Cyclin D3-dependent control of the dNTP pool and HIV-1 replication in human macrophages

Abstract

Cyclins control the activation of cyclin-dependent kinases (CDK), which in turn, control the cell cycle and cell division. Intracellular availability of deoxynucleotides (dNTP) plays a fundamental role in cell cycle progression. SAM domain and HD domain-containing protein 1 (SAMHD1) degrades nucleotide triphosphates and controls the size of the dNTP pool. SAMHD1 activity appears to be controlled by CDK. Here, we show that knockdown of cyclin D3 a partner of CDK6 and E2 a partner of CDK2 had a major impact in SAMHD1 phosphorylation and inactivation and led to decreased dNTP levels and inhibition of HIV-1 at the reverse transcription step in primary human macrophages. The effect of cyclin D3 RNA interference was lost after degradation of SAMHD1 by HIV-2 Vpx, demonstrating the specificity of the mechanism. Cyclin D3 inhibition correlated with decreased activation of CDK2. Our results confirm the fundamental role of the CDK6-cyclin D3 pair in controlling CDK2-dependent SAMHD1 phosphorylation and dNTP pool in primary macrophages.

Introduction

The size of the intracellular deoxynucleotide triphosphate (dNTP) pool was described in the early 90s as a major limiting factor for HIV-1 reverse transcription (RT).^1,2 Drugs capable of decreasing the dNTP pool size were proposed as therapeutic alternatives.³

SAM domain and HD domain-containing protein 1 (SAMHD1) is an enzyme with phosphohydrolase activity,^4,5 identified as a major restriction factor for virus infections, including HIV-1.^6–12 SAMHD1-induced restriction may be counteracted by viral protein X (Vpx) encoded by SIV or HIV subtype-2 (HIV-2), which triggers its degradation by the ubiquitin system.⁶ Samhd1^−/− mice show increased intracellular levels of dNTP.^13,14

The dNTP pool size is regulated along the cellular division, and SAMHD1 has been shown to play a key role, adjusting the intracellular concentration of dNTP to the requirements of DNA replication. ¹⁵ SAMHD1 activity is also regulated through the progression of the cell cycle and HIV-1-induced restriction by SAMHD1 is controlled by its phosphorylation. Cyclin-dependent kinases (CDK) have been proposed as the ones directly phosphorylating SAMHD1.^16-19 We have shown that CDK6 upstream of CDK2 controls SAMHD1 phosphorylation function in primary lymphocytes and macrophages,^19-21 whereas factors modulating cell cycle progression, such as CDKN1A (p21) may control HIV-1 replication through SAMHD1.²²

According to the classical model of cell cycle progression, CDK4/6-cyclin D complex is essential for cell cycle initiation and differentiation of haematopoietic cells, leading to retinoblastoma protein phosphorylation (Rb), release of E2F transcription factor and allowing expression of genes required for cell cycle progression.^23-25 In contrast to D-type cyclins, cyclin E is regulated by E2F, and in complex with its partner CDK2, controls DNA replication at the G1 to S phase transition. Cyclin A is induced at the beginning of S phase, and has both CDK1 and CDK2 as catalytic partners. Cell growth prior to cell division is restricted by the activity cyclin B1/CDK1 complexes.²⁶

The identity of the cyclin/CDK complex responsible for SAMHD1 phosphorylation is a relevant issue to clearly pinpoint its mode and time of action in restricting virus replication and its potential role as antiviral factor.¹⁰ Here, we have identified cyclin D3 as a major contributor to SAMHD1 phosphorylation and HIV-1 restriction, confirming the role of CDK6 on SAMHD1-mediated restriction.

Results

Evaluation of cyclin and CDK expression and overlap in monocyte derived macrophages

Cyclins and CDKs may compensate and/or affect the expression of other cyclins/CDK in proliferating cells.²⁷ Differentiation of monocytes with M-CSF induces cells to enter the cell cycle as measured by Ki67 staining and phosphorylation (inactivation) of SAMHD1.¹⁹ Therefore, we first evaluated the effect of cyclin knockdown in the expression of other cyclins and/or CDK in primary macrophages from at least 5 independent donors (Fig. 1). In general, we have found that interference of cyclins associated to early steps (G0, G1) in the cell cycle (D-and E-type cyclins) increased cyclin B2 expression whereas knockdown of later cyclins (E1, A2, B1 and B2) upregulated cyclin D3, associated to G0 to G1 transition. Interference of cyclin D3 induced the dowregulation of cyclin E2 and A2, associated to CDK2. Interference of D2 and D3 cyclins induced the downregulation of CDK1 and downregulation of D1 induced the upregulation of CDK4 and CDK6. This effect was not observed after interference of cyclins D2 and D3. These results may indicate that D-type cyclins are the most important cyclins in controlling CDK expression associated to HIV-1 restriction in primary macrophages, namely, CDK1, CDK2 and CDK6.

Figure 1.

Cell cycle cyclin and CDK mRNA expression in M-CSF monocyte derived macrophages. (A) Relative mRNA levels of indicated cyclins or CDK were measured by 2-step quantitative RT-PCR and normalized to GAPDH mRNA expression using the ΔΔCt method. Mean ± SD of at least 5 independent donors performed in duplicate were calculated. Bold numbers indicate the corresponding mRNA result from each interfered cyclin. Values were considered significant and therefore colored (green, upregulated and red downregulated) if below or above one standard deviation from the mean. (B) Different phases of the cell cycle are driven by specific cyclins and their corresponding CDK. The figure depict the cyclins commonly associated to each phase of the cell cycle and the effect on cyclin or CDK expression of downregulating each cyclin through RNA interference as evaluated in (A).

RNA interference of cyclins D-type cyclins and cyclin E2 significantly impaired HIV-1 replication in macrophages

mRNA of cyclins involved in cell cycle progression was successfully downregulated (>60%) (Fig. 2A) in primary monocyte-derived macrophages (MDM) as measured by quantitative PCR (qPCR) in the absence of cell toxicity (Fig. 2B). Macrophages were then infected with a VSV-pseudotyped, GFP-expressing NL4–3 virus and infection was measured by flow cytometry 72 h later. RNA interference of D-type cyclins significantly (P < 0.05) reduced HIV-1 infection, being cyclin D3 the one showing the strongest effect (81.8±6.9% inhibition) (Fig. 2C). In addition, downregulation of cyclin E2, a partner of CDK2, also reduced HIV-1 infection (p = 0.0017; 66.1±12.2% inhibition) under conditions in which the reverse transcriptase (RT) inhibitor AZT (1 μM) completely blocked HIV-1 infection (98.1±1.5% inhibition). The effect on HIV-1 replication was associated to the blockade of viral DNA formation (Fig. 2D) in which both cyclin D3 and E2 showed the strongest effect, similar to that of AZT but unaffected by the integrase inhibitor raltegravir, indicating that the inhibition of HIV-1 replication was dependent on an early step (i.e. reverse transcription) of the HIV-1 replication cycle. The effect of siRNA targeting D3 was confirmed by an additional siRNA (D3#2) (Fig. 2E).

Figure 2.

Efficient RNA interference of cell cycle related-cyclins and inhibition of HIV-1 replication. (A) A panel of siRNA were transfected in MDM and mRNA of the corresponding cyclin was measured by quantitative PCR and normalized to GAPDH expression. mRNA levels are shown compared to a non-transfected sample (UNF); NT: non-targeting siRNA. (B) Primary cells were gated as living or dead, according to flow cytometry FSC and SSC parameters. Mean ± SD of 4 independent donors are shown. (C) Transfected MDM were infected with a VSV-pseudotyped, GFP-expressing HIV-1 and infection measured 72 h later by flow cytometry and expressed as the percentage to non-transfected cells. Mean ± SD of 4 independent donors are shown. Statistical significance was calculated by one-sample t test. p-values>0.05 are not shown. (D) Total viral DNA formation after 16 h infection with HIV-1 BaL of macrophages transfected with the indicated siRNA or treated with AZT (1 µM) or raltegravir (Ralt; 2 µM) or the corresponding cyclin siRNA. Data represent Mean±SD of at least 2 donors. (E) The effect of siRNA targeting cyclin D3 was confirmed by an additional siRNA. Data represent Mean±SD of at least 2 donors. Statistical significance was calculated by one-sample t test. p-values > 0.05 are not shown. Ralt:raltegravir, NT: non-targeting.

Cyclin D3 downregulation reduced SAMHD1 phosphorylation, viral DNA formation, decreased the intracellular dNTP pool and restricted HIV-1 infection in a SAMHD1-dependent manner

Western blot analysis showed that 2 independent siRNA successfully targeting cyclin D3 (Fig. 3A) decreased SAMHD1 phosphorylation at Threonine 592 (T592) when compared with non-transfected MDM (Mock) or MDM treated with a non-targeting siRNA (siNT) (Fig. 3A). Importantly, siRNA mediated knockdown of cyclin D3 led to a decrease in CDK2 activation, measured as the phosphorylation of Threonine160 (T160; Fig. 3B), providing support to the role of CDK2 in phosphorylating SAMHD1 in primary macrophages.^19-21 Macrophages treated with siRNA targeting cyclin A2 or cyclin D3 were lysed as indicated in Materials and Methods and dNTP concentration was quantified (Fig. 3B). In concordance with the observed impairment in the phosphorylation of SAMHD1, cyclin D3-depleted MDM showed altered levels of measured dNTPs. dTTP and dATP showed the highest degree of reduction when compared to non-transfected MDM (Mock), MDM transfected with a non-targeting siRNA (siNT) or siRNA targeting cyclin A2, while dCTP and dGTP concentrations were only modestly inhibited, indicating that dNTP modulation is different for each of the nucleotides in macrophages lacking cyclin D3.

Figure 3.

RNA interference of Cyclin D3 blocks SAMHD1 phosphorylation and CDK2 activation. (A) Western blot of lysates from non-transfected macrophages (Mock) or transfected with siRNA targeting cyclin D3 or a non-targeting sequence (siNT) showing. Representative blots of one of 3 independent donors are shown. Hsp90 was used as loading control. (B) The deoxy-nucleotide pool (dNTP pool) of MDM transfected or not (Mock) with siRNA targeting cyclin D3 or cyclin A2 and non-targeting sequence (siNT) as controls, was measured as indicated in Materials and Methods. Data represent mean and SD of 3 independent donors. Statistical significance was calculated using the unpaired, 2-tailed Student's t test.

Treatment of MDM with viral-like particles carrying the SIV protein Vpx (VLP_Vpx) induced the degradation of SAMHD1 (Fig. 4A). As expected, SAMHD1 degradation by Vpx was accompanied by an increase in HIV-1 replication (Fig. 4B). Importantly, in MDM treated with VLP_Vpx, knockdown of cyclin D3 did not have a significant effect on HIV-1 replication when compared to untreated MDM (Fig. 4B and C), suggesting that the effect of cyclin D3 interference on HIV-1 infection is dependent on SAMHD1.

Figure 4.

Cyclin D3 effect on HIV replication is dependent on SAMHD1 expression. (A) Macrophages transfected with a non-targeting siRNA (siNT) or a siRNA targeting cyclin D3 were treated (VLP_Vpx) or not (No VLP_Vpx) with viral-like particles carrying Vpx. Cells were lysed 24 h after the treatment and samples were subjected to SDS-PAGE followed by immunoblotting with the indicated antibodies. Representative blots of one out of 2 independent experiments are shown. (B) Representative dot plots of macrophages treated with the indicated siRNA or AZT (1 µM). Macrophages were treated (bottom panels) or not (top panels) with VLP_Vpx and were infected with a GFP-expressing HIV-1 and infection assessed by flow cytometry 72 h later. Dot plots correspond to a representative donor out of 3. (C) As in (A), siRNA-transfected macrophages were treated (right panel) or not (left panel) with viral-like particles carrying Vpx and then infected with GFP-NL4–3. Infection was measured 72 h later by flow cytometry and expressed as the percentage to non-transfected cells. Mean ± SD of 3 independent donors are shown. Statistical significance was calculated by one-sample t test. p-values >0.05 are not shown.

Discussion

The exact identification of host genes that affect susceptibility and resistance to HIV is key to unravel HIV-host interactions susceptible to drug and vaccine development.²⁸

We have addressed the role of cell cycle cyclins in mediating SAMHD1-controlled restriction of HIV-1 in order to further investigate the role of CDK in controlling HIV-1 restriction mediated by SAMHD1.

The binding of specific regulatory subunits (cyclins) during the cell cycle is required for the activity of CDK. Cyclins are synthesized and destroyed at specific times during the cell cycle, regulating CDK activity²⁷ and thus, pinpoint the specific time in which an event (e.g. SAMHD1 inactivation) occurs during the cell cycle. Throughout the cell cycle, expression/degradation of specific cyclins should lead to a coordinated degradation/expression of other cyclins, leading to progression of the cell cycle. Indeed, we have found that knockdown of each specific cyclin led to up or down-regulation of the expression of other cyclins or CDK (Fig. 1), adding complexity to the understanding of the control of SAMHD1 function during cell cycle progression. Our work may not allow drawing definitive conclusions of the multiple interactions between cyclins and their corresponding CDK, representing an intricate pattern of proteins that moreover can vary between cell types, but instead provides a glimpse on how the cell cycle events may be regulated in monocyte-derived macrophages. Importantly, results show that interference of cyclins associated to different stages of the cell cycle (G1 to M) in non-stable primary human cells led to an increase in D-type cyclins, particularly D3, associated to G0 to G1 transition and suggestive of an arrest or even a reset of the cell cycle. Gene silencing of cell cycle-related cyclins on immortalized cell lines like human embryonic kidney (HEK293T) cells did not significantly affect its abnormal rapid proliferation and phosphorylation-related events, thus globally not affecting HIV-1 restriction mediated by SAMHD1 regulation (data not shown). Major cell cycle-derived functions can be executed by a single CDK in immortalized cell lines in order to drive cell division in an aberrant context even in the absence of specific cyclins, therefore evidencing the convenience of a primary model when studying cellular proliferation and SAMHD1-mediated HIV-1 restriction.

RNA interference of the cyclin D family significantly inhibited HIV-1 replication with D3 having the strongest impact on HIV-1 replication. Moreover, cyclin D3 knockdown decreased SAMHD1 phosphorylation, reduced intracellular levels of dNTP and blocked viral DNA formation. Blockade of HIV-1 infection was lost when SAMHD1 was absent following HIV-2 Vpx expression, demonstrating the specificity of the effect. We also found that RNA interference of cyclin D3 led to reduced levels of CDK2 activation, therefore slowing or even arresting cell cycle progression at G1 phase. Although cyclin D3 function on HIV-1 replication could not be pharmacologically validated due to unavailable specific compounds, our previous observations showing that CDK6 interference or its pharmacological inactivation also reduced CDK2 activation,^19,21 as well as the increase on CDK2 phosphorylation and activation after the interference of the cell-cycle inhibitor p21²² strongly suggest that cyclin D3/CDK6 complex may play a key role in controlling SAMHD1 phosphorylation upstream of CDK2. CDK2 has been found inactive in lymphocytes from cyclin D3 knockout (Ccnd3^−/−) mice,²⁹ suggesting the need for cyclin D3/CDK6 for CDK2-dependent cell cycle progression.

In addition, our initial screening found that RNA interference of cyclin E2 in primary human macrophages led to HIV-1 restriction. While cyclins A and B are known to complex with CDK1 and CDK2 kinases, cyclin E2 has CDK2 as unique partner at the G1 and S phases. This result could reinforce the idea that CDK2 and not CDK1 is the kinase that phosphorylates SAMHD1 in primary macrophages¹⁹ leading to its inactivation. Nevertheless, we cannot rule out that another cyclin may act as a compensatory mechanism. In immortalized HEK293T cells CDK2 and cyclin A2 have been shown in complex with SAMHD1.²⁰ It has been also described that RNA interference of CDK1, CDK2 and cyclin A2 partially affected SAMHD1 phosphorylation in HEK293T cells.²⁰ However, we have shown that CDK1 downregulation did not affect SAMHD1 phosphorylation in primary cells,¹⁹ again suggesting that cell cycle regulation and dNTP control significantly differs between immortalized cell lines and primary cells such as macrophages.

The identity of the CDK responsible for SAMHD1 phosphorylation has remained controversial. Early reports using immortal cell lines pointed to CDK1 as the kinase controlling SAMHD1 phosphorylation.^16,18 A recent report, however, identified CDK2 as a partner of SAMHD1 in HEK293T cells.²⁰ Recently, we have identified CDK2 as the kinase phosphorylating SAMHD1 in macrophages and CD4+ T cells.¹⁹ Moreover, we have identified CDK6 as a kinase upstream of CDK2 controlling SAMHD1.^19,21 Palbociclib, a specific CDK4/CDK6 inhibitor³⁰ was able to block HIV-1 reverse transcription in a SAMHD1-dependent manner.²¹

It has been suggested that SAMHD1 exerts its anti-HIV activity through a RNase activity rather than reducing the intracellular dNTP pool. According to this model, phosphorylation of SAMHD1 regulates its RNase activity but not the dNTPase activity when overexpressed in U937 cells.³¹ However, Hoffmann et al.³² concluded that reversibility of SAMHD1 restriction in time-course experiments could only be compatible with dNTP limitation as the major restriction force but not with its nuclease activity. Reduction of SAMHD1 phosphorylation at site T592 by inhibition of CDK2 or CDK6 ^19,21 or, as shown in here, by siRNA-mediated interference of cyclin D3 (Fig. 4), correlated with a reduction of intracellular dNTPs. We cannot discard that regulation of the cell cycle may affect multiple factors associated to thecontrol of dNTP levels, including SAMHD1. Our observation of a differential effect on the level of dTTP>dATP>dCTP≥dGTP may indicate that dNTP modulation is different for every deoxynucleotide and possibly includes additional factors that facilitate a specific dNTP biosynthesis metabolism. As previously described, the treatment with the CDK6 inhibitor palbociclib strongly reduced the intracellular dNTP pool in macrophages particularly affecting dTTP and dATP,²¹ indicating an intracellular dNTP turnover dependent on SAMHD1 and probably on other cellular enzymes also regulated by cyclin/CDK partners. The dNTP pool has been described as a strong limiting resource for HIV infection,^1–3 so although we cannot rule out additional restriction mechanisms, a reduced dNTP pool appears to be a major constraint for HIV-1 during early stages of the cell cycle required for viral replication.

The activity of another cyclin, L2, has been shown to target SAMHD1 for degradation in a proteasome and DCAF1-dependent manner in MDM differentiated with M-CSF.³³ Thus, cyclin L2 controls the abundance of SAMHD1 as an alternative mechanism regulating SAMHD1-mediated restriction. We have not observed differences in total SAMHD1 mRNA expression between HIV-1 restricted and susceptible macrophages ³⁴ and SAMHD1 is ubiquitously expressed in monocytes, MDM, and in resting and activated CD4+ T cells.^8,19,22,35 While SAMHD1 inactivation through phosphorylation is apparent in HIV-1 susceptible primary cells, regulation of total SAMHD1 expression may be dependent on a fine-tune mechanism irrespective of the cell cycle but similarly controlled by other host factors including cyclin L2.

Materials and Methods

Cells

Peripheral blood mononuclear cells (PBMC) were obtained and processed as described before.³⁶ Briefly, cells were collected after Ficoll-Paque density gradient centrifugation and used for fresh purification of monocytes using a negative selection antibody cocktail (StemCell Technologies). Purity of the population was confirmed by CD14 staining (1:20; BD Biosciences) assessed by flow cytometry (>80%). Monocytes were resuspended in complete culture medium: RPMI 1640 medium (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco), penicillin and streptomycin (Gibco). Monocytes were differentiated for 4–5 d in the presence of monocyte-colony stimulating factor (M-CSF, Peprotech) at 100 ng/ml. The work was approved by the scientific committee of Fundació IrsiCaixa. PBMC were isolated from ‘buffy coats’ of healthy blood donors. Buffy coats were purchased from the Catalan Banc de Sang i Teixits (http://www.bancsang.net/en/index.html). The buffy coats received were totally anonymous and untraceable and the only information given was whether or not they have been tested for disease.

Compounds

3′-azido-3′-deoxythymidine (zidovudine; AZT), was purchased from Sigma-Aldrich. Raltegravir was obtained from Merck Sharp and Dome (Barcelona, Spain). All compounds were resuspended in DMSO and stored at −20°C until use.

RNA interference

siRNAs were purchased from Dharmacon (siGENOME SMARTpool; Dharmacon, Thermo-Scientific) or Sigma-Aldrich. Monocytes were transfected with 50 pmol of the corresponding siRNA using the Monocyte Amaxa Nucleofection kit (Lonza) as previously described.^37,38 Transfected monocytes were left untreated overnight and then differentiated to macrophages as described above. Cell viability was quantified by flow cytometry in a forward-versus side-scatter (FCS and SSC, respectively) plot as described by Blanco et al.³⁶

Viruses and virus infections

R5-tropic HIV-1 strain BaL was grown in stimulated PBMC and titrated for its use in MDM. Envelope-deficient HIV-1 NL4–3 clone encoding IRES-GFP (NL4-3-GFP) was pseudotyped with VSV-G by cotransfection of HEK293T cells using polyethylenimine (Polysciences) as previously described.³⁷ For the production of viral-like particles carrying Vpx (VLP_Vpx), HEK293T cells were cotransfected with pSIV3+ ³⁹ and a VSV-G expressing plasmid. Three days after transfection, supernatants were harvested, filtered and stored at −80°C. In some cases, viral stocks were concentrated using Lenti-X concentrator (Clontech).

MDM were pre-treated with drugs or VLP_Vpx 4 h before infection. MDM were then infected with NL4–3-GFP as previously described and HIV-1 infection was measured 2 d later by flow cytometry (LSRII, BD Biosciences). Infection range showed variability between donors, ranging from 6% to 23% GFP+ cells in the absence of treatment. For quantification of proviral DNA, a primer and probe set that is able to amplify both unintegrated and integrated viral DNA was used as described before.⁴⁰ Infections were stopped at 16 h to measure only early events of viral infection (reverse transcription). DNA was extracted using a DNA extraction kit (Qiagen) and proviral DNA quantifications were performed. Ct values for proviral DNA were normalized using RNaseP as house-keeping gene using the ΔΔCt method, and infections were normalized to an untreated control. For proviral DNA quantifications, samples treated with RT inhibitor AZT (1 µM) were run in parallel to ensure that proviral DNA measured was a product of infection and not result from DNA contamination of the viral stocks. Raltegravir (2 µM) was used to ensure that no post-RT steps were being quantified by the assay.

mRNA quantification

For relative mRNA quantification, RNA was extracted using the Qiagen RNeasy Mini Extraction kit (Qiagen), as recommended by the manufacturer, including the DNase I treatment step. Reverse transcriptase was performed using the High Capacity cDNA Reverse Transcription Kit (Life Technologies). mRNA relative levels of different genes were measured by 2-step quantitative RT-PCR and normalized to GAPDH mRNA expression using the ΔΔCt method. Mean ± SD of at least 5 independent donors performed in duplicate were shown. Primers and DNA probes were purchased from Life Technologies (TaqMan gene expression assays).

Immunoblotting

Treated cells were rinsed in ice-cold PBS and extracts prepared in lysis buffer (50 mM Tris HCl pH 7.5, 1 mM EDTA, 1 mM EGTA, 1 mM Na₃VO₄, 10 mM Na β-glycerophosphate, 50 mM NaF, 5 mM Na Pyrophosphate, 270 mM sucrose and 1% Triton X-100) supplemented with protease inhibitor cocktail (Roche) and 1 mM phenylmethylsulfonyl fluoride. Lysates were subjected to SDS-PAGE and transferred to a PVDF membrane (ImmunolonP, Thermo). The following antibodies were used for immunoblotting: anti-rabbit and anti-mouse horseradish peroxidase-conjugated secondary antibodies (1:5.000; Pierce); anti-human Hsp90 (1:1000; 610418, BD Biosciences); anti-SAMHD1 (1:2.500; ab67820, Abcam); anti-cyclin D3 (2936), phospho-CDK2 (Thr160; 2561) and anti-CDK2 (2546) (all 1:1.000; Cell Signaling Technologies). The anti-phospho-SAMHD1 Thr592 (pSAMHD1 T592), obtained by immunization of rabbit using a phosphorylated peptide has been described before.^18,19

Determination of dNTP intracellular levels

MDM were rinsed and lysed with trichloroacetic acid (TCA, 0.5 M). Cellular proteins were cleared by centrifugation and supernatant was neutralized with 0.5 M Tri-n-octylamine in 1,1,2-trichlorotrifluoroethane (Sigma-Aldrich). Recovered aquose phase was recovered and dried in a SpeedVac. Pellets were resuspended in Tris-HCl buffer (40 mM, pH7.4) and dNTP content determined using a polymerase-based method^19,34,41 with minor modifications. Briefly, 20 µL of reaction mixture contained 5 µL of dNTP extract in 40 mM Tris-HCl, pH 7.4, 10 mM MgCl₂, 5 mM dithiothreitol, 0.25 µM oligoprimer, 0.75 µM [8-³H]dATP, 12–21 Ci/mmol (or [methyl-³H]dTTP for the dATP assay) and 1.7 units of Thermo Sequenase DNA Polymerase (GE Healthcare). Reaction mixtures with aqueous dNTP standards were processed in parallel. After incubation at 48°C for 60 min, 18 µL of the mix was spotted on a Whatman DE81 paper and left to dry. The filters were washed 3 times for 10 min with 5% Na₂HPO₄, once with water, once with absolute ethanol, and left to dry again. The retained radioactivity was determined by scintillation counting, and dNTP amounts calculated from interpolation on the calibration curves. To ensure the reliability of the results, duplicates of 2 different dilutions of each dNTP extract (usually undiluted and 1:3 water-diluted) were processed in each independent experiment.

Statistical methods

Data were analyzed with the PRISM statistical package. If not stated otherwise, all data were normally distributed and expressed as mean ± SD and p-values were calculated using an unpaired, 2-tailed, t-student test. For normalized data from different donors, one-sample t-test against hypothetical value of 1 or 100 was applied.

Changes in CDK/cyclin mRNA profile were considered significant if obtained mRNA levels were higher or lower than one standard deviation of the mean of all equally-treated samples.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank the National Institutes of Health (AIDS Research and Reference Reagent Program) and the EU Program EVA Centralised Facility for AIDS Reagents, NIBSC, UK for reagents.

Funding

This work was supported in part by the Spanish Ministerio de Economía y Competitividad (MINECO) and Fondo de Investigación Sanitaria (FIS) projects BFU2012-31569 (JAE), FIS PI13–01083 (BC), FIS CP14/00016 (EB) and the HIV BioBank RIS (project RD06/0006/0035). AR, RB and EB are research fellows from MINECO/FIS.