Antiangiogenic Therapeutic Potential of Peptides Derived from the Molecular Motor KIF13B that Transports VEGFR2 to Plasmalemma in Endothelial Cells

Abstract

Vascular endothelial growth factor receptor 2 (VEGFR2) localized on the surface of endothelial cells (ECs) is a key determinant of the magnitude and duration of angiogenesis induced by vascular endothelial growth factor (VEGF). The kinesin family plus-end motor KIF13B transports VEGFR2 to the EC surface, and as such, specific inhibition of polarized VEGFR2 trafficking prevents angiogenesis. We designed a series of bioactive peptides based on deep analysis of VEGFR2-binding domain of KIF13B that compete specifically with VEGFR2 binding of KIF13B and thereby potently inhibit angiogenesis. Expression of these peptides by lentivirus prevents VEGF-induced capillary network formation in Matrigel plugs and neovascularization in vivo. A synthetic soluble, cell-permeable, 23-amino acid peptide termed kinesin-derived angiogenesis inhibitor (KAI) not only prevents interaction of VEGFR2 with KIF13B but also trafficking of VEGFR2 in the plus-end direction to the EC plasmalemma. Kinesin-derived angiogenesis inhibitor also inhibits VEGF-induced EC migration and tumor growth in human lung carcinoma xenografted in immunodeficient mice. Thus, we describe a novel class of peptides derived from the site of interaction of KIF13B with VEGFR2 that inhibit VEGFR2 trafficking and thereby starve cancer of blood supply.

Figure 1

Identification of KIF13B domains inhibiting the interaction of KIF13B with vascular endothelial growth factor receptor 2 (VEGFR2). A: Schematic of the domains of KIF13B and truncated domains used in the study. B: For the in vitro binding assay, the domains of KIF13B [domains of unknown function (DUFs) 2A, 2B, and 2C] were expressed in bacteria and tested for binding to VEGFR2 in human primary umbilical vein endothelial cell (HUVEC) lysates by the pull-down assay. VEGFR2 was detected by Western blotting using an antibody. C:In vitro two-dimensional capillary network formation of HUVECs treated with either control virus (FLAG vector) or virus encoding truncated mutants of FLAG-KIF13B in Matrigel supplemented with VEGF (2.2 nmol/L). D and E: Vascular endothelial growth factor (VEGF)– and sphingosine-1-phosphate (S1P)–induced invasion in three-dimensional collagen matrices. HUVECs were infected without virus or with control versus FLAG-DUF2C or FLAG-DUF2 for 2 days and tested for invasion into collagen in the presence of S1P or VEGF (2.2 nmol/L). Results are representative of three individual experiments. Quantification of the collagen invasion assay is given in D by measuring distance of invading endothelial cells. Data are expressed as means ± SEM (C and D). n = 15 and 9 (C); n = 50 sprouts (D). ^∗P < 0.05. Scale bars: 200 μm(C); 100 μm (D). aa, amino acid; Cap-Gly, cytoskeleton-associated protein glycine-rich; FHA, forkhead associated; GST, glutathione S transferase; MBS, membrane-associated guanylate kinase–binding stalk; pro, proline rich; S, oligopeptide derived from RNase A (alias S-tag).

Figure 2

Peptides derived from KIF13B inhibit vascular endothelial growth factor (VEGF)–induced angiogenesis. A: Schematic of the domains of KIF13B and truncated domains used in the study. Domains of unknown function (DUF) 2C1 [1202-1240 amino acids (aa)], C2 (1221-1260 aa), C3 (1241-1281 aa), C4 (1226-1251 aa), C5 (1238-1260 aa), C6 (1238-1254 aa), C7 (1261-1281 aa), C8 (1251-1268 aa), and C9 (1235-1252 aa) were expressed as recombinant proteins in bacteria and tested for binding to vascular endothelial growth factor receptor 2 (VEGFR2) by pull-down assay. O, X, and triangle indicate binding, no binding, and partial or unstable binding, respectively. Binding region is indicated as pink shadow. B and C:In vitro two-dimensional capillary network formation of human primary umbilical vein endothelial cells treated with either control virus or virus encoding truncated mutants of FLAG-KIF13B in Matrigel supplemented with VEGF. C_T (1528-1826 aa) was used as a negative control because it did not bind to VEGFR2.D and E: Hematoxylin and eosin staining of Matrigel plug was supplemented with 4.4 nmol/L VEGF, 50 ng/mL of basic fibroblast growth factor, 60 U of heparin, and 0.8 × 10⁸ IFU lentivirus (vector control, DUF2, or DUF2C5, or C_T) and injected s.c. in C57BL6 mice. Data are expressed as means ± SEM (C and E). n = 15, 9, 6, 6, 6, 6, 6, and 3 for the groups, respectively (C); n = 4, 4, 5, and 5 for vector control, DUF2, DUF2C5, and C_T, respectively (E). ^∗P < 0.05 (one-way analysis of variance). Scale bars = 100 μm (B and D). IFU, infectious unit; KAI, kinesin-derived angiogenesis inhibitor.

Figure 3

The peptide kinesin-derived angiogenesis inhibitor (KAI) inhibits vascular endothelial growth factor receptor 2 (VEGFR2) trafficking to endothelial cell surface by preventing binding of VEGFR2 cargo to KIF13B. A: Competitive binding assay performed in human primary umbilical vein endothelial cells (HUVECs). HUVECs were preincubated with KAI for 1 hour and stimulated with vascular endothelial growth factor (VEGF) (2.2 nmol/L) for 1 hour. Immunoprecipitation (IP) of VEGFR2 with anti-KIF13B antibody was detected with anti-VEGFR2 antibody. B and C: Effects of peptide on VEGFR2 transport to the cell surface after VEGF (2.2 nmol/L) treatment of HUVECs. Data are expressed as means ± SEM (A and B). n = 3 (A); n = 14, 11, and 23 for control, KAI 3 μmol/L, and KAI 10 μmol/L, respectively. ^∗P < 0.05 (t-test and one-way analysis of variance). Scale bars = 50 μm (C).

Figure 4

Kinesin-derived angiogenesis inhibitor (KAI) inhibits vascular endothelial growth factor (VEGF)–induced endothelial cell (EC) migration. A: Schematic of domains of KIF13B and synthesized peptides used in the study. B: Pull-down assay for interaction of peptide KAI to vascular endothelial growth factor receptor 2 (VEGFR2), whereas CT23 control peptide did not interact with VEGFR2. C: VEGF-mediated migration of human primary umbilical vein endothelial cells (HUVECs) determined by wound healing scratch assay in the presence of 2.2 nmol/L VEGF with or without KAI (1 μmol/L). At indicated times, cells were fixed and stained with hematoxylin. D: Specificity of inhibitory effect of KAI determined by EC migration induced by different stimuli in Transwell migration assay as described. HUVECs migrated toward 2.2 nmol/L VEGF, 50 ng/mL basic fibroblast growth factor (bFGF), or 1 μmol/L sphingosine-1-phosphate (S1P) were visualized by hematoxylin staining, and number of migrated cells were counted. Data are expressed as means ± SEM (C and D). n = 3 (C and D). ^∗P < 0.05 (one-way analysis of variance and Bonferroni multiple comparisons test). Scale bars = 200 μm (C and D). Cap-Gly, cytoskeleton-associated protein glycine-rich; FHA, forkhead associated; GST, glutathione S transferase; MBS, membrane-associated guanylate kinase–binding stalk; pro, proline rich; S, oligopeptide derived from RNase A (alias S-tag).

Figure 5

Kinesin-derived angiogenesis inhibitor (KAI) inhibits endothelial cell migration, capillary network formation, and sprouting. A and B:In vitro two-dimensional capillary network formation. Human primary umbilical vein endothelial cells (HUVECs) were treated with phosphate-buffered saline (PBS) vehicle, synthetic peptides KAI, or CT23 at the indicated doses and tested for capillary network formation in Matrigel in the presence of 2.2 nmol/L vascular endothelial growth factor (VEGF). C–E:In vitro three-dimensional sprouting assay in fibrin gel supplemented with VEGF (2.2 nmol/L). HUVECs were treated PBS, KAI, or CT23 1 μmol/L as indicated. Data are expressed as means ± SEM (B, D, and E). n = 10, 4, 6, 4, 4, 3, and 3 in the groups, respectively (B); n = 11, 9, and 6 for the control, KAI, and CT23 treated groups, respectively (D and E). ^∗P < 0.05 (one-way analysis of variance). Scale bars: 200 μm (A); 50 μm (C).

Figure 6

Kinesin-derived angiogenesis inhibitor (KAI) inhibits angiogenesis and tumor growth in vivo. A: Study of human lung carcinoma H460 xenograft in mice with severe combined immunodeficiency. Mice were treated with KAI (10 mg/kg) or phosphate-buffered saline (PBS), i.v. via tail vein, three times per week, and tumor size is shown in the graph. B and C: Number of vessels in tumor in PBS-treated control and peptide-treated tumor. Representative immunohistochemistry of VEGFR2 and von Willebrand factor (vWF) are shown to visualize blood vessels. D: Apoptotic cells visualized with terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay in tumor tissue treated with PBS vehicle or KAI. Data are expressed as means ± SEM (A, B, and D). n = 8 in each group. ^∗P < 0.05, ^∗∗P < 0.01 (t-test). Scale bars = 50 μm (C and D).