A new cellular system opposing HIV infection: implications for gene transfer?

One of the longest-standing mysteries in HIV research has been the function of the viral-encoded VIF protein. VIF is required for efficient viral infection in vivo [1–4], apparently because it overcomes the action of a host-encoded antiviral system [5,6]. A recent report in Nature describes a major breakthrough in understanding the cellular defense machinery that is overcome by HIV VIF [7]. In addition to its importance for HIV research, this discovery raises a variety of questions about whether manipulating this system might be helpful in optimizing gene transfer. The vif gene is one of the “auxiliary” genes of HIV, which are defined as being dispensable for efficient replication in at least some settings in cell culture. In vivo, however, vif is clearly required, as indicated, for example, by the finding that simian immunodeficiency viruses (primate relatives of HIVs) lacking vif are greatly attenuated [1]. The outcome of infections with vif-negative viruses in different cell lines was intriguing and ultimately quite informative. In vif-permissive cells, HIV derivatives mutated for vif grow as well as wild-type virus. In vif-nonpermissive cells, growth is greatly abrogated, but in a selective fashion. The nonpermissive phenotype is only seen when vif-negative virus is produced from nonpermissive cells. When vif-negative virus is produced from permissive cells, then used to infect nonpermissive cells, the virus infection proceeds normally. Tests of primary human cells revealed that they were vif nonpermissive, suggesting that VIF acts against this system normally during HIV infection. Elegant experiments from the laboratories of Malim and Kabat established that this could be explained by VIF overcoming a host-encoded antiviral system in virus producer cells. Both labs fused permissive cells with nonpermissive cells and asked whether or not the hybrid cells restricted growth of VIF-mutant HIV. They reasoned that if nonpermissive cells encoded an antiviral system, the fused cells should be nonpermissive because they would contain the interfering factor. The alternative was that nonpermissive cells might lack a factor required for efficient viral replication, in which case the fused cells would be permissive because the factor would be present. The experiments revealed that the fused cells were in fact nonpermissive, supporting the conjecture that restrictive cells express an antiviral system.