The Late Promoter of the Human Cytomegalovirus Viral DNA Polymerase Processivity Factor Has an Impact on Delayed Early and Late Viral Gene Products but Not on Viral DNA Synthesis

Abstract

Transcription of the DNA polymerase processivity factor gene (UL44) of human cytomegalovirus initiates at three distinct start sites, which are differentially regulated during productive infection. Two of these start sites, the distal and proximal sites, are active at early times, and the middle start site is active at only late times after infection (F. Leach and E. S. Mocarski, J. Virol. 63:1783-1791, 1989). Compared to the wild type, UL44 gene expression was lower for recombinant viruses with the distal or the middle TATA element mutated. The transcripts initiating from the distal or middle start site facilitated late viral gene expression. The level of viral DNA synthesis was affected by mutation of the distal TATA element. In contrast, mutation of the middle TATA element did not affect the level of viral DNA synthesis, but it did affect significantly the level of late viral gene expression. Recombinant viruses with the distal or middle TATA element mutated grew more slowly than the wild type at both low and high multiplicities of infection. Reduced expression of the UL44 gene from the late middle viral promoter correlated with decreased late viral protein expression and decreased viral growth.

Human cytomegalovirus (HCMV) is a member of the betaherpesvirus family. Like all herpesviruses, HCMV is an enveloped, double-stranded DNA virus. The genome of HCMV is 240,000 bp, with at least 150 known open reading frames (ORFs) (6). A majority of the ORFs are nonessential for viral replication in cell culture. These nonessential ORFs likely encode proteins with redundant functions or proteins that may be required for replication in the human host. In addition, several ORFs are beneficial but not required for replication. However, approximately one-quarter, or 41 ORFs, are absolutely required for viral replication (43). The UL44 gene is essential.

During productive infection, HCMV genes are expressed in a temporal cascade, designated immediate early (IE), delayed early, and late. The major IE (MIE) genes UL123 and UL122 (IE1/IE2) play a critical role in subsequent viral gene expression and the efficiency of viral replication (15-17, 24-26). The IE72 protein, the predominant product of the IE1 transcript, is encoded by exons 2 and 3 spliced to exon 4. The IE86 protein, the predominant product of the IE2 transcript, is encoded by exons 2 and 3 spliced to exon 5. Translation of the IE1 and IE2 transcripts begins in exon 2. The IE72 protein is not essential for viral replication at a high multiplicity of infection (MOI), but the IE86 protein is essential (22). The early viral genes encode proteins necessary for viral DNA replication (29). Following viral DNA replication, delayed early and late viral genes are expressed which encode structural proteins for virion production. Several early genes of HCMV have the unusual property of three promoters, two that initiate transcription early and one that initiates transcription late (2, 20). The reason for the late promoter is not understood. We selected the UL44 gene to determine the role of a late promoter during HCMV replication.

The UL44 protein (pUL44), which binds double-stranded DNA, is an essential accessory protein for viral DNA replication and interacts specifically with the viral DNA polymerase encoded by UL54 (30, 34). pUL44 increases processivity of the polymerase along the viral DNA template (8, 40, 45). pUL44 accumulates to strikingly high levels at late times after infection (10, 35). Its late kinetics of transcription and the high level of expression suggest an additional important role for viral replication. pUL44 is phosphorylated by the viral UL97 protein kinase (pUL97) in infected cells (18). Phosphorylation by pUL97 is not required for pUL44 to interact with the catalytic subunit of the viral DNA polymerase (8, 40).

The HCMV UL44 transcription unit initiates at three distinct sites, which are separated by approximately 50 nucleotides and are differentially regulated during productive infection. Two of these start sites, the distal and the proximal sites, are active at early times, whereas the middle start site is inactive until late times (20). Expression from the late start site is dependent upon viral DNA replication.

We investigated whether the late start site is necessary for efficient viral replication in human fibroblast cells. Here we report that the UL44 gene product from the late viral transcript is required for efficient viral gene expression rather than viral DNA synthesis. The reasons why the product from the late viral transcript facilitates late gene expression are discussed.

MATERIALS AND METHODS

Cells and virus titration.

Primary human foreskin fibroblasts (HFFs) were maintained in Eagle's minimal essential medium supplemented with 10% fetal calf serum (Sigma, St. Louis, MO), penicillin (100 U/ml), and streptomycin (100 μg/ml) at 37°C in 5% CO₂ as described previously (35). The titers of wild-type (wt) HCMV Towne and recombinant viruses were determined by standard plaque assays on HFFs as described previously (26). Viral DNA input was determined by infecting HFFs in 35- or 60-mm plates in triplicate and harvesting the cells at 4 h postinfection (p.i.) in PCR lysis buffer (10 mM Tris-HCl [pH 8.0], 1 mM EDTA, 0.001% Triton X-100, and 0.001% sodium dodecyl sulfate [SDS]) containing 50 μg/ml proteinase K. After 55°C for 100 min, the proteinase K was inactivated at 95°C for 10 min. The relative amount of input viral DNA was estimated by real-time PCR using HCMV gB primers and probes as described previously (17). For analysis of virus growth kinetics, cells were infected at a multiplicity of 0.01 or 1 PFU/cell with the wt and the recombinant viruses. At various times after infection, whole cells and supernatant including free virus were collected from infected cultures, and three freeze-thaw cycles were performed before titration. Virus titers were determined by the 50% tissue culture infectious dose (TCID₅₀) assay on HFFs as described previously (27, 28, 37) except that we counted green fluorescent protein foci instead of assessing cytopathic effect. We used the Reed-Muench method to calculate TCID₅₀. The wt and recombinant viruses contain the green fluorescent protein gene in place of the dispensable, 10-kb US1-US12 region (US, unique short).

Enzymes.

Restriction endonucleases were purchased from New England Biolabs Inc. (Beverly, MA). High-fidelity and expanded high-fidelity Taq DNA polymerases were purchased from Invitrogen (Carlsbad, CA) and Roche (Mannheim, Germany), respectively, and RNasin and RNase-free DNase were purchased from Promega (Madison, WI). The enzymes were used according to the manufacturers' instructions.

Mutagenesis of HCMV BAC DNA.

A rapid homologous recombination system in Escherichia coli expressing bacteriophage lambda recombination proteins exo, beta, and gam (provided by D. Court, National Institutes of Health, Bethesda, MD) was employed as described previously (7). Bacterial artificial chromosome (BAC) DNA of HCMV Towne was obtained from F. Liu (University of California, Berkeley) (6). Double-stranded DNAs for recombination contained a kanamycin-resistant gene flanked by the 34-bp minimal FLP recognition target (FRT) sites (5′-GAAGTTCCTATTCTCTAGAAAGTATAGGAACTTC-3′) and 70 bp of homologous viral DNA sequence. To generate deletions of the UL45 ORF and each TATA box of the UL44 promoter (see Fig. 1), the following primer pairs were used: BACdlUL45KanF (5′-AGCGCGGAAACCGAGACGGAGGAATCGTCGGCAGAGGTCGCCGCTGATACTATCGGGGGAAACTCAGCAAAAGTTCGATTTATTCAAC-3′) plus BACdlUL45KanR (5′-CACGTCGGGTAGATGCAGCTTCTGCTCGTCGCTACGCGCAAACACGCAGCGAGCCACGTTTAATGCTCTGCCAGTGTTACAACCA-3′), BACUL44dlTATA1KanFRTF (5′-TCGGGGATGACGCCCGACGTGCTTCTGGCCAGGATGCTCAAGTGGTACCACTGGCGCTTTAAGGTCGGAGGAAGTTCCTATTCTCTAGAAAGTATAGGAACTTCAACTCAGCAAAAGTTCGATTTATTCAAC-3) plus BACUL44dlTATA1KanFRTR (5′-GCGAAGACGCCCGGCGTCTAATAATACAGCCGCGCCGAGCCAGCGGGCCCCCGACTAAGAGGCACAGTACGAAGTTCCTATACTTTCTAGAGAATAGGAACTTCTAATGCTCTGCCAGTGTTACAACCA-3′), and BACUL44dlTATA2KanFRTF (5′-TGGCGCTTTAAGGTCGGAGTATATAAGTACTGTGCCTCTTAGTCGGGGGCCCGCTGGCTCGGCGCGGCTGGAAGTTCCTATTCTCTAGAAAGTATAGGAACTTCAACTCAGCAAAAGTTCGATTTATTCAAC-3′) plus BACUL44dlTATA2KanFRTR (5′-GCGCGAGCGAGCGAAAGTTTTATAGAGAGCACACACGACGACCGGGAACGCTGCGAAGACGCCCGGCGTCGAAGTTCCTATACTTTCTAGAGAATAGGAACTTCTAATGCTCTGCCAGTGTTACAACCA-3′). Amplification by PCR was as follows: 1 cycle of denaturation at 94°C for 2 min; 40 cycles of denaturation at 94°C for 15 s, annealing at 55°C for 30 s, and extension at 72°C for 5 min; 1 cycle of extension at 72°C for 7 min or at 94°C for 2 min; 30 cycles at 94°C for 2 min, at 55°C for 2 min, and at 72°C for 2 min; and 1 cycle at 72°C for 7 min. To remove residual template DNA, the PCR products were digested with DpnI at 37°C for 1.5 h. The DNAs were phenol-chloroform extracted and precipitated with 95% ethanol. Approximately 100 ng of each DNA fragment was subjected to electroporation into competent E. coli DY 380 containing HCMV Towne BAC DNA. Electroporation was performed with a Bio-Rad Gene Pulser III (2.5 kV, 200 Ω, and 25 μF) following the kind suggestion of W. Dunn and F. Liu (University of California, Berkeley) (5). Bacteriophage-encoded recombination proteins for homologous recombination were induced at 42°C for 15 min as described previously (6).

FIG. 1.

Structure of recombinant HCMV BAC DNAs. (a) Schematic illustrations of parental and recombinant BAC DNA. Part of the UL45 ORF was replaced with a kanamycin-resistant gene (kan) by a PCR-based rapid recombination system. (b) Recombinant viruses with TATA element mutations. UL44 transcription initiates at three distinct sites (labeled 1, 2, and 3), which are separated by approximately 50 nucleotides. Two of these start sites, the distal and proximal sites, were active at early times after infection, whereas the middle start site was inactive until late times. The distal or middle TATA element was replaced by kan with flanking FRT sequence. kan was excised by FLP-mediated recombination.

Excision of the kanamycin-resistant gene.

To delete the kanamycin-resistant gene, the recombinant HCMV BAC DNA was transformed into E. coli DH10B. Plasmid pCP20 (provided by G. Hahn, Max von Pettenkofer Institute, Munich, Germany) (13) was transformed into DH10B containing the recombinant HCMV BAC DNA. HCMV BAC DNA without kanamycin was selected on LB plates containing ampicillin and chloramphenicol.

Recombinant virus isolation.

HFFs were transfected with either 5 or 10 μg of each recombinant BAC in the presence of 2 g of plasmid pSVpp71 by the calcium phosphate precipitation method of Graham and Van der Eb (11). After 5 to 7 days of 100% cytopathic effect, the extracellular fluid-containing virus was stored at −80°C in 50% newborn calf serum until used.

PCR analysis.

To select the recombinant BAC DNA with deletion of the UL45 ORF and replacement of each TATA box of the UL44 promoter, PCR analysis was performed using the following primer pairs: UL44 inner primer (5′-GGATAGCCGTCTTGTACGGCTTCA-3′) plus KanR (5′-CGCATCAACAATATTTTCACCTGAATC-3′) and UL44promoterF (5′-GCGATCCAAAACGACGTGGAAATGGCG-3′) plus UL44promoterR (5′-TGAGCGCACGGATCACAGATCGC-3′). The PCR cycling program was 1 cycle of denaturation at 94°C for 2 min; 30 cycles of denaturation at 94°C for 15 s, annealing at 55°C for 30 s, and elongation at 72°C for 1 min; and 1 cycle of elongation at 72°C for 5 min. A PCR product was cloned into a pCR 2.1-TOPO TA cloning vector (Invitrogen) and sequenced to confirm the recombination and excision (Aichi Cancer Center Research Institute Central Facility).

Southern blot analysis.

Recombinant BAC DNAs were purified using a NucleoBond kit (Macherey-Nagel, Duren, Germany), digested with the restriction endonuclease XhoI, and subjected to 1.0% agarose gel electrophoresis as described previously (39). Southern blot analysis was performed as described previously (26). UL45 probe and UL44 promoter probe were amplified by PCR using the primer pairs UL45 detectF2 (5′-TCCACCTTCACCATGTCTACCGTGG-3′) plus UL45detectR2 (5′-CGTCCGTGTTCACGACCACCTTG-3′) and UL44 promoterF plus UL44 promoterR as described above, respectively. The amplified probes were labeled using the Megaprime DNA labeling system (Amersham, Piscataway, NJ) and [³²P]dCTP (Amersham).

RNase protection assay.

For construction of the antisense UL44 probe, a DNA fragment including an upstream region of the transcription start site of all the UL44 transcripts was amplified by PCR using the primer pair UL44 F2 RNase protection assay primer (5′-CCGCTGGCTCGGCGCGGCTG-3′) plus UL44 inner primer as described above and cloned into the TA cloning vector pCRII (Invitrogen). The resulting clone, pUL44 pro-5, was linearized with EcoRV and used as a template for SP6 RNA polymerase. Synthesis by SP6 RNA polymerase on linear pUL44 pro-5 DNA produced a ³²P-labeled antisense RNA probe in agreement with the predicted size. Cytoplasmic RNAs from mock-infected or HCMV-infected HFFs were isolated at various times after infection as described previously (2, 14). DNA replication was inhibited with 200 μg/ml phosphonoacetic acid (PAA) (Sigma, St. Louis, MO) added to the medium at the time of infection and maintained throughout infection. Twenty micrograms of RNA was hybridized to ³²P-labeled antisense UL44 promoter probe at 37°C overnight before digestion with RNase T1 (100 U) as described previously (15, 19). The protected RNA fragments were subjected to electrophoresis in denaturing 6% polyacrylamide gels followed by autoradiography on Hyperfilm MP (Amersham).

Western blot analysis.

Cells were harvested at various times p.i., washed with PBS, and treated with lysis buffer as described previously (19, 20). Twenty-microgram aliquots of proteins were loaded in each lane for SDS-10% polyacrylamide gel electrophoresis and then transferred to polyvinylidene difluoride membranes. To detect the proteins pIE72 and pIE86 (encoded by IE1 and IE2, respectively), pp52 (encoded by UL44), and pp28 (encoded by UL99), the primary mouse monoclonal antibodies NEA-9221 (Perkin Elmer, Boston, MA), M0854 (Dako, Carpinteria, CA), and CA 004 (EastCoast Bio, North Berwick, ME), respectively, were used. To detect cellular GAPDH as a loading control, the polyclonal antibody MAB376 (Chemicon, Temecula, CA) was used. Enhanced chemiluminescence detection reagents (Amersham) and secondary horseradish peroxidase-labeled anti-mouse immunoglobulin G antibody (Zymed, San Francisco, CA) were used according to the manufacturer's instructions. Signal intensities were quantified with a LumiVision Image analyzer (Aisin/Taitec Inc., Tokyo, Japan).

Assay for immunostaining.

Confluent HFFs in 12-well multiwell plates were infected with the wt and recombinant viruses at an MOI of approximately 1. After 1 day p.i., cells grown on glass coverslips were fixed with 2% paraformaldehyde for 60 min at 4°C, and immunostaining was performed. Cells were washed three times with high-salt TPBS (0.01 M sodium phosphate [pH 7.3], 0.5 M NaCl, 0.1% Tween 20, and 0.1% Triton X-100) and incubated with the primary monoclonal antibody NEA-9221 for 1 h at 37°C. After three washes with PBS at room temperature, the cells were incubated with secondary goat anti-mouse immunoglobulin G antibody conjugated with Alexa Fluor 594 (Invitrogen). The immunofluorescence-positive cells in one field (magnification, ×40) per sample were counted under an inverted UV microscope with a digital camera (Eclipse TE-2000-S; Nikon, Tokyo, Japan) using ImageJ software (version 1.37) (http://rsb.info.nih.gov/ij/download.html). The results are means ± standard deviations from four independent experiments.

Viral DNA replication assay.

After infection with an MOI of 1 or 0.01, cells were collected at 2, 3, and 4 days p.i. or 5 and 8 days p.i., respectively. Cells in 35-mm plates in triplicate were suspended in lysis buffer (50 mM Tris-HCl [pH 8.0], 10 mM EDTA, 1% SDS, and 20 μg/ml RNase A) containing 50 μg/ml proteinase K. The replicated viral DNA was quantitated by real-time PCR using HCMV gB primers and probes as described previously (16, 17). Real-time PCR with 18S primers and probe purchased from Applied Biosystems (Foster City, CA) was also performed as an internal control for input DNA. Data are averages of three independent experiments.

Northern blot analysis.

Twenty micrograms of cytoplasmic RNA was subjected to electrophoresis in a 1% agarose gel containing 2.2 M formaldehyde and transferred to maximum-strength Hybond N+ (Amersham). Northern blot analysis with the IE1 probe was performed as described previously (14, 15). UL44, UL99, and UL75 DNA was amplified by PCR using the primer pairs UL44 ORFF (5′-TGCAGGACATCTCGGACCTGTCGG-3′) plus UL44ORFR (5′-CCAGACGCTGCTCAATTGCGCCG-3′), UL99F (5′-TGTGAGTTCGGTACCACGCCCGG-3′) plus UL99R (5′-CGTCTAGGTCGTCCGTCTCCTGC-3′), and UL75F (5′-CTGCGAAAAAGATCGGTAGCTGGCC-3′) plus UL75R (5′-CGCTGGACCCTCACGCATTTCACCTA-3′), respectively. A radioactive probe was generated by labeling with [³²P]dCTP as described above.

Real-time RT-PCR analysis.

Reverse transcriptase (RT) (Roche Applied Science, Penzberg, Germany) was used according to the manufacturer's directions to generate first-strand cDNA from 2 μg of RNA and 250 ng of oligo(dT) primer (Roche) in a final volume of 40 μl. Samples were heat inactivated at 80°C for 5 min. Amplifications were performed in a final volume of 10 μl containing Platinum quantitative PCR Supermix-UDG cocktail (Invitrogen). Each reaction mixture contained 1 μg of the first-strand cDNA, 5 mM MgCl₂, 500 nM each primer, and 250 nM each probe. MIE primers and the MIE reporter probe were designed as described previously (24). HCMV UL44, UL92-99, and UL75 forward and reverse primers and reporter probes were designed using Primer Express (Applied Biosystems) as follows: UL44F, 5′-TTTTCTCACCGAGGAACCTTTC-3′; UL44R, 5′-CCGCTGTTCCCGACGTAAT-3′; UL44 probe, 5′-6-FAM-AGCGTGGCGATCCCTTCGACAA-tetramethyl rhodamine (TAMRA)-3′; UL99-255F, 5′-CCACGACGGCTCCAAGAA-3′; UL99-319R, 5′-TCGGTTTCGGAGCCTTGTC-3′; UL99-275T probe, 5′-6-FAM-ACGCGGTGCGCTCGACGTT-TAMRA-3′; UL75-272F, 5′-TCCATATGCCTCGATGTCTTTT-3′; UL75-339R, 5′-GGTCAGATCTACCTGGTTCAGAAAC-3′; and UL75-296T probe, 5′-6- FAM-TTGGGCAACCACCGCACTGAGG-TAMRA-3′ (Nihon Gene Research Laboratories Inc., Sendai, Japan). Thermal cycling conditions were an initial step at 50°C for 2 min and 95°C for 2 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Relative RNA was quantitated according to a standard curve analysis as described previously (24). Real-time PCRs with glucose-6-phosphate dehydrogenase primers and probe as described previously (41) were also performed to serve as an internal control for input RNA. Each real-time RT-PCR assay was performed in triplicate and standardized to threshold cycle values for each viral RNA from HFFs infected with the wt virus at 4 days p.i.

Plasmid construction.

Plasmid pCAT3-Bacic vector purchased from Promega has multiple cloning sites upstream of the chloramphenicol acetyltransferase (CAT) gene. The sequence of the UL75 (gH) promoter (from −453 to + 103) or the UL99 (pp28) promoter (from −601 to + 94) was amplified from BAC DNA of HCMV Towne as described above by PCR using the primer pairs NheIUL75promoterF plus BglIIUL75promoterR and NheIUL99promoterF plus BglIIUL99promoterR. The sequences of the primers are as follows: NheIUL75promoterF, 5′-CTAGCTAGCCTGCAGGCTGTGGGTGGCGTG-3′; BglIIUL75promoterR, 5′-GAAGATCTCATAGCGCGGCCGCGCGGCTG-3′; NheIUL99promoterF, 5′-CTAGCTAGCCTCGCCGGGCATCCAGTTCGG-3′; and BglIIUL99promoterR, 5′-GAAGATCTGCTCGGGCGTGTTTTCGTCCTCGAAAG-3′. The amplified products were confirmed by sequencing (Aichi Cancer Center Research Institute Central Facility), digested by restriction endonucleases NheI and BglII, and cloned into pCAT3-Basic vector at the corresponding restriction endonuclease sites.

Transfection and infection assay.

Transfections were done in triplicate on 60-mm-diameter plates of HFFs by the calcium phosphate precipitation method as described above. HFFs were transfected with 2 μg of the reporter plasmid driven by the UL75 or UL99 promoter, and 48 h later, the cells were infected with the wt and the recombinant viruses at an MOI of approximately 1 and harvested 48 and 72 h p.i. Cell lysates were then prepared and subjected to CAT assays as described previously (17). Acetylated and unacetylated [¹⁴C]chloramphenicols (Amersham Pharmacia Biotech, Piscataway, NJ) were separated by thin-layer chromatography in a chloroform-methanol (95:5) solvent. Signal intensity was quantitated with an image guider (BAS 2500; Fujifilm, Tokyo, Japan), and the percent conversion of unacetylated [¹⁴C]chloramphenicol to the acetylated form was calculated. Data are averages of three independent transfection-infection experiments.

RESULTS

The distal TATA element of the UL44 promoter.

The HCMV viral DNA processivity factor pUL44 interacts with HCMV DNA polymerase (pUL54) to decrease dissociation of the polymerase from the viral DNA template (8, 40, 45). pUL44 accumulates to strikingly high levels at late times after infection (10, 35). The late times of transcription and the high level of pUL44 expression suggest additional important roles for this protein. The UL44 gene has three spatially distinct transcriptional start sites, and these three viral promoters are regulated independently (20). Two are early promoters, and the third is a late promoter that depends on viral DNA replication for activity (Fig. 1b). The distal TATA element of the UL44 promoter is located in the UL45 ORF (Fig. 1b).

To determine the role of the distal TATA element in UL44 transcription, we determined whether the mutation of the UL45 ORF has an effect on UL44 transcription and viral growth (6, 12). Since a drug-resistant gene is necessary to select the recombinant BAC DNA from wt, we constructed a recombinant virus in which a part of the UL45 ORF (corresponding to amino acids 34 to 612) was replaced by the kanamycin-resistant gene (kan) (Fig. 1a). This recombinant virus still has the distal TATA element of the UL44 promoter. PCR and Southern blot hybridization analysis confirmed the correct recombination, and the recombinant BAC DNAs do not contain the UL45 gene (data not shown). HFFs were transfected with the recombinant HCMV BAC DNA, and recombinant virus was stored and assayed by TCID₅₀ assay on HFFs as described in Materials and Methods.

To compare the expression levels of the IE proteins pIE72 and pIE86, the early protein pUL44, and the late protein pp28, HFFs were infected with the wt and dlUL45 at an MOI of approximately 1 and harvested 1, 2, and 3 days p.i. Equal amounts of protein were subjected to SDS-polyacrylamide gel electrophoresis and Western blotted as described in Materials and Methods. Levels of IE, early, and late protein expression were similar for the wt and recombinant virus with the UL45 deletion (data not shown). To compare the growth of the wt and the recombinant virus, HFFs were infected at an MOI of approximately 1 or 0.01. Virus titers were determined by the TCID₅₀ assay on HFFs as described in Materials and Methods. dlUL45 grew similarly to the wt at high and low MOI (data not shown). As reported previously (6, 12), we confirm that UL45 ORF has no effect on viral gene expression and growth in cell culture.

Recombinant BACs with the distal or the middle TATA element of the UL44 promoter mutated.

To determine the role of the middle late UL44 promoter, we constructed recombinant viruses with the distal early or the middle late UL44 TATA elements mutated (Fig. 1a). After kan was excised by FLP-mediated recombination, 34 bp of FRT was substituted for each of the TATA elements (Fig. 1a). The distal, middle, and proximal transcriptional start sites were labeled 1, 2, and 3, respectively (Fig. 1b). All the recombinant and the parental BAC DNAs were digested with XhoI and immobilized for Southern blot hybridization as described in Materials and Methods. Since kan contains an XhoI site, an additional DNA fragment was detected in the recombinant BAC DNA (data not shown). After kan excision, the DNA fragment digested by XhoI was approximately the same size (in base pairs), and the additional DNA fragment was completely lost (data not shown). Moreover, PCR analysis revealed the correct sizes of 446 and 445 base pairs for dlTATA1 and dlTATA2, respectively (data not shown). Sequencing of the amplified DNA fragments confirmed correct recombination and excision (data not shown). Two separate isolations of the recombinant viruses gave identical results, as described below.

Evaluation of equal MOI input among wt and recombinant viruses.

After isolation of the recombinant viruses, the titers of stored wt human CMV Towne and recombinant viruses were determined by standard plaque assays on HFFs as described previously (15). Since the titer of stored dlTATA1 was lower than that of the wt and dlTATA2, the stored wt and dlTATA2 viruses were diluted with growth medium to adjust the viral titer. We were unable to infect cells with the stored viruses at an MOI higher than 1. To confirm the equivalent viral input of the wt and the recombinant viruses at the same MOI, we harvested the infected cells before viral DNA synthesis and compared the relative levels of IE1 protein (pIE72). The levels of pIE72 were similar for the wt and the recombinant viruses at 1 day p.i. (see Fig. 5). The number of IE1/2-positive cells was determined by immunofluorescence assay as described in Materials and Methods. As shown in Fig. 2, the numbers of cells that were IE1/2 positive were similar at 1 day p.i. From these results, we confirmed equivalent viral input at the same MOI.

FIG. 5.

Analysis of viral protein expression after infection with wt and recombinant viruses. HFFs were infected with the wt and the recombinant viruses at an MOI of 1 and analyzed for viral proteins. IE pIE72 (UL123), pIE86 (UL122), early ppUL44, and late pp28 proteins were analyzed 1, 2, and 3 days p.i. with monoclonal antibodies NEA-9221, M0854, and CA 004, respectively, as described in Materials and Methods. Anti-pGAPDH (p36) antibody was used to show equal protein loading.

FIG. 2.

Comparison of the IE1/2 protein expression by wt and recombinant viruses. Number of IE1/2-positive cells by immunostaining. HFFs were infected with the wt and the recombinant viruses at an MOI of approximately 1 and stained with monoclonal antibody NEA-9221 1 day p.i. The IE1/2-positive cells were counted under an inverted UV microscope with a digital camera (Nikon Eclipse TE-2000-S) using Image J software following documentation (http://rsb.info.nih.gov/ij/). The results are means plus standard deviations from four independent experiments.

Effect of mutated TATA elements on UL44 transcription.

To determine the effect of the distal or the middle TATA element on UL44 transcription, cytoplasmic RNA was harvested at 1, 2, and 3 days after infection with wt or the recombinant viruses at an MOI of approximately 1. To detect all the transcripts derived from the different start sites, an RNase protection assay was performed. Twenty micrograms of RNA was hybridized to a ³²P-labeled antisense RNA probe at 37°C overnight before digestion with RNase T1 (100 U) as described in Materials and Methods. The FRT sequence in the RNA transcript was digested by RNase T1, because the antisense UL44 probe does not contain an FRT sequence. Consistent with a previous report (20), three major transcripts initiating at the spatially distinct start sites were detected 2 and 3 days after infection with the wt (Fig. 3). Minor transcripts of the wt and recombinant viruses that may represent heterogeneous start sites were also detected, but there was no alternative transcript for dlTATA1 and dlTATA2 due to the mutation of the TATA element (Fig. 3). Transcripts initiating at start site 1 or 2 consisted of a doublet start site (Fig. 3). Levels of the distal transcripts were approximately similar in the wt and dlTATA2 at 1 and 3 days p.i. (Fig. 3, lanes 3 and 5 and lanes 9 and 11). Levels of the distal and proximal transcripts in the wt and dlTATA2 were also approximately similar in the presence of an inhibitor of viral DNA synthesis (PAA) at 48 h p.i. (Fig. 3, lanes 12 and 14). Replacement of the TATA2 element had little effect on the levels of the distal and proximal transcripts. A relatively small amount of the distal transcript was detected with dlTATA1 (Fig. 3, lanes 4, 7, 10, and 13). These results showed that the distal TATA element determined the level of the early distal transcript. Consistent with a previous report (20), the middle transcript was not detected with wt and dlTATA1 in infected cells maintained in the presence of PAA for 48 h (Fig. 3, lanes 12 and 13), indicating that the TATA2 element was activated upon viral DNA replication. The middle transcript was detected with dlTATA1 as well as the wt at 2 and 3 days p.i. but was not detected with dlTATA2 (Fig. 3, lanes 6 to 11). These results showed that replacement of the TATA2 element caused loss of the late transcript initiating at start site 2. The levels of the middle and the proximal transcripts derived from dlTATA1 were slightly lower than those of the wt at 2 days p.i. (Fig. 3, lanes 6 and 7). Since levels of the proximal transcripts were similar for the wt and dlTATA1 in the presence PAA for 48 h (Fig. 3, lanes 12 and 13), replacement of the TATA1 element might affect the transcription initiating at start sites 2 and 3 in a viral DNA replication-dependent manner. Taking these observations together, we conclude that mutation of the distal or middle TATA element reduced the transcription from the corresponding start sites.

FIG. 3.

Effect of substitution for each TATA element on the UL44 transcription in cells infected with wt and recombinant viruses. Cytoplasmic RNAs were harvested at 1, 2, and 3 days after infection with an MOI of approximately 1. Twenty micrograms of RNA was hybridized to ³²P-labeled antisense UL44 promoter probe at 37°C overnight before digestion with RNase T1. The antisense UL44 probe contains sequence upstream of the transcription start site of all the UL44 transcripts. The protected RNA fragments were subjected to electrophoresis in denaturing 6% polyacrylamide gels. UL44 probe is a probe lacking RNase T1. Arrows labeled 1, 2, and 3 indicate the transcripts initiating at start sites 1, 2, and 3, respectively.

Replacement of the distal and middle TATA elements had an impact on the viral delayed early and late gene transcription.

To compare the viral delayed early and late transcription of the recombinant viruses with that of the wt, cells were infected at an MOI of approximately 1 and assayed for IE, early, delayed early, and late gene transcripts by Northern blotting at 1, 2, and 3 days p.i. RNA from infected cells maintained in the presence of PAA for 2 days was also assayed. Twenty micrograms of RNA was subjected to agarose gel electrophoresis. Ethidium bromide staining of 28S and 18S rRNA confirmed that equal amounts of RNA were loaded in each lane (Fig. 4a and b). We also used real-time RT-PCR for a quantitative analysis of viral mRNA as described in Materials and Methods. The IE2 protein negatively autoregulates the MIE promoter in a DNA sequence- and position-dependent manner by binding to the cis repression sequence (3, 21, 33). Thus, as infection proceeds, regulation of IE RNA is very complex, and the amount of IE RNA is variable without an MOI for single-step growth. For these reasons, it is difficult to evaluate the effect of the mutated UL44 promoter on the MIE transcription. RNA analysis after PAA treatment for 48 h did not show a significant difference in IE1 RNAs between the wt, dlTATA1, and dlTATA2 (Fig. 4a, lanes 7 to 9, and 4e). In contrast, the level of IE RNAs from dlTATA1 was approximately twofold lower than that of the wt and dlTATA2 2 and 3 days p.i. (Fig. 4a, lanes 5 and 6 and lanes 10 and 11, and 4e). This might suggest that viral DNA replication modulated the IE transcription from the MIE promoter. A significant reduction of the UL44 transcript from PAA-treated cells was observed relative to transcript levels from untreated cells. In the absence of PAA, the amount of UL44 RNA from dlTATA1 was more than twofold lower than that from the wt at 1, 2, and 3 days p.i. (Fig. 4b, lanes 1 and 2, 4, and 5 and lanes 10 and 11, and 4e). There was no significant difference in the amount of UL44 RNA between the wt and dlTATA2 at 1 and 2 days p.i. (Fig. 4b, lanes 1 and 3 and lanes 4 and 6, and 4e). In contrast, the level of UL44 transcript from dlTATA2 was approximately twofold lower than that of the wt at 3 days p.i. (Fig. 4b, lanes 10 and 12, and 4e). The reduction of the UL44 transcript in the cells infected with dlTATA1 or dlTATA2 correlates with the loss of the corresponding early or late transcript initiating at start site 1 or 2, respectively (Fig. 3).

FIG. 4.

Analysis of UL44 and subsequent gene transcription after infection with the wt and recombinant viruses. HFFs were infected with an MOI of approximately 1, and cytoplasmic RNA was harvested 1, 2, and 3 days p.i. as described in Materials and Methods. (a to d) Northern blotting for IE1, UL44, UL99 (pp28), or UL75 (gH). 28S and 18S rRNA served as controls for an equal amount of RNA loading. (e) Real-time RT-PCR for analysis for IE1/2, UL44, UL92 to -99, and UL75 (gH) gene transcripts.

The UL99 ORF is located in a complex region of HCMV with a series of 3′-coterminal transcripts (42). All the transcripts utilize a common polyadenylation site downstream of ORF UL99 (42). Thus, real-time PCR analysis using the UL99 primers and probe detects the total amount of the UL92 to -99 transcripts. The pp28 tegument protein is translated from two mRNAs of 1.6 and 1.3 kb (1). The 1.3-kb mRNA is of relatively low abundance (1). Northern blot analysis using the UL99 probe showed that in the presence of PAA, the 1.6-kb mRNA from the UL99 promoter was not detectable; thus, UL99 is a true late gene (Fig. 4c, lanes 7 to 9) as reported previously (1). Consistent with a previous report (42), analysis of PAA-treated RNA also showed that mRNAs initiating upstream of each of the potential ORFs in this region contained late transcripts (Fig. 4c, lanes 7 to 9). We did not detect differences in the level of steady-state mRNAs between the wt and dlTATA2 at 1 and 2 days p.i., but we did detect it at 3 days p.i. (Fig. 4c, lanes 1 and 3, 4 and 6, and 10 and 12). In a quantitative analysis using real-time RT-PCR, the amount of the UL92 to -99 transcripts for dlTATA1 and dlTATA2 was more than twofold lower at 2 and 3 days p.i., respectively (Fig. 4e). These results indicate that the early or late transcripts initiating at start site 1 or 2 of the UL44 gene facilitated delayed early and late transcription.

To determine whether the product from the UL44 late transcript initiating at start site 2 affected other late gene expression, Northern blot analysis using the UL75 (gH) probe was also performed. Three mRNAs were detected (Fig. 4d). Two transcripts, 4.5 and 5.5 kbp, corresponded to the UL76 gene and the UL77 and UL78 genes, respectively, which utilize different transcription start sites as reported previously (38). An analysis of the transcripts of the recombinant virus with an UL76 ORF deletion determined the transcript from the UL76 promoter in the three mRNAs (unpublished observation). The direction of transcription of the UL75 ORF is opposite to that of the UL76, -77, and -78 ORFs. Thus, real-time RT-PCR analysis using the UL75 primers and probes detects only the UL75 transcript. The initiation site of the UL75 transcript was determined by primer extension as described previously (23). RNA analysis from PAA-treated cells for 48 h suggested that UL75 is a true late viral gene (Fig. 4d, lanes 7 to 9; also long exposure [data not shown]). Since small amounts of the transcripts from the UL76, UL77, and UL78 promoters were detected in PAA-treated cells after long exposure (data not shown), UL76, -77, and -78 were delayed early viral genes. Similar to the result obtained with the probe for UL99, the level of steady-state mRNAs from the UL76 promoter and the UL77 and UL78 promoters in the cells infected with dlTATA2 was lower than that in cells infected with the wt at 3 days p.i. (Fig. 4d, lanes 10 and 12). Likewise, the level of transcript from the UL75 promoter was also approximately twofold lower for dlTATA2 at 3 days p.i. (Fig. 4d, lanes 10 and 12, and 4e). The steady-state mRNAs from UL75, UL76, and UL77 and UL78 promoters in the cells infected with dlTATA1 were lower at 2 and 3 days p.i. (Fig. 4d, lanes 5 and 6 and lanes 10 and 11, and 4e). From these results, we conclude that the UL44 gene product from early or late transcripts initiating at start site 1 or 2 has an impact on viral delayed early and late gene transcription.

Replacement of the distal and middle TATA element reduced accumulation of late viral protein.

Since replacements of the distal and middle TATA elements affected UL44 and viral delayed early and late gene transcription, we wish to determine whether the reduction of these viral gene transcriptions affected expression of the viral late proteins in cells infected with recombinant viruses. HFFs were infected with the wt and the recombinant viruses at an MOI of 1, harvested 1, 2, and 3 days p.i., and analyzed for viral protein expression by Western blotting. As expected, levels of pUL44 were lower for dlTATA1 and dlTATA2 than for the wt at 2 and 3 days p.i. (Fig. 5, lanes 7 to 9). There was also a little difference in the levels of pIE72 and pIE86 between wt and recombinant viruses at 2 and 3 days p.i. (Fig. 5, lanes 4 to 9). However, the levels of the late viral protein pp28 (pUL99) were 16.7- and 7.4-fold lower for dlTATA1 and dlTATA2, respectively (Fig. 5, lanes 8 and 9). From these results, we conclude that replacement of either the distal or middle TATA element reduced accumulation of late viral protein.

Replacement of the middle element had no effect on viral DNA synthesis.

Since pUL44 is essential for viral DNA synthesis (30, 34), we compared viral DNA replication of the recombinant viruses with that of the wt at high or low MOI. HFFs were infected with wt and recombinant viruses at an MOI of approximately 1 or 0.01 and harvested 2, 3, and 4 days p.i. or 5 and 8 days p.i., respectively. DNA from an equal number of infected cells was collected, and the viral DNA was quantitated by real-time PCR using gB primer and probes as described in Materials and Methods. While the relative input of viral genomes for an MOI equivalent to approximately 1 ranged between 0.001 and 0.0007 for wt, dlTATA1, and dlTATA2 (Fig. 6a), viral DNA of dlTATA1 was 5 to 10 times lower than that of the wt after infection at an MOI of approximately 1 or 0.01 (Fig. 6a and b, respectively) and dlTATA2 was similar to wt in relative amount (Fig. 6a and b, respectively). We conclude that the reduction of late gene expression in the cells infected with dlTATA2 was not due to a reduced level of DNA template for transcription.

FIG. 6.

Analysis of viral DNA replication after high- or low-multiplicity infection of HFFs with wt and recombinant viruses. Viral DNA was quantified by real-time PCR with gB primers and probe as described in Materials and Methods. Real-time PCR with 18S primers and probe was also performed to serve as an internal control. Data are averages of three independent experiments. (a) HFFs were infected with the wt, dlTATA1, or dlTATA2 at an MOI of approximately 1 and harvested at 2, 3, and 4 days p.i. Each value was calculated and plotted relative to the level of the wt DNA at 4 days p.i. (b) HFFs were infected with the wt, dlTATA1, or dlTATA2 at an MOI of approximately 0.01 and harvested at 5 and 8 days p.i. Each value was calculated and plotted relative to the level of the wt DNA at 8 days p.i.

The middle element did not affect late viral gene expression independently of viral genome replication.

To determine whether the effect of the middle element on late viral gene expression is independent of viral DNA replication, we constructed a reporter plasmid driven by the UL75 or UL99 late promoter. Since late viral promoters in a plasmid are activated by HCMV infection in the absence of viral DNA replication (4, 5), we compared the effect of viral infection by the wt and dlTATA2 on late viral promoters in a plasmid. HFFs were transfected with 2 μg of the reporter plasmid, and 48 h later, the cells were infected with wt or dlTATA2 at an MOI of approximately 1 and harvested 2 and 3 days p.i. Cell lysates were then prepared and subjected to CAT assays as described in Materials and Methods. As shown in Fig. 7, when the reporter plasmid was driven by the UL75 or UL99 promoters independently of the viral genome, accumulation of the reporter gene in the cells infected with dlTATA2 was not reduced compared to that in cells infected with the wt. These results indicate that the effect of the middle element on late viral gene expression is linked to viral DNA replication.

FIG. 7.

Activation of the UL75 and UL99 promoters in a plasmid after wt or dlTATA2 infection. HFFs were transfected with 2 μg of the reporter plasmid, and 48 h later, the cells were infected with the wt and the recombinant viruses at an MOI of approximately 1. The cells were harvested 2 and 3 days p.i. Cell lysates were then prepared and subjected to CAT assays as described in Materials and Methods. Data are averages of three independent transfection-infection assays.

Growth kinetics of the recombinant viruses.

Since accumulation of late viral protein for the recombinant viruses was reduced, we wished to compare the growth of the recombinant viruses with wt at low or high MOI. HFFs were infected with wt or recombinant virus at an MOI of approximately 0.01 or 1. Viruses from infected cultures at 1, 5, 8, 11, and 14 days p.i. or 1, 4, 5, 7, and 9 days p.i., respectively, were assayed as described in Materials and Methods. Growth of dlTATA1 or dlTATA2 was slightly delayed compared to that of the wt at low and high MOI (Fig. 8a and b). We detected approximately a 5- to 10-fold difference in viral replication at low MOI between wt and dlTATA2 at 5 and 8 days p.i. (Fig. 8a), while DNA replication of dlTATA2 was similar at the same time (Fig. 6a and b). Likewise, at high MOI, the viral titer of dlTATA2 was approximately 10-fold lower than that of wt at 4 and 5 days p.i. (Fig. 8b), while viral DNA replication of dlTATA2 was similar to that of the wt (Fig. 6a and b). Therefore, we conclude that insufficient expression of the late gene product of UL44 was linked to delayed viral growth. The TATA elements for the first and second start sites play a critical role for pUL44 gene expression. Transcription from the second start site, which is the late promoter, had an effect on pUL44 levels and the level of late viral gene expression but not on viral DNA synthesis.

FIG. 8.

Growth of wt and recombinant viruses after low-multiplicity (MOI ≈ 0.01 [a]) or high-multiplicity (MOI ≈ 1) infection of HFFs. Virus titers were determined by the TCID₅₀ assay as described in Materials and Methods.

DISCUSSION

Transcriptional regulation of UL44 gene expression occurs at two levels. The two transcription start sites that are activated early in infection presumably respond to IE proteins. Each contains a conventional TATA element, which is the only region of significant homology between start sites 1 and 3. Late in infection, sequences that are dependent on DNA replication result in transcriptional activation at start site 2 (20). This is referred to as a late viral promoter. DNA replication of recombinant virus dlTATA1 was lower than that of the wt after infection at an MOI of 1 or 0.01. Since pUL44 is an essential viral protein that acts as a processivity factor for the catalytic subunit of the viral DNA polymerase (pUL54) (8, 30, 34, 40, 45), the loss of transcription start site 1 was responsible for the lower level of viral DNA replication of recombinant virus dlTATA1. Delayed early and late gene expressions derived from recombinant virus dlTATA2 were also decreased compared to those from the wt. Since DNA replication of recombinant virus dlTATA2 was similar to that of the wt, the reduction in viral gene expression was not due to a reduced DNA template for transcription. What ratio of newly replicated DNA serves as the template for transcription is not known. We cannot rule out the possibility of a slight increase in distal and proximal transcripts from dlTATA2 at 2 days p.i., which may have contributed to the similar viral DNA replication levels of dlTATA2 and the wt. However, while pUL44 accumulates to strikingly high levels at late times after infection (10, 35), viral DNA polymerase does not accumulate. Therefore, we propose that the late accumulation of pUL44 depends on the late UL44 promoter, which is required for efficient viral delayed early and late gene expression. Taken together, we conclude that sufficient pUL44 is required for viral gene expression and that mutation of start site 2 reduces viral gene expression.

The Epstein-Barr virus (EBV) viral DNA polymerase processivity factor, BMRF1, activates transcription of an EBV early promoter, BHLF1 (44). The stimulating effect of HCMV pUL44 on subsequent viral gene expression was for multiple delayed early and late viral gene promoters that depend on viral DNA replication. pUL44 is detected at HCMV DNA replication centers in the nucleus (31, 32). Viral replication centers also serve as foci for viral gene expression, presumably in part by concentrating templates for transcription with the proteins that carry out or regulate this process. The herpes simplex virus single-stranded DNA-binding protein ICP8 is also located at viral DNA replication centers in the nucleus and stimulates multiple late viral gene promoters (9). A recent report showed that ICP8 coprecipitates with chromatin remodeling factors (36). Whether pUL44 is also associated with ATP-dependent nucleosome remodeling proteins on the HCMV genome is not known, but it might be possible that pUL44 recruits activating chromatin remodeling factors to late viral promoters at late times after infection.

Our analysis confirmed that the middle TATA element is involved in late transcripts initiating at start site 2 (20). Why late promoters dominate over early promoters at late times after infection is unclear. An important parameter governing transcription initiation is the relative concentration of the viral DNA template and the TATA element. TATA sequence-specific transcription factors may be differentially activated during HCMV replication. Alternatively, a conformational change in TATA sequence resulting from DNA replication may allow late viral gene transcription. The increase in intracellular concentration of viral DNA could also alleviate promoter occlusion. Regardless of the mechanism at the viral DNA level, the level of UL44 gene expression affects viral DNA template concentration. However, the UL44 gene product also affects late viral promoter activation and late viral gene expression independently of the concentration of viral DNA template. Since the level of IE1 RNA from dlTATA1 was approximately twofold lower than that of the wt and dlTATA2 at 2 and 3 days p.i., we cannot rule out the possibility that UL44 gene product also affects the MIE promoter and the affected MIE gene products modulate delayed early and late gene expression. However, since mutation of the late UL44 promoter did not have a significant effect on the level of IE1 gene transcription, the stimulating effect of the UL44 late gene product on the subsequent gene transcription was not through the modulation of the MIE transcription. Therefore, the late UL44 gene product functions either directly or indirectly as a late viral transactivator dependent on viral DNA replication.

In conclusion, we have shown that late transcripts initiating from the middle start site of UL44 promoter are required for efficient viral growth at a low or high MOI. The complex structure of the UL44 promoter, with three TATA elements and one that is active at only late times after infection, is necessary for efficient viral growth during HCMV infection.