original article
A novel peripheral cannabinoid receptor 1 antagonist, BPR0912, reduces weight independently of food intake and modulates thermogenesis W.-C. Hsiao† , K.-S. Shia† , Y.-T. Wang, Y.-N. Yeh, C.-P. Chang, Y. Lin, P.-H. Chen, C.-H. Wu, Y.-S. Chao & M.-S. Hung Institute of Biotechnology and Pharmaceutical Research, National Health Research Institutes, Miaoli, Taiwan
Aim: To investigate the in vivo metabolic effects of treatment with BPR0912, a novel and potent peripheral cannabinoid receptor 1 (CB1R) antagonist, on both normal mice and diet-induced obese (DIO) mice. Methods: The acute peripheral effects of BPR0912 administration on gastrointestinal transit and energy metabolism in normal mice were investigated. The effects of chronic BPR0912 treatment were compared with those of rimonabant using DIO mice. Alterations to body weight and biochemical and metabolic variables were determined. Results: Acute treatment with BPR0912 did not alter food intake or energy metabolism, but efficiently reversed CB1R-mediated gastrointestinal delay. Chronic treatment of DIO mice with BPR0912 showed that BPR0912 exerts a food intake-independent mechanism, which contributes to weight loss. Genes involved in 𝛽-oxidation and thermogenesis were upregulated in white adipose tissue (WAT) in addition to increased lipolytic activity, whereas Ucp1 expression was induced in brown adipose tissue (BAT) and body temperature was elevated. Expression of the 𝛽2-adrenoceptor was specifically elevated in both WAT and BAT in a manner dependent on the BPR0912 dose. Lastly, chronic BPR0912 treatment was more efficacious than rimonabant in reducing hepatic triglycerides in DIO mice. Conclusion: BPR0912 exhibits significant in vivo efficacy in inducing food intake-independent weight loss in DIO mice, while tending to reduce their hepatic steatosis. The thermogenic effects of BPR0912, as well as its modulation of protein and gene expression patterns in WAT and BAT, may enhance its efficacy as an anti-obesity agent. The results of the present study support the benefits of the use of peripheral CB1R antagonists to combat metabolic disorders. Keywords: food intake, metabolic disorders, obesity, peripheral CB1R antagonist, thermogenesis, 𝛽2-adrenoceptor Date submitted 13 October 2014; date of first decision 9 November 2014; date of final acceptance 31 January 2015
Introduction Cannabinoid receptor 1 (CB1R) is a G protein-coupled receptor that is activated by lipid mediators, such as endocannabinoids, and regulates a broad range of physiological functions. The high levels of CB1R expression in the brain and the significant influence of this receptor on appetite control and weight management led to the clinical development of rimonabant for anti-obesity, which was subsequently withdrawn because of adverse psychological effects [1,2]; however, in-depth analysis of data from the ‘Rimonabant in Obesity’ trials has shown that CB1R inhibition confined to the body’s periphery has direct beneficial effects on human cardiometabolic risk factors, such as insulin resistance and elevated levels of plasma triglycerides and fasting insulin [3]. These observations have spurred the study of CB1R functions in the peripheral organs [2]. Correspondence to : Ming-Shiu Hung, PhD, Associate Investigator, Institute of Biotechnology and Pharmaceutical Research, National Health Research Institutes, No. 35, Keyan Road, Zhunan, Miaoli 350, Taiwan. E-mail: [email protected] † These
authors contributed equally to this work.
CB1R is expressed in several peripheral organs, including the liver, adipose tissue, skeletal muscle, pancreas and gastrointestinal tract. The level of CB1R expression on cells in these tissues varies according to nutritional and pathophysiological conditions [4–6]. Levels of endocannabinoids are similarly regulated by cellular state and are affected by pathophysiological changes [7–10]. In addition to the effects of peripheral CB1R on lipogenesis and glucose homeostasis, recent studies have shown that CB1R blockade promotes the survival of insulin-secreting 𝛽 cells [11,12]. These data imply that targeting of peripheral CB1R may be a therapeutic strategy for treating metabolic disorders such as obesity, type 2 diabetes and hepatic steatosis. Currently, the availability of peripheral CB1R antagonists is limited [13,14], and most were designed based on the templates of centrally acting CB1R antagonists such as rimonabant and SLV-319. The CB1R antagonist TM38837 (7TM Pharma, Denmark) is the most advanced of these peripherally acting compounds, having completed a phase I clinical trial [15]. TM38837 had no central effects at the anticipated therapeutic dose, setting the foundation for subsequent development of this new class of agents.
ORIGINAL ARTICLE
Diabetes, Obesity and Metabolism 17: 495–504, 2015. © 2015 John Wiley & Sons Ltd
original article
DIABETES, OBESITY AND METABOLISM
(A)
(B)
70
* ###
60 O
50
transit (%)
S
N
N NH
N Cl
###
40 30
***
20 10
Cl
0
F3C
WIN55212-2
-
+
+
+
-
BPR0912
-
-
1
3
3
(C)
(D) 2.0
3.0 2.5
control
**
% BW change
food intake (g/mouse)
3.5
R-10
2.0 1.5
912-3
*
912-10
1.0 0.5 0.0
Night
Day
(F)
control 912-10
25
12 h RQ (VCO2/VO2)
75
12 h VO2 (ml/kg/min)
control R-10
1.0
912-3 912-10
0.5 0.0
24 hr
(E)
50
1.5
1.00 0.75
control 912-10
0.50 0.25 0.00
0
Day
Night
Day
Night
Figure 1. Structure of BPR0912 and its acute effects on normal mice. (A) Structure of BPR0912. (B) BPR0912 (1 or 3 mg/kg, orally) was administered to normal mice before treatment with WIN55212-2 (1 mg/kg, intraperitoneally) to induce a delay in gastrointestinal transit. n = 6 mice per group, except for the group treated with BPR0912 (3 mg/kg) alone (n = 3). (C, D) BPR0912 (3 or 10 mg/kg; 912-3, 912-10), rimonabant (10 mg/kg; R-10) or vehicle control was administered to standard diet-fed normal mice just before the dark period (four cages per group; n = 3 mice per cage). Night, day and 24-h food intake (C) and percent changes in body weight (BW) (D) were measured. (E) Oxygen consumption (left) and respiratory quotient (RQ) values (right) were measured in normal mice with BPR0912 (10 mg/kg; n = 6 mice per group). For (B–E), data are the mean ± standard error of the mean. * Versus control group; # versus WIN55212-2-treated group. * P < 0.05, ** P < 0.01, ***### P < 0.001 by one-way analysis of variance, followed by Tukey’s post hoc test (B–D) or unpaired t-test (E).
We have rationally designed a novel series of aryl alkynylthiophene compounds that are predominantly distributed in the peripheral organs. Among these compounds, BPR0912 (Figure 1A) was found to be a CB1R-selective antagonist with a CB2R/CB1R selectivity index of 71 [16]. BPR0912 possesses potent in vitro properties, with half maximum inhibitory
496 Hsiao et al.
concentration values of 8.5 and 18.5 nM for binding affinity to CB1R and inhibition of GTP binding, respectively. Moreover, BPR0912 has exhibited satisfactory in vivo pharmacokinetic properties, with 28% oral bioavailability and minimal brain permeability (brain to plasma ratio of 3%) [16]. Even at a dose as high as 50 mg/kg, BPR0912 could not reverse
Volume 17 No. 5 May 2015
original article
DIABETES, OBESITY AND METABOLISM
the hypothermia and analgesia induced in mice by treatment with a brain-penetrating CB1R agonist [16]. Based on these properties, we consider BPR0912 to be a potential candidate for targeting CB1R specifically in the periphery. In the present study, we examined the in vivo effects of BPR0912 on variables related to regulating metabolism and obesity in normal and diet-induced obese (DIO) mice.
Materials and Methods Chemicals and Reagents BPR0912 and rimonabant were synthesized in-house as described in Hung et al. [16]. WIN55212-2 was purchased from Cayman Chemical (Ann Arbor, MI, USA). Charcoal and gum Arabic were purchased from Sigma-Aldrich Co. (St Louis, MO, USA). Compound names are listed in Table S1.
Animals and Diets All experiments were conducted in accordance with the Institutional Animal Care and Use Committee guidelines of the National Health Research Institutes, Taiwan. C57BL/6 male mice obtained from the National Laboratory Animal Centre, Taiwan, were housed in a controlled environment (21 ± 2 ∘ C, 50 ± 10% humidity, 12-h light:12-h dark cycle) and had free access to food and water. Mice were fed either a standard diet (SD; LabDiet 5053, 12% fat) or a high fat diet (HFD; Research Diet D12451, 45% fat). BPR0912 and rimonabant were dissolved in vehicle [DMSO/Tween 80/H2 O (1 : 1 : 8, vol/vol/vol)] at the doses indicated in the Figures and given by oral gavage.
administered intraperitoneally at 1 h post-BPR0912 treatment. Gastrointestinal transit was assessed histologically as detailed in File S1 and S2.
Diet-induced Obese Mouse Model Seven-week-old C57BL/6 male mice were fed a HFD or SD for 11 weeks before drug treatment. BPR0912 (3 or 10 mg/kg) or rimonabant (10 mg/kg) was administered orally once daily for 20 days from day 0. Food intake and body weight were measured daily after dosing. On day 20, mice were fasted for 5 h and blood samples were collected for analysis after the mice had been euthanized. Tissues, including the liver, interscapular brown adipose tissue (BAT), and inguinal and epididymal white adipose tissue (WAT) were dissected, weighed and frozen at −80 ∘ C for further analyses.
Body Temperature Measurements Body temperature readings of mice were recorded before drug treatment in the afternoon as detailed in File S1.
Metabolic, Biochemical and Tissue Analyses Kits to measure serum concentrations of glucose, triglycerides (both from Randox, Crumlin, UK), insulin (Mercodia, Uppsala, Sweden), leptin (R&D Systems, Minneapolis, MN, USA), and non-esterified fatty acids (Wako, Osaka, Japan) were used according to the manufacturers’ instructions. Liver triglycerides were determined as previously described [17] and as detailed in File S1.
Western Blotting Food Intake and Weight Assessments in Acute Studies BPR0912 (3 or 10 mg/kg) or rimonabant (10 mg/kg) was administered orally to 9-week-old male C57BL/6 mice (weight 23.3 ± 0.1 g) just before the start of the 12-h dark period. Food intake was calculated manually from food remaining in the container. Body weights were measured at 24 h after drug administration.
Indirect Calorimetry C57BL/6 mice, aged 8 weeks (weight 25.3 ± 0.3 g) and acclimatized to individual metabolic chambers, were administered vehicle or BPR0912 (10 mg/kg) orally before the start of the 12-h light period. Mice were subsequently monitored in the chambers for 24 h. Metabolic rate (VO2 , ml/kg/min) and respiratory quotient [calculated as VCO2 (ml/kg/min)/VO2 (ml/kg/min)] were measured using the PhysioScan Metabolic System (AccuScan Instruments, Columbus, OH, USA) according to the manufacturer’s instructions.
Gastrointestinal Transit BPR0912 (1 or 3 mg/kg) was administered orally to 8-week-old male C57BL/6 mice (weight 23.6 ± 0.4 g) after 16 h fasting. WIN55212-2 (1 mg/kg in 1% DMSO in saline) was
Volume 17 No. 5 May 2015
Adipose tissues were homogenized in RIPA lysis buffer (50 mM Tris-HCl, 0.25% sodium deoxycholate, 150 mM NaCl, 1% NP-40) with the addition of protease and phosphatase inhibitor cocktails (Roche, Indianapolis, IN, USA). The antibodies used are listed in Table S2.
RNA Isolation and Quantitative PCR
®
Total RNA was extracted from frozen tissues with TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized from DNase I-treated total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Expression patterns of specific genes in BAT and WAT were determined by quantitative PCR (Applied Biosystems 7900). The expression of the peptidylprolyl isomerase B (Ppib) gene was measured as the endogenous control gene for quantitation of relative mRNA levels. The primer sequences are shown in Table S3.
Statistical Analysis Data are expressed as mean ± standard error of the mean. Statistical analysis of experiments was performed using Prism software (GraphPad Software, San Diego, CA, USA), and comparisons were assessed using an unpaired t-test or one-way analysis of variance, followed by Tukey’s post hoc test.
doi:10.1111/dom.12447 497
original article Results Acute Effects of BPR0912 on Gastrointestinal Transit, Food Intake and Energy Metabolism in Normal Mice The objective of the present study was to investigate the in vivo properties of BPR0912. When we administered the cannabinoid receptor agonist WIN55212-2 to normal C57BL/6 mice, this agent was effective in delaying the gastrointestinal transit mediated by intestinal CB1R [18,19], as expected (Figure 1B, lane 2). This delay was efficiently reversed by prior application of BPR0912 at either 1 or 3 mg/kg (Figure 1B, lanes 3 and 4). Similar results were obtained using the ACEA (CB1R selective agonist)-induced gastrointestinal delay model (Figure S1). Treatment with 3 mg/kg BPR0912 alone mildly increased gastrointestinal transit (Figure 1B, lane 5), suggesting that BPR0912 functions as an inverse agonist of CB1R in vivo. To directly examine BPR0912’s effects on night and day food intake and on 1-day body weight change, 3 or 10 mg/kg BPR0912 or 10 mg/kg rimonabant were administered to normal mice just before the start of the 12-h dark period, and 12-h food intake and 24-h body weight change were examined. The results showed that, unlike rimonabant, which significantly decreased night food intake but increased day food intake, BPR0912 did not alter night or day food intake (Figure 1C, left). Accordingly, 24-h total food intake values were similar to controls for all groups of treated mice (Figure 1C, right). Although BPR0912 treatment at either 3 or 10 mg/kg tended to decrease body weight, unlike rimonabant, these differences were not statistically significant (Figure 1D). Lastly, acute BPR0912 treatment did not alter metabolic rate or respiratory quotient (Figure 1E). These data suggest that BPR0912 can increase gastrointestinal transit without affecting overall food intake, body weight or metabolism.
Effects of Chronic BPR0912 Treatment on DIO Mice To examine the effects of chronic BPR0912 treatment on mice with a metabolic disorder, we administered this compound to mice fed for 11 weeks on a HFD to induce obesity (DIO mice). As in our study of acute BPR0912 treatment of normal mice, BPR0912 at a dose of 3 or 10 mg/kg did not suppress food intake by DIO mice on the first day of treatment and exerted only mild dose-dependent suppression on food intake from days 4 to 6 (Figure 2A). By contrast, treatment of DIO mice with 10 mg/kg rimonabant resulted in a strong suppressive effect that lasted until day 3. No dramatic suppression of food intake was observed for either agent after day 6 until the end of the treatment period. In addition to daily food intake, cumulative intake was analysed. Daily BPR0912 treatment for 6 days reduced cumulative food intake but only when given at the 10 mg/kg dose (Figure 2B). Six days of 10 mg/kg BPR0912 also induced a greater weight reduction than the 3 mg/kg dose (Figure 2B); however, by day 13, the cumulative food intake by both BPR0912-treated groups was similar to that of the rimonabant-treated mice and significantly lower than in controls (Figure 2C). After day 13 until the end of the treatment, no further change in food intake was observed, and BPR0912-treated mice exhibited a dose-dependent change in
498 Hsiao et al.
DIABETES, OBESITY AND METABOLISM
body weight (Figure 2C, D); thus, BPR0912 has no more severe an effect on food intake than does rimonabant and is effective in reducing body weight. When we extended our daily body weight measurements to 19 days, we found that BPR0912 treatment steadily and progressively reduced body weight in DIO mice, in contrast to the abrupt initial weight loss followed by steady-state maintenance of this weight in rimonabant-treated mice (Figure 3A). As a result, both abdominal (e.g. epididymal) and subcutaneous (e.g. inguinal) fat pads decreased in mass in a dose-dependent manner in BPR0912-treated mice (Figure 3B). At a dose of 10 mg/kg, DIO mice treated with BPR0912 or rimonabant exhibited a similar loss of fat mass, although BPR0912 induced a greater body weight loss (Figure 3B). Serum levels of leptin showed a dose-dependent decrease in BPR0912-treated DIO mice, as did serum levels of insulin and triglycerides (Figure 3C); however, only negligible changes in serum glucose and free fatty acids were observed in these mice (Figure 3C). Strikingly, 10 mg/kg BPR0912 markedly reduced hepatic lipids, and the efficacy of this agent at 3 mg/kg was similar to that of rimonabant at 10 mg/kg (Figure 3C).
BPR0912 Enhances Lipolysis and 𝛽-oxidation in White Adipose Tissue The loss of fat pad mass in BPR0912-treated DIO mice prompted us to examine lipolytic enzymes in WAT. Protein kinase A (PKA) activity, demonstrated by phosphorylation of PKA substrates, was slightly elevated in the WAT of BPR0912-treated DIO mice, although the difference from controls was not statistically significant (Figure 4A); however, hormone-sensitive lipase (HSL) was highly phosphorylated in BPR0912- and rimonabant-treated DIO mice compared with controls (Figure 4B). We then examined mRNA levels of genes related to triglyceride lipolysis and fat oxidation. The mRNA expression levels of all three 𝛽-adrenoceptors (encoded by the Adrb1, Adrb2 and Adrb3 genes), as well as peroxisomal proliferator-activated receptor 𝛼 (Ppar𝛼), Ppar𝛾 coactivator 1𝛼 (Pgc1𝛼), and mitochondrial uncoupling protein 1 (Ucp1), were significantly upregulated in DIO mice treated with 10 mg/kg BPR0912 compared with untreated controls (Figure 4C). In the rimonabant-treated group, expression levels of Adrb3, Ucp1 and Pgc1𝛼 were all significantly elevated (Figure 4C); thus, BPR0912 enhances lipolysis and 𝛽-oxidation in the WAT of DIO mice, accounting (at least in part) for the observed reduction in fat mass.
BPR0912 Induces Thermogenesis To determine the effects of BPR0912 on thermogenesis, the core body temperatures of DIO mice that had been treated with 3 or 10 mg/kg BPR0912 for 17–18 days were measured on each afternoon, a time of day when the mice would be in a basal metabolic state. We found that BPR0912 dose-dependently increased body temperature, with an elevation of 0.8 ∘ C over controls observed in mice that had been chronically treated with 10 mg/kg BPR0912 (Figure 5A). This increase was markedly greater than that observed in rimonabant-treated DIO mice.
Volume 17 No. 5 May 2015
original article
DIABETES, OBESITY AND METABOLISM
(B)
4
40
15
Control R-10
2
912-3 912-10
1 0 1
3
5
7
9
11
13
15
17
**
9 6
*
**
*
12
30
BW on Day 6
3
6-day cumulative intake (g)
Daily food intake (g)
(A)
control R-10
20
912-3 912-10
10
3
0
0
Day
(D) 50
Control R-10 912-3 912-10
40 30
6 3
BW change (%)
cumulative food intake (g)
(C)
20 10
1
3
5
7
9
11
13
15
17
Days
R-10
–3
912-3
–6
912-10
–9
***
–12 –15
0
control
0
–18
*** *** $$
Figure 2. Effects of chronic BPR0912 treatment on food intake and weight loss in diet-induced obese (DIO) mice. BPR0912 at 3 or 10 mg/kg (912-3, 912-10), or rimonabant at 10 mg/kg (R-10), was administered orally to DIO mice (weight 35.8 ± 0.6 g) once daily for 18 days. Daily food intake (A), 6-day cumulative food intake and body weight (BW) on day 6 (B), cumulative food intake (C), and BW change on day 18 (D) were measured. Data are mean ± standard error of the mean (four cages per group; n = 2–3 mice per cage). * Versus DIO control group; $ 912-3 versus 912-10. * P < 0.05; **$$ P < 0.01; *** P < 0.001 by one-way analysis of variance, followed by Tukey’s post hoc test.
Ucp1 upregulation is associated with thermogenesis, which prompted us to investigate the in vivo relevance of our observation that Ucp1 was elevated in the BAT of BPR0912-treated DIO mice. As shown in Figure 5B, Ucp1 protein was significantly increased in DIO mice treated with 10 mg/kg BPR0912, similar to the tendency observed in the rimonabant-treated group. When we examined the mRNA expression of 𝛽-adrenoceptor genes, levels of all three 𝛽-adrenoceptor mRNAs were increased in DIO mice treated with 3 mg/kg BPR0912 (Figure 5C); however, in DIO mice treated with 10 mg/kg BPR0912 or rimonabant, only the Adrb2 adrenoceptor mRNA was specifically upregulated (Figure 5C). These data suggest that an activation of Ucp1 that is mediated primarily by 𝛽2-adrenoceptors contributes to the thermogenesis induced by BPR0912.
Discussion Second-generation CB1R antagonists with peripheral distribution may represent powerful potential therapeutics for patients with cardiometabolic disorders. Naturally, the validation of any such chemical agent requires pharmacological evidence, biodistribution information and the definition of the physiological responses elicited in vivo in animal models. In rodents, central activation of CB1R induces a tetrad of responses comprising hypothermia, analgesia, hypolocomotion and catalepsy [20]. In previous work, we showed that our
Volume 17 No. 5 May 2015
candidate peripheral CB1R antagonist BPR0912 penetrated the blood–brain barrier to only a limited extent and did not reverse CP55940-induced hypothermia and analgesia in mice, even at a dose as high as 50 mg/kg [16]. Another outcome of peripheral CB1R activation is depression of gastrointestinal motility mediated by enteric neurons [21,22]. We have shown previously that mice receiving BPR0912 have a high oral drug exposure [16]. In the present study, we found that BPR0912 was able to reverse a peripheral CB1R-mediated delay in gastrointestinal transit at a dose as low as 1 mg/kg. The marked difference in the dose of BPR0912 required to affect these central and peripheral responses, for example, 50 mg/kg versus 1 mg/kg, indicates that BPR0912 is an effective peripheral CB1R antagonist. The biased biodistribution of BPR0912 was also evident in the food intake experiments of the present study. Our previous examination of a different peripheral CB1R antagonist, BPR0697, suggested that acute blockade of peripheral CB1R was not involved in the food intake suppression observed in animals that were on a SD and treated with this agent [17]. This result has been further supported by investigations of JD5037 [23], a peripheral CB1R antagonist, and also by our examination of BPR0912 in the present study. Vijayakumar et al. [17] showed in mice on SD that, unlike rimonabant, BPR0697 at a brain to plasma ratio of 4.3% was not able to elevate the metabolic rate, although it was capable of shifting
doi:10.1111/dom.12447 499
original article Control
38
R-10
36
912-3 912-10
34 32
*
5 4 3
***
***
2 1
0 1 2 3 4 5 6 7 8 9 10111213141516171819
(C)
Day
5 0 1 2 3 4 5 6 7 8 9 10111213141516171819 Day
2.5
*
2.0 1.5 1.0
Control R-10
2
***
912-3
***
912-10
1
60
***
***
* 200
100
50 control
40
R-10 912-3
30
0.0
0
90 80 70 60 50 40 30 20 10 0
0.8
912-10
*
20 10
0.5
–5
*
300
3.0
insulin (ng/ml)
10
3.5
glucose (mg/dL)
28
3
0
0
30
**
**
0
–10
–20 –25
0.7
*
0.6 0.5 0.4 0.3 0.2 0.1 0.0
liver TG (mg/g)
–15
triglycerides (mg/dL)
BW Change (%)
6
NEFA(mEq/L)
BW (g)
40
4
7
leptin (ng/ml)
42
Inguinal fat (% BW)
(B)
44
Epididymal fat (% BW)
(A)
DIABETES, OBESITY AND METABOLISM
90 80 70 60 50 40 30 20 10 0
**
** ***
Figure 3. Effects of chronic BPR0912 treatment on metabolic variables of diet-induced obese (DIO) mice. BPR0912 at 3 or 10 mg/kg (912-3, 912-10), or rimonabant at 10 mg/kg (R-10), was administered orally to DIO mice (weight 38.7 ± 0.6 g) once daily for 19 days. Body weight (BW) and percent BW change (A), epididymal and inguinal fat mass values (B), and serum levels of insulin, glucose, leptin, triglycerides, non-esterified fatty acids (NEFA) and hepatic triglycerides (TG) (C) were measured. Data are the mean ± standard error of the mean (n = 6 mice per group). * P < 0.05, ** P < 0.01, *** P < 0.001 versus DIO control group by one-way analysis of variance, followed by Tukey’s post hoc test.
the respiratory quotient toward lipid oxidation as effectively as rimonabant. In the present study, we found that BPR0912 at a brain to plasma ratio of 3% did not have a significant effect on energy metabolism. These results suggest that peripheral CB1R antagonists lack the ability to acutely modulate the food intake and metabolism of normal mice, which further differentiates these agents from central CB1R antagonists. We observed that the suppressive effect of rimonabant on food intake lasted
A novel peripheral cannabinoid receptor 1 antagonist, BPR0912, reduces weight independently of food intake and modulates thermogenesis W.-C. Hsiao† , K.-S. Shia† , Y.-T. Wang, Y.-N. Yeh, C.-P. Chang, Y. Lin, P.-H. Chen, C.-H. Wu, Y.-S. Chao & M.-S. Hung Institute of Biotechnology and Pharmaceutical Research, National Health Research Institutes, Miaoli, Taiwan
Aim: To investigate the in vivo metabolic effects of treatment with BPR0912, a novel and potent peripheral cannabinoid receptor 1 (CB1R) antagonist, on both normal mice and diet-induced obese (DIO) mice. Methods: The acute peripheral effects of BPR0912 administration on gastrointestinal transit and energy metabolism in normal mice were investigated. The effects of chronic BPR0912 treatment were compared with those of rimonabant using DIO mice. Alterations to body weight and biochemical and metabolic variables were determined. Results: Acute treatment with BPR0912 did not alter food intake or energy metabolism, but efficiently reversed CB1R-mediated gastrointestinal delay. Chronic treatment of DIO mice with BPR0912 showed that BPR0912 exerts a food intake-independent mechanism, which contributes to weight loss. Genes involved in 𝛽-oxidation and thermogenesis were upregulated in white adipose tissue (WAT) in addition to increased lipolytic activity, whereas Ucp1 expression was induced in brown adipose tissue (BAT) and body temperature was elevated. Expression of the 𝛽2-adrenoceptor was specifically elevated in both WAT and BAT in a manner dependent on the BPR0912 dose. Lastly, chronic BPR0912 treatment was more efficacious than rimonabant in reducing hepatic triglycerides in DIO mice. Conclusion: BPR0912 exhibits significant in vivo efficacy in inducing food intake-independent weight loss in DIO mice, while tending to reduce their hepatic steatosis. The thermogenic effects of BPR0912, as well as its modulation of protein and gene expression patterns in WAT and BAT, may enhance its efficacy as an anti-obesity agent. The results of the present study support the benefits of the use of peripheral CB1R antagonists to combat metabolic disorders. Keywords: food intake, metabolic disorders, obesity, peripheral CB1R antagonist, thermogenesis, 𝛽2-adrenoceptor Date submitted 13 October 2014; date of first decision 9 November 2014; date of final acceptance 31 January 2015
Introduction Cannabinoid receptor 1 (CB1R) is a G protein-coupled receptor that is activated by lipid mediators, such as endocannabinoids, and regulates a broad range of physiological functions. The high levels of CB1R expression in the brain and the significant influence of this receptor on appetite control and weight management led to the clinical development of rimonabant for anti-obesity, which was subsequently withdrawn because of adverse psychological effects [1,2]; however, in-depth analysis of data from the ‘Rimonabant in Obesity’ trials has shown that CB1R inhibition confined to the body’s periphery has direct beneficial effects on human cardiometabolic risk factors, such as insulin resistance and elevated levels of plasma triglycerides and fasting insulin [3]. These observations have spurred the study of CB1R functions in the peripheral organs [2]. Correspondence to : Ming-Shiu Hung, PhD, Associate Investigator, Institute of Biotechnology and Pharmaceutical Research, National Health Research Institutes, No. 35, Keyan Road, Zhunan, Miaoli 350, Taiwan. E-mail: [email protected] † These
authors contributed equally to this work.
CB1R is expressed in several peripheral organs, including the liver, adipose tissue, skeletal muscle, pancreas and gastrointestinal tract. The level of CB1R expression on cells in these tissues varies according to nutritional and pathophysiological conditions [4–6]. Levels of endocannabinoids are similarly regulated by cellular state and are affected by pathophysiological changes [7–10]. In addition to the effects of peripheral CB1R on lipogenesis and glucose homeostasis, recent studies have shown that CB1R blockade promotes the survival of insulin-secreting 𝛽 cells [11,12]. These data imply that targeting of peripheral CB1R may be a therapeutic strategy for treating metabolic disorders such as obesity, type 2 diabetes and hepatic steatosis. Currently, the availability of peripheral CB1R antagonists is limited [13,14], and most were designed based on the templates of centrally acting CB1R antagonists such as rimonabant and SLV-319. The CB1R antagonist TM38837 (7TM Pharma, Denmark) is the most advanced of these peripherally acting compounds, having completed a phase I clinical trial [15]. TM38837 had no central effects at the anticipated therapeutic dose, setting the foundation for subsequent development of this new class of agents.
ORIGINAL ARTICLE
Diabetes, Obesity and Metabolism 17: 495–504, 2015. © 2015 John Wiley & Sons Ltd
original article
DIABETES, OBESITY AND METABOLISM
(A)
(B)
70
* ###
60 O
50
transit (%)
S
N
N NH
N Cl
###
40 30
***
20 10
Cl
0
F3C
WIN55212-2
-
+
+
+
-
BPR0912
-
-
1
3
3
(C)
(D) 2.0
3.0 2.5
control
**
% BW change
food intake (g/mouse)
3.5
R-10
2.0 1.5
912-3
*
912-10
1.0 0.5 0.0
Night
Day
(F)
control 912-10
25
12 h RQ (VCO2/VO2)
75
12 h VO2 (ml/kg/min)
control R-10
1.0
912-3 912-10
0.5 0.0
24 hr
(E)
50
1.5
1.00 0.75
control 912-10
0.50 0.25 0.00
0
Day
Night
Day
Night
Figure 1. Structure of BPR0912 and its acute effects on normal mice. (A) Structure of BPR0912. (B) BPR0912 (1 or 3 mg/kg, orally) was administered to normal mice before treatment with WIN55212-2 (1 mg/kg, intraperitoneally) to induce a delay in gastrointestinal transit. n = 6 mice per group, except for the group treated with BPR0912 (3 mg/kg) alone (n = 3). (C, D) BPR0912 (3 or 10 mg/kg; 912-3, 912-10), rimonabant (10 mg/kg; R-10) or vehicle control was administered to standard diet-fed normal mice just before the dark period (four cages per group; n = 3 mice per cage). Night, day and 24-h food intake (C) and percent changes in body weight (BW) (D) were measured. (E) Oxygen consumption (left) and respiratory quotient (RQ) values (right) were measured in normal mice with BPR0912 (10 mg/kg; n = 6 mice per group). For (B–E), data are the mean ± standard error of the mean. * Versus control group; # versus WIN55212-2-treated group. * P < 0.05, ** P < 0.01, ***### P < 0.001 by one-way analysis of variance, followed by Tukey’s post hoc test (B–D) or unpaired t-test (E).
We have rationally designed a novel series of aryl alkynylthiophene compounds that are predominantly distributed in the peripheral organs. Among these compounds, BPR0912 (Figure 1A) was found to be a CB1R-selective antagonist with a CB2R/CB1R selectivity index of 71 [16]. BPR0912 possesses potent in vitro properties, with half maximum inhibitory
496 Hsiao et al.
concentration values of 8.5 and 18.5 nM for binding affinity to CB1R and inhibition of GTP binding, respectively. Moreover, BPR0912 has exhibited satisfactory in vivo pharmacokinetic properties, with 28% oral bioavailability and minimal brain permeability (brain to plasma ratio of 3%) [16]. Even at a dose as high as 50 mg/kg, BPR0912 could not reverse
Volume 17 No. 5 May 2015
original article
DIABETES, OBESITY AND METABOLISM
the hypothermia and analgesia induced in mice by treatment with a brain-penetrating CB1R agonist [16]. Based on these properties, we consider BPR0912 to be a potential candidate for targeting CB1R specifically in the periphery. In the present study, we examined the in vivo effects of BPR0912 on variables related to regulating metabolism and obesity in normal and diet-induced obese (DIO) mice.
Materials and Methods Chemicals and Reagents BPR0912 and rimonabant were synthesized in-house as described in Hung et al. [16]. WIN55212-2 was purchased from Cayman Chemical (Ann Arbor, MI, USA). Charcoal and gum Arabic were purchased from Sigma-Aldrich Co. (St Louis, MO, USA). Compound names are listed in Table S1.
Animals and Diets All experiments were conducted in accordance with the Institutional Animal Care and Use Committee guidelines of the National Health Research Institutes, Taiwan. C57BL/6 male mice obtained from the National Laboratory Animal Centre, Taiwan, were housed in a controlled environment (21 ± 2 ∘ C, 50 ± 10% humidity, 12-h light:12-h dark cycle) and had free access to food and water. Mice were fed either a standard diet (SD; LabDiet 5053, 12% fat) or a high fat diet (HFD; Research Diet D12451, 45% fat). BPR0912 and rimonabant were dissolved in vehicle [DMSO/Tween 80/H2 O (1 : 1 : 8, vol/vol/vol)] at the doses indicated in the Figures and given by oral gavage.
administered intraperitoneally at 1 h post-BPR0912 treatment. Gastrointestinal transit was assessed histologically as detailed in File S1 and S2.
Diet-induced Obese Mouse Model Seven-week-old C57BL/6 male mice were fed a HFD or SD for 11 weeks before drug treatment. BPR0912 (3 or 10 mg/kg) or rimonabant (10 mg/kg) was administered orally once daily for 20 days from day 0. Food intake and body weight were measured daily after dosing. On day 20, mice were fasted for 5 h and blood samples were collected for analysis after the mice had been euthanized. Tissues, including the liver, interscapular brown adipose tissue (BAT), and inguinal and epididymal white adipose tissue (WAT) were dissected, weighed and frozen at −80 ∘ C for further analyses.
Body Temperature Measurements Body temperature readings of mice were recorded before drug treatment in the afternoon as detailed in File S1.
Metabolic, Biochemical and Tissue Analyses Kits to measure serum concentrations of glucose, triglycerides (both from Randox, Crumlin, UK), insulin (Mercodia, Uppsala, Sweden), leptin (R&D Systems, Minneapolis, MN, USA), and non-esterified fatty acids (Wako, Osaka, Japan) were used according to the manufacturers’ instructions. Liver triglycerides were determined as previously described [17] and as detailed in File S1.
Western Blotting Food Intake and Weight Assessments in Acute Studies BPR0912 (3 or 10 mg/kg) or rimonabant (10 mg/kg) was administered orally to 9-week-old male C57BL/6 mice (weight 23.3 ± 0.1 g) just before the start of the 12-h dark period. Food intake was calculated manually from food remaining in the container. Body weights were measured at 24 h after drug administration.
Indirect Calorimetry C57BL/6 mice, aged 8 weeks (weight 25.3 ± 0.3 g) and acclimatized to individual metabolic chambers, were administered vehicle or BPR0912 (10 mg/kg) orally before the start of the 12-h light period. Mice were subsequently monitored in the chambers for 24 h. Metabolic rate (VO2 , ml/kg/min) and respiratory quotient [calculated as VCO2 (ml/kg/min)/VO2 (ml/kg/min)] were measured using the PhysioScan Metabolic System (AccuScan Instruments, Columbus, OH, USA) according to the manufacturer’s instructions.
Gastrointestinal Transit BPR0912 (1 or 3 mg/kg) was administered orally to 8-week-old male C57BL/6 mice (weight 23.6 ± 0.4 g) after 16 h fasting. WIN55212-2 (1 mg/kg in 1% DMSO in saline) was
Volume 17 No. 5 May 2015
Adipose tissues were homogenized in RIPA lysis buffer (50 mM Tris-HCl, 0.25% sodium deoxycholate, 150 mM NaCl, 1% NP-40) with the addition of protease and phosphatase inhibitor cocktails (Roche, Indianapolis, IN, USA). The antibodies used are listed in Table S2.
RNA Isolation and Quantitative PCR
®
Total RNA was extracted from frozen tissues with TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized from DNase I-treated total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Expression patterns of specific genes in BAT and WAT were determined by quantitative PCR (Applied Biosystems 7900). The expression of the peptidylprolyl isomerase B (Ppib) gene was measured as the endogenous control gene for quantitation of relative mRNA levels. The primer sequences are shown in Table S3.
Statistical Analysis Data are expressed as mean ± standard error of the mean. Statistical analysis of experiments was performed using Prism software (GraphPad Software, San Diego, CA, USA), and comparisons were assessed using an unpaired t-test or one-way analysis of variance, followed by Tukey’s post hoc test.
doi:10.1111/dom.12447 497
original article Results Acute Effects of BPR0912 on Gastrointestinal Transit, Food Intake and Energy Metabolism in Normal Mice The objective of the present study was to investigate the in vivo properties of BPR0912. When we administered the cannabinoid receptor agonist WIN55212-2 to normal C57BL/6 mice, this agent was effective in delaying the gastrointestinal transit mediated by intestinal CB1R [18,19], as expected (Figure 1B, lane 2). This delay was efficiently reversed by prior application of BPR0912 at either 1 or 3 mg/kg (Figure 1B, lanes 3 and 4). Similar results were obtained using the ACEA (CB1R selective agonist)-induced gastrointestinal delay model (Figure S1). Treatment with 3 mg/kg BPR0912 alone mildly increased gastrointestinal transit (Figure 1B, lane 5), suggesting that BPR0912 functions as an inverse agonist of CB1R in vivo. To directly examine BPR0912’s effects on night and day food intake and on 1-day body weight change, 3 or 10 mg/kg BPR0912 or 10 mg/kg rimonabant were administered to normal mice just before the start of the 12-h dark period, and 12-h food intake and 24-h body weight change were examined. The results showed that, unlike rimonabant, which significantly decreased night food intake but increased day food intake, BPR0912 did not alter night or day food intake (Figure 1C, left). Accordingly, 24-h total food intake values were similar to controls for all groups of treated mice (Figure 1C, right). Although BPR0912 treatment at either 3 or 10 mg/kg tended to decrease body weight, unlike rimonabant, these differences were not statistically significant (Figure 1D). Lastly, acute BPR0912 treatment did not alter metabolic rate or respiratory quotient (Figure 1E). These data suggest that BPR0912 can increase gastrointestinal transit without affecting overall food intake, body weight or metabolism.
Effects of Chronic BPR0912 Treatment on DIO Mice To examine the effects of chronic BPR0912 treatment on mice with a metabolic disorder, we administered this compound to mice fed for 11 weeks on a HFD to induce obesity (DIO mice). As in our study of acute BPR0912 treatment of normal mice, BPR0912 at a dose of 3 or 10 mg/kg did not suppress food intake by DIO mice on the first day of treatment and exerted only mild dose-dependent suppression on food intake from days 4 to 6 (Figure 2A). By contrast, treatment of DIO mice with 10 mg/kg rimonabant resulted in a strong suppressive effect that lasted until day 3. No dramatic suppression of food intake was observed for either agent after day 6 until the end of the treatment period. In addition to daily food intake, cumulative intake was analysed. Daily BPR0912 treatment for 6 days reduced cumulative food intake but only when given at the 10 mg/kg dose (Figure 2B). Six days of 10 mg/kg BPR0912 also induced a greater weight reduction than the 3 mg/kg dose (Figure 2B); however, by day 13, the cumulative food intake by both BPR0912-treated groups was similar to that of the rimonabant-treated mice and significantly lower than in controls (Figure 2C). After day 13 until the end of the treatment, no further change in food intake was observed, and BPR0912-treated mice exhibited a dose-dependent change in
498 Hsiao et al.
DIABETES, OBESITY AND METABOLISM
body weight (Figure 2C, D); thus, BPR0912 has no more severe an effect on food intake than does rimonabant and is effective in reducing body weight. When we extended our daily body weight measurements to 19 days, we found that BPR0912 treatment steadily and progressively reduced body weight in DIO mice, in contrast to the abrupt initial weight loss followed by steady-state maintenance of this weight in rimonabant-treated mice (Figure 3A). As a result, both abdominal (e.g. epididymal) and subcutaneous (e.g. inguinal) fat pads decreased in mass in a dose-dependent manner in BPR0912-treated mice (Figure 3B). At a dose of 10 mg/kg, DIO mice treated with BPR0912 or rimonabant exhibited a similar loss of fat mass, although BPR0912 induced a greater body weight loss (Figure 3B). Serum levels of leptin showed a dose-dependent decrease in BPR0912-treated DIO mice, as did serum levels of insulin and triglycerides (Figure 3C); however, only negligible changes in serum glucose and free fatty acids were observed in these mice (Figure 3C). Strikingly, 10 mg/kg BPR0912 markedly reduced hepatic lipids, and the efficacy of this agent at 3 mg/kg was similar to that of rimonabant at 10 mg/kg (Figure 3C).
BPR0912 Enhances Lipolysis and 𝛽-oxidation in White Adipose Tissue The loss of fat pad mass in BPR0912-treated DIO mice prompted us to examine lipolytic enzymes in WAT. Protein kinase A (PKA) activity, demonstrated by phosphorylation of PKA substrates, was slightly elevated in the WAT of BPR0912-treated DIO mice, although the difference from controls was not statistically significant (Figure 4A); however, hormone-sensitive lipase (HSL) was highly phosphorylated in BPR0912- and rimonabant-treated DIO mice compared with controls (Figure 4B). We then examined mRNA levels of genes related to triglyceride lipolysis and fat oxidation. The mRNA expression levels of all three 𝛽-adrenoceptors (encoded by the Adrb1, Adrb2 and Adrb3 genes), as well as peroxisomal proliferator-activated receptor 𝛼 (Ppar𝛼), Ppar𝛾 coactivator 1𝛼 (Pgc1𝛼), and mitochondrial uncoupling protein 1 (Ucp1), were significantly upregulated in DIO mice treated with 10 mg/kg BPR0912 compared with untreated controls (Figure 4C). In the rimonabant-treated group, expression levels of Adrb3, Ucp1 and Pgc1𝛼 were all significantly elevated (Figure 4C); thus, BPR0912 enhances lipolysis and 𝛽-oxidation in the WAT of DIO mice, accounting (at least in part) for the observed reduction in fat mass.
BPR0912 Induces Thermogenesis To determine the effects of BPR0912 on thermogenesis, the core body temperatures of DIO mice that had been treated with 3 or 10 mg/kg BPR0912 for 17–18 days were measured on each afternoon, a time of day when the mice would be in a basal metabolic state. We found that BPR0912 dose-dependently increased body temperature, with an elevation of 0.8 ∘ C over controls observed in mice that had been chronically treated with 10 mg/kg BPR0912 (Figure 5A). This increase was markedly greater than that observed in rimonabant-treated DIO mice.
Volume 17 No. 5 May 2015
original article
DIABETES, OBESITY AND METABOLISM
(B)
4
40
15
Control R-10
2
912-3 912-10
1 0 1
3
5
7
9
11
13
15
17
**
9 6
*
**
*
12
30
BW on Day 6
3
6-day cumulative intake (g)
Daily food intake (g)
(A)
control R-10
20
912-3 912-10
10
3
0
0
Day
(D) 50
Control R-10 912-3 912-10
40 30
6 3
BW change (%)
cumulative food intake (g)
(C)
20 10
1
3
5
7
9
11
13
15
17
Days
R-10
–3
912-3
–6
912-10
–9
***
–12 –15
0
control
0
–18
*** *** $$
Figure 2. Effects of chronic BPR0912 treatment on food intake and weight loss in diet-induced obese (DIO) mice. BPR0912 at 3 or 10 mg/kg (912-3, 912-10), or rimonabant at 10 mg/kg (R-10), was administered orally to DIO mice (weight 35.8 ± 0.6 g) once daily for 18 days. Daily food intake (A), 6-day cumulative food intake and body weight (BW) on day 6 (B), cumulative food intake (C), and BW change on day 18 (D) were measured. Data are mean ± standard error of the mean (four cages per group; n = 2–3 mice per cage). * Versus DIO control group; $ 912-3 versus 912-10. * P < 0.05; **$$ P < 0.01; *** P < 0.001 by one-way analysis of variance, followed by Tukey’s post hoc test.
Ucp1 upregulation is associated with thermogenesis, which prompted us to investigate the in vivo relevance of our observation that Ucp1 was elevated in the BAT of BPR0912-treated DIO mice. As shown in Figure 5B, Ucp1 protein was significantly increased in DIO mice treated with 10 mg/kg BPR0912, similar to the tendency observed in the rimonabant-treated group. When we examined the mRNA expression of 𝛽-adrenoceptor genes, levels of all three 𝛽-adrenoceptor mRNAs were increased in DIO mice treated with 3 mg/kg BPR0912 (Figure 5C); however, in DIO mice treated with 10 mg/kg BPR0912 or rimonabant, only the Adrb2 adrenoceptor mRNA was specifically upregulated (Figure 5C). These data suggest that an activation of Ucp1 that is mediated primarily by 𝛽2-adrenoceptors contributes to the thermogenesis induced by BPR0912.
Discussion Second-generation CB1R antagonists with peripheral distribution may represent powerful potential therapeutics for patients with cardiometabolic disorders. Naturally, the validation of any such chemical agent requires pharmacological evidence, biodistribution information and the definition of the physiological responses elicited in vivo in animal models. In rodents, central activation of CB1R induces a tetrad of responses comprising hypothermia, analgesia, hypolocomotion and catalepsy [20]. In previous work, we showed that our
Volume 17 No. 5 May 2015
candidate peripheral CB1R antagonist BPR0912 penetrated the blood–brain barrier to only a limited extent and did not reverse CP55940-induced hypothermia and analgesia in mice, even at a dose as high as 50 mg/kg [16]. Another outcome of peripheral CB1R activation is depression of gastrointestinal motility mediated by enteric neurons [21,22]. We have shown previously that mice receiving BPR0912 have a high oral drug exposure [16]. In the present study, we found that BPR0912 was able to reverse a peripheral CB1R-mediated delay in gastrointestinal transit at a dose as low as 1 mg/kg. The marked difference in the dose of BPR0912 required to affect these central and peripheral responses, for example, 50 mg/kg versus 1 mg/kg, indicates that BPR0912 is an effective peripheral CB1R antagonist. The biased biodistribution of BPR0912 was also evident in the food intake experiments of the present study. Our previous examination of a different peripheral CB1R antagonist, BPR0697, suggested that acute blockade of peripheral CB1R was not involved in the food intake suppression observed in animals that were on a SD and treated with this agent [17]. This result has been further supported by investigations of JD5037 [23], a peripheral CB1R antagonist, and also by our examination of BPR0912 in the present study. Vijayakumar et al. [17] showed in mice on SD that, unlike rimonabant, BPR0697 at a brain to plasma ratio of 4.3% was not able to elevate the metabolic rate, although it was capable of shifting
doi:10.1111/dom.12447 499
original article Control
38
R-10
36
912-3 912-10
34 32
*
5 4 3
***
***
2 1
0 1 2 3 4 5 6 7 8 9 10111213141516171819
(C)
Day
5 0 1 2 3 4 5 6 7 8 9 10111213141516171819 Day
2.5
*
2.0 1.5 1.0
Control R-10
2
***
912-3
***
912-10
1
60
***
***
* 200
100
50 control
40
R-10 912-3
30
0.0
0
90 80 70 60 50 40 30 20 10 0
0.8
912-10
*
20 10
0.5
–5
*
300
3.0
insulin (ng/ml)
10
3.5
glucose (mg/dL)
28
3
0
0
30
**
**
0
–10
–20 –25
0.7
*
0.6 0.5 0.4 0.3 0.2 0.1 0.0
liver TG (mg/g)
–15
triglycerides (mg/dL)
BW Change (%)
6
NEFA(mEq/L)
BW (g)
40
4
7
leptin (ng/ml)
42
Inguinal fat (% BW)
(B)
44
Epididymal fat (% BW)
(A)
DIABETES, OBESITY AND METABOLISM
90 80 70 60 50 40 30 20 10 0
**
** ***
Figure 3. Effects of chronic BPR0912 treatment on metabolic variables of diet-induced obese (DIO) mice. BPR0912 at 3 or 10 mg/kg (912-3, 912-10), or rimonabant at 10 mg/kg (R-10), was administered orally to DIO mice (weight 38.7 ± 0.6 g) once daily for 19 days. Body weight (BW) and percent BW change (A), epididymal and inguinal fat mass values (B), and serum levels of insulin, glucose, leptin, triglycerides, non-esterified fatty acids (NEFA) and hepatic triglycerides (TG) (C) were measured. Data are the mean ± standard error of the mean (n = 6 mice per group). * P < 0.05, ** P < 0.01, *** P < 0.001 versus DIO control group by one-way analysis of variance, followed by Tukey’s post hoc test.
the respiratory quotient toward lipid oxidation as effectively as rimonabant. In the present study, we found that BPR0912 at a brain to plasma ratio of 3% did not have a significant effect on energy metabolism. These results suggest that peripheral CB1R antagonists lack the ability to acutely modulate the food intake and metabolism of normal mice, which further differentiates these agents from central CB1R antagonists. We observed that the suppressive effect of rimonabant on food intake lasted
