Influence of NanoBiT tags on VEGFR2-mediated signalling.
We confirmed that the NanoBiT fragments did not interfere with VEGFR2 signalling using an NFAT reporter gene assay (Kilpatrick et al., 2017). Concentration-response curves for VEGF165a were compared between cells stably expressing VEGFR2 tagged at the N-terminus with LgBiT, HiBiT or SmBiT (Figure 5). Each receptor exhibited a concentration-dependent increase in NFAT gene transcription in response to increasing concentrations of VEGF165a. Each cell line had a similar potency derived for VEGF165a (LgBiT-VEGFR2 pEC50 = 9.95 ± 0.11; HiBiT-VEGFR2 pEC50= 10.06 ± 0.12; SmBiT-VEGFR2 pEC50 = 10.23 ± 0.23; n=5 for each). These were comparable to potency values derived for VEGF165a at wild type VEGFR2 (Kilpatrick et al., 2017).

Nanomolar affinity of fluorescent VEGF-A at a defined VEGFR2/NRP1 complex

Fluorescent VEGF-A ligand binding was monitored at full-length VEGFR2 and NRP1 tagged at their N-terminus with LgBiT and HiBiT, respectively. Since uncomplemented receptors cannot oxidise furimazine, luminescence was confined to proteins where complementation from a defined heteromeric VEGFR2/NRP1 NanoBiT complex had occurred (Figure 6a). BRET therefore only derived from the receptor/co-receptor complex and the fluorescent VEGF-A acceptor. We have previously demonstrated that VEGF165b-TMR selectively binds to NanoLuc-VEGFR2 (and not NRP1), whereas VEGF165a-TMR can bind to both NanoLuc-VEGFR2 or NanoLuc-NRP1 with nanomolar affinity (Peach et al., 2018a). At the complemented HiBiT complex, there was saturable binding in the presence of increasing concentrations of VEGF165b-TMR (Figure 6b) or VEGF165a-TMR (Figure 6c). This was displaced by a high concentration of unlabelled ligand, demonstrating low non-specific binding. Both fluorescent ligands had equilibrium dissociation constants (Kd) in the nanomolar range at the VEGFR2/NRP1 complex (VEGF165b-TMR Kd = 16.26 ± 3.81 nM, pKd = 7.82 ± 0.11; VEGF165a-TMR Kd = 2.53 ± 0.49, pKd = 8.61 ± 0.09; n=3 for both). Estimated ligand binding affinities were similar to those derived at isolated receptors tagged with full-length NanoLuc (Peach et al., 2018a).

Real-time kinetics of fluorescent VEGF-A isoforms at a heteromeric VEGFR2/NRP1 NanoBiT complex

Taking advantage of the NanoBiT approach to monitor real-time ligand binding at 37°C to a complex, we compared the kinetics of ligand binding of VEGF165b-TMR with that of VEGF165a-TMR at the VEGFR2/NRP1 NanoBiT complex in living cells. The kinetic binding profile of VEGF165b-TMR (which should only bind to VEGFR2; Peach et al., 2018a) continued to increase over the full 90 minute time course in intact cells, producing a classic ligand binding association maintained for each concentration of VEGF165b-TMR (Figure 7a). Fitted to a global association curve (Table 1), VEGF165b-TMR had a slightly slower association rate constant (kon) for the VEGFR2/NRP1 complex (2.29 x 106 ± 0.30 x 106min-1.M-1) compared to NanoLuc-VEGFR2 alone (7.29 x 106min-1.M-1; Peach et al., 2018a). We then directly compared the real-time binding profile for a saturating concentration of VEGF165b-TMR between the NanoBiT complex and cells expressing NanoLuc-tagged receptors alone in matched time course experiments (Figure 7b). Compared to NanoLuc-VEGFR2, the small decline in BRET signal after a peak at 20 minutes in intact cells was absent when monitored at the NanoBiT complex for VEGF165b-TMR. There was no BRET detected between VEGF165b-TMR and NanoLuc-NRP1, however this selective ligand had a distinct long-term kinetic profile at the VEGFR2/NRP1 complex compared to VEGFR2 alone (Figure 7b).
Kinetic experiments were repeated with four concentrations of VEGF165a-TMR (Figure 7c). Unlike VEGF165b-TMR, there was a small decline in BRET ratio between 30-60 minutes for VEGF165a-TMR at the HiBiT complex (Figure 7c). Association binding curves were globally fitted to kinetic data from the initial 20 minutes due to this decline (Table 1). VEGF165a-TMR had a slower dissociation rate constant (koff) at the HiBiT complex (0.046 ± 0.007 min-1; Table 1) compared to that previously reported for NanoLuc-NRP1 expressed alone (0.26 min-1; Peach et al., 2018a). As a consequence, the kinetic binding profile for 10 nM VEGF165a-TMR was directly compared between the NanoBiT complex and either NanoLuc-VEGFR2 or NanoLuc-NRP1 (Figure 7d). VEGF165a-TMR association kinetics at the NanoBiT complex in the initial 20 minutes were more comparable to NanoLuc-VEGFR2 than NanoLuc-NRP1 (NanoBiT kobs = 0.33 ± 0.04 min-1, NanoLuc-VEGFR2 kobs = 0.31 ± 0.03 min-1, NanoLuc-NRP1 kobs = 0.93 ± 0.09 min-1; n=5 per group). These observed rate constants were significantly slower at the complex than NRP1 alone (repeated-measures ANOVA and Holm-Šidák’s multiple comparisons;P< 0.05, n=5 for each). These data suggest that the ligand binding profile for VEGF165a-TMR at the NanoBiT complex reflected VEGFR2 binding kinetics, as opposed to the faster binding observed at NRP1.

Fluorescent VEGF-A kinetics were similar for the SmBiT Complex

Considering the distinct kinetic observations at the HiBiT complex, we further probed ligand binding kinetics at the SmBiT complex to explore possible influences of the NanoBiT tag characteristics (Dixon et al., 2016). Using four concentrations of VEGF165b-TMR, binding was monitored over 90 minutes (Figure 8a). The binding profile remained elevated throughout the time course with similarities to kinetics observed with the HiBiT complex. Kinetic data were globally fitted to a simple exponential association (Table 1). VEGF165b-TMR had a slower dissociation rate (koff) from the SmBiT complex compared to the HiBiT complex (Kruskal-Wallis test, P< 0.05, n=5 per group). Plotting the individual observed association rate constants (kobs) against VEGF165b-TMR concentration, there was a linear relationship observed at both HiBiT and SmBiT complexes (Figure 8b). The interaction between VEGF165b-TMR and the NanoBiT complex can therefore be defined as a first order reaction. Binding kinetics were also monitored at the SmBiT complex using four concentrations of VEGF165a-TMR (Figure 8c). Fitted data from the initial 20 minute period using a global fit, there were no differences between the association kinetic parameters derived for VEGF165a-TMR for the HiBiT and SmBiT complexes (Kruskal-Wallis test and Dunn’s multiple comparisons test,P> 0.05, n=5 per group). There was a linear relationship between the derived observed association rate (kobs) constants and VEGF165a-TMR concentration (Figure 8d). Despite having the potential to bind to both receptors within the complex, the interaction between VEGF165a-TMR and the NanoBiT complex could also be defined by a first order reaction.

Similar complex pharmacology using a binding-dead mutant of NRP1

In addition to comparing binding between selective and non-selective fluorescent VEGF-A isoforms, site-directed mutagenesis was an alternative approach to probe the contribution of NRP1 engagement to the pharmacological characteristics of the VEGFR2/NRP1 complex. Using a previously characterised binding-dead NRP1 mutant (Y297A; Herzog et al., 2011; Fantin et al., 2014; Peach et al., 2018a), comparisons were made using the same ligand in the absence of interactions between VEGF165a-TMR and NRP1 within the heteromeric NanoBiT complex (Figure 9a). Upon co-expression of LgBiT-VEGFR2 and either HiBiT- or SmBiT-NRP1 (Y297A), there were high luminescence emissions resulting from NanoBiT complementation (Figure 9b). Luminescence emissions from this NanoBiT complex were comparable to wild type NRP1, therefore this amino acid residue was not required for constitutive VEGFR2/NRP1 complex formation. NanoBiT constructs expressed in isolation from their complementary fragment also had minimal luminescence emissions in the presence of furimazine (Figure 9b). Isolating ligand binding from this VEGFR2/NRP1 Y297A complex, VEGF165a-TMR exhibited saturable binding at the NanoBiT complex (Figure 9c). This was displaced by a high concentration of unlabelled VEGF165a, confirming that there was low non-specific binding. Derived equilibrium dissociation constants were in the nanomolar range and similar to the wild type NanoBiT complex (VEGF165a-TMR/NanoBiT Y297A Kd = 1.55 ± 0.38; pKd 8.84 ± 0.11; n=3). Binding kinetics at the mutant NanoBiT complex were then monitored using four concentrations of VEGF165a-TMR (Figure 9d). This had an identical profile compared to VEGF165a-TMR at the wild type HiBiT complex (Figure 7c), whereby there was a small decline in BRET ratio following 30-60 minutes. Association kinetics were derived from the initial 20 minutes using a global fit (kon = 3.71x107 ± 0.21x107min-1M-1; koff = 0.054 ± 0.008 min-1; kinetic pKd = 8.85 ± 0.04; n=5). These data suggest that VEGF165a-TMR bound the NanoBiT complex with similar kinetics, regardless of the ability to simultaneously engage NRP1.