Fig 1.
Construction of the SIVmac239 and SIVmac316 X-del and X-Q76A Vpx mutants.
(A) The codon (TCA) for the second residue (serine) of SIV Vpx was changed to a stop codon (TAA) to generate the X-del Vpx mutant. This change does not alter the amino acid sequence of the overlapping Vif protein. In the X-Q76A Vpx mutant, amino acid 76 (glutamine) was changed to alanine by altering the CAA codon to GCA. (B) The intracellular expression of Vpx proteins and their incorporation into virions was evaluated by transfecting HeLa cells with WT and Vpx mutant full-length SIV plasmids. Forty-eight hours later, cell lysates or cell-free culture supernatants, pelleted through 20% sucrose, were evaluated by immunoblotting using anti-SIV plasma, anti-Vpx antibodies, and anti-tubulin antibodies.
Fig 2.
The replication of Vpx defective SIV mutants is attenuated in rhesus macaque mononuclear cells.
ConA-activated rhesus PBMCs (A and B), rhesus MDM (C and D) and SupT1-R5 cells (E and F) were infected with WT and Vpx mutant SIV stocks produced in transfected 293T cells. The indicated cell cultures were infected with equivalent amounts of virus inocula, based on particle-associated 32P -RT activity (approximately 5 × 106 cpm) and progeny virion production was monitored by measuring the RT activity released into the culture medium. The results in panels A and B, are shown as mean +/—s.e.m (n = 4). The parametric unpaired t test was performed using PRISM software. The significant p values (* p<0.05, ** p<0.01) for WT versus the X-del and X-Q76A Vpx mutants refer to the single time points at peak virus production. A representative result from at least two experiments is shown in panels C and D.
Fig 3.
The presence of wild type Vpx correlates with reduced levels of endogenous SAMHD1 in ConA-activated macaque CD4+ T lymphocytes.
CD8+ T cell-depleted rhesus PBMCs were activated with ConA for 24 h, cultured with IL-2 containing medium for 48 h, negatively selected for CD4+ T cells, and infected with SIVmac239 (MOI = 0.2). Cells and supernatant media were collected daily from the infected or mock-infected “enriched” CD4+ T lymphocyte cultures. (A) Progeny virions released into the medium were detected by 32P RT assay. (B) Levels of endogenous SAMHD1 present in SIVmac239 or mock infected cell extracts from 3 x 105 cells were determined by immunoblotting following polyacrylamide gel electrophoresis using anti-SAMHD1 antibody. GAPDH was used as a loading control. (C) Individual CD4+ T lymphocyte cultures were infected with WT SIVmac239, WT SIVmac316, or corresponding Vpx derived mutants (MOI = 0.2). Based on the results shown in Panels A and B, cells were harvested on day 3 PI and levels of endogenous SAMHD1 present in virus or mock infected cell extracts were determined by immunoblotting. (D) SAMHD1 phosphorylation status in macaque CD4+ T lymphocytes was determined by electrophoresis in acrylamide gels with/without Phos-Tag and analyzed by immunoblotting. Whole-cell extracts (from 3 × 105 cells) of non-activated and ConA-activated CD4+ T lymphocytes from two uninfected animals were separated on acrylamide gels with/without Phos-Tag and analyzed by immunoblotting using anti-SAMHD1 and GAPDH antibodies.
Fig 4.
WT SIVmac239 degrades endogenous SAMHD1 in memory CD4+ T cells during the acute infection of rhesus macaques.
(A) Levels of endogenous SAMHD1 expression were examined by immunoblotting rhesus macaque naïve and memory CD4+ T cells in blood and spleen, and myeloid lineages (CD14+ cells, MDM, and alveolar macrophages) or in two control human T cell leukemia lines (PM1 and SupT1-R5). Whole-cell extracts (from 3 × 105 cells/lane) from each indicated source were separated on 10% acrylamide gels and stained using anti-SAMHD1 antibody. GAPDH was used as a loading control. The numbers below each lane indicate relative densitometric intensities of SAMHD1 bands relative to that of GAPDH and normalized to that present in PM1 cells. (B) SAMHD1 phosphorylation status in macaque memory CD4+ T lymphocytes was determined by electrophoresis in acrylamide gels with Phos-Tag and analyzed by immunoblotting. Whole-cell extracts (3 × 105 cells) from PM1 cells, freshly collected rhesus CD4+ T cells or sorted rhesus memory CD4+ T cells from blood or spleen were separated on acrylamide gels with Phos-Tag and analyzed by immunoblotting using anti-SAMHD1 antibody. (C) Levels of SAMHD1 in sorted rhesus memory CD4+ T cells infected with VSV-G pseudotyped WT SIVmac239, SIVmac239 X-del or SIVmac239 X-Q76A were assessed by immunoblotting using anti-SAMHD1 antibody. (D) Acute phase viremia in two macaques inoculated intravenously with 10,000 TCID50 of SIVmac239. (E) Memory CD4+ T cells were prepared by cell sorting from PBMC or spleen from an uninfected or an infected rhesus macaque at day 9 PI following IV inoculation with 10,000 TCID50 of SIVmac239. The plasma viral load on day 9 post infection in the macaque inoculated with WT SIVmac239 was 7.8 x 107 RNA copies/ml; viral RNA in the uninfected animals were below levels of detection (< 100 RNA copies/ml). Whole-cell extracts from 3 x 105 sorted memory CD4+ T cells from each source or from PM1 cells were analyzed by immunoblotting using anti-SAMHD1 and anti-GAPDH antibodies.
Fig 5.
SIVs deficient in degrading SAMHD1 in memory CD4+ T cells exhibit an attenuated replication phenotype in inoculated rhesus macaques.
Rhesus macaques were inoculated intrarectally with 1 x 104 TCID50 of SIVmac239 WT or SIVmac239 X-Q76A derivatives (A and C) or 1 x 103 TCID50 of SIVmac316 WT or the SIVmac316X-Q76A derivatives (B and D). The infectious virus titers in the inocula were determined by end-point dilution using SAMHD1 negative SupT1-R5 cells to avoid suppressive effects of SAMHD1 restriction. Plasma viral copies/ml are shown in panel A and C. Memory CD4+ T cell counts/μl are shown in panel C and D. Black curves: WT virus; blue curves: putative revertant Vpx mutants; red curves: non-revertant Vpx mutants. (E) Amino acid substitutions present in the starting Q76A Vpx mutant virus or in the putative revertant virus populations present in the plasmas of macaques K42 and JAX4 at week 35 PI, based on SGA (see Fig 6) are shown. The locations of the three helical domains of SIVmac Vpx are indicated.
Fig 6.
Alignment of Vpx amino acid sequences amplified from the plasma of monkeys inoculated with SIV Vpx X-Q76A mutants.
SGA plus sequencing was used to generate vpx gene sequences present in the plasma of (A) SIVmac239 X-Q76A infected animals (K42, K2M, JWR, and JHL) and (B) SIVmac316 X-Q76A infected animals (JA4X, DX39, JLP, and K31) at week 35 PI. The sequence of WT SIVmac Vpx is shown at the top and the animal identifications are indicated on the left. Amino acids highlighted in red represent changes conferring putative revertant phenotypic changes. Amino acid substitutions in black are not thought to contribute to revertant phenotype because when present alone and in the absence of a change in the starting Q76A Vpx mutation (e.g. in the K2M virus), they fail to restore WT properties. The nucleotide changes corresponding to the putative revertant amino acid substitutions at Vpx residue 76 or residue 32 are shown at the bottom of each panel.
Table 1.
Infection of rhesus macaques with WT SIVmac and Q76A Vpx SIVmac mutants.
Fig 7.
Analyses of Vpx revertant viruses emerging in macaques K42 and JA4X.
ConA-activated rhesus PBMCs were infected at a MOI = 0.01 with (A) the WT SIVmac239, SIVmac239 X-Q76A Vpx mutant, or the virus recovered at week 40 PI from infected macaque K42 or (B) the WT SIVmac316, SIVmac316 X-Q76AVpx mutant, or the virus recovered at week 40 PI from infected macaque JA4X. Virus replication was monitored by measuring the particle-associated 32P RT activity released into culture supernatants. Results are shown as mean +/- s.e.m. The parametric unpaired t test was performed using PRISM software. The significant p values (* p<0.05, ** p<0.01) for putative revertant viruses compared to the starting X-Q76A Vpx mutant refer to the indicated time points during the infection. (C) ConA- activated enriched rhesus CD4+ T lymphocytes were infected with the indicated viruses at MOI = 0.2. Whole cell lysates were prepared on day 3 PI and levels of endogenous SAMHD1 present in virus or mock infected cell extracts from 3 x 105 cells/lane were determined by immunoblotting as described in Fig 4C. The numbers below each lane indicate relative densitometric intensities of SAMHD1 bands relative to that of GAPDH and normalized to that present in Mock infected cells. (D) Rhesus PBMC were infected with the WT SIVmac316 or the molecularly cloned SIVmac316 derivatives containing the X-Q76A, Q76S, or the Q76A/I32T Vpx substitutions, using virus stocks prepared in transfected 293T cells. Cell cultures were infected with equivalent amounts of virus inocula, based on particle-associated RT activity (approximately 5 × 106 32P cpm). Progeny virion production is shown as mean +/—s.e.m. The parametric unpaired t test was performed using PRISM software. The significant p values (* p<0.05) for SIVmac316 X-Q76S or SIVmac316 X-Q76A/I32T versus the starting SIVnac316 X-Q76A Vpx mutant are indicated. Representative results from at least two experiments are shown.
Fig 8.
Mapping of Vpx mutations at the Vpx-DCAF1-SAMHD1 interface.
(A) Crystal structure (PDB ID 4CC9) of the c-terminal domain of SAMHD1 (green) in complex with WT Vpx (blue) and DCAF1 (salmon). (B) Zoomed-in view displaying the interactions of Gln 76Vpx at the Vpx-DCAF1 interface. Models of the Q76A Vpx mutation (C), the Q76S Vpx substitution (D) in the revertant recovered from macaque K42, and the Q76A/I32T Vpx double substitution in the revertant recovered from macaque JAX4 (E). Hydrogen bonds are indicated by dashed lines and the red spheres denote a crystallographic water molecule observed the structure.