Active immunization against the vascular endothelial growth factor receptor flk1 inhibits tumor angiogenesis and metastasis

Abstract

The vascular endothelial growth factor (VEGF) receptor fetal liver kinase 1 (flk1; VEGFR-2, KDR) is an endothelial cell-specific receptor tyrosine kinase that mediates physiological and pathological angiogenesis. We hypothesized that an active immunotherapy approach targeting flk1 may inhibit tumor angiogenesis and metastasis. To test this hypothesis, we first evaluated whether immune responses to flk1 could be elicited in mice by immunization with dendritic cells pulsed with a soluble flk1 protein (DC-flk1). This immunization generated flk1-specific neutralizing antibody and CD8+ cytotoxic T cell responses, breaking tolerance to self-flk1 antigen. Tumor-induced angiogenesis was suppressed in immunized mice as measured in an alginate bead assay. Development of pulmonary metastases was strongly inhibited in DC-flk1-immunized mice challenged with B16 melanoma or Lewis lung carcinoma cells. DC-flk1 immunization also significantly prolonged the survival of mice challenged with Lewis lung tumors. Thus, an active immunization strategy that targets an angiogenesis-related antigen on endothelium can inhibit angiogenesis and may be a useful approach for treating angiogenesis-related diseases.

Figures

Figure 1.

Flk1-specific neutralizing antibody induced by vaccination with DCs pulsed with soluble flk1 protein. (A) Mice were immunized three times with flk1-AP-pulsed DCs (DC-flk1), AP-pulsed DCs (DC-AP), or PBS. Postimmunization sera (1:100 dilution) were analyzed for anti-flk1 antibody using ELISA in plates coated with flk1-His protein. Mice immunized with DC-flk1 exhibited significantly higher levels of anti-flk1 antibody compared with control groups. (B) Serially diluted sera from immunized mice were coincubated with soluble flk1-AP protein and then tested for binding to VEGF by ELISA. Results indicated that sera from DC-flk1–immunized mice blocked binding of VEGF to soluble flk1 receptor. (C) Serially diluted sera from immunized mice were incubated with flk1-expressing bEND.3 cells at 4°C for 4 h in a total volume of 0.4 ml. In parallel, the cells were incubated with a decreasing amount of unlabeled VEGF (serially diluted from a 400 ng/ml solution). After incubation, cells were washed and incubated with 10 nCi [125I]VEGF for 1 h at room temperature. The cells were washed and counted in a γ counter. Results suggested that sera from DC-flk1–immunized mice blocked binding of VEGF to flk1 expressed at the surface of endothelial cells.

Figure 2.

Flk1-specific CTL responses induced by vaccination with DC-flk1. (A) Mice were immunized three times with DC-flk1, DC-AP, or PBS. Splenocytes were harvested and restimulated in vitro with flk1-AP–pulsed DCs for 5 d. CTL activity against flk1+ endothelial cells was assessed in a standard 51Cr release assay using H5V endothelial cell line as target. Higher CTL activity was detected in DC-flk1–immunized mice compared with DC-AP–immunized or PBS control mice. (B) T cells cultured from DC-flk1–immunized mice were further analyzed for their antigen specificity in the 51Cr release assay using the following targets: flk1-AP-pulsed DCs, AP-pulsed DCs, flk1-negative tumor cell lines Lewis lung (D122–96) and EL4, and YAC-1. Results indicated that CTLs induced in DC-flk1–immunized mice were specific to flk1 antigen and showed no cytotoxicity against flk1-negative tumor cell lines.

Figure 3.

DC-flk1 vaccination inhibits tumor-induced angiogenesis. Mice were immunized three times with DC-flk1, or DC-AP. Alginate beads containing 5 × 104 Lewis lung tumor cells were then implanted subcutaneously into the mice. 12 d later, mice were injected intravenously with FITC-dextran. Beads were then surgically removed and FITC-dextran quantitated. Mean ± SE of n = 12 mice/group. In a negative control, beads containing no tumor cells (blank) were implanted in naive mice. On the right panel, pictures of representative samples of beads from each group were shown.

Figure 4.

Lewis lung tumor metastasis can be inhibited by DC-flk1 vaccination. (A) Mice were immunized three times with DC-flk1, DC-AP, or PBS and then challenge with Lewis lung tumor cells intrafootpad. The primary tumor was surgically removed when it reaches ∼5 mm in diameter. Mice were killed based on the metastatic death in the control groups, and lungs were weighed and assessed for tumor load. P < 0.01 by Student's t test. (B) Pictures of representative lung samples from each group. (C) Histological examination (H&E staining) of lung samples from each group. Original magnification: ×100.

Figure 5.

Survival advantage in mice immunized with DC-flk1 after a tumor challenge. Animals (n = 10) were immunized with DC-flk1, DC-AP, or PBS. The experiment is similar to that described Fig. 3 A, except that the animals were not killed but were monitored for survival. The data is expressed as percentage of survival as function of time. Survival for the mice immunized with DC-flk1 was significantly prolonged over the DC-AP and PBS controls.

Figure 6.

Inhibition of B16 tumor metastasis by DC-flk1 vaccination. (A) Mice were immunized with DC-flk1, DC-AP, or PBS, before challenged with intravenous injection of B16 tumor cells. Mice were killed 30 d after tumor inoculation and the lungs were assessed for tumor load. P < 0.01 by Student's t test. (B) Pictures of representative lung samples from each group. (C) Histological esamination of lung samples from each group. Original magnification: ×100.

Figure 7.

CD8+ CTL play a major role in antitumor response induced by DC-flk1. Mice were treated with anti-CD4, CD8, or control Abs before they were immunized DC-flk1. In an additional control group, mice were not immunized. Mice were then challenged with Lewis lung tumor intrafootpad. The primary tumor was surgically removed when it reached ∼5 mm in diameter. Mice were killed based on the metastatic death in the control groups. P < 0.01 by Student's t test.

Figure 8.

Cutaneous wound healing is not affected by DC-flk1 vaccination. A full-thickness excisional wound was created in DC-flk1 and untreated mice. Wound areas were measured twice weekly until wounds were completely healed (A). Wound scar tissues at day 15 from both groups were analyzed by histology (B). Top: hematoxylin and eosin staining; bottom: trichrome staining for collagen. Original magnification: ×100. No difference was found between these tissues.

Figure 8.

Cutaneous wound healing is not affected by DC-flk1 vaccination. A full-thickness excisional wound was created in DC-flk1 and untreated mice. Wound areas were measured twice weekly until wounds were completely healed (A). Wound scar tissues at day 15 from both groups were analyzed by histology (B). Top: hematoxylin and eosin staining; bottom: trichrome staining for collagen. Original magnification: ×100. No difference was found between these tissues.