PLoS Biol. 2019 Aug 20;17(8):e3000097. doi: 10.1371/journal.pbio.3000097. eCollection 2019 Aug.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells.

Buenaventura T¹, Bitsi S¹, Laughlin WE¹, Burgoyne T², Lyu Z^1,³, Oqua AI¹, Norman H¹, McGlone ER³, Klymchenko AS⁴, Corrêa IR Jr⁵, Walker A⁶, Inoue A⁷, Hanyaloglu A⁶, Grimes J^8,⁹, Koszegi Z^8,⁹, Calebiro D^8,⁹, Rutter GA¹, Bloom SR³, Jones B³, Tomas A¹.

Author information

1: Section of Cell Biology and Functional Genomics, Division of Diabetes, Endocrinology and Metabolism, Department of Medicine, Imperial College London, London, United Kingdom.
2: Department of Cell Biology, Institute of Ophthalmology, University College London, London, United Kingdom.
3: Section of Investigative Medicine, Division of Diabetes, Endocrinology and Metabolism, Department of Medicine, Imperial College London, London, United Kingdom.
4: Laboratoire de Bioimagerie et Pathologies, UMR CNRS 7021, University of Strasbourg, Illkirch-Strasbourg, France.
5: New England Biolabs, Ipswich, Massachusetts, United States of America.
6: Department of Surgery and Cancer, Imperial College London, London, United Kingdom.
7: Tohoku University, Sendai, Japan.
8: Institute of Metabolism and Systems Research, University of Birmingham, Birmingham, United Kingdom.
9: Centre of Membrane Proteins and Receptors (COMPARE), Universities of Birmingham and Nottingham, Birmingham and Nottingham, United Kingdom.

Abstract

The glucagon-like peptide-1 receptor (GLP-1R), a key pharmacological target in type 2 diabetes (T2D) and obesity, undergoes rapid endocytosis after stimulation by endogenous and therapeutic agonists. We have previously highlighted the relevance of this process in fine-tuning GLP-1R responses in pancreatic beta cells to control insulin secretion. In the present study, we demonstrate an important role for the translocation of active GLP-1Rs into liquid-ordered plasma membrane nanodomains, which act as hotspots for optimal coordination of intracellular signaling and clathrin-mediated endocytosis. This process is dynamically regulated by agonist binding through palmitoylation of the GLP-1R at its carboxyl-terminal tail. Biased GLP-1R agonists and small molecule allosteric modulation both influence GLP-1R palmitoylation, clustering, nanodomain signaling, and internalization. Downstream effects on insulin secretion from pancreatic beta cells indicate that these processes are relevant to GLP-1R physiological actions and might be therapeutically targetable.

PMID:: 31430273
PMCID:: PMC6716783
DOI:: 10.1371/journal.pbio.3000097

Free PMC Article

Images from this publication.See all images (8)Free text

Fig 1

Agonist-induced SNAP-GLP-1R clustering and recruitment to membrane nanodomains.

(A) SNAP-GLP-1R distribution within TMFs, DRFs, and DSFs isolated from MIN6B1 cells stably expressing SNAP-GLP-1R treated with vehicle (“Veh”) or 100 nM exendin-4 (“Ex4”) for 2 min, with flotillin as a marker of membrane raft enrichment. Inset shows quantification of poly-glycosylated SNAP-GLP-1R levels in DRFs; individual results normalized to flotillin (raft loading control) and expressed in AU, n = 6, paired t test. (B) Electron micrographs of gold-labeled SNAP-GLP-1Rs (arrows) from 2D plasma membrane sheets isolated from MIN6B1 cells stably expressing SNAP-GLP-1R following SNAP-tag gold labeling and treatment with vehicle or 100 nM exendin-4 for 2 min. An example of a receptor cluster is shown for the exendin-4-treated condition. Size bars, 100 nm. Mean distances to the nearest neighbor, quantified from a minimum of n = 1,000 gold particles per condition, are shown, unpaired t test. (C) Representative heatmaps of plasma membrane FLAG-GLP-1R localization following TIRF-PALM imaging and reconstruction with QuickPALM plugin (Fiji) in HEK293 cells stably expressing FLAG-GLP-1Rs labeled with anti-FLAG CAGE 500 antibody prior to stimulation with vehicle or 100 nM exendin-4 for 2 min. Size bars, 1 μm. Average cluster sizes from n = 11 regions per condition, paired t test. (D) Agonist-induced SNAP-GLP-1R clustering in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells dual-labeled with Lumi4-Tb and SNAP-Surface 647, treated with vehicle or 100 nM exendin-4, TR-FRET displayed as fold increase relative to individual baseline, n = 4. (E) GαS-YFP distribution within TMFs, DRFs, and DSFs isolated from MIN6B1 SNAP-GLP-1R cells transiently transfected with GαS-YFP prior to treatment with vehicle or 100 nM exendin-4 for 2 min, with flotillin as a marker of membrane raft enrichment. (F) Level of β-arrestin-2 (“βarr2”) recruitment to the GLP-1R in DSFs versus DRFs of CHO-PathHunter (DiscoverX) cells expressing ProLink-tagged human GLP-1R and β-arrestin-2-EA. β-arrestin-2 recruitment was calculated as fold increase of exendin-4 over vehicle condition, normalized to GLP-1R levels in each membrane fraction, and expressed as the fraction of total β-arrestin-2 recruitment for each experimental repeat, n = 4, paired t test. *p < 0.05, ***p < 0.001, by statistical test indicated in the text. All data are shown as mean ± SEM. Underlying raw data for all the panels included in this figure can be found in , and uncropped blots from this figure can be found in . AU, arbitrary unit; DRF, detergent-resistant fraction; DSF, detergent-soluble fraction; GαS, Gs alpha subunit; GLP-1R, glucagon-like peptide-1 receptor; HEK293, human embryonic kidney 293; PALM, photoactivatable localization microscopy; TIRF, total internal reflection fluorescence; TMF, total membrane fraction; TR-FRET, time-resolved Förster resonance energy transfer; YFP, yellow fluorescent protein.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 2

Inhibition of nanodomain compartmentalization reduces GLP-1R signaling and trafficking responses.

(A) Equilibrium binding affinity measurements for exendin-4 (“Ex4”)-K12-FITC in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells pretreated with indicated concentration of MβCD, n = 5, one-way randomized block ANOVA with Dunnett’s test versus vehicle. (B) Effect of 45-min cholesterol depletion with indicated MβCD concentration on SNAP-GLP-1R clustering induced by 100 nM exendin-4 in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells, expressed as the difference between AUC for exendin-4 versus vehicle-induced signal over 30 min, n = 5, 3-parameter logistic fit of pooled data shown. (C) Dose-response curves for exendin-4-induced clustering in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells with or without cholesterol depletion with MβCD (10 mM, 45 min), measured by HTRF and quantified as AUC for each concentration tested, 4-parameter logistic fit of pooled data shown and used to quantify E_max and log EC₅₀ (vertical dotted lines), paired t tests performed on parameter estimates from n = 5 experimental repeats. HTRF traces shown in . (D) cAMP dose response to exendin-4 in wt INS-1 832/3 cells, 10-min stimulation with 500 μM IBMX, normalized to FSK response (10 μM), 4-parameter logistic fit of pooled data shown and used to quantify E_max and log EC₅₀ (vertical dotted lines), paired t tests performed on parameter estimates from n = 5 experimental repeats. (E) Confocal analysis of SNAP-GLP-1R internalization in MIN6B1 SNAP-GLP-1R cells labeled with SNAP-Surface 549 probe (red) for 30 min and then incubated with or without MβCD (10 mM, 1 h) before stimulation with 100 nM exendin-4 for 15 min. Nuclei (DAPI), blue; size bars, 10 μm. (F) As in (E) but with INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells. (G) Effect of indicated concentration of MβCD on uptake of exendin-4-K12-FITC in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells, expressed as AUC relative to baseline for each concentration, n = 5, 3-parameter logistic fit of pooled data shown. (H) Level of exendin-4-K12-TMR uptake in INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells, calculated as AUC of time-lapse confocal microscopy kinetic traces (shown in ) after pretreatment with the indicated MβCD concentration for 1 h, one-way ANOVA with Dunnett’s test, n = 6 traces analyzed from three time-lapse acquisitions per condition. (I) TIRF plasma membrane microscopy analysis of INS-1 832/3 GLP-1R^−/− SNAP-GLP-1R cells transiently transfected with μ2-HA-WT, which codes for the μ2 domain of the clathrin adaptor AP2 fused to an HA tag (AP2-HA, red), labeled with SNAP-Surface 488 (green) and stimulated for 1 min with 100 nM exendin-4 prior to fixation and HA-tag immunofluorescence; size bar, 10 μm. (J) Representative electron micrographs depicting different stages of CCP endocytosis of gold-labeled SNAP-GLP-1Rs (red arrows) in MIN6B1 SNAP-GLP-1R cells stimulated for 1 min with 100 nM exendin-4; size bars, 100 nm. *p < 0.05, ***p < 0.001, “ns” indicates nonsignificant, by statistical test indicated in the text. All data are shown as mean ± SEM, with individual replicates indicated where relevant. Underlying raw data for all the panels included in this figure can be found in , and a dose-response summary for this figure is included in . AP2, adaptor protein 2; AUC, area under the curve; cAMP, cyclic adenosine monophosphate; CCP, clathrin-coated pit; FITC, fluorescein isothiocyanate; FSK, forskolin; GLP-1R, glucagon-like peptide-1 receptor; HA, hemagglutinin; HTRF, homogenous time-resolved fluorescence; IBMX, isobutylmethylxanthine; MβCD, methyl-β-cyclodextrin; TIRF, total internal reflection fluorescence; wt, wild type.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 3

β-arrestins are dispensable for GLP-1R clustering and endocytosis.

(A) SNAP-GLP-1R internalization measured by DERET in wt and β-arrestin-less (“KO”) HEK293 cells stably expressing SNAP-GLP-1R, treated with 100 nM exendin-4 or vehicle (“Veh”), expressed relative to baseline, n = 4. Inset shows AUC, paired t test. (B) Dose responses for SNAP-GLP-1R internalization measured by DERET, analogous experiments to (A), quantified as AUC for each dose tested, 4-parameter logistic fit of pooled data shown, n = 4. (C) Confocal analysis of SNAP-GLP-1R internalization in wt and β-arrestin-less HEK293 cells stably expressing SNAP-GLP-1R, labeled with SNAP-Surface 549 (red) and treated with or without MβCD before stimulation with 100 nM exendin-4 for 15 min. Nuclei (DAPI), blue; size bars, 10 μm. (D) SNAP-GLP-1R clustering measured by HTRF in wt and β-arrestin-less HEK293 cells stably expressing SNAP-GLP-1R, treated with 100 nM exendin-4 or vehicle, expressed relative to baseline, n = 4. Inset shows AUC, paired t test. (E) TIRF microscopy analysis of plasma membranes from wt and β-arrestin-less HEK293 cells stably expressing SNAP-GLP-1R and transiently transfected with AP2-HA, labeled with SNAP-Surface 488 (green) and stimulated for 1 min with 100 nM exendin-4 prior to fixation and HA-tag immunofluorescence (red); size bars, 10 μm, “ns” indicates nonsignificant, by statistical test indicated in the text. All data are shown as mean ± SEM. Underlying raw data for all the panels included in this figure can be found in , and a dose-response summary for this figure is included in . AP2, adaptor protein 2; AUC, area under the curve; DERET, diffusion-enhanced resonance energy transfer; GLP-1R, glucagon-like peptide-1 receptor; HA, hemagglutinin; HEK293, human embryonic kidney 293; HTRF, homogenous time-resolved fluorescence; MβCD, methyl-β-cyclodextrin; TIRF, total internal reflection fluorescence; wt, wild type.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 5

Reduced nanodomain segregation, signaling, and internalization of the palmitoylation-deficient GLP-1R C438A mutant.

(A) Cartoon showing C-terminal cysteine residues in the GLP-1R cytoplasmic tail, with C438 highlighted as the putative palmitoylation site. (B) Analysis of SNAP-GLP-1R palmitoylation levels in CHO SNAP-GLP-1R wt or C438A cells treated with 100 nM exendin-4 for 10 min. (C) Equilibrium dissociation binding constant of exendin-4-K12-FITC determined from HTRF kinetic binding experiments in INS-1 832/3 GLP-1R^−/− stably expressing wt or C438A SNAP-GLP-1R, n = 4, paired t test. Binding traces shown in . (D) Dose responses for exendin-4-induced clustering, measured by HTRF in INS-1 832/3 GLP-1R^−/− stably expressing wt or C438A SNAP-GLP-1R, expressed as AUC for each concentration tested after segmentation into indicated time windows, 4-parameter logistic fits of pooled data shown with E_max and log EC₅₀ (vertical dotted lines), paired t tests comparing parameter estimates for n = 5 repeats. Traces shown in . (E) Representative electron micrographs of gold-labeled SNAP-GLP-1Rs (arrows) from 2D plasma membrane sheets isolated from HEK293 cells stably expressing wt or C438A mutant SNAP-GLP-1R following SNAP-tag gold labeling and treatment with 100 nM exendin-4 for 2 min; size bars, 100 nm. Average distance to the nearest neighbor is shown from a minimum of n = 300 gold particles per condition, unpaired t test. (F) SNAP-GLP-1R wt versus C438A distribution within TMFs, DRFs, and DSFs isolated from INS-1 832/3 GLP-1R^−/− cells stably expressing each type of SNAP-GLP-1R and treated with 100 nM exendin-4 for 2 min, with flotillin as a marker of membrane raft enrichment. (G) Confocal analysis of SNAP-GLP-1R wt versus C438A internalization in INS-1 832/3 GLP-1R^−/− cells stably expressing each type of SNAP-GLP-1R following labeling with SNAP-Surface 488 for 30 min and stimulation with 100 nM exendin-4 for 15 min. Nuclei (DAPI), blue; size bars, 10 μm. (H) Dose responses for wt or C438A SNAP-GLP-1R internalization induced by exendin-4 in INS-1 832/3 GLP-1R^−/− cells stably expressing each SNAP-receptor type, measured by DERET, quantified as AUC for each concentration tested, 4-parameter logistic fits of pooled data shown with E_max and log EC₅₀ (vertical dotted lines), paired t tests performed on parameter estimates from n = 5 repeats. (I) Dynamic internalization profile, assessed as decrease in plasma membrane signal, from time-lapse confocal microscopy data of INS-1 832/3 GLP-1R^−/− cells stably expressing wt or C438A SNAP-GLP-1R following labeling with SNAP-Surface 549 for 30 min and stimulation with 100 nM exendin-4, n = 4, data normalized to baseline for every individual trace. Inset shows AUC calculated from main graph, unpaired t test. (J) Exendin-4-induced cAMP in INS-1 832/3 GLP-1R^−/− cells stably expressing wt or C438A SNAP-GLP-1R, 10-min exendin-4 stimulation with 500 μM IBMX, n = 6, 3-parameter fits shown and used to quantify E_max and log EC₅₀ (vertical dotted lines), paired t tests. (K) Dose-response curves of β-arrestin-2 recruitment to the GLP-1R in HEK293 β-arrestin-2-EA cells transiently transfected with wt or C438A SNAP-GLP-1R-PK, normalized to basal response, n = 5, E_max compared by paired t test. (L) Exendin-4-induced insulin secretion in INS-1 832/3 GLP-1R^−/− cells stably expressing wt or C438A SNAP-GLP-1R, 16-h stimulation at 11 mM glucose, expressed relative to 11 mM glucose alone, 3-parameter fits of pooled data shown, E_max from n = 4 repeats compared by paired t test. *p < 0.05, **p < 0.01, ****p < 0.0001, “ns” indicates nonsignificant, by statistical test indicated in the text. All data are shown as mean ± SEM, with individual replicates indicated where relevant. Underlying raw data for all the panels included in this figure can be found in , and a dose-response summary for this figure is included in ; uncropped blots from this figure can be found in . AUC, area under the curve; cAMP, cyclic adenosine monophosphate; cBF, cleaved bound fraction (corresponding to the palmitoylated pool); CHO, Chinese hamster ovary; cUF, cleaved unbound fraction; DERET, diffusion-enhanced resonance energy transfer; DRF, detergent-resistant fraction; DSF, detergent-soluble fraction; FITC, fluorescein isothiocyanate; GLP-1R, glucagon-like peptide-1 receptor; HEK293, human embryonic kidney 293; HTRF, homogenous time-resolved fluorescence; IF, input fraction; pBF, preserved bound fraction; pUF, preserved unbound fraction; TMF, total membrane fraction; wt, wild type.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 6

GLP-1R palmitoylation, nanodomain clustering, and spatial control of signaling with biased agonists.

(A) Total input (“I”) and palmitoylated (“P”) SNAP-GLP-1R fractions from CHO-K1 cells stably expressing SNAP-GLP-1R wild type or C438A and treated with vehicle (“Veh”) or 100 nM of exendin-4 (“Ex4”), exendin-phe1 (“Ex-phe1”), or exendin-asp3 (“Ex-asp3”) for 10 min. (B) SNAP-GLP-1R distribution within TMFs, DRFs, and DSFs isolated from MIN6B1 SNAP-GLP-1R cells treated with vehicle or indicated agonist (100 nM) for 2 min, with TfR as a marker of membrane non-raft and flotillin as a marker of membrane raft enrichment. Inset shows quantification of poly-glycosylated SNAP-GLP-1R levels in DRFs, with individual results normalized to flotillin (raft loading control), n = 3, one-way repeat-measures ANOVA with Dunnett’s test versus exendin-4. (C) Dose responses showing SNAP-GLP-1R clustering induced by each agonist in INS-1 GLP-1R^−/− SNAP-GLP-1R cells, expressed as AUC for each concentration tested, 4-parameter logistic fits of pooled data shown with E_max and log EC₅₀ (vertical dotted lines), one-way repeat-measures ANOVA with Tukey’s test comparing parameter estimates from n = 5 repeats. Traces shown in . (D) Representative snapshots of single-molecule TIRF image sequences acquired from plasma membranes of CHO-K1 cells transiently expressing SNAP-GLP-1R labeled with SNAP-Surface 549 prior to stimulation with vehicle or 100 nM of the indicated agonist, including overlaid individual trajectories (red). Size bars, 5 μm, and 2 μm for insets. (E, top) Frequency of immobile versus mobile single-molecule SNAP-GLP-1R trajectories following the indicated treatment, based on the generalized diffusion coefficient D. (E, bottom) Frequency of subdiffusion, normal diffusion, and superdiffusion of single-molecule SNAP-GLP-1Rs following the indicated treatment, based on the anomalous diffusion exponent α. Trajectories (6,100–11,000 for each group) analyzed with u-track (MATLAB) and further categorized using custom algorithms (see section). (F) Summary of relative potency changes in CHO SNAP-GLP-1R cells treated with the indicated agonist measured using the FRET biosensors ^TEpac^VV, AKAR4-NES, and AKAR4-Lyn. Bias was determined using the relative potency ratio approach by first subtracting log EC₅₀ values for exendin-phe1 and exendin-asp3 from that of exendin-4 and then normalizing values for AKAR4-Lyn from those of AKAR4-NES or ^TEpac^VV generated in the same experiment for each agonist, n = 5, two-way repeat-measures ANOVA with Sidak’s test comparing exendin-phe1 with exendin-asp3. For further details, see . Traces are shown in . *p < 0.05 by statistical test indicated in the text. All data are shown as mean ± SEM. Underlying raw data for all the panels included in this figure can be found in , and a dose-response summary for this figure is included in ; raw trajectory coordinates and MSDs used to calculate generalized diffusion coefficient D and anomalous diffusion exponent α in (E) can be downloaded from https://doi.org/10.6084/m9.figshare.c.4592000; uncropped blots from this figure can be found in . AUC, area under the curve; CHO, Chinese hamster ovary; DRF, detergent-resistant fraction; DSF, detergent-soluble fraction; FRET, Förster resonance energy transfer; GLP-1R, glucagon-like peptide-1 receptor; MSD, mean squared displacement; TfR, transferrin receptor; TIRF, total internal reflection fluorescence; TMF, total membrane fraction.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 7

Allosteric control of GLP-1R nanodomain segregation, signaling, and trafficking.

(A) SNAP-GLP-1R clustering measured by HTRF in INS-1 GLP-1R^−/− SNAP-GLP-1R cells treated with 100 nM exendin-4 (“Ex4”), exendin-phe1 (“Ex-phe1”), or vehicle, ± BETP (10 μM), expressed relative to baseline, n = 5. Inset shows AUC for each condition, two-way repeat-measures ANOVA with Sidak’s test comparing each condition ± BETP. (B) Representative electron micrographs of gold-labeled SNAP-GLP-1Rs from 2D plasma membrane sheets isolated from MIN6B1 SNAP-GLP-1R cells following gold labeling of SNAP-tagged receptors and treatment with 100 nM exendin-phe1 with and without BETP (10 μM) for 2 min. Average distance to the nearest neighbor quantified from a minimum of n = 155 (− BETP) or n = 330 (+ BETP) gold particles per condition, unpaired t test. (C) SNAP-GLP-1R distribution within TMFs, DRFs, and DSFs isolated from MIN6B1 SNAP-GLP-1R cells treated with vehicle or 100 nM of the indicated agonist with or without BETP (10 μM) for 2 min, with flotillin as a marker of membrane raft enrichment. Inset shows quantification of poly-glycosylated SNAP-GLP-1R levels in DRFs with individual results normalized to flotillin (raft loading control) shown as vehicle fold increase, n = 3, one-way repeat-measures ANOVA with Dunnett’s test versus exendin-4. (D) Effect of BETP (10 μM) on equilibrium dissociation binding constants of exendin-4-K12-FITC and exendin-phe1-K12-FITC in INS-1 GLP-1R^−/− SNAP-GLP-1R cells, n = 5, two-way repeat-measures ANOVA with Sidak’s test comparing ± BETP. (E) cAMP responses in INS-1 832/3 cells with endogenous GLP-1R expression, stimulated for 10 min in presence of 500 μM IBMX with exendin-4 or exendin-phe1 ± BETP (10 μM), normalized to FSK response (10 μM), 4-parameter logistic fits of pooled data shown with log EC₅₀ (vertical dotted lines), two-way repeat-measures ANOVA with Sidak’s test comparing ± BETP from n = 5. (F) SNAP-GLP-1R internalization measured by DERET in INS-1 GLP-1R^−/− SNAP-GLP-1R cells treated with 100 nM exendin-4, exendin-phe1, or vehicle ± BETP (10 μM), expressed relative to baseline, n = 5. Inset shows AUC, two-way repeat-measures ANOVA with Sidak’s test comparing each condition ± BETP. (G) Confocal analysis of the indicated FITC agonist (green) and SNAP-GLP-1R (red) localization in MIN6B1 SNAP-GLP-1R cells following labeling with SNAP-Surface 549 and stimulation with 100 nM of the indicated FITC agonist with or without BETP (10 μM) for 10 min or with BETP alone. Nuclei (DAPI), blue; size bars, 10 μm. *p < 0.05, **p < 0.01, ***p < 0.001, “ns” indicates nonsignificant, by statistical test indicated in the text. All data are shown as mean ± SEM, with individual replicates indicated where relevant. Underlying raw data for all the panels included in this figure can be found in , and a dose-response summary for this figure is included in ; uncropped blots from this figure can be found in . AUC, area under the curve; BETP, 4-(3-benzyloxyphenyl)-2-ethylsulfinyl-6-(trifluoromethyl)pyrimidine; cAMP, cyclic adenosine monophosphate; DRF, detergent-resistant fraction; DSF, detergent-soluble fraction; FITC, fluorescein isothiocyanate; FSK, forskolin; GLP-1R, glucagon-like peptide-1 receptor; HTRF, homogenous time-resolved fluorescence; IBMX, isobutylmethylxanthine; TMF, total membrane fraction.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.

Fig 8

Mechanism of BETP-mediated GLP-1R endocytosis.

(A) Cartoon depicting sites of covalent modification of GLP-1R by BETP at C347 and C438, as per []. (B) Total input (“I”) and palmitoylated (“P”) SNAP-GLP-1R fractions from CHO SNAP-GLP-1R cells treated with 100 nM exendin-4 (“Ex4”) for 10 min in the presence or absence of 10 μM BETP. (C) Confocal analysis of SNAP-GLP-1R C438A (green) localization in HEK293 cells following labeling with SNAP-Surface 488 and stimulation with exendin-phe1 (“Ex-phe1”) with or without BETP (10 μM) for 10 min. Nuclei (DAPI), blue; size bars, 10 μm. (D) SNAP-GLP-1R internalization measured by DERET in HEK293 cells transiently expressing wt or C438A mutant SNAP-GLP-1R, treated with 100 nM exendin-phe1, 10 μM BETP, or both, expressed as fold increase from baseline, n = 6. (E) Confocal analysis of exendin-phe1-K12-FITC (green) and SNAP-GLP-1R (red) internalization in HEK293 β-arrestin-less (“βarr1/2 KO”) cells stably expressing SNAP-GLP-1R following labeling with SNAP-Surface 549 and stimulation with 100 nM of exendin-phe1-K12-FITC for 40 min with or without BETP (10 μM). Nuclei (DAPI), blue; size bars, 10 μm. (F) Confocal analysis of the indicated FITC agonist (green) and SNAP-GLP-1R (red) localization in MIN6B1 SNAP-GLP-1R cells following labeling with SNAP-Surface 549 and stimulation with 100 nM of the indicated FITC agonist with or without BETP (10 μM) for 10 min in the presence or absence of MβCD (10 mM, 1 h preincubation). Nuclei (DAPI), blue; size bars, 10 μm. (G) TIRF microscopy analysis of plasma membranes from MIN6B1 SNAP-GLP-1R cells transiently transfected with AP2-HA, following labeling with SNAP-Surface 488 (green) and stimulation for 1 min with 100 nM exendin-phe1 with and without BETP (10 μM), prior to fixation and HA-tag immunofluorescence (red); size bars, 2 μm. (H) Representative electron micrographs depicting different stages of CCP endocytosis of gold-labeled SNAP-GLP-1Rs (red arrows) in MIN6B1 SNAP-GLP-1R cells stimulated for 1 min with 100 nM exendin-phe1 with BETP (10 μM); size bars, 100 nm. All data are shown as mean ± SEM. Underlying raw data for all the panels included in this figure can be found in ; uncropped blots from this figure can be found in . AP2, activator protein 2; BETP, 4-(3-benzyloxyphenyl)-2-ethylsulfinyl-6-(trifluoromethyl)pyrimidine; CCP, clathrin-coated pit; FITC, fluorescein isothiocyanate; GLP-1R, glucagon-like peptide-1 receptor; HA, hemagglutinin; HEK293, human embryonic kidney 293; MβCD, methyl-β-cyclodextrin; TIRF, total internal reflection fluorescence; wt, wild type.

Agonist-induced membrane nanodomain clustering drives GLP-1 receptor responses in pancreatic beta cells

PLoS Biol. 2019 Aug;17(8):e3000097.