We used Sprague–Dawley male rats (age 2 months), male db/db mice (age 5 months) and age-matched C57BLKS/J mice, all obtained from the Experimental Animal Center, Shanghai Medical College of Fudan University. All experimental procedures were approved by the Shanghai Animal Care and Use Committee on Animals and followed the policies of the International Association for the Study of Pain on the use of laboratory animals. All efforts were made to minimise animal suffering and reduce the numbers of the animals used.
Northern blot analysis
Target gene fragments of human GMRP1 were cloned into pDrive vectors (Qiagen, Chatsworth, CA, USA). Digoxigenin-labelled probes were generated by transcription with SP6/T7 RNA polymerase using a digoxigenin RNA labelling kit (Roche Diagnostics, Indianapolis, IN, USA). Human tissues including human islets were obtained from Shanghai Body Donation Center. Use of those tissues was approved by the Medical Ethics Committee at Huashan Hospital, Fudan University. Northern blotting was performed using a kit (non-isotopic digoxigenin Northern Starter; Roche Diagnostics) as described previously .
Preparation of polyclonal antibody against human GMRP1
The complete cDNA sequence of human GMRP1 gene was subcloned into pET32 vector (Novagen, Madison, WI, USA). The recombinant pET32-GMRP1 vector was transformed into E. coli strain BL21. Recombinant GMRP1 protein was produced and purified, and polyclonal antibody against human GMRP1 generated by immunising rabbits as described previously .
Pancreatic tissues from (1) Sprague–Dawley rats infused with saline and high glucose, (2) db/db mice and (3) C57BLKS/J mice were fixed overnight in formalin at room temperature, along with human insulinoma tissues, and then dehydrated, embedded in paraffin and sectioned (5 μm). For GMRP1 and insulin double immunohistochemistry, rabbit anti-GMRP1 polyclonal antibody and goat anti-insulin monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used to incubate the same section sequentially and the sections were incubated with anti-rabbit IgG antibody conjugated with alkaline phosphatase (Pierce, Rockford, IL, USA) to stain for GMRP1, followed by incubation with horseradish peroxidase-conjugated anti-goat IgG antibody (Pierce) to stain for insulin. The sections were then observed and imaged using a microscope (Leica Microsystems, Wetzlar, Germany). For Ki67 and insulin immunofluorescence, guinea pig anti-insulin polyclonal antibody (Abcam, Cambridge, MA, USA) and rabbit monoclonal antibody against Ki67 (Abcam) were used as primary antibodies, followed by staining for insulin with FITC-conjugated anti-guinea pig IgG antibody (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) and for Ki67 with Rhodamine-conjugated anti-rabbit IgG antibody (Jackson ImmunoResearch). Images were taken under a fluorescence microscope equipped with a charge-coupled device camera (Leica Microsystems).
Western blot analysis
Total proteins were extracted from isolated islet cells and INS-1E cells using RIPA buffer (Pierce). Western blotting was performed as described previously .
INS-1E cells kindly supplied by P. Maechler (Department of Cell Physiology and Metabolism, Faculty of Medicine, University of Geneva, Geneva, Switzerland) , were cultured according to the protocol described . For experiments on regulation of GMRP1 by glucose or insulin, INS-1E cells were cultured to 60% confluence and treated with 5.6 or 25 mmol/l glucose for 24 or 48 h, after which 0, 10, 50 and 100 nmol/l insulin was administered for 24 h. GMRP1 protein levels were investigated by western blotting. Apoptosis of INS-1E cells was evaluated by determination of histone-complexed DNA fragments using a kit (Cell Death Detection ELISA; Roche Molecular Biochemicals, Mannheim, Germany). The relative frequency of apoptosis was photometrically determined by measuring the peroxidase activity of the immunocomplexes at 405 nm. Cell proliferation was assessed by [3H]thymidine incorporation according to the protocol described by El Ouaamari .
Islet isolation and treatment
Islets of Langerhans were isolated from adult Sprague–Dawley rats, saline/high glucose-infused rats, C57BLKS/J mice and db/db mice by in situ pancreas collagenase infusion and separated by density gradient centrifugation at 800 g according to the modified protocol described by Kinasiewicz . Isolated islets were transferred to 24-well plates and cultured in RPMI containing different glucose concentrations (5.6 and 25 mmol/l) for different periods (24 and 48 h) respectively. GMRP1 protein levels were investigated by western blotting and apoptosis of islet cells was evaluated by DNA fragmentation. Cell proliferation was assessed by [3H]thymidine incorporation.
Adenovirus generation and infection
Recombinant adenovirus containing full-length cDNA of rat Gmrp1 (Ad-GMRP1) was generated using the AdMax cloning system (Microbix Biosystems, Toronto, ON, Canada) according to the protocol provided by the manufacturer. Recombinant adenovirus expressing enhanced green fluorescent protein (GFP) (Ad-GFP) was used as a control adenovirus. INS-1E cells or rat islet cells were cultured in six-well plates to 60% confluence and then infected with Ad-GMRP1 or Ad-GFP for 90 min. Viruses were removed and fresh medium was added for an additional 72 h of incubation.
Small interfering RNA-mediated knockdown of Gmrp1
For Gmrp1 small interfering RNA (siRNA) experiments, 20- to 25-nucleotide stealth-prevalidated siRNA duplexes designed for rat Gmrp1 (Invitrogen, Carlsbad, CA, USA) were used. INS-1E cells were seeded in six-well plates at a density of ~5 × 105 cells per well in culture medium without antibiotics and transfected with Lipofectamine 2000 transfection reagent (Invitrogen) according to the manufacturer’s instructions. Cells were transfected for 6 h with the Gmrp1 siRNA at a final concentration of 50 nmol/l or with control siRNA (non-targeting siRNA) at the same concentration before switching to fresh culture medium including antibiotics. At 72 h after transfection, cells were lysed to extract total protein or to detect apoptotic cells by DNA fragmentation or to investigate cell proliferation by [3H]thymidine incorporation. After transfection for 72 h with Gmrp1 siRNA or control siRNA, INS-1E cells were treated with 5.6 or 25 mmol/l glucose for an additional 24 h. Cell proliferation was then assessed by [3H]thymidine incorporation or cells were stimulated for 10 min with 100 nmol/l insulin, followed by western blot to determine phosphorylation of Akt at Thr 308.
Male Sprague–Dawley rats (~250 g) were randomly divided into six groups as follows: (1) rats infused with saline for 24 h (0.45% [wt/vol.] NaCl infusion); (2) rats infused with saline for 48 h; (3) rats infused with saline for 72 h: (4) rats infused with glucose for 24 h (50% [wt/vol.] dextrose infusion, at 2 ml/h); (5), rats infused with glucose for 48 h; and (6) rats infused with glucose for 72 h. The infusion technique was similar to that described by Bonner-Weir et al. . At the end of the infusion period, animals were removed from the restraint system, returned to their cages and eventually killed for morphological assessment of beta cell mass and proliferation, or for islet isolation.
Beta cell replication
Beta cell replication was evaluated using measurement of Ki67-postive beta cells in islets of rats infused with saline or glucose. Pancreatic sections were double-stained with anti-Ki67 antibody and anti-insulin antibody as described above. On these sections, beta cells showed green cytosol and Ki67-postive cells appeared with red nuclei. Ki67-postive beta cells were counted by systematically sampling all of the beta cells identified by insulin staining in each section using a light microscope under high magnification (×1,000). The proportion of Ki67-postive beta cell nuclei to total beta cell nuclei was calculated.
Measurement of beta cell mass
To measure beta cell mass, pancreatic sections (5 μm) were incubated overnight at 4°C with guinea pig anti-insulin polyclonal antibody (1:200 dilution; Abcam) and then incubated with biotinylated anti-guinea pig IgG antibody and with avidin–biotin–peroxidase complex (Vector, Burlingame, CA, USA). Point-counting morphometry of insulin immunostained pancreatic sections as applied by Montaña  was used to calculate the relative beta cell volume as a measure of beta cell mass.
Pancreas sections of db/db or control mice were double immunostained for insulin and for fragmented DNA by TUNEL assay. TUNEL staining was carried out using a DNA fragmentation detection kit (FragEL; Merck, Darmstadt, Germany) according to the manufacturer’s instructions. Cells were counted (1,000 cells per sample) and the number of TUNEL-positive cells quantified and expressed as a percentage of the total.
Data are presented as means ± SEM. Statistical significance was determined by ANOVA, with p < 0.05 considered significant.