original article Diabetes, Obesity and Metabolism 14: 555 – 564, 2012. © 2012 Blackwell Publishing Ltd original article Ibipinabant attenuates β -cell loss in male Zucker diabetic fatty rats independently of its effects on body weight K. Rohrbach 1 , M. A. Thomas 1 , S. Glick 1 , E. N. Fung 2 , V. Wang 3 , L. Watson 3 , P. Gregory 4 , J. Antel 4 & M. A. Pelleymounter 1 1 Department of Metabolic Research, Bristol-Myers Squibb Co., Hopewell, NJ, USA 2 Department of Bioanalytical Sciences, Bristol-Myers Squibb Co., Hopewell, NJ, USA 3 Department of Toxicology, Bristol-Myers Squibb Co., Hopewell, NJ, USA 4 Solvay Pharmaceuticals Research Laboratories, Hannover, Germany Aim: To test the antidiabetic efficacy of ibipinabant, this new cannabinoid receptor 1 (CB1) antagonist was compared with food-restriction- induced weight loss, rosiglitazone (4 mg/kg) and rimonabant (3 and 10 mg/kg), using parameters of glycaemic control in male Zucker diabetic fatty (ZDF) rats. Methods: Body weight, food and water intake, fasted and non-fasted glucose and insulin, glucose tolerance and glycosylated haemoglobin (HbA1c) were all assessed over the course of the 9-week study. Pancreatic insulin content and islet area were also evaluated. Results: At the end of the study, vehicle-treated ZDF rats were severely hyperglycaemic and showed signs of β -cell decline, including dramatic reductions in unfasted insulin levels. Ibipinanbant (10 mg/kg) reduced the following relative to vehicle controls: fasting glucose (−61%), glucose excursion area under the curve (AUC) in an oral glucose tolerance test (OGTT, −44%) and HbA1c (−50%). Furthermore, non-fasting insulin, islet area and islet insulin content were all increased (71, 40 and 76%, respectively) relative to vehicle controls by the end of the study. All of these effects were similar to those of rimonabant and rosiglitazone, where ibipinabant was slightly more effective than rimonabant at the lowest dose and somewhat less effective than rosiglitazone at all doses. These antidiabetic effects appear independent of weight loss because none of the parameters above were consistently improved by the comparable weight loss induced by food restriction. Conclusions: Ibipinabant may have weight loss-independent antidiabetic effects and may have the potential to attenuate β -cell loss in a model of progressive β -cell dysfunction. Keywords: β cell, antidiabetic drug, antiobesity drug, cannabinoids, Zucker diabetic fatty rat Date submitted 18 October 2011; date of first decision 16 November 2011; date of final acceptance 16 January 2012 Introduction Ibipinabant is a new, potent [Ki (CB1) = 7.8 ± 1.4 nM] and selective [Ki (CB2) = 7.943 ± 126 nM] CB1 antagonist [pA2 for arachidonic acid release in CHO cells = 9.9 ± 0.6] with in vitro pharmacological characteristics similar to rimona- bant [1], including inverse agonism [2] and brain penetrance. Although tainted by their potential to induce mood disor- ders, CB1 antagonists have been characterized as appetite suppressants which induce reliable weight loss, both in ani- mals [3–12] and in humans [13–15]. These agents have also been shown to improve glycaemic parameters in ani- mals [16–21] and in humans [22–24]. Interestingly, there is a growing body of evidence suggesting that the prototypic CB1 antagonist, rimonabant, may have direct effects on β -cell function, which are independent of its effects on appetite and body weight. Functional CB1 receptors have been localized in human and rodent pancreas [25–28]. Furthermore, there Correspondence to: Mary Ann Pelleymounter, Department of Metabolic Research, 311 Pennington-Rocky Hill Road, Hopewell, NJ 08534, USA. E-mail: maryann.pelleymounter@bms.com are data suggesting that CB1 antagonists may directly mod- ulate insulin secretion [2,29]. In the study of Getty-Kaushik et al. [29], rimonabant decreased the basal insulin hypersecre- tion observed in islets from 7- to 11-week-old obese Zucker or Zucker diabetic fatty (ZDF) rats without decreasing the fold increase in glucose stimulated insulin secretion. This could suggest that CB1 receptors may have a role in modulating insulin secretion, particularly under conditions of metabolic dysfunction. As progressive increases in basal insulin secretion could lead to insulin resistance and diabetes, it is possible that CB1 antagonists like ibipinabant could delay or attenuate the progression of type II diabetes. Therefore, we tested the hypothesis that ibipinabant could attenuate the progression of type II diabetes by characterizing it in a model of progressive type II diabetes, the ZDF rat. The ZDF rat is selectively bred from Zucker fa/fa rats, which have a dysfunctional, shortened form of the leptin receptor as a result of a spontaneous missense mutation [30]. Insulin levels are abnormally high during the intial stages of the syndrome, reaching their peak at 5–6 weeks of age. Insulin levels then begin to decline, coincident with the onset of hyperglycaemia,