Silky, sticky chimeras-designer VEGFs display their wares.

To the casual observer of the VEGF field, the panoply of players involved already looks sufficiently intimidating; 5 VEGF genes (A,B,C,D and E) with multiple splicing variants encoding proteins of varied functionality, 3 high-affinity tyrosine kinase receptors (VEGF-R1, 2 and 3), and at least 2 nontyrosine kinase receptors (neuropilin-1 and 2). One would imagine that with such a variety we would have encountered by now all possible variations on the theme of VEGF control of new vessel growth, arterial or lymphatic. But one would be wrong. Two reports in this issue of Circulation Research by Kari Alitalo and colleagues demonstrate how various VEGF domains dictate hitherto unappreciated biological nuances of vascular growth and how these can be combined into “designer” VEGFs with new activities not present in parent molecules.1,2 Principal VEGF-A domains include a signal peptide sequence, a VEGF homology domain (VHD) common to all VEGF but dictating specific VEGF receptor binding properties (VEGF-R1 and VEGF-R in the case of VEGF-A VHD), and a heparin binding domain (HBD) encoded by exons 6 and 7 in VEGF189 or a shorter HBD encoded only by exon 7 in VEGF-A165. Exon 7 also encodes a neurolin-1 binding domain present in both VEGF-A165 and A189 (Figure). Similarly, VEGF-C principal domains include the signal peptide domain and VEGF-C VHD (in this case dictating VEGF-R and VEGF-R3 binding) flanked by Nand C-terminal domains that normally undergo proteolytic process to form the mature protein (Figure). The C-terminal domain of VEGF-C, because of its cysteine-rich sequences eliciting comparisons with a protein component of silk, has been termed a silk homology domain. Although the significance of VHD in determining binding to a specific VEGF receptor and HBD in enhancing matrixbinding properties of VEGF-A have been long appreciated, the biological function of the silk homology domain of VEGF-C has not been determined. Likewise, it was also unclear how specific functions of one VEGF might be affected by introduction of non-VHD sequences from another VEGF. These questions are addressed in present studies. A chimeric protein consisting of the VEGF-A VHD and VEGF-C N-terminal and C-terminal cysteine-rich silk domain (A-VHD/CAC) induces formation of a strikingly abundant network of pericyte-covered vessels, dramatically distinct from VEGF-A-induced capillary network. At the same time, chimeras, consisting of the VEGF-C VHD and VEGF-A heparin binding domain (C-VHD/HBD) that includes exon 7 required for neuropilin binding, induce growth of lymphatics localized along tissues borders and basement membrane planes that differ equally dramatically from VEGF-C-induced lymphatics (Table). These results bring up several interesting questions and point to a new potential in “custom” growth factor design. Clearly, we still have much to learn about VEGF domain complexity. Various VEGF family members possess, in addition to their specific receptor binding domains, other structural motif that appear to have important biological function. The silk domain of VEGF-C and VEGF-D is one such example. Unexpectedly, introduction of a VEGF-C silk domain into VEGF-A significantly altered the latter’s biological activity without apparently altering its binding to its principal receptors, VEGF-R1 and 2. At the same time, the A-VHD/CAC chimera bound with much higher affinity to neuropilins 1 and 2, an interaction that was greatly facilitated by heparin. This greater binding to neuropilins was associated with a distinctly different type of a vascular network induced by the silk chimera, a dense web of capillaries enveloped by pericytes and extracellular matrix. This type of vasculature is in sharp contrast to the typical VEGF-A-induced vessels that tend to be larger in diameter, less packed and very permeable. In fact, increased permeability is a hallmark of VEGF-A-induced vasculature as seen, for example, in tumors, ischemic or fresh granulation tissues. That the VEGF-C silk domain should have such an effect is rather unexpected, as the lymphatic network induced by VEGF-C is typically neither dense nor impermeable. Silk proteins (spidroins and fibroins) are assembled into well-defined insoluble nanofibrills that have a pronounced tendency to self-aggregation.3 Silk domains in nonsilk proteins are a poorly understood structure that may promote formation of silk-like disulfide-bonded multimers that may physically facilitate cell-cell association accounting for increased network density. However, the capacity of the silk domain to promote VEGF-neuropilin (Nrp) interaction is rather unexpected. Furthermore, Nrp-1 signaling is reported to regulate VEGF-induced permeability increase.4 Given then decreased permeability of the A-VHD/CAC chimera-induced vasculature, one could speculate that the chimera somehow binds to but does not activate Nrp-1, or perhaps even suppress its signaling. Alternatively, because Nrp-1 is expressed in The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Angiogenesis Research Center, Section of Cardiology, Departments of Medicine and Pharmacology and Toxicology, Dartmouth Medical School, Dartmouth-Hitchcock Medical Center, Lebanon, NH. Correspondence to Michael Simons, MD, Section of Cardiology, DHMC, One Medical Center Drive, Lebanon, NH 03756. Phone: E-mail [email protected] (Circ Res. 2007;100:1402-1404.) © 2007 American Heart Association, Inc.