reviews - University of Bath

REVIEWS EPSTEIN–BARR VIRUS: 40 YEARS ON Lawrence S. Young and Alan B. Rickinson Abstract | Epstein–Barr virus (EBV) was discovered 40 years ago from examining electron micrographs of cells cultured from Burkitt’s lymphoma, a childhood tumour that is common in sub-Saharan Africa, where its unusual geographical distribution — which matches that of holoendemic malaria —indicated a viral aetiology. However, far from showing a restricted distribution, EBV — a γ-herpesvirus — was found to be widespread in all human populations and to persist in the vast majority of individuals as a lifelong, asymptomatic infection of the B-lymphocyte pool. Despite such ubiquity, the link between EBV and ‘endemic’ Burkitt’s lymphoma proved consistent and became the first of an unexpectedly wide range of associations discovered between this virus and tumours. HUMAN LEUKOCYTE ANTIGEN CLASS II MOLECULES

A subset of histocompatability antigens that are mainly expressed on cells of the immune system, including B cells, and are involved in the presentation of antigens to CD4+ T cells. LYTIC REPLICATION

The full cycle of virus infection, leading to the production of new virus progeny and, eventually, lysis of the infected cell.

Cancer Research UK Institute for Cancer Studies, University of Birmingham, Birmingham, B15 2TT, UK. Correspondence to L.S.Y. e-mail: [email protected] doi:10.1038/nrc1452

NATURE REVIEWS | C ANCER

Epstein–Barr virus (EBV)1 preferentially infects B lymphocytes through the binding of the major viral envelope glycoprotein gp350 to the CD21 receptor on the surface of B cells2, and through the binding of a second glycoprotein, gp42, to human leukocyte antigen (HLA) 3 CLASS II MOLECULES as a co-receptor . Infection of other cell types (principally epithelial cells) is much less efficient and occurs through separate, as yet poorly defined, pathways3. Cell tropism can be modified to some extent, however, by the cell type from which viral preparations are made: virions that are produced in HLA-class-II-positive B cells are relatively depleted of gp42, and therefore target the HLA-class-II co-receptor less efficiently3. Importantly, EBV has the unique ability to transform resting B cells into permanent, latently infected lymphoblastoid cell lines (LCLs), an in vitro system that has provided an invaluable, albeit incomplete, model of the lymphomagenic potential of the virus. By contrast, infection of epithelial cells in vitro does not activate the full growth-transforming programme of the virus, and rarely — if ever — achieves full LYTIC REPLICATION. B-cell transformation to LCLs therefore remains the dominant in vitro model of infection. EBV latent genes and transformation

In EBV-transformed LCLs, every cell carries multiple extrachromosomal copies of the viral episome (FIG. 1a) and constitutively expresses a limited set of viral gene products, the so-called latent proteins, which comprise

six EBV nuclear antigens (EBNAs 1, 2, 3A, 3B, 3C and -LP) and three latent membrane proteins (LMPs 1, 2A and 2B)4 (FIG. 1b). Transcripts from the BamHIA region of the viral genome (so-called BART transcripts; see later) are also detected in LCLs. In addition to the latent proteins, LCLs also show abundant expression of the small, non-polyadenylated (and therefore non-coding) RNAs, EBER1 and EBER2; the function of these transcripts is not clear, but they are consistently expressed in all forms of latent EBV infection4. This pattern of latent EBV gene expression, which appears to be activated only in B-cell infections, is referred to as ‘latency III’ (FIG. 1). LCLs show high levels of expression of the B-cell activation markers CD23, CD30, CD39 and CD70, and of the cell-adhesion molecules lymphocyte-function-associated antigen 1 (LFA1; also known as CD11a/18), LFA3 (also known as CD58) and intercellular cell-adhesion molecule 1 (ICAM1; also known as CD54)5,6. These markers are usually absent or expressed at low levels on resting B cells, but are transiently induced to high levels when these cells are activated into short-term growth by antigenic or mitogenic stimulation, indicating that EBV-induced immortalization can be elicited through the constitutive activation of the same cellular pathways that drive physiological B-cell proliferation. The ability of EBNA2, EBNA3C and LMP1 to induce LCL-like phenotypic changes when expressed individually in human B-cell lines indicates that these viral proteins

VOLUME 4 | OCTOBER 2004 | 7 5 7

REVIEWS

Summary • Epstein–Barr virus (EBV) infection is implicated in the aetiology of several different lymphoid and epithelial malignancies. • EBV-encoded latent genes induce B-cell transformation in vitro by altering cellular gene transcription and constitutively activating key cell-signalling pathways. • EBV exploits the physiology of normal B-cell differentiation to persist within the memory-B-cell pool of the immunocompetent host. • Immunosuppressed transplant patients are at risk of developing fatal EBVtransformed B-cell proliferations, presenting as ‘post-transplant lymphomas’. • Other EBV-associated tumours show more restricted forms of latent gene expression, reflecting a more complex pathogenesis that involves additional cofactors. • Pharmacological and immunotherapeutic approaches are being developed to treat or prevent EBV-associated tumours.

are key effectors of the immortalization process7. The role of EBV latent genes in the in vitro transformation of B cells has been confirmed more recently by the generation of recombinant forms of EBV that lack individual latent genes. Studies using these viruses have confirmed the absolute requirement for EBNA2 and LMP1 in the transformation process, and have highlighted a crucial role for EBNA1, EBNA-LP, EBNA3A and EBNA3C4. The EBV-encoded nuclear antigens

HODGKIN’S AND REED–STERNBERG CELLS

The malignant cells of Hodgkin’s lymphoma, named after the pathologists who first identified them as characteristic markers of this particular tumour. In EBVassociated Hodgkin’s lymphoma lesions, only these malignant cells express EBV latent-cycle antigens. GERMINAL CENTRES

Structures in peripheral lymphoid tissues that arise through clonal proliferation of antigen-stimulated B cells (germinal centroblasts) whose immunoglobulin genes undergo somatic hypermutation. A small fraction of cells expressing immunoglobulins of higher affinity for antigen are selected as memory B cells.

758

EBV-infected cells express a group of nuclear proteins that influence both viral and cellular transcription. EBNA1 is expressed in all virus-infected cells, in which its role in the maintenance and replication of the episomal EBV genome is achieved through sequence-specific binding to the plasmid origin of viral replication, OriP4 (FIG. 1b,c). EBNA1 can also interact with certain viral promoters, thereby contributing to the transcriptional regulation of the EBNAs (including EBNA1 itself ) and of LMP1. EBNA1 is separated into amino- and carboxy-terminal domains by a Gly-Ala repeat sequence, the main function of which seems to be to stabilize the mature protein — preventing its proteasomal breakdown8 — rather than functioning in its originally suggested role as an immune-evasion domain9–11. Gene-knockout studies indicate that EBNA1 does not have a crucial function in in vitro B-cell transformation beyond the maintenance of the viral genome12; on the other hand, a more direct involvement in oncogenesis is indicated by the ability of B-cell-directed EBNA1 expression to produce B-cell lymphomas in transgenic mice13, and by its possible contribution to the survival of Burkitt’s lymphoma cells in vitro14. The inability of one EBV strain — P3HR-1, which carries a deletion of the gene that encodes EBNA2 and the last two exons of that for EBNA-LP — to transform B cells in vitro was the first indication of the crucial role of EBNA2 in the transformation process4. Restoration of the EBNA2 gene in P3HR-1 has unequivocally confirmed the importance of EBNA2 in B-cell transformation and has allowed the functionally relevant domains of EBNA2 to be identified15,16. EBNA2 interacts with a

| OCTOBER 2004 | VOLUME 4

sequence-specific DNA-binding protein, Jκ-recombination-binding protein (RBP-Jκ), to transcriptionally activate cellular genes such as CD23 and the key viral genes LMP1 and LMP2A4,17,18 (FIG. 2). EBNA-LP interacts with EBNA2 and is required for the efficient outgrowth of virus-transformed B cells in vitro19,20. The transcriptional activation that is mediated by EBNA2 in conjunction with EBNA-LP is modulated by the EBNA3 family of proteins, which repress transactivation21,22 (FIG. 2). An essential role for EBNA3A and EBNA3C in B-cell transformation in vitro has been shown using EBV recombinants23. EBNA3C can cooperate with RAS in rodent-fibroblast transformation assays and disrupt cell-cycle checkpoints24,25. These effects are partly explained by the interaction of EBNA3C with factors that modulate transcription (for example, histone deacetylase 1, nonmetastatic protein 23-homologue 1 and C-terminal binding protein) or influence cell-cycle progression (for example, cyclin A)26. The EBV-encoded latent membrane proteins

LMP1 is the main transforming protein of EBV; it functions as a classic oncogene in rodent-fibroblast transformation assays and is essential for EBV-induced B-cell transformation in vitro 27,28. LMP1 has pleiotropic effects when it is expressed in cells, resulting in the induction of cell-surface adhesion molecules and activation antigens7, and upregulation of anti-apoptotic proteins (for example, BCL2 and A20)31,32. LMP1 functions as a constitutively activated member of the tumour necrosis factor receptor (TNFR) superfamily, and activates several signalling pathways in a ligand-independent manner33–35 (FIG. 3). Functionally, LMP1 resembles CD40 — another member of the TNFR superfamily — and can partially substitute for CD40 in vivo, providing both growth and differentiation signals to B cells36. The LMP1 protein activates several downstream signalling pathways that contribute to the many phenotypic consequences of LMP1 expression, including the induction of various genes that encode anti-apoptotic proteins and cytokines37 (see the legend for FIG. 3 for details). The LMP2 proteins, LMP2A and LMP2B, are not essential for EBV-induced B-cell transformation in vitro 38 (FIG. 4). However, expression of LMP2A in B cells in transgenic mice abrogates normal B-cell development, allowing immunoglobulin (Ig)-negative cells to colonize peripheral lymphoid organs39. This indicates that LMP2A can drive the proliferation and survival of B cells in the absence of signalling through the B-cell receptor (BCR). LMP2A can transform epithelial cells and enhance their adhesion and motility, effects that might be mediated by the activation of the phosphatidylinositol-3-kinase–AKT pathway 40. Repressive effects of LMP2A expression have recently been reported in human and murine B cells, and many of these target B-cell-specific factors, resulting in a phenotype that is similar to those of malignant HODGKIN’S AND REED STERNBERG (HRS) CELLS in Hodgkin’s lymphoma and 41,42 GERMINAL-CENTRE B cells . In addition to these effects, LMP2A was found to induce expression of a range of genes that are involved in cell-cycle induction, inhibition of apoptosis and suppression of cell-mediated immunity.

www.nature.com/reviews/cancer

REVIEWS a EBV electron micrograph

b EBV genome: latent genes EBER1

EBER2

LMP2A LMP2B

BamHIA region

BA RF 0

BA

Cp or Wp

1 RF TR LMP

OriP

1

Double-stranded DNA episome

EBNA1 EBNA-LP Qp

EBNA2

EBNA3C EBNA3B EBNA3A

c Open reading frames for the EBV latent proteins Nhet

OriP

a

CW W W W W W Y H TR

F QU P O M

S

e1 –e3 Z R L E

K

c bT d B

G D

XV I

W W W W W W EBNA-LP

Nhet A TR

EBNA2

EBNA3A EBNA3B EBNA3C

EBNA1

LMP1

Figure 1 | The Epstein–Barr virus genome. a | Electron micrograph of the Epstein–Barr virus (EBV) virion. b | Diagram showing the location and transcription of the EBV latent genes on the double-stranded viral DNA episome. The origin of plasmid replication (OriP) is shown in orange. The large green solid arrows represent exons encoding each of the latent proteins, and the arrows indicate the direction in which the genes encoding these proteins are transcribed. The latent proteins include the six nuclear antigens (EBNAs 1, 2, 3A, 3B and 3C, and EBNA-LP) and the three latent membrane proteins (LMPs 1, 2A and 2B). EBNA-LP is transcribed from a variable number of repetitive exons. LMP2A and LMP2B are composed of multiple exons, which are located on either side of the terminal repeat (TR) region, which is formed during the circularization of the linear DNA to produce the viral episome. The blue arrows at the top represent the highly transcribed nonpolyadenylated RNAs EBER1 and EBER2; their transcription is a consistent feature of latent EBV infection. The long outer green arrow represents EBV transcription during a form of latency known as latency III (Lat III), in which all the EBNAs are transcribed from either the Cp or Wp promoter; the different EBNAs are encoded by individual mRNAs that are generated by differential splicing of the same long primary transcript. The inner, shorter red arrow represents the EBNA1 transcript, which originates from the Qp promoter during Lat I and Lat II. Transcripts from the BamHIA region can be detected during latent infection, but no protein arising from this region has been definitively identified. The locations of the BARF0 and BARF1 coding regions are shown here. c | Location of open reading frames for the EBV latent proteins on the BamHI restriction-endonuclease map of the prototype B95.8 genome. The BamHI fragments are named according to size, with A being the largest. Lowercase letters indicate the smallest fragments. Note that the LMP2 proteins are produced from mRNAs that splice across the terminal repeats (TRs) in the circularized EBV genome. This region is referred to as Nhet, to denote the heterogeneity in this region due to the variable number of TRs in different virus isolates and in different clones of EBV-infected cells. b and c modified with permission from REF. 138 © (2003) Nature Publishing Group.

INFECTIOUS MONONUCLEOSIS

A transient illness, associated with hyperactivation of the CD8+ T-cell response, that occurs in some individuals whose primary EBV infection is delayed until the second or third decade of life. Primary infection during childhood is almost always asymptomatic.


Other EBV latent transcripts

In addition to the latent proteins, the two small nonpolyadenylated (non-coding) RNAs — EBER1 and EBER2 — are expressed in all forms of latency. However, the EBERs are not essential for the EBVinduced transformation of primary B lymphocytes4. The EBERs assemble into stable ribonucleoprotein particles with the autoantigen La and ribosomal

protein L22, and bind the interferon-inducible, doublestranded-RNA-activated protein kinase PKR43. PKR has a role in mediating the antiviral effects of the interferons, and it has been suggested that EBER-mediated inhibition of PKR function might be important for viral persistence44. Expression of the EBERs in Burkitt’s lymphoma cell lines has been found to increase tumorigenicity, promote cell survival and induce interleukin-10 (IL-10) expression43,45. Such studies indicate that EBV genes that were previously shown to be dispensable for transformation in B-cell systems might make more important contributions to the pathogenesis of some EBV-associated malignancies, and to EBV persistence, than was previously appreciated. A group of abundantly expressed RNAs that are encoded by the BamHIA region of the EBV genome were originally identified in nasopharyngeal carcinoma (NPC), but were subsequently found to be expressed in other EBV-associated malignancies, such as Burkitt’s lymphoma, Hodgkin’s lymphoma and nasal T-cell lymphoma, as well as in the peripheral blood of healthy individuals46–48. These highly spliced transcripts are commonly referred to as either BamHIA rightward transcripts (BARTs) or complementary-strand transcripts (CSTs)49,50. The protein products of these open reading frames remain to be conclusively identified. Another transcript that is generated from the BamHIA region is BARF1, which encodes a 31-kDa protein that was originally identified as an early antigen expressed on induction of the EBV lytic cycle. Recent studies have shown that BARF1 is a secreted protein that is expressed as a latent protein in EBV-associated NPC and gastric carcinoma51,52. BARF1 shares limited homology with the human colony-stimulating factor 1 receptor (the FMS oncogene) and displays oncogenic activity when it is expressed in rodent fibroblasts and simian primary epithelial cells53. EBV infections in immunocompetent hosts

In contrast to in vitro studies of EBV infection and latent-gene function, our understanding of the biology of EBV infection in vivo (FIG. 5) is still rudimentary. Primary infection (by oral transmission) is usually asymptomatic, but if it is delayed until adolescence it occasionally presents as INFECTIOUS MONONUCLEOSIS (IM). Patients with acute IM shed high titres of infectious virus in the throat from lytic infection at oropharyngeal sites. It is possible that this occurs in local mucosal B cells but, from evidence of virus replicative lesions that are seen in the oral mucosa of immunocompromised patients, it is likely that this also involves the oropharyngeal epithelium. At the same time, large numbers of latently infected B cells — at least some of which represent transformed EBNA2+, LMP1+ lymphoblasts54,55 — appear in tonsillar (and possibly other) lymphoid tissues. In vitro, both naive and MEMORY B CELLS seem equally susceptible to EBV infection. However, although some of the infected tonsillar cells in IM tonsils have a naive Ig genotype, the expanding clones preferentially involve cells with mutated Ig


REVIEWS a Repression SIN3A HDAC1/2

SMRT

SAP30 SKIP CIR Off

RBP-Jκ GTGGGAA

SIN3A EBNA2

SMRT

HDAC1/2 SAP30 CIR

b Activation EBNA2 EBNA3A EBNA3B

EBNA-LP

SKIP RBP-Jκ BTM

On

GTGGGAA EBNA3C

MEMORY B CELLS

B cells that have experienced antigen stimulation and, usually, somatic hypermutation and germinal-centre transit, before subsequent selection into a pool of long-lived recirculating cells. These cells rapidly respond to a later re-challenge with their specific antigen, mounting an efficient secondary antibody response.

Figure 2 | The EBV-encoded nuclear antigens. a | Epstein–Barr virus (EBV)-encoded nuclear antigen 2 (EBNA2) functions as a transcriptional activator by interacting with the DNA-binding Jκ-recombination-binding protein (RBP-Jκ) and relieving the transcriptional repression that is mediated by a large multiprotein complex consisting of SMAT, SIN3A, histone deacetylase 1 (HDAC1) and HDAC2 (REFS 4, 17, 18). SKIP (Ski interacting protein) is another RBP-Jκinteracting protein that also interacts with the SMRT–HDAC corepressor complex. EBNA2 abolishes RBP-Jκ mediated repression by competing for the SMRT–HDAC corepressor complex through binding to both RBP-Jκ and SKIP130. b | The acidic domain of EBNA2 then recruits the basal transcription machinery (TFIIB, TFIIH and p300; not shown) to activate transcription. EBNA-LP cooperates with EBNA2 in RBP-Jκ-mediated transcriptional activation by interacting with the acidic activation domain of EBNA2 (REF. 4). The EBNA3 family of proteins modulate EBNA2-mediated RBP-Jκ activation by interacting with RBP-Jκ and competing for binding and activation by EBNA2. The RBP-Jκ homologue in Drosophila is involved in signal transduction from the Notch receptor, a pathway that is important in cell-fate determination in Drosophila and has also been implicated in the development of T-cell tumours in humans131. EBNA2 can functionally replace the intracellular region of Notch132. BTM, basal transcription machinery; CIR, CBFI (RBP-Jκ)-interacting corepressor; SAP30, SIN3-associated protein 30.

LYMPHOCYTOSIS

A marked expansion of lymphocyte numbers in the blood, caused by proliferation of EBV-specific CD8+ T cells in the blood of patients with infectious mononucleosis. T-CELL IMMUNOCOMPROMISED

A state of immune T-cell impairment seen, for instance, in transplant patients receiving high doses of T-cell-suppressive drugs to prevent rejection of the transplant and in late-stage AIDS patients; in both situations, immune control over persistent viral infections, such as EBV, is impaired.

760

sequences that are typical of antigen-selected memory cells. This is consistent with the important finding that, in the blood of patients with acute IM and subsequently of long-term virus carriers, EBV-infected cells are concentrated in the IgD–CD27+ memory-B-cell subset56,57. Furthermore these cells have by that time returned to a resting state and have downregulated the expression of most, and possibly all, viral proteins58. The precise route of entry into memory is still a subject of much debate (FIG. 5). This reservoir of infected cells is then stably maintained, and seems to be subject to the same physiological controls as the general mucosa-associated memory-B-cell pool59. Such a strategy brings with it the possibility of fortuitous, antigen-driven recruitment of


infected cells into germinal centres, leading to progeny that either re-enter the circulating memory pool or differentiate to become plasma cells that might migrate to mucosal sites. The different forms of latency that are seen in virus-associated malignancies might represent latency programmes that have evolved to accommodate such changes in host-cell physiology. Germinal-centre transit therefore seems to activate a latency programme in which only the genome-maintenance protein EBNA1 is expressed, whereas exit from germinal centres is possibly linked to the transient activation of LMP1 and LMP2 expression58. Similarly, a commitment to plasmacytoid differentiation is thought to trigger these cells to undergo lytic viral replication, providing a source of low-level virus shedding into the oropharynx. There might also be circumstances in which infected cells in the reservoir can become reactivated to produce further proliferative latency III infections. Primary EBV infection elicits strong cellular immune responses that then bring the infection under control. The LYMPHOCYTOSIS that typifies acute IM therefore reflects the hyperexpansion of cytotoxic CD8 + T cells that are reactive to both lytic- and latent-cycle viral antigens, reactivities that are subsequently maintained in the CD8 + T-cell memory at levels that, collectively, might constitute up to 5% of the total circulating CD8+ T-cell pool60. This level of commitment to a single virus, which is apparent even in EBV carriers who have no prior history of IM, implies a crucial role for immune T-cell surveillance in controlling persistent EBV infection. Virus-associated B-cell lymphomas

There are three histologically and clinically distinct types of EBV-associated B-cell lymphoma that show different patterns of latent gene expression and seem, from Ig-gene sequencing, to derive from cells at different positions in the B-cell differentiation pathway. Here, we describe the main features of these tumours and discuss the role of EBV in their pathogenesis. Lymphomas in immunosuppressed individuals. T-CELLpatients are at high risk of developing B-cell lymphomas, and those that arise in transplant patients are the best studied of these lymphomas. Most ‘post-transplant lymphomas’ (PTLs) arise as polyclonal or monoclonal lesions within the first year of allografting, when immunosuppression is most severe. Almost all of these early-onset tumours are EBV-positive and (on the basis of positive EBNA2 and LMP1 staining) express the latency III programme, which identifies them as virus-transformed B cells that grow out in the absence of effective T-cell surveillance61. Some of the lymphomas that are seen in highly immunocompromised AIDS patients, particularly central nervous system (CNS) lesions, show essentially the same phenotype. Transplant cohorts continue to show a significant, albeit lower, risk of developing lymphomas well beyond the first year, but these late-onset tumours are a more heterogeneous group and — as with the non-CNS lymphomas of IMMUNOCOMPROMISED


REVIEWS

Lipid raft

EBV LMP1 AKT PI3K TRAFs

CTAR1

NH2

RHO GTPase p38, ERK?

MAPKKK NIK IKKα CTAR2

IKKα IKKγ IKKβ

TRADD TRAF2

MAPKKK

p100

IκB p65 p50

JNK, ERK? p65 p52

p65 p50

NF-κB2

NF-κB1

Figure 3 | Structure and function of LMP1. The Epstein–Barr virus latent membrane protein 1 (LMP1) is an integral membrane protein of 63 kDa and can be subdivided into three domains: first, an amino-terminal cytoplasmic tail (amino acids 1–23), which tethers LMP1 to the plasma membrane and orientates the protein; second, six hydrophobic transmembrane loops, which are involved in self aggregation and oligomerization (amino acids 24–186); third, a long carboxyterminal cytoplasmic region (amino acids 187–386), which possesses most of the signalling activity of the molecule. Two distinct functional domains referred to as C-terminal activation regions 1 and 2 (CTAR1 and CTAR2) have been identified on the basis of their ability to activate the nuclear factor-κB (NF-κB) transcription-factor pathway133. The signalling effects of LMP1 result from the ability of tumour necrosis factor receptor (TNFR)-associated factors (TRAFs) to interact either directly with CTAR1 or indirectly by interacting with the death-domain-containing protein TRADD, which binds to CTAR2 (REF. 37). These adaptor proteins subsequently recruit a multiprotein catalytic complex containing the NF-κB-inducing kinase (NIK) and the IκB kinases (IKKs). This results in the activation of both the classic IκBα-dependent NF-κB pathway (involving p50–p65 heterodimers) and the processing of p100 NF-κB2 to generate p52–p65 heterodimers134. Other kinases are recruited to LMP1 through interactions with TRAF molecules including the mitogen-activated protein kinase kinase kinases (MAPKKKs) TPL2 and TAK1, and these contribute to the activation of the NF-κB, MAPK and phosphatidylinositol 3-kinase (PI3K) pathways. ERK, extracellular signal-regulated kinase; JNK, c-JUN amino-terminal kinase.

late-stage AIDS patients — the proportion of EBVassociated cases can fall below 50%. EBV-positive lateonset PTLs are typically monoclonal tumours, some of which mimic the latency III phenotype of classic early-onset disease, whereas others are EBNA2- and LMP1-negative or, occasionally, EBNA2-negative but LMP1-positive for a proportion of cells 62,63. Such tumours might therefore have evolved from EBVtransformed LCL-like lesions through the acquisition of additional cellular genetic changes that render certain viral functions redundant. However, with the exception of BCL6 mutations 63 (which might be a coincidental consequence of germinal-centre transit), no specific change occurs consistently or can be linked to tumours of a particular phenotype. As described in BOX 1, Ig-gene sequencing has shown that PTLs can arise from a range of positions on the B-cell differentiation pathway, including some that indicate pathogenetic similarities with other EBVassociated B-lymphomas.


Hodgkin’s lymphoma. In this unusual tumour, the clone of malignant HRS cells is vastly outnumbered by a nonmalignant infiltrate, the appearance of which distinguishes the nodular sclerosing (NS), mixed cellularity (MC) and rarer lymphocyte-depleted (LD) subtypes. Approximately 40% of cases of classic Hodgkin’s lymphoma in the developed world are associated with EBV; this figure includes most MC and LD cases but, ironically, only a minority of the NS tumours that make up the marked peak of Hodgkin’s lymphoma incidence that is seen in the third decade of life in the developed world, and that first raised the issue of an infectious aetiology. Hodgkin’s lymphoma in the developing world seems to show an even higher overall association with EBV, reflecting at least in part the absence of this peak in young adults64. The existence of EBV-negative disease raises the question of whether EBV, when present, is simply a ‘passenger’. Although this remains formally possible, the likelihood of Hodgkin’s lymphoma having arisen in an EBV-positive target cell by chance seems remote, given that infected cells normally make up a tiny fraction (1–100 cells per million) of the total Bcell pool59. Furthermore, in EBV-positive tumours, the viral genome is present in every HRS cell and expresses a particular subset of latent-cycle proteins — EBNA1, LMP1 and LMP2, the so-called latency II programme 47. The identification of HRS cells as failed products of germinal-centre reactions by Ig genotyping65,66 (BOX 1) does indeed indicate a plausible pathogenetic role for the virus, based on rescuing such tumour progenitors from apoptosis. LMP1 is therefore capable of constitutively activating the CD40 pathway, thereby replacing a signal that is normally provided by cognate T cells during memorycell selection, and LMP2A can mimic signalling from surface Ig, replacing the usual requirement for highaffinity binding to cognate antigen. Whether EBV still contributes to the malignant phenotype at the time of tumour presentation is more difficult to determine, particularly as there are no EBV-positive HRS cell lines available as in vitro models that retain the classic latency II form of infection. Because the downstream components of surface Ig signalling have been eliminated in HRS cells67, LMP2A might no longer be operational, at least in the context of that pathway. By contrast, HRS cells continue to show many characteristics of LMP1-induced phenotypic changes, including the strong activation of nuclear factor-κB (NF-κB) and its associated downstream effects. Interestingly, the HRS cells of EBV-negative Hodgkin’s lymphoma show a very similar phenotype, which, in at least some cases, seems to have been caused by inactivation of inhibitor of κBα (IκBα), the physiological regulator of NF-κB activity, or by amplification of the REL gene, which encodes an NF-κB family member66. This parallel indicates that NF-κB deregulation is an important feature of the pathogenesis of Hodgkin’s lymphoma, and that EBVpositive and EBV-negative tumours have achieved the same end point by different routes.


REVIEWS

Lipid raft EBV LMP2A

Tyr112

LYN AKT PI3K

Tyr85

RHO GTPase

SRC/SYC

Tyr74

MAPKKK PY PY

ERK

Cell survival and motility

JUN?

Ubiquitylation of LMP2A and LMP2A-associated proteins

NEDD4

NH2

Figure 4 | Structure and function of LMP2. The structures of the Epstein–Barr virus (EBV) latent membrane proteins LMP2A and LMP2B are similar; both have 12 transmembrane domains and a 27-amino-acid cytoplasmic carboxyl terminus. In addition, LMP2A has a 119-amino-acid cytoplasmic amino-terminal domain that contains eight tyrosine residues, two of which (Tyr74 and Tyr85) form an immunoreceptor tyrosine-based activation motif (ITAM)38. The phosphorylated ITAM recruits members of the SRC family of protein tyrosine kinases and the SYK tyrosine kinase and negatively regulates their activities. A membrane-proximal tyrosine residue (Tyr112) binds the LYN tyrosine kinase and mediates the constitutive phosphorylation of the other tyrosine residues in LMP2A38. The LMP2A ITAM blocks signalling from the B-cell receptor (BCR) by sequestering these tyrosine kinases and by blocking the translocation of the BCR into lipid rafts135. LMP2A also recruits NEDD4-like ubiquitin protein ligases through phosphotyrosine (PY) motifs, and these promote the degradation of LYN and LMP2A by a ubiquitin-dependent mechanism136. LMP2A interacts with the extracellular signal-regulated kinase 1 (ERK1) mitogen-activated protein kinase (MAPK), and this results in the phosphorylation of two serine residues (Ser15 and Ser102) in LMP2A, and might contribute to LMP2A-induced activation of JUN137. MAPKKK, MAPK kinase kinase; PI3K, phosphatidylinositol 3-kinase.

SOMATIC HYPERMUTATION

Point mutations that occur in the immunoglobulin-gene variable regions (and some other genes) during B-cell differentiation. LYMPHADENOPATHY

A marked swelling of peripheral lymphoid tissues in situations arising from chronic antigenic stimulation by an infectious agent.

762

Burkitt’s lymphoma. EBV is present in all cases of ‘endemic’ Burkitt’s lymphoma — the high-incidence form of the tumour that affects children in areas of Africa and New Guinea in which malaria is holoendemic — and in up to 85% of cases in areas of intermediate incidence such as Brazil and North Africa, but in only 15% of the low-incidence ‘sporadic’ tumours that are seen in children in the developed world. Remarkably, Burkitt’s lymphoma is also common among adult human immunodeficiency virus (HIV) carriers in the developed world and often arises as the first AIDS-defining illness in relatively immunocompetent patients; some 30–40% of these tumours are EBV-associated5. All Burkitt’s lymphomas carry one of three characteristic chromosomal translocations that place the MYC oncogene under the control of the Ig heavy chain or one of the light-chain loci. The primacy of MYC deregulation as the key factor in the pathogenesis of Burkitt’s lymphoma is clear from a range of experimental systems68–70. In addition, many tumours have TP53 mutations or other defects in the p53–ARF pathway, as well as mutations in the putative tumour-suppressor gene retinoblastoma-like (REF. 71). Irrespective of their EBV status, the phenotype of Burkitt’s lymphoma cells (CD10+, CD77+, BCL6+) is remarkably similar to that of germinal centroblasts, and the detection of ongoing Ig-gene mutation in tumour cells72–74 (BOX 1) supports the suggestion that they originate in germinal centres. Indeed, the MYC translocation itself is likely to have occurred as an error of the SOMATIC HYPERMUTATION process.


At presentation, the vast majority of EBV-positive tumours show a highly restricted latency I form of infection, with viral antigen expression limited to that of EBNA1 (REFS 6,75). So, how EBV contributes to the pathogenesis of Burkitt’s lymphoma remains a matter of speculation. The virus might have an initiating role in which growth-transforming B-cell infections establish a pool of target cells that are at risk of a subsequent MYC translocation, a process that has been successfully modelled in vitro68. This latter study highlighted the apparent incompatibility of the EBV latency-III-driven and MYC-driven growth programmes in B cells76, indicating that the evolution to Burkitt’s lymphoma can only occur if the EBV programme is suppressed. Indeed, the strength of selection against full expression of viral latent genes is illustrated by a subset of Burkitt’s lymphomas in which the transcriptional features of latency III are retained, but the resident viral genome has deleted the EBNA2 gene and therefore abrogated the conventional B-cell transforming function75. Alternatively, EBV might contribute to the Burkitt’s lymphoma phenotype through the latencyI-active genes themselves. EBNA1 is an obvious candidate, but its reported oncogenicity in mouse transgene assays13 remains controversial, and its contribution to virus-induced B-cell transformation in vitro seems to be limited to maintenance of the viral genome 12,77; however, recent experiments in which EBNA1 function was blocked in EBV-positive Burkitt’s lymphoma cell lines have indicated that the protein does promote cell survival14. Other studies, based on infection or transfection of spontaneous EBV genome-loss derivatives of the Akata Burkitt’s lymphoma cell line, have also shown increased survival mediated either by virus-induced upregulation of the TCL1 oncogene78 or through induction of the expression of the IL-10 cytokine by the non-coding EBER RNAs 43. The wider relevance of these effects beyond the Akata Burkitt’s lymphoma model system remains to be determined. Any comprehensive view of Burkitt’s lymphoma pathogenesis must take into account the marked increases in tumour incidence that are associated with holoendemic malaria and HIV infections. In this regard, both agents can act as chronic stimuli of the B-cell system and, at least for HIV carriage, the persistent generalized LYMPHADENOPATHY that results from such stimulation is characterized by exaggerated germinal-centre activity79. We suggest that this greatly increases the chances of productive MYC translocations occurring. Furthermore, both agents disturb the EBV–host balance and probably increase the numbers of EBV-infected B cells that are at risk of being recruited into germinal-centre reactions80. There are differences between the two situations, however, and it will be interesting to see whether malaria has its own specific effects on the EBV–host balance as, unlike HIV, the 100-fold increase in the incidence of Burkitt’s lymphoma seen in patients with malaria consists entirely of EBV-positive disease5.


REVIEWS a Primary infection

Virus-associated T-cell and NK-cell lymphomas

Lytic replication

Latency III

Latency I/II

Latency 0 Memory B-cell reservoir

Naive B cell

Germinal centre PrimaryT-cell response

Epithelium

Memory B cell

b Persistent infection Latency 0

Latency I/II

MemoryB-cell reservoir MemoryT-cell response

Germinal centre

Lytic replication

Latency III

Naive B cell

Memory B cell

EBV is so markedly B-lymphotropic when it is exposed to human lymphocyte preparations in vitro that the association of the virus with rare but specific types of T-cell and natural killer (NK)-cell lymphomas81,82 was completely unexpected. How the virus accesses these cell lineages in vivo is still uncertain, but most evidence indicates that such infections are rare and, when they do occur, confer a high risk of lymphoma development. One of the best examples is an EBV-genome-positive T-cell lymphoma, which is seen worldwide but is most common in southeast Asian populations. This type of lymphoma arises either after acute primary infection, manifesting as virus-associated haemophagocytic syndrome (VAHS), or in the setting of a chronic active EBV infection with VAHS-like symptoms81,83. These are monoclonal tumours of CD4+ or CD8+ T-cell origin, in which viral gene expression is restricted to the production of EBNA1 and LMP2, with variable levels of LMP1 being detectable in only a fraction of cells (designated latency I/II). The tumours seem to arise rapidly from a pre-malignant pool of EBV-infected T cells that, uniquely, are present in the blood of patients who have VAHS-like disease84 and might, through cytokine release, drive macrophage activation and haemophagocytosis. A second example, which is again most common in south-east Asia, is an extranodal lymphoma that usually presents as an erosive lesion (‘lethal midline granuloma’) in the nasal cavity 82. Some of these are CD3+CD56– T-cell lymphomas, but most are CD3–CD56+ tumours of NK-cell origin, again with latency I/II gene expression85. Carcinomas associated with the virus

Plasma cell Epithelium

Figure 5 | Putative in vivo interactions between Epstein–Barr virus and host cells. a | Primary infection. Incoming virus establishes a primary focus of lytic replication in the oropharynx (possibly in the mucosal epithelium), after which the virus spreads throughout the lymphoid tissues as a latent (latency III) growth-transforming infection of B cells. Many of these proliferating cells are removed by the emerging latent-antigen-specific primary-T-cell response, but some escape by downregulating antigen expression and establishing a stable reservoir of resting viral-genomepositive memory B cells, in which viral antigen expression is mostly suppressed (latency 0). Different views of these events are shown. One view is that naive B cells are the main targets of new EBV infections in vivo. In this scenario, viral transformation drives naive cells into memory by mimicking the physiological process of antigen-driven memory-cell development in lymphoid tissues, a process involving somatic immunoglobulin-gene hypermutation during transit through a germinal centre. However, this is difficult to reconcile with the finding that EBV-infected B cells in tonsils from patients with infectious mononucleosis (IM) localize to extrafollicular areas — not to germinal centres — and show no evidence of ongoing hypermutation within expanding clones. An alternative view therefore envisages infection of pre-existing memory cells as a direct route into memory; this is consistent with the above observations on IM tonsils, but still leaves unexplained the apparent disappearance of the infected naive cell population. b | Persistent infection. The reservoir of EBVinfected memory B cells becomes subject to the physiological controls governing memory-B-cell migration and differentiation as a whole. Occasionally, these EBV-infected cells might be recruited into germinal-centre reactions, entailing the activation of different latency programmes, after which they might either re-enter the reservoir as memory cells or commit to plasma-cell differentiation — possibly moving to mucosal sites in the oropharynx and, in the process, activating the viral lytic cycle. Virions produced at these sites might initiate foci of lytic replication in permissive epithelial cells, allowing low-level shedding of infectious virus in the oropharynx, and might also initiate new growthtransforming latency III infections of naive and/or memory B cells; these new infections might possibly replenish the B-cell reservoir, but are more likely to be efficiently removed by the now wellestablished memory-T-cell response.


Nasopharyngeal carcinoma. An important consequence of epithelial infection with EBV is malignant transformation, resulting in the development of NPC, a subset of gastric adenocarcinomas and certain salivarygland carcinomas5,86. The EBV-associated, undifferentiated form of NPC — World Health Organization (WHO) type III — shows the most consistent worldwide association with EBV and is particularly common in areas of China and south-east Asia, reaching a peak incidence of around 20–30 cases per 100,000 (REF. 87). Incidence rates are high in individuals of Chinese descent, irrespective of where they live, and particularly in Cantonese males. In addition to this genetic predisposition, environmental cofactors such as dietary components (for example, salted fish) are thought to be important in the aetiology of NPC88. NPC tumours are characterized by the presence of undifferentiated carcinoma cells and a prominent lymphocytic infiltrate, and this interaction between tumour cells and lymphocytes seems to be crucial for the continued propagation of the malignant component. EBV latent-gene expression in NPC is predominantly restricted to the EBNA1 nuclear antigen, the latent membrane proteins (LMP2A and LMP2B) and the BamHIA transcripts, with ~20% of tumours also expressing the oncogenic LMP1 protein86. Southern-blot hybridization of DNA from NPC tissues demonstrates the monoclonality of


REVIEWS

Box 1 | Cellular origins of EBV-positive B-cell lymphomas

Post-transplantation lymphomas From immunoglobulin variable (IgV) gene sequencing, post-transplantation lymphomas (PTLs) seem able to arise from naive cells, memory cells or — more commonly in late onset cases — cells that have atypical or non-functional mutations that are normally inconsistent with cell survival; a minority of these latter cases even show ongoing hypermutation. Epstein–Barr virus (EBV) might create a pool of nonmalignant, but atypical, EBV-carrying B cells in vivo, either by directly rescuing cells from within germinal centres or possibly by activating somatic hypermutation in cells outside the germinal-centre environment. The latter scenario is supported by evidence from recent studies of EBV-infected non-malignant B-cell clones, both in a posttransplantation Hodgkin’s lymphoma biopsy125 and in infiltrating an EBV-negative T-cell lymphoma of the angioimmunoblastic lymphadenopathy type126.

Hodgkin’s lymphoma All Hodgkin’s and Reed-Sternberg cells within a tumour are part of the same clone and almost always carry a mutated IgV sequence that lacks intraclonal diversity. Sequences that contain nonsense mutations or deletions are seen in 25% of cases. As this almost certainly underestimates the true incidence of gene inactivation, most cases of Hodgkin’s lymphoma probably derive from crippled germinal-centre cells that have been rescued from the germinal-centre reaction. There are clear analogies here with EBV-positive PTLs that carry inactivated Ig genes or other apparently non-antigenselected mutations.

Burkitt’s lymphoma IgV sequencing of Burkitt’s lymphoma biopsy cells and derived cell lines (from both EBV-positive and EBV-negative cases) has identified multiple mutations and, in many cases, ongoing sequence diversification within the malignant clone. This confirmed Burkitt’s lymphoma as a tumour of germinal-centroblast origin. Burkitt’s lymphoma cell lines that had switched to a latency III infection and an LCL-like surface phenotype continued to show Ig-gene diversification. EBV therefore does not suppress ongoing hypermutation, even when the activation of the viral growth-transforming programme eliminates other markers of germinal centroblast origin.

the resident viral genomes, indicating that EBV infection takes place before the clonal expansion of the population of malignant cells89. Studies of normal nasopharyngeal tissue and pre-malignant biopsies indicate that genetic events occur early in the pathogenesis of NPC, and that these might predispose to subsequent EBV infection (BOX 2). Extensive serological screening has identified increased EBV-specific antibody titres in high-incidence areas; in particular, IgA antibodies to the EBV capsid antigen and early antigens have proved useful in diagnosis and in monitoring the effectiveness of therapy90. More recent studies using real-time quantitative PCR to measure circulating tumour-derived EBV DNA in the blood of patients with NPC have shown that the level of pre-treatment EBV DNA is strongly associated with overall survival, and that posttreatment EBV DNA levels predict progression-free and overall survival91. Association of EBV with the other more differentiated forms of NPC (WHO types I and II) has been shown, particularly in those geographical regions with a high incidence of undifferentiated NPC92. Carcinomas that have similar features to undifferentiated NPC have been described at other sites, including the thymus, tonsils, lungs, stomach, skin and uterine cervix, and are often referred to as ‘undifferentiated carcinomas of nasopharyngeal type’ (UCNT) or ‘lymphoepitheliomas’. There is geographical variation in the extent of the association of EBV with UCNTs.

764


Gastric carcinoma. EBV is also found in ~10% of more typical gastric adenocarcinomas, accounting for up to 75,000 new cases per year93,94. These tumours display a restricted pattern of EBV latent-gene expression (resulting expression of EBERs, EBNA1, LMP2A, BARTs and BARF1), similar to that seen in NPC95. There is significant geographical variation in the association of EBV with gastric carcinoma, which might be due to ethnic and genetic differences. EBV-positive gastric carcinomas have distinct phenotypic and clinical characteristics compared with EBV-negative tumours, including loss of expression of INK4A (also known as p16) and improved patient survival96,97. As in NPC, the precise role of EBV in the pathogenesis of gastric carcinoma remains to be determined, but the absence of EBV infection in pre-malignant gastric lesions supports the suggestion that viral infection is a relatively late event in gastric carcinogenesis98. Is EBV associated with other common epithelial malignancies? Detection of the EBERs by in situ hybridization has become the standard method of detecting EBV infection in the routine processing of tumour tissues. Although the EBERs were previously considered to be expressed in all forms of EBV latency, two studies have raised the possibility that EBER-negative forms of latency might exist in previously unrecognized EBV-associated malignancies, such as carcinomas of the breast and liver99,100. Difficulties in confirming these associations have raised concerns about the use of PCR analysis to detect EBV infection, and have also questioned the specificity of monoclonal-antibody reagents101. It is our contention that the definitive designation of a tumour as ‘EBV-associated’ should require unequivocal demonstration of the presence of the EBV genome or viral gene products within most of the tumour-cell population. Novel therapeutic approaches

Given the significant burden of EBV-associated tumours worldwide, an important priority is to design novel therapies that specifically target viral proteins or otherwise exploit the presence of the virus in malignant cells. Pharmacological approaches. One potential approach is the use of gene-therapy constructs to express either cytotoxic or inhibitory proteins selectively in tumour cells. For example, OriP-based constructs that are responsive to endogenous EBNA1 in infected cells have been used to express cytotoxic proteins (for example, FAS ligand) or wild-type p53 in in vitro and in vivo models of NPC102,103. Other approaches are based on the induction of the EBV lytic cycle, either by pharmacological agents or by delivery of EBV immediate-early genes, thereby inducing virusencoded kinases (EBV thymidine kinase and BGLF4, a protein kinase) that phosphorylate the nucleoside analogue gancyclovir to produce its active cytotoxic form104,105. Demethylating agents such as 5-azacytidine are able to de-repress lytic, as well as potentially immunogenic, latent genes106 and are now in early-stage clinical trials in patients with NPC, Hodgkin’s lymphoma and AIDS-associated lymphoma107. Another common chemotherapeutic agent, hydroxyurea, is able to induce


REVIEWS the loss of EBV episomes in in vitro models and has shown some limited clinical efficacy in patients with EBV-positive AIDS-related CNS lymphoma108,109. More focused pharmacological approaches aim to abrogate the functions of individual EBV proteins. In model systems, LMP1 effector function has been targeted directly, using single-chain antibodies or antisense RNA approaches110,111, and indirectly, by the genetic or pharmacological interception of its downstream effects

on NF-κB112. More recently, LCL growth in vitro has been impaired by blocking the transactivating function of EBNA2 using a short-peptide mimic of the RBP-Jκinteraction domain of the viral protein113 (FIG. 2). EBNA1 — the one viral protein that is expressed in all EBV-positive tumours — is a particularly attractive target, and a dominant-negative form of the protein that blocks its genome-maintenance function might have therapeutic potential114.

Box 2 | Role of Epstein–Barr virus in the pathogenesis of nasopharyngeal carcinoma In both nasopharyngeal carcinoma (NPC) and Epstein–Barr virus (EBV)-positive gastric carcinoma, the tumour cells carry monoclonal viral genomes, which indicates that EBV infection must have occurred prior to expansion of the malignant cell clone89. However, the difficulty of detecting EBV-infected epithelial cells in normal nasopharyngeal biopsies from individuals who are at high risk of developing NPC argues against a pre-existing normal reservoir of epithelial cell infection from which virus-positive carcinomas arise. Indeed, EBV infection has been detected both by in situ hybridization to the EBV-encoded RNAs (EBERs) and by the presence of monoclonal EBV genomes in high-grade pre-invasive lesions (severe dysplasia and carcinoma in situ) in the nasopharynx, but not in low-grade disease127. Similar results have been obtained from EBV-positive gastric carcinoma, in which associated normal gastric mucosa, inflamed mucosa and pre-malignant lesions are EBV-negative98. Multiple genetic changes have been found in NPC, with frequent deletion of regions on chromosomes 3p, 9p, 11q, 13q and 14q and promoter hypermethylation of specific genes on chromosomes 3p (RASSF1A and retinoic-acid receptor β2) and 9p (genes that encode INK4A, INK4B, ARF and deathassociated protein kinase)128. Deletions in both 3p and 9p have been identified in low-grade dysplastic lesions and in normal nasopharyngeal epithelium from individuals who are at high risk of developing NPC in the absence of EBV infection, indicating that genetic events occur early in the pathogenesis of NPC and that these might cause predisposition to subsequent EBV infection128. This possibility is supported by in vitro data showing that the stable infection of epithelial cells by EBV requires an altered, undifferentiated cellular environment129.

Normal epithelium

Low-grade pre-invasive lesion

High-grade pre-invasive lesion

Nasopharyngeal carcinoma

H/E

CIS? Metastasis EBER

LOH on chromosome 3p and 9p Inactivation of RASSF1A and CDKN2A EBV latent infection Telomerase dysregulation

BCL2 overexpression

LOH on 14q, 11q, 13q and 16q Inactivation of EDNRB and TSLC1? Other genetic changes (for example, in TP53 and E-cadherin)

The above scheme (see figure; images show stained epithelial sections) has been proposed, in which loss of heterozygosity (LOH) occurs early in the pathogenesis of NPC, possibly as a result of exposure to environmental cofactors such as dietary components (such as salted fish). This results in low-grade pre-invasive lesions that, after additional genetic and epigenetic events, become susceptible to EBV infection. Once cells have become infected, EBV latent genes provide growth and survival benefits, resulting in the development of NPC. Additional genetic and epigenetic changes occur after EBV infection. CIS, carcinoma in situ; EDNRB, endothelin receptor B; H/E, staining with haematoxylin and eosin; TSLC1, tumour suppressor in lung cancer 1.



REVIEWS Immunotherapy. A large body of work attempting to target EBV-positive malignancies with T cells that are specific for EBV antigens is important both in its own right and as proof of principle for tumour immunotherapy in general. The approach was first used to target EBV-positive PTLs. Bone-marrow-transplant patients were infused with EBV latent-antigen-specific effector T cells that were prepared from the bone marrow donor by autologous LCL stimulation and expansion in vitro. This strategy was highly effective, both as a therapy for the treatment of existing disease, and in prophylaxis115. Similar adoptive-transfer approaches have now been used to treat PTLs in solid-organ transplant settings using T cells that are expanded in vitro and are prepared either from the patient116 or, where necessary, from a partially HLA-matched donor117. However, these LCL-stimulated effector preparations tend to be dominated by CD8+ T cells that are specific for the immunodominant EBNA3A, EBNA3B and EBNA3C proteins — antigens that are not expressed in all EBV-associated tumours. For Hodgkin’s lymphoma and NPC, therefore, clinical trials with LCL-stimulated effectors118 represent just a first step. Strategies are now being developed either to generate T-cell preparations for transfer that are enriched in CD8+ — and possibly CD4+ — reactivities to available sub-dominant targets (such as LMP2A and EBNA1)119,120, or to immunize the patient with appropriate antigenic constructs to boost these particular responses in vivo121.

1.

Epstein, M. A., Barr, Y. M. & Achong, B. G. Virus particles in cultured lymphoblasts from Burkitt’s lymphoma. Lancet 15, 702–703 (1964). The discovery of EBV — the first description of a virus based on the use of the electron microscope. 2. Nemerow, G. R., Mold, C., Schwend, V. K., Tollefson, V. & Cooper, N. R. Identification of gp350 as the viral glycoprotein mediating attachment of Epstein–Barr virus (EBV) to the EBV/C3d receptor of B cells: sequence homology of gp350 and C3 complment fragment C3d. J. Virol. 61, 1416–1420 (1987). 3. Borza, C. M. & Hutt-Fletcher, L. M. Alternate replication in B-cells and epithelial cells switches tropism of Epstein–Barr virus. Nature Med. 8, 594–599 (2002). 4. Kieff, E. & Rickinson, A. B. in Fields Virology (eds Knipe, D. M. & Howley, P. M.) 2511–2574 (Lippincott Williams and Wilkins, Philadelphia, 2001). 5. Kieff, E. & Rickinson, A. B. in Fields Virology (eds Knipe, D. M. & Howley, P. M.) 2575–2627 (Lippincott Williams and Wilkins, Philadelphia, 2001). 6. Rowe, M. et al. Differences in B cell growth phenotype reflect novel patterns of Epstein–Barr virus latent gene expression in Burkitt’s lymphoma cells. EMBO J. 6, 2743–2751 (1987). First description of distinct forms of EBV latency in Burkitt’s lymphoma cells, as compared to EBVtransformed B-cell lines. 7. Wang, F. et al. Epstein–Barr virus latent membrane protein (LMP1) and nuclear proteins 2 and 3C are effectors of phenotypic changes in B lymphocytes: EBNA-2 and LMP1 cooperatively induce CD23. J. Virol. 64, 2309–2318 (1990). 8. Levitskaya, J. et al. Inhibition of antigen processing by the internal repeat region of the Epstein–Barr virus nuclear antigen-1. Nature 375, 685–688 (1995). Demonstration that the Gly-Ala-repeat domain of EBNA1 protects the protein from proteasomemediated degradation. 9. Voo, K. S. et al. Evidence for the presentation of major histocompatibility complex class I-restricted Epstein–Barr virus nuclear antigen 1 peptides to CD8+ T lymphocytes. J. Exp. Med. 199, 459–470 (2004). 10. Lee, S. P. et al. CD8 T cell recognition of endogenously expressed Epstein–Barr virus nuclear antigen 1. J. Exp. Med. 199, 1409–1420 (2004).

766


These tumours might also be capable of evading or suppressing T-cell immune attack — in the case of Hodgkin’s lymphoma, possibly through immunosuppressive cytokines that are produced by the HRS cells themselves122. More work is needed to determine the influence of the tumour microenvironment on T-cell attack and to explore ways of modifying the cytokine milieu, such as the use of EBV-specific T cells to deliver immunostimulatory cytokines123. Conclusions

Since its discovery in 1964, EBV has moved from being a bit-part player in the story of an obscure African tumour to its present leading role as the prime example of a human tumour virus that is aetiologically linked to an unexpectedly diverse range of malignancies. Much early work concentrated on the epidemiology of infection in human populations and on the extent of the associations of EBV with tumours. The publication of the EBV genome sequence in 1984 (REF. 124) allowed the molecular analysis of the virus, work that continues to illuminate the mechanisms of action of the viral proteins that contribute to tumorigenesis. Now, the challenge is to exploit these mechanistic insights both to gain a better understanding of the biology of EBV infection in vivo and to develop novel therapies for treating virus-associated disease.

11. Tellam, J. et al. Endogenous presentation of CD8+ T cell epitopes from Epstein–Barr virus-encoded nuclear antigen 1. J. Exp. Med. 199, 1421–1431 (2004). 12. Humme, S. et al. The EBV nuclear antigen1 (EBNA1) enhances B cell immortalisation several thousandfold. Proc. Natl Acad. Sci. USA 100, 10989–10994 (2003). 13. Wilson, J. B., Bell, J. L. & Levine, A. J. Expression of Epstein–Barr virus nuclear antigen-1 induces B cell neoplasia in transgenic mice. EMBO J. 15, 3117–3126 (1996). 14. Kennedy, G., Komano, J. & Sugden, B. Epstein–Barr virus provides a survival factor to Burkitt’s lymphomas. Proc. Natl Acad. Sci. USA 100, 14269–14274 (2003). 15. Hammerschmidt, W. & Sugden, B. Genetic analysis of immortalizing functions of Epstein–Barr virus in human B lymphocytes. Nature 340, 393–397 (1989). 16. Cohen, J. I., Wang, F., Mannick, J. & Kieff, E. Epstein–Barr virus nuclear protein 2 is a key determinant of lymphocyte transformation. Proc. Natl Acad. Sci. USA 86, 9558–9562 (1989). 17. Grossman, S. R., Johannsen, E., Tong, R., Yalamanchili, R. & Kieff, E. The Epstein–Barr virus nuclear antigen 2 transactivator is directed to response elements by the Jκ recombination signal binding protein. Proc. Natl Acad. Sci. USA 91, 7568–7572 (1994). 18. Hsieh, J. J. & Hayward, S. D. Masking of the CBF1/RBPJκ transcriptional repression domain by Epstein–Barr virus EBNA2. Science 268, 560–563 (1995). This and the previous paper were the first to show that EBNA2 influences gene transcription by affecting the notch/RBP–Jκ pathway. 19. Mannick, J. B., Cohen, J. I., Birkenbach, M., Marchini, A. & Kieff, E. The Epstein–Barr virus nuclear protein encoded by the leader of the EBNA RNAs is important in B-lymphocyte transformation. J. Virol. 65, 6826–6837 (1991). 20. Sinclair, A. J., Palmero, I., Peters, G. & Farrell, P. J. EBNA-2 and EBNA-LP cooperate to cause G0 and G1 transition during immortalization of resting human B lymphocytes by Epstein–Barr virus. EMBO J. 13, 3321–3328 (1994). 21. Robertson, E., Lin, J. & Kieff, E. The amino-terminal domains of Epstein–Barr virus nuclear proteins 3A, 3B and 3C interact with RBPJκ. J. Virol. 70, 3068–3074 (1996). 22. Zhao, B., Marshall, D. R. & Sample, C. E. A conserved domain of the Epstein–Barr virus nuclear antigens 3A and

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

3C binds to a discrete domain of Jκ. J. Virol. 70, 4228–4236 (1996). Tomkinson, B., Robertson, E. & Kieff, E. Epstein–Barr virus nuclear proteins EBNA-3A and EBNA-3C are essential for B-lymphocyte growth transformation. J. Virol. 67, 2014–2025 (1993). Parker, G. A. et al. Epstein–Barr virus nuclear antigen (EBNA)3C is an immortalizing oncoprotein with similar properties to adenovirus E1A and papillomavirus E7. Oncogene 13, 2541–2549 (1996). Parker, G. A., Touitou, R. & Allday, M. J. Epstein–Barr virus EBNA3C can disrupte multiple cell cycle checkpoints and induce nuclear division divorced from cytokinesis. Oncogene 19, 700–709 (2000). Radkov, S. A. et al. Epstein–Barr virus nuclear antigen 3C interacts with histone deacetylase to repress transcription. J. Virol. 73, 5688–5697 (1999). Wang, D., Liebowitz, D. & Kieff, E. An EBV membrane protein expressed in immortalised lymphocytes transforms established rodent cells. Cell 43, 831–840 (1985). First study to show that LMP1 functions as a classic oncogene in a rodent cell-transformation assay. Kaye, K. M., Izumi, K. M. & Kieff, E. Epstein–Barr virus latent membrane protein 1 is essential for B-lymphocyte growth transformation. Proc. Natl Acad. Sci. USA 90, 9150–9154 (1993). Henderson, S. et al. Induction of bcl-2 expression by Epstein–Barr virus latent membrane protein 1 protects infected B cells from programmed cell death. Cell 65, 1107–1115 (1991). Laherty, C. D., Hu, H. M., Opipari, A. W., Wang, F. & Dixit, V. M. The Epstein–Barr virus LMP1 gene product induces A20 zinc finger protein expression by activating nuclear factor κB. J. Biol. Chem. 267, 24157–24160 (1992). Eliopoulos, A. G., Gallagher, N. J., Blake, S. M., Dawson, C. W. & Young, L. S. Activation of the p38 mitogen-activated protein kinase pathway by Epstein–Barr virus-encoded latent membrane protein 1 coregulates interleukin-6 and interleukin-8 production. J. Biol. Chem. 274, 16085–16096 (1999). Eliopoulos, A. G. et al. Epstein–Barr virus-encoded LMP1 and CD40 mediate IL-6 production in epithelial cells via an NF-κB pathway involving TNF receptor-associated factors. Oncogene 14, 2899–2916 (1997).


REVIEWS 33. Mosialos, G. et al. The Epstein–Barr virus transforming protein LMP1 engages signalling proteins for the tumour necrosis factor receptor family. Cell 80, 389–399 (1995). Important study showing that LMP1 functions as a constitutively activated member of the TNF receptor family by interacting with common accessory proteins. 34. Gires, O. et al. Latent membrane protein 1 of Epstein–Barr virus mimics a constitutively active receptor molecule. EMBO J. 16, 6131–6140 (1997). 35. Kilger, E., Kieser, A., Baumann, M. & Hammerschmidt, W. Epstein–Barr virus-mediated B-cell proliferation is dependent upon latent membrane protein 1, which simulates an activated CD40 receptor. EMBO J. 17, 1700–1709 (1998). 36. Uchida, J. et al. Mimicry of CD40 signals by Epstein–Barr virus LMP1 in B lymphocyte responses. Science 286, 300–303 (1999). In vivo demonstration that LMP1 can partially substitute for absence of the CD40 receptor. 37. Eliopoulos, A. G. & Young, L. S. LMP1 structure and signal transduction. Semin. Cancer Biol. 11, 435–444 (2001). 38. Longnecker, R. Epstein–Barr virus latency: LMP2, a regulator or means for Epstein–Barr virus persistence? Adv. Cancer Res. 79, 175–200 (2000). 39. Caldwell, R. G., Wilson, J. B., Anderson, S. J. & Longnecker, R. Epstein–Barr virus LMP2A drives B cell development and survival in the absence of normal B cell receptor signals. Immunity 9, 405–411 (1998). 40. Scholle, F., Bendt, K. M. & Raab-Traub, N. Epstein–Barr virus LMP2A transforms epithelial cells, inhibits cell differentiation, and activates Akt. J. Virol. 74, 10681–10689 (2000). 41. Portis, T., Dyck, P. & Longnecker, R. Epstein–Barr virus (EBV) LMP2A induces alterations in gene transcription similar to those observed in Reed-Sternberg cells of Hodgkin lymphoma. Blood 102, 4166–4178 (2003). 42. Portis, T. & Longnecker, R. Epstein–Barr virus LMP2A interferes with global transcription factor regulation when expressed during B-lymphocyte development. J. Virol. 77, 105–114 (2003). 43. Takada, K. & Nanbo, A. The role of EBERs in oncogenesis. Semin. Cancer Biol. 11, 461–467 (2001). 44. Nanbo, A., Inoue, K., Adachi-Takasawa, K. & Takada, K. Epstein–Barr virus RNA confers resistance to interferon-αinduced apoptosis in Burkitt’s lymphoma. EMBO J. 21, 954–965 (2002). 45. Ruf, I. K., Rhyne, P. W., Yang, C., Cleveland, J. L. & Sample, J. T. Epstein–Barr virus small RNAs potentiate tumourigenicity of Burkitt lymphoma cells independently of an effect on apoptosis. J. Virol. 74, 10223–10228 (2000). 46. Hitt, M. M. et al. EBV gene expression in an NPC-related tumour. EMBO J. 8, 2639–2651 (1989). 47. Deacon, E. M. et al. Epstein–Barr virus and Hodgkin’s disease: transcriptional analysis of virus latency in the malignant cells. J. Exp. Med. 177, 339–349 (1993). 48. Chen, H., Smith, P., Ambinder, R. F. & Hayward, S. D. Expression of Epstein–Barr virus BamHI-A rightward transcripts in latently infected B cells from peripheral blood. Blood 93, 3026–3032 (1999). 49. Karran, L., Gao, Y., Smith, P. R. & Griffin, B. E. Expression of a family of complementary-strand transcripts in Epstein–Barr virus-infected cells. Proc. Natl Acad. Sci. USA 89, 8058–8062 (1992). 50. Smith, P. R. et al. Structure and coding content of CST (BART) family RNAs of Epstein–Barr virus. J. Virol. 74, 3082–3092 (2000). 51. Decaussin, G., Sbih-Lammali, F., de Turenne-Tessier, M., Bouguermouh, A. & Ooka, T. Expression of BARF1 gene encoded by Epstein–Barr virus in nasopharyngeal carcinoma biopsies. Cancer Res. 60, 5584–5588 (2000). 52. zur Hausen, A. et al. Unique transcription pattern of Epstein–Barr virus (EBV) in EBV-carrying gastric adenocarcinomas: expression of the transforming BARF1 gene. Cancer Res. 60, 2745–2748 (2000). 53. Sheng, W., Decaussin, G., Sumner, S. & Ooka, T. N-terminal domain of BARF1 gene encoded by Epstein–Barr virus is essential for malignant transformation of rodent fibroblasts and activation of BCL-2. Oncogene 20, 1176–1185 (2001). 54. Anagnostopoulos, I., Hummel, M., Kreschel, C. & Stein, H. Morphology, immunophenotype and distribution of latently and/or productively Epstein–Barr virus-infected cells in acute infectious mononucleosis: implications for the interindividual infection route of Epstein–Barr virus. Blood 85, 744–750 (1995). 55. Kurth, J. et al. Epstein–Barr virus-infected B cells in infectious mononucleosis: viral strategies for spreading in the B cell compartment and establishing latency. Immunity 13, 485–495 (2000).


56. Babcock, G. J., Decker, L. L., Volk, M. & Thorley-Lawson, D. A. EBV persistence in memory B cells in vivo. Immunity 9, 395–404 (1998). Important series of papers from the same group suggesting that EBV exploits the physiology of normal B-cell differentiation to persist within the memory-Bcell pool of the immunocompetent host. 57. Hochberg, D. et al. Acute infection with Epstein–Barr virus targets and overwhelms the peripheral memory B-cell compartment with resting, latently infected cells. J. Virol. 78, 5194–5204 (2004). 58. Babcock, G. J., Hochberg, D. & Thorley-Lawson, D. A. The expression pattern of Epstein–Barr virus latent genes in vivo is dependent upon the differentiation stage of the infected B cell. Immunity 13, 497–506 (2000). 59. Laichalk, L. L., Hochberg, D., Babcock, G. J., Freeman, R. B. & Thorley-Lawson, D. A. The dispersal of mucosal memory B cells: evidence from persistent EBV infection. Immunity 16, 745–754 (2002). 60. Hislop, A. D., Annels, N. E., Gudgeon, N. H., Leese, A. M. & Rickinson, A. B. Epitope-specific evolution of human CD8+ T cell responses from primary to persistent phases of Epstein–Barr virus infection. J. Exp. Med. 195, 893–905 (2002). 61. Young, L. et al. Expression of Epstein–Barr virus transformation-associated genes in tissues of patients with EBV lymphoproliferative disease. N. Engl. J. Med. 321, 1080–1085 (1989). First demonstration of EBV latent protein expression in post-transplantation lymphomas. 62. Timms, J. M. et al. Target cells of Epstein–Barr virus (EBV)positive post-transplant lymphoproliferative disease: similarities to EBV-positive Hodgkin’s lymphoma. Lancet 361, 217–223 (2003). 63. Capello, D. et al. Molecular histogenesis of posttransplantation lymphoproliferative disorders. Blood 102, 3775–3785 (2003). 64. Hsu, J. L. & Glaser, S. L. Epstein–Barr virus-associated malignancies: epidemiologic patterns and etiologic implications. Crit. Rev. Oncol. Hematol. 34, 27–53 (2000). 65. Kanzler, H., Kuppers, R., Hansmann, M. L. & Rajewsky, K. Hodgkin and Reed-Sternberg cells in Hodgkin’s disease represent the outgrowth of a dominant tumour clone derived from (crippled) germinal center B cells. J. Exp. Med. 184, 1495–1505 (1996). One of the first studies to isolate the malignant cells of Hodgkin’s lymphoma and definitively determine the origin of these cells by immunoglobulin-gene sequencing. 66. Kuppers, R. Molecular biology of Hodgkin’s lymphoma. Adv. Cancer Res. 84, 277–312 (2002). 67. Schwering, I. et al. Loss of the B lineage-specific gene expression program in Hodgkin and Reed-Sternberg cells of Hodgkin lymphoma. Blood 101, 1505–1512 (2003). 68. Polack, A. et al. c-myc activation renders proliferation of Epstein–Barr virus (EBV)-transformed cells independent of EBV nuclear antigen 2 and latent membrane protein 1. Proc. Natl Acad. Sci. USA 93, 10411–10416 (1996). 69. Kovalchuk, A. L. et al. Burkitt lymphoma in the mouse. J. Exp. Med. 192, 1183–1190 (2000). 70. Li, Z. et al. A global transcriptional regulatory role for c-myc in Burkitt’s lymphoma. Proc. Natl Acad. Sci. USA 100, 8146–8169 (2003). 71. Lindstrom, M. S. & Wiman, K. G. Role of genetic and epigenetic changes in Burkitt lymphoma. Semin. Cancer Biol. 12, 381–387 (2002). 72. Chapman, C. J., Mockridge, C. I., Rowe, M., Rickinson, A. B. & Stevenson, F. K. Analysis of V-H genes used by neoplastic B cells in endemic Burkitts lymphoma shows somatic hypermutation and intraclonal heterogeneity. Blood 85, 2176–2181 (1995). 73. Klein, U., Klein, G., Ehlin-Henriksson, B., Rajewsky, K. & Kuppers, R. Burkitt’s lymphoma is a malignancy of mature B cells expressing somatically mutated V-region genes. Mol. Med. 1, 495–505 (1995). 74. Harris, R. S., Croom-Carter, D., Rickinson, A. B. & Neuberger, M. S. Epstein–Barr virus and the somatic hypermutation of immunoglobulin genes in Burkitt’s lymphoma cells. J. Virol. 75, 10488–10492 (2001). 75. Kelly, G., Bell, A. & Rickinson, A. B. Epstein–Barr virusassociated Burkitt lymphomagenesis selects for downregulation of the nuclear antigen EBNA1. Nature Med. 8, 1098–1104 (2002). 76. Pajic, A. et al. Antagonistic effects of c-myc and Epstein–Barr virus latent genes on the phenotype of human B cells. Int. J. Cancer 93, 810–816 (2001). 77. Kang, M. S., Hung, S. C. & Kieff, E. Epstein–Barr virus nuclear antigen 1 activates transcription from episomal but not integrated DNA and does not alter lymphocyte growth. Proc. Natl Acad. Sci. USA 98, 15233–15238 (2001). 78. Kiss, C. et al. T cell leukaemia I oncogene expression depends on the presence of Epstein–Barr virus in the virus-

79. 80.

81.

82.

83.

84.

85.

86. 87. 88.

89.

90. 91.

92.

93.

94. 95.

96.

97.

98.

99.

100.

101.

102.

103. 104.

105.

106.

carrying Burkitt lymphoma lines. Proc. Natl Acad. Sci. USA 100, 4813–4818 (2003). Boshoff, C. & Weiss, R. AIDS-related malignancies. Nature Rev. Cancer 2, 373–382 (2002). Araujo, I. et al. Frequent expansion of Epstein–Barr virus (EBV) infected cells in germinal centres of tonsils from an area with a high incidence of EBV-associated lymphoma. J. Pathol. 187, 326–330 (1999). Jones, J. F. et al. T-cell lymphomas containing Epstein–Barr viral DNA in patients with chronic Epstein–Barr virus infections. N. Engl. J. Med. 318, 733–741 (1988). Harabuchi, Y. et al. Epstein–Barr virus in nasal T-cell lymphomas in patients with lethal midline granuloma. Lancet 335, 128–130 (1990). Kanegane, H., Nomura, K., Miyawaki, T. & Tosato, G. Biological aspects of Epstein–Barr virus (EBV)-infected lymphocytes in chronic active EBV infection and associated malignancies. Crit. Rev. Oncol. Haematol. 44, 239–249 (2002). Kikuta, H. et al. Epstein–Barr virus genome-positive T lymphocytes in a boy with chronic active EBV infection associated with Kawaski-like disease. Nature 333, 455–457 (1988). Kanavaros, P. et al. Nasal T-cell lymphoma: a clinicopathologic entity associated with peculiar phenotype and with Epstein–Barr virus. Blood 81, 2688–2695 (1993). Raab-Traub, N. Epstein–Barr virus in the pathogenesis of NPC. Semin. Cancer Biol. 12, 431–441 (2002). Yu, M. C. & Yuan, J. M. Epidemiology of nasopharyngeal carcinoma. Semin. Cancer Biol. 12, 421–429 (2002). Yu, M. C., Ho, J. H., Lai, S. H. & Henderson, B. E. Cantonese-style salted fish as a cause of nasopharyngeal carcinoma: report of a case–control study in Hong Kong. Cancer Res. 46, 956–961 (1986). Raab-Traub, N. & Flynn, K. The structure of the termini of the Epstein–Barr virus as a marker of clonal cellular proliferation. Cell 47, 883–889 (1986). Zeng, Y. Seroepidemiological studies on nasopharyngeal carcinoma in China. Adv. Cancer Res. 44, 121–138 (1985). Chan, A. T. C. et al. Plasma Epstein–Barr virus DNA and residual disease after radiotherapy for undifferentiated nasopharyngeal carcinoma. J. Natl Cancer Inst. 94, 1614–1619 (2002). Pathmanathan, R. et al. Undifferentiated, non-keratinising, and squamous cell carcinoma of the nasopharynx. Variants of Epstein–Barr virus-infected neoplasia. Am. J. Pathol. 146, 1355–1367 (1995). Shibata, D. & Weiss, L. M. Epstein–Barr virus-associated gastric adenocarcinoma. Am. J. Pathol. 140, 769–774 (1992). Tokunaga, M. et al. Epstein–Barr virus in gastric carcinoma. Am. J. Pathol. 143, 1250–1254 (1993). Imai, S. et al. Gastric carcinoma: monoclonal epithelial malignant cells expressing Epstein–Barr virus latent infection protein. Proc. Natl Acad. Sci. USA 91, 9131–9135 (1994). Schneider, B. G. et al. Loss of p16/CDKN2A tumour suppressor protein in gastric adenocarcinoma is associated with Epstein–Barr virus and antomic location in the body of the stomach. Hum. Pathol. 31, 45–50 (2000). Lee, H. S., Chang, M. S., Yang, H. K., Lee, B. L. & Kim, W. H. Epstein–Barr virus-positive gastric carcinoma has a distinct protein expression profile in comparison with Epstein–Barr virus-negative carcinoma. Clin. Cancer Res. 10, 1698–1705 (2004). zur Hausen, A. et al. Epstein–Barr virus in gastric carcinomas and gastric stump carcinomas: a late event in gastric carcinogenesis. J. Clin. Pathol. 57, 487–491 (2004). Bonnet, M. et al. Detection of Epstein–Barr virus in invasive breast cancers. J. Natl Cancer Inst. 91, 1376–1381 (1999). Sugawara, Y. et al. Detection of Epstein–Barr virus (EBV) in hepatocellular carcinoma tissue: a novel EBV latency characterised by the absence of EBV-encoded small RNA expression. Virology 256, 196–202 (1999). Murray, P. G. et al. Reactivity with a monoclonal antibody to Epstein–Barr virus (EBV) nuclear antigen 1 defines a subset of aggressive breast cancers in the absence of the EBV genome. Cancer Res. 63, 2338–2343 (2003). Li, J.-H. et al. Tumour-targeted gene therapy for nasopharyngeal carcinoma. Cancer Res. 62, 171–178 (2002). Li, J.-H. et al. Efficacy of targeted FasL in nasopharyngeal carcinoma. Mol. Ther. 8, 964–973 (2003). Israel, B. F. & Kenney, S. C. Virally targeted therapies for EBV-associated malignancies. Oncogene 22, 5122–5130 (2003). Feng, W.-H., Hong, G., Delecluse, H. J. & Kenney, S. C. Lytic induction therapy for Epstein–Barr virus-positive B-cell lymphomas. J. Virol. 78, 1893–1902 (2004). Ambinder, R. F., Robertson, K. D. & Tao, Q. DNA methylation and the Epstein–Barr virus. Semin. Cancer Biol. 9, 369–375 (1999).


REVIEWS 107. Chan, A. T. C. et al. Azacitidine induces demethylation of the Epstein–Barr virus genome in tumours. J. Clin. Oncol. 22, 1373–1381 (2004). 108. Chodosh, J. et al. Eradication of latent Epstein–Barr virus by hydroxyurea alters the growth-transformed cell phenotype. J. Infect. Dis. 177, 1194–1201 (1998). 109. Slobod, K. S. et al. Epstein–Barr virus-targeted therapy for AIDS-related primary lymphoma of the central nervous system. Lancet 356, 1493–1494 (2000). 110. Piche, A., Kasono, K., Johanning, F., Curiel, T. J. & Curiel, D. T. Phenotypic knockout of the latent membrane protein 1 of Epstein–Barr virus by an intracellular single-chain antibody. Gene Ther. 5, 1171–1179 (1998). 111. Kenney, J. L., Guinness, M. E., Curiel, T. & Lacy, J. Antisense to the Epstein–Barr virus (EBV)-encoded latent membrane protein 1 (LMP-1) suppresses LMP-1 and Bcl-2 expression and promotes apoptosis in EBV-immortalised B cells. Blood 92, 1721–1727 (1998). 112. Cahir-McFarland, E. D., Davidson, D. M., Schauer, S. L., Duong, J. & Kieff, E. NFκB inhibition causes spontaneous apoptosis in Epstein–Barr virus-transformed lymphoblastoid cells. Proc. Natl Acad. Sci. USA 97, 6055–6060 (2000). 113. Farrell, C. J. et al. Inhibition of Epstein–Barr virus-induced growth proliferation by a nuclear antigen EBNA2-TAT peptide. Proc. Natl Acad. Sci. USA 101, 4625–4630 (2004). 114. Kirchmaier, A. L. & Sugden, B. Dominant-negative inhibitors of EBNA-1 of Epstein–Barr virus. J. Virol. 71, 1766–1775 (1997). 115. Rooney, C. M. et al. Use of gene-modified virus-specific T lymphocytes to control Epstein–Barr virus-related lymphoproliferation. Lancet 345, 9–13 (1995). First study to use adoptive transfer of EBV-specific T cells for the treatment of lymphoproliferative disease. 116. Khanna, R. et al. Activation and adoptive transfer of Epstein–Barr virus-specific cytotoxic T cells in solid organ transplant patients with post-transplant lymphoproliferative disease. Proc. Natl Acad. Sci. USA 96, 10391–10396 (1999). 117. Haque, T. et al. Treatment of Epstein–Barr virus-positive post-transplantation lymphoproliferative disease with partly HLA-matched allogeneic cytotoxic T cells. Lancet 360, 436–442 (2002). 118. Roskrow, M. A. et al. Epstein–Barr virus (EBV)-specific cytotoxic T lymphocytes for the treatment of patients with EBV-positive relapsed Hodgkin’s disease. Blood 91, 2925–2934 (1998). 119. Wagner, H. J. et al. Expansion of EBV latent membrane protein 2a specific cytotoxic T cells for the adoptive immunotherapy of EBV latency type 2 malignancies: influence of recombinant IL12 and IL15. Cytotherapy 5, 231–240 (2003).

768


120. Paludan, C. et al. Epstein–Barr nuclear antigen 1-specific CD4+ Th1 cells kill Burkitt’s lymphoma cells. J. Immunol. 169, 1593–1603 (2002). 121. Taylor, G. S. et al. Dual stimulation of Epstein–Barr virus (EBV)-specific CD4+/– and CD8+/– T cell responses by a chimeric antigen construct: Potential therapeutic vaccine for EBV-positive nasopharyngeal carcinoma. J. Virol. 78, 768–778 (2004). 122. Frisan, T. et al. Local suppression of Epstein–Barr virus (EBV)-specific cytotoxicity in biopsies of EBV-positive Hodgkin’s disease. Blood 86, 1493–1501 (1995). 123. Wagner, H. J. et al. A strategy for treatment of Epstein–Barr virus-positive Hodgkin’s disease by targeting interleukin 12 to the tumour environment using tumour antigen-specific T cells. Cancer Gene Ther. 11, 81–91 (2004). 124. Baer, R. et al. DNA sequence and expression of the B95-8 Epstein–Barr virus genome. Nature 310, 207–211 (1984). Sequencing of the prototype strain of EBV — at the time, this was the largest region of DNA ever sequenced and the first human virus to be fully sequenced. 125. Brauninger, A. et al. Epstein–Barr virus (EBV)-positive lymphoproliferations in post-transplant patients show immunoglobulin V gene mutation patterns suggesting interference of EBV with normal B cell differentiation processes. Eur. J. Immunol. 33, 1593–1602 (2003). 126. Brauninger, A. et al. Survival and clonal expansion of mutating ‘forbidden’ (immunoglobulin receptor-deficient) Epstein–Barr virus-infected B cells in angioimmunoblastic T cell lymphoma. J. Exp. Med. 194, 927–940 (2001). 127. Pathmanathan, R., Prasad, U., Sadler, R., Flynn, K. & RaabTraub, N. Clonal proliferations of cells infected with Epstein–Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N. Engl. J. Med. 333, 693–698 (1995). First demonstration of clonal EBV infection in premalignant lesions of the nasopharynx. 128. Lo, K.-W. & Huang, D. P. Genetic and epigenetic changes in nasopharyngeal carcinoma. Semin. Cancer Biol. 12, 451–462 (2002). Important review summarizing the genetic and epigenetic events that contribute to the pathogenesis of NPC and indicating that some of these occur before EBV infection. 129. Knox, P. G., Li, Q. X., Rickinson, A. B. & Young, L. S. In vitro production of stable Epstein–Barr virus-positive epithelial cell clones which resemble the virus: cell interaction observed in nasopharyngeal carcinoma. Virology 215, 40–50 (1996). 130. Zhou, S., Fujimuro, M., Hsieh, J. J., Chen, L. & Hayward, S. D. A role for SKIP in EBNA2 activation of CBF1-repressed promoters. J. Virol. 74, 1939–1947 (2000). 131. Artavanis-Tsakonas, S., Matsuno, K. & Fortini, M. E. Notch signaling. Science 268, 225–232 (1995).

132. Sakai, T. et al. Functional replacement of the intracellular region of the Notch1 receptor by Epstein–Barr virus nuclear antigen 2. J. Virol. 72, 6034–6039 (1998). 133. Huen, D. S., Henderson, S. A., Croom-Carter, D. & Rowe, M. The Epstein–Barr virus latent membrane protein-1 (LMP1) mediates activation of NF-κB and cell surface phenotype via two effector regions in its carboxyterminal cytoplasmic domain. Oncogene 10, 549–560 (1995). 134. Eliopoulos, A. G. et al. Epstein–Barr virus-encoded latent infection membrane protein 1 regulates the processing of p100 NFκB2 to p52 via an IKKγ/NEMOindependent signalling pathway. Oncogene 22, 7557–7569 (2003). 135. Dykstra, M. L., Longnecker, R. & Pierce, S. K. Epstein–Barr virus coopts lipid rafts to block the signalling and antigen transport functions of the BCR. Immunity 14, 57–67 (2001). 136. Ikeda, M., Ikeda, A., Longan, L. C. & Longnecker, R. The Epstein–Barr virus latent membrane protein 2A PY motif recruits WW domain-containing ubiquitin-protein ligases. Virology 268, 178–191 (2000). 137. Chen, S.-Y., Lu, J., Shih, Y.-C. & Tsai, C.-H. Epstein–Barr virus latent membrane protein 2A regulates c-Jun protein through extracellular signal-regulated kinase. J. Virol. 76, 9556–9561 (2002). 138. Young, L. S. & Murray, P. G. Epstein–Barr virus and oncogenesis: from latent genes to tumours. Oncogene 22, 5108–5121 (2003).

Acknowledgements The authors apologize to colleagues whose primary research papers are not cited because of the limited number of references. The authors thank D. Huang, P. Murray and G. Niedobitek for assistance with the figures and D. Williams for secretarial support. The authors’ studies are supported by Cancer Research UK, the Leukaemia Research Fund and the Medical Research Council, UK.

Competing interests statement The authors declare no competing financial interests.

Online links DATABASES The following terms in this article are linked online to: Cancer.gov: http://www.cancer.gov Hodgkin’s lymphoma | nasopharyngeal cancer Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene BCR | CD21 | INK4A | MYC | NF-κB | PKR | TP53 Access to this links box is available online.
