Modification of a retroviral envelope protein to achieve cell type specific targeted entry

Retroviruses have been used for gene delivery vectors in human gene therapy. The lack of efficient, entry-targeted vectors has hampered the development of in vivo gene therapy protocols. In this study, I characterized a number of different modifications of the envelope protein (Env) from Moloney murine leukemia virus (MoMLV) in order to develop new methods for achieving cell type specific transduction. Initially, conventional methods of Env modification were employed in which clinically relevant ligands such as IL-13 and erythropoietin were inserted at the amino-terminus or in the proline-rich region (PRR) of Env. Although retroviral vectors containing modified Env bound to cells through the target receptor, they failed to transduce these cells indicating that a post-binding block exists. Fusion enhancing mutations did not overcome the post-binding block. Further analysis using a fluorescence-based cell-cell fusion assay indicated that the site of ligand insertion influences the post-binding block. IL-13 insertion into the amino-terminus blocked an early step in membrane fusion prior to fusion peptide exposure, whereas insertion into the PRR could give fusion peptide exposure but was blocked at a subsequent step prior to fusion pore formation. Thus, a primary reason for failure of these designs is the inability of target receptor binding to activate chimeric Env for membrane fusion. Based on these observations, a new design for the modification of Env was proposed in which the ligand sequences replace the sequences at the natural receptor binding site in Env. This design was based on the hypothesis that full activation of chimeric Env upon target receptor binding will result when the ligand lies in the position of the natural receptor binding site and will give efficient cell type specific transduction. A novel chimeric Env based on this new concept was constructed by replacing residues in the natural receptor binding site of MoMLV Env with the peptide ligand somatostatin (Sst). The prototype chimera transduced approximately half the receptor positive target cells and the transduction was specific to Sst receptors. Taken together, these results establish proof-of-principle that the new concept is a method that gives targeted transduction with a clinically relevant short ligand.