The endogenous soluble VEGF receptor-2 isoform suppresses lymph node metastasis in a mouse immunocompetent mammary cancer model

Abstract

Background: Cancer metastasis contributes significantly to cancer mortality and is facilitated by lymphangiogenesis and angiogenesis. A new splicing variant, endogenous soluble vascular endothelial growth factor receptor-2 (esVEGFR-2) that we recently identified is an endogenous selective inhibitor of lymphangiogenesis. To evaluate the antimetastatic potential of esVEGFR-2, gene therapy with vector expressing esVEGFR-2 (pesVEGFR-2) or endostatin (pEndo) as a positive control was conducted on murine metastatic mammary cancer.

Methods: Syngeneic inoculated metastatic mammary cancers received direct intratumoral injection of pesVEGFR-2, pEndo or pVec as control, once a week for six weeks. In vivo gene electrotransfer was performed on the tumors after each injection.

Results: Deaths from metastasis were much lower in the pesVEGFR-2 and pEndo groups than in those of the pVec. Tumor volume was significantly lower in the pesVEGFR-2 and the pEndo groups throughout the study. Multiplicity of lymph node and lung metastatic nodules was significantly suppressed in the pesVEGFR-2 and pEndo groups. Moreover, the total number of overall metastasis including the other organs was also decreased in these groups. However, pesVEGFR-2 was not able to decrease the number of lungs, ovaries, kidneys and adrenals with metastasis as counted by unilateral or bilateral metastasis. The number of CD34+/Lyve-1⁻ blood microvessels was significantly decreased in the pEndo group, while the number of CD34⁻/Lyve-1+ lymphatic vessels was significantly decreased in the pesVEGFR-2 and pEndo groups. In addition, a significant reduction in the number of dilated lymphatic vessels containing intraluminal cancer cells was observed in the pesVEGFR-2 and pEndo groups. Levels of apoptosis were significantly increased in the pEndo group, whereas the rates of cell proliferation were significantly decreased in the pesVEGFR-2 and pEndo groups.

Conclusions: Our data demonstrate that esVEGFR-2 can inhibit mainly lymph node metastasis. The antimetastatic activity of esVEGFR-2 may be of high clinical significance in the treatment of metastatic breast cancer because lymph node involvement is a most important prognostic factor in cancer patients.

Figure 1

Survival rates, body weights, tumor volumes and bioluminescent imaging in mice receiving esVEGFR-2 gene therapy. Survival rates (A), body weights (B) and tumor volumes (C) in mammary carcinomas from female BALB/c mice transfected with pVec (control), pesVEGFR-2 and pEndo vectors using in vivo gene electrotransfer. Each group consisted of 10 mice. A. Survival rates were much lower in the pVec group than in those of the pesVEGFR-2 and pEndo groups. B. Body weights were similar between pVec group and pesVEGFR-2 or pEndo groups with exception of the pEndo group at Week 2. C. From Week 1 on, rates of tumor growth (tumor volumes) in the pesVEGFR-2 and pEndo groups were significantly decreased as compared to the control values, the differences becoming even more pronounced by the termination of the experiment (Week 6) (**P < 0.01 as compared to the pVec group; †P < 0.01 as compared to the pesVEGFR-2 group). Data presented are means ± SD values. D-F. Bioluminescent imaging of five mice representative of each group. Bioluminescent imaging showed a tendency for decreases in extension of metastasis in the pesVEGFR-2 (E) and pEndo (F) groups as compared to the pVec group (D). G. Quantification of the average bioluminescent signals was significantly decreased in those of the pesVEGFR-2 and pEndo groups (**P < 0.01) as compared to the control values. As compared to the pesVEGFR-2 group, the levels were further decreased in the pEndo group (†P < 0.01).

Figure 2

Histopathological findings in mice receiving esVEGFR-2 gene therapy. The implanted mammary carcinomas proved to be moderately differentiated adenocarcinoma (A-C, scale bar = 50 μm). Histopathologically, no apparent differences in mammary carcinomas were found between the pVec (A) and pesVEGFR-2 (B) groups. However, mammary carcinomas in the pEndo group showed spoke-like cell death regions (asterisks) within viable tumor cells (C). The cell death region was observed around blood vessels (asterisk) and apoptotic bodies (arrows) were seen along with the cell death area (C). p53 immunohistochemistry of mammary carcinoma induced by BJMC3879luc2 cell inoculation (D, scale bar = 25 μm). Note nuclear staining for abnormal p53 protein, indicating that these cells carry mutant p53. Metastasis to lymph node in the pVec (E), pesVEGFR-2 (F) and pEndo (G) (E-G, scale bar = 50 μm). Metastatic carcinoma cells were presented in medullary cord and intraluminal space of the blood vessel (E, arrows). Metastatic carcinoma cells were filled with subcapsular sinus (asterisk) and cortical sinus (arrow) (F). Fewer lymph nodes with metastasis were found in the therapeutic groups. Metastatic tumor cells were observed in subcapsular sinus (G, arrow). Metastatic foci in the lung of the pVec, pesVEGFR-2 and pEndo groups (H-J, scale bar = 200 μm). Many metastatic foci and nodules with small to large were seen in the pVec group (H). Metastatic lung foci were much smaller in the pesVEGFR-2 (I) group than in the control group (J). No metastatic foci were observed in the lung of mouse given pEndo (J). A-C and E-J, H&E stain; D, p53 immunohistochemistry.

Figure 3

Quantitative analyses of metastasis, vascular density, apoptosis, cell proliferation in mammary carcinomas. A. Multiplicity of lymph node metastasis was significantly decreased in the pesVEGFR-2 and pEndo groups (**P < 0.01 as compared to the pVec group; †P < 0.01 as compared to the pesVEGFR-2 group). B. Multiplicity of lung metastatic foci > 250 μm was significantly reduced in the pesVEGFR-2 and pEndo groups (**P < 0.01 as compared to the pVec group; †P < 0.01 as compared to the pesVEGFR-2 group). C. Quantitation of blood and lymphatic microvessels were conducted using CD34 and Lyve-1 immunohistochemistry. The number of CD34⁺/Lyve-1^- blood vessels was significantly decreased in the pEndo group, but not in the pesVEGFR-2 group, as compared to the pVec group. The number of CD34^-/Lyve-1⁺ lymphatic microvessels was significantly decreased in the pesVEGFR-2 and pEndo groups. D. The mammary tumors were immunohistochemically stained for another lymphatic endothelial marker podoplanin. The number of dilated lymphatic microvessels containing intraluminal tumor cells was significantly lower in pesVEGFR-2 and pEndo groups than those observed in the control pVec group. E. Apoptotic cell death, assessed by TUNEL assay, was significantly increased in the pEndo group. F. Cell proliferation, inferred by BrdU labeling indices, was significantly decreased in the pesVEGFR-2 and pEndo groups. *P < 0.05 and **P < 0.01 as compared to the values of the pVec group. Data presented are means ± SD values.

Figure 4

Immunofluorescence for blood and lymphatic vessels in mammary carcinomas. Double immunohistochemical staining with CD34 for blood microvessels (red) and Lyve-1 for lymphatic (green) microvessels and their merged images in mammary tumors. As can be seen in their merged image (C), some microvessels showed both expressions of CD34 (A) and Lyve-1 (B). Such microvessels having CD34⁺/Lyve-1⁺ characteristics were excluded from quantitation. The number of lymphatic microvessels was lower in the pesVEGFR-2 (G-I, merge in I) and pEndo (J-L, merge in L) groups than in the pVec group (D-F, merge in F). A-L, scale bar = 50 μm.

Figure 5

Dilated lymphatic vessels with cancer cell invasion, apoptosis, and cell proliferation in mammary carcinomas. Podoplanin-positive lymphatic microvessels of a tumor in a control mouse were often dilated and filled with tumor cells (A). The pesVEGFR-2 (B) and pEndo (C) groups showed a reduction in the numbers of dilated lymphatic microvessels containing intralumenal tumor cells (A-C, scale bar = 50 μm). Whereas some TUNEL-positive cells are seen in the tumor of a control mouse (D) and a mouse treated with pesVEGFR-2 (E), many more TUNEL-positive cells are observed in the tumor of a mouse treated with pEndo (F) (D-F, scale bar = 50 μm). Higher magnification in F showed many TUNEL-positive cells are observed in along with brownish necrotic vessels (G, arrows) (scale bar = 25 μm). Apoptosis of the tumor cells may be due to injury of the blood vessels (blockage of oxygen supply and nutrition to tumor cells). The number of BrdU-labeled cells tended to be lower in the pesVEGFR-2 (I) and pEndo (J) groups than in the pVec group (H) (H-J, scale bar = 50 μm). A-C, podoplanin immunohistochemistry; D-G, TUNEL stain; H-J, BrdU immunohistochemistry.