A novel DCAF1-binding motif required for Vpx-mediated degradation of nuclear SAMHD1 and Vpr-induced G2 arrest

Authors

  • Wei Wei,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author
    • These authors contributed equally.
  • Haoran Guo,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author
  • Xue Han,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author
  • Xianjun Liu,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author
  • Xiaohong Zhou,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    2. Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205, USA
    Search for more papers by this author
  • Wenyan Zhang,

    1. First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author
  • Xiao-Fang Yu

    Corresponding author
    1. Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205, USA
    • First Hospital of Jilin University, Institute of Virology and AIDS Research, Changchun, Jilin Province, China
    Search for more papers by this author

For correspondence. E-mail xfyu@jhsph.edu; Tel. (+1) 410 955 3768; Fax (+1) 410 955 0105.

Summary

HIV-2 and closely related SIV Vpx proteins are essential for viral replication in macrophages and dendritic cells. Vpx hijacks DCAF1–DDB1–Cul4 E3 ubiquitin ligase to promote viral replication. DCAF1 is essential for cell proliferation and embryonic development and is responsible for the polyubiquitination of poorly defined cellular proteins. How substrate receptors recruit the DCAF1-containing E3 ubiquitin ligase to induce protein degradation is still poorly understood. Here we identify a highly conserved motif (Wx4Φx2Φx3AΦxH) that is present in diverse Vpx and Vpr proteins of primate lentiviruses. We demonstrate that the Wx4Φx2Φx3AΦxH motif in SIVmac Vpx is required for both the Vpx–DCAF1 interaction and/or Vpx-mediated degradation of SAMHD1. DCAF1-binding defective Vpx mutants also have impaired ability to promote SIVΔVpx virus infection of myeloid cells. Critical amino acids in the Wx4Φx2Φx3AΦxH motif of SIV Vpx that are important for DCAF1 interaction maintained the ability to bind SAMHD1, indicating that the DCAF1 and SAMHD1 interactions involve distinctive interfaces in Vpx. Surprisingly, VpxW24A mutant proteins that were still capable of binding DCAF1 and SAMHD1 lost the ability to induce SAMHD1 degradation, suggesting that Vpx is not a simple linker between the DCAF1–DDB1–Cul4 E3 ubiquitin ligase and its substrate, SAMHD1.VpxW24A maintained the ability to accumulate in the nucleus despite the fact that nuclear, but not cytoplasmic, mutant forms of SAMHD1 were more sensitive to Vpx-mediated degradation. The Wx4Φx2Φx3AΦxH motif in HIV-1 Vpr is also required for the Vpr–DCAF1 interaction and Vpr-induced G2 cell cycle arrest. Thus, our data reveal previously unrecognized functional interactions involved in the assembly of virally hijacked DCAF1–DDB1-based E3 ubiquitin ligase complex.

Introduction

Vpx is a virion-associated viral accessory protein present in HIV-2 and selected SIV lineages (Henderson et al., 1988; Yu et al., 1988). Although Vpx has only a moderate effect on viral replication in CD4+ T lymphocytes, with some degree of individual variation (Guyader et al., 1989; Kappes et al., 1991; Marcon et al., 1991; Yu et al., 1991), it is essential for efficient viral replication in macrophages (Yu et al., 1991) and dendritic cells (Mangeot et al., 2002; Goujon et al., 2007). In SIV-infected monkeys, Vpx is required for viral dissemination and disease progression (Gibbs et al., 1995; Hirsch et al., 1998).

Vpx is packaged through a specific interaction with the carboxyl-terminus of the Gag polyprotein (Paxton et al., 1993; Wu et al., 1994; Accola et al., 1999; Selig et al., 1999). The presence of abundant Vpx in virions (Henderson et al., 1988; Yu et al., 1988) suggests a role during the early steps of the viral life cycle prior to de novo viral protein synthesis. Consistent with this concept, Vpx-deficient viruses exhibit defects in the accumulation of viral DNA during reverse transcription soon after the infection of myeloid cells (Fujita et al., 2008; Sharova et al., 2008; Srivastava et al., 2008; Gramberg et al., 2010). It is conceivable that either DNA synthesis or reverse transcription is blocked, or that newly synthesized viral DNA is degraded in the absence of Vpx.

Heterokaryon experiments have suggested that myeloid cells express one or more unique retrovirus inhibitors that are inactivated by Vpx (Kaushik et al., 2009). Recent studies have revealed that SAMHD1 is a myeloid cell-type specific inhibitor of HIV-1 and SIVmac (Berger et al., 2011; Hrecka et al., 2011; Laguette et al., 2011; Lahouassa et al., 2012). Indeed, SAMHD1 is linked to the reduction of viral DNA accumulation in HIV-1 virus-infected myeloid cells. Depletion of SAMHD1 from macrophages allows HIV-1 to replicate more efficiently. In addition to its role in suppressing SAMHD1 function in the cytoplasm (the site of viral reverse transcription), Vpx has also been reported to suppress the anti-viral activity of APOBEC3A (Berger et al., 2010) and to facilitate viral replication by suppressing interferon induced anti-viral activity in the nucleus (Pertel et al., 2011).

Vpx inactivates SAMHD1 by triggering its degradation through proteasome-dependent activity (Laguette et al., 2011; Hrecka et al., 2011; Ahn et al., 2012; Lim et al., 2012; Laguette et al., 2012). Vpx binds DCAF1 (also known as VPRBP), which in turn binds to DDB1 and Cul4 to form an E3 ubiquitin ligase complex (Le Rouzic et al., 2007; Sharova et al., 2008; Srivastava et al., 2008; Bergamaschi et al., 2009). Vpx mutants that cannot associate with DCAF1 are defective for the induction of SAMHD1 degradation. Knockdown of endogenous DCAF1 expression also compromises Vpx-mediated SAMHD1 degradation. However, how Vpx and DCAF1 interact with each other is not fully understood. The HIV-1 Vpr protein also interacts with DCAF1, which is required for Vpr-induced G2 cell cycle arrest (Battey et al., 2007; Belzile et al., 2007; DeHart et al., 2007; Hrecka et al., 2007; Le Rouzic et al., 2007; Schrofelbauer et al., 2007; Wen et al., 2007).

Here, we identify a novel DCAF1-binding motif in SIVmac Vpx and HIV-1 Vpr that is important for their activity: it is important for the interaction of SIVmac Vpx with DCAF1 and for Vpx-induced degradation of human SAMHD1. This DCAF1-binding motif in HIV-1 Vpr is also required for Vpr-induced G2 cell cycle arrest. The identification of this novel interaction between Vpx/Vpr and DCAF1 may facilitate the discovery of other DCAF1-binding cellular substrate receptor proteins and viral hijackers of Cul4-based E3 ubiquitin ligases.

Results

Identification of a highly conserved DCAF1-binding motif in Vpx

Both Vpx and Vpr from various HIV and SIV strains interact with DCAF1, suggesting the existence of a potential conserved mechanism for DCAF1 interactions. To determine whether there is a common functional motif in HIV/SIV Vpx and Vpr molecules that is involved in their interaction with DCAF1, we aligned and examined Vpx sequences from various HIV-2 subtypes and related SIV strains, as well as Vpr from HIV-1 and various SIV viruses. Sequence analysis of HIV/SIV Vpr/Vpx proteins identified a highly conserved Wx4Φx2Φx3AΦxH motif (Fig. 1A) in the α-helix 1 of Vpx proteins that may participate in the assembly of Vpx with the DCAF1–DDB1–Cul4 E3 ubiquitin ligase complex. To determine whether this motif is important for Vpx function, we generated a series of SIVmac Vpx mutant constructs in which individual or double residues within this motif were mutated (Fig. 1B). HEK293T cells were then transfected with the wild-type (WT) or one of the mutant Vpx–HFA expression vectors as indicated (Fig. 1C). The interaction of WT and mutant Vpx–HFA molecules with DCAF1 was evaluated by co-immunoprecipitation analysis. The Vpx–HFA proteins were immunoprecipitated from the cell lysates with a monoclonal antibody against the HA tag; co-precipitation of endogenous DCAF1 with various Vpx–HFA molecules was examined by immunoblotting with an antibody against DCAF1. DCAF1 was efficiently co-precipitated with WT SIVmac Vpx (Fig. 1C, lane 2). As expected, DCAF1 was not detected in the absence of Vpx (Fig. 1C, lane 1), indicating a specific interaction between Vpx and DCAF1. Several mutant Vpx proteins interacted poorly with DCAF1, including VpxV29S (Fig. 1C, lane 4), VpxI32S (lane 6), VpxA36S/V37S (lane 8) and VpxH39A (lane 9). As previously reported (Hrecka et al., 2011; Laguette et al., 2011), VpxQ76R also had a defect in DCAF1 interaction (Fig. 1C, lane 10). Since VpxV29S and VpxI32S were immunoprecipitated to a lesser amount than was WT Vpx (Fig. 1C), we repeated this experiment. Again, the interaction of VpxV29S with DCAF1 was reduced, but not abolished, when compared with the WT Vpx–DCAF1 interaction (Fig. 1D). In repeated experiments, VpxV29S, VpxI32S, VpxA36S/V37S and VpxH39A mutants showed a 50–90% reduction in DCAF1 binding when compared with the WT Vpx (Fig. 1E). These data suggest that helix1 of SIVmac Vpx indeed contains a critical DCAF1-binding motif.

Figure 1.

Identification of a conserved motif in Vpx required for DCAF1 binding.

A. Conserved amino acids in helix 1 of HIV-2 and SIV Vpx molecules. The N-terminal region of HIV-2 and related SIV Vpx sequences were aligned. A highly conserved motif, Wx4ΦxxΦx3AΦxH, corresponding to amino acids 24–39 of SIVmac Vpx was identified.

B. SIVmac Vpx mutant constructs.

C. The Wx4ΦxxΦx3AΦxH motif is required for Vpx–DCAF1 interaction. HEK293T cells were transfected with a control vector (lane 1, VR), WT Vpx (lane 2), or one of the indicated Vpx mutants (lanes 3–10). Cell lysates were prepared 48 h after transfection and immunoprecipitated with an anti-HA antibody. The interaction of endogenous DCAF1 with WT or mutant Vpx molecules was detected by immunoblotting with anti-HA antibody to detect Vpx–HFA and anti-DCAF1 antibody to detect endogenous DCAF1.

D. The interaction of endogenous DCAF1 with WT or certain mutant Vpx molecules was repeated.

E. Relative binding of WT Vpx (100%) and Vpx mutants. Error bars represent the standard deviation from triplicate experiments.

The DCAF1-binding motif of Vpx is critical for SIVmac Vpx-mediated degradation of SAMHD1

SIVmac Vpx targets SAMHD1 to DCAF1–DDB1–Cul4 E3 ubiquitin ligase complexes for polyubiquitination and proteasome-mediated degradation (Hrecka et al., 2011; Laguette et al., 2011). To establish our Vpx-mediated SAMHD1 degradation system, we transfected HEK293T cells with a SAMHD1–HA expression vector plus a control vector (Fig. 2A, lanes 1 and 2) or a vector expressing WT Vpx–HFA (lanes 3 and 4). Coexpression of Vpx and SAMHD1 in transfected HEK293T cells resulted in the depletion of intracellular SAMHD1 (Fig. 2A, lane 3) when compared with SAMDH1 levels in the absence of Vpx (lane 1). Addition of the proteasome inhibitor MG132 blocked the Vpx-induced depletion of SAMHD1 (Fig. 2A, lane 4), indicating that Vpx induced proteasome-mediated SAMHD1 degradation in our assay system.

Figure 2.

Effects of amino acid mutations in the highly conserved Wx4ΦxxΦx3AΦxH motif on Vpx function.

A. Effect of proteasome inhibitor MG132 on Vpx-mediated degradation of SAMHD1. HEK293T cells were co-transfected with SAMHD1–HA plus a control vector or Vpx–HFA expression vector. At 36 h after transfection, the cells were DMSO-treated or treated with MG132 for 16 h. Cell lysates were harvested and analysed by SDS-PAGE, followed by immunoblotting to detect SAMHD1–HA, with β-actin as loading control.

B. Effect of WT and mutant Vpx proteins on SAMHD1 degradation. HEK293T cells were co-transfected with a SAMHD1–HA expression vector plus control vector (lane 1, VR), WT Vpx (lane 2), or one of the indicated Vpx mutants (lanes 3–10). Expression of SAMHD1–HA and Vpx–HFA was assessed by immunoblotting with anti-HA antibody, with β-actin as loading control.

C. Interaction of WT and mutant Vpx proteins with SAMHD1. HEK293T cells were co-transfected with a SAMHD1–FLAG expression vector plus control vector (lane 1, VR), WT Vpx (lane 2), or one of the indicated Vpx mutants (lanes 3–6). Cell lysates were prepared 48 h after transfection and immunoprecipitated with anti-HA antibody. The interaction of SAMHD1 with WT or mutant Vpx molecules was detected by immunoblotting with anti-HA antibody to detect Vpx–HFA and anti-FLAG antibody to detect SAMHD1–FLAG.

To analyse the function of the Vpx mutants, we transfected HEK293T cells with the SAMHD1–HA expression vector plus a control vector (Fig. 2B, lane 1) or a vector expressing WT Vpx–HFA (lane 2), VpxW24A (lane 3), VpxV29S (lane 4), VpxE30A/E31A (lane 5), VpxI32S (lane 6), VpxE35A (lane 7), VpxA36S/V37S (lane 8), VpxH39A (lane 9) or VpxQ76R (lane 10). SAMHD1 expression was evaluated by immunoblotting. As compared with the WT Vpx, the DCAF1 binding-defective mutant Vpx proteins had a reduced ability to induce SAMHD1-degradation (Fig. 2B). In repeated experiments, VpxV29S (Fig. 2B, lane 4), VpxI32S (lane 6), VpxA36S/V37S (lane 8) and VpxH39A (lane 9) showed an impaired ability to promote SAMHD1 degradation when compared with the WT Vpx (Fig. 2B, lane 2). As previously reported (Hrecka et al., 2011; Laguette et al., 2011), VpxQ76R also had a defect in its ability to mediate SAMHD1 degradation (Fig. 2B, lane 10).

The DCAF1 binding-defective Vpx proteins still interact with SAMHD1

To promote SAMHD1 ubiquitination, the Vpx molecule requires at least two functional interactions, one with DCAF1 and another with SAMHD1. Mutations of amino acids in the Wx4Φx2Φx3AφxH motif of Vpx that affected DCAF1 binding should not affect the Vpx–SAMHD1 interaction unless these mutations have some general effect(s) on Vpx folding. We therefore used co-immunoprecipitation to evaluate the interaction of WT and mutant Vpx–HFA molecules with the target molecule SAMHD1–FLAG. HEK293T cells were transfected with an SAMHD1–FLAG expression vector plus a control vector or vector expressing WT or mutant Vpx–HFA, as indicated (Fig. 2C). SAMHD1 proteins were overexpressed to prevent significant VPX-induced degradation. The Vpx–HFA proteins were immunoprecipitated from the cell lysates with a monoclonal antibody against the HA tag; co-precipitation of SAMHD1–FLAG with various Vpx–HFA molecules was examined by immunoblotting using an antibody against FLAG to detect SAMHD1–FLAG. The ability of VpxV29S (Fig. 2C, lane 3), VpxI32S (lane 4), VpxA36S/V37S (lane 5) and VpxH39A (lane 6) interact with SAMHD1 was similar to that of WT Vpx (lane 2). The fact that these Vpx mutations did not affect the interaction with SAMHD1 suggests that they do not alter overall Vpx folding.

The DCAF1-binding defective Vpx proteins still accumulate in the nucleus

SAMHD1 is primarily found in the nucleus (Rice et al., 2009). Vpx is also known to be translocated to the nucleus (Hirsch et al., 1998); nuclear localization of Vpx may be required for efficient binding and Vpx-mediated degradation of SAMHD1. To determine whether Vpx mutants have a defect in nuclear targeting, we expressed WT Vpx, VpxW24A, VpxV29S and VpxH39A in transfected HEK293T cells. At 48 h post transfection, the cells were fixed and stained with a mouse anti-HA antibody followed by a secondary Texas Red anti-mouse antibody for examination by deconvolution microscopy. Consistent with previous results, Vpx could be detected in both the nucleus and the cytoplasm (Fig. 3). Nuclear localization of VpxW24A, VpxV29S and VpxH39A was also observed (Fig. 3). Thus, the fact that VpxW24A, VpxV29S and VpxH39A have a reduced ability to mediate SAMHD1 degradation cannot be explained by an inability to target to the nucleus.

Figure 3.

Cellular localization of Vpx and Vpx mutants. HEK293T cells were transfected with various Vpx constructs. Two days after transfection, cells were fixed, permeabilized, and stained with a mouse anti-HA antibody followed by a secondary Texas red-conjugated anti-mouse antibody. The nucleus was stained with DAPI. The cells were examined by deconvolution microscopy.

Cytoplasmic SAMHD1 shows higher resistance to Vpx-mediated degradation

The first 148 amino acids of SAMHD1 are sufficient for nuclear targeting (Rice et al., 2009). According to the PSORT-II program, amino acids 11–14 of SAMHD1 (KRPR) are predicted to contain a nuclear targeting signal, with K11 being a critical residue. As expected, SAMHD1K11A had a defect in nuclear targeting (Fig. 4A). A fusion protein containing mCherry–SAMHD1 was primarily nuclear (Fig. 4A), while mCherry alone was detected throughout the cell (Fig. 4A). On the other hand, mCherry–SAMHD1K11A was predominantly cytoplasmic (Fig. 4A). Vpx was able to induce efficient degradation of mCherry–SAMHD1 (Fig. 4B). However, of Vpx-induced degradation of mCherry–SAMHD1K11A was significantly impaired (Fig. 4B). These results were also confirmed by immunoblot analysis (Fig. 4C). Consistent with previous reports (Ahn et al., 2012; Lim et al., 2012; Laguette et al., 2012), we observed that the C-terminal region of SAMHD1 was important for Vpx-mediated degradation (Fig. 4B and C).

Figure 4.

Live cell imaging of mCherry–tagged SAMHD1 and SAMHD1 mutant proteins as well as Vpx-mediated SAMHD1 depletion.

A. Live cell imaging of mCherry, mCherry–SAMHD1, mCherry–SAMHD1K11A and mCherry–SAMHD1-1-547.

B. Live cell imaging of mCherry, mCherry–SAMHD1, mCherry–SAMHD1K11A and mCherry–SAMHD1-1-547 in the absence or in the presence of Vpx.

C. Influence of Vpx on the expression of mCherry–SAMHD1, mCherry–SAMHD1K11A and mCherry–SAMHD1-1-547, as revealed by immunoblotting with an anti-HA antibody. Vpx–HA was also detected with the anti-HA antibody.

The interaction with DCAF1 and SAMHD1 is not sufficient to promote Vpx-mediated degradation of SAMHD1

Amino acids in helix 1 of SIVmac Vpx (V29, I32, V36 and H39) that are important for DCAF1 binding are predicted to be present on the same face of the α-helix (Fig. 5). W24 of Vpx is predicted to be on the other face of helix 1. Unlike VpxV29S, VpxI32S, VpxA36S/V37S and VpxH39A, VpxW24A maintained the ability to interact with DCAF1 (Fig. 1C) and to accumulate in the nucleus (Fig. 3). However, VpxW24A was defective in inducing SAMHD1 degradation (Fig. 2B). It is possible that VpxW24A had lost the ability to interact with SAMHD1. We therefore transfected HEK293T cells with an SAMHD1–FLAG expression vector plus a control vector or a vector expressing WT or mutant VpxW24A–HFA, as indicated (Fig. 6A). Surprisingly, we found that VpxW24A could still interact with SAMHD1–FLAG (Fig. 6A, lane 3). These data suggest that in addition to SAMHD1 and DCAF1 binding, Vpx W24 may be involved in another functional activity that is required for Vpx-mediated degradation of SAMHD1.

Figure 5.

Vpx protein structure model based on homology modelling, as described in Experimental procedures. Models of Vpx were generated using both the program Modeller and the automated modelling server I-TASSER based on the HIV-1 Vpr template (PDB ID: 1M8L, chain A). Amino acids important for DCAF1 binding are shown in red. Trp24 is labelled in light blue.

Figure 6.

Characterization of the VpxW24A interaction with SAMHD1 and promotion of viral infectivity by Vpx mutants in myeloid cells.

A. Interaction of WT Vpx and VpxW24A proteins with SAMHD1. HEK293T cells were co-transfected with the SAMHD1–FLAG expression vector plus control vector (lane 1, VR), WT Vpx (lane 2), or VpxW24A (lanes 3). Cell lysates were prepared 48 h after transfection and immunoprecipitated with anti-HA antibody. The interaction of SAMHD1 with WT or mutant Vpx–HFA molecules was detected by immunoblotting with anti-HA antibody to detect Vpx–HFA and anti-FLAG antibody to detect SAMHD1–FLAG.

B. Promotion of viral infectivity by Vpx mutants in PMA-induced THP-1 cells. HEK 293T cells were co-transfected with the SIVmac239ΔENVΔVpx–GFP, pCMV coding for VSV-G, and the WT or mutant Vpx expression vector. Two days after transfection, virus was harvested. Virus infection of PMA-treated (100 ng ml−1 for 2 days) THP-1 cells was examined using an Olympus IX51 microscope 3 days after infection.

Vpx W24 as well as residues important for DCAF1 binding (V29, I32, H39) were found to be important for Vpx promotion of SIVΔVpx virus infectivity in PMA-treated THP-1 cells (Fig. 6B). Thus, DCAF1 recruitment is important for Vpx function in myeloid cells.

The DCAF1 binding motif is important for HIV-1 Vpr-induced G2 arrest

Sequence analysis of HIV-1 Vpr proteins indicated the presence of a Wx4Φx2Φx3AΦxH motif (Fig. 7A). To examine whether several conserved amino acids in this motif could be important for DCAF1 interaction, we constructed VprL22S/L23S and VprA30S/V31S mutants (Fig. 7A). We found that VprL22S/L23S and VprA30S/V31S mutants were indeed defective for DCAF1 binding. WT Vpr, VprL22S/L23S and VprA30S/V31S proteins were detected in the transfected HEK293T cells, and comparable levels of WT Vpr, VprL22S/L23S and VprA30S/V31S proteins were immunoprecipitated from the cell lysates (Fig. 7B). DCAF1 was efficiently co-immunoprecipitated with WT Vpr, but not with VprL22S/L23S or VprA30S/V31S (Fig. 7B). As expected, DCAF1 was also not co-immunoprecipitated with VprQ65R (lane 5), which has been reported to be defective for DCAF1 binding (Battey et al., 2007; Belzile et al., 2007; DeHart et al., 2007; Hrecka et al., 2007; Le Rouzic et al., 2007; Schrofelbauer et al., 2007; Wen et al., 2007). These data suggest that primate lentiviral Vpr and Vpx proteins contain a conserved DCAF1-binding motif.

Figure 7.

Conserved amino acids in the Wx4ΦxxΦx3AΦxH motif of HIV-1 Vpr are important for DCAF1 binding.

A. HIV-1 Vpr mutant constructs.

B. The Wx4ΦxxΦx3AΦxH motif mediates the HIV-1 Vpr–DCAF1 interaction. HEK293T cells were transfected with control vector (lane 1, VR), WT Vpr (lane 2) or one of the indicated Vpr mutants (lanes 3–5). Cell lysates were prepared 48 h after transfection and immunoprecipitated with anti-HA antibody. The interaction of DCAF1 with WT or mutant Vpr molecules was detected by immunoblotting with anti-HA antibody to detect Vpr-HA and anti-DCAF1 antibody to detect endogenous DCAF1.

We next examined the effect of DCAF1-binding mutants of Vpr in inducing G2 cell cycle arrest. Transfection of WT Vpr into HEK293T cells induced significant G2 cell cycle arrest when compared with cells transfected with the control vector (Fig. 8). On the other hand, VprL22S/L23S and VprA30S/V31S (Fig. 8) had an impaired ability to induce G2 arrest when compared with WT Vpr. We found that VprL22S/L23S and VprA30S/V31S were attenuated by 50–80% in terms of the induction of G2 arrest when compared with the WT Vpr. Together, these results suggest that the Wx4Φx2Φx3AΦxH motif in HIV-1 Vpr is important for the interaction of Vpr with DCAF1 and for Vpr function.

Figure 8.

Conserved amino acids in the Wx4ΦxxΦx3AΦxH motif of HIV-1 Vpr are important for Vpr-induced G2/M cell cycle arrest. HEK293T cells were transfected with control vector, WT Vpr, or one of the indicated Vpr mutants. The cells were harvested 48 h after transfection, fixed with ice-cold ethanol, then subjected to RNase A treatment and propidium iodide staining before flow cytometric analysis. The G2/G1 ratios were calculated by dividing the proportion of cells in G2/M by the proportion of cells in G1. Analysis of the flow cytometry data was carried out using FlowJo software.

Discussion

Previous research has shown that HIV-2 and SIV Vpx recruit Cul4–DDB1–DCAF1 E3 ubiquitin ligase to support viral replication (Sharova et al., 2008; Srivastava et al., 2008). In particular, Vpx induces the degradation of the newly identified host anti-viral factor SAMHD1 to promote viral replication in macrophages and dendritic cells (Hrecka et al., 2011; Laguette et al., 2011; 2012; Ahn et al., 2012; Lim et al., 2012). However, the molecular mechanism underlying the Vpx-mediated hijacking of Cul4–DDB1–DCAF1 E3 ubiquitin ligase is still unclear. We now report the identification of a novel DCAF1-binding motif (Wx4Φx2Φx3AΦxH) that is highly conserved among diverse HIV/SIV Vpx proteins. This motif is critical for the interaction of Vpx with DCAF1, and therefore for the assembly of viral Cul4–DDB1–DCAF1 E3 ubiquitin ligase complex (Fig. 1). This motif is also important for the SIVmac Vpx-mediated degradation of human SAMHD1 (Fig. 2) and the promotion of myeloid cell infection by SIV (Fig. 4).

This newly identified Wx4Φx2Φx3AΦxH motif (amino acids 24–39) of SIVmac Vpx is predicted to form an α-helix (helix 1). Amino acids in helix 1 of SIVmac Vpx that are important for DCAF1 binding (V29, I32, A36 and H39) are predicted to be on the same face of the helix (Fig. 5). Mutations of these residues did not affect Vpx's interaction with SAMHD1, suggesting that they did not alter the overall folding of Vpx. Furthermore, these mutant Vpx proteins could still be targeted into the nucleus (Fig. 3). Interestingly, Vpx may preferentially target nuclear SAMHD1 for degradation. SAMHD1 tagged with mCherry was detected mainly in the nucleus in live cells and was efficiently depleted in the presence of Vpx (Fig. 4). On the other hand, the cytoplasmic form of SAMHD1 mutant proteins (SAMHD1K11A) appeared to be more resistant to Vpx-mediated degradation (Fig. 4).

Vpx should contain at least two functional domains: one interacting with DCAF1 and another with target proteins such as SAMHD1. DCAF1 binding-defective Vpx mutants were also defective in inducing SAMHD1 degradation. Vpx also contains a functional domain upstream of helix 1 that could be involved in target protein interactions (Gramberg et al., 2010). We have observed that the extent of the Vpx-mediated degradation of SAMHD1 is not strictly proportional to Vpx's ability to bind DCAF1 or SAMHD1. For example, VpxW24A bound DCAF1 (Fig. 1) and SAMHD1 (Fig. 6) as efficiently as did WT Vpx. However, the degradation of SAMHD1 induced by VpxW24A was less efficient than that mediated by WT Vpx (Fig. 2). Thus, the interaction of Vpx with DCAF1 and SAMHD1 is necessary but not sufficient for the degradation of SAMHD1, suggesting that Vpx may not simply be a linker between the substrate SAMHD1 and the DCAF1–Cul4–DDB1 E3 ubiquitin ligase. It is still possible that VpxW24 contributes to an interaction between Vpx and DCAF1 or between Vpx and SAMHD1 that is not detected under our assay conditions. Alternatively, VpxW24 may be involved in an uncharacterized step that follows Vpx–DCAF1 or Vpx–SAMHD1 binding and is required for Vpx-induced polyubiquitination and/or degradation of SAMHD1. Vpx may be involved in positioning SAMHD1 in a unique orientation or in inducing a conformational change in SAMHD1 that makes it a better target for polyubiquitination and/or proteasome-mediated degradation. It is also possible that additional factor involved in the Vpx-induced ubiquitination of SAMHD1 may interact with Vpx and that its interaction with Vpx is affected by the W24A mutation.

The Wx4Φx2Φx3AΦxH motif is also highly conserved among diverse Vpr proteins from various HIV-1 and SIV strains. Mutations of conserved hydrophobic residues in this motif interfered with HIV-1 Vpr interaction with DCAF1. Furthermore, HIV-1 Vpr mutants in this motif had a reduced ability to induce G2 cell cycle arrest when compared with the WT Vpr. Thus, Vpx and Vpr presumably bind DCAF1 through some conserved features to promote substrate ubiquitination.

Our identification of a novel DCAF1-binding motif in Vpx and Vpr raises the possibility of the existence of additional DCAF1-binding proteins (cellular or viral) that could assemble with Cul4–DDB1–DCAF1 CRL4 E3 ubiquitin ligases through similar mechanisms. DCAF1 is essential for cell proliferation, DNA replication and embryonic development (McCall et al., 2008) and has been shown to be involved in the polyubiquitination of cellular proteins. However, the cellular targets of Cul4–DDB1–DCAF1-containing E3 ubiquitin ligases remain to be identified. The identification of a novel DCAF1-binding motif that recruits Cul4–DDB1–DCAF1 may be useful for the identification of other DCAF1-binding proteins.

Experimental procedures

Plasmid construction

SIVmac239 pVpx–HFA in the pCG vector was a gift from J. Skowronski. Plasmids pVpxW24A, pVpxV29S, pVpxE30A/E31A, pVpxI32S, pVpxN33A, pVpxE35A, pVpxA36S/V37S, pVpxH39A and pVpxQ76R were made from pVpx–HFA by site-directed mutagenesis. The HIV-1 Vpr expression vector (pVpr–HA) has been described previously (Battey et al., 2007). pVpr–L22S/L23S, pVpr–A30S/V31S and pVpr–Q65R were constructed from pVpr–HA by PCR-based site-directed mutagenesis. The human SAMHD1 expression vector was obtained from Changchun Boyi Biological Science and Technology Company. The coding sequence of SAMHD1 was amplified using the following primers: forward, 5′-GTCGACACCATGCAGCGAGCCGAT-3′ and reverse, 5′-TCTAGATCACTTGTCATCGTCGTCCTTGTAGTCCATTGGGTCATCTTTAA-3′, containing SalI and XbaI sites, respectively, and a C-terminal FLAG tag. The PCR product was cloned into VR1012 to generate SAMHD1–FLAG. SAMHD1–HA was amplified with the following primers: forward, 5′-GTCGACACCATGCAGCGAGCCGAT-3′ and reverse, 5′-TCTAGATCAGGCGTAATCTGGAACATCGTATGGGTACATTGGGTCATCTTTAA-3′, containing SalI and XbaI sites, respectively, and a C-terminal HA tag. The PCR product was cloned into VR1012 to generate SAMHD1–HA. To generate an expression vector encoding mCherry–SAMHD1–mCherry fusion protein, the SAMHD1–HA fragment was digested with SalI and XbaI and cloned into pmCherry–C1 to generate pmCherry–SAMHD1–HA. pmCherry–SAMHD1K11A–HA was generated by PCR-based site-directed mutagenesis and its sequence confirmed.

Antibodies and cell culture

The following antibodies were used: anti-HA monoclonal antibody (MAb, Covance, MMS-101R), anti-Vprbp (DCAF1, Shanghai Genomics, SG4220-28), anti-FLAG M2 antibody (Sigma, F1804), anti-β-tubulin monoclonal antibody (Covance, MMS-410P), anti-histone H3 (Genscript, A01502) and anti-actin monoclonal antibody (Sigma, A3853). HEK293T cells (AIDS Research Reagents Program) were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum and penicillin/streptomycin. Human monocytic cell line THP-1 (ATCC) cells were cultured in RPMI medium with 10% FBS. All cultured cell lines were maintained at 37°C in a humid atmosphere containing 5% CO2.

Transfection, co-immunoprecipitation and Immunoblotting

DNA transfection was carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. HEK293T cells were harvested at 48 h after transfection, washed twice with cold PBS, and lysed in lysis buffer [150 mM Tris, pH 7.5, with 150 mM NaCl, 1% Triton X-100 and complete protease inhibitor cocktail tablets (Roche)] at 4°C for 30 min, then centrifuged at 10 000 g for 30 min. For haemagglutinin (HA) tag immunoprecipitation, precleared cell lysates were mixed with anti-HA antibody-conjugated agarose beads (Roche, catalogue number 190–119) and incubated at 4°C for 3 h or overnight. Samples were then washed eight times with washing buffer (20 mM Tris, pH 7.5, with 100 mM NaCl, 0.1 mM EDTA and 0.05% Tween 20). The beads were eluted with elution buffer (0.1 M glycine-HCl, pH 2.0). The eluted materials were then analysed by SDS-PAGE and immunoblotting with the appropriate antibodies as previously described (He et al., 2008; Zhang et al., 2012).

Immunofluorescent staining

HEK293T cells were transfected with various Vpx expression plasmids, passed onto coverslips 24 h after transfection, and allowed to reattach to the coverslip for another 24 h. The cells were then fixed with 4% paraformaldehyde, permeabilized with 0.5% Triton, and blocked with 10% BSA. HA-tagged Vpx variants were stained with a mouse anti-HA antibody (Sigma, H9658) followed by a Texas Red-conjugated anti-mouse antibody. The cell nucleus was stained with DAPI and examined by deconvolution microscopy using a Nikon i90 microscope.

Viral particle production and infection

SIVmac239ΔENVΔVpx–GFP was a gift from J. Skowronski and has been previously described (Srivastava et al., 2008). To generate Vpx-containing VLPs, HEK 293T cells were co-transfected with the SIVmac239ΔENVΔVpx–GFP, pCMV coding for VSV-G, and the WT or mutant Vpx expression vector. Two days after transfection, cell culture medium was harvested. Cell debris was removed by centrifugation at 9838 g for 5 min and stored at −80°C. Virus input was adjusted using THP-1 as the target cells. Virus infection of PMA-treated (100 ng ml−1 for 2 days) THP-1 cells was examined using an Olympus IX51 microscope 3 days after infection.

Cell cycle analysis

HEK293T cells were harvested and washed with PBS at 48 h post transfection. Cells were resuspended in 1% paraformaldehyde at room temperature for 60 min, then fixed with 95% ethanol at 4°C overnight. Following RNase A treatment, propidium iodide staining (2 × 106 cells ml−1), and incubation in the dark on ice for 30 min, cells expressing the internal membrane-anchored GFP were analysed for DNA content using FACSCailbur flow cytometry (Becton Dickinson Company). At least 10 000 GFP-positive cells were analysed for their distribution in the various phases of cell cycle using Modfit software.

Modelling

Homology modelling of SIVmac Vpx was carried out as previously described (Stanley et al., 2008) using both the program Modeller and the automated modelling server I-TASSER, which has been ranked as the best server in the recent CASP7 and CASP8 modelling contests (Battey et al., 2007). Both Modeler and I-TASSER produced very similar models, suggesting convincing modelling results. Modelling by Modeller was carried out according to the following steps: prior to the modelling, multiple sequence alignment was carried out for HIV-1 Vpr and SIVmac Vpx. The available structures of HIV-1 Vpr (Schuler et al., 1999; Morellet et al., 2003) were then used to validate and improve the alignment manually. Subsequently, the aligned sequences of Vpr and Vpx were input into SWISS-MODEL for homology modelling in alignment mode using the structure of Vpr (PDB ID: 1M8L, chain A) as template. Confidence in the homology modelling was imparted by (i) the high degree of sequence homology between Vpx and the template protein Vpr and (ii) the similarity of the resulting models produced by Modeller and I-TASSER. The figures were prepared with Pymol (The PyMOL Molecular Graphics System, Version 1.1, Schrödinger, LLC)

Acknowledgements

We thank Drs Jacek Skowronski and Mario Stevenson for critical reagents, Juan Du, Ke Zhao, Ying Guo and Chunyan Dai for technical assistance, and Dr Deborah McClellan for editorial assistance. This work was supported in part by funding from the Chinese Ministry of Science and Technology (No. 2012CB911100), the Chinese Ministry of Education (No. IRT1016), and the Key Laboratory of Molecular Virology, Jilin Province (No. 20102209), China.

Ancillary