Author Response: Bevacizumab Suppression of Establishment of Micrometastases in Experimental Ocular Melanoma

Abstract

Bevacizumab is a recombinant humanized monoclonal IgG1 antibody that binds to and inhibits the biological activity of human vascular endothelial growth factor (VEGF) in in vitro and in vivo assay systems. It contains human framework regions and the complementarity-determining regions of a murine antibody that binds to VEGF. When comparing the sequence of human VEGF protein with that of mouse VEGF protein, we found that 86% of the protein sequence in mouse VEGF164 is similar to that of human VEGF165.1 We believe that it is possible for bevacizumab to block human and mouse VEGF. A blocking antibody is defined as an antibody that does not have a reaction when combined with an antigen, but prevents other antibodies from combining with that antigen.2,3 We believe that bevacizumab fits this definition and explains why the neutralized VEGF is not detected by ELISA. The human or mouse VEGF immunoassay (Quantikine; R&D Systems, Minneapolis, MN) that we used in our experiment can detect human VEGF165 and VEGF121 or mouse VEGF164 and VEGF120, but does not detect VEGF189 and VEGF206. VEGF121 and VEGF165 are diffusible proteins that are secreted into the medium. VEGF189 and VEGF206 have high affinity for heparin and are mostly bound to heparin-containing proteoglycans in the extracellular matrix. Bevacizumab binds to and neutralizes all isoforms of VEGF-A and bioactive proteolytic fragments. Bock et al.4 performed three different experimental approaches using Western blot analysis, ELISA, and Biacore (surface plasmon resonance; GE Health Care, Piscataway, NJ) analysis to assess whether bevacizumab binds to murine VEGF-A. In the Western blot analyses, recombinant murine VEGF-A164, a 24-kDa protein consisting of 164 amino acid residues, and the 19.1-kDa homologous human VEGF165, which consists of 165 amino acid residues, were used. In ELISA and the Biacore analyses, the recombinant human and the homologous murine VEGF-A were used as well. Western blot analysis showed that bevacizumab bound murine and human VEGF-A. To confirm these results, Bock et al. performed two different experimental approaches for analysis of the binding of bevacizumab to murine VEGF-A. The first assay, an ELISA, analyzed the interaction of soluble bevacizumab with VEGF-A immobilized on a microtiter plate surface. In the second assay, the surface plasmon resonance technology (Biacore) was used for the analysis of interaction of soluble VEGF-A with bevacizumab immobilized on the sensor chip surface. Both assays allow a qualitative characterization of the interaction between bevacizumab and human or murine VEGF-A. In the ELISA binding assay, bevacizumab bound to both human and murine VEGF-A; however, a 1000-fold higher bevacizumab concentration was necessary to reach a similar level of binding to murine VEGF-A compared with that needed for binding to the human protein. There was no significant binding of isotype control IgG1k to murine VEGF-A at this concentration. During the Biacore surface plasmon resonance experiments, different concentrations of human and murine VEGF-A were injected, to analyze their binding to immobilized bevacizumab. A more than 10-fold higher concentration of murine VEGF-A than of human VEGF-A was necessary to reach similar resonance units. ELISA is more sensitive than Western blot analysis. It can detect a protein of at picogram concentrations; Western blot can only detect a protein at nanogram concentrations. Studies have also indicated that bevacizumab inhibits experimental choroidal and corneal neovascularization in a mouse model,5,6 endothelial cell proliferation in a retinopathy of prematurity mouse model,7 and tumor progression in a A/J Tsc2+/− C57BL/6 background mouse model of pancreatic tumor.8,9 When we initially used bevacizumab in a mouse model to treat mouse melanoma cells, we found that it slightly decreased the tumor size and number of metastases. We then performed bevacizumab experiments with human uveal melanoma cell lines. Our experiments showed that bevacizumab had much more effect on human melanoma than on mouse melanoma. We acknowledge point three of Sharma et al. and reply that we meant that 10 and 100 μg/mL of bevacizumab reduced VEGF levels by 8.51 and 15.99 pg/mL, respectively. Bevacizumab acts as a tumor suppressor by inhibiting angiogenesis and lymphangiogenesis. Lymphangiogenesis not only plays an important role in mediating immune reactions, but also facilitates tumor metastasis10. In our study, we investigated only the antiangiogensis function of bevacizumab.