HIV encodes a protein that forms a conical shell, called a capsid, that surrounds its genome. The capsid has been shown to protect the viral genome from innate immune sensors in the cell, to help transport the genome towards and into the nucleus, to keep the components of reverse transcription together for conversion of the RNA genome into DNA, and to target viral DNA integration into specific regions of the host genome.
In this study, we show that HIV hijacks two host proteins to bind to capsid sequentially to choreograph the precise order and timing of these virus replication steps.
We used a virus with mutations in the CA protein, called AC-1, that binds with higher affinity to the host protein, cyclophilin A (CypA) during infection. We demonstrated this by infecting Owl monkey cells, which express a fusion of TRIM5a with CypA that is restrictive to infection for viruses that can bind to the CypA domain. With increasing amounts of wild-type (WT) or CA mutant viruses, we see that WT HIV-1 can overcome TRIMCyp restriction, but AC-1 cannot unless we use a drug (CsA) or make an additional mutation in CA (P90A) that prevent CypA binding to capsid:
Zac found that increased CypA binding inhibits HIV-1 infection in a manner that is dependent on another host protein, CPSF6. Infectivity of AC-1 virus is rescued when CsA is added to prevent CypA binding or when an additional CA mutation, N74D, is added. N74D prevents CPSF6 from binding to capsid.
The restriction of HIV-1AC-1 was at the step of nuclear import of the viral genome.
Increased CypA binding to HIV-1 capsid (AC-1) led to reduced CPSF6 binding to capsid. When CsA was used to disrupt binding of CypA, Zac saw more CPSF6-GFP binding during HIV-1 infection.
Using CypA knockout cells, Zac added back either normal CypA, which is most expressed in the cell cytoplasm, or CypA with a nuclear localization signal so that it is exclusively expressed in the nucleus. The infection phenotype of AC-1 with and without CsA treatment was only seen when CypA was in the cytoplasm, suggesting that cytoplasmic but not nuclear CypA regulates HIV-1 nuclear import and infection.
HIV-1AC-1 capsids that enter the nucleus, perhaps during mitosis, appeared not to engage nuclear CPSF6, leading to misintegration near the nuclear membrane in lamin-associated domains. With a loss of cytoplasmic CypA binding (either with CsA treatment or a mutation in capsid), HIV-1AC-1 integrated preferentially into gene-dense regions and speckle-associated domains, similar to WT HIV-1.
Our model of optimal HIV-1 replication suggests that CypA binds to the incoming capsid at the periphery of the cell and the availability of CypA decreases near the nucleus, allowing CPSF6 to engage with the capsid, leading to a competitive exchange of capsid binding host factors that is spatiotemporally regulated.
Recent Comments