pubmed.ncbi.nlm.nih.gov

VEGF (Vascular Endothelial Growth Factor) Induces NRP1 (Neuropilin-1) Cleavage via ADAMs (a Disintegrin and Metalloproteinase) 9 and 10 to Generate Novel Carboxy-Terminal NRP1 Fragments That Regulate Angiogenic Signaling - PubMed

  • ️Invalid Date

VEGF (Vascular Endothelial Growth Factor) Induces NRP1 (Neuropilin-1) Cleavage via ADAMs (a Disintegrin and Metalloproteinase) 9 and 10 to Generate Novel Carboxy-Terminal NRP1 Fragments That Regulate Angiogenic Signaling

Vedanta Mehta et al. Arterioscler Thromb Vasc Biol. 2018 Aug.

Abstract

Objective- NRP1(neuropilin-1) acts as a coreceptor for VEGF (vascular endothelial growth factor) with an essential role in angiogenesis. Recent findings suggest that posttranslational proteolytic cleavage of VEGF receptors may be an important mechanism for regulating angiogenesis, but the role of NRP1 proteolysis and the NRP1 species generated by cleavage in endothelial cells is not known. Here, we characterize NRP1 proteolytic cleavage in endothelial cells, determine the mechanism, and investigate the role of NRP1 cleavage in regulation of endothelial cell function. Approach and Results- NRP1 species comprising the carboxy (C)-terminal and transmembrane NRP1 domains but lacking the ligand-binding A and B regions are constitutively expressed in endothelial cells. Generation of these C-terminal domain NRP1 proteins is upregulated by phorbol ester and Ca2+ ionophore, and reduced by pharmacological inhibition of metalloproteinases, by small interfering RNA-mediated knockdown of 2 members of ADAM (a disintegrin and metalloproteinase) family, ADAMs 9 and 10, and by a specific ADAM10 inhibitor. Furthermore, VEGF upregulates expression of these NRP1 species in an ADAM9/10-dependent manner. Transduction of endothelial cells with adenoviral constructs expressing NRP1 C-terminal domain fragments inhibited VEGF-induced phosphorylation of VEGFR2 (VEGF receptor tyrosine kinase)/KDR (kinase domain insert receptor) and decreased VEGF-stimulated endothelial cell motility and angiogenesis in coculture and aortic ring sprouting assays. Conclusions- These findings identify novel NRP1 species in endothelial cells and demonstrate that regulation of NRP1 proteolysis via ADAMs 9 and 10 is a new regulatory pathway able to modulate VEGF angiogenic signaling.

Keywords: cytokines; endothelial cells; phosphorylation; proteolysis; signal transduction.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.

Endothelial expression of C-terminal NRP1 (neuropilin-1) fragments. A, Lysates of cytoplasmic (Cyt, C) and nuclear extracts (Nuc, N) of human umbilical vein endothelial cells (HUVECs) were immunoblotted with NRP1 antibody specific to the cytoplasmic domain of NRP1 (C-term; antibody C-19 from Santa Cruz Inc). This antibody recognized several low (30 kDa and below) molecular weight species predominantly in the cytoplasmic fraction. B, HUVECs were transfected with control scrambled (siScr) and NRP1-specific siRNAs, and whole cell protein lysates immunoblotted with NRP1 antibody specific either to the cytoplasmic (C-term) or extracellular (N-term; antibody AF387) domains of NRP1. Knockdown of endogenous NRP1 using siRNA resulted in diminished expression of not only the full-length NRP1 band but also of the cytoplasmic fragments. C, HUVECs were transduced with adenoviruses encoding wild-type (WT) NRP1 (WT), an NRP1ΔC mutant lacking the cytoplasmic domain (ΔC), or GFP (green fluorescent protein), and 48 h later cytoplasmic (C) or nuclear (N) cell lysates were immunoblotted with antibodies either specific for the NRP1 cytoplasmic domain (C-term), or specific to the NRP1 extracellular domain (N-term), or for β-actin, glyceraldehyde 3-phosphate dehydrogenase (GAPDH; cytoplasmic marker), or lamin B (nuclear marker). Overexpression of WT NRP1 by an adenovirus (WT), but not of the NRP1ΔC mutant (lacking the cytoplasmic domain; ΔC) results in the generation of low molecular weight cytoplasmic fragments that can be detected only by antibody specific for the NRP1 cytoplasmic domain, but not by antibody specific for the NRP1 extracellular domain.

Figure 2.
Figure 2.

The generation of NRP1 (neuropilin-1) cytoplasmic fragments in human umbilical vein endothelial cells (HUVECs) is mediated by metalloproteinase activity. A, Increasing concentrations of marimastat, a broad-spectrum metalloproteinase inhibitor, reduces the expression of the 10 and 15 kDa fragments detected by antibody specific for the NRP1 cytoplasmic domain (C-term) in a dose-dependent manner. B, The dose-response curve for the effect of marimastat on generation of the 10 kDa NRP1 cytoplasmic fragment indicates an IC50 of 10 μmol\L, similar to reported values for marimastat. C, HUVECs, either infected with Ad.NRP1WT or uninfected (UI) were treated for 24 h with Phorbol ester (PMA), or Ionomycin (IM), or vehicle (DMSO) control (VC), at the indicated concentrations, and cell lysates were then prepared and immunoblotted as shown.

Figure 3.
Figure 3.

NRP1 (neuropilin-1) cleavage in endothelial cells is mediated by ADAMs (a disintegrin and metalloproteinases) 9 and 10. Human umbilical vein endothelial cells (HUVECs) were transfected with control siRNA (siScr) or siRNAs specific for ADAM9 (A) or ADAM10 (B), and after 72 h were then treated with 25 ng/mL VEGF (vascular endothelial growth factor) for 60 min; cell lysates were then immunoblotted with the antibodies indicated. C, HUVECs were treated for 24 h with either no additions, or with DMSO (vehicle), or with the indicated concentrations of GI254023X, a specific inhibitor of ADAM10, and cell lysates were then immunoblotted with the antibodies indicated; quantification of the 10 kDa NRP1 cytoplasmic domain band is shown, *P<0.05 vs DMSO treatment, n≥3. D, Effects of single and double knockdown of ADAM9 and ADAM10 on the expression of NRP1 cytoplasmic domain fragments. The blots shown are representative of 4 different experiments. E, Quantification of the 10 kDa NRP1 cytoplasmic fragment from experiments in D, normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression, after knockdown of ADAMs 9 or 10 using single siRNAs at 200 nmol/L (black symbols), or double knockdown of ADAM9 plus ADAM10 (unfilled symbols) using combinations of siRNAs (total siRNA concentration 400 nmol/L); *P<0.05 vs Scr siRNA (200 nm) or Scr siRNA (400 nm) as appropriate. Values are presented as a scatterplot. Differences between samples were analyzed using 1-way ANOVA with the Bonferroni correction for multiple pairwise comparisons after testing for normality and equal variance using the Shapiro-Wilk and Levene tests, respectively.

Figure 4.
Figure 4.

ADAMs (a disintegrin and metalloproteinases) 9 and 10 mediate VEGF (vascular endothelial growth factor)-induced NRP1 (neuropilin-1) proteolytic cleavage. A, Human umbilical vein endothelial cells (HUVECs) were transfected with siRNAs as indicated, and after 72 h were then treated with 25 ng/mL VEGF-A for 60 min, and lysates were then immunoblotted as indicated. B, Quantification of the 10 kDa NRP1 cytoplasmic domain fragment in experiments in A in which HUVECs were treated without (C, unfilled symbols) or with VEGF-A (V, black symbols); VEGF significantly enhanced the generation of NRP1 cytoplasmic fragments in siScr-treated cells and this was blocked by siADAM9 (left graph) or siADAM10 (right graph); #P<0,05 vs siScr C, *P<0.05; **P<0.01; ***P<0.005 vs siScr plus VEGF (V; n≥3). C, HUVECs were either transfected with control (siScr) and ADAM10 specific siRNAs and were then either untreated (C, control, unfilled symbols)) or treated with 25 ng/mL VEGF (V, black symbols) for 60 min. Cell supernatant was then removed, concentrated, and immunoblotted NRP1 antibody specific for the extracellular domain with (AF3870) to detect sNRP1 (soluble NRP1). Quantification of results from 3 independent experiments is shown below; **P<0.01 for siADAM10 vs siScr. D, HUVECs were untreated (C, control, unfilled symbols), or were treated for 60 min with 25 ng/mL VEGF plus either DMSO (V, black symbols) or with 25 ng/mL VEGF plus an equal volume of GI254023X (V+G, black symbols). Quantification of results from 3 independent experiments is shown below; *P<0.05 for VEGF plus GI254023X (10 μmol/L) vs VEGF plus DMSO. In this figure, values are presented as a scatterplot. Differences between samples were analyzed using 1-way ANOVA with the Bonferroni correction for multiple pairwise comparisons after testing for normality and equal variance using the Shapiro-Wilk and Levene tests, respectively.

Figure 5.
Figure 5.

Protein sequencing of NRP1 (neuropilin-1) cleavage products. A, Human umbilical vein endothelial cells (HUVECs) were transduced with Ad.NRP1WT and after 48 h were treated with 30 ng/mL PMA for 24 h in the absence or presence of the ADAM (a disintegrin and metalloproteinase) 10 inhibitor, GI254023X (10 μmol/L). Cells were lysed, immunoprecipitated using antibody specific for the NRP1 cytoplasmic domain (IP: NRP1 C-term; C-19), and proteins were stained with Ponceau S to detect bands. In parallel, HUVECs transduced with Ad.NRP1WT for 48 h were lysed and lysates incubated with IgG control. Whole cell lysate of PMA-treated Ad.NRP1WT-infected cells is also shown for comparison. B, HUVECs were transduced with Ad.NRP1WT and after 48 h were treated with 30 ng/mL PMA for 24 h. Cells were then lysed and full-length NRP1 and the low molecular weight cleavage products were immunoprecipitated using the C-19 antibody specific for the NRP1 cytoplasmic domain, and immunoprecipitated proteins separated by SDS-PAGE. Proteins were transferred onto a PVDF membrane and stained with Ponceau S to detect bands. The low molecular weight cleavage products, indicated as bands 1 to 4, were cut and sent for N-terminal sequencing by Edman degradation. C, Amino acid sequence of NRP1, N-terminal sequences detected by N-terminal sequencing by Edman degradation are highlighted in red within the NRP1 sequence. A schematic diagram representing the putative sites of NRP1 cleavage by members of the ADAM family, resulting in the generation of cytoplasmic fragments is shown.

Figure 6.
Figure 6.

Overexpression of NRP1 (neuropilin-1) cytoplasmic domain fragments in human umbilical vein endothelial cells (HUVECs) inhibits VEGF (vascular endothelial growth factor)-induced angiogenesis. A, Schematic diagram representing adenoviral NRP1 constructs comprising the cytoplasmic and transmembrane domains (Ad.NRP1Cyt-TM, residues 860–923), the cytoplasmic, transmembrane, and juxtamembrane regions (Ad.NRP1Cyt-JM, residues 797–923), and the cytoplasmic, transmembrane, juxtamembrane, and MAM domains (Ad.NRP1Cyt-MAM, residues 638–923). B, HUVECs transfected with adenoviruses overexpressing LacZ, wild-type NRP1 (Ad.NRP1WT), or the cytoplasmic domain species (Ad.NRP1Cyt-TM, Cyt-JM and Cyt-MAM, respectively) were used in a Transwell migration assay with (+, black symbols) and without (−, unfilled symbols) VEGF treatment (25 ng/mL, 4 h); the means±SEM of results from 3 independent assays are shown, P<0.05 vs Ad.NRP1WT plus VEGF. C, Aortic rings were incubated with the indicated adenoviruses in Opti-MEM overnight. The aortic ring assay was performed as detailed in the Materials and Methods with no treatment (−, unfilled symbols) or with VEGF-A165 treatment (+, black symbols). Quantification of the number of branch points (left graph) and network area (right graph) are shown below the representative figures; **P<0.01, ***P<0.001 vs Ad.NRP1 plus VEGF, n=3 (each n includes aortic rings from 4 mice, to have sufficient sample to set up each condition using 6 replicate aortic rings). In this figure, values are presented as a scatterplot. Differences between samples were analyzed using 2-way ANOVA with the Bonferroni correction for multiple pairwise comparisons after testing for normality and equal variance using the Shapiro-Wilk and Levene tests, respectively.

Figure 7.
Figure 7.

Overexpression of NRP1 (neuropilin-1) cytoplasmic fragments in human umbilical vein endothelial cells (HUVECs) inhibits KDR (kinase domain insert receptor) activation but not VEGF (vascular endothelial growth factor)-induced NRP1 association with KDR. A, HUVECs were transfected with adenoviruses overexpressing LacZ, wild-type NRP1 (Ad.NRP1WT), or the low molecular weight species (Ad.NRP1Cyt-TM, Cyt-JM, and Cyt-MAM, respectively). After treatment without (−, unfilled symbols) or with VEGF-A165 (+, black symbols) for 10 min, cells were lysed and immunoblotted with the indicated antibodies. Quantification of the blots is shown on the right for KDR (n=3; *P<0.05 vs Ad.NRP1WT plus VEGF). B, HUVECs overexpressing WT and mutant NRP1s as detailed above were stimulated with VEGF for 10 min and KDR was immunoprecipitated. Immunoprecipitates were immunoblotted with an antibody to NRP1. Blots demonstrating equal KDR immunoprecipitation for each condition are shown on the right-hand side. In this figure, values are presented as a scatterplot. Differences between samples were analyzed using 2-way ANOVA with the Bonferroni correction for multiple pairwise comparisons after testing for normality and equal variance using the Shapiro-Wilk and Levene tests, respectively.

Similar articles

Cited by

References

    1. Kawasaki T, Kitsukawa T, Bekku Y, Matsuda Y, Sanbo M, Yagi T, Fujisawa H. A requirement for neuropilin-1 in embryonic vessel formation. Development. 1999;126:4895–4902. - PubMed
    1. Kitsukawa T, Shimizu M, Sanbo M, Hirata T, Taniguchi M, Bekku Y, Yagi T, Fujisawa H. Neuropilin-semaphorin III/D-mediated chemorepulsive signals play a crucial role in peripheral nerve projection in mice. Neuron. 1997;19:995–1005. - PubMed
    1. Pellet-Many C, Frankel P, Jia H, Zachary I. Neuropilins: structure, function and role in disease. Biochem J. 2008;411:211–226. doi: 10.1042/BJ20071639. - PubMed
    1. Pan Q, Chathery Y, Wu Y, Rathore N, Tong RK, Peale F, Bagri A, Tessier-Lavigne M, Koch AW, Watts RJ. Neuropilin-1 binds to VEGF121 and regulates endothelial cell migration and sprouting. J Biol Chem. 2007;282:24049–24056. doi: 10.1074/jbc.M703554200. - PubMed
    1. Gerhardt H, Ruhrberg C, Abramsson A, Fujisawa H, Shima D, Betsholtz C. Neuropilin-1 is required for endothelial tip cell guidance in the developing central nervous system. Dev Dyn. 2004;231:503–509. doi: 10.1002/dvdy.20148. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources