Fig 1.
Interaction between TASOR and DCAF1 is stabilized in the presence of Vpx.
(A) HA-Vpx WT or DCAF1 binding-deficient HA-Vpx Q76R proteins were co-expressed with TASOR-Flag in HeLa cells, then an anti-HA immunoprecipitation was performed. (B) TASOR-Flag and HA-Vpx WT were co-expressed in HeLa cells, then an anti-Flag immunoprecipitation was performed. (C) SAMHD1-Flag and HA-Vpx WT were co-expressed in HeLa cells, then an anti-Flag immunoprecipitation was performed. (D) HeLa cells were treated with either siRNA CTL or siRNA DCAF1. After 24h, HA-Vpx WT and TASOR-Flag were co-expressed for 48h, then an anti-HA immunoprecipitation was performed. In each panel, the indicated proteins were revealed by western blot. For this figure, each shown immunoblot is representative of at least 3 independent experiments.
Fig 2.
TASOR PARP-like domain is involved in DCAF1 binding.
(A) Schematic representation of HA-tagged TASOR or Flag-tagged TASOR constructs. 1–1670: TASOR long isoform. 1–1512: TASOR short isoform. 1–931: N-terminal fragment of TASOR. 630–1512: C-terminal fragment of TASOR short isoform. WT: TASOR (short isoform). ΔPARP: TASOR (short isoform) deleted of its PARP-like domain (106–319 aa). (B) Indicated HA-TASOR constructions were expressed in HeLa cells, then an anti-HA immunoprecipitation was performed. (C) Flag-Vpx WT and indicated HA-TASOR constructs were co-expressed in HeLa cells, then an anti-Flag immunoprecipitation was performed. (D) TASOR-Flag WT or TASOR-ΔPARP-Flag were co-expressed with HA-Vpx WT in HeLa cells, then an anti-Flag immunoprecipitation was performed. In each panel, the indicated proteins were revealed by western blot. For this figure, each shown immunoblot is representative of at least 3 independent experiments.
Fig 3.
Vpx Q47AV48A loss of activity in J-Lat A1 T cells results from the V48A mutation.
(A and B) HIV-2.Gh1 Vpx WT or indicated mutants were tested for TASOR degradation (A) and viral reactivation in J-Lat A1 T cells (B). J-Lat A1 T cells were treated with Vpx-containing VLPs. After overnight treatment with TNF-α, cells were analyzed by flow cytometry and whole-cell extracts by western blot. For each mutant, reactivation assay was performed at least 3 times and the immunoblot shown is representative of at least 3 independent VLP productions, SD is shown. (C) HIV-2.Gh1 Vpx WT or indicated mutants were tested for SAMDH1 degradation. Non-differentiated THP-1 cells were treated for 24h with VLPs and whole-cell extracts were analyzed by western blot. The shown immunoblot is representative of at least 3 independent VLP productions. Empty: VLP in which Vpx is not incorporated. QV: Vpx Q47A-V48A double mutant. (D and E) HA-Vpx WT or indicated mutants were co-expressed along with TASOR-Flag in HeLa cells, then an anti-HA (D) or anti-Flag (E) immunoprecipitation was performed. shown immunoblots are representative of 3 independent experiments.
Fig 4.
The integrity of a set of Vpx exposed residues from α-helix 1 and 2 and the C-ter tail is required for HUSH antagonism.
(A) The representation of the crystal structure of CtD-huDCAF1/Vpx SIVsmm/CtD-huSAMDH1 ternary complex, resolved by Schwefel et al in 2014 [42] (PDB: 4CC9), has been adapted here to highlight Vpx residues that we have tested in this study regarding SAMHD1 and TASOR degradations. Top: sequence alignment of HIV-2.Gh1 Vpx and SIVsmm Vpx. Our study is dedicated to HIV-2 Vpx, while the structure was obtained with SIVsmm Vpx. The two sequences share 79.81% identity and 92% similarity. Green and Yellow marks indicate residues of SIVsmm Vpx involved in DCAF1 and SAMDH1 binding respectively, according to Schwefel et al.[42] Stars: interaction mediated by the lateral chain. Dot: Interaction mediated by the principal chain. The integrity of residues shown in red is important for HUSH degradation, while this is not the case for residues in blue. Residues in bold dark are important for HUSH and SAMDH1 degradation. Bottom: Two different views of the CtD-huDCAF1/Vpx SIVsmm/CtD-huSAMDH1 complex, resolved by Schwefel et al in 2014[42] (PDB: 4CC9). The huDCAF1 and huSAMHD1 C-terminal domains are shown in green and yellow, respectively. Vpx SIVsmm is shown in light pink. Tested residues involved (red) or not (blue) in HUSH antagonism are indicated. (B) HIV-2.Gh1 Vpx WT and mutants were tested for TASOR degradation and viral reactivation. J-Lat A1 T cells were treated with VLPs. After overnight treatment with TNF-α, HIV-1 LTR-driven GFP expression was analyzed by flow cytometry and whole-cell extracts by western blot. For each mutant, reactivation assay was performed 2, 3 or more times and shown immunoblots are representative of at least 2 independent experiments. Empty: VLP in which Vpx is not incorporated. QV: Vpx Q47A-V48A double mutant. RQV: Vpx R42A-Q47A-V48A triple mutant. (C) HIV-2.Gh1 Vpx WT and mutants were tested for SAMDH1 degradation. Non-differentiated THP-1 cells were treated 24h with VLPs and whole-cell extracts were analyzed by western blot. For each mutant, shown immunoblots are representative of 2 independent experiments.
Fig 5.
Vpx R34A-R42A, which induces SAMHD1 but not TASOR degradation, is characterized by a reduced binding affinity for DCAF1.
(A) HA-Vpx WT or indicated mutants and TASOR-Flag were co-expressed in HeLa cells, then an anti-HA immunoprecipitation was performed. (B) Graph representing Vpx binding affinity to DCAF1. Co-immunoprecipitated DCAF1 and immunoprecipitated HA-Vpx (Vpx WT (n = 6), Vpx R34A (n = 6), R42A (n = 6) and RR (n = 4)) were quantified from independent experiments. Co-immunoprecipitated DCAF1 signal over immunoprecipitated Vpx signal ratios were calculated and reported to Vpx WT (ratio 1). (C) TASOR-Flag and HA-Vpx WT or indicated mutants were co-expressed in HeLa cells, then an anti-Flag immunoprecipitation was performed. Immunoblot is representative of at least 3 independent experiments. (D and E) HIV-2.Gh1 Vpx WT and mutants were tested for viral reactivation (D) and TASOR degradation (E). J-Lat A1 T cells were treated with VLPs. After overnight treatment with TNF-α, HIV-1-LTR-driven GFP expression was analyzed by flow cytometry and whole-cell extracts were analyzed by western blot. For each mutant, reactivation assay was tested at least 3 times (SEM is shown), and immunoblots are representative of at least 3 independent experiments. (F) HIV-2.Gh1 Vpx WT and mutants were tested for SAMDH1 degradation, THP-1 cells were treated with VLP overnight and then whole-cell extracts were analyzed by western blot. RR: Vpx double mutant R34A-R42A. For each mutant, immunoblots are representative of 3 independent experiments.
Fig 6.
Vpx R34A-R42A and Vpx R42A-Q47A-V48A mutants are impaired in inducing TASOR, but not SAMHD1, degradation in macrophages.
Long (A) and short (B) kinetics of TASOR and SAMHD1 degradations by HIV-2.Gh1 Vpx WT or mutants brought by VLP in Monocyte-derived-Macrophages (MDM). Purified monocytes from an healthy donor were differenced 7 days with GM-CSF and M-CSF. After differentiation, MDM were transduced with indicated Vpx-containing VLPs and harvested at indicated times. Whole cell extracts were analyzed by western-blot. QV; Vpx Q47A-V48A double mutant. RQV: Vpx R42A-Q47A-V48A triple mutant. RR: Vpx R34A-R42A double mutant. For this figure, TASOR and SAMHD1 degradation were tested in tree independent donors (see S7 Fig.).
Fig 7.
(A) TASOR antagonism. In the absence of Vpx, TASOR can be found in association with DCAF1. TASOR PARP-like domain is involved in this interaction. By co-binding to TASOR and to DCAF1, Vpx WT would create another indirect contacting point between TASOR and DCAF1, increasing the interaction affinity between TASOR and DCAF1. Structural conformation changes would trigger the degradation of TASOR by the whole ubiquitin ligase complex and TASOR poly-ubiquitination. Vpx RR or Vpx RQV mutants bind TASOR but present a lower affinity for DCAF1, which is represented by a black line; in turn, the interaction affinity between TASOR and DCAF1 is not sufficient to orientate the whole ubiquitin ligase complex in order to trigger TASOR degradation. Of note, whether TASOR-DCAF1 and TASOR-Vpx interactions are direct or not is unknown. (B) SAMHD1 antagonism. SAMDH1 does not interact with DCAF1 in the absence of Vpx. Vpx bridges the two cellular proteins, allowing the recruitment of the whole ubiquitin ligase complex, SAMHD1 ubiquitination and degradation. Vpx RR or RQV are both able to induce SAMHD1 degradation, despite their apparent low binding affinity for DCAF1.