Defining the interactions and role of DCAF1/VPRBP in the DDB1-cullin4A E3 ubiquitin ligase complex engaged by HIV-1 Vpr to induce a G2 cell cycle arrest

Abstract

HIV viral protein R (Vpr) induces a cell cycle arrest at the G2/M phase by activating the ATR DNA damage/replication stress signalling pathway through engagement of the DDB1-CUL4A-DCAF1 E3 ubiquitin ligase via a direct binding to the substrate specificity receptor DCAF1. Since no high resolution structures of the DDB1-DCAF1-Vpr substrate recognition module currently exist, we used a mutagenesis approach to better define motifs in DCAF1 that are crucial for Vpr and DDB1 binding. Herein, we show that the minimal domain of DCAF1 that retained the ability to bind Vpr and DDB1 was mapped to residues 1041 to 1393 (DCAF1 WD). Mutagenic analyses identified an α-helical H-box motif and F/YxxF/Y motifs located in the N-terminal domain of DCAF1 WD that are involved in exclusive binding to DDB1. While we could not identify elements specifically involved in Vpr binding, overall, the mutagenesis data suggest that the predicted β-propeller conformation of DCAF1 is likely to be critical for Vpr association. Importantly, we provide evidence that binding of Vpr to DCAF1 appears to modulate the formation of a DDB1/DCAF1 complex. Lastly, we show that expression of DCAF1 WD in the absence of endogenous DCAF1 was not sufficient to enable Vpr-mediated G2 arrest activity. Overall, our results reveal that Vpr and DDB1 binding on DCAF1 can be genetically separated and further suggest that DCAF1 contains determinants in addition to the Vpr and DDB1 minimal binding domain, which are required for Vpr to enable the induction of a G2 arrest.

Figure 1. Delineation of the minimal domain of DCAF1 that interacts with HIV-1 Vpr and DDB1.

A. Schematic representation of Myc-DCAF1 WT (1-1507), Myc-DCAF1 WD (1041-1393) and Myc-DCAF1 1377 (1041-1377). The different domains of DCAF1 with their amino-acid positions are highlighted. Additionally, the region targeted by the full length DCAF1-specific bp3 siRNA is highlighted in red (see below). B-C. HEK293T cells were mock-transfected (lanes 1 and 2) or transfected with Myc-DCAF1 (1-1507) (lanes 3 and 4), Myc-DCAF1 WD (1041-1393) (lanes 5 and 6) or with Myc-DCAF1 1377 (1041-1377) (lanes 7 and 8) -encoding plasmids in the presence of empty vector (lanes 1, 3, 5 and 7) or HA-tagged Vpr-expressing plasmid (lanes 2, 4, 6, and 8). Total amounts of DNA were adjusted with empty vector so that similar quantities of plasmids were transfected in each sample. B. Immunoprecipitations using anti-Myc antibody were performed on cell extracts using protein-A sepharose beads. The levels of HA-Vpr, endogenous DDB1, Myc-DCAF1 proteins and actin were monitored in cell extracts as well as, when applicable, in immunoprecipitated fractions by Western Blot using specific antibodies. C. Quantitation of DDB1 binding efficiency. Band signals corresponding to DDB1 in immunocomplexes were scanned by laser densitometry. The ratio of DDB1 signal over that of precipitated Myc-DCAF1 1507 or Myc-DCAF1WD was calculated and expressed as the percentage of that obtained in the absence of Vpr, which was assigned a value of 100%. Error bars indicate the standard error of the mean (SEM) from the quantitative analysis of three independent experiments. Statistical analysis was performed as described in the Experimental Procedures (p<0.05; ns; non significant). D. Immunoprecipitation using anti-HA antibody was performed on cell extracts using anti-HA antibody-coupled agarose beads. The levels of HA-Vpr, endogenous DDB1, endogenous DCAF1, Myc-DCAF1 proteins and actin were monitored in cell extracts as well as, when applicable, in immunoprecipitated fractions by Western Blot using specific antibodies. The data shown here are representative of results obtained in three independent experiments. * denotes the light chain of the IgG used for immunoprecipitation. # represents non-specific immunoprecipitated proteins. E. Structural and molecular features of the DCAF1 WD minimal domain. Consensus secondary structure prediction of DCAF1 WD 1041-1393 was generated using the PSI-PRED server and structural data obtained from the 3D modelization. Orange lines highlight the predicted β-sheet structures while the green line and the green amino-acid residues highlight α-helices and the putative H-box motif (see below), respectively. The F/YxxF/Y repeats are highlighted in purple whereas the WDxR motifs are highlighted in red.

Figure 2. Mutations in the putative H-box motif of DCAF1 abrogate DDB1 binding without affecting Vpr interaction.

A. Sequence alignment of the H-box motifs of indicated cellular DCAF’s and viral proteins. HBX and WHX indicate viral protein X from Hepatitis B virus and Woodchuck Hepatitis virus, respectively, while SV5-V represents viral protein V encoded by paramyxovirus simian virus 5. Residues of the H-box motif previously reported to contact DDB1 are boxed. Alignment adapted from Li & al. . The asterisks indicate the position of substitution mutation (L1054 and R1057). B. HEK293T cells were mock-transfected (lanes 1 and 2) or transfected with Myc-DCAF1 WD (lanes 3 and 4), Myc-DCAF1 WD L1054P (lanes 5 and 6) or with Myc-DCAF1 WD R1057E (lanes 7 and 8) -encoding plasmids in the presence of empty vector (lanes 1, 3, 5 and 7) or HA-tagged Vpr-expressing plasmid (lanes 2, 4, 6, and 8). Total amounts of DNA were adjusted with empty vector so that similar quantities of plasmids were transfected in each sample. Immunoprecipitations were performed on cell extracts using anti-Myc antibodies. The levels of HA-Vpr, endogenous DDB1, Myc-DCAF1 WD (WT and mutants) and actin were monitored in cell extracts as well as, when applicable, in immunoprecipitated fractions by Western Blot using specific antibodies. * denotes the light chain of the IgG used for immunoprecipitation. C. Quantitation of DDB1 and HA-Vpr binding to Myc-DCAF1 WD. The percentage of DDB1 or HA-Vpr bound to Myc-DCAF1 WD was determined by evaluating the ratio of DDB1 or HA-Vpr band signal over that of Myc-DCAF1 WD WT or mutant in the immunoprecipitated fractions. Ratios obtained with Myc-DCAF1 WD WT were assigned a value of 100%. Error bars indicate the SEM from the quantitative analysis of at least 3 independent experiments. Statistical analysis was performed as described in the Experimental Procedures (p<0.05; ns: non significant).

Figure 3. Mutation of DCAF1 WD WDxR motifs differentially affects Vpr and DDB1 binding.

A. HEK293T cells were mock-transfected (lanes 1, 2) or transfected with Myc-DCAF1 WD (lanes 3 and 4), Myc-DCAF1 WD R1247A (lanes 5 and 6), Myc-DCAF1 WD R1283A (lanes 7 and 8) or with Myc-DCAF1 WD R1247A/R1283A (lanes 9 and 10)-encoding plasmid in the presence of empty vector (lanes 1, 3, 5, 7, and 9) or HA-Vpr expressing plasmid (lanes 2, 4, 6, 8 and 10). Immunoprecipitations and Western Blot detection were performed as described in figure 2B. * denotes the light chain of the IgG used for immunoprecipitation. # represents non-specific immunoprecipitated proteins. B. Quantitation of DDB1 and HA-Vpr binding to DCAF1 WD (WT or mutants). Quantitation was determined as described in figure 2C.

Figure 4. Effect of mutations in the N-terminal F/YxxF/Y motifs of DCAF1 on Vpr and DDB1 binding.

A. HEK293T cells were mock-transfected (lanes 1 and 2) or transfected with Myc-DCAF1 WD (lanes 3 and 4), Myc-DCAF1 WD F1060A/Y1063A (lanes 5 and 6), Myc-DCAF1 WD F1077A/F1080A (lanes 7 and 8) or with Myc-DCAF1 WD Y1120A/F1123A (lanes 9 and 10)-encoding plasmids in the presence of empty vector (lanes 1, 3, 5, 7 and 9) or HA-Vpr-expressing plasmid (lanes 2, 4, 6, 8 and 10). Immunoprecipitations and Western Blot detection were performed as described in figure 2B. * denotes the light chain of the IgG used for immunoprecipitation. B. Quantitation of the DDB1 and HA-Vpr binding. Quantitation was determined as described in figure 2C.

Figure 5. The FDKF motif at position 1255-1258 is not required for efficient recruitment of Vpr.

A. HEK293T cells were mock-transfected (lanes 1 and 2) or transfected with Myc-DCAF1 WD (lanes 3 and 4), Myc-DCAF1 WD F1255A/F1258A at two different concentrations (lanes 5 to 8), or with Myc-DCAF1 WD F1077A/F1080A (lanes 9 and 10)-encoding plasmid in the presence of empty vector (lanes 1, 3, 5, 7 and 8) or HA-Vpr-expressing plasmids (lanes 2, 4, 6, 8, and 10). Immunoprecipitations and Western Blot detection were performed as described in figure 2B. * denotes the light chain of the IgG used for immunoprecipitation. # represents non-specific immunoprecipitated proteins. B. Quantitation of the DDB1 and HA-Vpr binding. Quantitation was determined as described in figure 2C.

Figure 6. DCAF1 does not solely act as a bridge to engage Vpr to the DDB1-CRL4A E3 ubiquitin ligase and induce G₂ cell cycle arrest.

A-B. HEK293T cells were mock-transfected (lanes 2 and 3) or transfected with HA-Vpr-expressing plasmid (lanes 4 and 5) or transfected with DCAF1 bp3R (DCAF1 1-1507 bp3 siRNA resistant) (lanes 6 and 7), Myc-DCAF1 WD (lanes 8 and 9) or Myc-DCAF1 1377 (lanes 10-11)-encoding plasmids in the presence of HA-Vpr-expressing plasmid. All cells were also transfected with a plasmid encoding GFP and treated with either non-targeting siRNA (NT siRNA) (lanes 2, 4, 6, 8 and 10) or specific DCAF1 bp3 siRNA (bp3 siRNA) (lanes 3, 5, 7, 9 and 11). Non-transfected HEK293T cells were used as negative control (lane 1). A. Non-transfected or transfected HEK293T cells were lysed in 0.5% Triton lysis buffer and subjected to anti-HA or anti-Myc immunoprecipitation and further resolved on SDS-PAGE. The level of HA-Vpr, actin, exogenous and endogenous DCAF1, endogenous DDB1, Myc-DCAF1 WD and Myc-DCAF1 1377 were monitored in cell extracts as well as in the immunoprecipitated fractions by Western Blot using specific antibodies. * denotes the light chain of the IgG used for immunoprecipitation. # represents non-specific immunoprecipitated proteins. B. Cell cycle profile of transfected cells (GFP+) analyzed in A. G₂/M:G₁ ratios were determined using the Modfit Software. C. The graph depicts the mean G₂/M:G₁ ratios obtained in two independent experiments. Errors bars represent SEM.