Transcriptional Regulation of the Human Cytomegalovirus US11 Early Gene

Abstract

The human cytomegalovirus (HCMV) US11 early gene encodes a protein involved in the down-regulation of major histocompatibility complex class I cell surface expression in HCMV-infected cells. Consequently, this gene is thought to play an important role in HCMV evasion of immune recognition. In this study, we examined the transcriptional regulation of US11 gene expression. Analysis of deletions within the US11 promoter suggests that two sequence elements are important for activation by the viral immediate-early (IE) proteins. Deletion of a CREB site located at −83 relative to the cap site resulted in a reduction in promoter activity to 50% of the wild-type level. Deletion of an additional ATF site immediately upstream of the TATA box resulted in abrogation of responsiveness to the IE proteins. To confirm the role of the CREB and ATF sites within the US11 promoter, mutagenesis of these two sites, both individually and in combination, was carried out. Results indicate that both the CREB element and the ATF site were required for full promoter activity, with the ATF site critical for US11 promoter activation. The loss of transcriptional activation correlated with a loss of cellular proteins binding to the mutated US11 promoter elements. In combination with the viral IE proteins, the HCMV tegument protein pp71 (UL82) was found to up-regulate the US11 promoter by six- to sevenfold in transient assays. These results suggest that pp71 may contribute to the activation of the US11 promoter at early times after infection. Up-regulation by pp71 required the presence of the CREB and ATF sites within the US11 promoter for full activation. The role of the ATF and CREB elements in regulating US11 gene expression during viral infection was then assessed. The US11 gene is not required for replication of HCMV in tissue culture. This property was exploited to generate US11 promoter mutants regulating expression of the endogenous US11 gene in the natural genomic context. We generated recombinant HCMV that contained the US11 promoter with mutations in either the CREB or ATF element or both regulating the expression of the endogenous US11 gene. Northern blot analysis of infected cell mRNA revealed that mutation of the CREB element reduced US11 mRNA expression to approximately 25% of that of the wild-type promoter, with identical kinetics of expression. Mutation of the ATF site alone reduced US11 mRNA levels to 6% of that of the wild-type promoter, with mRNA detectable only at 8 h after infection. Mutation of both the CREB and ATF elements in the US11 promoter reduced US11 gene expression to undetectable levels. These results demonstrate that the CREB and ATF sites cooperate to regulate the US11 promoter in HCMV-infected cells.

The regulation of viral gene expression in human cytomegalovirus (HCMV)-infected cells relies on a complex interplay between cellular and viral factors. This process is initiated by the binding of HCMV to its cellular receptor, resulting in the enhanced expression of cellular transcription factors such as c-Jun, c-Fos, and NF-κB, required for the initiation of viral gene expression (3, 4, 49). Once the virus particle penetrates the cell, HCMV gene expression follows an ordered and sequential pathway which can be broken down into three broad classes: immediate-early (IE), early, and late (5). The majority of IE gene expression is directed by a complex promoter, the major IE promoter (MIEP), which can be activated by cellular transcription factors such as AP-1, NF-κB, and ATF/CREB (reviewed in reference 34). An HCMV tegument protein, pp71, functioning via ATF and AP-1 sites, enhances the activation of the MIEP (31). The IE proteins of HCMV are responsible for the activation of subsequent viral gene expression (43, 44). The HCMV IE2 protein, IE86, can bind directly to DNA, and binding sites for this protein are involved in the regulation of the UL112-113 early promoter (1, 35, 38, 39). In addition to its DNA binding function, the IE86 protein of HCMV can interact with TATA binding protein (TBP), TFIIB, and TBP-associated factors, as well as cellular transcription factors such as p53, c-Jun, and CREB (6, 21, 30, 32, 37, 39, 40).

In HCMV-infected cells, transcriptional activation of early genes relies on both cellular transcription factors and the viral IE proteins (13, 23, 28, 30, 37, 39–41). Several subclasses of HCMV early gene kinetics have been defined, indicating an additional level of complexity in early gene regulation (43). One factor that may contribute to the kinetic complexity of early gene regulation is the presence of additional HCMV gene products able to stimulate viral early gene expression (8, 14, 15, 24, 31, 36, 42, 48). Previous studies assessing the activation of early promoters have typically relied on transient assays such as cotransfection experiments to assess promoter activity (13, 23, 28, 33, 37, 39, 41, 46). More recently, strategies have been developed to assess HCMV early gene regulation in the context of the viral genome (24–26, 29, 35). These strategies have used a region of the HCMV genome dispensable for growth of the virus in tissue culture (19, 20). Promoter constructs regulating the expression of reporter genes such as the chloramphenicol acetyltransferase (CAT) gene have been inserted into the viral genome in the unique short region between open reading frames (ORFs) US9 and US10. These studies have demonstrated that gene expression in the context of the viral genome is much more complex than that found in transient assays. For example, sequence elements that are not required for promoter activation in transient assays can play critical roles in regulating gene expression in the context of the viral genome (24, 25). In addition, such studies have enabled the identification of HCMV promoter elements that are involved in temporal regulation of early gene expression (24, 25, 29, 35). However, these studies are hindered by the inability to assess gene expression in the natural genomic context. Due to the overlapping nature of HCMV transcriptional units, it is possible that the transcription of upstream ORFs can influence transcription of downstream genes. We have begun to address this problem by assessing transcriptional regulation of genes within the nonessential US gene region (19, 20). With this strategy, we can study the regulation of endogenous genes within their natural gene environment.

The present study examines the regulation of the HCMV early gene, US11 (18, 20). This gene is nonessential for replication in tissue culture but likely plays a critical role in HCMV pathogenesis (16, 17, 20, 47). The product of the US11 gene is an endoplasmic reticulum glycoprotein that causes the rapid destruction of major histocompatibility complex class I proteins, resulting in the down-regulation of cell surface expression of class I (16, 17, 47). Expression of the US11 gene can be detected within 2 h after infection of permissive cells with HCMV (18). Interestingly, US11 is down-regulated at late times after infection, which places it in the E1 subclass of viral early genes (43). In this study, we focused on the transcriptional regulation of US11 gene expression. The results demonstrate that two ATF/CREB binding sites within the US11 promoter play significant roles in transactivation by the viral IE proteins. The consensus ATF site immediately adjacent to the TATA element was critical for US11 promoter activation in transient assays. The pp71 protein was found to stimulate activation of the US11 promoter by the IE proteins. Enhanced transcription by pp71 was also dependent on the presence of the ATF/CREB sites within the US11 promoter. To fully assess the role of ATF/CREB in the regulation of this early gene, US11 promoter mutants regulating the expression of the endogenous US11 gene were introduced into the HCMV genome by homologous recombination. Analysis of US11 mRNA revealed that both ATF/CREB sites within the US11 promoter contribute to the transcriptional activation of this gene.

MATERIALS AND METHODS

Plasmids.

Plasmids pSVH, pSVOd, and pSVOCAT (9) were obtained from R. M. Stenberg (Eastern Virginia Medical School, Norfolk). pSVH contains the major IE genomic region expressed under the control of the MIEP in the vector pSVOd. Mark F. Stinski (University of Iowa, Iowa City) kindly provided plasmid pCMV71 (31), which contains the pp71 ORF downstream of the MIEP. In addition, pCMV71dlPvuI, a carboxy-terminal deletion mutant of pCMV71 which is unable to transactivate the MIEP, was used as a control for these experiments (31). The plasmid used for generating recombinant viruses, pUS115′3′, was kindly provided by T. R. Jones (Wyeth-Ayerst Research, Pearl River, N.Y.). This plasmid was modified by the elimination of a HindIII site within the polylinker to generate pUS115′3′dH. The US11 promoter was cloned from the HindIII site at nucleotide (nt) 205020 of the HCMV genome to the XbaI site at nt 204555 (7) as a HindIII fragment into the HindIII site of the reporter vector, pSVOCAT, to generate pUS11CAT.

Generation of US11 promoter variants.

Nested deletions within the US11 promoter were generated as previously described (43). Briefly, the HindIII site at the 3′ end of the US11 promoter in pUS11CAT was eliminated to generate pUS11dHCAT. This plasmid was then partially digested with either DpnI, NlaIV, NspI, or RsaI, and the linear DNA was isolated. The linearized DNA was then ligated to HindIII linkers, the resultant DNA was digested with HindIII, and the plasmid was recircularized. This procedure results in the removal of sequences between the HindIII site 5′ of the promoter and the inserted linker. Deletions within the promoter region were selected by restriction enzyme digestion and confirmed by DNA sequencing.

Site-directed mutagenesis of the CREB site was performed by overlapping PCR mutagenesis using plasmid pUS11CAT with oligonucleotides CAT1 (5′ GCTCTGATGCCGCATAGTTAAGCC), CAT2 (5′ GCGGGCAAGAATGTGAATAAAGGCCGG), US11Cm1 (5′ CACCGTGTTCTCCCGACGATATCACTAGATCACCACCCTG), and US11Cm2 (5′ CAGGGTGGTGATCTAGTGATATCGTCGGGAGAACACGGTG). The mutated nucleotides are underlined. PCR using Vent DNA polymerase (New England Biolabs, Beverly, Mass.) was performed on pUS11CAT, using the primer pairs CAT1-US11Cm2 and CAT2-US11Cm1. This PCR resulted in two fragments that overlap by 40 bp. The two fragments were then combined and PCR performed with Taq DNA polymerase (Gibco-BRL, Gaithersburg, Md.), using the CAT1-CAT2 primer pair. The resultant fragment was digested with HindIII, and the mutated US11 promoter fragment was recloned into pSVOCAT to generate pUS11CmCAT. The presence of the CREB mutation was assessed by restriction enzyme digestion as well as DNA sequencing. The construct with the mutation in the US11 ATF site, pUS11AmCAT, was generated in a similar manner, using primers US11Am1 (5′ CCACCCTGTTCCCCGTGAATTCCAAGACTACATGCTATAAG) and US11Am2 (5′ CTTATAGCATGTAGTCTTGGAATTCACGGGGAACAGGGTGG).

Mutation of both the ATF and CREB sites within the US11 promoter was achieved by subjecting plasmid pUS11AmCAT to a second round of PCR mutagenesis using primers US11Cm1 and US11Cm2. The resultant fragment was then cloned as a HindIII fragment into pSVOCAT to generate pUS11CAmCAT. All mutations were confirmed by restriction enzyme digestion and DNA sequencing.

Transfections.

Cotransfection analysis of the US11 promoter variants was performed as previously described (9, 45). Briefly, primary human foreskin fibroblasts (HFFs) were transfected with 10 μg of plasmid DNA by the DEAE-dextran method. Cells were harvested 48 h after infection and assessed for CAT enzyme activity (9).

Gel shift assays.

Gel shift analysis of proteins binding to the CREB and ATF sites of the US11 promoter were performed essentially as described previously (25). Briefly, the oligonucleotides US11ATF-1 (5′ CCTGTTCCCCGTGACGTGCAAGACTACAT), US11ATF-2 (5′ CATGTAGTCTTGCACGTCACGGGGAACAG), US11CREB-1 (5′ GTGTTCTCCCGACGTCACTAGATCACC), and US11CREB-2 (5′ GGGTGATCTAGTGACGTCGGGAGAACA) were annealed in 20 mM Tris-HCl (pH 7.5)–50 mM NaCl–10 mM MgCl₂. The probes were end labeled with [γ-³²P]ATP and T4 polynucleotide kinase and purified by nondenaturing polyacrylamide gel electrophoresis followed by electroelution. Nuclear extracts were isolated from uninfected HFFs by a modified Dignam procedure (10, 23). Assays were performed in 0.5× buffer D in the presence of 5 μg of poly(dI-dC) · poly(dI-dC) and 30,000 to 40,000 cpm of probe DNA (equivalent to 0.5 to 1 ng of probe DNA). Electrophoresis was performed on a 4% polyacrylamide gel in 0.5× Tris-borate-EDTA buffer. Competitor DNAs containing mutated ATF and CREB elements were generated by annealing US11Cm1 and US11Cm2 or US11Am1 and US11Am2 as described above.

Genomic transfections and purification of recombinant viruses.

Recombinant viruses were generated by homologous recombination via a modified calcium phosphate transfection protocol (20, 29). US11 promoter mutants generated in plasmid pUS11CAT were digested with HindIII and NspI, and the US11 promoter fragment was isolated. This fragment was then cloned into an intermediate vector containing the HindIII-to-XbaI US11 promoter fragment cloned into pBluescript KS(+) (Stratagene, La Jolla, Calif.). The mutant promoter regions were then cloned as HindIII-to-XbaI fragments into the pUS115′3′dH recombination vector digested with HindIII and XbaI. This results in the replacement of the wild-type US11 promoter sequences with the mutated promoter fragments, upstream of the endogenous US11 ORF. The resultant plasmids were then linearized with either PstI or EcoRI and cotransfected into primary HFFs with RV699 genomic DNA. RV699 contains the β-glucuronidase gene inserted into the US11 ORF (20). Homologous recombination will result in the replacement of the β-glucuronidase gene region with the US11 ORF, regulated by mutant US11 promoters. RV699 DNA for transfection was isolated as previously described (29) or by using a GenomicPrep kit (Pharmacia Biotech, Piscataway, N.J.). The recombinant virus, RVUS11, which reconstructs the original AD169 genotype, was generated by the same procedure except that the wild-type US11 promoter was used. Some transfection experiments were performed with the addition of a plasmid, pCMV71, that expresses the pp71 tegument protein. Studies have shown that pp71 can enhance the infectivity of HCMV DNA (2). Recombinant virus pools were screened for the presence of the recombinant virus by either Southern blotting or PCR. PCR screening was performed with Platinum Taq (Gibco Life Technologies, Grand Island, N.Y.) on genomic DNA isolated by the GenomicPrep kit, using the primers US11-1 (5′ GTAATGCTTATTCTAGCCCTCTGGG) and US11-2 (5′ AATCACTGCCACCATCATCAGC). These primers will amplify a region of the US11 ORF present in the recombinant viruses but absent in RV699. Pools that showed evidence of the presence of recombinant viruses were then screened for individual viruses lacking the β-glucuronidase gene as previously described (29). The presence of the appropriate mutations in the isolated recombinant viruses was confirmed by Southern blot analysis.

Northern blot analysis.

Total-cell RNA was isolated from infected cells by using the RNeasy system (Qiagen, Chatsworth, Calif.). Equal quantities of RNA (5 μg) were subjected to Northern blot analysis and hybridized to ³²P-radiolabeled probe for US11. RNA levels were quantitated by PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.) analysis. Multiplicity of infection was assessed by stripping the Northern blots and reprobing with a ³²P-radiolabeled probe to the endogenous pp28 gene (UL99).

RESULTS

Identification of US11 promoter elements required for promoter activity.

Analysis of the US11 promoter sequence demonstrated the presence of several putative transcription factor binding sites that could be important for promoter activity (Fig. 1). To assess the role of these sites in US11 promoter activation, we generated a series of nested deletions within the US11 promoter (Fig. 2A) and analyzed their effects on US11 promoter activity in the presence of the viral major IE proteins (Fig. 2B). Deletion of US11 promoter sequences from −470 to −124 had little effect on US11 promoter activity, indicating that promoter elements downstream of −124 were sufficient for US11 promoter activation by the IE proteins. Additional deletion of US11 promoter sequences from −124 to −70 resulted in a drop in promoter activity to 50% of the level for the wild-type promoter. This region contains a consensus CREB binding site at −83, suggesting that this CREB site may contribute to activation of the US11 promoter. Further deletion to −40 abrogated responsiveness of the US11 promoter to the IE proteins. This deletion removes all sequences upstream of the TATA element, including a consensus binding site for the ATF transcription factor located at −53. Together, these data strongly suggest that the CREB and ATF sites within the US11 promoter were required for full promoter activity, consistent with a previous study suggesting a role for this region in US11 promoter activation (27).

FIG. 1.

Sequence of the US11 promoter. The TATA element is indicated in bold. Putative binding sites for the transcription factors ATF, CREB, and CF1 are boxed. In addition, two copies of a direct repeat element (DR1) as well as a palindromic sequence (P1) are indicated.

FIG. 2.

(A) Schematic diagram of the US11 promoter. Putative promoter elements located within the US11 promoter (Fig. 1) and nested deletions generated within the US11 promoter are indicated. (B) Activation of the US11 promoter by the IE proteins. The US11 promoter and the indicated deletion mutants were cotransfected into HFFs with a construct (pSVH) expressing the IE gene region under the control of the MIEP. Cells were harvested 48 h after transfection and assessed for CAT activity. Activity is expressed relative to the wild-type promoter at 100%. Except for the d40 deletion, the average (± standard deviation) of at least three experiments is represented. Results for the d40 deletion represent data from two experiments. ∗, P < 0.001 (student’s t test).

To confirm the role of the CREB and ATF sites in US11 promoter activity, these two elements were mutated within the context of the entire promoter, both individually and in combination. Mutation of the CREB site within the US11 promoter resulted in the insertion of 3 nt (underlined) within the CREB site (CCCGACGTCACTAG to CCCGACGATATCACTAG). Mutagenesis of the ATF site within the US11 promoter resulted in a 3-nt substitution (CCGTGACGTGCAA to CCGTGAATTCCAA). The effect of mutating the US11 ATF and CREB elements on promoter activity was then assessed (Fig. 3). Mutation of the CREB site at −83 in the context of the entire promoter resulted in a decrease in promoter activity to approximately 20% of the wild-type level. These findings strongly suggest that the CREB site is the functional element within this region. The mutation of the CREB element within the entire US11 promoter resulted in a greater loss of promoter activity than was observed upon deletion of this site, most likely due to an effect of the promoter context; similar effects have been observed with other HCMV early promoters (23). The mutation of the ATF site located at −53 abrogated the responsiveness of the US11 promoter to the viral IE proteins. This finding demonstrates that the ATF site is critical for US11 promoter activity. Mutation of both the ATF and CREB sites also resulted in the elimination of US11 promoter activation by the IE proteins. Together, these data show that both the US11 CREB and ATF sites contribute to activation of the US11 promoter by the IE proteins.

FIG. 3.

Activation of the US11 promoter mutants by the IE proteins. The US11 promoter and promoter variants containing mutations in either the CREB site (Cm), the ATF site (Am), or both (CAm) were assessed by cotransfection into HFFs with the IE proteins. Cells were harvested 48 h after transfection and assessed for CAT activity. Activity is expressed relative to the wild-type promoter at 100%. The average (± standard deviation) of two experiments is represented. ∗, P < 0.003.

As the US11 promoter is activated very early after infection (20), we anticipated that the cellular proteins required for activating this promoter were likely present in uninfected cells. Figure 4 demonstrates that proteins present in uninfected nuclear extracts were capable of binding to oligonucleotides containing either the US11 ATF site or the CREB element. Competition analysis with oligonucleotides containing the wild-type CREB and ATF sites was performed. These studies showed that DNA fragments containing the wild-type CREB and ATF sites were capable of efficiently competing for binding to the appropriate probe DNA. Similarly, the US11 ATF and CREB oligonucleotides were also able to compete for binding to other known ATF/CREB binding sites (22). However, oligonucleotides containing the mutated ATF and CREB sites exhibited a greatly reduced ability to compete for the wild-type DNA probes. The reduction in protein binding affinity to the oligonucleotides containing the mutated ATF and CREB sites correlates with the reduced activation exhibited by these mutants in cotransfection experiments. To identify the cellular proteins binding to the CREB and ATF sites, gel supershift analysis was performed with specific antibodies to ATF-1, CREB, and ATF-2 (22). However, none of these antibodies resulted in a change in mobility of the proteins associated with the CREB or ATF site. It was noted that the protein-DNA complexes formed with either the CREB or ATF oligonucleotides migrated to the same position on the gel. In addition, it was found that the CREB oligonucleotide was capable of efficiently competing for binding to the ATF site and vice versa (22). Therefore, these two sites may bind similar factors present in uninfected cell extracts.

FIG. 4.

Gel shift analysis of proteins binding to the US11 CREB and ATF sites. Gel shift assays were performed with probes containing the CREB or ATF site from the US11 promoter. Extracts were obtained from uninfected HFFs. Competition analysis was performed with either a 10-, 25-, or 50-fold excess of cold competitor DNA consisting of either the wild-type binding site or the mutated version of this site. The image was generated with a Hewlett-Packard ScanJet IIcx with Hewlett-Packard HP Deskscan II software (version 2.3.1) and labeled with Microsoft PowerPoint.

Role of the UL82 tegument protein in activating the US11 promoter.

US11 mRNA can be detected within 2 h of viral infection (18). However, this promoter exhibited a weaker response to viral IE proteins than did other viral early promoters (22). This level of activation appears inconsistent with the kinetics of US11 gene expression. We therefore wished to determine if other viral proteins were capable of contributing to US11 promoter activation. One candidate viral protein that could be involved in US11 promoter activation is the viral tegument protein pp71 (UL82). This protein is postulated to enter the cell upon viral infection and has been shown to stimulate the MIEP via ATF and AP-1 sites (31). We accordingly tested the ability of the pp71 protein to stimulate US11 promoter activity (Fig. 5). Figure 5 shows that in the presence of IE proteins, pp71 could stimulate US11 promoter activity six- to sevenfold. To confirm the specificity of the promoter activation, we performed cotransfection experiments using plasmid pCMV71dlPvuI, which expresses a truncated version of the pp71 protein and lacks the ability to activate the MIEP (31). Cotransfection of the IE proteins with plasmid pCMV71dlPvuI did not result in increased US11 promoter activity.

FIG. 5.

(A) Activation of US11 promoter deletions by pp71. pUS11dHCAT and the indicated deletion mutants were cotransfected into HFFs with pSVH and either pCMV71 or pCMV71dlPvuI as described for Fig. 2. Cells were harvested 48 h after transfection and assessed for CAT activity. The results are expressed as fold activation relative to the activation by the IE proteins in the presence of the control plasmid pCMV71dlPvuI and represent the average of three independent experiments. ∗, P = 0.01. (B) Activation of the US11 promoter mutants by the pp71 tegument protein. The US11 promoter and promoter variants containing mutations in either the CREB site (Cm), the ATF site (Am), or both (CAm) were assessed by cotransfection into HFFs with the IE proteins and either pCMV71 or pCMV71dlPvuI as described for Fig. 3. Cells were harvested 48 h after transfection and assessed for CAT activity. The results are expressed as fold activation relative to the activation by the IE proteins in the presence of the control plasmid pCMV71dlPvuI and represent the average of three independent experiments. ∗, P ≤ 0.05.

Previous studies have demonstrated that pp71 activates promoters via ATF and AP-1 sites (31). To determine the US11 promoter elements responsible for pp71 stimulation, we tested the ability of pp71 to activate the US11 deletion mutants. Figure 5A demonstrates that deletion of the CREB site at −83 (pUS11d70CAT) resulted in an approximately twofold drop in the stimulation by pp71 of US11 promoter activity (P = 0.03). Further deletion of the ATF site immediately upstream of the TATA element resulted in a loss of pp71 stimulation. This result suggests that pp71 activates the US11 promoter via the ATF site located at −53. To confirm this, the US11 promoter mutants were tested for the ability to be activated by pp71 (Fig. 5B). Mutation of the CREB site at −83 resulted in a threefold drop in US11 promoter activation by pp71 (P = 0.05), suggesting that this element can contribute to promoter stimulation. Cotransfection of the pp71 protein with the US11 promoter containing a mutation in the ATF site resulted in the loss of US11 promoter stimulation compared to the IE proteins alone. Mutation of both the CREB and ATF elements in the US11 promoter also completely abrogated the ability of this promoter to be activated by pp71. These findings strongly suggest that pp71 can stimulate transcription via the ATF and CREB sites in the US11 promoter, with the ATF site being critical for activation.

Analysis of US11 gene expression in the context of the viral genome.

Our previous results demonstrated a critical role for the two ATF/CREB sites within the US11 promoter in regulating promoter activity in transient assays. To assess the role of the CREB and ATF sites in regulating US11 gene expression during viral infection, we generated recombinant viruses that contained the US11 promoter mutants regulating expression of the endogenous US11 ORF. This approach was possible because the US11 gene is not required for replication of the virus in tissue culture (20). Four recombinant viruses were generated by homologous recombination with the virus RV699, which contains the β-glucuronidase gene inserted into the US11 ORF (20). The first virus, RVUS11, restored the wild-type US11 promoter upstream of the US11 ORF and was used as a control for these experiments. Three additional viruses, RVUS11Cm, RVUS11Am, and RVUS11CAm, contained the US11 promoter with mutations in the CREB site, the ATF site, and both, respectively. The integrity of the viruses was confirmed by Southern blot analysis (22). US11 mRNA expression was then assessed in HFFs infected with the recombinant viruses and compared to the expression of US11 mRNA in AD169-infected cells (Fig. 6). In all cases, data were corrected for multiplicity of infection by assessing the levels of UL99 mRNA expression (Table 1). Levels of expression of the US11 mRNA in AD169-infected cells were high at 8 h after infection and declined thereafter, in accordance with previously reported analysis of US11 mRNA expression (18). Expression of the US11 gene in cells infected with RVUS11 was indistinguishable from that in AD169-infected cells. This result confirms that the RVUS11 virus restores the AD169 phenotype. US11 mRNA expression in RVUS11Cm-infected cells at 8 h after infection was reduced to approximately 24% of that from the wild-type promoter, similar to the result observed when the CREB mutation was assessed in transient assays. Although the overall level of US11 mRNA was reduced, the kinetics of expression in RVUS11Cm-infected cells was identical to that of the wild-type construct. Mutation of the ATF site within the US11 promoter (RVUS11Am) dramatically reduced US11 mRNA expression. Detectable levels of US11 mRNA were observed only at 8 h after infection. At this time point, US11 mRNA was 6% of that observed in AD169-infected cells; thereafter, US11 mRNA levels dropped to below detectable levels. Assessment of the RVUS11CAm virus revealed that mutation of both the CREB and ATF elements within the US11 promoter reduced US11 gene expression to undetectable levels. For all mutant virus constructs, as well as the wild-type RVUS11 virus, expression of the downstream US10 gene was unaffected by the mutations in the US11 promoter region (22). These results clearly demonstrate that the CREB and ATF sites within the US11 promoter cooperate to activate US11 gene expression in the context of the viral genome.

FIG. 6.

Northern blot analysis of recombinant HCMV. HFFs were infected with 5 PFU of the indicated viruses per cell, harvested at the indicated times, and assessed for US11 RNA levels by Northern blot analysis utilizing a ³²P-labeled probe to the US11 gene. RNA levels were quantitated by PhosphorImager analysis. (A) Representative Northern blot analysis of the 1.5-kb US11 mRNA in cells infected with the recombinant viruses, as well as US11 mRNA levels in AD169-infected cells. Values shown in panel B (average of results from two replicate experiments) were corrected for multiplicities of infection by stripping the blot and reprobing with a probe to the UL99 gene and expressed relative to the level of US11 mRNA expression in AD169-infected cells at 8 h after infection.

TABLE 1.

Northern blot analysis of US11 and UL99 mRNA levels expressed in virus-infected cells as quantitated by PhosphorImager analysis

Expt	Virus	mRNA (arbitrary units)
		US11a	UL99b
1	AD169	59.3	71.4
	RVUS11	59.5	81.1
	RVUS11Cm	21.3	126.3
2	RVUS11	108.3	52.5
	RVUS11Cm	34.1	61.2
3	AD169	120.0	140.3
	RVUS11CAm	0.0	66.6
4	AD169	44.0	46.1
	RVUS11CAm	1.9	83.2
	RVUS11Am	7.4	134.5
5	AD169	96.9	18.5
	RVUS11Am	17.4	51.1

^aAt 8 h after infection. 

^bAt 72 h after infection. 

DISCUSSION

Transcriptional regulation in HCMV-infected cells relies on a complex interaction between cellular and viral transactivators (13, 23, 28, 37, 41, 43, 46). Several studies have implicated a role for the transcription factors ATF/CREB in early gene regulation (25, 30, 35, 37, 39). For example, several early promoters can be regulated by ATF/CREB sites in transient assays (30, 35, 37, 39). In addition, a role for ATF/CREB in the activation of the UL54 and UL112-113 promoters at early times in the context of the viral genome has been demonstrated (25, 35). Our present analysis of the US11 promoter revealed that expression of this early gene is also regulated by two ATF/CREB sites within the promoter. The primary regulatory element of the US11 promoter, both in transient assays and in the context of the viral genome, is an ATF site located immediately upstream of the TATA element. In addition to the ATF site, the CREB site at −83 was also involved in US11 promoter activation. In the context of the viral genome, both elements were required for full activation of US11 mRNA expression. These studies therefore add to the growing evidence for a role of ATF/CREB in HCMV early gene regulation.

There are at least two potential mechanisms that could contribute to activation by cellular ATF/CREB at early times in HCMV-infected cells. First, the pp71 or UL82 viral tegument protein could influence HCMV gene activation through ATF/CREB sites at the initial stages of infection. This protein has previously been shown to up-regulate the MIEP via ATF and AP-1 sites within the MIEP (31). In our present study, pp71 in combination with the IE proteins resulted in enhanced activation of the US11 promoter. Analysis of the US11 mutants revealed that pp71 activation of this promoter was also dependent on the presence of the ATF and CREB sites within this promoter. Recently, it has been demonstrated that pp71 protein from input virus can survive in the nucleus for at least 3 h postinfection (12), a time sufficient to activate some early promoters such as US11. However, it is also possible that the pp71 protein could enhance US11 promoter activity by increasing the level of the IE proteins via its effect on the MIEP. Indeed, preliminary data indicate that pp71 can increase steady-state levels of IE proteins approximately twofold (22). However, at this stage we cannot rule out the possibility that pp71 acts via a more direct mechanism to enhance IE activation of the US11 promoter.

A second factor that could influence HCMV activation via ATF/CREB proteins is the IE86 protein. The IE86 protein has been shown to interact with the CREB protein in vitro (30, 37). This interaction could contribute to activation via CREB sites at early times after infection (35, 37). However, little is known regarding the affinity of IE86 interactions with other ATF/CREB proteins. Our present study found that the ATF/CREB proteins binding to the US11 promoter could not be supershifted by antibodies to ATF-1, ATF-2, or CREB. Further studies will be required to identify the ATF/CREB family member(s) involved in US11 promoter activation and to determine if this protein can also interact with IE86. However, it is clear that IE86 has the potential to play a critical role in transcriptional activation via ATF/CREB sites within early gene promoters at early times after infection. The role of IE86 interactions with ATF/CREB in regulating gene expression at later times of infection is less straightforward. High levels of the IE86 protein can be found in infected cells at late times after infection (43). In addition, our previous studies have demonstrated that the DNA binding activity of one ATF/CREB subtype, ATF-1, is increased at late times after infection (25). However, analysis of the UL54 and UL112-113 promoters in the context of the viral genome revealed that the ATF/CREB sites are less critical for the activation of these promoters at late times (25, 35). In addition, results of the present study demonstrate that two ATF/CREB sites within the US11 promoter are insufficient to activate the US11 promoter at late times after infection. Thus, even though ATF/CREB factors and IE86 are present in infected cells at late times, IE86 interactions with ATF/CREB do not appear to be involved in the activation of early gene promoters at late times. One possibility is that differential affinity of IE86 for ATF/CREB subtypes may play a role in regulating transcriptional activation at various stages of infection. Alternatively, IE86 function could be modified at late times after infection. For example, it is known that IE86 is phosphorylated in infected cells (11). Differential phosphorylation could potentially influence the ability of IE86 to interact with ATF/CREB subtypes. The interaction of IE86 with IE86-related proteins such as the p40 protein present in late infected cells (15) could also modify the activity of IE86 at this stage of infection.

These studies analyzing the role of ATF/CREB in early gene activation in the context of the viral genome have revealed that ATF/CREB plays an important role in transcriptional activation at early but not late times (24, 25, 35). This finding suggests that some early gene promoters utilize differential mechanisms of promoter activation at early and late times after infection. Two of these studies have focused on early genes in the E2 subclass of HCMV early genes, UL54 and UL112-113 (24, 25, 35). The E2 subclass of early genes is characterized by mRNA expression at relatively constant levels throughout the course of infection (48). Analysis of the UL54 promoter revealed that cellular factors such as the IR1 binding protein, which contains Sp1 as one of its components, and ATF/CREB, were required for the activation of this promoter at early times after infection (24, 25). However, these factors are not required for activation of this promoter at late times after infection. Similarly, the UL112-113 promoter is regulated primarily by a CREB site and an IE86 binding site at early times after infection (35). At late times, the CREB site within this promoter plays a less significant role in UL112-113 promoter regulation. Thus, early promoters of the E2 subclass rely on existing cellular factors such as ATF/CREB and Sp1 for activation at early times, with additional factors required at late times after infection. There is some evidence for the use of alternate promoter start sites for both the UL54 and UL112-113 promoters in late transcription (22, 35). This finding suggests that different regions of the promoter may be required for activation of transcription at late times. In addition, other viral transactivators could also play a role in the activation of these early promoters at late times (14, 24). In contrast to the E2 promoters, the US11 gene has been characterized as an E1 early gene, in that mRNA levels are high at early times after infection and decline at late times (43). Our analysis of the US11 promoter revealed that two ATF/CREB sites were sufficient for the activation of this promoter at early times. Thus, the E1 subclass of early genes may represent a class characterized by very simple promoters, with existing cellular factors, IE proteins, and virion components such as pp71 sufficient for transcriptional activation. For these promoters, no additional mechanisms for transcriptional activation would be proposed to function at the late stages of infection. Thus, these studies are allowing us to begin to dissect the mechanisms of regulation of the different kinetic subclasses of HCMV early genes.