The retroviral restriction ability of SAMHD1, but not its deoxynucleotide triphosphohydrolase activity, is regulated by phosphorylation

Abstract

SAMHD1 is a cellular enzyme that depletes intracellular deoxynucleoside triphosphates (dNTPs) and inhibits the ability of retroviruses, notably HIV-1, to infect myeloid cells. Although SAMHD1 is expressed in both cycling and noncycling cells, the antiviral activity of SAMHD1 is limited to noncycling cells. We determined that SAMHD1 is phosphorylated on residue T592 in cycling cells but that this phosphorylation is lost when cells are in a noncycling state. Reverse genetic experiments revealed that SAMHD1 phosphorylated on residue T592 is unable to block retroviral infection, but this modification does not affect the ability of SAMHD1 to decrease cellular dNTP levels. SAMHD1 contains a target motif for cyclin-dependent kinase 1 (cdk1) ((592)TPQK(595)), and cdk1 activity is required for SAMHD1 phosphorylation. Collectively, these findings indicate that phosphorylation modulates the ability of SAMHD1 to block retroviral infection without affecting its ability to decrease cellular dNTP levels.

Figure 1. Phosphorylation State of SAMHD1 in Cycling and Noncycling Cells

(A) Endogenously expressed SAMHD1 was immunoprecipitated using anti-SAMHD1 antibodies from PMA-treated (noncycling) or untreated (cycling) human monocytic THP-1 cells. Proteins were separated by SDS-PAGE, and a Coomassie-blue-stained band containing the SAMHD1 protein was subjected to phosphopeptide mapping by mass spectrometry. The ratios of phosphorylated to unphosphorylated peptides are shown. The ability of the different cells to restrict HIV-1 is presented. Similar analysis was performed for determining the phosphorylation state of the endogenously expressed SAMHD1 in human MDMs. In parallel, PMA-treated or untreated human monocytic U937 cells stably expressing SAMHD1-FLAG were used to immunoprecipitate and determine the phosphorylation state of SAMHD1, as described in Experimental Procedures. Mass-spectrometry analysis was performed three times, and a representative experiment is shown. See also Figure S1. (B) Wild-type (WT) human SAMHD1 protein is depicted showing the numbers of the amino acid residues at the boundaries of each domain. The residue T592 is depicted in the consensus motif ⁵⁹²TPQK⁵⁹⁵ for recognition and phosphorylation by cdk1.

Figure 2. Phosphorylation State of SAMHD1 in Different Human Cell Lines and Primary Cells

(A) 293T or HeLa cells transfected with a plasmid expressing SAMHD1-FLAG or pLPCX were lysed 24 hr after transfection and treated or not with λPP, which is a protein phosphatase with activity toward phosphorylated serine, threonine, and tyrosine residues. Protein samples were separated by SDS-PAGE containing Phos-tag (+Phos-tag), a ligand that shifts the mobility of phosphorylated proteins, and analyzed by western blotting using FLAG antibodies. We also analyzed the phosphorylation state of endogenously expressed SAMHD1 proteins using anti-SAMHD1 antibodies in different cells: human THP-1 cells, human primary MDMs, human primary resting CD4⁺ T cells, and replicating CD4⁺ T cells. Protein samples were also separated in SDS-PAGE gels without Phos-tag (–Phos-tag). (B) Similarly, we analyzed the phosphorylation state of SAMHD1-FLAG stably expressed in U937 cells in the presence or absence of PMA. As a loading control, cell lysates were western blotted using GAPDH antibodies. (C) The phosphorylation state of SAMHD1 in resting and replicating CD4⁺ T cells from two donors was analyzed by using a specific antibody that recognizes a phosphorylated SAMHD1 at position T592 (anti-phospho-T592-SAMHD1). As a control, we analyzed the samples by western blotting using anti-SAMHD1 and anti-GAPDH antibodies. Similar results were obtained in three independent experiments, and a representative experiment is shown.

Figure 3. Ability of SAMHD1 Phosphorylation Variants to Restrict HIV-1

(A) Human 293T cells were transfected with wild-type or the indicated SAMHD1 variant. Cells were lysed 24 hr after transfection, and protein samples were separated using SDS-PAGE gels containing Phostag (+Phos-tag). SAMHD1 variants were detected by western blotting using anti-FLAG antibodies. The indicated sample was treated with λPP for detecting the migration of the unphosphorylated SAMHD1. In parallel, we performed a similar analysis using the empty vector pLPCX. Protein samples were also separated in SDS-PAGE gels without Phos-tag (–Phos-tag). As a loading control, cell lysates were western blotted using GAPDH antibodies. (B) The different SAMHD1 variants expressed in 293T cells were also analyzed by western blotting using anti-phospho-T592-SAMHD1 antibodies, which only recognize a phosphorylated SAMHD1 protein at residues T592. (C) Human monocytic U937 cells were stably transduced with wild-type SAMHD1 and the indicated SAMHD1 variant. PMA-treated stable cells were analyzed for SAMHD1 expression by western blotting using anti-FLAG antibodies. Similarly, western blot analysis of GAPDH was used as loading control. (D and E) PMA-treated human monocytic U937 cells stably expressing the indicated SAMHD1 variants were challenged with increasing amounts of HIV-1-GFP. Infection is shown as the percentage of GFP-positive cells 48 hr postinfection measured by flow cytometry. As a control, U937 cells stably transduced with the empty vector pLPCX were challenged with increasing amounts of HIV-1-GFP. See also Figure S2. Experiments were performed in triplicate, and a representative result is shown. WT, Wild-type.

Figure 4. Role of the N-Terminal Residues 1–112 in the Ability of Phosphorylation to Regulate Retroviral Restriction

(A) Human 293T cells were transfected with the indicated SAMHD1 variant. Cells were lysed 24 hr after transfection, and protein samples were separated using SDS-PAGE gels containing Phos-tag (+Phos-tag). SAMHD1 variants were detected by western blotting using anti-FLAG antibodies. Protein samples were also separated in SDS-PAGE gels without Phos-tag (–Phos-tag). As a loading control, cell lysates were western blotted using GAPDH antibodies. (B) Human monocytic U937 cells were stably transduced with wild-type, and the indicated SAMHD1 variants were analyzed for SAMHD1 expression by western blotting using anti-FLAG antibodies. Western blot analysis of GAPDH was used as a loading control. (C) Human monocytic U937 cells stably expressing the different SAMHD1 variants were challenged with increasing amounts of HIV-1-GFP. Infection is shown as the percentage of GFP-positive cells 48 hr postinfection as measured by flow cytometry. As a control, U937 cells stably transduced with the empty vector pLPCX were challenged with increasing amounts of HIV-1-GFP. See also Figure S3. Experiments were performed in triplicate, and a representative result is shown.

Figure 5. Analysis of Enzymatic Activity of the Different SAMHD1 Variants

(A) Quantification of dATP, dTTP, and dGTP levels from PMA-treated U937 cells expressing the indicated SAMHD1 variants was performed with a primer-extension assay, as described in Experimental Procedures. Similar results were obtained in three separate experiments, and a representative experiment is shown. (B) Immunoprecipitated SAMHD1 variants from human 293T cells were normalized by western blotting using anti-FLAG antibodies. WT, wild-type; WB, western blot; IP, immunoprecipitation. (C) Thin-layer chromatography analysis of the dTTP-triphosphohydrolase activity of the different immunoprecipitated SAMHD1 variants. For this purpose, we incubated radio-labeled α-[³²P]dTTP in the presence of the indicated SAMHD1 variants. Products from the α-[³²P]dTTP hydrolysis were separated by thin-layer chromatography using polyethyleneimine cellulose. Hydrolysis of α-[³²P]dTTP yields dT and α-[³²P]PP, which are visualized by using a phosphoimager. As a control, we also measured the dTTP-triphosphohydrolase activity in the presence of the dsRNA analog ISD-PS. The results of three independent enzymatic reactions per treatment are shown. (D) Recombinant SAMHD1 and SAMHD1-T592D proteins were purified from baculovirus-infected Sf9 insect cells. Proteins were separated by SDS-PAGE gels and stained with Coomassie blue. (E) Similarly, the dTTPase activity of baculovirus recombinant SAMHD1 and SAMHD1-T592D proteins was determined by measuring the hydrolysis of radio-labeled α-[³²P]dTTP of the indicated recombinant SAMHD1 protein. Reactions were stopped at the indicated times and separated by thin-layer chromatography using polyethyleneimine cellulose, as described in Experimental Procedures. See also Figure S4. A representative result of three independent experiments is shown.

Figure 6. Involvement of cdk1 in the Phosphorylation of SAMHD1

(A) Effect of the Cdk1 dominant-negative mutant D146N (Cdk1-D146N) on the phosphorylation levels of SAMHD1. Cell lysates from human 293T cells transfected with plasmids expressing the SAMHD1-FLAG and Cdk1-D146N-HA proteins were separated by SDS-PAGE containing Phostag (+Phos-tag). SAMHD1-FLAG and Cdk1-D146N-HA were detected by western blotting using anti-FLAG and anti-HA antibodies, respectively. As a control, the levels of SAMHD1 phosphorylation were determined in the presence of the empty vector pCMV. Protein samples were also separated in SDS-PAGE gels without Phos-tag (–Phos-tag). Loading was normalized by western blotting using GAPDH antibodies. Similar results were obtained in three independent experiments, and a representative experiment is shown. (B) In vitro phosphorylation of recombinant GST-SAMHD1 purified from bacteria by cdk1 complex, which was purified from baculovirus-infected cells using γ-[³²P]ATP as a phosphate donor. As a positive control, we incubated the cdk1 kinase complex with the human histone H1, which is a known substrate for this kinase complex. As a negative control, we incubated a similar amount of purified GST protein with the kinase complex cdk1.