Cellular restriction factors and mechanisms of intracellular resistance to HIV-1

I. The role of APOBEC3 family of cytidine deaminases in virus restriction.
APOBEC3G: a powerful restriction factor counteracted by Vif. Mammalian cells resist viruses through a collection of mechanisms termed innate or intrinsic immunity. These mechanisms differ from the classical adaptive immune response in which specialized B, T helper and cytolytic T cells recognize foreign antigens and clonally expand upon engagement of their antigen receptor. Intrinsic and innate mechanisms are more generalized. They are present in many different cell types and some are constitutively active in the cell. In some cases, innate immune mechanisms are activated by signals from cytokines such as interferons or toll-like receptors that warn the cell of the presence of a foreign invader. Viruses have responded over the course of evolution by developing remarkable and diverse ways to escape the adaptive and the innate response. A major focus of our research is to understand how innate immune mechanisms restrict retroviruses.

APOBEC3, which in the human includes APOBEC3A, B, C, D/E, F, G and H, is a family of antiviral cytidine deaminases. These enzymes act as DNA mutators that target the genome of viruses by deaminating cytosine nucleotides to uracil. APOBEC3G (and to a lesser extent, APOBEC3F) is of particular importance because it specifically restricts HIV-1. APOBEC3G is expressed in the CD4+ T cells and monocytes that are the target of the virus. HIV-1 is able to productively replicate in such cells by virtue of the Vif accessory protein, which binds to APOBEC3G and causes its degradation. Infection of a cell with Δvif HIV-1 that cannot encode a functional Vif, results in an abortive infection. Virions are produced by the Δvif HIV-1 infected cell but these are noninfectious. The virions contain packaged APOBEC3G molecules that attack the viral genome as it is copied from RNA into DNA in the next round of replication. In a sort of evolutionary biological warfare, the virus developed Vif to bind APOBEC3G before it can be packaged and then shunt it to the proteasome for degradation. Vif contains a zinc binding motif and a conserved SOCS box that interacts with a Cul5-based E3 ubiquitin ligase that mediates the ubiquitination of APOBEC3G which flags it for proteasomal degradation. We are addressing several aspects of how this system operates. These include defining the interaction domains of Vif, APOBEC3 and the E3 ubiquitin ligase; developing small molecule inhibitors of the interaction; and understanding how the APOBEC3 genes are transcrptionally regulated.

Figure 1. Vif, Vpr/Vpx and Vpu counteract cellular restriction factors. Vif and Vpr/Vpx and Vpu associate with E3 Ubiquitin ligases to induce the proteasomal degradation of restriction factors (the E3 for Vpu is not shown). Vpr/Vpx overcome a post-entry block to reverse transcription by targeting a yet unidentified restriction factor indicated by question marks. Vif induces the degradation of APOBEC3G to prevent its packaging into the virion and the subsequent C→U deamination of the reverse transcribed viral DNA. Vpu antagonizes tetherin and CAML (not shown) which hold the virus onto the cell surface.

Physiological roles of APOBEC3 in the mouse. In contrast to the human, the mouse genome contains only one APOBEC3 gene. Although mice are not subjected to lentivirus infection, the mouse APOBEC3 protein (mo-APOBEC3) is a potent inhibitor of HIV-1 when introduced into human cells in culture. Interestingly, mo-APOBEC3 is not recognized by HIV-1 Vif and as a result, the murine enzyme restricts the replication of wild-type HIV-1. The presence of only a single gene in the mouse is convenient because APOBEC3 knock-out mice are then completely deficient. We have generated such mice and, consistent with other groups, that there is no obvious phenotype. As shown by others, APOBEC3 knock-out mice are more susceptible to murine retroviruses such as mouse mammary tumor virus and Friend murine leukemia virus. We are currently using the knock-out mouse to study the mechanisms by which APOBEC3 protects against these viruses. Interestingly, it does not appear to be due to deamination of their genomes.

APOBEC3A: potent inhibitor of retrotransposons and parvovirus. The roles of the various APOBEC3 family members are not yet known but an attractive hypothesis is that they act against different viruses. In support of this possibility, we have studied APOBEC3A. APOBEC3A has no activity against HIV-1 or SIV, but we found that it is a potent inhibitor of retrotransposons. Retrotransposons are a class of endogenous genetic elements that are present in thousands of copies in mammalian genomes. They resemble retroviruses in genome structure and some have the ability to transpose, or jump in the genome, an event that occasionally cause diseases such as muscular dystrophy or hemophilia. Our collaborators at the Salk Institute have found that APOBEC3A also inhibits adeno-associated virus, a small DNA virus that replicates in the nucleus through a single strand DNA intermediate. APOBEC3 proteins are active against single-stranded DNA and APOBEC3A localizes to the nucleus, allowing it to be in the vicinity of reverse transcribing retroelements and replicating parvovirus. APOBEC3A is strongly induced by type-1 interferon, further supporting its role as an antiviral mechanism. We have produced pure, catalytically active recombinant APOBEC3A in milligram quantities and found that the enzyme acts nonprocessively and strongly binds to single-stranded DNA. We are interested to understand how it works, both in vitro and in vivo. We have generated a panel of point mutants and are testing their activity in various assays including retrotransposition, cytidine deaminase activity and DNA binding. In addition, we have generated APOBEC3A transgenic mice to determine whether expression of the protein has an effect on the mouse, preventing the jumping of endogenous retroelements or causing mutations in the mouse genome.

II. The search for novel viral restriction factors
We are also trying to uncover new mechanisms of innate immunity. APOBEC3 and TRIM5α are the first identified lentiviral restriction factors but there is reason to believe that there are others that are not yet identified. We are using gene transfer and microarray approaches to identify these factors.

III. Vpr/Vpx: viral accessory proteins of unknown function
HIV-1 encodes another accessory protein, Viral protein R (Vpr), that may play a role in neutralizing an innate resistance mechanism. Vpr is conserved in HIV-1 isolates but its role in HIV-1 replication and pathogenesis is not well understood. It is not required for the virus to replicate on cultured cells, but its importance for virus replication and disease pathogenesis in vivo has been demonstrated in the rhesus macaque model. We have that found, Vpr, like Vif, is associated with an E3 ubiquitin ligase. These multi-subunit complexes are molecular machines that target specific host proteins for degradation. Vpr is the only viral regulatory protein that is present in HIV-1 particles. It is specifically targeted to the virus by binding to the P6 region of the Gag polyprotein as the virus assembles. The presence of the protein in the virion suggests that it acts early post-entry. Its association with the E3 ubiquitin ligase suggests that it may act in the target cell to stave off attack from a yet unidentified host antiviral protein. We are trying to identify the antiviral protein and to understand how it blocks virus replication. Understanding this mechanism may uncover yet another target for antiretroviral drug development. Interestingly, SIV encodes Vpx, a protein derived by duplication of the Vpr open reading frame. Vpx is required for efficient infection of macrophages. If Vpx protein is introduced into macrophages prior to infection by treatment of the cells with virus-like particles, it relieves the block to infection of Δvpx SIV. There is reason to believe that Vpx, which also interacts with the E3 ubiquitin ligase, neutralizes an antiviral protein. Whether that protein is the same or related to the target of Vpr, is a current focus of our studies.

IV. Nuclear import of the HIV-1 preintegration complex
One of the longstanding questions in HIV research is how the virus penetrates the cell nucleus to access the cellular genomic DNA where integration occurs. Lentiviruses such as HIV-1 differ from the simpler retroviruses that wait in the cytoplasm for the nucleus to breakdown during mitosis before they can get to the cellular DNA. In contrast, lentiviruses are able to infect nondividing cells. This allows them to infect macrophages that are terminally differentiated cells that do not divide. Recent genome-wide siRNA knock-down screens from three groups turned up an interesting clue as to how lentiviruses infect nondividing cells. These screens showed that successful infection of a cell by HIV-1 requires components of the nuclear pore and nuclear import proteins. We have confirmed the finding that transportin-3 (TNPO3), a nuclear shuttle protein that serves to import RNA splicing proteins into the nucleus, is required for HIV-1 nuclear import. We have found that SIV also uses this pathway. We are currently addressing the question of how TNPO3 and nuclear pore proteins mediate entry of the viral preintegration complex. We are determining which component of the preintegration complex interacts with TNPO3 and are testing a panel of TNPO3 point mutants to dissect the functional domains of the protein.