Class-Wide Analysis of Frizzled-Dishevelled Interactions Using BRET Biosensors Reveals Functional Differences among Receptor Paralogs

Abstract

Wingless/Int-1 (WNT) signaling is mediated by WNT binding to 10 Frizzleds (FZD_1-10), which propagate the signal inside the cell by interacting with different transducers, most prominently the phosphoprotein Dishevelled (DVL). Despite recent progress, questions about WNT/FZD selectivity and paralog-dependent differences in the FZD/DVL interaction remain unanswered. Here, we present a class-wide analysis of the FZD/DVL interaction using the DEP domain of DVL as a proxy in bioluminescence resonance energy transfer (BRET) techniques. Most FZDs engage in a constitutive high-affinity interaction with DEP. Stimulation of unimolecular FZD/DEP BRET sensors with different ligands revealed that most paralogs are dynamic in the FZD/DEP interface, showing distinct profiles in terms of ligand selectivity and signal kinetics. This study underlines mechanistic differences in terms of how allosteric communication between FZDs and their main signal transducer DVL occurs. Moreover, the unimolecular sensors represent the first receptor-focused biosensors to surpass the requirements for high-throughput screening, facilitating FZD-targeted drug discovery.

Keywords: Dishevelled; Frizzled; GPCR; WNT; bioluminescence resonance energy transfer; class F; transducer.

Figure 1

Class-wide profiling of constitutive FZD-DEP interactions. (A) Schematic showing the principle of the direct BRET assay between a C-terminally Nanoluciferase (Nluc)-tagged receptor and the mVenus-tagged isolated DEP domain of human DVL2. Created with biorender.com. (B) Assessment of the constitutive recruitment of DEP-mVenus to FZD_x-Nluc/SMO-Nluc via bioluminescence resonance energy transfer (BRET)-based acceptor titration experiments. Experiments were performed in transiently transfected HEK293A cells. Shown data represent values from three to four independent experiments (superimposed) ± SD, where each experiment was performed in triplicate. (C, D) log BRET₅₀ (C) and BRET_max (D) values extracted from DEP titration experiments shown in (B). Data represent mean values ± SEM from three to four experiments performed in triplicate.

Figure 2

Basal validation of unimolecular FZD-DEP biosensors. (A) Architecture of the unimolecular FZD-DEP sensors, consisting of the respective FZD sequence, Nanoluciferase (Nluc), a 10 nm E/RK linker, mVenus, and the DEP domain of human DVL2. Created with biorender.com (B) Surface expression analysis of unimolecular FZD-DEP biosensors as determined by surface ELISA in transiently transfected HEK293A cells. Data show mean values ± SEM from three to four (eight for pcDNA) independent experiments, performed in triplicate. Statistical significance compared to an empty-vector transfection (pcDNA) was assessed using one-way ANOVA followed by Fisher’s Least Significant Differences analysis. (C) Basal BRET ratios of unimolecular FZD-DEP biosensors (wild-type (wt) or L445E) in the absence of ligand. The black dashed line represents the average BRET ratio of the three surface-expressed FZD-DEP (L445E) sensors. Experiments were performed in transiently transfected HEK293A cells. Data show mean values ± SEM from three to four independent experiments performed in triplicate. Statistical significance was assessed using one-way ANOVA followed by Sidák’s multiple comparison test. (D) TOPFlash response of selected unimolecular FZD-DEP sensors upon stimulation with vehicle or WNT-3A (300 ng/mL for FZD₁ and FZD₅; 1 μg/mL for FZD₄). Experiments were conducted in transiently transfected ΔFZD_1–10 HEK293T cells. Data show mean values (normalized to vehicle) ± SEM from three independent experiments (each performed in triplicate). Statistical significance between vehicle- and WNT-3A treated conditions was assessed using multiple t tests (two-tailed). ns: not significant; *: p < 0.05; **: p < 0.01; ****: p < 0.0001.

Figure 3

Kinetic profiles of unimolecular FZD-DEP (wt)-Clamps upon WNT stimulation. (A–I) Kinetic BRET responses of the FZD₁- (A), FZD₂- (B), FZD₄- (C), FZD₅- (D), FZD₆ Δ559- (E), FZD₇- (F), mFZD₈- (G), FZD₉- (H), or FZD₁₀-DEP-Clamp (I) upon stimulation with WNT-3A, WNT-5A, WNT-5B, or WNT-16B. Experiments were conducted in HEK293A cells transiently transfected with the indicated FZD-DEP-Clamp. Data present mean values ± SEM from three to five independent experiments performed in triplicate. (J) Heatmap depicting the peak ΔBRET values (ΔBRET_max) for every FZD-DEP-Clamp–WNT combination.

Figure 4

Determination of the Z′ factor for the stably expressed FZD₅-DEP-Clamp. (A) Schematic showing the experimental setup of Z′ factor determination. Created with biorender.com. (B) Representative results from one plate treated with 500 ng/mL WNT-3A or vehicle. Experiments were performed in HEK293A cells stably expressing the FZD₅-DEP-Clamp.