|Home | About | Journals | Submit | Contact Us | Français|
A long-standing question in the HIV field is why HIV-1 fails to replicate in resting CD4+ T cells. A new study shows that the host deoxynucleoside triphosphate (dNTP) triphosphohydrolase (dNTPase) sterile alpha motif and histidine/aspartic domain-containing protein 1 (SAMHD1), previously shown to block HIV infection in myeloid cells, also restricts HIV replication in resting CD4+ T cells by hydrolyzing dNTPs, which are needed for reverse transcription of the virus (aaa-bbb).
One of the puzzles of HIV-1 biology is why the virus barely replicates in resting CD4+ T lymphocytes, whereas activated CD4+ T lymphocytes support robust HIV-1 replication. In HIV-infected people, very few T lymphocytes in the peripheral blood or lymph node are actively infected (even in the absence of antiviral treatment) and viral replication is mostly restricted to activated CD4+ T cells. In resting CD4+ T cells, reverse transcription, which copies the viral genomic RNA to cDNA soon after viral entry prior to nuclear import and chromosomal integration, is very inefficient and other barriers to viral production appear at later steps in the viral life cycle1,2.
The study by Baldauf et al.3 in this issue of Nature Medicine provides a simple and elegant explanation for HIV-1 restriction in resting T cells (Figure). The authors identify SAMHD1, a dNTPase that removes the triphosphate from dNTPs, as a host restriction factor for HIV-1 in resting CD4+ T cells. Mutations in SAMHD1 are associated with Aicardi-Goutières syndrome (AGS), an autoinflammatory genetic brain disorder of newborns that mimics congenital infection and is characterized by elevated amounts of type I interferons4. SAMHD1 was identified in 2011 as the long-sought restriction factor for HIV productive infection of myeloid dendritic cells, monocytes and macrophages5,6. SAMHD1 limits HIV-1 replication in these cells by depleting intracellular dNTPs, thus blocking HIV DNA synthesis by reverse transcription7. High dNTP concentrations are not needed in nondividing cells and are kept low; myeloid cells have about 100-fold lower concentrations of dNTPs than activated T cells. Although HIV-1 reverse transcriptase very efficiently utilizes dNTPs, the dNTP concentration in myeloid cells is rate-limiting. The small lentiviral accessory protein Vpx (from SIV or HIV-2, which is missing in HIV-1) targets SAMHD1 for ubiquitylation and degradation. Supplying Vpx or knocking down SAMHD1 at the time of HIV-1 infection enhances dNTP concentrations within hours, promotes reverse transcription and drastically improves HIV-1 replication in myeloid cells.5–7
The original papers describing SAMHD15,6 assumed that this factor was not operating in T cells because the authors did not detect SAMHD1 expression in T cell lines. However, Baldauf et al.3 found that supplying Vpx before or at the time of HIV-1 challenge greatly enhanced reverse transcription and HIV gene expression in resting T cells, as it had in myeloid cells. Consistent with these results, they showed that SAMHD1 is highly expressed in resting primary human CD4+ T cells, similar to in the THP-1 monocyte cell line, but is not expressed in Jurkat T cells (as previously reported). Vpx does not alter the quiescent state of resting T cells, but elevates cellular dNTP concentrations to facilitate HIV reverse transcription by activating proteasomal degradation of SAMHD1. To confirm that SAMHD1 is the restriction factor overcome by Vpx, the authors knocked down SAMHD1 in resting CD4+ T cells and observed increased HIV replication3. The incubation of resting CD4+ T cells with dNTPs also enhanced their infection with HIV-1. The authors also found that resting CD4+ T cells from an AGS patient with a nonsense mutation in SAMHD1 are permissive for HIV-1. This result is consistent with a previous study that found that SAMHD1 deficiency significantly enhanced HIV-1 replication compared with normal controls in PBMCs, but only when the PBMCs were not activated8.
Baldauf et al.3 also made several unexpected observations about SAMHD1 expression. First, SAMHD1 expression was similar in resting and activated CD4+ T cells. As T cell activation is linked to cell proliferation and dividing cells need dNTPs for DNA replication, one would expect reduced SAMHD1 expression following T cell activation. This is not the case3, suggesting that dNTP concentrations are maintained in activated T cells by other enzymes, that SAMHD1 activity is regulated post-translationally or it requires interaction with other host factors, or that its expression is tightly controlled during the cell cycle and is only degraded transiently around the time of DNA synthesis when elevated dNTP concentrations are needed (only a fraction of activated T cells are in S phase at any time). Consistent with this latter idea, the authors found that a proliferation marker (Ki67) and SAMHD1 were never coexpressed in immune cells in human tonsil sections3. However, a more careful analysis of SAMHD1 expression and dNTP concentrations (and HIV-1 susceptibility) in activated T cells at different phases of the cell cycle is also warranted.
Another unexpected finding by Baldauf et al.3 was that SAMHD1 localizes to both the nucleus and cytoplasm in T cells and monocyte-derived macrophages, as previous studies suggested that SAMHD1 is mostly located in the nucleus4,8–10. Vpx arrives in the cytoplasm with the incoming virion, but is actively transported to the nucleus. Where in the cell Vpx and SAMHD1 interact needs to be clarified, especially given that there are conflicting reports about whether leptomycin B, which inhibits nuclear export of proteins, interferes with Vpx-mediated SAMHD1 degradation9,10.
But how much does SAMHD1 contribute to the overall restriction of HIV-1 replication in resting CD4+ T cells? Most of the experiments in this study3 and in the previous myeloid cell papers5–7,9–13 compared the effect of supplying Vpx, exogenous dNTPs or knocking down SAMHD1 on early steps in the viral life cycle by measuring the amounts of early reverse transcripts or HIV-1 long terminal repeat (LTR)-driven transcription from a single round of replication of an integrated reporter virus. Baldauf et al.3 found that HIV-1 transcription from Vpx-treated resting CD4+ T cells is about 4 fold less robust than from Vpx-untreated activated CD4+ T cells from the same donor. The residual advantage to HIV-1 in activated T cells is likely explained by the greater nuclear abundance of transcription factors, such as NFAT, FOS, JUN or nuclear factor-κB, that are activated by T cell activation and are also required to activate the HIV-1 LTR. However, viral reverse transcription, integration and transcription is not the whole story. The authors could not detect viral production in resting CD4+ T cells infected with Vpx-containing HIV-1, thus, unknown Vpx-insensitive block(s) to HIV production exist at a later stage in the viral life cycle. It is unclear whether this might also be the case in myeloid cells. Although two studies found that Vpx-treated monocyte-derived dendritic cells are more efficient at transmitting the virus to activated CD4+ T cells in in vitro co-culture11,12, it is uncertain whether cell-free viruses were produced by the infected dendritic cells.
With this new study by Baldauf et al.3, SAMHD1 is becoming an increasingly important host restriction factor for HIV-1 in both myeloid cells and resting CD4+ T cells. During its evolution from primate lentiviruses, the accessory gene that counters SAMHD1 (vpx) was lost. Do loss of vpx and persistent SAMHD1 expression and the corresponding limit to infection in myeloid cells and resting CD4+ T cells provide an advantage to HIV-1? It would seem so. Although the virus strain SIVmac239, which lacks vpx, is slightly less pathogenic than the wild type SIV virus in rhesus monkeys13, in humans, HIV-1 (which lacks vpx) is much more pathogenic than HIV-2 (which resembles macaque SIV and contains vpx). Although these findings are at first counter-intuitive, the answer to why HIV-1 has lost vpx may lie in avoiding triggering an antiviral innate immune response in SAMHD1-restricted cells. When dendritic cells are infected with HIV-1 supplemented with Vpx, interferon-β - a cytokine that orchestrates a global antiviral response - is induced11,12. It is not clear how Vpx-mediated degradation of SAMHD1 leads to interferon induction in myeloid cells. HIV-1 reverse transcripts also trigger interferon production in T cells and macrophages when another AGS-linked gene TREX1 is knocked out or inactivated by genetic mutation14. Trex1 is a cytosolic DNase that digests excess non-productive HIV reverse transcripts to avoid triggering interferon responses in activated CD4+ T cells and macrophages.
Mutations in both TREX1 and SAMHD1 are linked to AGS and both enzymes target HIV-1 reverse transcription – SAMHD1 controls DNA synthesis (via dNTP concentrations) whereas Trex1 controls degradation of reverse transcripts, much like a kitchen sink analogy in which SAMHD1 controls the faucet and Trex1 controls the drain. Therefore, HIV-1 may take advantage of both molecules to achieve an optimal viral cDNA concentration to promote viral transmission while evading innate immunity. Further studies are needed to understand whether and how HIV-1 restriction might enable HIV to avoid triggering antiviral innate immunity and how this might enhance viral transmission and infection.
Nan Yan, Department of Internal Medicine and Department of Microbiology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.
Judy Lieberman, Program in Cellular and Molecular Medicine, Boston Children’s Hospital and Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, USA.