Host SAMHD1 protein restricts endogenous reverse transcription of HIV-1 in nondividing macrophages
SAM domain and HD domain containing protein 1 (SAMHD1) is a host anti-HIV-1 restriction factor known to suppress viral reverse transcription in nondividing myeloid cells by its dNTP triphosphorylase activity that depletes cellular dNTPs. However, HIV-2 and some SIV strains rapidly replicate in macrophages due to their accessory protein, viral protein X (Vpx), which proteosomally degrades SAMHD1 and elevates dNTP levels. Endogenous reverse transcription (ERT) of retroviruses is the extra-cellular reverse transcription step that partially synthesizes proviral DNAs within cell-free viral particles before the viruses infect new cells. ERT activity utilizes dNTPs co-packaged during budding from the virus-producing cells, and high ERT activity is known to enhance HIV-1 infectivity in nondividing cells. Here, since Vpx elevates cellular dNTP levels in macrophages, we hypothesize that HIV-2 should contain higher ERT activity than HIV-1 in macrophages, and that the Vpx-mediated dNTP elevation should enhance both ERT activity and infectivity of HIV-1 particles produced in macrophages.
Here, we demonstrate that HIV-2 produced from human primary monocyte derived macrophages displays higher ERT activity than HIV-1 produced from macrophages. Also, HIV-1 particles produced from macrophages treated with virus like particles (VLPs) containing Vpx, Vpx (+), displayed large increases of ERT activity with the enhanced copy numbers of early, middle and late reverse transcription products within the viral particles, compared to the viruses produced from macrophages treated with Vpx (−) VLPs. Furthermore, upon the infection with an equal p24 amount to fresh macrophages, the viruses produced from the Vpx (+) VLP treated macrophages demonstrated higher infectivity than the viruses from the Vpx (−) VLP treated macrophages.
This finding identifies the viral ERT step as an additional step of HIV-1 replication cycle that SAMHD1 restricts in nondividing myeloid target cells.
KeywordsHIV-1 HIV-2 Reverse transcription Endogenous reverse transcription SAMHD1 Vpx dNTPs Macrophages
human immunodeficiency virus type 1
human immunodeficiency type 2
simian immunodeficiency virus
viral protein X
SAM domain and HD domain containing protein 1
endogenous reverse transcription
virus like particles
All lentiviruses including type 1 and type 2 human immunodeficiency virus (HIV-1 and HIV-2) and simian immunodeficiency virus (SIV) infect both activated/dividing CD4+ T cells and terminally differentiated/nondividing myeloid cells such as macrophages and microglia during the course of their pathogenesis [1, 2]. However, while HIV-1 rapidly replicates in activated CD4+ T cells, HIV-1 replication in nondividing myeloid cells is kinetically suppressed [3, 4, 5]. We previously demonstrated that this restricted replication kinetics of HIV-1 in myeloid cells is due to the extremely low dNTP concentrations found in this nondividing cells type, which blocks the viral reverse transcription step that consumes cellular dNTPs during proviral DNA synthesis . A series of recent studies reported that a host dNTPase, sterile alpha motif (SAM) domain and histidine-aspartate (HD) domain containing protein 1 (SAMHD1), which hydrolyzes dNTPs to dNs and triphosphates, is responsible for the poor dNTP availability in macrophages, suggesting that SAMHD1 is a myeloid cell specific host restriction factor against HIV-1 [7, 8].
Unlike HIV-1, HIV-2 and some SIV strains replicate rapidly even in macrophages due to their unique accessary protein, viral protein X (Vpx) , which counteracts the anti-viral activity of SAMHD1 [7, 10]. Vpx recruits SAMHD1 to the CRL4 (DCAF1) E3-ligase complex for the proteosomal degradation of SAMHD1, leading to the elevation of cellular dNTP levels and rapid reverse transcription kinetics of these Vpx encoding lentiviruses in macrophages [11, 12]. However, the replication of HIV-1 in activated CD4+ T cells is not significantly affected by the dNTPase activity of SAMHD1 because SAMHD1 in dividing cells is phosphorylated at its C-terminal site (Threonine at the residue 592) , which blocks the formation of the enzymatically active tetramer form of SAMHD1 [14, 15]. Furthermore, the dNTP biosynthesis, which is closely tied to cell cycle, is active in dividing cells, and cellular dNTPs in activated/dividing CD4+ T cells are highly abundant (1–5 μM), compared to nondividing macrophages (20–40 nM) , which can support the robust replication kinetics of HIV-1 in this dividing cell type.
In addition to the reverse transcription step of HIV-1, the DNA gap filling step of the HIV-1 integration requires cellular dNTPs. The integration step of HIV-1, which is mediated by virally encoded integrase, leaves 2–3 nucleotide single stranded (ss) DNA gaps at both 5′ sides of the partially integrated double stranded proviral HIV-1 DNAs . These ss DNA gaps can be repaired by host DNA polymerases, which also consume cellular dNTPs. Indeed, recent biochemical studies suggested that the HIV-1 DNA gap repair is dependent on the cellular dNTP availability and SAMHD1-mediated dNTP depletion can kinetically delay the gap filling reaction during viral integration process in nondividing cells [17, 18].
Mature HIV-1 particles can initiate reverse transcription even before infection if dNTP substrates are available within viral particles. Indeed, it was previously reported that cell free mature HIV-1 particles contain partially synthesized proviral DNAs even before they infect cells, suggesting that the viral cores contain dNTPs and these dNTP are likely co-packaged during virus budding. This cell-free reverse transcription of HIV-1 is called endogenous reverse transcription (ERT) . With this ERT activity, HIV-1 can proceed proviral DNA synthesis through not only the early 1st strong-stop DNA product but also intermediate/late (−) strand DNA products before infection. Furthermore, HIV-1 particles with high ERT activity, which already initiated the proviral DNA synthesis, display higher viral infectivity particularly in macrophages  likely because these viruses already completed some of the reverse transcription steps that is rate limiting in this nondividing cell type due to the SAMHD1-mediated dNTP depletion.
In this study, since the Vpx-induced dNTP elevation in macrophages can allow the budding viral particles to co-package more dNTPs, we hypothesize that HIV-2 produced from macrophages should have higher ERT activity than HIV-1 produced from macrophages. We further hypothesize that Vpx can enhance the ERT activity of HIV-1 in macrophages and the infectivity of the produced viruses to new macrophages. Indeed, this study demonstrates that HIV-2 harbors higher ERT activity than HIV-1 in macrophages, and that Vpx was able to increase both ERT activity and infectivity of HIV-1 in macrophages. These findings support that the higher ERT activity of HIV-2 may contribute to its efficient replication kinetics in macrophages, and that the ERT step is an additional step of HIV-1 replication cycle that is restricted by host SAMHD1 protein in nondividing myeloid cells.
ERT activity comparison between HIV-1 and HIV-2 produced from human primary monocyte derived macrophages
Primary monocyte derived macrophages infected with HIV-1 89.6 or HIV-2 Rod10 were washed to remove the remaining uninfected viruses, and then the culture media were collected at days 2, 3, and 4 post infection for the isolation of the total viral nucleic acids. The ERT activity of HIV-1 89.6 and HIV-2 Rod10 for the three different regions of the viral gene were determined. Indeed, as shown in Fig. 1b–d, HIV-2 Rod10 displayed 11-15 times higher ERT activity in all three regions than HIV-1 89.6 produced from macrophages. We also infected activated CD4+ T cells isolated from the same donors with the same amount of HIV-1 89.6 and HIV-2 Rod10 as used for the macrophage infection. For activated CD4+ T cells, the infected cells were cultured for 3 days and the produced viral particles were collected at days 1, 2 and 3 post infection for the ERT activity assay. As shown in Fig. 1e, unlike the viruses produced from macrophages, both HIV-1 89.6 and HIV-2 Rod10 showed very similar ERT activity in activated CD4 + T cells that harbor abundant cellular dNTPs. The data presented in Fig. 1 demonstrates that HIV-2 has higher ERT activity than HIV-1 in nondividing macrophages, which is likely due to Vpx that elevates cellular dNTPs and then enhance the co-packaging of cellular dNTPs into the produced HIV-2 viral particles.
ERT activity of HIV-1 89.6 in activated CD4+ T cells and macrophages with and without Vpx treatment
To test the hypothesis that the treatment of macrophages with Vpx can enhance the ERT activity of HIV-1, as illustrated in Fig. 2b, we pre-treated macrophages from 5 donors with Vpx (−) and Vpx (+) VLPs for 24 h, and then infected cells with HIV-1 89.6. The remaining uninfected viruses in the media were extensively washed at 9 h post infection, and the culture media containing the produced viral particles were collected at every 24 h for 4 days. The total nucleic acids extracted from the collected media were applied for the ERT assay for the early region. As shown in Fig. 2c, the Vpx treatment of macrophages enhanced the ERT activity of HIV-1 89.6 by 5 folds. HIV-1 89.6 harvested from activated CD4+ T cells at day 3 showed 23 times higher ERT activity than the viruses harvested at day 4 from macrophages treated with Vpx (−) VLPs (Fig. 2c). The order of ERT activities of the viruses analyzed for the early region is: activated CD4+ T cells > Vpx (+) VLP-treated macrophages > Vpx (−) VLP-treated macrophages. Importantly, the order of the ERT activity among the viral particles produced from these three conditions is correlated with the order of the cellular dNTP concentrations found in these three conditions (Fig. 2a): activated CD4+ T cells (3–5 µM) > Vpx (+) VLP treated macrophages (200–700 nM) > Vpx (−) VLP treated macrophages (20–60 nM). However, the ERT activity for the middle and late regions were not detected, presumably because the detection of the middle and late products may need longer virus culture.
Effect of Vpx on the ERT activity during long-term culture of macrophages repeatedly treated with Vpx
Effect of dN treatment on HIV-1 ERT activity In macrophages
Comparison of replication kinetics and Infectivity of HIV-1 89.6 produced from activated CD4+ T cells and macrophages with and without Vpx treatment
While HIV-1 replication is kinetically suppressed, HIV-2 and some SIV strains are able to rapidly replicate even in macrophages. Vpx of HIV-2 and some SIV strains (i.e. SIVsm) as well as Vpr of other SIV strains (i.e. SIVagm)  enables these lentiviruses to overcome the SAMHD1-mediated dNTP depletion in macrophages. So far, two major steps of lentivirus life cycle, reverse transcription and DNA gap filling during integration, were known to be affected by SAMHD1, and the Vpx-induced SAMHD1 degradation and subsequent dNTP level elevation can kinetically facilitate both reverse transcription and DNA gap filling specifically in macrophages. Now, the data presented in this study support that the higher ERT activity of HIV-2 may also contribute to the rapid replication kinetic of HIV-2 in macrophages.
The dNTP concentration found in macrophages is below the Km values of HIV-1 RT, and therefore, the enzymatic proviral DNA synthesis rate of HIV-1 RT at the macrophage dNTP concentration is below the maximal rate of HIV-1 RT . However, Vpx enhances the cellular dNTP concentration (Fig. 2a) above the Km values of HIV-1 RT in macrophages, which can accelerate the reverse transcription kinetics in macrophages. Furthermore, our data the dNTP elevation by Vpx in macrophages enhances the ERT activity, which also elevate the infectivity of the viruses produced from the Vpx (+) VLP treated macrophages to fresh macrophages. Interestingly, the macrophages dNTP concentration enhanced by Vpx is still lower than the dNTP concentrations found in activated CD4+ T cells (3–5 μM, Fig. 2a) , and this high dNTP concentration found in activated CD4+ T cells contributed to the highest ERT activity and infectivity. It was previously reported that the early ERT product can occur independently of intracellular dNTP levels, instead by using exogenous dNTPs that could exist in the natural microenvironments . However, our tissue culture experimental setting does not include the exogenous dNTPs. Overall, the data presented in this study support that the cellular dNTP concentrations in the virus-producing cells can contribute to the infectivity of the produced HIV-1 particles to macrophages by modulating the ERT activity of the produced viruses that consumes the dNTPs co-packaged during viral budding.
Figure 5c illustrates three steps of HIV-1 replication cycle that are suppressed by the dNTPase activity of the host SAMHD1 protein because these three steps all consume dNTPs. First, the reverse transcription step was the first identified step of HIV-1 replication cycle in nondividing myeloid cells that host SAMHD1 restricts by reducing the cellular dNTP concentration [11, 12]. Second, since the ssDNA gap filling step of HIV-1 integration requires cellular dNTPs, the DNA gap filling step can be controlled by SAMHD1 [17, 18]. Finally, the data presented in this study identified the ERT step of HIV-1 as the third step of HIV-1 replication cycle that is restricted by host SAMHD1 in nondividing myeloid cells. Vpx, which counteracts SAMHD1, can derepress these three steps of HIV-1 replication cycle restricted by SAMHD1. In conclusion, this study suggests that SAMHD1 is a highly effective anti-HIV-1 restriction factor that can suppress three different steps of viral replication cycle by modulating cellular dNTP levels in nondividing myeloid cells.
Cells, viruses and virus like particles
Human primary macrophages were prepared by the GMCSF-mediated differentiation of monocytes isolated from buffy coats as previously described [24, 25]. Briefly, monocytes were isolated by positive selection (CD14 microbeads; Miltenyi Biotec) using an automated magnetic cell sorter (AutoMacs, Miltenyi Biotec) from human buffy coats of 5 healthy donors (LifeSouth Community Blood Center, Dunwoody, GA). Monocytes were pooled in an equal number per donor and then differentiated to macrophages with GMCSF (10 ng/ml, Miltenyl Biotec) for 7 days prior to use with all experiments. HIV-1 89.6 viruses were initially prepared from the transfection of HIV-1 89.6 molecular clone plasmid (kindly provided from NIH AIDS Reagent Program, Division of AIDS, NIAID, NIH, https://www.aidsreagent.org/reagentdetail.cfm?t=viruses&id=704) to 293 FT cells (Invitrogen). The produced viruses were propagated for 8 passages in CEMx174 cells until the beta-lactamase gene encoded in the plasmid was not detected in the PCR amplification in the viral nucleic acid extracted from the culture media; the absence of this gene demonstrated a plasmid-free culture. HIV-1 p24 ELISA (Advanced Bioscience Laboratories Inc.) was used for monitoring and quantitating the produced viruses. HIV-2 Rod was also prepared as described for HIV-1 89.6, and the viral production was monitored by p27 ELISA kit (Advanced Bioscience Laboratories Inc.). HIV-1 BaL was also obtained from NIH AIDS Reagent Program. Vpx (−) and Vpx (+) virus like particles (VLPs) were prepared as previously described . Briefly, 293 FT cells were transfected with pSIV Vpx (−) or pSIV Vpx (+) and pVSV-G, and the media were collected in day 3. After removing the cells and other debris by low-speed centrifugation and filtration, the produced VLPs were collected by ultracentrifugation (28,000 rpm for 90 min). The pellets were re-suspended to serum free DMEM and aliquots were stored at − 80 °C.
Assay for SAMHD1 degradation by Vpx
2 × 106 macrophages were treated with Vpx (−) or Vpx (+) VLPs in 6 well plates, and the cell lysates were prepared with RIPA buffer at 24 h post treatment. Human SAMHD1 and GAPDH levels in the lysates were determined by western blots. Human anti-SAMHD1 (Abcam) and anti-GAPDH (Abcam) as well as secondary anti-Mouse and anti-Rabbit antibodies (GE Healthcare) were used for the visualization.
Assay for cellular dNTP levels
dNTPs in 2 × 106 macrophages treated with Vpx (−) or Vpx (+) VLPs for 24 h were extracted with 65% methanol, and the extracted dNTPs were determined by the primer extension based dNTP assay as previously described . The amount of each of four dNTPs per 1 × 106 cells as well as dNTP concentration was calculated for comparison.
Assay for ERT activity of HIV-1
1 × 106 macrophages pre-treated with Vpx (−) or Vpx (+) VLPs for 24 h and 2.5 mM dNs for 3 h and activated CD4+ T cells were infected in triplicates with an equal p24 level of HIV-1 89.6. After 9-h incubation, the remaining viruses were removed by washing the cells three times. The infected cells were cultured and the viruses in the media were collected at varying time points. The total viral nucleic acids were isolated by QIAGEN EZ1 virus mini kit v2.0. The regions of HIV-1 89.6 gene , early, mid and late regions, were chosen for the determination of viral RNA and DNA copy numbers in the isolated total viral nucleic acids. Primers for HIV-1 89.6 ERT assay: early region (559–643) primers: 5′-GTGCCCGTCTGTTGTGTGAC-3′ and 5′-GGGCGCCACTGCTAGAGATTT-3′, early region probe: 5′-CTAGAGATCCCTCAGACCATCC TAGTTAGTGTAG-3′. Middle region (8782–8928) primers: 5′-CTATAAGATGGGAGGCAAGTG G-3′ and 5′-CTTGTGATAGCTCCATGTCTCG-3′. Middle region probe: 5′-AAACGTAGGGCAG AGGGATGGC-3′. Late region (556–698) primers: 5′-TGTGTGCCCGTCTGTTGTGT-3′ and 5′-GAGTCCTGCGTCGAGAGATC-3′, late region probe: 5′-CAGTGGCGCCCGAACAGGGA-3′. Primers for HIV-2 Rod ERT assay: early region (225–309) primers: 5′-TCACCTGAGTAACAAGACCC-3′ and 5′-GGCGCCAACCTGCTAGG GATTT-3, early region probe: 5′-TTCTTGCTTTGG GAAACCGAGGC-3′. Middle region primers (8681–8780): 5′-TCCAAGAAGGATCAGACAGG-3′ and 5′-TCGCC CTCCTGTGAGGGACGG-3′, middle region probe: 5′-ATCGCCCTCCTGTGAG G-3′. Late region primers (241–352): 5′-ACCCTGGTCTGTTAGGACCC-3′ and 5′-GCCGTGTTC CAAGACTTCTCAG-3′, late region probe: 5′-TCCCTAGCATGGCGCCTG-3′. For both RNA and DNA copy determination, Q-RT PCR (LightCycler® 480 RNA Master Hydrolysis Probes) was used, and, for DNA copy number, the same protocol and reagents were used except skipping the reverse transcription step of the PCR protocol with the inactivation of RT contained in the RT-PCR kits by preheating at 98 °C. p89.6 plasmid was used as a copy number control, and Q-RT PCR and Q-PCR were conducted in triplicates per viral nucleic acid sample. ERT activity was determined by ratio between DNA copy number and (RNA + DNA) copy number in a single viral nucleic acid sample extracted. The same primers and probes were used from the analysis of the early ERT activity of HIV-1 BaL. Middle region primers for BaL: 5′TGGGTTATGAACTCCATCCTGAT′3 and 5′TGTCATTGACAG TCCAGCGTCT′3, middle region probe 5′TTTCTGGC AGCACTATAGGCTGTACTGTCCATT.
Quantitative 2LTR circle DNA PCR
Macrophages from 5 healthy donors were infected with an equal p24 amount of HIV-1 89.6 collected at day 4 from the Vpx (−) VLP or Vpx (+) VLP treated macrophages as well as at day 3 from the activated CD4+ T cells (Fig. 2) in triplicates. The infected cells (in triplicates) were collected and the total cellular genomic DNA was isolated from the cells (Wizard® Genomic DNA Purification Kit), and used for quantitative HIV-1 2LTR circle DNA PCR. The protocol, primers and probe for the quantitative HIV-1 89.6 2LTR circle PCR were the same as the one previously used for NL4-3 .
Viral production assay by Q-RT PCR
The viruses in the culture media of the infected cells (triplicates) were collected and the total nucleic acids were extracted from the viruses as described above. The collected viruses were quantitated by Q-RT PCR for the copy numbers of the early reverse transcription product in the extracted total viral nucleic acids as described above.
Statistical analyses were performed with an unpaired t test (Figs. 1b–e, 3a–c, 4a–c) comparing each time point independently in each figure. The statistical analysis was also performed using an ordinary one-way analysis of variance (ANOVA) with multiple comparisons comparing each column to the Vpx- or NT control (Figs. 2a, c, 5a, b). Time point Days 3 and 4 were analyzed together in Fig. 2a. The statistical significance of differences between data is indicated as follows: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.
BM: acquisition of data and analysis and interpretation of data. CG: acquisition of data. DHK: conception, RFS: conception, and BK: conception, design, drafting and finalizing manuscript. All authors read and approved the final manuscript.
We appreciate Dr. James Kohler for reading the manuscript.
The authors declare that they have no competing interests.
Availability of data and materials
All data generated or analyzed during this study are included in this published article (and its additional file).
Consent for publication
Ethics approval and consent to participate
No human subjects were used in this study.
This work was supported by NIH AI136581 (B.K.), GM104198 (B.K.), and MH1166950 (R.F.S.).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 11.Lahouassa H, Daddacha W, Hofmann H, Ayinde D, Logue EC, Dragin L, Bloch N, Maudet C, Bertrand M, Gramberg T, et al. SAMHD1 restricts the replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside triphosphates. Nat Immunol. 2012;13(3):223–8.CrossRefGoogle Scholar
- 15.Bhattacharya A, Wang Z, White T, Buffone C, Nguyen LA, Shepard CN, Kim B, Demeler B, Diaz-Griffero F, Ivanov DN. Effects of T592 phosphomimetic mutations on tetramer stability and dNTPase activity of SAMHD1 can not explain the retroviral restriction defect. Sci Rep. 2016;6:31353.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.