Study of an interleukin-2 immunoconjugate as a novel tumor vasoactive agent in animal models

Abstract

The elevated levels of the HER2/neu protein in 60% of patients with ductal carcinoma in situ and in 30% of patients with invasive breast cancer make this oncogene product an attractive target for tumor-specific therapeutic agents. Despite some successes to date, vascular inaccessibility and tumor heterogeneity are two main factors that hindered the efficacy of anti-HER-2 antibody therapy. There is ample evidence suggesting that recombinant interleukin-2 (rIL-2) has profound antitumor properties, but was limited by its severe systemic toxicity known as vascular leakage syndrome (VLS). A fusion protein H520C9sFv-rhIL-2 has been constructed by combining the humanized VH and VLportions of a murine monoclonal antibody 520C9 specific for human HER2 with human rIL-2 to deliver effective concentration of IL-2 to HER2 positive tumors whilst avoiding toxic effects on normal tissues. In this study in vitro characteristics of the H520C9sFv-rhIL-2 were investigated. Western blot of partially purified conditioned medium from 293 cells transfected with the cDNA of the H520C9sFv-rhIL-2 revealed one major protein band with a molecular weight of 45 kDa, indicating that the fusion protein was expressed in the transfected 293 cells. The H520C9sFv-rhIL-2 was shown to preserve both immuno-stimulatory activities of IL-2 as measured by an IL-2 dependent CTLL2 cell proliferation assay and antigen binding specificity against HER2 positive SKOV3 and B16/neu cells. These in vitro results have stimulated interest in the in viva application of the H520C9sFv-rhIL-2. Present results demonstrated that a single intravenous dose of H520C9sFv-rhIL-2 to mice bearing HER2 positive tumors could preferentially increase the permeability of blood vessels in the tumors in a time- and dose-dependent manner. Using C57/BL mice bearing B16/neu s.c. tumors as a model, 18 ug of H520C9sFv-rhIL-2 and 24 h post injection of a vascular permeability tracer resulted in maximal tumor: nontumor uptake ratios of the tracer. Compared to a saline injection control group, the tumor: nontumor uptake ratios in bone, blood, kidney, lung, muscle and spleen ranged from 2.1 to 12.2. Nine ug of H520C9sFv-rhIL-2 only resulted in 4.5-fold and 2.3-fold increase respectively for tumor: bone and tumor: muscle uptake ratios when compared to the saline group 24 h post tracer injection. At 12-h and 72-h post tracer injection, no significant difference in tumor: nontumor uptake ratios was observed between the 9 ug H520C9sFv-rhIL-2 group and the saline group. Some fatality was observed at 18 ug dose of H520C9sFv-rhIL-2. Using nude mice carrying SKOV3 s.c. tumors as another model, a single intravenous dose of 9 ug H520C9sFv-rhIL-2 could not preferentially increase the permeability of blood vessels in the tumors 24 h post injection of a vascular permeability tracer, although a single intra-tumor dose of 9 ug of H520C9sFv-rhIL-2 could significantly increase tumor: nontumor uptake ratios for bone, kidney and spleen relative to a saline injection group. Although the mechanisms underlying IL-2 mediated VLS have not been investigated in detail in this study, present results suggest that at least one of the mechanisms depends on T cells. Another mechanism does not depend on T cells because vascular leak could be induced in the athymic nude mice lacking functional T cells by intra-tumoral injection of the fusion protein. Taken together, these encouraging in vitro and in vivo results of the H520C9sFv-rhIL-2 suggest that this fusion protein might provide a good approach to selectively increasing the permeability of tumor vasculature. It is hoped that this will translate into further investigation into its clinical application.