Publication Date


Document Type

Doctoral Dissertation

Academic Program

Immunology and Microbiology



First Thesis Advisor

John L. Sullivan M.D.


Herpesvirus 4, Human, T-Lymphocytes, Cytotoxic, Immunity, Cellular


Epstein-Barr virus (EBV) is a ubiquitous human herpesvirus which causes acute infectious mononucleosis and is etiologically associated with malignant lymphoproliferative disorders including Burkitt's lymphoma, nasopharyngeal carcinoma, B-cell lymphomas in immunocompromised hosts, Hodgkin's disease, T cell lymphomas, and smooth muscle tumors in allograft recipients. The medical significance of EBV is underscored by its potent growth transforming effects on human B-lymphocytes in-vitro and the potentially oncogenic consequences of infection in-vivo. The majority of EBV-associated malignancies occur in the setting of chronic infection and strong virus-specific humoral immunity, suggesting that cellular immunity is primarily responsible for preventing the outgrowth of EBV-transformed B cells in-vivo. Similarly, primary EBV infection in adolescents and adults stimulates an intense cytotoxic-T-lymphocyte (CTL) response which coincides with a marked reduction in the number of infected B cells in the peripheral blood. Evidence of previous EBV infection can be confirmed by the presence of EBV-specific, HLA-restricted memory T cells in the peripheral blood which inhibit the outgrowth of newly EBV-transformed B cells and efficiently lyse established autologous B-lymphoblastoid cell lines.

Worldwide, EBV is responsible for substantial morbidity, comparable to measles, mumps and hepatitis virus, for which vaccines exists. Accordingly, the potential public health impact of an EBV vaccine has reinforced our efforts to identify the immunodominant virus-encoded T-cell epitopes which stimulate naive CTL effectors during acute infection and maintain memory CTL surveillance during convalescence. The EBV-encoded antigens against which the memory CTL response is directed have been partially defined, and include most of the EBV latent proteins (EBNA-2, 3a, 3b, 3c, LP, and LMP-l, 2a, 2b) consistently expressed by in-vitro EBV-transformed B lymphocytes (type-III latency). Importantly, all EBV-associated malignancies express EBNA-1, and as yet no EBNA-1-specific memory CTL have been convincingly demonstrated. Additionally, many EBV-specific CTL lines and clones have been described which do not recognize any of the known latent proteins or other EBV protein antigens tested thus far. Thus while much is known about CTL-mediated immunity against EBV, our knowledge of EBV-derived CTL epitopes remains incomplete. In contrast to the EBV-specific memory CTL response, very little is known about the source of viral epitopes recognized during the primary CTL response to EBV. In this regard, acute infectious mononucleosis represents an ideal model system to study virus-specific, cell-mediated immunity. Acute IM is a self-limited illness characterized by the appearance of "atypical" lymphocytes (CD3+/CD8+/HLADR+), including both virus-specific and alloreactive CTL, which undoubtedly contribute to virus elimination and provide CTL precursors for life-long immunity to EBV.

Like other herpesvirus, EBV can undergo either lytic or latent cycle replication. During primary EBV infection many lytic cycle genes are expressed which are likely responsible for stimulating the intense cellular immune response associated with acute infectious mononucleosis. During convalescence a minor population of circulating B cells remain latently infected, harbor multiple EBV episomes, and express only EBNA-1 and possibly LMP-2a (type-I latency). Thus, latency type-I infected B cells in-vivo express a much more restricted spectrum of latent proteins and are therefore not subject to elimination by the same virus-specific CTL as are type-III EBV latently infected cells. Accordingly, many mechanisms have been proposed to explain EBV persistence including; restricted expression of EBV latent genes, reduced levels of cellular adhesion molecules, downregulation of MHC class-I molecules, absence of EBNA-1 T-cell-epitopes, and most recently, EBNA-1-mediated inhibition of antigen processing. While these mechanisms may contribute to ineffective T cell surveillance against latency type-I EBV- infected cells, B cells expressing the full spectrum of latent proteins (type-III) also exist transiently in vivoand maintain detectable humoral and CTL responses to most latent proteins.

Our first goal was to identify the virus-encoded immunodominant antigens recognized by in-vivoactivated MHC class-I restricted CTL isolated from college students experiencing primary EBV infection, manifested as acute IM. Following a prodromal period of several weeks, newly EBV infected patients present with signs and symptoms of acute IM, including elevated numbers of activated CD8+ T cells in their peripheral blood, many of which, like memory CTL, are EBV-specific and HLA-restricted. In order to address the issue of EBV persistence and the immune control of EBV-induced lymphoproliferation, we also studied the long-term EBV-specific memory CTL response in these same individuals.

Blood from acute IM patients and healthy EBV seropositive donors served as a source of peripheral blood lymphocytes to generate bulk CTL cultures and autologous target cells. The infecting strain of EBV was determined for each patient by DNA-PCR amplification of virus from saliva. Lymphocytes were isolated from whole blood by Ficoll-Paque density centrifugation and T- and B-cell enriched populations were obtained by AET-sheep red cell rosette selection. Autologous B cell blasts served as a source of target cells and recombinant vaccinia virus constructs were used to introduce individual EBV latent genes into target cells. Expression of individual EBV genes in target cells was confirmed by both western blot and immunofluorescence. Primary CTL responses to EBV were evaluated in standard 5lCr release assays using freshly isolated, T-cell enriched PBL from acute IM patients as effector cells. EBV-specific memory CTL responses were evaluated with bulk CTL culture generated by in-vitro restimulation with autologous B-LCLs. FACS analyses were routinely performed on bulk cultures of effector CTL populations in order to more clearly characterize their phenotype. Lastly, monoclonal antibody blocking studies and cold target competition assays were performed in order to accurately identify the viral antigen and MHC components responsible for target cell recognition.

Our results based upon evaluation of 35 acute IM patients and 32 convalescent patients demonstrate that the virus-specific primary CTL response is broadly directed against the full spectrum of latent proteins, including EBNA1 and the viral coat glycoprotein gp350, while the memoryCTL response, which essentially lacks EBNA1 reactivity, is directed primarily against the EBNA 3 family of proteins (3A, 3B, 3C). Importantly, the immunodominant response by both primary and memory CTL was directed against the EBNA3 proteins.

CTL from 7 of the 35 acute IM patients evaluated recognized EBNA1 expressing targets, and in 4 of these 7 patients, EBNA1 was an immunodominant antigen. Similarly, CTL from 7 of 35 acute IM patients recognized gp350 transfected targets, while no gp350-specific memory CTL responses were observed.

While the phenotype of in-vivo primed CTL effectors were CD8+/HLA-DR+/CD11b+, the major subpopulation of memory CTL were CD8+/HLA-DR+/CD11b-. The CD11b "memory marker" reached peaked levels on the first sample day for all patients and gradually declined to baseline levels over a period of several months. In contrast, the CD11b marker was quickly shed from in vitropropogated CTL, over a period of 5-10 days.

Target cell lysis by in-vivoactivated CTL was almost completely blocked by antibody directed againt [against] class-I molecules (BBM.1), whereas the effect of blocking target cell lysis by anti-CD8 mAb varied between 40-75%. These findings are consistent with an absolute need for class-I restricted antigen presentation, and imply that CD8 was variably required, likely for the lower affinity TCR/ Ag combinations. Cell lysis mediated by in-vitro-restimulated memory CTL was also largely inhibited by anti-class-I mAb, while anti-CD8 mAb was only mild/moderately effective in blocking target cell lysis, in keeping with the concept that memory CTL bear higher avidity TCR which can recognize antigen independent of CD8.

Our detection of only one EBNA1-specific memory CTL response among the 32 patients tested supports the theory that latently infected B cells in-vivo, expressing only EBNA1, escape CTL recogition and thus might serve as a reservoir for viral persistence and/or reactivation. The rare ability to detect an EBNA1-specific memory CTL responses remains a relatively unexplained phenomenon and may involve a number of tolerizing mechanisms including the induction of anergy by presentation of EBNA-1 in the absence of costimulation, clonal deletion of low affinity T cells, the absence of dominant T cell epitopes within EBNA1 or a result of the recently described inhibiting properties of EBNA-1 on antigen processing and presentation.

Alternatively, the absence of detectable EBNA1-specific memory CTL may be the result of insufficient or inappropriate restimulation of memory CTL in vitro. We addressed this possibility by attempting to selectively restimulate and expand EBNA1-specific CTL from acute IM patients by using EBNA1 expressing B cells blasts as a stimulus. Effector cells generated in this manner killed target cells in an MHC class-I restricted manner but were specific for an unspecified vaccinia antigen. Interestingly, the phenotype of the effector cells was predominantly CD3+/CD4-/CD8-/γδ T cells.

In summary, our findings suggest that a multitude of previously unrecognized, EBV-specific CTL are present in the peripheral blood during acute IM, and include EBNA-1-specific CTL. The importance of accurately defining the in-vivo immune response to EBV is underscored by the ever-growing list of EBV associated malignancies. In addition to providing insights into the oncogenesis and potential treatment of NPC, a newly described link between precursor lesions and EBV infection raises the possibility that heightened immunity to EBV or EBV-infected cells may prevent the development of NPC. An obvious expectation would include extension of such knowledge to other EBV associated malignancies such as B and T cell lymphomas, Hodgkin's lymphomas, and smooth muscle tumors. First however, existing gaps in knowledge regarding the immune response to EBV and EBV-associated malignancies must be closed. Details about the viral gene products which are involved in stimulating a broadly protective, virus-specific immune response in a large number of individuals is fundamental to the design of an effective EBV vaccine. Since the presence of activated CD8+ T cells correlates with the rapid decline of EBV infected B cells in the peripheral blood, a concise description of the EBV-specific CTL response in the setting of acute infection will be necessary for the rational design of an effective acute IM vaccine. Increased understanding of viral escape mechanisms is also likely to contribute to therapeutic modalities to treat autoimmune disorders.


Some images did not scan well. Please see print version for images if desired.

Rights and Permissions

Copyright is held by the author, with all rights reserved.