HIV-1 replicates itself by integrating into its host cell's DNA. Studies in cell culture reveal that nuclear-membrane proteins aid engagement of the viral DNA with that of its host before integration.
Rather than going to the trouble of replicating its own genome, human immunodeficiency virus type 1 (HIV-1) inserts a DNA copy of its genome into that of its host so that the virus is reproduced during host-cell division (Fig. 1). The virus genome consists of RNA, so a DNA copy must first be made from the RNA template. The newly synthesized viral DNA forms part of a large complex in the cyto-plasm of an infected cell, the pre-integration complex (PIC). This contains the viral integrase enzyme that splices the viral DNA into cellular DNA, as well as other viral and cellular proteins. To reach the cellular DNA, the PIC must cross the nuclear envelope that separates the cytoplasm and nucleoplasm. On page 641 of this issue, Jacque and Stevenson1 show that HIV-1 PICs target a specific nuclear-membrane protein called emerin, and that this is required for the viral DNA to engage with the host DNA before integration.
Unlike many viruses, HIV-1 and other lentiviruses can infect cells that are not dividing, so the viral DNA must cross the intact nuclear envelope to gain access to the cellular chromatin (the DNA and associated proteins that package it into chromosomes). This presents a formidable challenge, because the PIC is huge on a molecular scale — far larger than the diameter of nuclear pores. Thus, nuclear entry may be facilitated by first jettisoning some components of the PIC. Many protein components of the PIC carry signalling sequences that enable them to enter the nucleus, but the roles of these proteins and the mechanism by which the PIC is imported into the nucleus remain controversial. Many cellular proteins have been reported to associate with PICs2, but in most cases the functional consequences are unclear.
The nuclear envelope is composed of the inner and outer nuclear membranes, embedded pore complexes, and the lamina and associated proteins3. The nuclear lamina, which lies just inside the inner nuclear membrane, is a network of lamin proteins that, as part of the nuclear ‘matrix’, provides structural support for the nucleus. Proteins associated with the lamina include the lamin B receptor (LBR), lamin-associated polypeptides 1 and 2 (LAP1 and LAP2), emerin and MAN1. LAP2, emerin and MAN1 share a conserved structural domain (called the LEM domain) that binds to a protein known as the ‘barrier-to-auto-integration’ factor (BAF). BAF was first identified as a factor that blocks the self- destructive integration of murine leukaemia virus (MLV) DNA into its own genome in vitro4. This factor is a nonspecific DNA-binding protein that can bridge strands of duplex DNA and may link chromatin to LEM-domain proteins at the nuclear envelope5. BAF is required for progression through the cell-division cycle6, and lack of BAF causes dividing cells to die7.
Jacque and Stevenson1 examined the ability of HIV-1 to infect cells in which individual nuclear-envelope-associated proteins had been selectively depleted using a method called RNA interference (RNAi). This technique selectively targets messenger RNAs for degradation, stopping production of the encoded protein. The authors shrewdly studied the roles of the nuclear-envelope proteins in primary macrophages, one of the immune cell types infected by HIV. These cells do not divide, so this strategy bypassed the lethal effects of loss of BAF, and sidestepped potential cell-division roles of the other nuclear-envelope proteins tested.
Depletion of either emerin or BAF severely impaired HIV-1 replication in infected macrophages. Notably, expression of these proteins using RNAi-resistant mRNAs restored replication, supporting the conclusion that the defect was the direct consequence of depleting each protein. The emerin-depleted cells were fully susceptible to infection with a different retrovirus (MLV), further indicating that the replication defect is specific to HIV-1 and not due to a widespread disruption of the normal cell physiology. Replication of MLV required BAF and LAP2a, but not emerin1,8. So it would seem that, although both retroviruses recruit BAF, each virus also enlists different LEM-domain proteins.
What is the viral replication defect in cells depleted of either emerin or BAF? In these cells, viral DNA was synthesized at normal levels but failed to integrate into the cellular genome. Biochemical subcellular fractionation experiments indicated that depletion of emerin or BAF did not prevent the viral DNA from entering the nucleus, but that the viral DNA became associated with different nuclear fractions from normal. In control macrophages, most of the viral DNA was associated with the soluble chromatin fraction, whereas in cells depleted of BAF or emerin it was mainly in the insoluble nuclear-matrix fraction. The identical infectivity defects caused by depletion of either emerin or BAF suggest that these proteins have a cooperative role in HIV-1 infection, consistent with the known interaction between BAF and the LEM domain of emerin.
The mechanism by which emerin and BAF facilitate the proper nuclear localization of the HIV-1 PICs remains unknown. The association of emerin with the PIC depends on BAF, which probably interacts with the viral DNA. But how does emerin then influence the association of the HIV-1 PIC with the host chromatin? It may be that the work required to answer this question will also uncover other features of nuclear architecture. Although the organization of chromatin has been extensively studied at the level of the nucleosome (the smallest unit of DNA packaging), the global organization of chromatin within the nucleus is not well understood. However, the nucleus is clearly both highly compartmentalized and dynamic, and chromatin is intimately associated with the nuclear envelope. Perhaps we should not be surprised that there is more to accessing chromatin than simply crossing the nuclear envelope.