We previously showed that Tat is phosphorylated by recombinant CDK2/cyclin E in vitro and that Tat's Ser16 was a potential phosphorylation site 20. Indeed recombinant CDK2/cyclin E efficiently phosphorylates Tat (Fig. 1A, lane 1). Tat can also be phosphorylated by HeLa nuclear extract (Fig. 1A, lane 2). Immunodepletion of CDK2 from HeLa nuclear extract completely abolished Tat phosphorylation (Fig. 1A, lane 3) suggesting that under these conditions Tat was largely phosphorylated by CDK2. Upon analysis of the sequence of Tat, the (S/T)₀P₁K₂(K/R)₃ consensus motif for serine phosphorylation by CDK2 2728 was not found, but several sequences were found that partially matched this motif: S¹⁶QP(K/R)¹⁹ , S⁴⁶YGR⁴⁹ , S⁶⁸LSK⁷¹ . To determine potential phosphorylation sites, Tat phosphorylated in vitro was analyzed by Mass spectrometry. For this purpose, purified recombinant Tat was phosphorylated with recombinant CDK2/cyclin E in vitro followed by immunoprecipitation, SDS-PAGE purification, in-gel digestion with trypsin, and HPLC purification. Analysis of HPLC eluates showed presence of two peaks in the digest generated from phosphorylated Tat that were absent in digest of non-phosphorylated Tat (Fig. 1B). These two peaks were collected and subjected to MALDI-TOF mass spectrometry. Masses of the peptides over 900 Da were acquired and analyzed with FindPep tool 29 using the sequence of Tat (MEPVDPNLEPWKHPGSQPRTACNNCYCKKCCFHCYACFTRKGLGISYGRKKRRQRRRAPQDSQTHQASLSKQ) as input. The masses of peptides that did not match to Tat were further compared with Tat peptides from which we subtracted 18 Da, assuming β-elimination of phosphoric acid which is likely to occur during MALDI-TOF analysis 30. Matched peptides shown in Table 1, indicate that Peak I contains peptides with Serine 46 as a potential phosphorylation site, whereas Peak II contains peptides with Serine 16 as a potential phosphorylation site. All matched peptides contain internal lysine or arginine residues, and thus they are apparently products of incomplete digestion by trypsin, which could be a result of incomplete in-gel digestion. Also acquisition of peptide with masses over 900 Da would only allow detection of relatively large peptides. The data suggest that Tat is phosphorylated in vitro by CDK2 and that this phosphorylation might take place at Ser¹⁶ or Ser⁴⁶ residues within S¹⁶QPR¹⁹ and S⁴⁶YGR⁴⁹ sequences that partially match to the (S/T)P×(K/R) consensus sequence for CDK2 phosphorylation 28.

Previous attempts to detect Tat phosphorylation in vivo were unsuccessful 25. We hypothesized that low level of Tat expression after transfection and/or rapid de-phosphorylation by cellular phosphatases might prevent detection of Tat phosphorylation in vivo. To overcome these difficulties, we expressed Flag-tagged Tat which we found to be expressed to a higher level in COS-7 cells than untagged Tat (Fig. 2, compare lanes 3 and 4). To facilitate expression of Flag-Tat in HeLa cells we used adenovirus-mediated expression of Tat 31. HeLa cells were infected with Adeno-Tat and incubated 48 hours post infection to allow expression of Tat. Then cells were pulsed with (³²P)-labeled orthophosphate, and Tat was immunoprecipitated from cellular lysates using monoclonal anti-Flag or polyclonal anti-Tat antibodies. Immunoprecipitated proteins were resolved by SDS-PAGE on 15% Tris-Tricine gel 32 and transferred to PVDF membrane. The membrane was probed with monoclonal anti-Tat antibodies (Fig. 3A) using 3,3'-Diaminobenzidine enhancer system (DABM, Sigma), and also exposed to a PhosphoImager screen (Fig. 3B). Both antibodies precipitated a well detectable amount of Tat protein (Fig. 3A, lanes 3, 4, 6 and 7). Under these experimental conditions, Tat was phosphorylated (Figs. 3B and 3C, compare lane 3 to lane 1 and lane 6 to lane 5). Next we treated cells with okadaic acid, an inhibitor of PPP-type phosphatases, to prevent rapid dephosphorylation of Tat in the cells and during the lysis procedure. Treatment with okadaic acid did not change the amount of precipitated Tat (Fig. 3A, lanes 3, 4, 6 and 7). In contrast, Tat phosphorylation was significantly enhanced when cells were treated with okadaic acid (Figs.3B and 3C, lanes 4 and 7). Taken together, these results indicate that Tat is phosphorylated in vivo and that Tat phosphorylation is enhanced when cells are treated with the inhibitor of PPP-phosphatases.

We next investigated whether Tat phosphorylation was mediated by CDK2 in vivo using CDK2-directed RNA interference 33. HeLa cells were infected with Adeno-Tat and subsequently transfected with siRNAs against CDK2. Transfection of HeLa cells with siRNAs against CDK2 decreased the level of expression of CDK2 by 2.5-fold (Figs. 4A and 4B, lane 3). A control non-targeting siRNA pool did not affect expression of CDK2 (Figs. 4A and 4B, lane 2). The non-targeting siRNA control was used to ensure that transfection itself did not affect CDK2 expression. Western blot analysis of tubulin and CDK9 was used as control for the specificity of siRNAs. As shown in Fig. 4A transfection with both siRNAs did not affect the level of α-tubulin expression. To ensure that CDK2-directed siRNA decreased the enzymatic activity of cellular CDK2, CDK2 was immunoprecipitated from cells transfected with non-targeting or CDK2-directed siRNA and assayed for its enzymatic activity using recombinant Tat as a substrate. The activity of CDK2 was decreased in the cells transfected with CDK2-directed siRNA (Fig. 4C, lane 2) as compared to the cells transfected with non-targeting siRNA (Fig. 4C, lane 1). Next the cells were infected with Adeno-Tat, transfected with non-targeting or CDK2-directed siRNAs and pulse-labeled with (³²P). In this experiment no okadaic acid was used. Inhibition of CDK2 by siRNA reduced the level of Tat phosphorylation by 3-fold (Figs. 5A,B and 5C, compare lanes 2 and 3). Thus CDK2 is likely to mediate Tat phosphorylation in cultured cells. To analyze whether Tat and CDK2 might be present in the same molecular weight complex, we analyzed sedimentation of Tat and CDK2 by ultracentrifugation on a glycerol gradient (Fig. 6). Both Tat and CDK2 co-migrated in fractions 1–8 (Fig. 6). The CDK9 was present in most of the fractions with the peak in fractions 4–5 and 9–10 (Fig. 6), which is likely to correspond to low and high molecular weight P-TEFb complexes. HEXIM1, and Brd4 were mostly present in fractions 2–6, although HEXIM1 was also detected in higher molecular weight fractions 10 and 11 (Fig. 6). Neither Tat nor CDK2 co-migrated with RNAPII which was present in fractions 1–13 (Fig. 6). We further analyzed association of Tat with CDK2 by immunoprecipitation. Flag-Tat was expressed in HeLa cells by infection with Adeno-Tat and precipitated from cellular extracts with anti-Flag antibodies (Fig. 7A). CDK2 co-precipitated with Tat (Fig. 7A, lane 4). CDK2 was not precipitated with anti-Tat antiserum from non-infected cells (Fig. 7A, lane 3) or not with non-specific preimmune serum from adeno-Tat infected cells (Fig. 7A, lane 5). Inhibition of CDK2 expression by CDK2-specific RNAi significantly reduced CDK2 co-precipitated to Tat apparently due reduction of the expressed CDK2 (Fig. 7B, compare lane 4 to lane 2). Association of Tat with CDK9 was slightly reduced (Fig. 6B) but this reduction correlated to the decreased amount of Tat precipitated by anti-Flag antibodies. Binding of Tat-cyclin T1 was not reduced (Fig. 7C, lanes 3 and 5). The cyclin T2 did not precipitate with Tat (Fig. 7C), which indicated a specificity of the immunoprecipitation. Taking together, these results suggest that CDK2 associates with Tat in cultured cells, and that inhibition of CDK2 expression prevents Tat phosphorylation. Thus, CDK2 is likely to phosphorylate Tat directly in cultured cells.

We next determined whether serine, threonine or tyrosine residues of Tat were phosphorylated in vivo. HeLa cells were infected with Adeno-Tat, labeled with (³²P), and treated with okadaic acid to achieve a higher level of Tat phosphorylation. Tat was immunoprecipitated with anti-Flag antibody and resolved on 15% SDS Tris-Tricine PAGE (Fig. 8A, lane 1). Phosphoamino acid analysis of radioactive Tat extracted from the gel (see Materials and Methods) showed presence of phospho-serines but not phospho-threonines or phospho-tyrosines in (³²P)-labeled Tat (Fig. 8B). Thus, only serine residues of Tat were phosphorylated in vivo.

We next investigated a possibility that Tat might be phosphorylated on S¹⁶ or S⁴⁶ residues in vivo. We generated mutants of Flag-Tat in which either or both Ser residues were substituted by Ala. 293T cells were transfected with WT and mutant Tat-expressing vectors, Tat was precipitated with anti-Flag antibodies and analyzed on 15% SDS Tris-Tricine PAGE followed by PhosphoImager analysis. Expression of Tat was verified by Western blotting (Fig. 8C). While we could detect phosphorylation of WT Tat (Fig. 8C, lane 2), the Tat S16A mutant and Tat S46A mutant were about 2–3 fold less phosphorylated (Fig. 8C, middle and lower panels, lanes 3 and 4). The Tat S16,46A double mutant was even less phosphorylated (Fig. 8C, lane 5). Our results indicate that both S¹⁶ and S⁴⁶ are likely to be phosphorylated in vivo.

We next investigated the functional relevance of Tat's S¹⁶ and S⁴⁶ residues in HIV-1 transcription. We generated mutants of Tat in which either or both Ser residues were substituted by Ala. To ensure expression of the mutants, COS-7 cells were transfected with WT and mutant Tat-expressing vectors and cellular lysates were analyzed on 15% SDS Tris-Tricine PAGE followed by Western blot with anti-Tat antibodies. As shown in Fig. 9A, all Tat mutants were expressed, with the level of expression of Tat mutants higher than the WT Tat. The higher expression level of non-tagged Tat mutants was a reproducible effect and was not a consequence of the difference in the amount of transfected DNA. The effect of Tat mutations on the ability of Tat to activate HIV-1 LTR promoter was analyzed in HeLa cells co-transfected with Tat-expression vectors and HIV-1 LTR-LacZ reporter plasmid (Fig. 9B). Non-mutated Tat (WT) increased the level of transcription by 400-fold (Fig 9B). HIV-1 transcription induction by the Tat S16A mutant was approximately 75% that of WT Tat (Fig. 9B), while transactivation by the Tat S46A mutant was about 2 times lower than with the WT Tat (Fig. 9B) and induction by the double S16, 46A Tat mutant was 3-times lower than that of WT Tat (Fig. 9B). Thus, mutation of either Ser¹⁶ or Ser⁴⁶ of Tat interferes with the level of Tat-transactivation and mutation of both residues has an additive effect.

We determined whether mutations of Tat S16A and/or S46A have an effect on the ability of Tat to induce HIV-1 transcription from an integrated HIV-1 provirus. We used HLM-1 cells (AIDS Research and Reference Reagent Program) that were derived from HeLa-CD4+ cells containing an integrated copy of HIV-1 proviral genome with a Tat-defective mutation (termination linker at the first AUG). HLM-1 cells are negative for virus particle production, but can be induced to express high levels of infectious HIV-1 after transfection with Tat. We transfected the HLM-1 cells with wild type or mutant Tat vectors and tested supernatants for the presence of HIV-1 particles using p24 gag antigen ELISA at day 0, day 1, day 2, day 7 and day 14 posttransfection. Neither S16A nor S46A mutants of Tat efficiently induced HIV-1 viral production (Fig. 10A). The double S16, 46A mutant also had a reduced activity (Fig. 10A). To phosphorylate Tat during virus replication, we pulsed HLM-1 cells transfected with WT and mutant Tat with (³²P) orthophosphate and also treated the cells with okadaic acid. Tat was immunoprecipitated with anti-Flag antibodies, resolved on 15% SDS PAGE and its phosphorylation was detected by PhosphoImager. While WT Tat was phosphorylated, the mutants were not phosphorylated efficiently (Fig. 10B). These data indicate that the S16A and S46A mutations of Tat interfere with the ability of Tat to activate integrated HIV-1 provirus, and prevent Tat phosphorylation during one round of viral replication.

As we discussed above, analysis of the sequence of Tat for the presence of the (S/T)₀P₁K₂(K/R)₃ consensus motif for serine phosphorylation by CDK2 2728 showed that several sequences partially matched this motif: ¹⁶SQP(K/R)¹⁹ , ⁴⁶ SYGR⁴⁹ , ⁶⁸SLSK⁷¹ . We determined conservancy of S¹⁶, S⁴⁶ and S⁶⁸ residues. We analyzed 158 sequences of Tat isolates deposited in the PubMed database. Both S¹⁶ and S⁴⁶ were highly conserved with an occurrence of 100% or nearly 100% (Fig. 11). In contrast, S⁶⁸ was present in 41%, and S⁷⁰ – in 60% of the Tat isolates and all other serines were present in less than 50% (Fig. 11). This analysis indicates that S¹⁶ and S⁴⁶ residues might be critical for HIV-1 replication. Because S¹⁶ and S⁴⁶ residues were 100% conserved, it was not possible to analyze the effect of the mutation of these residues on the HIV-1 progression. We took advantage of the notion that a lysine or arginine present in the fourth position of the (S/T)₀P₁K₂(K/R)₃ consensus motif is critical for the recognition of the motif by CDK2 2728. Thus we analyzed whether (K/R)¹⁹ and R⁴⁹ residues within ¹⁶SQP(K/R)¹⁹ and ⁴⁶ SYGR⁴⁹ sequences of Tat were conserved among different HIV-1 isolates and whether mutations in these residues correlated with apparent progression of disease in the patient from whom the isolates were obtained. In this investigation, we studied 105 sequences of Tat deposited to PubMed database 34. Of these, 55 were obtained from patients who were not on antiretroviral treatment and were classified as being healthy, and 50 were from patients who were classified as being ill and who were on antiretroviral treatment. We found that R⁴⁹ was absolutely conserved, and thus no correlation could be obtained. In contrast, Table 2 shows that 29 (53%) of the "healthy" HIV patients had a (K/R)¹⁹ mutation compared to only 4 (8%) of the "ill" HIV patients (Pearson Chi-square correlation 24.312, df = 1; P < 0.001). This result suggests physiological importance of mutation of position 19 of Tat for progression of HIV-1 disease, and are consistent with the possibility that ¹⁶SQP(K/R)¹⁹ sequence is a putative CDK2 recognition site in Tat.

293T cells and COS-7 cells were purchased from ATCC (Manassas, VA). HeLa-MAGI cells 53, HLM-1 cells 54, anti-Tat rabbit polyclonal (HIV-1 BH10 Tat antiserum, 55) and monoclonal (NT3 2D1.1, courtesy of Dr. Jonathan Karn) antibodies were received from the AIDS Research and Reference Reagents Program (NIH). Anti-Flag monoclonal antibodies, anti-α-tubulin antibodies, protein (G) and protein (A) agarose and okadaic acid were purchased from Sigma (Atlanta, GA). All radioactive reagents were purchased from GE Health Care Life Sciences. The 3,3'-diaminobenzidine enhanced liquid substrate system for membrane ELISA (DABM) was purchased from Sigma (St Louis, MO). Antibodies for CDK2, CDK9, cyclin T1 and cyclin T2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against RNAPII were from Babco. Anti-Brd4 and anti-HEXIM1 antibodies were a gift from Dr.Q. Zhou (University of California, Berkeley). HIV-1 Tat was expressed in Escherichia coli and purified on Aquapore RP-300 column (Applied Biosystems, Foster City, CA) by reverse-phase chromatography as we described 20.

Tat expression plasmid was a gift from Dr. Ben Berkhout (University of Amsterdam) 48. The Flag-Tat was cloned into the adeno-CMV-link vector as described below and verified by sequencing. The S16A and S46A mutations of the sequence of Tat were made according to the Quick-Change site-directed mutagenesis protocol of Stratagene, using the appropriate primers and templates. The sequences of the DNA constructs were verified by sequencing using a commercial service from Macrogen (Seoul, Korea).

CDK2 and cyclin E were purified from lysates of Sf9 insect cells infected with baculoviruses producing CDK2 and cyclin E. Proteins were purified essentially as described earlier 56. Briefly, 1 ml of cells (from 250 ml of culture) was lysed with 16 ml of lysis buffer (50 mM Tris-HCl, pH 8.0, 10 mM 2-mercaptoethanol, 10% glycerol and with PMSF), homogenized on ice and centrifuged at 45,000 g for 1 hr at 4°C. Supernatant was loaded on Mono-Q 10/10 column (Amersham, USA). Two separate cell cultures, one infected with CDK2-expressing baculovirus and the other one infected with cyclin E-expressing baculovirus were used for purification. The Mono-Q fractions containing CDK2 or cyclin E were mixed 1:1 and loaded onto Superdex column (Sephadex H200, Amersham, USA). Purity of CDK2/cyclin E was checked on 12% PAGE followed by Coomassie staining (Additional file 1). We also analyzed the Superdex fractions by immunoblotting with andti-CDK2 antibodes and assayed their enzymatic activity using histone H1 and purified Tat proteins as substrates. Fractions containing CDK2/cyclin E were concentrated using Microcon tubes (Amicon, USA).

CDK2 kinase assays were performed at 30°C for 30 min in kinase assay buffer (50 mM HEPES-KOH, pH 7.9, 10 mM MgCl₂, 6 mM EGTA, 2.5 mM DTT) containing histone H1 or purified Tat protein, 200 μM ATP and (γ-P³²)ATP. A mixture was incubated for 30 min at 30°C, reaction was stopped with 8 μl of 4× SDS buffer and resolved on 12% PAGE.

293T cells were transfected with CDK2-directed or non-targeting siRNA. At 48 hours after transfection cells were lysed in whole cell lysis buffer. About 50 μg of protein lysate was subjected to precipitation with anti-CDK2 rabbit antibodies (600 ng/IP) on 40 μl protein A agarose beads (50% slurry) (Sigma). As a control, rabbit preimmune serum was used. Precipitation was carried out for 2 hrs at 4°C. The beads were washed with TNN buffer, then with TTK buffer and supplemented with 20 μl of kinase mixture containing TTK buffer, 100 μM ATP, 0.5 μCi of γ-(³²P)ATP and 1 μg of purified Tat. The kinase reaction was carried out for 15 min at 30°C, reaction was stopped with 8 μl of 4× SDS-loading buffer and resolved on 12% Tris-Tricine gel. The gel was stained with Coomassie blue, dried and exposed to Phosphor Imager screen.

293T cells were infected with Adeno-Tat virus with MOI 1, and incubated overnight. The cells in 100 mm plate were lysed with 0.5 ml of whole cell lysis buffer (50 mM Tris-HCl, pH 7.5, 0.5 M NaCl, 1% NP-40, 0.1% SDS) supplemented with protease cocktail (Sigma) and RNasin (Amersham). Cell lysates were clarified by centrifugation for 30 min at 10,000 g and loaded on top of 10% to 30% glycerol (9 ml) gradient. Glycerol gradient buffer contained 20 mM HEPES-KOH, pH 7.9, 150 mM KCl, 200 μM EDTA

The gradient was spun in SW 41Ti rotor (Beckman) at 38,000 rpm for 18 hours. Fractions (0.5 ml) were collected through a needle inserted to the bottom of the tube and analyzed by Immunoblotting.

Non-phosphorylated and phosphorylated Tat were resolved by 15% Tris-Tricine SDS-PAGE. A gel piece containing Tat was crushed and incubated with 1 μg of porcine trypsin (Promega, Madison, WI) in 0.1 M NaHCO₃ (pH 7.9) overnight at 37°C. The extracted peptides were lyophilized, dissolved in 0.1% TCA and separated by reverse-phase chromatography on a μRPC C2/C18 ST 4.6/100 column (Amersham Pharmacia Biotech) using AKTA purifier (AmershamPharmaciaBiotech). Eluent A was 0.1% TCA in water and eluent B was 0.1% TCA in 90% acetonitrile. The gradient was 5% B for 2 column volumes (CVs), 5–50% B for 20 CVs, 50–100% B for 12 CVs and 100% B for 4 CVs. The flow rate was 0.5 ml/min.

Fractions from the HPLC separation described above were lyophilized. Multiple peptide sequences were determined in a single run by Applied Biosystems Maldi-TOF/TOF 4700 proteomics analyzer (25–30,000 resolution in reflectron mode, 5 ppm accuracy with IS, subfmole sensitivity for peptide mass fingerprints (PMF), fmole sensitivity for PSD/CID capacity of >36,000 PMF per 24 hours). Data were analyzed utilizing a local multi-processor installation of MASCOT MS PMF and MS/MS PSD, CID peptide identification software. GPS also supports an integrated MS-MS/MS mode and a chromatographic separation mode with SEQUEST-like capability.

The E1-deleted recombinant Ad carrying Tat was generated as previously described 31. Briefly, a cDNA fragment encoding the full length HIV-1 Tat protein was cloned into the plasmid pCXN was subcloned in the pAd.CMV link plasmid. The Tat-coding DNA corresponds to the HIV-1 isolate P9.3 from United Kingdom (Accession number AF324447). The DNA sequence is ATGGACTACAAGGACGACGA TGACAAAGAA TTCATGGAGC CAGTAGATCCTAGACTAGAG CCCTGGGAGC ATCCAGGAAG TCAGCCTAAG ACTGCTTGTACCCCTTGCTA TTGTAAAAAG TGTTGCTTTC ATTGCCAAGT TTGTTTCACAACAAAAGGCT TAGGCATCTC CTATGGCAGG AAGAAGCGGA GACAGCGACGAAGAGCTCCT CAAGACAGTC AGACTCATCA GGCTTCTCTA TCAAAGCAATCCCTACCCCA AACCCAGAGG GACTCGACAG GCCCGGAAGA ATCGAAGAAGGAGGTGGAGA GCAAGGCAGAGACAGATCGA TTCGATTA. It corresponds to the protein sequence of Tat containing Flag epitope at its N-terminus-MDYKDDDDKEFMEPVDPRLEPWEHPGSQPKTACTPCYCKKCCFHCQVCFTTKGLGISYGRKKRRQRRRAPQDSQTHQASLSKQSLPQTQRDSTGPEESKKEVESKAETDRFD. pAd.CMVlink and Cla I digested adenoviral DNA were co-transfected into HEK293 cells at a ratio of 3:1 by calcium phosphate precipitation to allow the recombination. The virus was purified through three rounds of plaque-purification. Viruses were replicated in HEK293 cells and were purified from a cell lysate by two rounds of CsCl density gradient centrifugation. The purified virus was desalted on a Bio-Gel P-6 desalting column (Bio-Rad Laboratories, Hercules, CA) equilibrated with PBS. The titer of the virus preparation was determined both by absorbency at 260 nm and by plaque assay. The particle to plaque forming unit ratio was less than 100. Purified viruses were suspended in PBS at the desired concentrations.

HeLa cells were infected with recombinant Adenovirus carrying Flag-tagged Tat prepared as we described 31. The purified virus had a particle to plaque forming unit (Pfu) ratio of less than 100. We added approximately 10 Pfu per cell to achieve high level of Tat expression in infected HeLa cells. At 48 hours post infection the media was changed for 1 hour to a phosphate-free DMEM media (Life Technologies, Rockville, MD) containing no serum. Then the media was changed to phosphate-free DMEM supplemented with 0.5 mCi/ml of (³²P)-orthophosphate and cells were further incubated for 2 hours at 37°C. Where indicated, 1 μM okadaic acid (Sigma) was added to block cellular PPP-phosphatases. Cells were washed with PBS and lysed in whole cell lysis buffer (50 mM Tris-HCl, pH 7.5, 0.5 M NaCl, 1% NP-40, 0.1% SDS) supplemented with protease cocktail (Sigma). After 10 min on ice, cellular material was scraped and then centrifuged at 14,000 rpm, 4°C for 30 min. The supernatant was recovered and immediately used for immunoprecipitation. Tat was precipitated with anti-Flag monoclonal antibodies coupled to protein G agarose and with polyclonal anti-Tat antibodies coupled to protein A agarose for 2 h at 4°C in a TNN Buffer containing 50 mM Tris-HCl, pH 7.5, 0.15 M NaCl, and 1% NP-40. The immunoprecipitated Tat was recovered by heating for 2 min at 100°C in Tricine SDS-loading buffer, resolved on 15% Tris-Tricine SDS-PAGE 32 and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Allen, TX). The membrane was analyzed with anti-Tat monoclonal antibodies using 3,3'-Diaminobenzidine enhancer system (Sigma) and was also exposed to Phosphor Imager screen (Packard Instruments, Wellesley, MA).

The analysis was carried out on the Hunter thin layer peptide mapping electrophoresis system (CBS, Del Mar, CA) according to the manufacturer's recommendations. Briefly, phosphorylated Tat, prepared as described above was resolved on 15% SDS-Tris-Tricine PAGE and the gel was dried on a Whatman paper. The portion of the gel containing Tat was excised, rehydrated and treated overnight with trypsin to elute Tat. The eluted peptides were boiled in 5.7 M HCl at 110°C to liberate (³²P)-labeled phospho-amino acids as described 57. After the hydrolysis, the sample was lyophilized and resuspended in the buffer for pH 1.9 electrophoresis also containing the phosphoamino acid standards at 0.06 mg/ml. The sample was resolved by thin-layer electrophoresis on a cellulose plate (CBS, Del Mar, CA) at pH 1.9 in the first direction and at pH 3.5 in the second direction. Cold standards were visualized by staining the plate with 0.25% ninhydrine dissolved in ethanol. Positions of labeled phosphoamino acids were analyzed with Phosphor Imager (Packard Instruments, Wellesley, MA).

HLM-1cells were derived from HeLa-T4+ cells integrated with one copy of HIV-1 genome containing a Tat-defective mutation. The mutation was introduced as a triple termination linker (TTL) at the first AUG of Tat gene 54. HLM-1 cells are negative for virus particle production HLM1 cells are negative for virus particle production. HLM1 cells can be induced to express non-infectious HIV-1 particles after transfection with Tat cDNA, or by treatment with mitogens such as TNF-α or sodium butyrate. HLM1 cells were grown in DMEM media containing 100 μg/ml of G418, plus 1% streptomycin, penicillin antibiotics and 1% L-Glutamine (Gibco/BRL). The cells grown up to 75% confluence were transfected with Tat expression vectors, including wild type Tat, S16A, S46A, and S16A/S46A mutant Tat plasmids using the calcium phosphate method. The transfected cells were washed after four hrs and fresh complete DMEM media with 10% fetal bovine was added for the remainder of the experiment. The p24 gag antigen was detected in the supernatants of transfected cells using a standard ELISA kit (Abbott).

CDK2-directed siRNA pool (M-003236-03-005) and negative control pool (D-001206-13-05) were purchased from Dharmacon (Dallas, TX). The siRNAs were transfected at final concentration of 100 nM using Lipofectamin reagent (Invitrogen) according to the manufacturer's recommendations. The siRNAs were incubated with cells for 2 days before cells were labeled with ³²P or lysed for Western blotting analysis.