Absence of cannabinoid 1 receptor in beta cells protects against high-fat/high-sugar diet-induced beta cell dysfunction and inflammation in murine islets

Aims/hypothesis The cannabinoid 1 receptor (CB1R) regulates insulin sensitivity and glucose metabolism in peripheral tissues. CB1R is expressed on pancreatic beta cells and is coupled to the G protein Gαi, suggesting a negative regulation of endogenous signalling in the beta cell. Deciphering the exact function of CB1R in beta cells has been confounded by the expression of this receptor on multiple tissues involved in regulating metabolism. Thus, in models of global genetic or pharmacological CB1R blockade, it is difficult to distinguish the indirect effects of improved insulin sensitivity in peripheral tissues from the direct effects of inhibiting CB1R in beta cells per se. To assess the direct contribution of beta cell CB1R to metabolism, we designed a mouse model that allows us to determine the role of CB1R specifically in beta cells in the context of whole-body metabolism. Methods We generated a beta cell specific Cnr1 (CB1R) knockout mouse (β-CB1R−/−) to study the long-term consequences of CB1R ablation on beta cell function in adult mice. We measured beta cell function, proliferation and viability in these mice in response to a high-fat/high-sugar diet and induction of acute insulin resistance with the insulin receptor antagonist S961. Results β-CB1R−/− mice had increased fasting (153 ± 23% increase at 10 weeks of age) and stimulated insulin secretion and increased intra-islet cAMP levels (217 ± 33% increase at 10 weeks of age), resulting in primary hyperinsulinaemia, as well as increased beta cell viability, proliferation and islet area (1.9-fold increase at 10 weeks of age). Hyperinsulinaemia led to insulin resistance, which was aggravated by a high-fat/high-sugar diet and weight gain, although beta cells maintained their insulin secretory capacity in response to glucose. Strikingly, islets from β-CB1R−/− mice were protected from diet-induced inflammation. Mechanistically, we show that this is a consequence of curtailment of oxidative stress and reduced activation of the NLRP3 inflammasome in beta cells. Conclusions/interpretation Our data demonstrate CB1R to be a negative regulator of beta cell function and a mediator of islet inflammation under conditions of metabolic stress. Our findings point to beta cell CB1R as a therapeutic target, and broaden its potential to include anti-inflammatory effects in both major forms of diabetes. Data availability Microarray data have been deposited at GEO (GSE102027). Electronic supplementary material The online version of this article (10.1007/s00125-018-4576-4) contains peer-reviewed but unedited supplementary material, which is available to authorised users.


Animal care
All animal procedures and care followed US National Institute of Health guidelines and were approved by the National Institute on Aging Animal Care and Use Committee. Mice were housed in groups of 4 using 12 hrs dark/light cycles, provided with water and fed ad libitum. Male mice were fed standard diet (SD; 16.7% kJ fat and 12.4% kJ sugar wt/wt) or high fat-high sugar diet (HFHS; 49.2% kJ fat and 32.2% kJ sugar wt/wt) to induce obesity. Global CB1R knockout (CB1KO) mice backcrossed to a C57Bl/6J background were bred as previously described. At the end of the diet study, in vivo tests were performed. Tissues, including pancreas, liver, epididymal and subcutaneous fat, were collected, weighed, and flash frozen or fixed for immunohistochemistry.

Method of randomization
Age and sex matching littermate mice were randomly assigned to vehicle or S961 and to SD or HFHS groups.

Assessment of body composition using nuclear magnetic resonance (NMR) spectroscopy
Mice were weighed and immediately placed on a Bruker mini-NMR (Billerica, MA). Data are presented as % of body weight.

Insulin resistance induction by S961
Miniosmotic pumps (7 days, 1 µl/h) were filled with a total of 10 nmoles of S961. Pumping rate was 0.05 nmoles of S961/h (1.2nmoles/day) (3). Implantation of miniosmotic pumps was performed as previously described (4) and according to Alzet guidelines. Tail vein blood was collected daily until day 6 when mice were injected with 5-Bromo-2′-deoxyuridine (BrdU; 0.1 nmoles/g of body weight) and sacrificed the following day.

Respiratory exchange ratio
Animals were placed in metabolic cages for 48 hrs (Columbus Instruments Comprehensive Lab Monitoring System, Columbus Instruments, Columbus, OH), with ad lib access to food and water and controlled light. Metabolic measurements included the respiratory exchange ratio (RER = CO2 produced / O2 consumed), activity levels and food and water intake.

Glucose and insulin tolerance tests
Mice were fasted for 4 hrs or overnight and given free access to water, for insulin (ITT) or glucose tolerance tests (IPTT and OGTT) respectively. For ITT responses to pharmacological CB1R blockade, mice were treated with one single dose of JD-5037 [1 mg/kg in a mixture of saline: Tween-80: treatment (94:1:4); (5)] 30 minutes prior to blood draw. Mice were injected i.p. with 1.5 U/kg of insulin and blood glucose was measured at 0, 15, 30, 60 and 90 minutes. For IPGTT and OGTT mice were given i.p. or orally a bolus of 1.5 g/kg glucose and tail vein blood was collected at the timepoints described above. Area under the curve (AUC) was calculated using GraphPad Prism program.

Blood glucose and hormones measurements
Blood glucose from tail vein was measured utilizing an Easygluco blood glucose meter. Blood was collected from tail vein or terminal retro-orbital bleed and heparin or EDTA, aprotinin and DPP4 inhibitors were added. Plasma insulin, GLP-1 and GIP were quantified by ELISA.

Immunohistochemistry of pancreas and islet size quantification
Immunohistochemistry was carried out as previously described (6) with slight modifications.

Immunoblotting
Protein samples extracted from tissues using RIPA buffer containing protease and phosphatase inhibitor cocktails were immunoprecipitated using anti-IRS2 or anti-IR. Samples were then subjected to Tris-glycine PAGE, immunoblotted with anti-IRS2 or anti-p-Tyr and visualized by ECL. Densitometry of bands were quantified using ImageJ.

Islet isolation
Islets were isolated as previously described (8). Mice were sacrificed and pancreas immediately perfused with 0.7 mg/ml collagenase P containing DNAse I in HBSS without calcium and magnesium. Pancreata were incubated 9 min at 37°C with gentle agitation. The enzymatic reaction was quenched with cold HBSS containing 1% (wt/vol) horse serum. Disaggregated pancreata were spun and washed with cold HBSS, and islets were then handpicked. Intra-islet insulin content was extracted by acid ethanol extraction as described previously (9) and static insulin secretion was normalized to content.

Quantification of cAMP in isolated islets
100 islets were lysed with cold 0.1 mol/l HCl and supernatant was used for cAMP detection by Direct cAMP ELISA kit. Response to incretin (exendin 4; Ex-4; 0.33 nM) was performed as previously described (10).

Islet perifusion
Islet perifusion experiments were performed using a mini-perifusion system previously described by us [13]. One hundred islets were placed into insulin secretion assay buffer in polyacrylaminde P4 BioGel fine. The islet-gel mixture was placed into the perifusion column and connected to the perifusion system. Fractions were collected in a 96 well plate every minute for 30 min after switching to the stimulation conditions. Islets were perifused for 1 h in low glucose (2 mmol/l) at a rate of 100 μl/min, followed by perifusion with stimulatory glucose (7.5 mmol/l; a typical postprandial glucose level) for 15 min at which point the columns were switched back to basal conditions (2mmol/l glucose). Fractions were collected in a 96 well plate every minute for 30mins from the start of the stimulatory conditions. A dead volume equivalent to 14 mins is noted in the measurement of the glucose concentrations. Insulin concentrations were determined by ELISA.

Oxygen consumption and extracellular acidification rate quantification
Freshly isolated islets (30 islets per well) were plated on an Islet Capture Microplate and measurements of oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were performed per manufacturer's protocol using a XF e 24 Seahorse Analyzer (Agilent Technologies, Santa Clara, CA).

Mitochondrial reactive oxygen species measurement
Freshly isolated islets from overnight fasted mice were stained for using MitoSOX (5 µmol/l) for 15 min at 37ºC as previously described (11).

Real Time PCR analysis
Total RNAs were isolated using Trizol or PicoPure RNA Isolation Kit from dissected pancreatic islets. Total RNA concentrations and quality were measured by Nanodrop (Thermo Fisher Scientific). Reverse transcription was performed using SuperScript III First-Strand Synthesis System. Relative expression of selected genes was assayed using TaqMan Fast Advanced Master Mix and FAM-labeled TaqMan Gene Expression Assays on a StepOnePlus Real-Time PCR System (Thermo Fisher Scientific). Duplex reactions were performed using VIC-labeled βactin for endogenous control.

Phosphoprotein and protein cleavage array
Protein phosphorylation and protein cleavage in isolated islets was performed using PathScan® Array Kit following manufacturer's instructions. Briefly 200 islets were lysed and diluted up to 0.2 mg/ml. Samples were incubated for 2 hrs at room temperature and images were captured using LI-COR Odyssey Sa Infrared Imaging System. Densitometry was analysed using Odyssey Sa Imaging Software.

Cytotoxicity assay on isolated islets
Islets were treated as previously described (12). Briefly, freshly isolated islets were incubated for 18 hours in the absence or prescence of a mixture of cytokines (10 ng/ml IL-1β, 50 ng/ml TNF-α and 50 ng/ml IFN-γ). Cytotoxicity was assayed using MultiTox-Fluor Multiplex Cytotoxicity Assay per manufacturer's protocol.

Microarray analysis
Microarray experiments and analysis were performed and analyzed as previously described (13).
Briefly, RNA was extracted from dissected pancreatic islets using Trizol as per the manufacturers recommendations, and RNA concentration and quality evaluated using Nanodrop and the Agilent Bioanalyzer RNA 6000 Chip. Two-hundred ng total RNA was labeled using the Agilent Low-Input QuickAmp Labeling Kit, and was purified and quantified according to the manufacturer's recommendations. A total of 600ng Cy3-labeled cRNA was hybridized for 17hrs to Agilent SurePrint G3 8×60K mouse oligo microarrays (G4852A). Following posthybridization rinses, arrays were scanned using an Agilent SureScan microarray Scanner at 3 micron resolution, and hybridization intensity data extracted from the scanned images using Agilent's Feature Extraction Software. Significant genes were selected by the z-test p value ≤ 0.05, fdr ≤ 0.30 and z-ratio ≥ 1.5. Gene regulatory network and canonic pathway analysis was performed by using Ingenuity Pathway Analysis and gene heatmap by JMP program. The data has been deposited at GEO (GSE102027).

Endocannabinoid (EC) levels in plasma
20µl of plasma was added to 500 µl of cold methanol/Tris buffer [50mmol/l, pH8] containing internal standards. Subsequently to each mixture, 1.5 ml ice-cold methanol-chloroform mixture (2:1) with 0.5 ml of Tris buffer [50mmol/l, pH8] was added. The mixture was centrifuged at 500g for 2 minutes, and the chloroform phase was removed to a glass tube. The extraction was repeated twice. The combined extract was dried and reconstituted in chloroform. Acetone was added to all the samples and the solution was centrifuged at 16,000 x g for 5 minutes. The

Flow Cytometry
Pancreatic lymphocytes were isolated after pancreas perfusion as described above for islet isolation. Cells were washed 3 times with media (RPMI) by centrifuging 5 min at 1500 x rpm.
Then, erythrocytes were lysed by incubating cells incubated with ACK lysis buffer for 3 min at 24ºC. After washing 3 times with media, cells were plated at a density of 2x10 5 cells per well in a 96-well plate, washed 2 times with FACS buffer containing PBS, 0.05% BSA and 0.5% and NaN3, and blocked for 5 min with 0.1% goat serum. The staining was carried out by using PE-anti mouse CD8a and FITC anti-mouse CD69 that were incubated for 20 min at 24ºC. Appropriate isotype controls were used in all the experiments. Populations were determined by using BD FACS Canto II and the software FACS Diva.