Dynamics of VEGF-A binding at VEGFR2 and NRP1

Vascular Endothelial Growth Factor A (VEGF-A) is a key mediator of angiogenesis, a process dysregulated in tumour development. Alternative splicing of the Vegfa gene leads to distinct endogenous VEGF-A isoforms with differential signalling and expression. These include pro-angiogenic VEGF165a, ‘anti-angiogenic’ VEGF165b and freely diffusible VEGF121a. VEGF-A primarily signals via its cognate receptor tyrosine kinase (RTK), VEGF Receptor 2 (VEGFR2). Signalling can be potentiated by its co-receptor Neuropilin-1 (NRP1), a transmembrane protein also found at high levels in malignant tumours. Despite numerous approved anti-cancer therapeutics targeting VEGF-A/VEGFR2 signalling, there is limited quantitative pharmacological understanding of VEGF-A isoforms at full-length VEGFR2 or NRP1 in living cells. This thesis explored the spatial and temporal dynamics of ligand binding to VEGFR2 and its co-receptor NRP1. First, VEGF165a, VEGF165b and VEGF121a were stoichiometrically labelled with tetramethylrhodamine (TMR) in collaboration with Promega Corporation. VEGFxxxx-TMR ligand binding was quantified in real-time at 37°C using bioluminescence resonance energy transfer (BRET) with VEGFR2 or NRP1 tagged with NanoLuciferase (NanoLuc). This technique was used to discriminate between VEGF-A binding to two distinct classes of membrane protein expressed in isolation in HEK293T cells. While all VEGF-A isoforms had similar nanomolar affinities at NanoLuc-VEGFR2, not all isoforms were able to interact with NanoLuc-NRP1. This also revealed marked differences in the kinetic binding profiles of VEGF165a-TMR for NRP1 and VEGFR2, despite similar binding affinities. Using live cell imaging, we identified differences between the localisation of HaloTag-VEGFR2 and HaloTag-NRP1. Whereas NRP1 remained at the plasma membrane, VEGFR2 was subject to constitutive and ligand-driven endocytosis in HEK293T cells. Second, we investigated the relationship between receptor localisation and ligand binding given the complex trafficking of VEGFR2 observed in Chapter 3. Each fluorescent VEGF-A isoform was internalised with VEGFR2 within 30 minutes. At NanoLuc-VEGFR2, there was a decline in BRET signal for each fluorescent VEGF-A isoform following a peak at 20 minutes in living cells. This was absent for ligand binding at NanoLuc-NRP1. We further exploited these techniques to gain insight into how inhibition of VEGFR2 phosphorylation influenced ligand binding and endocytosis using a tyrosine phosphorylation deficient receptor mutant. In the absence of phosphorylation, there was an elevation in the BRET signal upon stimulation with fluorescent VEGF-A. VEGFR2 phosphorylation at Y951, Y1054, Y1059, Y1175 or Y1214 was not required for endocytosis. Membrane preparations were then used to probe VEGF-A/VEGFR2 binding in the absence of this endocytic component. Here, ligand binding profiles were maintained for 90 minutes and reached equilibrium. This assay was exploited to directly probe how ligand/receptor interactions were influenced by the acidic pH in the endosomal microenvironment. Interestingly, VEGF-TMR had a shorter residence time at NanoLuc-VEGFR2 at a pH similar to that in endosomes. Third, we investigated the effect of co-expressing VEGFR2 and NRP1 in the same living cell. Colocalisation was monitored between HaloTag-VEGFR2 and SnapTag-NRP1 in live cells upon stimulation with VEGF165b or VEGF165a. Using receptor-receptor BRET, we confirmed that VEGFR2 and NRP1 were in close proximity in the absence of ligand. We isolated the real-time pharmacology of VEGFxxxx-TMR at a defined VEGFR2/NRP1 complex using split NanoLuc Binary Technology (NanoBiT). As NanoBiTs require complementation to emit luminescence, BRET can only occur from a heteromeric complex of LgBiT-VEGFR2 and HiBiT-NRP1. Despite having faster kinetics at NRP1 in isolation, VEGF165a-TMR bound to the VEGFR2/NRP1 complex with dynamics comparable to those of NanoLuc-VEGFR2. VEGF165b-TMR had a ligand binding profile that largely remained elevated in cells over 90 minutes, despite being selective for VEGFR2. This thesis applied quantitative technologies to monitor real-time ligand binding at receptors that contribute to physiological and patho-physiological angiogenesis. These findings have implications for how NRP1 modulates VEGFR2 as a potential target in drug discovery.