The dNTP triphosphohydrolase activity of SAMHD1 persists during S-phase when the enzyme is phosphorylated at T592

Abstract

SAMHD1 is the major catabolic enzyme regulating the intracellular concentrations of DNA precursors (dNTPs). The S-phase kinase CDK2-cyclinA phosphorylates SAMHD1 at Thr-592. How this modification affects SAMHD1 function is highly debated. We investigated the role of endogenous SAMHD1 phosphorylation during the cell cycle. Thr-592 phosphorylation occurs first at the G1/S border and is removed during mitotic exit parallel with Thr-phosphorylations of most CDK1 targets. Differential sensitivity to the phosphatase inhibitor okadaic acid suggested different involvement of the PP1 and PP2 families dependent upon the time of the cell cycle. SAMHD1 turn-over indicates that Thr-592 phosphorylation does not cause rapid protein degradation. Furthermore, SAMHD1 influenced the size of the four dNTP pools independently of its phosphorylation. Our findings reveal that SAMHD1 is active during the entire cell cycle and performs an important regulatory role during S-phase by contributing with ribonucleotide reductase to maintain dNTP pool balance for proper DNA replication.

Keywords: Deoxynucleotide metabolism; Sterile alpha motif and HD domain containing protein1 (SAMHD1); cell cycle.

Figure 1.

dNTP pools in different phases of the cell cycle in SAMHD1-KO and parental THP-1 cells. A. cell proliferation curve and cell cycle analysis of THP-1 cells knock-out for SAMHD1 (KO – dashed line) and the parental control (wt – solid line). Bars: Mean ± standard error for n = 3 values. B. Representative image of the cell cycle distribution of THP1 cells before (A = asynchronous population) and after centrifugal elutriation (G1, S and G2/M enriched sub-populations). C. Immunoblotting for SAMHD1, RNR subunits and cell cycle markers in each enriched sub-population of wt and KO THP-1 cells. Beta-actin: loading control. SAMHD1 was detected using a mouse polyclonal antibody D. Phosphorylated SAMHD1 was detected after electrophoresis in a phosgel using an antibody directed against pT592 (α-pSAMHD1) and a rabbit polyclonal antibody recognizing both the phosphorylated and the non-phosphorylated form (α-SAMHD1). Before electrophoresis, parallel samples were treated with (+) or without (-) lambda phosphatase (lambda PP). E. Comparison of dNTP pool sizes in wt and KO THP-1 cells in the indicated phases of the cell cycle. In the individual phases the level of each dNTP was calculated as detailed in Experimental procedures. The fold increase of each dNTP in KO cells relative to the wt is reported for each phase of the cycle. Bars: Mean ± standard deviation for n = 6 values. F. The composition of the dNTP pools in each phase of the cycle for wt and KO cells, evaluated from the ratios between pmoles of dATP, dTTP or dCTP and pmoles of dGTP under each condition.

Figure 2.

Phosphorylation status of SAMHD1 during cell cycle progression in normal human fibroblasts (A) and transformed cells (B). A. Cycling lung fibroblasts were synchronized by double thymidine block and released. Confluent fibroblasts were maintained for 48 h in medium with 10 or 0.1% serum. The top panel shows the cell cycle profiles at the indicated times after release and at confluence. The lower panels show the expression levels of SAMHD1 detected by an antibody directed against phosphoT592 (pSAMHD1) and a mouse monoclonal antibody recognizing both the phosphorylated and non- phosphorylated forms of SAMHD1 (SAMHD1). Cyclin E, A2 and B were used as cell cycle markers and beta-actin as a loading control. B. Whole cell extracts of HeLa cells synchronized by double thymidine block were analyzed for expression of SAMHD1 with the same mouse monoclonal recognizing both forms of SAMHD1. The phosphorylated form appears as a slower migrating band (empty arrow) compared to the non-phosphorylated form (black arrow). The cell cycle markers cyclin E, cyclin A and histone H3-S10-phosphorylation are those analyzed in the same experiment and data for these markers was published in Figure 2B of Kara et al [25].

Figure 3.

Cyclin E/CDK2 phosphorylates SAMHD1 in vitro. A. In vitro interaction between SAMHD1 and cyclin/CDK complexes. Top panel: schematic representation of GST-SAMHD1 constructs (GST in grey, SAMHD1 in white) used for in vitro pull down assays with cyclin E or cyclin A2. The full-length GST-SAMHD1 or the indicated fragments were bound to Glutathione agarose resin and then incubated with in vitro translated [S³⁵]-labeled cyclin E or cyclin A2. Beads were isolated and bound proteins were separated by gel electrophoresis and visualized by autoradiography or Coomassie staining. B. In vitro kinase assay. Purified recombinant his- tagged wild type SAMHD1 (wt) or the non-phosphorylatable mutant T592A were incubated with recombinant cyclin A2/CDK2 or cyclin E/CDK2 in the presence of 1μCi [γ-³²P]-ATP. Reactions were separated by SDS-PAGE and visualized by autoradiography or Coomassie staining.

Figure 4.

T592 phosphorylation favors SAMHD1 interaction with Skp2. U2OS cells were transiently transfected with a plasmid coding for GFP-tagged wild type SAMHD1 (wt) or non-phosphorylatable T592A SAMHD1 or the empty vector (GFP). After 48h cells were lysed and proteins immunoprecipitated (IP) from detergent extracts via the GFP tag. SAMHD1 and the indicated proteins were detected in immune-complexes and extracts by immunoblotting. Red Ponceau is used as loading control for protein extracts and shows the immune-complexes after IP.

Figure 5.

Turn-over of SAMDH1 in proliferating U2OS cells. A. The expression of GFP-tagged SAMHD1 was induced in U2OS cells stably transfected with a tetracyclin-inducible vector by treatment with different doses of tetracycline (tet) for 24 h and analyzed by immunoblotting. The lower panel reports a representative immunoblot of endogenous SAMHD1 (Endo-SAMHD1) and ectopically expressed GFP-SAMHD1. B. After tet removal the decline of the induced mRNA was followed by RT-PCR. The level of mRNA is reported as fold increase relative to that of not induced transfected cells (dashed line). C. The amount of GFP-SAMHD1 was determined by immunoblotting using antibodies against SAMHD1 and GFP. Representative immunoblots of total GFP-SAMHD1 (α-SAMHD1) and phosphorylated GFP-SAMHD1 (α-pSAMHD1) in samples induced for 24 h with 1 µg/ml tet and chased in the absence of tet for the indicated times. Beta actin: loading control. NI = not induced. D. Densitometric analysis was performed for GFP-SAMHD1, normalized for beta-actin and expressed as percentage relative to the protein level at 0 time. Bars in A.B. and D: Mean ± standard error, n = 3.

Figure 6.

Kinetics of SAMHD1 dephosphorylation during mitotic exit in hRPE-1 cells. A. hRPE-1 cells were synchronized in prometaphase by 16 h nocodazole treatment (100 ng/ml), mitotic cells were detached by gentle shake-off and seeded in fresh medium. Samples were collected without nocodazole synchronization (A), immediately after shake-off (M) and after 2 h in fresh medium (early G1) and analyzed by immunoblotting. SAMHD1 was detected by a mouse monoclonal antibody recognizing both the phosphorylated (empty arrow) and the non-phosphorylated form (black arrow) as indicated by a band shift. Cyclin A2, cyclin B and R2, the S-phase induced small subunit of ribonucleotide reductase, are used as cell cycle markers. GAPDH: loading control. B. Schematic diagram of the protocol used to synchronize mitotic exit in hRPE-1 cells. Samples were collected in prometaphase and during mitotic exit at the indicated time points after RO3306 addition. The proteasome inhibitor MG132 was present 15 min before RO3306 addition and during mitotic exit. C. hRPE-1 cells were synchronized as in B and whole cell extracts were immunoblotted for cyclin B, securin, phospho-Histone H3(Ser10) (pH3S10), pThr-CDK substrates, pSer-CDK substrates and beta-actin (loading control). In the same experiment SAMHD1 phosphorylation was analyzed as in A. by electrophoresis in a phosgel using the antibody against phospho T592 (pSAMHD1) after pre-treatment in the presence (+) or absence (-) of lambda phosphatase (PP). Empty arrow: pSAMHD1, black arrow non-phosphorylated SAMHD1. Decay of protein phosphorylation was evaluated from the relative intensities of pSAMHD1, pThr-CDK substrates and pH3S10 normalized for beta-actin and the level of phosphorylation at 0 min, taken as 100%. Bars: Mean ± standard error for n = 4 values. D. hRPE-1 cells were synchronized as in B. After RO3306 addition one set of cultures was treated with proteasome inhibitor MG132 (RO3306 +MG132). Samples were collected at 0–5-15–30 min during mitotic exit and immunoblotted for cyclin B, securin, pThr-CDK substrates, pSer-CDK substrates and beta-actin (loading control). SAMHD1 was analyzed as described in C. *non-specific band

Figure 7.

Okadaic acid treatment maintains SAMHD1 phosphorylation in normal and transformed cells. Okadaic acid was added to hRPE-1 cells during synchronous mitotic exit (A-C), and to asynchronous populations of hRPE-1 (D) or THP-1 cells (E). A. Schematic diagram of the experiment for hRPE-1 cells. Prometaphases of nocodazole synchronized cells were incubated with okadaic acid (OKA) before RO3306 addition and during mitotic exit in the presence (+MG132) or absence (-MG132) of proteasome inhibitor MG132. Samples were collected at the indicated times. B. hRPE-1 cells were synchronized and treated with 400 or 1600 nM OKA as indicated in A. Whole cell extracts were analyzed by immunoblotting using the antibody against phospho T592 (pSAMHD1) after pre-treatment with (+) or without (-) lambda phosphatase (PP) and electrophoresis in a phosgel. C. The mitotic markers (pThr-CDK substrates, cyclin B, securin, pH3S10) and loading control (beta-actin) were analyzed in parallel. D. asynchronous proliferating hRPE-1 cells and E. THP-1 cells were treated with OKA as indicated. Whole cell extracts were analyzed for pThr-CDK substrates or for CK2alpha kinase as positive controls for OKA treatment. SAMHD1 was detected either by an antibody recognizing both the phosphorylated (empty arrow) and the non-phosphorylated form (black arrow) or by anti pSAMHD1 antibody after electrophoresis in a phosgel after pretreatment with (+) or without (-) lambda phosphatase (PP).