Specific and direct modulation of the interaction between adhesion GPCR GPR56/ADGRG1 and tissue transglutaminase 2 using synthetic ligands

Figures

• 

  Figure 1. The GPR56-TG2 interaction is direct. (A) Structures of GPR56 (left) and TG2 (right, PDB: 2Q3Z) are colored by domain. The ECR of GPR56 (PDB: 5KVM) is shown in the context of a modelled 7TM based on the structure of the glucagon receptor (PDB: 46LR). (B) Schematic of a GPR56-TG2 binding assay between purified soluble GPR56 ECR and M280 beads coated with purified TG2 FL or TG2 D3D4. Blue star indicates fluorescent label. (C, D) GPR56-TG2 binding assays using setup outlined in (B). TG2 median fluorescence intensity (MFI) is plotted on the y-axis as TG2 binding signal. (C) Binding of TG2-coated beads to 100 nM GPR56 ECR, which was tetramerized12,14 to increase avidity. (D) Concentration titrations of monomeric GPR56 ECR constructs binding to TG2 D3D4-coated M280 beads. The H89A mutation was previously shown to confer a GPR56 loss-of-function phenotype in a zebrafish model12. GAIN domain is included as a negative control. Curves were fit to a simple one-to-one binding model to determine the dissociation constant, KD, of each interaction: mouse ECR, 440 ± 20 nM; human ECR, 330 ± 15 nM; mouse ECR H89A, 880 ± 80 nM; and human ECR H89A, 790 ± 11 nM. Error bars in (C) and (D) indicate…

• 

  Figure 2. The conserved patch on the PLL domain mediates TG2 binding. (A, B) Cartoon of the conserved patch on the PLL domain (PDB: 5KVM). (A) Structure is colored by residue conservation score51,52. Figure adapted from reference12. (B) Sidechains are colored based on TG2 binding signal upon their mutation. The Asn148-linked glycan, which is expected to be removed with the S150A mutation, is shown as transparent yellow. (C) Schematic of the GPR56-TG2 binding assay using purified TG2 D3D4 and GPR56 FL constructs expressed on HEK293T cells. The binding signal of monobody β3, which binds the GAIN domain, was used to quantify GPR56 cell-surface expression. Green and blue stars represent different fluorescent labels. (D, E) Quantification of cell-surface expression of each mutant (D) and TG2 D3D4 binding signal (E) using the setup described in (A). Binding signal represents MFI. (E) Binding signal of 5 nM TG2 D3D4 (tetramerized) is measured within a defined range of cell-surface expression as detailed in Figure S4. Error bars in (D) and (E) indicate S.E.M. of n = 3 independent measurements. Significance levels calculated by 2-way ANOVA with Bonferroni correction for multiple comparisons.…

• 

  Figure 3. PLL-binding monobodies block TG2 binding. (A) Schematic of the monobody competition assay using GPR56 FL expressed on HEK293T cells and purified TG2 D3D4. Before and during incubation with labeled TG2 D3D4, cells were incubated with excess unlabeled monobody competitor (i.e. β7 or β12). Blue stars indicate fluorescent label. (B) HEK293T cells transfected with indicated GPR56 constructs were stained with purified TG2 D3D4 in the presence of excess unlabeled monobody competitor as described in (A). Binding signal represents MFI. Error bars indicate S.E.M. of n = 3 independent measurements. Significance levels shown were calculated by 2-way ANOVA with Bonferroni correction for multiple comparisons. n.s., not significant. (C) Schematic of the “sandwich” format binding assay: M280 beads coated with an immobilized protein (step 0) are incubated with unlabeled human GPR56 ECR (step 1) and then with a fluorescently labeled protein (step 2). By design, any protein pairs from steps 0 and 2 that that bind overlapping sites on hGPR56 ECR result in low binding signal. Conversely, protein pairs with non-overlapping binding sites on hGPR56 produce high binding signal. Blue stars indicate…

• 

  Figure 4. Modulation of the GPR56-TG2 interaction. Summary of the results presented are depicted. (Left) Depiction of the interaction between wt FL GPR56 and TG2. (Middle) Mutations to the PLL domain (red star) block mTG2 binding. (Right) PLL-binding monobodies disrupt the interaction between GPR56 and mTG2.