Endothelial barrier disruption by VEGF-mediated Src activity potentiates tumor cell extravasation and metastasis

Abstract

VEGF is unique among angiogenic growth factors because it disrupts endothelial barrier function. Therefore, we considered whether this property of VEGF might contribute to tumor cell extravasation and metastasis. To test this, mice lacking the Src family kinases Src or Yes, which maintain endothelial barrier function in the presence of VEGF, were injected intravenously with VEGF-expressing tumor cells. We found a dramatic reduction in tumor cell extravasation in lungs or livers of mice lacking Src or Yes. At the molecular level, VEGF compromises the endothelial barrier by disrupting a VE-cadherin-beta-catenin complex in lung endothelium from wild-type, but not Yes-deficient, mice. Disrupting the endothelial barrier directly with anti-VE-cadherin both amplifies metastasis in normal mice and overcomes the genetic resistance in Yes-deficient mice. Pharmacological blockade of VEGF, VEGFR-2, or Src stabilizes endothelial barrier function and suppresses tumor cell extravasation in vivo. Therefore, disrupting Src signaling preserves host endothelial barrier function providing a novel host-targeted approach to control metastatic disease.

Figure 1.

Tumor cell accumulation hours-to-days after tumor cell injection. To determine a time line for CT26 tumor cell extravasation after i.v. inoculation, we prepared lungs for analysis by transmission EM. By 1–3 h, tumor cells (T) had extended protrusions toward endothelial cells (EC) (A) or extended processes between EC junctions to contact the underlying basal lamina (BL) (B). (C) At 3 h, we observed individual extravasated tumor cells immediately outside the blood vessels where they were typically lodged in the extracellular space between an EC and a pneumocyte (PC). (D) This ultimately gave way to metastatic foci containing numerous tumor cells by day 4. Bars: (A–C) 1 μm; (D) 5 μm.

Figure 2.

VEGF-expressing tumor cells show enhanced metastasis. ID8 murine ovarian carcinoma cells, which stably express GFP, were injected i.v. to form pulmonary metastatic lesions. (A) Fresh lung tissue was examined using laser scanning confocal microscopy to detect GFP-positive metastatic tumor cells. After 9 d, considerably more metastatic lesions were visible in lungs of mice injected with cells expressing VEGF (ID8-VEGF-GFP) compared with cells expressing GFP alone (ID8-GFP). (B) Lung weight due to tumor was significantly increased. * indicates P < 0.05; n = 8 each bar. Bar, 1 mm.

Figure 3.

Gene-targeted deletion of individual Src kinases protects against tumor cell extravasation. (A–C) Tumor burden, computed as the increase in lung/heart weight ratio over that for normal mice, was measured 12 d after i.v. inoculation with metastatic tumor cells. Mice lacking Src (A) or Yes (B and C) are protected from experimental pulmonary metastasis of CT26 (A and B) or D121 (C) cells. (D) This phenomenon is not specific to the lung, because Yes-deficient mice are also protected from experimental hepatic metastasis of CT26 or D121 cells. (E) Growth of primary tumors on the flank is not different between genotypes for CT26 or D121 cells. (F) Mice lacking Fyn, which show a normal vascular permeability response to VEGF, are not protected from metastasis. * indicates P < 0.05; n = 8 each bar. Bar, 1 mm.

Figure 4.

Disruption of cadherin-mediated adhesion enhances tumor cell extravasation. (A) Direct i.v. injection of VEGF induces rapid and transient dissociation of β-catenin from VE-cadherin in lung from wild-type, but not Yes-deficient mice, determined by immunoprecipitation and immunoblotting of mouse lung homogenates. Representative data from three experiments are shown. (B) Treating mice before tumor cell inoculation with VE-cadherin–disrupting antibody BV13 induces vascular permeability, facilitates CT26 tumor cell extravasation, and increases metastases. VE-cadherin antibody E4G10 which does not produce permeability has no impact on metastasis. Tumor burden represents increase in lung/heart weight ratio over control. * indicates P < 0.05; n = 8 each bar (A) and n = 4 each bar (B).

Figure 5.

Pharmacological Src, Flk, or VEGF blockade suppresses tumor cell extravasation. (A) Pretreatment with VEGF inhibitor Cyclo-VEGI provides a similar extent of protection from pulmonary metastases as Src or Flk inhibition, further suggesting the permeability-inducing effects of VEGF contribute to tumor cell extravasation. (B) VEGF receptor 2 signaling is required for metastasis of CT26 cells because a single pretreatment with Flk inhibitor SU1498 reduces lung tumor burden after 12 d. (C) Pretreatment with Src inhibitor provides a dose-dependent blockade of VEGF-induced vascular leak in the skin evaluated using the Miles assay. (D) Src inhibitor treatment administered twice daily during days 0–3 after i.v. introduction of CT26 or D121 cells significantly reduces lung tumor burden after 12 d. (E) Preventing VEGF-induced endothelial barrier breakdown via VEGF, Flk, or Src blockade preserves cadherin-mediated endothelial cell adhesion and limits tumor cell extravasation. Tumor burden represents mean ± SEM for increase in lung/heart weight ratio over control. * indicates P < 0.05; n = 8 each bar (A, B, and D); n = 4 each bar (C).