Brd4 Is Required for E2-Mediated Transcriptional ... - Journal of Virology

2 downloads 0 Views 1MB Size Report
May 30, 2006 - Bromodomain protein 4 (Brd4) has been identified as the cellular binding target through which the E2 protein of bovine papillomavirus type 1 ...
JOURNAL OF VIROLOGY, Oct. 2006, p. 9530–9543 0022-538X/06/$08.00⫹0 doi:10.1128/JVI.01105-06

Vol. 80, No. 19

Brd4 Is Required for E2-Mediated Transcriptional Activation but Not Genome Partitioning of All Papillomaviruses† M. G. McPhillips, J. G. Oliveira, J. E. Spindler, R. Mitra, and A. A. McBride* Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Received 30 May 2006/Accepted 20 July 2006

Bromodomain protein 4 (Brd4) has been identified as the cellular binding target through which the E2 protein of bovine papillomavirus type 1 links the viral genome to mitotic chromosomes. This tethering ensures retention and efficient partitioning of genomes to daughter cells following cell division. E2 is also a regulator of viral gene expression and a replication factor, in association with the viral E1 protein. In this study, we show that E2 proteins from a wide range of papillomaviruses interact with Brd4, albeit with variations in efficiency. Moreover, disruption of the E2-Brd4 interaction abrogates the transactivation function of E2, indicating that Brd4 is required for E2-mediated transactivation of all papillomaviruses. However, the interaction of E2 and Brd4 is not required for genome partitioning of all papillomaviruses since a number of papillomavirus E2 proteins associate with mitotic chromosomes independently of Brd4 binding. Furthermore, mutations in E2 that disrupt the interaction with Brd4 do not affect the ability of these E2s to associate with chromosomes. Thus, while all papillomaviruses attach their genomes to cellular chromosomes to facilitate genome segregation, they target different cellular binding partners. In summary, the E2 proteins from many papillomaviruses, including the clinically important alpha genus human papillomaviruses, interact with Brd4 to mediate transcriptional activation function but not all depend on this interaction to efficiently associate with mitotic chromosomes.

multifunctional viral protein, E2 (5, 25, 31, 42, 47). BPV1 E2 protein noncovalently tethers extrachromosomal viral genomes to cellular mitotic chromosomes through an interaction with a cellular factor. One component of this complex was recently identified as bromodomain-containing protein 4, Brd4 (61). Consistent with this finding, we have shown that mutated BPV1 E2 proteins that are unable to associate with mitotic chromosomes do not interact with Brd4 in vitro and that expression of Brd4 reconstitutes E2-mediated plasmid segregation in Saccharomyces cerevisae (7, 11). Moreover, BPV1 E2 greatly stabilizes the association of Brd4 with chromatin both in interphase and in mitosis, showing that BPV1 E2 actively modulates its chromosome binding partner and does not passively associate with an opportune cellular factor (34). Brd4 is a member of the BET family of bromodomain proteins, which remain bound to condensed chromosomes throughout mitosis by binding to acetylated lysine residues of histones H3 and H4 (14, 17). Recent reports also described that Brd4 binds and positively regulates transcription elongation factor b (P-TEFb), indicating that Brd4 plays a role in regulating the general RNA polymerase II-dependent transcription machinery (27, 59). In addition to its role in the extrachromosomal maintenance of PV genomes, E2 also functions as a transcriptional regulator of viral gene expression and is important in the initiation of viral DNA replication, in complex with the viral E1 protein (33). The E2 protein comprises three regions: an amino-terminal transactivation domain that is separated from the carboxy-terminal DNA binding and dimerization domain by a flexible, nonconserved hinge region (33). The transactivation domain of BPV1 E2 is required for interaction with condensed chromosomes and associates with the carboxy-terminal domain (CTD) of Brd4 (61). The Brd4 CTD acts as a dominant-

To date, more than 100 different types of papillomaviruses, representing 17 distinct genera, have been identified in a wide range of mammals as well as in birds and reptiles (10, 13). Members of the Papillomaviridae family have a specific tropism for squamous epithelium. For example, mucosal human papillomavirus (HPV) types, such as HPV6, HPV11, HPV16, and HPV31, infect genital and sometimes oral epithelia. Cutaneous HPV types, such as HPV1, -2, -3, and -4, cause plantar and palmar warts, whereas HPV5 and HPV8 are associated with epidermodysplasia verruciformis. Certain fibropapillomaviruses, such as bovine papillomavirus type 1 (BPV1), can also infect fibroblasts. Despite this diversity, all papillomaviruses are capable of establishing persistent infections by maintaining their viral genomes as low-copy-number extrachromosomal elements in mitotically active cutaneous and mucosal epithelial cells. Papillomavirus genomes do not contain centromeric sequences and hence require an alternative and efficient means of retaining extrachromosomal viral genomes within the nucleus and distributing them to daughter cells. The mechanism of papillomavirus genome segregation and chromosomal attachment in BPV1 has been extensively studied (25, 31, 47). BPV1-transformed mouse cells maintain extrachromosomal viral genomes indefinitely and therefore provide a useful system in which to study genome maintenance (30). Faithful segregation of BPV1 genomes is achieved through the actions of the

* Corresponding author. Mailing address: Laboratory of Viral Diseases, NIAID, NIH, Building 4, Room 137, 4 Center Dr., MSC 0455, Bethesda, MD 20892-0455. Phone: (301) 496-1370. Fax: (301) 4801497. E-mail: [email protected]. † Supplemental material for this article may be found at http://jvi .asm.org/. 9530

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

negative inhibitor to disrupt the interaction of E2 with fulllength Brd4, resulting in the dissociation of BPV1 E2 and viral genomes from mitotic chromosomes (61, 62). Furthermore, Brd4 CTD expression inhibits BPV1 transformation of mouse cells and reduces the number of BPV1 viral genomes in previously transformed cells (61, 62). Brd4 and BPV1 E2 interact throughout the cell cycle, and so, Brd4 may be involved in additional processes critical to the viral life cycle. Recent studies showed that Brd4 mediates the transcriptional activation function of E2 as well as plays a role in the viral DNA replication process, although the latter function appears to be independent of its ability to associate with BPV1 E2 (26, 34, 45). Viral protein-mediated tethering of genomes to mitotic chromosomes is also a characteristic of genome segregation and maintenance in other extrachromosomal viruses, such as Epstein-Barr virus and human herpesvirus 8 (4, 18, 19). We recently showed that, similar to those of BPV1, many additional animal and human papillomavirus E2 proteins also bind to mitotic chromosomes (37). Interestingly, these studies further indicated that while the association of E2 proteins with mitotic chromosomes is a common theme among diverse PV E2s, the E2 chromosomal binding targets may be different for different viruses (37). In the present study, we used the same panel of animal and human PV E2 proteins to examine the role of Brd4 in the association of PV E2 proteins with mitotic chromosomes. We found that all PV E2s tested thus far can bind Brd4 in vitro. Moreover, expression of the dominant-acting negative inhibitor Brd4 CTD reduces transcriptional activation by all E2 proteins. All E2 proteins examined associated with mitotic chromosomes, although prefixation treatment of cells expressing alpha PV E2s was necessary to detect these E2s on condensed chromosomes, as previously observed (37). Several of the E2 proteins did colocalize on mitotic chromosomes with Brd4, as has been observed with BPV1 E2. However, a subset of E2 proteins, including members of the alpha papillomaviruses, associated with condensed chromosomes in the absence of detectable chromosome-bound Brd4. Furthermore, point mutations in E2 that disrupt the E2-Brd4 interaction did not affect the mitotic association of these proteins. Thus, while Brd4 is required for the transcriptional activation function of E2, it is not the sole chromosomal binding target for all papillomaviruses.

MATERIALS AND METHODS Plasmids. The pMEP E2 plasmid expression vector, from which all normal and mutated animal and human PV E2 genes are expressed, has been described previously (38), as has a description of their regulated expression from the metallothionein promoter. A detailed description of the generation of the FLAG-tagged animal and human E2 proteins (those from BPV1, European elk PV [EEPV], canine oral PV [COPV], cottontail rabbit PV [CRPV], rabbit oral PV [ROPV], deer PV (DPV), HPV1a, HPV4, HPV11, HPV16, HPV31, and HPV57), together with details of the codon optimization of HPV4, HPV11, HPV16, and HPV31, is also available elsewhere (37). BPV1 and HPV16 E2 proteins containing alanine substitutions in residues R37 and I73 were described previously (7, 48). Standard mutagenesis procedures were used to mutate HPV31 E2 residues R37 and I73 to alanines. For in vitro translation, the FLAG tag was removed from each of the E2 proteins, which where subsequently cloned into the pTZ19U vector by using conventional cloning techniques. HPV18 E2 protein, amino acids 2782 to 4794, was cloned into pTZ19R for in vitro translation. The C-terminal domain of the mouse Brd4 protein, amino acids 1051 to 1400, was PCR amplified from pBSK MCAP WT (kindly provided by Keiko Ozato) and

9531

cloned into pCEP4 (Invitrogen). The E2-responsive luciferase reporter plasmid pBS1073 contains the firefly luciferase gene downstream from four E2 binding sites and the herpes simplex virus type 1 thymidine kinase promoter (a gift from B. Sugden) (22). Additional luciferase reporter constructs contained the promoter regions from Rous sarcoma virus long terminal repeat, cytomegalovirus immediate-early region, ubiquitin C, and c/EBP␤ genes in pGL3-Basic (Promega) (gifts from T. Kristie) (28, 36). Establishment of stable pMEP-E2 cell lines. CV-1-derived lines were generated by transfection of the various pMEP-E2 plasmids using Fugene (Roche) and selection with 200 ␮g/ml hygromycin B (Roche). After 2 weeks, drug-resistant colonies were pooled and cultures were expanded. Immunoblot analysis showing the levels of E2 proteins expressed in these cell lines has been published previously (37). Indirect immunofluorescence. For analyses of interphase cells, cells were seeded directly onto glass slides (Superfrost Plus) and grown for 48 h. E2 expression was induced by the addition of 1.0 ␮M CdSO4 for 3 to 4 h prior to fixation. To analyze mitotic cells, cells were blocked in S phase by the addition of 2 mM thymidine for 14 to 16 h. The thymidine block was released, and the cells were cultured for 9 h, with E2 expression induced by the addition of 1.0 ␮M CdSO4 for the last 3 to 4 h. Cells were either fixed directly in 4% paraformaldehyde for 20 min at room temperature and permeabilized with 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 15 min at room temperature or, where indicated, permeabilized for 30 s in microtubule-stabilizing buffer (BRB80 [80 mM PIPES, pH 6.8, 1 mM EGTA, 1 mM MgCl2] plus 4 mM EGTA) (39) plus 0.1% Triton X-100, followed by a 1-min incubation in microtubule-stabilizing buffer (39) containing 4% paraformaldehyde and fixation in methanol for 30 min at ⫺20°C. E2 FLAG proteins were detected with Sigma monoclonal anti-FLAG M2 (1:500) and with a secondary antibody conjugated to fluorescein isothiocyanate (FITC) (1:100 dilution; Jackson Immunochemicals). Brd4 was detected with a 1:500 dilution of the rabbit polyclonal antibody 2290 (14, 15) and with goat anti-rabbit immunoglobulin G conjugated to Texas Red (1:100 dilution; Jackson Immunochemicals). Slides were mounted in Vectashield mounting medium (Vector Laboratories) containing 1 ␮g/ml of DAPI (4⬘,6⬘-diamidino-2-phenylindole). Immunofluorescent staining was detected, and digital images were captured with a Leica TCS-SP2 laser scanning confocal imaging system. Protein extraction. For analysis by immunofluorescence, slides were submerged in ice-cold CSK extraction buffer (10 mM PIPES, pH 6.8, 30 mM sucrose, 3 mM MgCl2, 1 mM EGTA, 0.5% Triton X-100, complete protease inhibitor [Roche]) (8) containing 300 mM NaCl for 15 min, followed by fixation in 4% paraformaldehyde-PBS for 20 min. Unextracted slides were fixed directly in 4% paraformaldehyde-PBS for 20 min, as described above. For Western blot analysis, cell monolayers were covered with ice-cold CSK extraction buffer (8) containing 300 mM NaCl for 15 min. Following removal of the salt solution, proteins were extracted in 50 mM Tris-HCl, pH 6.8, 2% (wt/vol) sodium dodecyl sulfate (SDS), 10% glycerol, and complete protease inhibitor (Roche). Cells were extracted directly in 50 mM Tris-HCl, pH 6.8, 2% (wt/vol) SDS, 10% glycerol, and complete protease inhibitor (Roche) for untreated samples. Total protein concentrations in the extracts were determined by the use of a bicinchoninic acid protein assay kit (Pierce). For each sample, 5 ␮g of total protein was separated by SDS-polyacrylamide gel electrophoresis (PAGE) and electrotransferred onto Immobilon-P polyvinylidene difluoride membranes (Millipore). Western blotting was performed according to standard protocols, and E2 proteins were detected with anti-M2 FLAG monoclonal antibody (Sigma) followed by horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (Pierce). Protein bands were detected by using the chemiluminescence reagent SuperSignal West Dura (Pierce). Data were collected on a Kodak Image Station 440CF, and signals were quantitated using Kodak one-dimensional image analysis software. Transactivation assay. CV-1 cells (2.5 ⫻ 105/dish) or C-33a cells (1 ⫻ 106/dish) (60) were plated onto 6-cm-diameter dishes. One day later, the cells were cotransfected with 100 ng of pMEP-E2 expression plasmid, 1 ␮g pCEP empty vector or 1 ␮g pCEP Brd4 CTD, and 1 ␮g of pBS1073, using 6 ␮l of Fugene transfection reagent per dish. Duplicate transfections were carried out for each sample. Cells were harvested at 48 h posttransfection. After being washed with PBS, cells were lysed with 1 ml of 1⫻ cell culture lysis reagent (Promega) per dish. Cell extracts were assayed for firefly luciferase activity with Promega’s luciferase assay system. Luciferase activity was measured in a Zylux Femtomaster FB12 luminometer, and measurements were carried out in duplicate for each sample. The luciferase activity, measured as relative light units per second, is expressed relative to the activity of wild-type BPV1 E2, which was set at 100% activity. Purification of baculovirus mouse Brd4 from SF-9 cells. Mouse Brd4 tagged with hexahistidine on the N terminus and FLAG on the C terminus was ex-

9532

MCPHILLIPS ET AL.

pressed and extracted from SF-9 cells as described previously (7, 32). Brd4 protein was purified by incubation with Ni-nitrilotriacetic acid agarose at 4°C and eluted in elution buffer (10% glycerol, 20 mM Tris HCl, pH 8.0, 0.5 M KCl, 200 mM imidazole, 5 mM dithiothreitol [DTT], 0.5 mM phenylmethylsulfonyl fluoride, and 1⫻ complete EDTA-free protease inhibitor cocktail [Roche]). Eluted protein was dialyzed in PBS with 1 mM DTT and 10% glycerol and diluted in extraction buffer (10% glycerol, 20 mM Tris HCl, pH 8.0, 0.2 mM EDTA, 0.1% Tween, 0.15 M NaCl, and 1⫻ complete EDTA-free protease inhibitor cocktail [Roche]) and added to an M2 FLAG agarose chromatography column (Sigma) for 14 h at 4°C. Brd4 protein was eluted in extraction buffer containing 150 ng/␮l FLAG peptide and further dialyzed against extraction buffer to remove the FLAG peptide. Brd4-E2 pull-down. Purified Brd4 protein was bound to M2 FLAG agarose by incubation in 20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.1% Tween 20, 1 mM DTT, and 1⫻ complete EDTA-free protease inhibitor cocktail (Roche) at room temperature for 1 hour. E2 proteins were synthesized and labeled with [35S]methionine using Promega’s TNT quick-coupled reticulocyte lysate system, and the concentrations of each were normalized to each other based upon the number of methionines. Equivalent amounts of 35S-labeled E2 proteins were incubated with the Brd4 agarose beads for 1 hour. Bound proteins were eluted in 2⫻ SDS sample buffer and separated by SDS-PAGE. For quantitative assays, the amount of 35S-E2 bound was measured directly by suspending the beads in CytoScint fluid (ICN Biomedicals) and counted in a 1450 Wallac Microbeta counter.

RESULTS All papillomavirus E2 proteins interact with Brd4 but with various efficiencies. Brd4 has been identified as the chromosomal binding partner for BPV1 E2 (61). Previously, using a panel of mutated BPV1 E2 proteins, we have shown a strong correlation between mitotic chromosome binding activity, transcriptional activation, and interaction with Brd4 in vitro (7). Furthermore, colocalization of Brd4 and BPV1 E2 in interphase cells requires an E2 protein with a transcriptionally functional transactivation domain (34). HPV16 E2 has also been shown to interact with Brd4 by coimmunoprecipitation assays, and in a series of mutated HPV16 E2 proteins, there was good correlation between amino acids important for transcriptional activation and Brd4 binding (45, 61). To further analyze the role of the E2-Brd4 interaction, a series of 14 different animal and human PV E2 proteins were tested for their abilities to associate with Brd4 in an in vitro binding assay. For this assay, FLAG-tagged Brd4 protein was prebound to agarose beads and then incubated with in vitrotranslated E2 protein. As shown in Fig. 1A, each E2 protein could associate with Brd4. To fully investigate the binding competence of each E2 protein for Brd4, increasing amounts of each translated E2 protein were added to the Brd4 beads, and bound 35S-labeled E2 proteins were counted directly. As shown in Fig. 1B, there was a wide range of affinities of each E2 protein for Brd4. In general, the alpha PV E2 proteins bound Brd4 with less efficiency than other PV E2 proteins. Previous experiments by Baxter et al. showed that a double substitution of BPV1 E2 residues R37 and I73 with alanine disrupted the interaction with Brd4 (7). Schweiger et al. also showed that mutation of the HPV16 E2 R37 residue to alanine reduced binding to Brd4 whereas mutation of I73 to alanine resulted in no binding (45). The effect of the double mutation on the Brd4 binding efficiencies of BPV1 E2, HPV16 E2, and HPV31 E2 was tested by comparing the mutated proteins with their wild-type counterparts. Each R37A I73A E2 protein bound Brd4 with much less efficiency than the wild-type coun-

J. VIROL.

terpart (Fig. 1C). Thus, the same E2 residues are important for Brd4 interaction in several E2 proteins. A dominant-negative domain of Brd4 represses the transcriptional activation function of all E2 proteins. Since all E2s are capable of interacting with Brd4 and there is a strong correlation between transactivation capability and Brd4 interaction, we wanted to determine whether this interaction was important for E2-mediated transcriptional activation. To do this, a portion of Brd4 was expressed to act as a dominantnegative protein and disrupt the E2-Brd4 complex. This consisted of the C-terminal 350 amino acids of mouse Brd4 (amino acids 1051 to 1400) constitutively expressed from the cytomegalovirus promoter. This protein showed a strictly nuclear localization by immunofluorescence analysis (data not shown) and efficiently bound to BPV1 E2 (data not shown). The Brd4 CTD was tested for its ability to abrogate E2-mediated transactivation from an E2-responsive promoter expressing the luciferase reporter gene. E2 activates transcription by binding directly to cis elements consisting of multiple E2 binding sites upstream from promoters (50). For this assay, the E2-responsive reporter plasmid, pBS1073, was cotransfected into CV-1 and C33a cells with expression plasmids for 12 different E2 proteins and either the empty control vector, pCEP, or pCEP expressing the Brd4 CTD. The results, averaged from several such experiments, are shown in Fig. 2A. As shown previously, all E2s could transactivate the E2-responsive promoter in both CV-1 cells (Fig. 2A) and C33a cells (37; data not shown). Expression of the Brd4 CTD reduced transcriptional activation by all E2 proteins, regardless of variations in their relative transactivation activities (Fig. 2A). The repressive effect of the Brd4 CTD was also evident in C33a cells (data not shown). To further confirm the important role of Brd4 in E2-mediated transactivation, the transactivation activity of E2 proteins containing the R37A I73A double mutation, shown to disrupt the E2-Brd4 interaction (Fig. 1), was tested. Previously, we have shown that mutation of these residues in BPV1 E2 reduces transactivation function, abrogates mitotic chromosome association, and disrupts the interaction of this E2 protein with Brd4 (6, 7). BPV1 E2 R37A I73A can still activate transcription to a limited extent (approximately 8% of wild-type levels) (Fig. 2B). HPV31 E2 R37A I73A also showed a reduction in transcriptional activation to approximately 20% of wild-type HPV31 E2 levels (Fig. 2B). In the presence of the Brd4 CTD, transactivation by the wild-type BPV1 and HPV31 E2 proteins was significantly reduced to approximately 20% of that of wildtype BPV1 E2 (Fig. 2B). However, the residual level of transactivation from the mutated E2 proteins remained relatively unaffected, showing that since these proteins are no longer able to interact with Brd4, they are resistant to the repressive effects of the dominant-negative Brd4 CTD (Fig. 2B). This observation is further confirmation that the interaction of E2 and Brd4 is important for transcriptional activation of all E2 proteins. As shown in Fig. 2C, the Brd4 CTD had no effect on transactivation from a number of viral and cellular promoters, confirming that the repressive effect of the Brd4 CTD is specific to E2-mediated transcriptional activation. Most E2 proteins stabilize the association of Brd4 with chromatin. The presence of BPV1 E2 greatly stabilizes the interaction of Brd4 with chromatin (34). Thus, the abilities of 12 of the E2 proteins to stably bind to chromatin and in turn

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

9533

FIG. 1. E2 proteins interact with Brd4 with various affinities. (A) 35S-labeled in vitro-translated E2 proteins were incubated with purified FLAG-Brd4 protein bound to anti-FLAG M2 agarose affinity gel. Shown is an autoradiograph of SDS-PAGE analysis of the bound E2 proteins. (B) Increasing amounts of 35S-labeled E2 proteins were incubated with a constant amount of Brd4 agarose gel. Bound E2 proteins were counted, and the amount of E2 bound is plotted against the amount added to the binding reaction. (C) A binding efficiency assay was carried out as described for panel B with wild-type and R37A I73A mutated versions of BPV1, HPV16, and HPV31 E2 proteins.

stabilize the chromatin association of Brd4 were assayed. Protein binding to chromatin was assessed by protein extraction (14, 34, 46). BPV1 E2 remains tightly bound to chromatin in moderate salt concentrations and stabilizes the chromatin association of Brd4, which is easily extracted under these conditions from cells that do not express E2 (14, 34). Cells were extracted using CSK buffer containing 300 mM NaCl, and retention of E2 and Brd4 proteins under these conditions was

measured by indirect immunofluorescence and by Western blotting. CV-1 cell lines stably expressing each of the E2s were grown directly on glass slides prior to extraction. For indirect immunofluorescence, E2 proteins were detected using the monoclonal M2 anti-FLAG antibody, whereas Brd4 was detected using the 2290 anti-Brd4 polyclonal antiserum. Representative images of BPV1, HPV8, and HPV16 E2 proteins are shown in

9534

MCPHILLIPS ET AL.

FIG. 2. Expression of a Brd4 CTD represses E2 transcriptional activation. (A) CV-1 cells were cotransfected with 1 ␮g pBS1073 (E2responsive reporter), 100 ng of the pMEP E2 expression plasmids, and either 1 ␮g pCEP empty vector (light-gray bars) or 1 ␮g pCEP Brd4 CTD dominant-negative expression plasmid (dark-gray bars). (B) CV-1 cells were cotransfected with 1 ␮g pBS1073, 100 ng of BPV1 E2, BPV1 E2 R37I73, HPV31 E2, or HPV31 E2 R37I73, and 1 ␮g pCEP (light-gray bars) or 1 ␮g pCEP Brd4 CTD (dark-gray bars). Luciferase activity is expressed relative to the activity of wild-type BPV1 E2, which was set at 100% activity. (C) CV-1 cells were cotransfected with 1 ␮g pRSV-luc, pCMV-luc, pUB-luc, or pCEPB/B-luc and either 1 ␮g pCEP (light-gray bars), which was set at 100% activity, or 1 ␮g pCEP Brd4 CTD (dark-gray bars).

J. VIROL.

Fig. 3A. The data obtained from the entire panel of E2 proteins is shown in Fig. S1 in the supplemental material. Each E2 protein showed distinct nuclear localization (Fig. 3A, left side). The Brd4 protein was also detected only in the nucleus and was localized in a speckled pattern, regardless of E2 expression (Fig. 3A). Following extraction in 300 mM NaCl-CSK buffer, BPV1 E2, tagged with an N-terminal FLAG tag, was tightly bound to chromatin and stabilized the chromatin association of Brd4, in agreement with our previous results using nontagged BPV1 E2 (Fig. 3A, right side) (34). Therefore, the N-terminal FLAG epitope does not interfere with the behavior of the BPV1 E2 protein, as shown previously by Oliveira et al. (37). Similar results were obtained with EEPV, COPV, CRPV, ROPV, HPV1a, HPV4, and HPV8 E2 proteins; each protein was tightly bound to chromatin and stabilized the association of Brd4 with chromatin (Fig. 3) (see Fig. S1 in the supplemental material). Notably, only members of the alpha papillomaviruses, HPV 11, HPV16, HPV31, and HPV57 E2 proteins, were not detected following extraction in 300 mM NaCl-CSK buffer (Fig. 3) (see Fig. S1 in the supplemental material). Moreover, little or no Brd4 protein could be detected in these extracted cells. This suggests that the alpha PV E2 proteins are not tightly bound to chromatin and cannot stabilize the chromatin association of Brd4. Parallel salt extraction experiments with C33a cell lines stably expressing each of the human papillomavirus E2 proteins gave similar results (data not shown). To ensure that the inability to detect alpha PV E2s by immunofluorescence following extraction was not the result of masking of the FLAG epitope, the retention of E2 protein was also examined by Western blot analysis. CV-1 cell lines expressing BPV1, HPV4, HPV8, HPV11, HPV16, HPV31, and HPV57 E2 proteins were extracted as a monolayer, using the same extraction buffer as that described above. Equivalent amounts of protein lysate prepared from untreated cell monolayers or monolayers that had previously been extracted with 300 mM NaCl-CSK buffer were separated by SDS-PAGE and immunoblotted using the anti-FLAG antibody, M2. The retention of E2 protein following extraction can be compared for each cell line in the representative immunoblot shown in Fig. 3B. Equal amounts of total protein lysate were analyzed, and so, the chromatin/nuclear matrix-associated protein fraction is enriched approximately eightfold relative to that of the unextracted protein samples. In agreement with the results of salt extraction on slides, BPV1, HPV4, and HPV8 E2 proteins were resistant to salt extraction. In each of these cases, more E2 protein was detected in the retained samples than in the unextracted samples (Fig. 3B). HPV11, HPV16, HPV31, and HPV57 E2 proteins were easily extracted under these conditions, and no enrichment was detected in the retained samples relative to that detected in the corresponding untreated controls. Therefore, these E2 proteins are not retained after salt extraction and thus are not tightly bound to chromatin (Fig. 3B). In three such experiments (Fig. 3B, lower panel), this equates to an 18 to 60% retention of BPV1, HPV4, and HPV8 E2 proteins compared to a 2 to 9% retention of the alpha papillomavirus E2s. Various patterns of colocalization of E2 and Brd4 on mitotic chromosomes. Previously, we have shown that BPV1 E2 and Brd4 colocalize specifically in distinct speckles on

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

9535

FIG. 3. (A) E2 protein was detected by an FITC-labeled antibody (green), Brd4 expression was detected by a Texas Red-labeled antibody (red), and cellular DNA was detected by DAPI staining (blue). The panels on the left show untreated cells, and those on the right show cells after extraction in 300 mM NaCl-CSK buffer. (B) (Top) Western blots of protein extracts from CV-1 cell lines stably expressing E2 proteins, as indicated, before (U) and after (R) salt extraction. The pMEP lanes indicate cells containing the empty vector. E2 protein was detected with anti-FLAG M2 antibody. (Bottom) Quantification of E2 protein retention. The percentage of retention of E2 protein was averaged from three experiments by dividing the protein retained (R) by the corresponding untreated control (U).

mitotic chromosomes in monkey CV-1 cells and mouse C127 cells and that E2 expression resulted in relocalization and stabilization of the Brd4 protein into punctate complexes on the mitotic chromosomes (34). Our group has also previously shown that all papillomavirus E2 proteins can bind to mitotic chromosomes (37). To investigate the interaction of all the E2 proteins in our panel with Brd4 on mitotic chro-

mosomes, the localization of Brd4 and E2 was assessed by immunofluorescence analysis of mitotic CV-1 cells expressing each E2 protein. Representative images of the location of each papillomavirus E2 protein in mitotic cells are shown in Fig. 4A to C. However, a close examination of the mitotic Brd4 binding pattern reveals distinct differences among the group of E2

9536

MCPHILLIPS ET AL.

J. VIROL.

FIG. 4. The E2 proteins show various patterns of colocalization with Brd4 on mitotic chromosomes. (A) Cells expressing the empty pMEP vector, BPV1 E2, EEPV E2, ROPV E2, and HPV1a E2. (B) Cells expressing HPV8 E2, HPV4 E2, CRPV E2, and COPV E2. (C) Cells expressing HPV11 E2, HPV16 E2, HPV31 E2, and HPV57 E2. E2 protein, as detected by an FITC-labeled antibody, is shown in green. Brd4 protein, as detected by a Texas Red antibody, is shown in red. Colocalization of these proteins appears as yellow in the merged images. Cellular DNA was detected by DAPI staining and is shown in blue.

proteins. On the basis of these differences, we have assigned the E2 proteins into four distinct and separate groups. E2 proteins from EEPV, ROPV, and HPV1a were similar to BPV1 E2 on the basis of their ability to bind mitotic chromo-

somes and colocalize perfectly with Brd4 (Fig. 4A). For comparison, a representative image of the empty pMEP control is also shown in Fig. 4A. In agreement with previous observations, some Brd4 staining was detected in the cytosol but there

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

9537

FIG. 4—Continued.

was no specific association with mitotic chromosomes in the absence of E2 protein binding. E2 proteins from papillomaviruses CRPV, HPV4, and HPV8 were also easily detected on mitotic chromosomes, but the E2 staining patterns differed somewhat. For example, BPV1 E2 is observed to be distributed in speckles over the arms of the chromosomes in what appears to be a fairly random manner, while HPV8 E2 binds to the pericentromeric region of mitotic chromosomes (37). This suggested that the chromosomal binding target was different for this group of E2 proteins. This hypothesis is supported by an examination of the Brd4 binding pattern in mitotic cells in the presence of these E2 proteins (Fig. 4B). In the HPV8 E2 image shown in Fig. 4B, Brd4 protein colocalizes with a minor subpopulation of E2 speckles on mitotic chromosomes but the majority of the HPV8 E2 protein binds to the pericentromeric regions of mitotic chromosomes in the absence of detectable Brd4. A sim-

ilar pattern was observed for HPV4 E2 and CRPV E2. In both cases, a small amount of chromosome-bound E2 protein colocalizes with Brd4 but, for the majority of E2, there was no obvious association with Brd4. Therefore, HPV8 E2, HPV4 E2, and CRPV E2 can associate with mitotic chromosomes independently from Brd4 binding. The COPV E2 protein seems to exist in a group of its own due to an unusual staining pattern of E2 and Brd4 on mitotic chromosomes (Fig. 4B). Lower levels of COPV E2 protein are detected in mitosis than those of most other E2s. However, despite the reduced E2 expression level, Brd4 was always easily and abundantly detected on any mitotic chromosomes to which COPV E2 was also bound. Furthermore, Brd4 was often detected on mitotic chromosomes in the absence of any detectable colocalized E2 protein. COPV E2 is the only protein that displays this particular pattern of enhanced Brd4 staining in the absence of clearly observable E2 protein.

9538

MCPHILLIPS ET AL.

J. VIROL.

FIG. 4—Continued.

The fourth group consists of the four E2 proteins from the alpha papillomaviruses, HPV11, HPV16, HPV31, and HPV57. Representative metaphase images are shown in Fig. 4C for each of the alpha PV E2s. As shown previously, these E2s did not stably interact with chromosomes throughout mitosis (37). In each case, E2 protein was expressed in the metaphase cell but was dispersed throughout the cytosol and showed no specific association with the condensed chromosomes. Moreover, although some cytosolic staining was detected, Brd4 was excluded from the chromosomes in cells expressing the alpha PV E2 proteins, similar to the empty pMEP vector control (Fig. 4A). While each of these proteins could be observed in close association with mitotic chromosomes at the beginning (prophase) and end (telophase) of mitosis, they could not be detected on metaphase or anaphase chromosomes when cells were directly fixed in paraformaldehyde. Brd4 was also occasionally visible on cells in the early stages of prophase and late

stages of telophase, presumably because they are transitioning from and to interphase, where Brd4 levels are high (34). However, colocalization of these E2 proteins with Brd4 could not be detected on mitotic chromosomes even at these stages of mitosis (Fig. 5). Stabilization of binding of alpha papillomavirus E2 proteins to mitotic chromosomes does not require Brd4 binding. While alpha PV E2s cannot be detected on condensed chromosomes following direct fixation, we have previously shown that if mitotic cells expressing these E2 proteins are first treated with a mitotic stabilization buffer containing 0.1% Triton X-100 prior to fixation, then the alpha E2s can be observed to be associated with mitotic chromosomes in the form of large speckles, reminiscent of the HPV8 E2 chromosome staining pattern (37). This prepermeabilization technique enables detection of the alpha PV E2s on the pericentromeric region of mitotic chromosomes (37, 39).

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

FIG. 5. The alpha papillomavirus E2 proteins do not colocalize with Brd4 at early (prophase) and late (telophase) stages of mitosis. The E2 and Brd4 proteins were detected in CV-1 cells expressing HPV16 E2, as described for Fig. 4.

To investigate the interaction of the mitotic chromosomebound alpha PV E2 proteins with Brd4, mitotic cells were prepermeabilized before fixation. Representative images of BPV1, HPV8, HPV11, and HPV31 E2 proteins are shown in Fig. 6. As previously reported, this prefixation treatment did not alter the pattern of mitotic chromosomal localization of BPV1 E2 (37). Moreover, Brd4 was still clearly associated with chromosomes and colocalized with E2, as observed in directly fixed cells (Fig. 4A and 6), while there was no significant chromosomal Brd4 staining in the absence of E2. Thus, BPV1 E2 stabilizes the association of Brd4 with mitotic, as well as interphase, chromatin. The mitotic staining pattern of HPV8 E2 was also unchanged by the prefixation step (compare Fig. 4B and 6), as

9539

shown previously by Oliveira et al. (37). However, the mitotic association of Brd4 was disrupted and chromosome-bound Brd4 was no longer readily detected on mitotic chromosomes, even in the presence of HPV8 E2. As described above, in directly fixed cells, a small amount of the total chromosomebound HPV8 E2 protein colocalized with Brd4, while the majority of HPV8 E2 associated with chromosomes without detectable Brd4. However, this minor population of E2associated Brd4 was not observed on chromosomes after permeabilization of cells. Therefore, the minor population of HPV8 E2 and Brd4 is not tightly bound to mitotic chromosomes and the population of HPV8 E2 protein that strongly associates with chromosomes does so independently from Brd4. Given the high efficiency of the HPV8 E2-Brd4 interaction in vitro (Fig. 1) and the ability of HPV8 E2 to stabilize the chromatin association of Brd4 in interphase (Fig. 3), it is tempting to speculate that the minor population of colocalizing HPV8 E2-Brd4 chromosomal speckles observed following direct fixation may be carried through to mitosis as the result of the tight interaction of these proteins in interphase. However, this complex is not tightly associated with mitotic chromosomes as it is easily removed during the preextraction step. Therefore, a high-efficiency E2-Brd4 association is not sufficient for mitotic chromosome association and other factors must be involved in the stable chromosomal E2 complex of other papillomaviruses, such as BPV1. The alpha E2 proteins (HPV11 and HPV31 E2 proteins are shown as examples) were clearly detected on mitotic chromosomes following prepermeabilization treatment (Fig. 6). The E2 mitotic staining pattern of these proteins was similar to that of HPV8 E2 and very distinct from that of BPV1 E2. Moreover, chromosome-bound Brd4 was not visible on condensed chromosomes on which alpha E2s were clearly localized. These results indicate that alpha papillomaviruses can associate with mitotic chromosomes but that they do not require Brd4 as their binding partner. To confirm this observation, we analyzed the ability of the HPV31 E2 R37A I73A mutated protein, which does not interact with Brd4, to bind to mitotic chromosomes. As clearly shown in Fig. 6, this E2 protein retains its ability to associate with mitotic chromosomes despite being unable to interact with Brd4. Moreover, chromosome-bound Brd4 was not detected. The equivalent mutation in BPV1 E2 abrogates the mitotic association of this protein (7). Therefore, the alpha papillomavirus E2 proteins use a cellular binding target other than Brd4 to mediate their association with mitotic chromosomes. DISCUSSION To successfully establish a persistent infection that continually synthesizes and releases mature infectious virions, papillomaviruses, like other persistent viruses, face many challenges. The viral genome must be stably established in mitotically active host cells. Papillomaviruses infect the dividing basal cells of stratified epithelia, and low-level expression of the viral proteins, E1 and E2, results in limited replication of the viral genome. The viral genome must be maintained as a stable episome in order for the virus to complete a productive life cycle. Carcinogenic HPVs can maintain their genomes at low copy number in mitotically active

9540

MCPHILLIPS ET AL.

J. VIROL.

FIG. 6. Stabilization of binding of alpha papillomavirus E2 proteins to mitotic chromosomes does not require Brd4 binding. Cells expressing the E2 proteins shown were treated with mitotic stabilization buffer containing 0.1% Triton X-100 before fixation. E2 proteins, as detected by an M2 FLAG antibody, are labeled with an FITC-labeled antibody (green). Brd4 protein was detected by a Texas Red antibody (red). Colocalization of these proteins appears as yellow in the merged images. Cellular DNA was detected by DAPI staining (blue).

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

cells for decades, and persistence of infection highly correlates with risk of cervical cancer (reviewed in reference 44). Perpetuation of the extrachromosomal genome is achieved through the actions of the viral E2 protein. For BPV1, the E2 protein associates with binding sites within the viral genome and acts as a tether, linking the viral DNA to cellular mitotic chromosomes through interactions with cellular binding partners, one of which has been identified as Brd4 (61). We recently reported that many other papillomavirus E2 proteins are also capable of localizing to mitotic chromosomes, although with differences in the chromosomal target (37). In this study, we have analyzed this association further and have shown that some E2 proteins, including those from HPV8 and the alpha PVs, can efficiently associate with mitotic chromosomes in the absence of Brd4. As previously reported, treatment of mitotic cells with a microtubule-stabilizing buffer was necessary to allow visualization of the alpha PV E2s on condensed chromosomes (37). The resulting pattern is strikingly similar to that of HPV8 E2. However, although this treatment stabilizes and enhances staining of the mitotic spindle, we do not observe colocalization of the alpha PV E2s with the spindle apparatus, as has previously been reported for HPV11 E2 (58). Using HPV31 E2, we show that the ability of this protein to associate with chromosomes is unaffected by mutations that abrogate the interaction of E2 with Brd4, further confirming that all papillomavirus E2 proteins do not require Brd4 to mediate their association with mitotic chromosomes. BPV1 E2 is tightly bound to chromatin and stabilizes the chromatin association of Brd4 (14, 34). In this study, we show that most E2 proteins are also tightly associated with the nucleus in interphase cells and confer stability to the interaction of Brd4 with chromatin. Members of the alpha papillomaviruses are an exception; these proteins are easily eluted from interphase nuclei and do not stabilize the association of Brd4 with chromatin. The difference in chromatin binding observed with the alpha PV E2s compared to that observed with other E2s could be due to a missing cellular or viral factor. This factor could be the cellular environment of a specific cell type or a cellular or viral protein that is required to stabilize the alpha PV E2 complex on chromatin. Notably, the alpha papillomavirus genomes can be maintained stably only in keratinocytes grown in specialized media with fibroblast feeders. Unfortunately, the alpha E2 proteins are not well expressed or tolerated under these conditions. HPV11 E2 proteins have been shown previously to associate with the detergent-insoluble nuclear matrix (53, 63). However, this could be because the procedure used in these studies utilized less stringent extraction buffers of lower salt concentration than those used here. Notably, the alpha PV E2 proteins interact with Brd4 in vitro with lower efficiencies than all other E2 proteins, suggesting that there may be a correlation between in vitro Brd4 binding and stabilization of Brd4 in interphase nuclei. While Brd4 is not necessary for the mitotic chromosome association of all E2s, this is not due to an inability to bind Brd4 since we show here that all E2 proteins can bind Brd4, albeit with variations in efficiency. The alpha papillomavirus E2 proteins, from HPV11, -16, -31, and -57, cluster together with the lowest binding affinities. Moreover, as discussed above, these E2 proteins also associate with chromosomes in the absence of Brd4. It is tempting to speculate that the pattern

9541

of mitotic association correlates with Brd4 binding efficiency; however, this is not the case for HPV8 E2. HPV8 E2 has one of the highest affinities for Brd4 in vitro, but the majority of HPV8 E2 localizes to mitotic chromosomes without detectable Brd4. Nevertheless, despite differences in binding efficiency, we have shown that the same E2 residues in the transactivation domain are important for Brd4 interaction in several E2 proteins. Alanine substitution of residues 37 and 73 in BPV1 E2, HPV16 E2, and HPV31 E2 proteins abrogates the binding of each protein to Brd4. Collectively, these results indicate that the interaction of E2 and Brd4 is conserved among papillomaviruses and likely plays a key role in some E2 function in the viral life cycle. To this end, we demonstrate that Brd4 is required for the transactivation activity of all E2 proteins tested. We had previously shown that BPV1 E2-mediated stabilization of Brd4 binding to chromatin required a transcriptionally competent transactivation domain and postulated that this interaction was important for viral transcriptional regulation (34). Furthermore, two recent studies also showed that Brd4 is important for the transcription activation function of BPV1 E2 and HPV 16 E2 (45). Both studies made use of a C-terminal domain of Brd4 which acts as a dominant-negative inhibitor of the E2-Brd4 interaction (26, 45, 61). The region of E2 interaction on Brd4 has been mapped to the extreme C terminus of the protein and is the only function ascribed to this region to date (61). Given the large size of full-length Brd4, the negative effects of Brd4 overexpression on progression of the cell cycle, and the essential nature of Brd4, the use of this short inhibitory domain is more suitable (15, 24, 32). Brd4 is a positive regulatory component of the P-TEFb complex and therefore likely plays a role in regulating the general RNA polymerase IIdependent transcriptional machinery (27, 59). Thus, changes in levels of full-length Brd4 are predicted to have global effects on transcription whereas the Brd4 CTD is specific for the E2-Brd4 interaction. Comprehensive experiments reported by Ilves et al., Schweiger et al., and You et al. have established that Brd4 CTD expression does not affect growth or cell cycle progression in a range of cell types (26, 45, 61, 62). In this study, we show that the transactivation activity of all E2s is down-regulated in the presence of the Brd4 CTD, and this inhibitory effect is specific to E2-mediated transactivation. However, there was no obvious correlation between the degree of CTDmediated transcriptional down-regulation and E2-Brd4 binding efficiency. Point mutations in the E2 protein that disrupt the interaction of E2 and Brd4 had greatly reduced transcriptional activation function, but a residual level of transactivation was resistant to the effects of the Brd4 CTD. This observation builds on the previous results of Ilves et al. and Schweiger et al. and confirms that Brd4 is indeed important for the transcriptional activation function of all E2s (26, 45). If Brd4 is required for the transactivation function of all E2s, then why are there variations in E2-Brd4 binding efficiency and differences in the choice of cellular binding partners to mediate genome segregation and episomal maintenance? This diversity could be a reflection of variations in the endogenous expression of Brd4 in the natural host cells of the virus which in turn influence the role of Brd4 in the papillomavirus life cycle. Papillomaviruses have a strict tropism for epithelial cells that extends beyond merely host and tissue specificity to an affinity

9542

MCPHILLIPS ET AL.

for specific anatomical sites. The PV type-specific cellular tropism is most likely related to differences in transcriptional requirements of different PV types as well as alternate mechanisms of transmission. We have previously shown that the mitotic localization pattern of Brd4 varies between cell types; for example, Brd4 is found in a diffuse pattern on mitotic chromosomes in P19, C127, and NIH 3T3 cells but is either undetectable on mitotic chromosomes or cytosolic in CV-1 cells and in several human keratinocyte-derived cell lines (34). However, irrespective of this, BPV1 E2 stabilizes the chromatin association of Brd4 in interphase and in mitotic cells in all lines tested thus far (34). The association of Brd4 with chromatin is a somewhat labile interaction; Brd4 is highly mobile within the nucleus and associates transiently with chromatin, binding only acetylated histones (14, 40). Therefore, in some cell types, Brd4 may be an unreliable binding target, and so, different papillomaviruses have chosen alternate, cell-type-specific binding partners to mediate chromosome attachment of viral genomes. Another potentially significant difference among the papillomaviruses is the importance of the role of E2 transactivation in the virus life cycle. The BPV1 E2 protein was the first to be characterized as an enhancer-binding protein capable of transiently transactivating E2-responsive synthetic promoters (3, 21, 50). Several promoters in the viral genome are also transactivated by BPV1 E2 in a binding site-dependent manner (20, 23, 49, 54). The E2 proteins of the human papillomaviruses also function as transcriptional transactivators of heterologous promoters (12, 29, 35, 37, 41, 43, 51). Yet, in the context of the homologous E6 promoter, most HPV E2s act as transcriptional repressors (9, 16, 55–57). Studies carried out in the context of the HPV31 genome showed that the transactivation function of E2 is not required for the expression of viral proteins, the maintenance of episomal genomes, or differentiation-dependent genome amplification, suggesting that, at least for oncogenic HPVs, E2 transactivation is not an essential function (52). Mutational analysis of BPV1 E2 has not cleanly separated the genome maintenance and transactivation functions of E2, suggesting that the same cellular interactions, including association with Brd4, are important for both functions (1, 2, 7). The results presented here also confirm that for a number of papillomaviruses, the interaction of E2 and Brd4 is important not only for E2-mediated transactivation but also for the maintenance of extrachromosomal viral genomes. The E2-Brd4 interaction is therefore a promising potential target for antiviral therapies designed to cure infected cells of viral genomes. However, as we have shown, there are a number of papillomaviruses, including the clinically relevant alpha HPVs, which do not use Brd4 for genome maintenance and segregation, and so, antiviral therapeutics directed at the E2-Brd4 interaction would not disrupt mitotic chromosome association. Nevertheless, since Brd4 is important for E2-mediated transactivation of all PV E2s, loss of this function may be sufficient to eventually clear the virus. The analysis of antiviral compounds should clearly be carried out both in the BPV1 model and in additional HPV models. However, an HPV partitioning system to test this has yet to be developed. In summary, these data show that while the E2-Brd4 interaction is important for E2-mediated transactivation of all E2s,

J. VIROL.

Brd4 is not an essential chromosome binding partner of all papillomaviruses. ACKNOWLEDGMENTS We are grateful to Karl Munger for critical reading of the manuscript. This research was supported by the Intramural Research program of the NIAID, NIH. REFERENCES 1. Abroi, A., I. Ilves, S. Kivi, and M. Ustav. 2004. Analysis of chromatin attachment and partitioning functions of bovine papillomavirus type 1 E2 protein. J. Virol. 78:2100–2113. 2. Abroi, A., R. Kurg, and M. Ustav. 1996. Transcriptional and replicational activation functions in the BPV1 E2 protein are encoded by different structural determinants. J. Virol. 70:6169–6179. 3. Androphy, E. J., D. R. Lowy, and J. T. Schiller. 1987. Bovine papillomavirus E2 trans-activating gene product binds to specific sites in papillomavirus DNA. Nature 325:70–73. 4. Ballestas, M. E., P. A. Chatis, and K. M. Kaye. 1999. Efficient persistence of extrachromosomal KSHV DNA mediated by latency-associated nuclear antigen. Science 284:641–644. 5. Bastien, N., and A. A. McBride. 2000. Interaction of the papillomavirus E2 with mitotic chromosomes. Virology 270:124–134. 6. Baxter, M. K., and A. A. McBride. 2005. An acidic amphipathic helix in the bovine papillomavirus E2 protein is critical for DNA replication and interaction with the E1 protein. Virology 332:78–88. 7. Baxter, M. K., M. G. McPhillips, K. Ozato, and A. A. McBride. 2005. The mitotic chromosome binding activity of the papillomavirus E2 protein correlates with interaction with the cellular chromosomal protein, Brd4. J. Virol. 79:4806–4818. 8. Berezney, R., and D. S. Coffey. 1974. Identification of a nuclear protein matrix. Biochem. Biophys. Res. Commun. 60:1410–1417. 9. Bernard, B. A., C. Bailly, M. C. Lenoir, M. Darmon, F. Thierry, and M. Yaniv. 1989. The human papillomavirus type 18 (HPV18) E2 gene product is a repressor of the HPV18 regulatory region in human keratinocytes. J. Virol. 63:4317–4324. 10. Bernard, H. U. 2005. The clinical importance of the nomenclature, evolution and taxonomy of human papillomaviruses. J. Clin. Virol. 32(Suppl 1):S1–S6. 11. Brannon, A. R., J. A. Maresca, J. D. Boeke, M. A. Basrai, and A. A. McBride. 2005. Reconstitution of papillomavirus E2-mediated plasmid maintenance in Saccharomyces cerevisiae by the Brd4 bromodomain protein. Proc. Natl. Acad. Sci. USA 102:2998–3003. 12. Cripe, T. P., T. H. Haugen, J. P. Turk, F. Tabatabai, P. G. Schmid III, M. Du ¨rst, L. Gissmann, A. Roman, and L. P. Turek. 1987. Transcriptional regulation of the human papillomavirus-16 E6-E7 promoter by a keratinocyte-dependent enhancer, and by viral E2 trans-activator and repressor gene products: implications for cervical carcinogenesis. EMBO J. 6:3745–3753. 13. de Villiers, E. M., C. Fauquet, T. R. Broker, H. U. Bernard, and H. zur Hausen. 2004. Classification of papillomaviruses. Virology 324:17–27. 14. Dey, A., F. Chitsaz, A. Abbasi, T. Misteli, and K. Ozato. 2003. The double bromodomain protein Brd4 binds to acetylated chromatin during interphase and mitosis. Proc. Natl. Acad. Sci. USA 100:8758–8763. 15. Dey, A., J. Ellenberg, A. Farina, A. E. Coleman, T. Maruyama, S. Sciortino, J. Lippincott-Schwartz, and K. Ozato. 2000. A bromodomain protein, MCAP, associates with mitotic chromosomes and affects G2-to-M transition. Mol. Cell. Biol. 20:6537–6549. 16. Dong, G., T. R. Broker, and L. T. Chow. 1994. Human papillomavirus type 11 E2 proteins repress the homologous E6 promoter by interfering with the binding of host transcription factors to adjacent elements. J. Virol. 68:1115– 1127. 17. Florence, B., and D. V. Faller. 2001. You bet-cha: a novel family of transcriptional regulators. Front. Biosci. 6:D1008–D1018. 18. Grogan, E. A., W. P. Summers, S. Dowling, D. Shedd, L. Gradoville, and G. Miller. 1983. Two Epstein-Barr viral nuclear neoantigens distinguished by gene transfer, serology, and chromosome binding. Proc. Natl. Acad. Sci. USA 80:7650–7653. 19. Harris, A., B. D. Young, and B. E. Griffin. 1985. Random association of Epstein-Barr virus genomes with host cell metaphase chromosomes in Burkitt’s lymphoma-derived cell lines. J. Virol. 56:328–332. 20. Haugen, T. H., T. P. Cripe, G. D. Ginder, M. Karin, and L. P. Turek. 1987. Trans-activation of an upstream early gene promoter of bovine papilloma virus-1 by a product of the viral E2 gene. EMBO J. 6:145–152. 21. Hawley-Nelson, P., E. J. Androphy, D. R. Lowy, and J. T. Schiller. 1988. The specific DNA recognition sequence of the bovine papillomavirus E2 protein is an E2-dependent enhancer. EMBO J. 7:525–531. 22. Heino, P., J. Zhou, and P. F. Lambert. 2000. Interaction of the papillomavirus transcription/replication factor, E2, and the viral capsid protein, L2. Virology 276:304–314.

VOL. 80, 2006

ROLE OF Brd4 IN DIFFERENT PAPILLOMAVIRUS LIFE CYCLES

23. Hermonat, P. L., B. A. Spalholz, and P. M. Howley. 1988. The bovine papillomavirus P2443 promoter is E2 trans-responsive: evidence for E2 autoregulation. EMBO J. 7:2815–2822. 24. Houzelstein, D., S. L. Bullock, D. E. Lynch, E. F. Grigorieva, V. A. Wilson, and R. S. Beddington. 2002. Growth and early postimplantation defects in mice deficient for the bromodomain-containing protein Brd4. Mol. Cell. Biol. 22:3794–3802. 25. Ilves, I., S. Kivi, and M. Ustav. 1999. Long-term episomal maintenance of bovine papillomavirus type 1 plasmids is determined by attachment to host chromosomes, which is mediated by the viral E2 protein and its binding sites. J. Virol. 73:4404–4412. 26. Ilves, I., K. Maemets, T. Silla, K. Janikson, and M. Ustav. 2006. Brd4 is involved in multiple processes of the bovine papillomavirus type 1 life cycle. J. Virol. 80:3660–3665. 27. Jang, M. K., K. Mochizuki, M. Zhou, H. S. Jeong, J. N. Brady, and K. Ozato. 2005. The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19:523–534. 28. Khurana, B., and T. M. Kristie. 2004. A protein sequestering system reveals control of cellular programs by the transcriptional coactivator HCF-1. J. Biol. Chem. 279:33673–33683. 29. Kovelman, R., G. K. Bilter, E. Glezer, A. Y. Tsou, and M. S. Barbosa. 1996. Enhanced transcriptional activation by E2 proteins from the oncogenic human papillomaviruses. J. Virol. 70:7549–7560. 30. Law, M. F., D. R. Lowy, I. Dvoretzky, and P. M. Howley. 1981. Mouse cells transformed by bovine papillomavirus contain only extrachromosomal viral DNA sequences. Proc. Natl. Acad. Sci. USA 78:2727–2731. 31. Lehman, C. W., and M. R. Botchan. 1998. Segregation of viral plasmids depends on tethering to chromosomes and is regulated by phosphorylation. Proc. Natl. Acad. Sci. USA 95:4338–4343. 32. Maruyama, T., A. Farina, A. Dey, J. Cheong, V. P. Bermudez, T. Tamura, S. Sciortino, J. Shuman, J. Hurwitz, and K. Ozato. 2002. A mammalian bromodomain protein, Brd4, interacts with replication factor C and inhibits progression to S phase. Mol. Cell. Biol. 22:6509–6520. 33. McBride, A. A., and G. Myers. 1997. The E2 proteins: an update. In G. Myers, C. Baker, K. Munger, F. Sverdrup, A. McBride, and H.-U. Bernard (ed.), Human papillomaviruses 1997. Los Alamos National Laboratory, Los Alamos, N.Mex. 34. McPhillips, M. G., K. Ozato, and A. A. McBride. 2005. Interaction of bovine papillomavirus E2 protein with Brd4 stabilizes its association with chromatin. J. Virol. 79:8920–8932. 35. Muller, A., A. Ritzkowsky, and G. Steger. 2002. Cooperative activation of human papillomavirus type 8 gene expression by the E2 protein and the cellular coactivator p300. J. Virol. 76:11042–11053. 36. Narayanan, A., M. L. Nogueira, W. T. Ruyechan, and T. M. Kristie. 2005. Combinatorial transcription of herpes simplex virus and varicella zoster virus immediate early genes is strictly determined by the cellular coactivator HCF-1. J. Biol. Chem. 280:1369–1375. 37. Oliveira, J. G., L. A. Colf, and A. A. McBride. 2006. Variations in the association of papillomavirus E2 proteins with mitotic chromosomes. Proc. Natl. Acad. Sci. USA 103:1047–1052. 38. Penrose, K. J., and A. A. McBride. 2000. Proteasome-mediated degradation of the papillomavirus E2-TA protein is regulated by phosphorylation and can modulate viral genome copy number. J. Virol. 74:6031–6038. 39. Pepper, D. A., and B. R. Brinkley. 1979. Microtubule initiation at kinetochores and centrosomes in lysed mitotic cells. Inhibition of site-specific nucleation by tubulin antibody. J. Cell Biol. 82:585–591. 40. Phair, R. D., P. Scaffidi, C. Elbi, J. Vecerova, A. Dey, K. Ozato, D. T. Brown, G. Hager, M. Bustin, and T. Misteli. 2004. Global nature of dynamic proteinchromatin interactions in vivo: three-dimensional genome scanning and dynamic interaction networks of chromatin proteins. Mol. Cell. Biol. 24:6393– 6402. 41. Phelps, W. C., and P. M. Howley. 1987. Transcriptional trans-activation by the human papillomavirus type 16 E2 gene product. J. Virol. 61:1630–1638. 42. Piirsoo, M., E. Ustav, T. Mandel, A. Stenlund, and M. Ustav. 1996. Cis and

43.

44. 45. 46.

47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

59. 60. 61. 62. 63.

9543

trans requirements for stable episomal maintenance of the BPV-1 replicator. EMBO J. 15:1–11. Sakai, H., T. Yasugi, J. D. Benson, J. J. Dowhanick, and P. M. Howley. 1996. Targeted mutagenesis of the human papillomavirus type 16 E2 transactivation domain reveals separable transcriptional activation and DNA replication functions. J. Virol. 70:1602–1611. Schiffman, M., and P. E. Castle. 2003. Human papillomavirus: epidemiology and public health. Arch. Pathol. Lab. Med. 127:930–934. Schweiger, M. R., J. You, and P. M. Howley. 2006. Bromodomain protein 4 mediates the papillomavirus e2 transcriptional activation function. J. Virol. 80:4276–4285. Sears, J., J. Kolman, G. M. Wahl, and A. Aiyar. 2003. Metaphase chromosome tethering is necessary for the DNA synthesis and maintenance of oriP plasmids but is insufficient for transcription activation by Epstein-Barr nuclear antigen 1. J. Virol. 77:11767–11780. Skiadopoulos, M. H., and A. A. McBride. 1998. Bovine papillomavirus type 1 genomes and the E2 transactivator protein are closely associated with mitotic chromatin. J. Virol. 72:2079–2088. Soeda, E., M. C. Ferran, C. C. Baker, and A. A. McBride. 2006. Repression of HPV16 early region transcription by the E2 protein. Virology 351:29–41. Spalholz, B. A., P. F. Lambert, C. L. Yee, and P. M. Howley. 1987. Bovine papillomavirus transcriptional regulation: localization of the E2-responsive elements of the long control region. J. Virol. 61:2128–2137. Spalholz, B. A., Y. C. Yang, and P. M. Howley. 1985. Transactivation of a bovine papilloma virus transcriptional regulatory element by the E2 gene product. Cell 42:183–191. Steger, G., and S. Corbach. 1997. Dose-dependent regulation of the early promoter of human papillomavirus type 18 by the viral E2 protein. J. Virol. 71:50–58. Stubenrauch, F., A. M. Colbert, and L. A. Laimins. 1998. Transactivation by the E2 protein of oncogenic human papillomavirus type 31 is not essential for early and late viral functions. J. Virol. 72:8115–8123. Swindle, C. S., N. Zou, B. A. Van Tine, G. M. Shaw, J. A. Engler, and L. T. Chow. 1999. Human papillomavirus DNA replication compartments in a transient DNA replication system. J. Virol. 73:1001–1009. Szymanski, P., and A. Stenlund. 1991. Regulation of early gene expression from the bovine papillomavirus genome in transiently transfected C127 cells. J. Virol. 65:5710–5720. Tan, S. H., B. Gloss, and H. U. Bernard. 1992. During negative regulation of the human papillomavirus-16 E6 promoter, the viral E2 protein can displace Sp1 from a proximal promoter element. Nucleic Acids Res. 20:251–256. Thierry, F., and P. M. Howley. 1991. Functional analysis of E2-mediated repression of the HPV18 P105 promoter. New Biol. 3:90–100. Thierry, F., and M. Yaniv. 1987. The BPV1-E2 trans-acting protein can be either an activator or a repressor of the HPV18 regulatory region. EMBO J. 6:3391–3397. Van Tine, B. A., L. D. Dao, S. Y. Wu, T. M. Sonbuchner, B. Y. Lin, N. Zou, C. M. Chiang, T. R. Broker, and L. T. Chow. 2004. Human papillomavirus (HPV) origin-binding protein associates with mitotic spindles to enable viral DNA partitioning. Proc. Natl. Acad. Sci. USA 101:4030–4035. Yang, Z., J. H. Yik, R. Chen, N. He, M. K. Jang, K. Ozato, and Q. Zhou. 2005. Recruitment of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol. Cell 19:535–545. Yee, C., I. Krishnan-Hewlett, C. C. Baker, R. Schlegel, and P. M. Howley. 1985. Presence and expression of human papillomavirus sequences in human cervical carcinoma cell lines. Am. J. Pathol. 119:361–366. You, J., J. L. Croyle, A. Nishimura, K. Ozato, and P. M. Howley. 2004. Interaction of the bovine papillomavirus E2 protein with Brd4 tethers the viral DNA to host mitotic chromosomes. Cell 117:349–360. You, J., M. R. Schweiger, and P. M. Howley. 2005. Inhibition of E2 binding to Brd4 enhances viral genome loss and phenotypic reversion of bovine papillomavirus-transformed cells. J. Virol. 79:14956–14961. Zou, N., B. Y. Lin, F. Duan, K. Y. Lee, G. Jin, R. Guan, G. Yao, E. J. Lefkowitz, T. R. Broker, and L. T. Chow. 2000. The hinge of the human papillomavirus type 11 E2 protein contains major determinants for nuclear localization and nuclear matrix association. J. Virol. 74:3761–3770.