Abstract
Vascular endothelial growth factor-A isoform 165 (VEGF-A165), a critical regulator of angiogenesis, has been extensively investigated as a therapeutic agent for ischaemic diseases. Nevertheless, therapies based on recombinant human VEGF-A165 proteins have failed to achieve clinical success. Glycoengineering has emerged as a tool for improving the utility of therapeutic proteins. Here, recombinant and synthetic forms of VEGF-A165, with no, natural, and engineered glycans, were used to investigate the impact of glycosylation on VEGF-A165 binding to its receptors.
Recombinant human VEGF-A165 proteins with no and native glycosylation were purified using chromatographic methods from mammalian (human embryonic kidney cells) and bacterial (Escherichia coli) production systems, respectively. Methods for folding VEGF-A165 dimers were optimised using the bacterially-produced protein, then applied to chemically-synthesised VEGF-A165 polypeptides with a single neutral (glucose, galactose, mannose) or charged glycan (sialic acid) attached at Asn-75 (site of native VEGF-A165 N-linked glycosylation). The extent of VEGF-A165 dimerisation was assessed through electrophoretic, protein staining, and Western blotting analyses. VEGF-A165 dimer activity was confirmed using competitive VEGF receptor (VEGFR) binding assays. The impact of glycosylation on VEGF-A165-induced cell proliferation was then assessed using VEGFR-specific bioassay cell lines.
This study marks the first successful folding of synthetic VEGF-A165 dimers, with and without glycans attached. A dialysis method which transitioned the synthetic VEGF-A165 proteins from a buffer containing the denaturant urea, cosolvent glucose and thiol reagents reduced glutathione/oxidized glutathione into acetic acid then Milli-Q water supported 70-86 % dimerisation. While the highest and lowest folding extents were achieved with the sialic acid and mannose VEGF- A165 glycoforms, respectively. The synthetic VEGF-A165 dimers exhibited similar efficacy and potency to bacterial- and mammalian-produced VEGF-A165 in an in vitro assay utilising VEGFR-1- or VEGF-2-immunoglobulin G fusion proteins. However, the synthetic VEGF-A165 with sialic acid attached was more potent at inducing proliferation of bioassay cells expressing chimeric VEGFR-1 or VEGFR-2 than the synthetic proteins with no or neutral glycans. By contrast, no differences in cell proliferation were observed between the recombinant VEGF-A165 proteins with no or native glycosylation.
These findings provide the first evidence that the glycan type or charge rather than the presence or absence of glycosylation may impact VEGF-A165 bioactivity. However, further analyses are needed to ensure equivalent disulfide bond formation between synthetic proteins, establish the mechanism by which sialic acid may enhance VEGFR interactions, and elucidate the ideal glycan chain length for VEGF-A165 activity. But the findings suggest that the pharmacological and biological attributes of sialic acid glycoforms of VEGF-A165 should be explored further. This study was therefore a successful first step in the use of glycoengineering to improve the utility of VEGF-A165 as a pro- angiogenic therapy.