Specific targeting of HIV-1 infected cells by novel recombinant defective interfering HIV-1 particles is not only at the center of our antiviral strategy against HIV-1, but also serves as a model for the targeting of enveloped viruses to selected cell types in general. This FY we succeeded in generating defective HIV-1 particles that can specifically infect cells expressing the HIV-1 Env protein on the cell surface. Virus particles were collected from the supernatant of HeLa T4 cells that constitutively express both, the HIV-1 receptor CD4 and the CXCR4 co-receptor. Pseudotyped HIV-1 (CD4/CXCR4) particles were generated after transfection of the cells with HIV-1 DNA in which the gene for the viral envelope protein ENV was replaced by a neomycin-resistance gene. Virus particles released from the cells were noninfectious since they lacked fusogenic envelope proteins. The same particles which after co- transfection also carried the vesicular stomatitis virus (VSV) G protein, readily infected and stably transduced a variety of different cell types at high efficiency. Using HIV-1 (CD4/CXCR4) particles, we previously tried unsuccessfully to infect several different human and hamster cell lines which constitutively express HIV-1 or HIV-2 Env protein on the cell surface. In contrast, a human lymphoblastoid cell line that expresses HIV-1 Env at high levels could now be infected with the same particles. In fact, the IU/particle ratios were similar to those of HIV-1 itself. The cells were stably transduced and they could be selected for neomycin resistance. This is the first example of a targeted stable transduction using defective HIV-1 particles. We found earlier that HIV-1 particles that carry CD4 unexpectedly gain a new cell adhesion activity. This activity is only observed when CD4 is in the viral envelope and not when CD4 is in the cell membrane. Surprisingly, the binding is as strong as CD4-Env binding. The ligand seems to be ubiquitous and neither cell type nor species specific. Competition experiments using defined ratios of infected and uninfected cells as targets demonstrated that the cell adhesion strongly competes with the CD4-Env binding and the targeting. In fact even at ratios of 1:1 Env(+) to Env(-) cells transduction efficiencies were reduced by fifty percent. When HIV-1 expressing cells are less abundant, this new binding activity does misguide the particles to uninfected cells. HIV-1(CD4/CXCR4) pseudotypes specifically infected HIV-1 infected cells at efficiencies comparable to HIV-1 itself. This clearly demonstrates that after a role reversal of viral and cellular membrane, membrane fusion is still possible. For CD4 pseudotype particles to target HIV-1 infected cells in mixed cell populations, cell adhesion by CD4 virus must be abolished without affecting CD4-Env binding.
