Epstein-Barr Virus in Burkitt’s Lymphoma: a role for Latent Membrane Protein 2A

Abstract

Burkitt’s lymphoma (BL) is characterized by translocation of the MYC gene to an immunoglobulin locus. Transgenic mouse models have been used to study the molecular changes that are necessary to bypass tumor suppression in the presence of translocated MYC. Inactivation of the p53 pathway is a major step to tumor formation in mouse models that is also seen in human disease. Human BL is often highly associated with Epstein-Barr virus (EBV). The EBV latency protein latent membrane protein 2A (LMP2A) is known to promote B cell survival by affecting levels of pro-survival factors. Using LMP2A transgenic mouse models, we have identified a novel mechanism that permits lymphomagenesis in the presence of an intact p53 pathway. This work uncovers a contribution of EBV to molecular events that have documented importance in BL pathogenesis, and may underlie the poorly understood link between EBV and BL.

The link between Epstein-Barr virus (EBV) and human cancer is constantly evolving. Uncovering the mechanisms that underlie this association has lead not only to insight into the biology of the virus, but also to discovery of novel concepts relevant to all of cancer research. The relationship between EBV and cancer is historically important, and is responsible for the initial discovery of EBV as a novel herpesvirus. In 1957, Denis Burkitt, an Irish surgeon working in Uganda, began to observe jaw tumors in children in Kampala.^{1, 2} Further investigation showed that these tumors were relatively common in young children who lived in regions with specific temperature and rainfall conditions. The distribution pattern suggested to Dr. Burkitt that an infectious agent may be responsible for the malignancies known as Burkitt’s lymphoma (BL).³ Tumor samples were sent to the UK where Anthony Epstein, Yvonne Barr, and Bert Achong cultured cell lines from the tumors ⁴, and identified a novel herpesvirus in these cultured cells, which we now call Epstein-Barr virus.^{3, 5}

Since then, we have learned much more about the transformative capability of EBV, and a role for the virus in additional lymphocyte and epithelial cell malignancies has been revealed. We have also progressed in our understanding of the specific functions of various components of this large DNA virus. However, the molecular characteristics and epidemiology of BL tumors make the relevant viral functions that are important in development of this particular malignancy difficult to identify. As a result, the exact mechanism or mechanisms that explain the part played by EBV in BL lymphomagenesis are still not entirely clear.

Burkitt’s Lymphoma

Based on epidemiologic data, specific forms of Burkitt’s lymphoma (BL) can be described. Denis Burkitt observed a form of BL that is now known as endemic BL and is most commonly seen in regions of sub-Saharan Africa. Regions of endemic BL have a very high frequency of disease, about 5–10 cases per 100,000 children.⁶ Facial tumors, especially of the jaw, are the most common presentation of endemic BL, but late-stage patients can also develop abdominal tumors with CNS involvement.^{7, 8} The geography of endemic BL is strikingly similar to the geography of holoendemic malaria. The distributions of these diseases are so similar, in fact, that it was initially suggested that BL may be a mosquito-borne disease.^{9, 10} Malaria infection likely does contribute to BL pathogenesis; although the mechanisms that link the two diseases are debated. Epstein-Barr virus infection, however, is indisputably linked to endemic BL. Viral genomes can be found in nearly 100% of endemic BL tumors.⁶

BL occurs worldwide at a much lower incidence than endemic BL in a form known as sporadic BL. Like endemic BL, sporadic BL is also seen primarily in children.^{6, 7} However, sporadic BL has a lower association with EBV infection, which varies by region from 15 to 85% of tumors containing viral genomes.¹¹ Interestingly, areas with higher BL incidence tend to have a stronger EBV association.⁶ A third form of the tumor is seen in HIV patients. BL tumors make up about 30% of AIDS-associated lymphomas. EBV is found more often in HIV-associated BL tumors than in sporadic BL tumors, at least in Western nations, where 30–40% of HIV-associated BL tumor cells harbor EBV.⁶

The various forms of BL are broadly categorized as “non-Hodgkin’s lymphomas”, distinguishable from other non-Hodgkin’s lymphomas by histological analysis. Lymphoma cells in BL tumors tend to be round, monomorphic, and highly proliferative. Tumor cells express germinal center markers, including CD10 and Bcl-6.⁶ The tumors have a histological appearance that is described as a “starry sky” pattern caused by phagocytosis of apoptotic debris by infiltrating macrophages.

Regardless of EBV status, all forms of BL have acquired a translocation between the proto-oncogene MYC and an immunoglobulin locus, leading to deregulated expression of MYC in B lymphocytes. Three translocations are observed in BL tumor cells. The most common is t(8:14) which places MYC upstream of the immunoglobulin heavy chain (IgH) locus.⁶ Expression of MYC from kappa or lambda light chain loci, t(2:8) and t(8:22) are also observed in BL tumor cells.⁶ BL tumors are thought to be derived from germinal center B cells. Double stranded DNA breaks are common during B cell maturation in the germinal center, where class switch recombination and somatic hypermutation occur.¹² It is reasonable that chromosomal translocations occur in this environment.

MYC is a transcription factor containing a basic-helix-loop-helix-zipper (bHLH-zip) domain that binds DNA containing E-box sequences (CACGTG).^{13, 14} MYC regulates expression of many genes, generally acting to promote cell growth and proliferation and inhibit cell cycle arrest.¹⁵ MYC activity is highly regulated through interactions with MAX and MAD proteins, and disruption of these regulatory mechanisms leads to tumorigenesis in mouse models.^{16, 17} Furthermore, deregulation of MYC through various mechanisms including chromosomal translocations, is implicated in many human cancers.

In addition to induction of cell proliferation, expression of MYC also sensitizes cells to apoptosis triggered by a variety of stimuli. Early studies showed an accelerated apoptotic response in cells with deregulated MYC after serum or growth factor withdrawal,^{18, 19} and MYC-induced apoptosis can also occur in cells exposed to hypoxia, DNA damage, and signaling through TRAIL receptors.²⁰

The major tumor suppressor pathway, p53, is responsible for much of the apoptotic signaling attributed to deregulated MYC (Figure 1). Like MYC, p53 is an important transcription factor, modulating the expression of many cell growth and survival genes. The outcome of p53 activation is cell cycle arrest or apoptosis. In normal cells, levels of p53 are low, but various forms of stressful stimuli trigger a cascade that leads to p53 protein stabilization. ARF and MDM2 are two important players that regulate p53 induction. Activation of a proto-oncogene such as MYC generates mitogenic signals that induce expression of ARF.²¹ ARF inhibits the activity of MDM2, an ubiquitin ligase that negatively regulates p53 by targeting the protein for degradation.²² Activation of ARF, therefore, stabilizes p53 by releasing the tumor suppressor from negative control by MDM2. This system contains feedback loops, including the ability of p53 to transcriptionally target MDM2, increasing expression of its own negative regulator (Figure 1).

Figure 1. Early events in the apoptotic cascade induced by deregulated MYC.

Deregulation of MYC causes mitogentic signals that trigger ARF. ARF inhibits MDM2, a ubiquitin ligase that targets p53 for protein degradation. Nutlin 3 is an inhibitor that blocks the interaction between MDM2 and p53, leading to activation of p53 and downstream effectors. Activated p53 can bind DNA and induce expression of pro-apoptotic proteins, including PUMA, NOXA, and BAX. In addition, p53 transcriptionally upregulates MDM2, and inhibits ARF, mechanisms that downregulate the p53 response. PUMA, NOXA, and BIM are pro apoptotic BH-3 only proteins that can directly bind to pro-survival Bcl-2 family members, such as Bcl-2, Bcl-X_L, and Mcl-1. Pro-survival Bcl family members prevent insertion of BAX into the mitochondrial membrane. BAX insertion triggers release of cytochrome c from the mitochondria, leading to activation of downstream caspases and, ultimately, apoptosis.

Transcriptional regulation by active p53 triggers apoptosis or cell cycle arrest. Interestingly, MYC opposes cell cycle arrest by inhibiting the activity and expression of p21, a major p53 transcriptional target.^23–25 Therefore apoptosis is the major functional outcome of p53 activation in the context of deregulated MYC. To trigger apoptosis, p53 activates expression of pro-apoptotic BH-3 only proteins, including PUMA and NOXA.^{26, 27} Active p53 can directly bind the promoter for PUMA or NOXA^{28, 29}, however, PUMA seems to be most functionally important in lymphocytes. Both PUMA and NOXA promote apoptosis by binding Bcl-2 family members such as Bcl-2, Bcl-X_L and MCL-1, inhibiting their pro-survival functions (Figure 1).^28–31 The balance of pro-apoptotic to pro-survival Bcl family members ultimately regulates survival versus apoptosis.³²

Burkitt’s Lymphoma and p53

Evasion of apoptosis is a hallmark of cancer, and often occurs through lesions in the p53-ARF-MDM2 pathway. Among hematological malignancies, p53 is most frequently mutated in BL, and the site of mutations differs from those observed in solid tumors.³³ The human p53 gene, tp53, contains 393 codons, and the gene product contains five domains: an RNA polymerase transactivation domain, a proline-rich domain, a central DNA-binding domain, a tetramerization domain, and a carboxy terminal regulatory domain. Most p53 mutations in human tumors are missense and cluster within four highly conserved regions in exons 5–8 located within the central DNA-binding domain.^{34, 35} Two classes of mutations within the DNA-binding domain of p53 have been identified in human lymphomas.^{36, 37} One class disrupts the structure of p53 by displacing zinc molecules; in the second class p53 retains its structure but cannot contact DNA. The latter has a worse prognosis in lymphomas.³⁸ Evidence also suggests that mutant p53 protein accumulates within cells, allowing detection by immunoblotting techniques.³⁹

Mutations in tp53 are found in 30–40% of BL biopsies and greater than 50% of BL-derived cell lines, indicating the role of p53 mutations in tumor progression.^40–43 In human BL, most missense mutations within p53 cluster within a stretch of 33 amino acids (codons 213 to 248), located within the central DNA binding domain.^{33, 40} Despite the identification of both p53 mutations and EBV genomes in BL biopsies, there appears to be no clear link between the two as evidenced from a number of small studies on non-endemic BL.^44–47 In particular, it is unclear whether endemic BL, with a near 100% association with EBV, has higher frequencies of p53 mutation.

Several mouse models of BL have been useful in helping define the role of the p53-ARF-MDM2 pathway in development of human BL. The λ-MYC and Eμ-MYC transgenic mouse models express MYC under the control of the Ig light chain and Igμ heavy chain promoter and enhancer, respectively.^{48, 49} Both mouse models develop spontaneous pre-B and B cell cervical lymph node tumors at a high penetrance and with striking histopathological similarity to human BL.^{50, 51} In addition, p53 mutations in the mouse are comparable to humans; mutations occur within the central DNA-binding domain at similar frequency.¹⁶

Mutation of trp53 in Eμ-MYC lymphomas has been well documented. Approximately 80% of tumor cells contain lesions within the p53-ARF-MDM2 pathway, with p53 missense mutations found in 28%.¹⁶ Deletion of ARF was seen exclusively in tumors with wild-type p53, indicating that MYC activates the p53 pathway through ARF while also selecting for cells that have sustained a p53 mutation or biallelic ARF deletion. However, 20% of Eμ-MYC tumors do not contain p53-ARF-MDM2 lesions, which implies that other apoptotic pathways may also be affected.

Generation of Eμ-MYC mice lacking alleles for BAX, BIM and PUMA accelerates MYC-induced lymphomagenesis, and a reduction in p53 pathway alterations are found in these rapidly arising tumors.^52–54 Loss of PUMA in PUMA^−/− Eμ-MYC tumors results in a lower frequency of p53 mutations compared to PUMA^+/+ Eμ-MYC tumors.⁵⁴ In BAX^−/− Eμ-MYC tumors, no p53 mutations are detected, unlike BAX ^+/− Eμ-MYC tumors.⁵³ A similar pattern of wild-type ARF is seen in BIM^−/− Eμ-MYC tumors compared to Eμ-MYC and BIM ^+/+ Eμ-MYC tumors.⁵² These data suggest that selective pressure on the p53 pathway may be removed by mutations in downstream pro-apoptotic targets of p53 for lymphomagenesis.

Epstein-Barr virus and Burkitt’s Lymphoma

Analyses of EBV-positive cell lines and EBV-infected tumor biopsies have shown that latent infection can take one of several forms in which a different pattern of latent genes are expressed. EBV is highly efficient at transforming primary B cells in culture into proliferating, latently infected lymphoblastoid cell lines (LCLs). In LCLs, EBV expresses the full spectrum of latent genes; two small nuclear RNAs (EBERs), the highly spliced BamHI rightward transcripts (BARTs), three integral latent membrane proteins, (LMP1, -2A and -2B) and six EBV nuclear antigens (EBNA1, -2, -3A, -3B, -3C, and EBNA-LP).^{55, 56} This pattern of gene expression is also observed in EBV-associated immunoblastic lymphoma.

A common feature of all EBV-associated malignancies is the expression of one or more latent membrane proteins. Of the three latent membrane proteins, LMP1 is known to have oncogenic activity in vitro and in vivo via constitutive CD40-like signaling.^57–61 LMP1 is expressed in nasopharyngeal carcinoma and Hodgkin’s lymphoma (HL). In the latter, LMP2A is also expressed, along with EBNA1, BARTs and EBERs, suggesting that LMP2A may play a role in tumorigenesis in some cell types.

Conventionally, BL biopsies and tumor-derived cell lines were thought to express EBNA1 only.^{62, 63} However, a number of studies suggested that BL biopsies express additional latent proteins, including LMP1, LMP2A and EBNA2.^64–66 LMP2A expression was recently confirmed in endemic BL biopsies using a sensitive RT-PCR assay, however, LMP1 and EBNA2 transcripts were not detected by this method.⁶⁷ These data support the possibility of a functional role for LMP2A in human BL.

LMP2A signal transduction and function

Much of our knowledge concerning LMP2A signal transduction and function in B lymphocytes derives from experiments using EBV LCLs in vitro and LMP2A-transgenic mice.^68–75 LMP2A has a 118 amino terminal cytoplasmic domain that contains 8 tyrosine residues, two of which form an immunoreceptor tyrosine-based activation motif (ITAM).⁷⁶ Data indicate that LMP2A uses the tyrosines in its cytoplasmic tail to mimic low levels of B cell receptor (BCR) signaling by activating proteins normally utilized by the BCR. For example, initial studies using LCLs show that LMP2A constitutively phosphorylates Lyn and Syk with Lyn binding to tyrosine 112 and Syk binding to the ITAM motif (tyrosines 74 and 85) of LMP2A.^{77, 78} Additional studies using LMP2A-Tg mice demonstrate that LMP2A constitutively phosphorylates and activates many of the proteins induced by the BCR, such as Lyn, Syk, BLNK, BTK, Ras, PI3K, NF-kB, and MAP kinases.^79–84 Using both experimental models, these results indicate that LMP2A signal transduction mimics the signal used by the BCR.

More importantly, multiple studies indicate that LMP2A functionally acts as a BCR mimic, since murine and human B cells that are negative for a BCR survive in the presence of LMP2A.^{75, 81, 85} The LMP2A-transgenic mouse (TgE) line generates B cells that survive in the absence of a BCR, suggesting that LMP2A functionally mimics a BCR to promote B cell development.^{75, 81} In a study by Mancao and Hammerschmidt, human BCR-negative B cells infected with wildtype EBV, but not recombinant EBV lacking LMP2A, were rescued from apoptosis.⁸⁵ This study further confirms that LMP2A acts as a BCR mimic in human cells.

As suggested above, one common function of LMP2A in multiple experimental models is the protection of B cells from apoptosis ^{81, 85–87}. Protection of cells from apoptosis is a sensitive balance between the production of anti-apoptotic and pro-apoptotic factors.³² Therefore, for LMP2A to protect B cells from apoptosis, LMP2A should modulate the levels of factors that influence the induction of apoptosis. For example, studies identified that LMP2A increases levels of the anti-apoptotic protein survivin^{88, 89} that may regulate protection of cells from apoptosis. Additionally, LMP2A increases levels of NF-kB⁷⁹ to increase levels of Bcl family members that are required for the protection of resting and activated B cells from apoptosis (M. Swanson-Mungerson and R. Longnecker, manuscript in preparation). Finally, LMP2A increases the levels of RAS/PI3K/AKT pathway, which induces an increase in Bcl-X_L levels that are required for LMP2A-dependent protection from apoptosis.⁸¹ This finding is interesting in light of the finding that LMP2A protects B lymphocytes overexpressing c-MYC from apoptosis and increases levels of Bcl-X_L in these cells.⁹⁰ Taken together, the ability of LMP2A to increase the levels of anti-apoptotic Bcl family members⁸¹ may shift the balance of pro-apoptotic and anti-apoptotic factors towards B cell survival and tumor development.

LMP2A and Burkitt’s Lymphoma

The ability of LMP2A to influence the balance of survival factors in B lymphocytes may be functionally important BL. The link between EBV and BL is strong enough to suggest a functional role for EBV in lymphomagenesis; however the mechanistic role for the virus is unidentified. Detection of low levels of LMP2A transcripts in fresh tumor biopsies⁹¹ further supported our hypothesis that LMP2A protects B cells from apoptosis induced by deregulated MYC in human BL.

To begin to test this hypothesis, we turned to our transgenic mouse lines. As a model for BL, we used the λ-MYC mouse model, in which MYC is expressed from the immunoglobulin lambda locus.⁴⁸ Interestingly, when our TgE LMP2A transgenic line was crossed with the λ-MYC transgenic mice, the double transgenic LMP2A-TgE/λ-MYC mice developed lymph node tumors more quickly than λ-MYC mice alone.⁹⁰ In addition, prior to tumor onset, spleen size was greatly enlarged in the double transgenic mice. This enlargement could be attributed to a dramatic increase in the number of B220+ B cells in the spleens of the LMP2A-TgE/λ-MYC mice. Cell cycle analysis showed a decrease in the proportion of apoptotic cells and an increase in proliferating cells when comparing B cells from LMP2A-TgE/λ-MYC mice to λ-MYC B cells.⁹⁰ Increased levels of the pro-survival protein Bcl-X_L could be detected in pretumor spleens as well as tumor samples from the LMP2A-TgE/λ-MYC mice when compared to pretumor spleens and tumor samples from λ-MYC mice.⁹⁰ These data support our hypothesis that the pro-survival function of LMP2A protects cells from apoptosis induced by translocated MYC in BL, and the ability of LMP2A to induce Bcl-X_L ^{81, 90} may underlie the contribution of LMP2A, and EBV, to BL.

The level of LMP2A expression detected in BL biopsies is quite low, and multiple lines of evidence suggest that levels of expression of LMP2A can dramatically alter experimental outcomes.^{68, 92, 93} To address the possibility that levels of LMP2A may affect the ability of the protein to function in the context of deregulated MYC, we used our Tg6 LMP2A transgenic line, which expresses lower levels of LMP2A and does not display the LMP2A-induced developmental phenotypes that are characteristic of the TgE LMP2A transgenic line.⁶⁸ Upon crossing the Tg6 LMP2A line with λ-MYC mice, we found that the double transgenic LMP2A-Tg6/λ-MYC mice also demonstrate a rapid time to tumor that is considerably faster than λ-MYC mice.⁹⁴ In addition, increased spleen size can be observed in the LMP2A-Tg6/λ-MYC mice as early as 3 weeks of age.

We were interested in further examining the mechanism that underlies the phenotypes observed for the LMP2A/λ-MYC mice. An important characteristic of the tumors that develop in MYC transgenic mice is inactivation of the ARF-MDM2-p53 tumor suppressor pathway.¹⁶ This lead us to investigate the status of the p53 pathway in tumors from our LMP2A-Tg6/λ-MYC mice. We examined stabilization of ARF and p53 protein by immunoblot and sequenced exons within the p53 DNA binding domain in tumor DNA. In line with previous studies, we identified mutations in p53 in the majority of the spontaneous tumors from λ-MYC mice. Interestingly, we found no p53 lesions in tumors from the LMP2A-Tg6/λ-MYC mice.⁹⁴ To analyze p53 pathway function in tumor cells, we used the inhibitor Nutlin 3, which blocks the interaction between MDM2 and p53, thereby inhibiting negative regulation of p53 (Figure 1). Nutlin 3 has been shown to activate downstream p53 targets and induce apoptosis.⁹⁵ Tumor cells from λ-MYC mice containing a p53 mutation were unaffected by Nutlin treatment. In contrast, LMP2A-Tg6/λ-MYC tumors which lacked p53 lesions were sensitive to Nutlin 3 and unable to survive in high concentrations of the inhibitor.⁹⁴ Furthermore, freshly isolated LMP2A-Tg6/λ-MYC tumor cells contain increased levels of PUMA, a major p53 target, when compared to λ-MYC tumors, further supporting that the p53 pathway is intact.⁹⁴ We can conclude from these studies that even the weaker LMP2A signal found in our Tg6 LMP2A transgenic line is able to bypass the MYC-induced pro-apoptotic p53 response.

Expression of LMP2A is a novel mechanism that protects cells from apoptosis in the context of MYC deregulation, accelerating tumor development. We have developed a model to relate our results to development of BL in mice and humans. In previously existing models for mouse BL, MYC is expressed as a transgene throughout B cell development. Spontaneous tumors develop in these mice after period of time that allows for an inactivating mutation to occur in some step of the p53 pathway (Figure 2A). In our LMP2A mouse model for BL, both the LMP2A and MYC transgenes are expressed throughout B cell development. Deregulated expression of MYC triggers the p53 pathway, but the LMP2A survival signal protects the B cells from apoptosis, resulting in an initial expansion of B cells, which is observed as splenomegaly in pre-tumor mice. LMP2A, in the presence of deregulated MYC, also promotes accelerated onset of tumors, bypassing p53 pathway inactivation (Figure 2B).

Figure 2. LMP2A in a mouse model of Burkitt’s lymphoma.

(A) In the λ-MYC mouse model, MYC is expressed from an immunoglobulin promoter in B cells. Expression of deregulated MYC activates the p53 pathway, triggering apoptosis. Burkitt’s lymphoma-like tumors are eventually observed in these mice upon mutation or inactivation of one or more components of the p53 pathway. (B) Both LMP2A and MYC are expressed in B cells throughout development in the LMP2A/λ-MYC transgenic mouse. The MYC transgene activates the p53 pathway, but LMP2A counteracts this activation, likely at some step that is downstream of PUMA, possibly through upregulation of anti-apoptotic proteins. LMP2A causes an initial expansion of cells, observed as splenomegaly in pre-tumor animals. In the absence of immune selection, LMP2A accelerates tumor onset in this model, allowing tumorigenesis in the presence of an intact p53 pathway.

Our data suggest that LMP2A may play a functional role in development of human BL. One major hurdle in comparing our transgenic mouse models to human BL is a lack of epidemiological data reporting a decrease in the frequency of p53 mutations in EBV-positive compared to EBV-negative BL. However, we still believe that LMP2A is functionally important in BL tumors, and have developed a model to propose a role for LMP2A in human disease. First, in EBV-negative BL, (Figure 3A), the MYC translocation occurs in a germinal center. Similar to our MYC mouse model, we know that MYC induces a p53 response, triggering apoptosis. It is not surprising then, that mutations in the p53 pathway are frequently seen in EBV-negative BL. In EBV-positive BL (Figure 3B), an EBV-infected cell undergoes a germinal center reaction and acquires a MYC translocation. Although MYC activates p53, LMP2A provides survival signals, such as increasing levels of anti-apoptotic proteins, which protect the cell from apoptosis, leading to an expansion of pretumor cells similar to the splenomegaly we observe in our double transgenic mice. Although the frequency of p53 mutation in BL tumors is not dependent on EBV status ^{33, 42}, p53 mutations are detected in a higher percentage of BL cell lines than in fresh biopsies.³³ This likely reflects an association of p53 mutation with tumor progression. In our model, the expansion of cells mediated by MYC and LMP2A increases the probability of acquiring a mutation in p53 in an EBV-positive cell. Mutation in p53 leads to tumor progression, after which weak immune surveillance not present in our transgenic mouse model may select against cells expressing high levels of LMP2A. The p53 bypass function of LMP2A is no longer needed once a p53 mutation has occurred, consistent with the low levels of LMP2A detected in BL biopsies (Figure 3B).⁹¹

Figure 3. LMP2A in human Burkitt’s lymphoma.

(A) In EBV-negative human BL, translocation of MYC to an immunoglobulin locus occurs during a germinal center reaction. MYC induces p53 pathway activation, triggering apoptosis. Cells which have inactivated the p53 pathway grow into lymphoma. (B) In EBV-positive human BL, LMP2A is present during the germinal center reaction and MYC translocation. LMP2A enhances the survival of cells with de-regulated MYC early in lymphomagenesis, allowing expansion of cells that overexpress MYC. The expansion of cells containing deregulated MYC increases the probability of acquisition of a secondary mutation, which leads to tumor progression. After tumor progression, immune regulation selects against high levels of LMP2A in tumor cells, resulting in low levels of LMP2A in tumor biopsies. LMP2A is indicated by the multi spanning transmembrane protein. Constitutive MYC expression either by translocation or transgene construct is indicated by black/grey line in nucleus.

*Figure 2 and Figure 3 are adapted from ref 94.

Conclusion

Investigation into the mechanism of BL pathogenesis has driven EBV research, while studies targeted toward a better understanding of EBV have lead to advances in cancer biology. Even so, the exact mechanisms that underlie tumor development in BL are still in question. The absence of transcripts for viral proteins with a known role in viral oncogenesis has confounded efforts to understand this question. EBV LMP2A may be an important component of the missing link between EBV and BL. Transcripts of the EBV latency protein LMP2A are present at low levels in BL biopsies. LMP2A increases the levels of pro-survival Bcl family members in B lymphocytes, shifting the balance of pro-apoptotic and anti-apoptotic factors to promote cell survival. This function of LMP2A allows for bypass of p53 inactivation in a MYC tumor model. We propose a role for LMP2A early in development of BL, where the survival signal allows for expansion of cells that contain a MYC translocation. The expanded cells increase the probability of acquiring a p53 mutation, which leads to tumor progression. After the p53 mutation, the tumor cells become less dependent on LMP2A and immune selection may explain the low levels of LMP2A present in tumor biopsies. The history, epidemiology, and molecular characteristics of Burkitt’s lymphoma make it a fascinating topic that is sure to be continually investigated in the future.

Abbreviations

BL

Burkitt’s lymphoma

EBV

Epstein-Barr virus

LMP2A

latent membrane protein 2A

LCL

lymphoblastoid cell line

Ig

immunoglobulin

BCR

B cell receptor