Aaron Diamond AIDS Research Center and Laboratory of Retrovirology, The Rockefeller University, New York, New York, USA.
Address correspondence to: Paul D. Bieniasz, Aaron Diamond AIDS Research Center, 455 First Avenue, New York, New York 10016, USA. Phone: (212) 448-5070; Fax: (212) 725-1126; E-mail: firstname.lastname@example.org.
First published February 1, 2007 - More info
See the related article beginning on page 473.
HSCs are one of only a few cell types that resist HIV-1 infection despite the presence of HIV-1 receptors. An increasing number of genes have been identified that can reduce the sensitivity of cultured cells to retrovirus infection, and in this issue of the JCI, Zhang et al. identify p21Waf1/Cip1/Sdi1 (p21) as a gene product that can influence the sensitivity of HSCs to HIV-1 infection (see the related article beginning on page 473). Strikingly, p21 appears to alter the fate of nuclear HIV-1 DNA, promoting the formation of circular viral DNA forms rather than functional proviruses.
For many years, the ability of a particular retrovirus to colonize a given target cell type or species was thought to be governed solely by its ability to exploit required cellular cofactors provided to it by a candidate target cell. HIV-1, for example, can only infect cells that express CD4 and a chemokine receptor because those molecules are required to mediate the fusion of virion and target cell membranes. Similarly, HIV cannot replicate in rodent fibroblasts even when they are engineered to express HIV-1 receptors because of an incompatibility between the viral and host factors required for efficient gene expression. These host cell–specific blocks have proved extremely useful in enabling researchers to infer and subsequently discover and validate the existence of host cell factors that are required for HIV-1 replication.
However, what was not appreciated until quite recently is that evolution has equipped cells with a variety of genes whose major and perhaps only role is to prevent retrovirus replication (1, 2). The products of these inhibitory genes, termed restriction factors, are nearly as important as required cofactors in determining the cellular host range of HIV-1 and other retroviruses. The best known and characterized of the restriction factors are encoded by the TRIM5 and APOBEC3G genes (3, 4), which, based on their potency, specificity, and evident species-specific positive selection as well as the clear existence of viral countermeasures, seem likely to have evolved for the sole purpose of combating viral infection.
In addition to a number of known host-encoded antiretroviral defense mechanisms, an increasing number of genes that can reduce the sensitivity of cultured cells to retrovirus infection in a less obviously directed way have also been identified (5–9). It can be shown that the overexpression of these genes can inhibit susceptibility to retroviruses or that their underexpression can enhance susceptibility to infection. However, whether these genes directly block some critical step in the retrovirus life cycle or affect virus growth indirectly by affecting some aspect of cell physiology is not apparent. Some of these genes play important and obvious roles in the life of cells independent of retroviral infection, and so it is not clear whether their apparent antiretroviral activity is incidental.
In this issue of the JCI(10) and in a previous study (11), Zhang et al. have identified p21Waf1/Cip1/Sdi1 (p21) as a gene product that can influence the sensitivity of HSCs to HIV-1 infection. HSCs are one of a few cell types, others being undifferentiated monocytes and unstimulated CD4+ T cells, that resist HIV-1 infection despite the presence of HIV-1 receptors (12–14). In the case of HSCs, CD4 expression appears low, and previous work indicates a major block to HIV-1 infection of HSCs is at virus entry, since infection can be achieved using pseudotyped HIV-1 virions that carry a different viral envelope (14). However, later work also suggests that a second block to HIV-1 replication is imposed by p21 because its depletion could enhance sensitivity to pseudotyped HIV-1 vectors by about 2- to 4-fold (11). The new work in this issue extends this finding and shows that low-level–spreading HIV-1 replication can be obtained in HSCs if these cells are transfected with siRNAs that deplete p21. Additionally, Zhang et al. present a number of experiments designed to illuminate the molecular mechanism underlying the enhancement of HIV-1 replication in HSCs upon p21 depletion (10).
So, is the effect of p21 on HIV-1 infection a direct effect on the incoming virus or a secondary effect of modulating cell physiology? The normal cellular functions of the p21 protein suggest that either could be the case. Notably, p21 is a cyclin-dependent kinase (CDK) inhibitor (15) and particularly targets CDKs that are active in the G1 phase of the cell cycle. In doing so, p21 inhibits cell-cycle progression in some cellular types, including HSCs (16). Although cell-cycle progression is not absolutely required for HIV-1 infection, quiescent cells (in G0 phase) are known to be very poor targets for HIV-1 infection, at least in part because reverse transcription is not completed (17). Additionally, p21 binds to factors involved in DNA repair pathways (18). The requirement for DNA repair activities in HIV-1 infection, particularly at integration, has been debated for some years, and recent findings suggest that some DNA repair activities may actually inhibit HIV-1 infection, perhaps by targeting viral DNA prior to integration into the target cell genome (19). Overall, therefore, there are at least 2 plausible models by which p21 could inhibit HIV-1 infection.
Zhang et al. (10) argue for a direct effect of p21 on incoming HIV-1. Data in the study hint that p21 binds to an incoming subviral complex containing HIV-1 matrix and integrase, known as the HIV-1 preintegration complex (PIC). This is consistent with the notion that the effect could be a direct one on viral components. However, caution is warranted here because this type of experiment is notoriously difficult and artifact prone. Nonetheless, it appears that the effect of p21 depletion is specific to HIV-1 infection, as low-level replication in HSCs of the simian immunodeficiency virus SIVmac251 is unaffected by p21 depletion. The lack of effect on a related virus is more difficult, albeit not impossible, to reconcile with a model in which the effect of p21 depletion on HIV-1 replication has an indirect effect on cell physiology. It would be useful to know whether SIVmac251 PICs associate with p21 and what viral determinants govern the apparently discordant effects of p21 on HIV versus SIVmac251 replication — this would provide strong clues as to an underlying mechanism. If the effects of p21 are direct, it might be possible to show that resistance can be induced in other cellular contexts by ectopic expression of p21. For now, the central role of p21 in cell physiology, and likely a complex cascade of events that accompany its depletion, urge circumspection in formulating models to explain why it appears to inhibit HIV-1 replication. The authors do show that p21-depleted cells do not begin to proliferate until some time after siRNA transfection, but this does not compellingly refute the possibility that p21’s effects on HIV-1 infection are indirect because the physiological changes that lead to DNA synthesis and cell division obviously precede the events themselves.
These caveats aside, the most striking and provocative finding in the current study (10), and one that suggests a specific effect on the incoming HIV-1 PIC, is the rather dramatic difference in the fate of nascent viral DNA that was observed in HSCs upon p21 depletion. In fact, the effects of p21 depletion on the fate of HIV-1 DNA appear much greater than the effects of p21 depletion on infection (as measured by the expression of a reporter gene embedded in the HIV genome). Specifically, control-infected HSCs appear to accumulate circular forms of HIV-1 DNA while infection of p21-depleted cells results in far more integrated proviruses and very low–level circle formation. Excessive circle formation is a hallmark of failed integration; pharmacological or mutational inhibition of HIV-1 integrase catalytic activity induces precisely this phenotype. Reasonably, the authors speculate that DNA repair systems that interact with p21 (18) might be responsible for this phenomenon. The fate of HIV-1 DNA following its entry into the nucleus is a step in the viral life cycle that is increasingly recognized to be influenced by host factors; for example, LEDGF, a chromatin-associated host protein that directly binds HIV-1 integrase, appears to be rather important for integration (20). DNA repair pathways are also increasingly recognized as inhibiting retroviral infection (19); thus, there are a number of potential mechanisms by which p21 depletion could promote integration at the expense of circle formation.
Based on the current findings (10), it would seem premature to dub p21 a bona fide restriction factor. Nonetheless, it does join the growing list of gene products that can influence cellular sensitivity to HIV-1 infection, and its effects on nascent HIV-1 DNA are unique and interesting. Further work will be required to determine precisely how p21’s effects on HIV-1 DNA are mediated and how p21’s effects relate to the function of other host proteins that determine the fate of HIV-1 DNA once it has entered the target cell nucleus.
Work in the author’s laboratory is supported by the NIH, amFAR, and the Elizabeth Glaser Pediatric AIDS Foundation.
Nonstandard abbreviations used: p21, p21Waf1/Cip1/Sdi1; PIC, HIV-1 preintegration complex.
Conflict of interest: The authors have declared that no conflict of interest exists.
Reference information: J. Clin. Invest.117:302–304 (2007). doi:10.1172/JCI31290.
See the related article beginning on page 473.