Retroviral targeting of tumor endothelium

Expanding tumor masses require an adequate supply of nutrients and oxygen which is provided by a newly established vasculature. The inner lining of blood vessels consists of so called endothelial cells, which undergo characteristic changes upon vessel formation such as the presentation of specific cell surface markers which are absent from the established vasculature or other cell types. In this respect, the tumor endothelium represents an attractive target to address therapeutic agents to tumor tissue. We have chosen an approach involving retroviruses, which are generally capable of infecting target cells and introducing a gene of interest (gene transfer). Thus, a "therapeutic gene" could be transferred to the tumor vasculature; the anti-cancer drug would be locally produced and secreted, thereby reaching nearby tumor cells. Applying retroviral vectors which could specifically bind to surface markers characteristic of tumor endothelium would thus offer the possibility to limit the treatment to expanding tumor tissues while keeping adverse effects (to the whole body system) at a minimum. Furthermore, these viruses would we able to trace and infect large tumor masses as well as smaller, dispersed foci of growing tumor cells i.e. would potentially offer the possibility for treatment of metastatic disease. The intrinsic host range of naturally occurring retroviruses is, however, rather broad, and is crucially determined by the viral envelope protein mediating interaction with host cell surface proteins. The aim of our study was directed towards modifying the envelope protein by introducing so called endothelial cell ligands (angiogenin, angiopoietin- 2) which would confer specific interaction with surface markers of tumor endothelium. Five different fusion modalities were designed (to either add ligand sequences to or substitute for parts of the envelope protein) since form and functionality of the resulting protein are difficult to predict and need to be varied to possibly obtain a functional fusion variant. We have thus constructed a total of 10 fusion variants. The resulting proteins were analyzed and characterized, and subsequently applied in the production of modified retroviral particles. Unfortunately, these retroviruses did not specifically infect human endothelial cells, i.e. no gene transfer could be observed. It thus seems likely that the applied fusion modalities did not result in a functional envelope protein mediating cell entry and they need to be further improved. Considering, however, the therapeutic potential discussed above, further efforts to generate retroviral vectors successfully targeted to tumor endothelium seem warranted.