Human Cytomegalovirus pUL97 Regulates the Viral Major Immediate Early Promoter by Phosphorylation-Mediated Disruption of Histone Deacetylase 1 Binding

Biological reagents.

The wild-type HCMV strain AD169 (ADwt) was obtained from the AD169 bacterial artificial chromosome (BAC) clone (42). A UL97 substitution mutant, ADdel97 (del97), was constructed using BAC recombineering (43, 44). The UL97 gene was replaced by inserting the Escherichia coli galK gene into the UL97 open reading frame of the AD169 wild-type BAC clone (forward primer, 5′-TCGGTGTGGTAGCTAGTGCAGCCTTAGGAACAGGGAAGACTGTCGCCACTCCTGTTGACAATTAATCATCGGCA-3′, and reverse, 5′-AGCCGCATGAGGCAAAAGCCCAGCACGTTACCCAGCGCCGACAGCTCCGATCAGCACTGTCCTGCTCCTT-3′) (Integrated DNA Technologies, Inc., Coralville, IA). AD169 viral stocks were prepared by infecting primary human fibroblasts (HFF) at a low multiplicity of infection (<0.1 PFU/ml). ADdel97 and matched AD169 stocks were prepared by cotransfecting BAC DNA with the pCGN-pUL82HA expression vector. To concentrate viral stocks, culture medium and cleared cell lysate were pelleted through a sorbitol cushion (20% d-sorbitol, 50 mM Tris-HCl [pH 7.2], 1 mM MgCl₂) at 55,000 × g for 1 h. Titers were defined as infectious units (IU) per ml and determined by using serial dilutions of viral stocks to infect HFFs in 12-well dishes, fixing cells at 48 hpi, staining with an antibody against pUL123 (IE1), and counting the UL123-positive cells per well. To validate viral titers, we infected HFFs with ADwt, ADwt plus 40 μM maribavir (MBV), or ADdel97 at 0.01 IU/ml, fixed them at 48 hpi, and stained them for pUL123-positive cells.

HFFs, MRC-5 lung fibroblast cells, U-2 OS osteosarcoma cells, and U373 astrocytoma cells were propagated in Dulbecco's modified Eagle medium (DMEM) (Life Technologies, Carlsbad, CA) supplemented with 7% fetal bovine serum (FBS) (Life Technologies) and 1% penicillin-streptomycin (Life Technologies). ADdel97 and matched AD169 viral stocks were propagated in DMEM supplemented with 15% FBS and 1% penicillin-streptomycin following transfection. All infections were completed using cells grown to ∼90% confluence and serum starved (0.5% serum) for 24 to 48 h.

Cells were infected at a multiplicity of infection of 1 IU per cell in DMEM containing 7% FBS and 1% penicillin-streptomycin and exposed to the viral inoculum for 2 h, unless otherwise noted. Compounds were added to the culture medium at the start of infection and reapplied after removing the inoculum. Cells were treated with 40 μM maribavir (Viropharma, Exton, PA), 300 nM trichostatin A (Sigma, St. Louis, MO), 100 μg/ml cycloheximide (MP Biomedicals, Santa Ana, CA), or dimethyl sulfoxide (DMSO) as a vehicle (Sigma). Electroporation of plasmid or BAC DNA was carried out in DMEM at 260 mV for 23 ms in a 4-mm electroporation cuvette. Cells were transfected using Fugene 6 (Roche, Indianapolis, IN) following the manufacturer's protocol. The pCGN-UL82HA, pCGN-UL97HA, and pCGN-UL97K355MHA plasmids were generously provided by Robert Kalejta (University of Wisconsin—Madison) (20). Vector controls were performed using a pCGN vector. The pBJ5-HDAC1Flag plasmid was graciously provided by Edward Seto (45). Point mutations of the pBJ5-HDAC1Flag were created using the Change-IT Multiple Mutation site-directed mutagenesis kit (Affymetrix, Santa Clara, CA). The following primers were used to generate the mutations: T208A, 5′-phospho-AGAGTACTTCCCAGGTGCGGGTGACCTACGGGATATC-3′; T208E, 5′-phospho-AGAGTACTTCCCAGGTGAGGGTGACCTACGGGATATC-3′; and T208D, 5′-phospho-AGAGTACTTCCCAGGTGACGGTGACCTACGGGATATC-3′ with the reverse primer 5′-phospho-CCAATGCTGAGGAGATGACCAAGTACCACAGC-3′. (Underlining identifies the locations of nucleotide substitutions.) E. coli was transformed with the generated plasmids, and transformants were screened by sequencing using an ABI 3500 Genetic Analyzer (ABI, Oyster Bay, NY).

Antibodies used for Western blot (WB), immunoprecipitation (IP), chromatin immunoprecipitation (ChIP), or immunofluorescence analysis (IF) include mouse anti-FLAG M2 (WB, IP, and IF) (Sigma), mouse anti-HDAC1 2E10 (WB and ChIP) (EMD Millipore, Billerica, MA), mouse anti-GAPDH (glyceraldehyde-3-phosphate dehydrogenase) clone 0411 (WB) (Santa Cruz Biotechnology, Santa Cruz, CA), mouse anti-acetyl-histone H3 Lys 9 (ChIP) (EMD Millipore), mouse anti-HA clone HA-7 (WB and IP) (Sigma-Aldrich), and mouse control IgG (Santa Cruz Biotechnology, Santa Cruz, CA). The antibodies against HCMV proteins were mouse anti-pUL123 clone 1B12 (WB and IF), mouse anti-pUL122 clone 3A9 (WB), mouse anti-pUL82 clone 2H10-9 (WB and IF), mouse anti-pUL83 clone 8F5 (WB and IF), mouse anti-pUL99 clone 10B4-29 (WB), and mouse anti-pUL38 clone 3D12 (WB) (46, 47), a gift from T. Shenk (Princeton University), mouse anti-pUL85 (WB), a gift from W. Gibson (The Johns Hopkins University School of Medicine), and mouse anti-pUL44 (Virusys, Taneytown, MD). Secondary antibodies for Western blot analysis were goat anti-mouse horseradish peroxidase (HRP) and donkey anti-rabbit HRP (Jackson ImmunoResearch, West Grove, PA), and for immunofluorescence donkey anti-mouse IgG (H+L) Alexa Fluor 488 (Life Technologies) was used.

Analysis of protein and RNA.

Western blot analysis of steady-state protein levels were performed on cells lysed by sonication in a radioimmunoprecipitation assay (RIPA) lysis buffer (50 mM Tris-Cl [pH 8.0], 150 mM NaCl, 0.5% NP-40). Protein from lysates were resolved by sodium dodecyl sulfate 8 or 10% polyacrylamide gel electrophoresis (SDS-PAGE) and transferred by semidry transfer to Protran nitrocellulose membrane (Whatman Inc., Piscataway, NJ). Blocking was completed in 5% milk in phosphate-buffered saline (PBS) containing 0.1% Tween 20 (PBS-T). Membranes were incubated in primary antibody followed by a secondary antibody conjugated to HRP, both diluted in 5% milk in PBS-T. Antibodies were detected with ECL (GE Healthcare, Piscataway, NJ) and film or FluorChem HD2 (ProteinSimple, Santa Clara, CA) for protein quantification. Immunoprecipitation experiments were initiated by lysing cells using sonication in a RIPA lysis buffer and using Dynabeads Protein G (Life Technologies). Lysates were spun for 2 min at ∼15,000 × g, and supernatants were precleared for 30 min at 4°C with beads. Following two washes with lysis buffer, beads were preconjugated to antibody for 30 min at room temperature and then washed twice with lysis buffer. Samples were incubated with antibody for 3 h at 4°C with rotation. The beads were washed five times with lysis buffer and resuspended in Laemmli sample buffer. The immunoprecipitated samples and lysate controls were evaluated by Western blotting. All buffers used for lysis or washes contained Complete Mini EDTA-free Protease Inhibitor Cocktail (Roche) and PhosStop (Roche).

Cellular and viral RNA was quantified using quantitative reverse transcriptase PCR (qRT-PCR). RNA was collected from samples using TRIzol Reagent (Life Technologies), and cDNA was synthesized using 2 μg of DNase-treated RNA, random hexamers, and SuperScript III reverse transcriptase (Life Technologies). qPCR of cDNA was performed using the following primers: UL123 (5′-GCCTTCCCTAAGACCACCAAT-3′ and 5′-ATTTTCTGGGCATAAGCCATAATC-3′), UL38 (5′-ACGGTGCTATCGTGCTGGAGTATT-3′ and 5′-AAGACCATCACCAGGTCGTCCATA-3′), UL44 (5′-GCCCGATTTCAATATGGAGTTCAG-3′ and 5′-CGGCCGAATTCTCGCTTTC-3′), UL54 (5′-CACCAAAGACACGTCGTT-3′ and 5′-GTCCTTTGCGACCAGAAT-3′), UL97 (5′-GACGCCGTCTAACACGTATACC-3′ and 5′-GGTTCCTACCTTCTCTGTTGCC-3′), and GAPDH (5′-ACCCACTCCTCCACCTTTGAC-3′ and 5′-CTGTTGCTGTAGCCAAATTCGT-3′). Quantification was accomplished by using FastStart Universal SYBR green Master (Roche) and the 7900HT Fast Real-Time PCR system. Relative abundance was determined using a standard curve consisting of 10-fold serial dilutions of one sample.

Cell cycle analysis.

For cell cycle analysis, HFFs were grown to ∼80% confluence in DMEM with 7% FBS. Cells were then serum starved in DMEM with 0.5% FBS and 40 μM MBV or DMSO for 48 h. Subsequently, cells were released from starvation into DMEM with 7% FBS and either MBV or DMSO. After 24 h postrelease, cells were harvested and fixed with 70% ethanol, and DNA content was analyzed by treating cells with Cell Cycle Reagent (EMD Millipore) for 30 min. Cell cycle analysis was performed on a Guava EasyCyte Mini flow cytometer (EMD Millipore). Flow cytometry data were analyzed using FlowJo Analysis Software (FlowJo, Ashland, OR). Flow cytometry data were analyzed and gated identically using FlowJo Analysis Software (FlowJo, Ashland, OR).

Cell doubling was performed from HFFs following release from serum starvation. Cells (7.5 × 10⁴) were plated onto 6-well dishes in DMEM with 7% FBS. Following attachment of cells to the plate, serum starvation was performed in DMEM with 0.5% FBS. After 48 h, cells were released from starvation into DMEM with 7% FBS containing either DMSO or 40 μM MBV. To perform counting, cells were collected by dissociating into 250 μl of 1% trypsin (Life Technologies, Carlsbad, CA) and counted on a hemocytometer (Hausser Scientific, Horsham, PA) using a Nikon Eclipse TS100 inverted microscope set at the 10× objective lens (Nikon Inc., Melville, NY). Each biological sample was counted in technical replicate, and results were averaged.

Analysis of transactivation and viral entry.

For transactivation assays, 2.5 μg ADdel97 BAC DNA was cotransfected with a pCGN expression vector into 1.5 × 10⁵ MRC5 cells using electroporation or U373 cells using Fugene 6. Cells were plated onto 12-well dishes. At 7 days postinfection (dpi) (MRC5 cells) or 48 hpi (U373 cells), cells were fixed using 4% paraformaldehyde, permeabilized using 100% methanol, and blocked using 3% bovine serum albumin (BSA) dissolved in PBS-T. Cells were then incubated for 1 h with mouse anti-pUL123 (1:500 dilution in 3% BSA PBS-T). Cells were washed three times with PBS-T, incubated for 1 h with anti-mouse IgG Alexa 488 (1:1,000 dilution in 3% BSA PBS-T), and then washed three times with PBS-T. Finally, 3% SlowFade Gold Antifade Reagent with DAPI (4′,6-diamidino-2-phenylindole) (Life Technologies) in PBS-T was added to the wells. pUL123-positive cells in each transfection were counted using a Nikon Eclipse TS100 inverted microscope set at the 20× objective lens (Nikon Inc., Melville, NY). Following quantification, cells were prepared for images of pUL123 and hemagglutinin (HA) double-positive cells by incubating cells for 1 h with mouse anti-HA conjugated Zenon Alexa Fluor 568. Conjugation was performed using the Zenon Mouse IgG1 Labeling kit (Life Technologies) following the manufacturer's protocol. To assess entry, HFFs were infected at 1 IU/cell for 1 h, after which time they were fixed and permeabilized as described above. Fixed cells were incubated for 1.5 h with antibody against pUL83 (1:500 dilution in 3% BSA PBS-T). Cells were washed with PBS-T and then incubated with anti-mouse IgG Alexa 488 (1:1,000 dilution in 3% BSA PBS-T) for 1 h. For images containing DAPI, SlowFade with DAPI was added to the cells for 20 min and then washed with PBS-T before imaging using a 40× objective lens (Nikon Inc., Melville, NY).

Chromatin immunoprecipitation assay.

HFFs (5 × 10⁶) were plated onto one 150-mm dish as described in “Biological reagents” above. Cells were fixed and protein-chromatin complexes cross-linked in 1% ChIP grade formaldehyde (Thermo Fisher Scientific, Waltham, MA) for 10 min and neutralized with 125 mM glycine for 5 min. Three washes with ice-cold phosphate-buffered saline were performed before cells were scraped, collected, and lysed in 500 μl SDS lysis buffer (1% sodium dodecyl sulfate, 10 mM EDTA, 50 mM Tris-HCl [pH 8.1]) containing Complete Mini EDTA-free Protease Inhibitor Cocktail (Roche). Lysates were sonicated to ∼1 kbp using a microtip sonicator. Lysates were centrifuged at 18,000 × g at 4°C for 10 min to remove insoluble material. A fraction of supernatant was kept for the input measurement and to test shearing efficiency. The supernatant was precleared with 10 μl Dynabeads Protein G (Life Technologies) without antibody.

The supernatant was diluted with 5 volumes ChIP Dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 1.67 M Tris-HCl [pH 8.1], 176 mM NaCl). Supernatant was combined with 10 μg of antibody and incubated overnight at 4°C with rotation. Antibody complexes were precipitated with 60 μl Dynabeads Protein G for 2 h with rotation at 4°C. The bound complexes were collected by magnet and washed once with low-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.1], 150 mM NaCl), once with high-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.1], 500 mM NaCl), once with LiCl wash (250 mM LiCl, 1% IGEPAL-CA630 [octylphenoxypolyethoxyethanol], 1% Na deoxycholate, 1 mM EDTA), and twice with TE buffer (10 mM Tris-HCl [pH 8.1], 1 mM EDTA). Complexes were eluted from beads twice in 250 μl elution buffer (40 mM NaHCO₃, 1% SDS, 10 mM dithiothreitol [DTT]) at room temperature for 15 min. The beads were removed, and the eluate was reverse cross-linked using NaCl (final concentration of 4%, wt/vol) at 65°C for 12 h. Eluted proteins were digested in 10 mM EDTA and 40 mM Tris-HCl (pH 6.5) using 40 μg proteinase K (Roche) at 45°C for 1 h. DNA was isolated using the Gel/PCR DNA Fragment Extraction kit (IBI Scientific, Peosta, IA) per manufacturer protocols and eluted using 30 μl Milli-Q filtered H₂O (Millipore). Viral and cellular DNA levels were determined using quantitative PCR with sequence specific primers, as described above. PCR primers used were against the major immediate early promoter (MIEP) (5′-CTTACGGGACTTTCCTACTTG-3′ and 5′-CGATCTGACGGTTCACTAA-3′) or beta-actin (5′-CATTGCCGACAGGATGCA-3′ and 5′-GCCGATCCACACGGAGTACT-3′).

HDAC activity assays.

For HDAC activity assays, immunoprecipitation was performed as stated above with the following modifications. Cells were lysed in Triton X-100 lysis buffer (20 mM HEPES-K, 100 mM KOAc, 2 mM MgCl₂, 0.1% Tween 20, 1 μM ZnCl₂, 1 μM CaCl₂, 0.5% Triton X-100, 500 mM NaCl). Following washes, the bead-antibody-protein complex or lysate was subjected to the HDAC Fluor-de-Lys Activity assay according to the manufacturer's protocol (Enzo, Plymouth Meeting, PA). Fluorescence was analyzed on a Synergy2 plate reader (BioTek, Winooski, VT) per HDAC Fluor-de-Lys Activity assay manufacturer recommendations at an excitation wavelength of 485 nm, emission wavelength of 528 nm, and gain of 50. Protein was eluted from beads in Laemmli sample buffer. Lysates and immunoprecipitated sample were analyzed by Western blotting and FluorChem HD2 (Proteinsimple) for protein quantification.

Phosphopeptide enrichment and mass spectrometry.

Mass spectrometry experiments were initiated by transfecting U-2 OS cells using Fugene 6 as described above. Immunoprecipitation was then performed using the protocol outlined in the HDAC Activity assay methods. Following immunoprecipitation, protein was eluted from beads with 0.4 M NH₄OH at 25°C for 20 min with agitation. Disulfide bonds were reduced in 10 mM DTT at 25°C for 30 min, and free cysteines were alkylated in 10 mM iodoacetamide at 25°C for 30 min. Protein was digested with 1 μg of Trypsin Gold, mass spectrometry grade (Promega, Madison, WI) at 37°C for 12 to 15 h. Digestion was stopped by trifluoric acid added to a final concentration of 0.1%. The sample was enriched for phosphopeptides using PHOS-select Iron Affinity Gel (Sigma) according to the manufacturer's protocol. Supernatant and washes from the PHOS-select Iron Affinity Gel phosphopeptide enrichment were saved and further enriched for monophosphorylated peptides using Proteo Extract Phosphopeptide Enrichment TiO2 (Calbiochem, Gibbstown, NJ) according to the manufacturer's protocol. The eluates from the PHOS-select Iron Affinity Gel and Calbiochem Proteo Extract Phosphopeptide Enrichment TiO2 enrichments were combined and cleaned using ZipTip C₁₈ (Millipore) according to the manufacturer's protocol.

Mass spectral analysis was performed on an LTQ-Orbitrap Velos mass spectrometer (Thermo Scientific, Waltham, MA, USA) coupled with a NanoAquity UPLC (ultraperformance liquid chromatographer) (Waters Corporation) equipped with an autosampler and interfaced with a nanoelectrospray ion source. Peptide separation took place on a 5-μm C₁₈ nanocolumn (50 μm inner diameter [i.d.]] by 150 mm). The fused silica capillaries (Polymicro Technologies, Phoenix, AZ, USA) for the columns were pulled by a micropipette puller P-2000 (Sutter Instrument Company, CA, USA) and packed with C₁₈ resin using a bomb-loader. The solvents A and B used for chromatographic separation of peptides were 2% acetonitrile in 0.1% formic acid and 98% acetonitrile in 0.1% formic acid, respectively. The peptides injected onto the microcapillary column were resolved at the rate of 200 nl/min, using the following gradient conditions: 0 to 10 min, 2 to 20% B; 120 min, 40% B; 135 min, 60% B; 150 min, 70% B; 158 min, 98%B, which was held for 4 min and then switched to 98% A and held for another 18 min. The ions eluted from the column were electrosprayed at a voltage of 1.75 kV. The ion transfer temperature was kept at 250°C. No auxiliary or sheath gas was used.

Survey full-scan mass spectrometry (MS) spectra (m/z, 300 to 2,000) were recorded in the Orbitrap analyzer at a resolution of 30,000 followed by tandem mass spectrometry (MS/MS) of the 10 most intense peptide ions in the linear ion trap analyzer. Neutral loss-triggered multistage activation for simultaneous fragmentation of neutral loss product and precursor was enabled at m/z of −98, −80, −49, 40, −32.6, and −26.7 relative to the precursor ion, corresponding to a neutral loss of phosphate moiety from +1, +2, and +3 charged ions. The chromatographic and mass spectral functions were controlled using the Xcalibur data system (ThermoFinnigan, Palo Alto, CA, USA). Peptides were identified using the Mascot and SEQUEST search algorithm for combined human cytomegalovirus and human protein database at a global false-discovery rate of 5%. The resulting files were evaluated using Visualize (48).

Inhibition of the HCMV pUL97 kinase disrupts immediate early and early gene expression.

HCMV pUL97 acts at multiple stages during viral replication (49). To determine whether pUL97 could influence viral gene expression, we completed infections in the presence of the pUL97 kinase inhibitor, maribavir (MBV) (50). MBV is in clinical trials and functions as an ATP-competitive UL97 kinase inhibitor (51, 52). For these studies, primary human fibroblasts were infected at a multiplicity of 1 infectious unit (IU) per cell using HCMV strain AD169 (ADwt) in the presence or absence of 40 μM MBV (53). At the indicated times, we determined changes in viral gene expression using quantitative RT-PCR and normalized expression to cellular GAPDH RNA (Fig. 1A). The addition of MBV at the time of infection resulted in a significant decrease in expression of the immediate early gene UL123, with an average decrease of 1.4-fold at 6 hpi, 3.7-fold at 10 hpi, and 2.7-fold at 24 hpi (Fig. 1A). We also evaluated expression of two early HCMV genes, UL38 and UL44. Expression of UL38 decreased 2.5-fold at 10 hpi and 1.7-fold at 24 hpi, while UL44 decreased 2-fold at 10 hpi and 3.5-fold at 24 hpi (Fig. 1A). These data suggest that pUL97 kinase activity does contribute to efficient HCMV gene expression at early times during infection.

To confirm that the observed defect is due to pUL97 and not an unknown off-target effect of the compound, we repeated our studies using a recombinant HCMV mutant that contains the bacterial galK gene in place of the UL97 gene (ADdel97). Although the ADdel97 virus grew slowly, which is consistent with studies by other laboratories (9, 10, 54), we prepared the viral stocks in the absence of complementation. This method of preparation was used to avoid pUL97 being packaged into viral particles. We infected fibroblasts using a multiplicity of 1 IU/cell with either ADdel97 or ADwt. Again via qRT-PCR, we observed significant decreases in UL123 gene expression at 6 and 24 hpi using ADdel97. Furthermore, we quantified significant decreases in both UL38 and UL44 expression at 24 hpi. These data show that HCMV pUL97 contributes to efficient HCMV gene expression at immediate early and early times of infection.

To determine whether pUL97 could also influence steady-state viral protein levels, we completed Western blot analysis on whole-cell lysates following infection using either ADwt, ADwt with MBV, or ADdel97 virus. We observed reduced levels of HCMV pUL123 (IE1) as well as pUL122 (IE2) at 6 and 8 hpi following inhibition of kinase activity or upon removal of the UL97 gene (Fig. 1C). Treatment of the ADdel97 viral infection with MBV did not further reduce pUL123 expression at 8 hpi (Fig. 1D). By 24 hpi, however, we observed wild-type levels of pUL123 and pUL122 protein expression under each condition. We also detected differences in pUL44 expression at 24 hpi (Fig. 1C). For pUL38, we observed a difference in expression between ADwt and both ADwt with MBV and ADdel97 at 8 hpi but only ADdel97 at 24 hpi (Fig. 1C). ADdel97 exhibited larger defects than MBV-mediated inhibition, which is likely due to the multiple roles of pUL97 including virion assembly at late times (9, 10, 16, 21, 23, 54). For all of these studies, we determined the titers of our viral stocks by staining for pUL123-postive cells at 48 hpi. To ensure that the transient defect in viral gene expression was not affecting our ability to accurately determine viral titers, we infected fibroblasts at 1 IU/cell using stocks of ADwt or ADdel97 for which titers had been determined, and we quantified pUL123-positive cells at 48 hpi. The infections were completed using ADwt, ADwt with MBV, or ADdel97. Under these conditions, equivalent numbers of cells were positive between infections (Fig. 1E), indicating that the method is valid. Next, we completed an additional control to determine whether 40 μM MBV could influence cell cycle progression. HCMV pUL97 is a viral cyclin-dependent kinase (v-CDK) (12, 13, 20), and IE expression is sensitive to changes in the cell cycle. For this control, we serum starved subconfluent fibroblasts using 0.5% FBS for 48 h and released cells using 7% FBS with and without 40 μM MBV. At 24 hpi, we evaluated DNA content using flow cytometry as an indicator of the cell cycle phases. As shown in Fig. 1F, 40 μM MBV did not result in observable changes in cellular DNA content compared to the DMSO-treated control cells. We also evaluated cell doubling following release, and the presence of MBV did not alter cell doubling compared to DMSO-treated cells (Fig. 1G). Overall, our studies support the conclusion that HCMV pUL97 contributes to efficient HCMV gene expression beginning at immediate early times of infection.

While we suspected that pUL97 was promoting immediate early gene expression, disruption of kinase activity or deletion of the kinase could also influence entry. To control for possible viral entry defects, we infected cells again with 1 IU/cell using ADwt, ADwt with MBV, or ADdel97 virus and quantified viral genomes relative to a cellular gene at 1 hpi (Fig. 2A). We detected similar levels of viral DNA associated with the infected cells, suggesting that the defect was not due to differences in viral entry. To provide additional evidence, we completed immunofluorescence analysis using an antibody against the major HCMV tegument protein, pUL83 (pp65) (55). pUL83 is delivered to the nucleus upon entry and participates in regulating cellular responses to infection (56–59). Again at 1 hpi and 1 IU/cell, we observed similar numbers of cells between infections that were positive for the tegument protein (Fig. 2B). However, for the ADdel97 virus, the signal intensity of pUL83 staining was lower, which is likely due to packaging defects occurring during preparation of the viral stock in the absence of pUL97 (21, 23, 24). To address this possibility, we evaluated the virion protein and DNA composition in equal infectious units between ADwt and ADdel97. Using Western blot analysis, we observed in ADdel97 virions increased levels of pUL82 (pp71), pUL85 (mCP), pUL44, and pUL99 (pp28) but reduced levels of pUL83 compared to equivalent infectious units of ADwt (Fig. 2C). Looking at virion DNA content, we observed similar levels of viral DNA using qPCR and primers to the major immediate early promoter (MIEP) and UL54 sequences between viruses (Fig. 2D). Primers designed against the UL97 gene failed to amplify a signal from ADdel97 virions, confirming the absence of the gene (Fig. 2D). These data demonstrate that deletion of UL97 impacts ADdel97 virion protein levels but not DNA content when comparing equivalent infectious units of ADwt. Since multiple tegument proteins influence IE gene expression, these differences likely contribute to the greater defect observed for ADdel97 compared to MBV-treated wild-type virus.

Expression of pUL97 is sufficient to enhance immediate early UL123 expression from the viral genome.

We were next interested in determining whether expression of pUL97 was sufficient to promote HCMV immediate early gene expression. For these studies, we electroporated ADdel97 viral genomic DNA into fibroblasts with equal amounts of either an empty vector, a vector expressing pUL97, or a vector expressing the kinase-inactive version, pUL97K355M (13) (Fig. 3A). As a positive control, we also included an expression vector encoding the transactivating protein pUL82 (32). After 7 days posttransfection, we evaluated the cells for the expression of pUL123 protein from the HCMV genome. Using immunofluorescence analysis and an antibody specific to pUL123, we quantified a significant increase in pUL123-positive cells upon transfection of pUL97 in comparison to either empty vector or kinase-inactive pUL97K355M (Fig. 3A). Furthermore, we observed similar numbers of pUL123-positive cells between pUL97 and pUL82 transfections (Fig. 3A). We repeated the experiments using U373 human astrocytoma cells, which have a higher capacity to express transfected DNA. Cells were transfected using lipid-based transfection with ADdel97 genome along with the different expression vectors, and pUL123-positive cells were identified at 48 h posttransfection. Again, we quantified a significant increase in pUL123-positive cells upon transfection of the pUL97 expression vector in comparison to both empty and pUL97K355M vectors (Fig. 3B). In U373 cells, however, pUL82 promoted higher levels of pUL123-positive cells than pUL97 (Fig. 3B). Representative fluorescent images of pUL123-positive cells, which were later costained using an antibody against the HA epitope in pUL97 and pUL82, are shown in Fig. 3C. Taken together, our data demonstrate that active pUL97 kinase is sufficient to promote increased HCMV pUL123 protein expression from the viral genome.

HCMV pUL97 is a component of virions, and we next investigated whether tegument-delivered pUL97 could influence immediate early gene expression. For these studies, we cultured cells in the presence or absence of cycloheximide starting 1 h prior to infection in order to inhibit viral protein translation (60). Cells were then infected at 1 IU/cell using ADwt, ADwt with MBV, or ADdel97. To ensure that viral protein synthesis was inhibited, we collected whole-cell lysates at 8 hpi and studied pUL123 protein levels using Western blot analysis. Under these conditions, cycloheximide treatment inhibited pUL123 protein synthesis to below detectable levels (Fig. 3D). In the absence of cycloheximide, we detected pUL123 during ADwt infection but reduced levels in MBV-treated ADwt and ADdel97 infections, as expected. To assess the impact of cycloheximide treatment on viral RNA levels, we pretreated cells as described above, infected them at 1 IU/cell, and harvested total RNA at 6 hpi. Using qRT-PCR, we observed significant decreases in UL123 RNA levels in the ADdel97 and MBV-treated ADwt infections compared to those in ADwt infection alone (Fig. 3D). These results indicate that tegument-derived pUL97 is contributing to the activation of UL123 IE gene expression upon infection.

pUL97 interacts with HDAC1 and alters histone H3 acetylation at the HCMV MIE promoter.

In our previous studies, we observed an association between pUL97 and HDAC1 during HCMV infection using mass spectrometry and HDAC1 as the bait protein (25). We were next interested in investigating the role of HDAC1 in pUL97-mediated regulation of HCMV gene expression. To determine whether the interaction between pUL97 and HDAC1 can also occur in the absence of infection, we transfected U-2 OS osteosarcoma cells with an HDAC1 expression vector containing the FLAG epitope along with vectors for either pUL97 or pUL97K355M containing the HA epitope. We immunoprecipitated HDAC1 and any associated proteins using an antibody against the FLAG epitope and completed Western blot analysis using an antibody against HA. We detected proteins corresponding to the correct molecular weight of pUL97 and pUL97K355M (Fig. 4A). We did not observe these proteins in the absence of either pUL97 expression vectors or HDAC1-FLAG following immunoprecipitation with the FLAG antibody (Fig. 4A). Our data demonstrate that pUL97 as well as inactive-kinase pUL97K355M can associate with HDAC1 and confirm our previous observation (25).

We postulated that pUL97 may be acting to increase gene expression through opposing HDAC1-mediated repression of immediate early transcription. To test this possibility, we determined the protein levels of pUL123 during infection under several conditions. Again, infected cells were treated with MBV to inhibit pUL97 kinase activity, resulting in a delay in HCMV pUL123 expression (Fig. 4B). To evaluate the impact of deacetylase activity, we included the pan-HDAC inhibitor, trichostatin A (TSA), which has been previously demonstrated to increase immediate early gene expression (61). When 300 nM TSA was combined with MBV, we observed a partial rescue of pUL123 expression, which was most evident at 8 hpi (Fig. 4B). To confirm these results, we repeated the experiment using the deacetylase inhibitor mocetinostat (MGCD0103), at 0.8 μM, which specifically disrupts HDAC1 and 2 activity (50% inhibitory concentration [IC₅₀] for HDAC1, 0.15 μM; for HDAC2, 0.29 μM; and for HDAC3, 1.66 μM) (62). Under these conditions, the addition of mocetinostat also resulted in a rescue of MBV-mediated or UL97 substitution-mediated disruption of pUL123 expression at 8 hpi (Fig. 4C). Together, these studies provide evidence that pUL97 is opposing the repressive effects of HDAC1 on HCMV IE gene expression.

One way in which HDACs repress transcription is by deacetylating histones associated with promoters (63–67). To determine whether pUL97 could alter the level of histone acetylation at the HCMV MIE promoter, we completed chromatin immunoprecipitation (ChIP). We focused our studies on histone H3 lysine 9 acetyl (H3K9ac), which is a mark of active transcription (68). For these experiments, we infected fibroblasts at a multiplicity of 1 IU/cell using ADwt, ADwt plus MBV, or ADdel97 and completed ChIP at 1 and 6 hpi using an antibody against H3K9ac. Infections involving both ADdel97 and MBV treatment of ADwt virus resulted in decreased MIE promoter sequence associating with H3K9ac at 1 and 6 hpi compared to untreated wild-type virus (Fig. 4D). We detected no differences between each infection using nonspecific IgG (Fig. 4D). As a control, we used primers to the beta-actin gene, and although the data were highly variable, we observed no statistically significant difference in H3K9ac binding at 1 hpi (Fig. 4D). However, we did observe elevated levels of beta-actin sequence in the MBV-treated and ADdel97 samples at 6 hpi (Fig. 4D). Possible explanations include loss of pUL97-mediated changes directly at this gene or indirect effects due to changes in immediate early gene expression. However, we did not observe changes in beta-actin protein levels at this time (Fig. 4B). Overall, our results suggest that disruption of pUL97 is associated with decreased amounts of H3K9ac associated with the HCMV MIE promoter. Furthermore, these data support the hypothesis that pUL97 is disrupting HDAC1-mediated repression.

pUL97 kinase activity alters HDAC1 association with the MIE promoter.

Park et al. (36) have reported that total HDAC2 deacetylase activity decreases early during HCMV infection. We were interested in determining the impact of pUL97 kinase activity on deacetylase activity in the presence and absence of infection. Using U-2 OS cells, we transfected HDAC1 containing the FLAG epitope in the presence of either empty, pUL97, or pUL97K355M expression vectors. HDAC1 was isolated by immunoprecipitation, and the associated deacetylase activity was quantified using a fluorometric HDAC activity assay (Fig. 5A) and normalized to the amount of HDAC1 protein isolated as measured by chemiluminescence from the Western blot (Fig. 6A). Although we detected an elevation in activity following transfection and immunoprecipitation of HDAC1, neither pUL97 nor pUL97K355M appeared to influence deacetylase activity (Fig. 5A). To evaluate activity during infection, we infected fibroblasts using ADwt, ADwt plus MBV, or ADdel97 virus and again immunoprecipitated HDAC1. We quantified deacetylase activity relative to the total amount of HDAC1 isolated as determined by chemiluminescence analysis (Fig. 5B, lower panel). At 1 hpi, we did not observe differences in HDAC1-associated activity between samples (Fig. 5B). At 6 hpi, we observed a significant decrease in activity when comparing infected samples to either the uninfected mock or mock plus MBV samples. However, no difference was observed between ADwt and ADwt plus MBV or ADdel97 infections. These data demonstrate that HDAC1-associated deacetylase activity is decreased at 6 hpi during HCMV infection but this occurs independently of HCMV pUL97.

In general, HDAC1 associates with complexes on chromatin but does not directly bind to DNA to mediate transcriptional repression (65). Furthermore, Reeves et al. have demonstrated that HDAC1 associates with the HCMV MIE promoter after 24 hpi in conjunction with IE2 to mediate repression at the cis-repressive sequence (CRS) (37). An alternative hypothesis explaining our observations is that pUL97 alters the association of HDAC1 with the promoter. To evaluate this possibility, we utilized ChIP analysis to quantify changes in endogenous HDAC1 associating with the MIE promoter at 6 hpi. Following the addition of MBV to ADwt infection, we quantified an average increase of 5.7-fold in MIE promoter sequence associating with HDAC1 compared to the untreated infection (Fig. 5C). We repeated the experiment using the ADdel97 virus and quantified a 16-fold increase relative to infection using ADwt (Fig. 5C). This response was not observed for the beta-actin gene (Fig. 5C). These data demonstrate that inhibition of pUL97 activity results in elevated levels of HDAC1 at the MIE promoter. These changes correlate with the observed decrease in H3K9ac levels at the promoter.

HCMV pUL97 kinase-dependent phosphorylation of HDAC1 and reduced DAXX interaction.

HDAC1 is recruited to DNA through interactions with multiple proteins (reviewed in references 26 and 69). With this in mind, we evaluated changes in HDAC1's association with several known binding partners upon pUL97 expression. The interactions of HDAC1 with MTA2, Sin3A, and Rbbp48 were unchanged when pUL97 was coexpressed in U-2 OS cells compared to either empty vector or expression of pUL97K355M (data not shown). HDAC1 has also been demonstrated to interact with the transcriptional repressor DAXX (33). DAXX acts to silence HCMV IE expression in a deacetylase-dependent manner (32, 33). To evaluate the DAXX interaction with HDAC1, we transfected U-2 OS cells with a vector expressing HDAC1-FLAG and either empty or pUL97 or pUL97K355M expression vectors (Fig. 6A). Following immunoprecipitation with an antibody against the FLAG epitope, we detected low levels of DAXX by Western blotting using empty vector or pUL97K355M (Fig. 6A). Although the effect is not robust, we consistently observed a reduction in DAXX associating with HDAC1 upon expression of active pUL97 kinase (Fig. 6A). During infection, steady-state levels of DAXX are decreased at 1 hpi and are substantially decreased by 2 hpi in a pp71-dependent manner (32). We therefore evaluated the influence of pUL97 activity on the interaction between DAXX and HDAC1 at 1 hpi. U373 cells were transfected with a HDAC1-FLAG expression vector and were either mock infected or infected at 3 IU/cell using ADwt virus. Furthermore, cells were treated with DMSO or with 40 μM MBV. Immunoprecipitation with an anti-Flag antibody followed by analysis using Western blot revealed a small but consistent increase in DAXX association with HDAC1 during MBV treatment of infected cells (Fig. 6B) compared to DMSO control. These data show that pUL97 promotes a reduction in the association between HDAC1 and DAXX.

In the above-described experiment, we observed a higher-molecular-weight band for HDAC1 in the infected but not the MBV-treated sample (Fig. 6B), suggesting the possibility that HDAC1 is being modified in a kinase-dependent manner. Furthermore, kinases from alphaherpesviruses have been shown to phosphorylate HDAC1 and/or 2 (40, 41, 70). Therefore, we decided to investigate whether HCMV pUL97 could influence the phosphorylation state of HDAC1. For these experiments, we completed Western blot analysis using whole-cell lysates from U-2 OS cells cotransfected with vectors expressing HDAC1 and either wild-type pUL97 or kinase-inactivated pUL97K355M (Fig. 6C). As a control for kinase activity, we observed substantial autophosphorylation occurring with the wild-type kinase but not K355M when using an antibody against the HA epitope in pUL97 (Fig. 6C). When we used an antibody against the FLAG epitope in HDAC1, we detected a band with increased mobility following transfection with wild-type pUL97 but not with the K355M mutant (Fig. 6C), which suggests that HDAC1 is being modified in the presence of pUL97.

To determine whether the modification is the result of pUL97-mediated phosphorylation, we repeated the experiment in the presence or absence of phosphatase treatment. For these studies, we cotransfected U-2 OS cells with a vector expressing HDAC1 and either wild-type pUL97, pUL97 plus MBV, or pUL97K355M expression vector. We isolated whole-cell lysates and treated one half of the sample with calf intestinal phosphatase. In the untreated sample half expressing both HDAC1 and pUL97, we again detected a modification of HDAC1 by Western blotting (Fig. 6D). Treatment of this sample with phosphatase resulted in a reduction in the band intensity, suggesting that this modification is sensitive to phosphatase treatment (Fig. 6D). We did not observe the higher-mobility species of HDAC1 either in the control transfection, in MBV-treated samples, or when using the inactive kinase (Fig. 6D). Our results show that HCMV pUL97 is promoting the phosphorylation of HDAC1.

Identification of amino acids in HDAC1 being phosphorylated by viral and cellular kinases.

We were next interested in trying to identify the amino acids in HDAC1 that might be differentially phosphorylated by the HCMV kinase. For these studies, we used a phosphopeptide enrichment strategy in-line with a high mass accuracy mass spectrometer (Fig. 7A). The expression vectors for pUL97 and HDAC1 were transfected into U-2 OS cells. As controls, we also transfected HDAC1 with either an empty vector or the pUL97 plus MBV or pUL97K355M expression vector. HDAC1 and associated proteins were isolated by immunoprecipitation followed by in-solution tryptic digestion and phosphopeptide enrichment using immobilized metal ion affinity chromatography (IMAC) and TiO₂. Sites of phosphorylation were identified using a high mass accuracy LTQ-Orbitrap mass spectrometer. Under these conditions, we observed peptides to pUL97 associating with HDAC1 under all conditions except when using the empty expression vector, providing additional evidence for the association of pUL97 with HDAC1 (data not shown).

Sites of phosphorylation were identified in peptides by the presence of an 80-Da (HPO₃⁻) or 98-Da (H3PO₄⁻) mass shift using the Sequest and Mascot algorithms. The experiments were completed in triplicate, and spectra containing the shift were manually validated. In cells receiving HDAC1 with the empty vector control, we detected 64% of the tryptic peptides representing 49% of the total HDAC1 protein sequence and observed several sites of phosphorylation (Fig. 7B). These included serine 85 (S85), S409 or S410 (S409/S410), S421, and S423 and are the result of cellular kinase activity, since pUL97 was not included in these samples (Fig. 7B). For S409/S410, we were unable to determine which of the two neighboring amino acids was phosphorylated. Phosphorylation at amino acids S409/S410, S421, and S423 has been previously observed by other laboratories (71–73), but S85 represents a newly identified site of phosphorylation. A representative spectrum for S85 is shown in Fig. 7C. The same amino acids were observed to be phosphorylated in samples with pUL97, MBV plus pUL97, and the inactive kinase (Fig. 7B). Unique to only the sample expressing the active pUL97 kinase, we observed phosphorylation of HDAC1 at threonine 208 (T208) (Fig. 7B). Examples of the spectra for this modification are presented in Fig. 7C, and this also represents a newly identified site of phosphorylation in HDAC1. We did observe the unmodified form of this peptide within the control, MBV-treated, and inactive-kinase samples (Fig. 7B). These studies have identified precise amino acids being phosphorylated in HDAC1 in the presence and absence of HCMV kinase activity, and this includes two previously unknown sites of phosphorylation, S85 and T208.

To assess the role of phosphorylation occurring at T208 in modulating the HDAC1-DAXX interaction, we constructed substitution mutants of HDAC1 at T208 using either alanine to prevent phosphorylation (T208A) or a glutamic acid (T208E) or aspartic acid (T208D) to mimic phosphorylation. These expression vectors were cotransfected into U-2 OS cells with either empty vector or the pUL97 kinase. We immunoprecipitated HDAC1 using an antibody against FLAG and evaluated changes in association between HDAC1 and DAXX by Western blotting (Fig. 7D). In the absence of pUL97 expression, changes to HDAC1 T208 did not alter the HDAC1-DAXX interaction. Upon pUL97 expression, we again observed a reduction in the amount of DAXX associating with HDAC1 (Fig. 7D). However, this reduction also occurred when using either the T208A or the T208E HDAC1 mutant (T208D was not tested) (Fig. 7D). Increased pUL97-dependent phosphorylation was observed for all constructs (Fig. 7D). These data confirmed the observations that expression of pUL97 increases phosphorylation of HDAC1 and disrupts the HDAC1-DAXX interaction yet phosphorylation at T208 is not necessary for the changes.