Regulation of transcription by the Epstein–Barr virus nuclear antigen EBNA 2

Richard D. Palermo, Helen M. Webb, Andrea Gunnell, Michelle J. West


The EBNA 2 (Epstein–Barr nuclear antigen 2) transcription factor is essential for B-cell transformation by the cancer-associated EBV (Epstein–Barr virus) and for the continuous proliferation of infected cells. EBNA 2 activates transcription from the viral Cp (C promoter) during infection to generate the 120 kb transcript that encodes all nuclear antigens required for immortalization by EBV. EBNA 2 contains an acidic activation domain and can interact with a number of general transcription factors and co-activators. It is now becoming clear, however, that the regulation of transcription elongation in addition to initiation by EBNA 2, at least in part through CDK9 (cyclin-dependent kinase 9)-dependent phosphorylation of the RNA polymerase C-terminal domain, is likely to play a crucial role in the mechanism of action of this key viral protein.

  • C-terminal domain phosphorylation (CTD phosphorylation)
  • cyclin-dependent kinase 9 (CDK9)
  • elongation
  • Epstein–Barr nuclear antigen 2 (EBNA 2)
  • Epstein–Barr virus (EBV)
  • transcription


Greater than 90% of the world's population carry EBV (Epstein–Barr virus) as a persistent asymptomatic infection usually established during childhood. Delayed primary infection can give rise to the benign lymphoproliferation, infectious mononucleosis. Despite its ubiquitous and generally harmless nature, EBV was first isolated from a Burkitt's lymphoma biopsy and has since been causally associated with numerous other cancers including nasopharyngeal carcinoma, Hodgkin's disease and PTLD (post-transplant lymphoproliferative disease). EBV has potent transforming activity in vitro and is able to infect resting B-cells and drive their uncontrolled growth. The virus establishes a latent infection in these immortalized B-cells and expresses only 11 viral genes, including six EBNAs (Epstein–Barr nuclear antigens; EBNAs 1, 2, 3A, 3B, 3C and LP) and three LMPs (latent membrane proteins; LMPs 1, 2A and 2B). All of the EBNAs are able to function as regulators of transcription and EBNAs 1, 2, 3A and 3C have been shown to be essential for the infection and immortalization process [14].


EBNA 2 activates cellular and viral transcription through interactions with the cellular DNA-binding proteins RBP-Jκ (recombination signal-binding protein 1 for Jκ) and PU.1 that bind consensus sites in its target promoters [59]. EBNA 2 activates the viral Cp (C promoter), from where the long (up to 120 kb) primary transcript encoding all six of the EBNAs initiates, in addition to the promoters of all three LMP genes [10,11]. EBNA 2 contains an acidic activation domain [12,13] and interacts with a number of general transcription factors and co-activators including TFIIB (transcription factor IIB), TAF40 [TBP (TATA-box-binding protein)­associated factor 40], the p62 and XPD (xeroderma pigmentosum group D) (p80) subunits of TFIIH, a 100 kDa protein that associates with TFIIE, hSNF5, p300, CBP [CREB (cAMP-response-element-binding protein)-binding protein] and PCAF (p300/CREB-binding protein-associated factor) [1418]. Consistent with the ability of EBNA 2 to recruit HATs (histone acetyltransferases) to promoters, differential histone H3 and H4 acetylation has been detected in the presence of EBNA 2 at target promoters in vivo [19]. The transcriptional function of EBNA 2 appears to be inhibited by the phosphorylation of EBNA 2 by CDK1 (cyclin-dependent kinase 1) at Ser-243 during mitosis [20], potentially as a result of reduced binding to one of its cellular DNA-binding partners, PU.1. Hyperphosphorylated EBNA 2 fails to activate the LMP 1 and Cps efficiently [20,21].

CTD (C-terminal domain) phosphorylation

Although polymerase recruitment and the generation of an ‘open’ chromatin state are crucial regulatory points in the control of transcription initiation, phosphorylation of the CTD of the largest subunit of pol II (RNA polymerase II) also plays a vital role in the regulation of transcription. The CTD in humans consists of 52 repeats of a heptapeptide sequence and becomes phosphorylated during transcription primarily on Ser-2 and Ser-5 in the sequence. Phosphorylation of the CTD is required for efficient initiation, promoter clearance, elongation and RNA processing, and can be regulated in a gene-specific and activator-specific manner (for a review, see [22]). A number of viral and cellular transcription factors have been shown to regulate CTD phosphorylation by recruiting or modulating the activity of the transcription-associated cyclin-dependent CTD kinases, CDK7, CDK8 and CDK9. These include HIV-1 Tat, adenovirus E1A, herpes simplex virus VP16 (viral protein 16), Myc, CIITA (class II transactivator), the androgen receptor and the aryl hydrocarbon receptor [2328]. CDK7 and cyclin H are subunits of the general transcription factor TFIIH; CDK8 and cyclin C are components of Mediator; CDK9 and cyclin T1 form the pTEF-b (positive transcription elongation factor-b).

EBNA 2 and CTD phosphorylation

Studies carried out in our laboratory have shown that efficient activation of transcription by EBNA 2 requires the activity of CDK9 [29]. Thus inhibition of CDK9 using a dominant-negative mutant or the CDK9 inhibitor, DRB (5,6-dichloro-1-β-D-ribofuranosylbenzimidazole), dramatically reduces EBNA 2-activated transcription from the Cps and LMP1 promoters, but has little effect on basal transcription. In addition, using a cell line expressing a conditionally active EBNA 2–oestrogen receptor fusion protein, we were able to detect EBNA 2-dependent increases in the levels of phospho-Ser-5 on the CTD. Although Ser-2 phosphorylation has been closely linked to CDK9 activity, CDK9 phosphorylates Ser-5 and Ser-2 in HIV-1 transcription elongation complexes [30] and has increased activity towards Ser-5 in the presence of HIV-1 Tat [31]. CDK9 has also been shown to have some preference for Ser-5 in peptide studies [32]. These results therefore indicate that the EBNA 2 activation mechanism also involves the potential recruitment of CDK9 to the promoter, leading to increased phosphorylation of Ser-5 and potentially Ser-2 of the CTD (Figure 1). Since CDK9-dependent CTD phosphorylation has been shown to stimulate transcriptional elongation in the HIV-1 Tat transcriptional activation mechanism [27], it is possible that EBNA 2 may activate transcriptional elongation in addition to initiation through CDK9. This is of particular significance if we consider the role played by EBNA 2 in establishing viral latency and driving the proliferation and immortalization of the infected cell. EBNA 2 is one of the first viral proteins to be expressed after infection (8–12 h) [33,34] with the mRNA encoding EBNA 2 initiating from the Wp (W promoter) under the control of cellular and B-cell-specific transcription factors [35]. Switching of promoter usage from Wp to the upstream Cp begins to occur approx. 48 h post-infection and appears to be a direct consequence of the activation of Cp by EBNA 2 via RBP-Jκ binding to consensus sites in the promoter regulatory region (Figure 2). Cp transcription results in the expression of the remaining nuclear antigens required for immortalization and is therefore a crucial step in the virus life cycle and the transformation process [36] (Figure 2). We speculate that the driving force behind this switch may be the need to promote processive transcription given that transcription from W leads to the expression of only the most promoter-proximal genes (EBNA 2 and LP) and may thus be predominantly abortive (Figure 2). EBNA 2 activation of the Cp may therefore promote the assembly of elongation-competent transcription complexes capable of efficiently transcribing full-length mRNAs. Interestingly, the Wp is located in the IR-1 (internal repeat 1) region of the EBV genome and is therefore present in multiple copies. This duplication may act to increase the number of initiating polymerases to ensure that sufficient transcripts encoding EBNA 2 are generated early in infection, since most of the transcription complexes may prematurely disengage from the template. This would, however, represent a very wasteful transcription system and a switch to a more efficient promoter would clearly be beneficial.

Figure 1 A model for the mechanism of transcriptional activation by EBNA 2

EBNA 2 binds HATs and chromatin remodellers (SWI/SNF) to disrupt chromatin structure at the promoter. It also promotes transcription complex assembly by interacting with general transcription factors, e.g. TFIIH and TFIIE. In addition, EBNA 2 promotes transcription initiation and elongation through CDK9-dependent phosphorylation of the CTD on Ser-5 and possibly also Ser-2. TFIIH will also phosphorylate Ser-5 of the CTD during initiation, but there is no evidence that EBNA 2 targets CDK7 to promoters.

Figure 2 A model for the role of EBNA 2 in driving promoter switching during infection

(1) Upon initial infection, poorly processive transcription from the Wp gives rise to some transcripts encoding EBNA LP and EBNA 2. (2) EBNA 2 activates an alternative EBNA promoter (Cp) upstream from Wp and promotes CDK9-dependent processive transcription, leading to the production of full-length mRNAs encoding the remaining EBNAs required for immortalization.

EBNA 2 as a drug target

The dependence of EBNA 2-activated transcription on CDK9 raises the possibility that CDK9-specific CDK inhibitors that have been investigated and found to be wide-spectrum anticancer drugs, such as flavopiridol, could be used as anti-EBV agents. Flavopiridol effectively inhibits CDKs at least 20-fold more selectively than other cellular kinases [37,38] and has been shown to block cell-cycle progression and promote apoptosis through a number of mechanisms, including the inhibition of cell-cycle CDKs, and the inhibition of transcription of anti-apoptotic proteins [39,40]. Significantly, flavopiridol displays most selectivity for CDK9 (IC50=3 nM) [41]. Given the requirement for CDK9 in the HIV-1 Tat transactivation mechanism [42,43], the potential use of flavopiridol as an anti-HIV agent has been suggested, and studies in mice reported a reduction in HIV-induced disease symptoms in the presence of the drug [44]. Our results identifying a requirement for CDK9 in the EBNA 2 activation mechanism pinpoint a stage in the latent infection process that could be targeted by drugs such as flavopiridol to inhibit EBV-induced B-cell proliferation and immortalization. Potential applications of CDK9-specific drugs may be in the treatment of EBNA 2-positive EBV-associated tumours, such as PTLD, or to block EBNA 2-dependent B-cell proliferation in severe cases of infectious mononucleosis where the use of traditional antivirals is ineffective.

PTLD arises as a direct result of the uncontrolled proliferation of EBV-infected B-lymphocytes, a process that is normally restricted in the presence of an effective T-cell response in healthy EBV-infected individuals. Thus organ or stem cell transplant patients undergoing immunosuppressive therapy are at risk for PTLD development, as are individuals with congenital or acquired immunodeficiencies, e.g. the Wiskott–Aldrich syndrome, ataxia telangiectasia, acute lymphoblastic leukaemia and AIDS. In a large study of solid organ transplant patients, PTLD was found to develop at a frequency of between 1.3 and 8.2%, depending on the organ type, and has been reported to occur in 30% of children undergoing small intestine transplantation [45]. Mortality rates from PTLD have been reported to be as high as 80% for bone-marrow transplant patients [45].

The tumour cells in PTLD resemble the permanently proliferating LCLs (lymphoblastoid cell lines) generated in vitro after latent infection of resting B-cells by EBV and therefore express all latent proteins including EBNA 2. EBNA 2 is required for the continued proliferation of EBV-infected cells, and can provide protection from Nur77-mediated apoptosis [4649]. Agents that specifically target EBNA 2 function would therefore be expected to block not only the initial infection and cellular transformation by EBV, but also the growth of EBV immortalized cell lines and tumour cells that proliferate in an EBNA 2-dependent manner. Indeed, peptide inhibitors that block the EBNA 2–RBP-Jκ interaction and prevent transcriptional activation by EBNA 2 have been successfully used to down-regulate EBNA 2 target gene expression, inhibit the growth of LCLs and block EBV transformation of B-cells in vitro [49]. First-line treatment for PTLD involves reducing immunosuppression, but this approach runs the risk of inducing transplant rejection. The use of anti-CD20 (B-cell) antibodies has shown reasonable success in phase II trials, but 5-year survival rates still remain unacceptably low [45]. The investigation of new EBV-specific treatment strategies could therefore still be of enormous benefit.

Future directions

Given the ability of EBNA 2 to promote CTD phosphorylation on Ser-5, it will be interesting to examine the effects of EBNA 2 on the phosphorylation of Ser-2, the main CDK9 CTD target during elongation. It would be likely that the most promoter-distal effects of EBNA 2 would be manifested through phosphorylation of this residue. Moreover, given that CDK9-dependent phosphorylation of the Spt5 subunit of the pausing and elongation regulator DSIF (DRB-sensitivity-inducing factor) plays a role in relieving promoter-proximal pausing induced by NELF (negative elongation factor) and preventing termination, it will be interesting to determine whether these factors are also involved in the EBNA 2 activation mechanism.


  • Transcription: A Biochemical Society Focused Meeting held at the University of Manchester, U.K., 26–28 March 2008 as part of the Gene Expression and Analysis Linked Focused Meetings. Organized and Edited by Stefan Roberts (Manchester, U.K.) and Robert White (Beatson Institute, Glasgow, U.K.).

Abbreviations: CREB, cAMP-response-element-binding protein; Cp, C promoter; CTD, C-terminal domain; DRB, 5,6-dichloro-1-β-D-ribofuranosylbenzimidazole; EBNA, Epstein–Barr nuclear antigen; EBV, Epstein–Barr virus; HAT, histone acetyltransferase; LCL, lymphoblastoid cell line; LMP, latent membrane protein; NELF, negative elongation factor; pol II, RNA polymerase II; pTEF-b, positive transcription elongation factor-b; PTLD, post-transplant lymphoproliferative disease; RBP-Jκ, recombination signal-binding protein 1 for Jκ; TFIIB, transcription factor IIB; Wp, W promoter


View Abstract