Development and optical properties of GLPLight1.

a) Structural model of GLPLight1 obtained using Alphafold (Mirdita et al., 2022). The human GLP1R is shown in gold, cpGFP in green and residue targets of mutagenesis are shown in magenta. b) Representative images showing GLPLight1 expression and fluorescence intensity change before (left) and after (center) addition of 10 μM GLP-1(7–37) as well as their respective pixel-wise ΔF/F0 images in HEK293T cells (top) and primary cortical neurons (bottom). Scale bars, 10 μm. c) Maximal fluorescence response of GLPLight1 expressed in the indicated cell types after the addition of 10 μM GLP1. n = 35 neurons and n = 30 HEK293T cells, from 3 independent experiments. d) One-photon excitation/emission spectra of GLPLight1-expressing HEK293T cells before (dark green and dark blue) and after (light green and light blue) addition of 10 μM GLP-1(7–37) normalized to the peak excitation and emission of the GLP-1-bound state of the sensor. Data were obtained from 3 independent experiments. Only mean values are shown. All data shown as mean ± SEM unless stated otherwise.

Pharmacological characterization of GLPLight1.

a) Absolute ΔF/F0 responses of GLPLight1 expressing HEK293T cells to various GLP-1-agonists, antagonist, or to other class-B1 neuropeptide ligands applied at 1 μM final (unless stated otherwise). n = 30 cells from three independent experiments for all conditions. b) ΔF/F0 responses from time-lapse imaging experiments in which 1 μM GLP-1 and 10 μM exendin-9 (Ex-9, a peptide antagonist of GLP1R) were subsequently bath-applied onto GLPLight1-expressing HEK293T cells. n = 30 cells from three independent experiments. c) Normalized dose-response curves showing the fluorescent responses of GLPLight1-expressing HEK293T cells and primary cortical neurons to GLP-1 (dark green and light green respectively) or GLPLight1-expressing HEK293T cells to glucagon (purple). The curves fit were performed using a 4-parameter equation and the mean EC50 values determined are shown next to the traces in the corresponding color. n=3,6 and 3 independent experiments for GLP-1(7–37) in neurons, GLP-1(7–37) and glucagon in HEK293T cells, respectively. d) Dose-response curves showing the fluorescent responses of GLPLight1-expressing HEK293T cells to alanine mutants of the GLP-1 peptide normalized to the maximum mean fluorescence response (FITC intensity) obtained for the WT GLP-1 peptide. n = 3 independent experiments for each peptide. All data are displayed as mean ± SEM.

Titration parameters of alanine scanned variants of GLP-1 peptide.

The Emax and pEC50 values were derived from the four-parameter non-linear fit for each peptide and the EC50 shift by comparison against WT GLP-1 peptide measured alongside.

All-optical visualization and control of human GLP1R activation.

a) Schematic representation of the N-terminal chemical caging strategy used to generate photo-GLP1. The peptide product (native GLP-1) after optical uncaging is shown. b) Time-lapses of fluorescence response for GLPLight1 or GLPLight-ctr expressing HEK293T cells before and after optical uncaging (purple vertical bar, 405 nm laser, scanning rate of 0.8 Hz and variable durations as specified below the graph). The values were normalized to the maximal response of GLPLight1 between t = 150 and 200 sec to 10 μM GLP-1 (purple trace). In all cases, cells were pre-incubated for 1–2 minutes with 10 μM photo-GLP1 before imaging and optical uncaging started. All fluorescent signals were analyzed within 20 μm distance from the uncaging area. n=11 to 19 cells from three independent experiments. c) Quantification of the normalized average fluorescence from b between t = 25 to 75 sec for uncaging experiments, and t = 150 to 200 sec for GLP-1 application experiments. All uncaging experiments on GLPLight1 were compared to the one with pre-incubation with Exendin-9 (Ex-9, see Figure 3-figure supplement 2) using Brown-Forsythe ANOVA test followed by Dunnett’s T3 multiple comparison. p = 0.0061; 0.0001; 0.0026 and 1,293 x 10-6 for 10, 20, 50 and 100 sec uncaging events, respectively. d) Fluorescence response of GLPLight1-expressing HEL293T cells to 10 μM GLP-1 either after pre-incubation with 10 μM photo-GLP1 (magenta) or in the absence of it (blue). The data were normalized to the maximal response of GLPLight1-expressing cells in the absence of photo-GLP1 and fitted with a non-linear mono exponential fit to determine τ values. Statistical analysis was performed using the extra sum-of-squares F test; ****p< 0.0001; n=18 and 17 cells from three independent experiments in the absence or presence of photo-GLP1 respectively. All data are displayed as mean ± SEM. e) Kymographs of representative cells from d showing the fluorescence intensity of a line drawn across a cell membrane over time in the absence (top) or presence (bottom) of photo-GLP1 (10 μM) in the bath. The timepoint of GLP-1 application is shown by the red dotted line. f) Representative images of multiple uncaging events performed at different locations across the field of view. Images show the basal fluorescence of GLPLight1-expressing HEK cells (left) in the presence (top) or absence (bottom) of 10 μM photo-GLP1, as well as the corresponding pixel-wise heatmap of SNR post-uncaging. Localized functional sensor responses to optical uncaging of photo-GLP1 are indicated by white arrows. Uncaging was performed for a duration of 40 sec in total for all the three areas shown as white squares using a scanning rate of 1.5Hz. Scale bars: 20 μm. g) Quantification of the timelapse of fluorescence response of GLPLight1 from f inside the uncaging areas. h) Same as f but with a sub-cellular localized uncaging region selected on the membrane of a GLPLight1-expressing cell with 1,5 s uncaging duration and a 25 Hz scanning rate. Scale bar 10 μm.

Effect of photo-GLP1 uncaging on intracellular signaling.

a) Representative images of HEK293T cells used in c,d. hmGLP1R expression was visualized using an Alexa-647 conjugated anti-FLAG antibody. The uncaging region is represented by the white square in the lower right area. b) Representative images from the pixel-wise fluorescence response from G-Flamp1 after uncaging (left) and bath-application of 1 nM WT GLP-1 (right). The white arrow indicates the localized area of uncaging. c) Fluorescence change during timelapse imaging of G-Flamp1/GLP1R co-expressing cells after addition of 1 nM photo-GLP1 and localized uncaging (405 nm, 2 sec and 32 Hz scanning rate), followed by bath-application of 1 nM WT GLP-1. The timepoints of ligand addition are represented using grey rectangles and the uncaging bout by the vertical black line. Quantification of the fluorescence response is shown separately for ‘target cells’ (blue, cells within the uncaging area) and for ‘distant cells’ (grey, cells positioned at least 10 μm away from the uncaging area). The fluorescent responses from G-Flamp1 were normalized to the maximal activation after addition of WT GLP-1. d) Quantification from the average normalized fluorescence in c between t=400 and t=450 sec using the 10 frames before uncaging as a baseline for each condition. n = 3 ‘target cells’ and 15 ‘distant cells’ from 3 independent experiments. *P = 0.03061 using a one-tailed Student’s t-test with Welch’s correction. All scale bars: 20 μm and all data displayed as mean ± SEM.

Screening process for the development of GLPLight1.

a) Maximal fluorescence response to 10 μM GLP-1 of the prototype GLP-1 sensor and the N-terminus deletion variants (16 or 23 first amino acids). b) Maximal fluorescence response to 10 μM GLP-1 of the ICL2 lysine variants compared to GLP-1 sensor prototype with deletion of the first 23 amino acids. The variant with the best response containing the mutation L260K was named GLPLight0.1. The top-right insert depicts the site that were mutated on the sensor’s ICL2 in black letters. c) Overview of the randomized screening of residues K336 and T343 on GLPLight0.1 backbone after addition of 10 μM GLP-1. The data were normalized to the variant showing the highest response. The two randomized residues are shown above in the diagram as the black letters: x. d) Comparison of the maximal response to 10 μM GLP-1 of GLPLight0.1 to the best variant from c, subsequently called GLPLight0.2. e) Comparison of the maximal response to 10 μM GLP-1 of GLPLight0.2 constructs with or without addition of an ER export sequence at the C-terminus. The construct containing the ER export sequence was called GLPLight0.3. f) Comparison of the maximal response to 10 μM GLP-1 of GLPLight0.3 with or without the cpGFP mutations as well as their basal brightness in g. The construct containing the mutations was called GLPLight0.4. h) Comparison of the maximal response to 10 μM GLP-1 of GLPLight0.4 with or without the alanine mutations of the C-terminal phosphorylation sites. The final version of the sensor was called GLPLight1. All data were acquired in HEK293T cells with multiple repeats shown as mean ± SEM.

Development of the control sensor GLPLight-ctr.

a) Maximal response of GLPLight1 binding pocket variants to 1 μM GLP-1(7–37). b) Structural model of GLPLight-ctr shown from the side (left) or top (right) perspective. The residues mutated to alanine compared to GLPLight1 are shown in magenta. c) Representative images from GLPLight-ctr expression and fluorescence intensity change before (left) and after (center) addition of 10 μM GLP-1(7–37) as well as the respective pixel-wise ΔF/F0 images in HEK293T cells (top) and primary cortical neurons (bottom). Scale bars: 10 μm for HEK cells and 20 μm for neurons. d) Maximal response of either GLPLight1 or GLPLight-ctr in HEK293T cells and primary cortical neurons after addition of 10 μM GLP-1(7–37).

Intracellular signaling characterization of GLP1R and GLPLight1.

Normalized timelapse recording of agonist induced miniG protein or beta arrestin recruitment in GLP1R or GLPLight1 expressing HEK293T cells along with quantification of the mean signal between t = 600– 700 sec for mini-Gs in a–b, mini-Gq in c–d, mini-Gi in e–f, mini-G12 in g–h, and β-arrestin-2 in i–j. The dashed line corresponds to the addition of 100 nM GLP-1(7-37). One-tailed unpaired t-test with Welch’s correction, mini-Gs: p = 0,0202; mini-Gq: p = 0,0074; mini-Gi: p = 0,0102; mini-G12: p = 0,0356 and β-arrestin-2: p = 0,0452. k) Titration of GLP-1 induced intracellular cAMP in HEK293T cells co-expressing GLO20F cAMP sensor and GLP1R or GLPLight1. The data were normalized to the maximum signal observed for GLP1R. Unpaired two-tailed t-test with Welch’s correction for both constructs after 100 nM GLP-1. P = 0,0041. l) Titration of GLP-1 on miniGs-recruitment assay in GLP1R expressing HEK293T cells. The curve fit was performed using a four parameter non-linear fit and the EC50 value is shown next to the trace. All data obtained from three independent experiments.

Detecting endogenous GLP-1 release from enterocrine cells using GLPLight1.

a) Representative fluorescence and brightfield images of HEK293T co-transfected with GLPLight1 and cytosolic mKate2 cultured overnight in the absence (top row) or presence (bottom row) of GLUTag cells. b) Fluorescence response of GLPLight1 expressing HEK cells cultured with (right) or without (left) GLUTag cells. The data were normalized to the average response of GLPLight1 from the HEK cells only population. Statistical analysis was done using a two-tailed unpaired t-test, ****P = 3.399 × 10-14; n = 32 and 43 cells from three independent experiments for HEK cells only or HEK+GLUTag cells, respectively.

Biochemical characterization of photo-GLP1.

a) Scheme illustrating the uncaging reaction of photo-GLP1 when exposed to UV light at 370–405 nm, producing WT GLP-1. b) LCMS chromatographic traces of (i) pure WT GLP-1, (ii) pure photo-GLP1, photo-GLP1 after irradiation at 370 nm for (iii) 20 sec, (iv) 40 sec, (v) 120 sec, (vi) 300 sec, and (vii) co- injection of photo-GLP1 after irradiation for 300 sec (50 µL, 80 µM) with pure WT GLP-1 (50 µL, 80 µM). Chromatographic peaks corresponding to WT GLP-1 are highlighted in blue/green, peaks corresponding to photo-GLP1 are highlighted in purple. c) UV-Vis spectra (λ = 200–450 nm) of photo- GLP1 (purple, 80 µM), WT GLP-1 (blue/green, 80 µM), and α-methyl-5-methoxy-2-nitro-4-(2-propyn-1- yloxy)benzyl alcohol (PC Building Block, grey, 80 µM). All UV-Vis measurements were carried out in HBSS buffer and processed with blank (HBSS buffer) subtraction. d) Amount of WT GLP-1 produced by irradiation of photo-GLP1 at 370 nm over time. Amount of WT GLP-1 at each time-point was estimated using area under the curve (AUC) of the corresponding chromatographic peak and given as a percentage of the total AUC value of peaks corresponding to photo-GLP1 and WT GLP-1.

Further characterization of photo-GLP1 uncaging.

a) Fluorescence response (ΔF/F0) of in GLPLight1 expressing HEK293T cells after bolus addition of 10 μM photo-GLP1. The time of addition is represented by the blue rectangle above; n = 16 cells. b) Fluorescence response of GLPLight1 expressing HEK293T cells normalized to the maximal response to 10 μM WT GLP-1 after 2 minutes pre-incubation with a mix of 10 μM Exendin-9 + 10 μM photo-GLP1 and following optical uncaging for 100 seconds with a 405 nm laser (represented by the magenta shaded area); n = 17 cells. All data shown as mean ± SEM and acquired from 3 independent experiments.

Control experiment for intracellular signaling characterization.

a) Timelapse imaging of G-Flamp1 response in HEK293T cells co-expressing GLP1R after addition of HBSS buffer (used for imaging and dilution of the ligands). The time point of buffer addition is indicated by the grey rectangle above. n = 15 cells from three independent experiments. b) Timelapse imaging of G-Flamp1 response in HEK293T cells co-expressing GLP1R after 2 seconds photo-activation (represented by the black vertical line) in the absence of photo-GLP1. n = 3 cells from three independent experiments.

LCMS Profile of crude core peptide GLP1[6-30]. (a) Chromatogram (λ = 214 nm) of core peptide GLP1[6-30]; Rt 5.246 min. (b) HRMS (ESI-TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass calcd. for C129H197N33O37 2800.4548, found 2800.7377.

UHPLC profile of [H1A]GLP1 crude product. Rt 14.168 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 58% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

UHPLC profile of [H1A]GLP1 pure product. Rt 14.144 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), >95% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

LCMS profile of crude peptide [H1A]GLP1. (a) Chromatogram (λ = 214 nm) of crude [H1A]GLP1; Rt 5.482 min. (b) ESI-TOF. (c) Deconvoluted HRMS; monoisotopic mass calcd. for C146H224N38O45 3229.6408, found 3229.9870.

LCMS profile of pure peptide [H1A]GLP1. (a) Chromatogram (λ = 214 nm) of pure [H1A]GLP1; Rt 5.523 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass calcd. for C146H224N38O45 3229.6408, found 3229.9545.

UHPLC profile of [E3A]GLP1 crude product. Rt 13.885 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 54% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

UHPLC profile of [E3A]GLP1 pure product. Rt 13.880 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), >95% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

LCMS profile of crude peptide [E3A]GLP1. (a) Chromatogram (λ = 214 nm) of crude [E3A]GLP1; Rt 5.210 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass calcd. for C147H224N40O43 3237.6571, found 3237.6608.

LCMS profile of pure peptide [E3A]GLP1. (a) Chromatogram (λ = 214 nm) of pure [E3A]GLP1; Rt 5.253 min. (b) HRMS (ESI-TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass calcd. for C147H224N40O43 3237.6571, found 3237.6512.

UHPLC profile of [G4A]GLP1 crude product. Rt 13.954 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 53% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

UHPLC profile of [G4A]GLP1 pure product. Rt 13.927 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 94% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

LCMS profile of crude peptide [G4A]GLP1. (a) Chromatogram (λ = 214 nm) of crude [G4A]GLP1; Rt 5.235 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C150H228N40O45 3309.6782, found 3309.6778.

LCMS profile of pure peptide [G4A]GLP1. (a) Chromatogram (λ = 214 nm) of pure [G4A]GLP1; Rt 5.294 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C150H228N40O45 3309.6782, found 3309.6696.

UHPLC profile of [T5A]GLP1 crude product. Rt 13.770 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 58% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

UHPLC profile of [T5A]GLP1 pure product. Rt 13.765 min (Agilent Zorbax 300SB-C18 column, 5 µm, 2.1 × 150 mm, 5-95% MeCN over 30 min, ca. 3% MeCN/min), 95% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 4.0–29 min.

LCMS profile of crude peptide [T5A]GLP1. (a) Chromatogram (λ = 214 nm) of crude [T5A]GLP1; Rt 5.172 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C148H224N40O44 3265.6520, found 3265.6516.

LCMS profile of pure peptide [T5A]GLP1. (a) Chromatogram (λ = 214 nm) of pure [T5A]GLP1; Rt 5.215 min. (b) HRMS (ESI- TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C148H224N40O44 3265.6520, found 3265.6399.

UHPLC profile of crude photo-GLP1; Rt 5.081 min (Agilent Zorbax Eclipse Plus C18 Rapid Resolution HD column, 1.8 µm, 2.1 × 50 mm, 5-95% MeCN over 9 min, ca. 10% MeCN/min), 49% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 2.0–9.5 min.

UHPLC profile of pure photo-GLP1; Rt 5.097 min (Agilent Zorbax Eclipse Plus C18 Rapid Resolution HD column, 1.8 µm, 2.1 × 50 mm, 5-95% MeCN over 9 min, ca. 10% MeCN/min), >95% purity based on Area Under Curve (AUC) at λ = 214 nm, accounting for all peaks between 2.0–9.5 min

LCMS profile of crude photo-GLP1. (a) Chromatogram (λ = 214 nm) of crude photo-GLP1; Rt 5.97 min. (b) HRMS (ESI-TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C164H240N42O52 3629.7427, found 3629. 2054.

LCMS profile of pure photo-GLP1. (a) Chromatogram (λ = 214 nm) of pure photo-GLP1; Rt 5.98 min. (b) HRMS (ESI-TOF) spectrum. (c) Deconvoluted HRMS; monoisotopic mass (ESI+) calcd. for C164H240N42O52 3629.7427, found 3629.2373.