Compounds and solutions
Unless otherwise stated, chemicals were obtained from Sigma-Aldrich (Poole, UK). Peptides were purchased from Bachem (Bubendorf, Switzerland). The GLP1R extracellular domain (ECD) was produced in Escherichia coli, purified as previously described [13] and biotinylated using EZ-link Sulfo-NHS-LC-Biotin (Thermo Fisher Scientific, Loughborough, UK).
Saline solution for imaging experiments contained (mmol/l): 138 NaCl, 4.5 KCl, 4.2 NaHCO3, 1.2 NaH2PO4, 2.6 CaCl2, 1.2 MgCl2 and 10 HEPES (pH 7.4, NaOH). KRB contained (mmol/l): 120 NaCl, 3.5 KCl, 1.2 KH2PO4, 1.2 MgSO4, 2.5 CaCl2, 25 NaHCO3 (pH 7.2, NaOH). Assay buffer for calcium measurements and ligand-binding assays was Hanks’ balanced salt solution supplemented with 25 mmol/l HEPES and 0.1% (wt/vol.) BSA (pH 7.4); for the cAMP homogenous time-resolved fluorescence (HTRF) assay, the buffer was also supplemented with 0.5 mmol/l 3-isobutyl-1-methylxanthine (IBMX).
Cell culture
All cell lines used in this study were mycoplasma negative. INS-1 832/3 cells were maintained in RPMI 1640 Media with GlutaMAX supplement (Thermo Fisher Scientific), 10% FBS, 10 mmol/l HEPES, 50 μmol/l 2-mercaptoethanol, 1 mmol/l sodium pyruvate and penicillin/streptomycin at 37°C in 5% CO2 [14]. Stably transfected cell lines overexpressing GPCRs were generated at AstraZeneca (Gothenburg, Sweden) or MedImmune (Cambridge, UK) using public-domain- or in-house-determined sequences for each receptor, with parental lines purchased from ATCC, ECACC or Invitrogen [15,16,17]. Overexpressing cell lines for experimental use were thawed from liquid nitrogen stocks into assay buffer on the day of the experiment for cAMP HTRF assays, and responses to agonist were assessed to confirm expected cell activity. A mirrorball system (TTP Labtech, Melbourn, UK) was used for the receptor ligand binding assay, described in detail in ESM Methods [18].
Phage display selections
The bone marrow vault library (BMVtrp), combined spleen library (CStrp) and DP47 library (DP47trp) were used for phage selections of single-chain variable fragments (ScFvs) that bound to the GLP1R. The first two rounds of selections, on 100 and 50 nmol/l soluble biotinylated human GLP1R ECD, respectively, were performed as previously described [16, 19]. For the third round of selection, cell surface selections were carried out on Chinese hamster ovary (CHO) cells overexpressing mouse GLP1R. Variable genes from antibodies of interest were cloned into pEU expression vectors for expression and purification of IgGs, as previously described [20].
cAMP HTRF accumulation assay
Cell-based cAMP HTRF accumulation assays were used to screen Glp1R0017 for activity in overexpressing cell lines. Serial dilutions of the antibody and control peptides were prepared in assay buffer and plated using an ECHO525 acoustic liquid handler (Labcyte, Sunnyvale, CA, USA) to give an 11 point dose–response curve in duplicate. Cells were resuspended in assay buffer and then combined with serially diluted antibodies/peptides for 15 min of incubation at room temperature. An agonist challenge based on GLP-1 dose–response curves (see electronic supplementary material [ESM] Fig. 1) was added to samples using the ECHO550 acoustic liquid handler (Labcyte), and samples were centrifuged at 150 g for 1 min for mixing. Following 30 min of incubation at room temperature, cellular cAMP levels were measured using a cAMP dynamic 2 HTRF kit (Cisbio, Codolet, France) according to the manufacturer’s recommendations. After 1 h, plates were read on an EnVision plate reader (PerkinElmer, Waltham, MA, USA). This assay was also used for Schild analysis of Glp1R0017 in the CHO mGLP1R cell line. EC50 values were calculated using non-linear regression in GraphPad Prism (San Diego, CA, USA); dose ratios were determined and plotted to calculate dissociation constants.
Calcium measurements
CHO cells overexpressing human GLP1R were cultured in black poly-d-lysine-coated 384-well plates (15,000 cells/well; Greiner Bio-One, Stonehouse, UK) at 37°C overnight. Cells were washed with assay buffer, loaded with Fluo-4 NW containing 2.5 mmol/l probenecid (Thermo Fisher Scientific) for 30 min (37°C) and 15 min (room temperature), and then incubated with antibody for 15 min at room temperature before adding GLP-1. Fluorescence was recorded using FLIPR Tetra (Molecular Devices, Wokingham, UK) every 0.5 s for 1 min after agonist addition, followed by every 3 s for a further 4 min. Individual responses were normalised to vehicle control, and average responses were calculated by subtracting the basal fluorescence from the peak intensity. Statistical significance was assessed by one-way ANOVA with post hoc Bonferroni test.
cAMP FRET measurements
Single-cell measurements of cAMP were made using the fluorescence resonance energy transfer (FRET)-based sensor Epac2-camps [21], based on the method described for GLUTag cells [22]. INS-1 832/3 cells were seeded into 35 mm plastic dishes and, when 70–80% confluent, were transfected with 3 μg Epac2-camps DNA probe using 3 μl lipofectamine 2000 for 24 h. Cells were trypsinised and re-seeded onto Matrigel-coated 35 mm glass-bottomed dishes for experiments 24–48 h later. For each experiment, cells were washed with saline and then continuously perfused with saline ± test reagents. For the antibody experiment, cells were preincubated with 100 μmol/l Glp1R0017 diluted in saline for 15 min prior to the start of imaging.
Cells were visualised with a ×40 oil immersion objective on an inverted microscope (IX71; Olympus Southend on Sea, UK). A xenon arc lamp coupled to a monochromator (Cairn Research, Faversham, UK) controlled by MetaFluor software (Molecular Devices) was used to excite the cells at 435 nm (200–225 ms excitations) every 5 s. Cyan fluorescent protein (CFP) emission (470 nm) and yellow fluorescent protein (YFP) emission (535 nm) were monitored using an Optosplit II beam splitter (Cairn Research) and an Orca-ER digital camera (Hamamatsu, Welwyn Garden City, UK), and expressed as the CFP/YFP fluorescence ratio. A sliding average across 30 s was used, and responses were calculated by subtracting the maximum ratio at baseline from the maximum ratio during application of the test reagent. Changes in the CFP/YFP emission ratio were calculated as mean ± SEM, and statistical significance was assessed by one-way ANOVA with post hoc Bonferroni test.
Insulin secretion assays
Twenty-four hours before insulin secretion, INS-1 832/3 cells were seeded in 24-well plates at 5 × 105 cells/well. Cells were washed in PBS and then incubated with or without Glp1R0017 in KRB containing 0.2% BSA for 1 h at 37°C in 5% CO2. Cells were then incubated for 2 h with test reagents in KRB. Supernatant fractions were collected and centrifuged to remove any cell debris. Insulin was measured in the supernatant fractions using a rat insulin assay (Meso Scale Discovery, Gaithersburg, MD, USA). Statistical significance was assessed by one-way ANOVA with post hoc Bonferroni test.
Animals
All animal care and experimental procedures were performed in accordance with the Animal (Scientific Procedures) Act 1986, local establishment usage guidelines and Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. The Project Licences authorising the work were approved by a local ethical review body (Animal Welfare and Ethical Review Body). C57/Bl6 mice were sourced from Charles River UK and group-housed in individually ventilated cages within a barrier unit with 12 h light/dark cycle and ad libitum access to chow diet and water. For the experiments, mice were randomised according to body weight.
Immunostaining
Pancreas tissue was fixed in 4% (wt/vol.) paraformaldehyde, dehydrated in 15% (wt/vol.) and 30% (wt/vol.) sucrose and frozen in optimal cutting temperature embedding media (VWR Chemicals, Radnor, PA, USA). Sections were cut (7–10 μm) using a cryostat and mounted directly on to Superfrost Plus glass slides (Thermo Fisher Scientific). Slides were incubated for 1 h in blocking solution containing 5% (vol./vol.) serum, 0.05% (vol./vol.) Tween-20 and 1% (wt/vol.) BSA, and then overnight in blocking solution with the primary antisera of interest against insulin, glucagon, GLP1R and tandem-RFP (ESM Table 1). Slides were washed with blocking solution and incubated with appropriate secondary antisera diluted to 1:300 for 1 h. Negative control slides were stained with secondary antisera alone. Cover slips were mounted using Hydromount (National Diagnostics, Atlanta, GA, USA) and DABCO before confocal microscopy (TCS SP8; Leica, Wetzlar, Germany).
Brain tissue was collected from mice perfused in situ with 4% (wt/vol.) paraformaldehyde and stored in 4% paraformaldehyde 30% (wt/vol.) sucrose prior to slicing at 25 μm. Sections were stained in suspension after sodium citrate antigen retrieval using protocols otherwise similar to those for pancreatic tissue, apart from the blocking solution, which did not contain any BSA.
Single-dose pharmacokinetics study
Two groups of six male C57/Bl6 mice (13 weeks old, mean weight 29.1 ± 0.2 g) were administered Glp1R0017 at 19.2 mg/kg either intraperitoneally or subcutaneously. A sparse sampling approach was used, collecting blood samples from animals across the groups at 0.5, 1, 2, 4 and 7 h after antibody dosing, and then every 24 h over a period of 5 days. At the 120 h endpoint, the animals were terminally anaesthetised using isoflurane inhalation, blood samples were collected via cardiac puncture, and death confirmed via cervical dislocation. Blood samples were collected into EDTA capillary tubes and centrifuged to obtain plasma. Quantitative analysis of plasma antibody concentration was performed in a Gyrolab assay according to the manufacturer’s guidelines (Gyros, Uppsala, Sweden), using a biotinylated idiotype antibody against IgG1 for capture (MedImmune, Cambridge, UK), and an Alexa-labelled sheep anti-human IgG for detection (The Binding Site, Birmingham, UK).
IPGTTs and OGTTs
GTTs were performed in 10-week-old male C57/Bl6 mice (mean weight: IPGTT 27.7 ± 0.2 g, OGTT 25.5 ± 0.3 g) after a 6 h fast, using n = 8 per group determined by power analysis where the effect size was 15%, and 80% power and significance <0.05. Depending on the group (detailed in Results), animals were dosed subcutaneously with antibody or saline 24 h prior to the GTT.
For the IPGTT, mice were dosed with 0.1 mg/kg liraglutide (Victoza; Novo Nordisk, Gatwick, UK) or vehicle subcutaneously 2 h before intraperitoneal glucose administration (2 g/kg). Blood glucose levels were determined from tail-prick blood samples using a hand-held glucometer (AlphaTrak; Zoetis, London, UK) at −120, 0, 15, 30, 45, 60, 90 and 120 min relative to the glucose challenge. For the OGTT, mice were orally gavaged with glucose (2 g/kg), and blood glucose levels were determined at 0, 15, 30, 45, 60, 90 and 120 min relative to the glucose challenge. At the end of the experiment, animals were euthanised using cervical dislocation. The AUC between 0 and 120 min was calculated, and statistical significance was assessed by one-way ANOVA with post hoc Bonferroni test.