Response to “Binding of the C‐terminal amino acids of VEGF 121 directly with neuropilin‐1 should be considered”

We have recently found that neuropilin-1 and neuropilin-2 enhance VEGF121 induced signal transduction mediated by the VEGFR-2 tyrosine kinase receptor (1). This observation was surprising because several previous studies have shown that VEGF121 does not bind to native neuropilins of endothelial cells or to recombinant neuropilins (2, 3). Our observation suggests that neuropilins may enhance VEGFR-2-mediated VEGF signaling even when the VEGF binding site of neuropilin-1 is not occupied by VEGF. In their letter, Drs. Von-Wronski, Tweedle, and Nunn point out that several small peptides containing sequences homologous to the sequence found at the C-terminal of exon-8 of the VEGF gene are able to inhibit the binding of VEGF165 to neuropilin-1 and inhibit VEGF165 induced signaling. These results suggest that all VEGF forms containing this exon, including VEGF121, should be able to bind to neuropilin-1. Knowing that VEGF121 does not normally bind to neuropilins the authors of the letter suggest that neuropilin-1 may display increased affinity towards VEGF121 when in complex with VEGFR-2, thereby enabling binding of VEGF121 to neuropilin-1. This alternative hypothesis predicts that these peptides should be able to inhibit neuropilin-1-mediated enhancement of VEGFR-2 signaling induced by VEGF121, provided that the mechanism by which these peptides inhibit the biological effect of neuropi-lin-1 on VEGF signaling is indeed dependent on their ability to inhibit VEGF binding to neuropilin-1. When we noticed that neuropilins enhance VEGFR-2-mediated VEGF121 signaling, we did think of the possibility suggested by the authors of the letter. We published a manuscript several years ago in which we showed that VEGF121 is able to bind to neuropilin-1 and neuropilin-2 in the presence of VEGFR-1 (4). Therefore, following our initial observation, we conducted cross-linking experiments in which we cross- linked 125I-VEGF121 to PAE cells coexpressing VEGFR-2 and either neuropilin-1 or neuropilin-2. We used 125I-VEGF121 at concentrations of up to 100 ng/ml. However, we could not detect 125I-VEGF121/np1 or 125I-VEGF121/np2 complexes in these experiments. We did not show a figure containing these data in the manuscript due to their negative nature, but we described these experiments and their outcome in the results section. The results of these experiments represent a major reason for favoring the model we proposed in the manuscript. Secondly, neuro-pilin-1 enhanced 125I-VEGF165 binding to VEGFR-2 at low 125I-VEGF165 concentrations, but at these concentrations we could not observe binding to neuropilin-1, arguing independently that the VEGF binding site does not need to be occupied for neuropilin-1 mediated enhancement. Nevertheless, it is still possible, although we think less likely, that when in complex with VEGFR-2, neuropilins assume a conformation that allows VEGF121 to bind to neuropilins, but does not allow the homobifunctional cross-linkers we used to cross-link the bound VEGF121 to neuropilins, resulting in incorrect conclusions. If that is what happens, than this neuropilin conformation must differ considerably from the neuropilin conformation induced by VEGFR-1 since in that case we were able to cross-link VEGF121 to neuropilins (4). Examining this possibility may be worthwhile but it may not be easy to design such an experiment. It should be noted that it is not sufficient to show that a peptide such as Tuftsin inhibits neuropilin-enhanced VEGF121 activity, because the mechanism by which these peptides inhibit the biological effect of neuropilin-1 may be unrelated to their effect on VEGF165 binding to neuropilin-1.