigated in this study for ethical reasons. No CB1R an-tagonist is currently in clinical use for humans and the sub-jects in this study already received the maximal acceptableannual radiation dose considered safe for healthy volunteers.Future studies need to be performed to investigate whetherperipheral CB1R antagonism with novel compounds couldactivate BAT in humans.

One limitation of this study is that the data exhibit onlyshort-term changes in the ECS and BAT, lacking the long-termeffects.When combining our data and others (22,29), it seemsthat, acutely, BAT can produce endocannabinoids and mod-ulate the density of CB1Rs available for binding them inorder to adapt to changes in sympathetic tone. We speculatethat during prolonged cold exposure, Cnr1mRNA expressionmay increase, but some desensitization of ECS signalingmay also occur, similar to long-term changes seen in obesity.Other long-term adaptations such as expansion of BATvolume and activity and browning of WAT may also affectthe response. Cold acclimation studies in humans or repeatedb3-AR stimulation experiments in rodents are required tounderstand possible chronic changes and adaptations of theECS and BAT function.

The CB1R PET radioligand [18F]FMPEP-d2 has previouslybeen used for neuropsychiatric brain studies, and its use asa surrogate biomarker for BAT was previously reported inrats (21), but the potential of using it to study human BATphysiology has previously been unexplored. Here, in acti-vated BAT of both humans and rats, we observed a strongincrease in [18F]FMPEP-d2 uptake. Binding was blockedby rimonabant, confirming a CB1R-mediated mechanism,whereas nonspecific or off-target binding is negligible. Inrats, in addition to increased CB1R binding, we estimate thatan increase in perfusion occurred based on the increase inglucose extraction rate. Increased perfusion of BAT is a conse-quence of increased oxidativemetabolism (25,40), and whereashigher perfusion could also result in more radioligand de-livery to the target tissue, perfusion alone could not explainthe sixfold increase in [18F]FMPEP-d2 binding. Furthermore,we observe active retention of the radioligand in BAT through-out the PET scan, in the form of increased VT (an index ofspecific binding). Therefore, the current PET data can beexplained by transiently increased expression of the CB1R, inresponse to cold or b3-AR–mediated activation in BAT.

In conclusion, acute adrenergic activation of BAT in-creases CB1R density in BAT in humans and rodents. Thisupregulationmay be a negative feedback response of the ECS

to inhibit excessive energy expenditure and restore homeo-stasis, and is likely mediated via the central nervous system.In overweight subjects, CB1R density in BAT, WAT, and thebrain was significantly lower compared with lean subjects,reflecting impairment of the endogenous cannabinoid sys-tem in obesity. We conclude that endocannabinoid signalingvia the CB1R is significant in the activation and regulation ofhuman BAT, and targeting CB1R could provide a prospectiveway to treat obesity and metabolism.

Acknowledgments. The authors thank all of the study subjects enrolled inthis study for cooperation and all the technical staff of Turku PET Centre forassistance.Funding. The study was financially supported by the Academy of Finland (307402,259926, 265204, 292839, and 269977), the Paulo Foundation, the InstrumentariumFoundation, the Turku University Hospital Research Funds, and the European Union(EU FP7 project 278373, DIABAT). The study was conducted within the Finnish Centreof Excellence in Cardiovascular and Metabolic Diseases, which was supported by theAcademy of Finland, the University of Turku, Turku University Hospital, and ÅboAkademi University. T.G. was supported by Deutsche Forschungsgemeinschaft (DFG)grant GN 108/1-1. A.P. was supported by DFG grant RTG 1873.Duality of Interest. No potential conflicts of interest relevant to this article werereported.Author Contributions.M.L. designed the studies, performed clinical imagingexperiments, analyzed data, and prepared the manuscript. O.E. designed the studies,obtained funding, conducted the preclinical imaging experiments and data analysis,and edited the manuscript. T.G. conducted preclinical in vitro studies and data analysisand edited the manuscript. V.O., M.B., K.K., J.T., M.U., and M.H.-S. processed andanalyzed data. J.H. supervised and contributed to image analysis. T.N. and M.T.conducted acquisition of human tissue biopsy samples. S.L. provided radioligands forimaging experiments. A.P. obtained funding, supervised the studies, and edited themanuscript. K.A.V. and P.N. designed the study, obtained funding, supervised theperformance of the studies and data analysis, and edited the manuscript. Allauthors contributed to the critical revision of the manuscript and approved the finalversion. P.N. is the guarantor of this work and, as such, had full access to all the datain the study and takes responsibility for the integrity of the data and the accuracy ofthe data analysis.Prior Presentation. Parts of this study were presented at the KeystoneSymposium on Obesity and Adipose Tissue Biology, Keystone, CO, 22–26 January2017.

References1. Cypess AM, Weiner LS, Roberts-Toler C, et al. Activation of human brownadipose tissue by a b3-adrenergic receptor agonist. Cell Metab 2015;21:33–382. Blondin DP, Tingelstad HC, Noll C, et al. Dietary fatty acid metabolism of brownadipose tissue in cold-acclimated men. Nat Commun 2017;8:141463. Chondronikola M, Volpi E, Børsheim E, et al. Brown adipose tissue activation islinked to distinct systemic effects on lipid metabolism in humans. Cell Metab 2016;23:1200–12064. Chondronikola M, Volpi E, Børsheim E, et al. Brown adipose tissue improveswhole-body glucose homeostasis and insulin sensitivity in humans. Diabetes 2014;63:4089–40995. Orava J, Nuutila P, Lidell ME, et al. Different metabolic responses of humanbrown adipose tissue to activation by cold and insulin. Cell Metab 2011;14:272–2796. Cannon B, Nedergaard J. Brown adipose tissue: function and physiologicalsignificance. Physiol Rev 2004;84:277–3597. Virtanen KA, Lidell ME, Orava J, et al. Functional brown adipose tissue in healthyadults. N Engl J Med 2009;360:1518–15258. Kajimura S, Spiegelman BM, Seale P. Brown and beige fat: physiological rolesbeyond heat generation. Cell Metab 2015;22:546–559

diabetes.diabetesjournals.org Lahesmaa and Associates 1235

9. Loh RKC, Kingwell BA, Carey AL. Human brown adipose tissue as a targetfor obesity management; beyond cold-induced thermogenesis. Obes Rev2017;18:1227–124210. Gnad T, Scheibler S, von Kügelgen I, et al. Adenosine activates brown adiposetissue and recruits beige adipocytes via A2A receptors. Nature 2014;516:395–39911. Silvestri C, Di Marzo V. The endocannabinoid system in energy homeostasisand the etiopathology of metabolic disorders. Cell Metab 2013;17:475–49012. Mazier W, Saucisse N, Gatta-Cherifi B, Cota D. The endocannabinoid system:pivotal orchestrator of obesity and metabolic disease. Trends Endocrinol Metab 2015;26:524–53713. Christopoulou FD, Kiortsis DN. An overview of the metabolic effects of rimo-nabant in randomized controlled trials: potential for other cannabinoid 1 receptorblockers in obesity. J Clin Pharm Ther 2011;36:10–1814. Christensen R, Kristensen PK, Bartels EM, Bliddal H, Astrup A. Efficacy and safetyof the weight-loss drug rimonabant: a meta-analysis of randomised trials. Lancet2007;370:1706–171315. Hsiao WC, Shia KS, Wang YT, et al. A novel peripheral cannabinoid receptor1 antagonist, BPR0912, reduces weight independently of food intake and modulatesthermogenesis. Diabetes Obes Metab 2015;17:495–50416. Takano A, Gulyás B, Varnäs K, et al. Low brain CB1 receptor occupancy bya second generation CB1 receptor antagonist TM38837 in comparison with rimo-nabant in nonhuman primates: a PET study. Synapse 2014;68:89–9717. Kunos G, Tam J. The case for peripheral CB₁ receptor blockade in the treatmentof visceral obesity and its cardiometabolic complications. Br J Pharmacol 2011;163:1423–143118. Terry GE, Hirvonen J, Liow J-S, et al. Biodistribution and dosimetry in humans oftwo inverse agonists to image cannabinoid CB1 receptors using positron emissiontomography. Eur J Nucl Med Mol Imaging 2010;37:1499–150619. Hirvonen J, Goodwin RS, Li C-T, et al. Reversible and regionally selectivedownregulation of brain cannabinoid CB1 receptors in chronic daily cannabis smokers.Mol Psychiatry 2012;17:642–64920. Hirvonen J, Zanotti-Fregonara P, Umhau JC, et al. Reduced cannabinoid CB1receptor binding in alcohol dependencemeasured with positron emission tomography.Mol Psychiatry 2013;18:916–92121. Eriksson O, Mikkola K, Espes D, et al. The cannabinoid receptor-1 is an imagingbiomarker of brown adipose tissue. J Nucl Med 2015;56:1937–194122. Krott LM, Piscitelli F, Heine M, et al. Endocannabinoid regulation in white andbrown adipose tissue following thermogenic activation. J Lipid Res 2016;57:464–47323. Donohue SR, Krushinski JH, Pike VW, et al. Synthesis, ex vivo evaluation, andradiolabeling of potent 1,5-diphenylpyrrolidin-2-one cannabinoid subtype-1 receptorligands as candidates for in vivo imaging. J Med Chem 2008;51:5833–584224. Hamacher K, Coenen HH, Stöcklin G. Efficient stereospecific synthesis ofno-carrier-added 2-[18F]-fluoro-2-deoxy-D-glucose using aminopolyether supportednucleophilic substitution. J Nucl Med 1986;27:235–238

25. U Din M, Raiko J, Saari T, et al. Human brown adipose tissue [(15)O]O2 PETimaging in the presence and absence of cold stimulus. Eur J Nucl Med Mol Imaging2016;43:1878–188626. Patlak CS, Blasberg RG. Graphical evaluation of blood-to-brain transfer constantsfrom multiple-time uptake data. Generalizations. J Cereb Blood Flow Metab 1985;5:584–59027. U Din M, Raiko J, Saari T, et al. Human brown fat radiodensity indicates un-derlying tissue composition and systemic metabolic health. J Clin Endocrinol Metab2017;102:2258–226728. Rodriguez AM, Pisani D, Dechesne CA, et al. Transplantation of a multipotentcell population from human adipose tissue induces dystrophin expression in theimmunocompetent mdx mouse. J Exp Med 2005;201:1397–140529. Kantae V, Nahon KJ, Straat ME, et al. Endocannabinoid tone is higher in healthylean South Asian than white Caucasian men. Sci Rep 2017;7:755830. Cani PD, Plovier H, Van Hul M, et al. Endocannabinoids–at the crossroadsbetween the gutmicrobiota and hostmetabolism. Nat Rev Endocrinol 2016;12:133–14331. Jacobson S, Marcus EM. Neuroanatomy for the Neuroscientist. Boston, MA,Springer, 200832. Blüher M, Engeli S, Klöting N, et al. Dysregulation of the peripheral and adiposetissue endocannabinoid system in human abdominal obesity. Diabetes 2006;55:3053–306033. Engeli S, Böhnke J, Feldpausch M, et al. Activation of the peripheral endo-cannabinoid system in human obesity. Diabetes 2005;54:2838–284334. Côté M, Matias I, Lemieux I, et al. Circulating endocannabinoid levels, abdominaladiposity and related cardiometabolic risk factors in obese men. Int J Obes 2007;31:692–69935. Di Marzo V, Côté M, Matias I, et al. Changes in plasma endocannabinoid levels inviscerally obese men following a 1 year lifestyle modification programme and waistcircumference reduction: associations with changes in metabolic risk factors. Dia-betologia 2009;52:213–21736. Boon MR, Kooijman S, van Dam AD, et al. Peripheral cannabinoid 1 receptorblockade activates brown adipose tissue and diminishes dyslipidemia and obesity.FASEB J 2014;28:5361–537537. Ruiz de Azua I, Mancini G, Srivastava RK, et al. Adipocyte cannabinoid receptorCB1 regulates energy homeostasis and alternatively activated macrophages. J ClinInvest 2017;127:4148–416238. Bajzer M, Olivieri M, Haas MK, et al. Cannabinoid receptor 1 (CB1) antagonismenhances glucose utilisation and activates brown adipose tissue in diet-induced obesemice. Diabetologia 2011;54:3121–313139. Verty ANA, Allen AM, Oldfield BJ. The effects of rimonabant on brown adiposetissue in rat: implications for energy expenditure. Obesity (Silver Spring) 2009;17:254–26140. Muzik O, Mangner TJ, Leonard WR, Kumar A, Janisse J, Granneman JG. 15O PETmeasurement of blood flow and oxygen consumption in cold-activated human brownfat. J Nucl Med 2013;54:523–531

1236 CB1Rs and Brown Adipose Tissue Diabetes Volume 67, July 2018