GSBS Dissertations and Theses

Publication Date


Document Type

Doctoral Dissertation

Academic Program

Immunology and Microbiology



First Thesis Advisor

Liisa K. Selin


Epstein-Barr Virus Infections, HLA-A2 Antigen, CD8-Positive T-Lymphocytes


A person is exposed to many pathogens throughout their lifetime, and with the resolution of each infection, there remains a pool of pathogen-specific immune cells that protect that person from re-infection with the same pathogen. However, there is a great deal of evidence to suggest that the pool of pathogen-specific memory cells can also participate in the immune response to future infections with unrelated pathogens. Many believe T cells to be cross-reactive in nature because of their interaction with self antigens during development in the thymus and their interaction with foreign antigens once in the periphery. There are many features of the interaction between a T cell and its ligand that facilitate this cross-reactive nature. Based on solved crystal structures, relatively few contacts are required for a stable interaction, and that interaction is often mediated by the flexible CDR3 loops of the T cell receptor that accommodate ligands of various structure. There is also evidence in the murine and human systems that subsets of virus-specific memory CD8 T cells take on an activated phenotype upon infection with an unrelated virus. In murine models, these memory T cell subsets could kill target cells, secrete several cytokines, and proliferate in response to a cross-reactive stimulation, suggesting that a cross-reactive T cell response could impact the outcome of a viral infection. In fact, upon heterologous infection, mice immune to a previous virus were often protected, having lower titers of the second unrelated virus, their epitope-specific and T cell receptor repertoires were often skewed, and they were more prone to immune-mediated pathologies. All of these observations coincided with the presence of cross-reactive T cell responses. Thus, we define heterologous immunity as changes in viral replication and the disease pathology associated with that viral infection as a result of the host's history of infection, and this can be mediated, in part, by cross-reactive CD8 T cell responses.

Since many human viral infections are associated with a wide range of disease states, we questioned whether cross-reactive CD8 T cell responses occurred as commonly as they appeared to occur in the murine models and whether they influenced the outcome of such infections. Epstein-Barr virus (EBV) infects over 90% of the U. S. population and has a large genome with the capacity to encode a multitude of T cell epitopes. The first part of this thesis research focuses on the identification of cross-reactive CD8 T cell responses with specificity for known epitopes derived from EBV, a common human virus. We directed our study to HLA-A2-restricted responses because of the common expression of this MHC Class I allele in the U. S. population. This study resulted in the detection of cross-reactive responses with five different specificities that involved either the immunodominant lytic EBV-BMLF1280 epitope or the latent EBNA 3A596epitope. Three of the cross-reactive responses had specificity for epitopes derived from another unrelated, but common, human virus, influenza A virus (IV). Each of these cross- reactive responses had the potential to participate in the collective immune response to acute EBV infection.

EBV is also well-suited as a model system to study heterologous immunity in humans, as infection at an early age is frequently asymptomatic, while the same infection during adolescence often results in an immune-mediated syndrome, infectious mononucleosis (IM). Since older individuals have presumably been exposed to more pathogens in their lifetime and, therefore, would have memory CD8 T cell pools with more extensive specificities, we hypothesized that acute EBV infection activated cross-reactive memory CD8 T cell responses that promoted the development of IM. In order to determine if the cross-reactive responses we identified above contributed to the immune response to acute EBV infection, we first screened the blood of IM patients for cross-reactive T cells with specificity for EBV-BMLFl280 and IV-M158. The total number of M1-specific T cells of 5 of 8 patients was increased at presentation with IM, which was suggestive of their specific activation during the EBV infection since a bystander mechanism would have resulted in 8 out of 8 patients having increased numbers of M1-specific T cells. Our hypothesis was further supported by the fact that we clearly detected cross-reactive T cells capable of recognizing both BMLF1 and M1 epitopes in the blood of 2 of the 5 IM patients with an augmented M1-specific T cell frequency. Furthermore, the M1-specific TCR repertoires of those two patients were dramatically skewed, which was an indication of cross-reactive M1-specific T cell expansions and, therefore, participation in the lymphoproliferation characteristic of IM. In addition, T cell lines derived from 3 out of 8 healthy donors with previous exposure to both viruses contained a subset of T cells that responded to both BMLF1 and M1 epitopes, suggesting that these cross-reactive cells are often maintained in memory. These cross-reactive T cells were cytotoxic and produced MIP-1β, IFNγ, and TNFα, functions which could potentially promote the symptoms of IM and, indeed, may have been contributed to the severe case of IM noted in one patient.

The final part of this thesis research focused on defining the structure of the cross-reactive TCR that recognized both BMLF1 and M1 epitopes, which have only 33% sequence similarity. In addition, we examined the cross-reactive TCR repertoire organization of multiple individuals to determine the breath and, therefore, the likelihood that this cross-reactive T cell response will occur. These studies revealed that a wide range of Vα and Vβ families can mediate interaction with both epitopes and that the cross-reactive TCR repertoire was unique to each individual, relying heavily on the T cell clones present in that individual's private BMLF1- and M1-specific repertoires. We also observed an increased frequency of TCRs with longer CDR3 regions within the cross-reactive repertoire, which were often extended by non-bulky amino acid residues that could provide these TCRs with more flexibility in order. to accommodate the two different epitope structures.

Given that we detected a cross-reactive T cell response with specificity for two immunodominant epitopes derived from two of the most common human viruses among people that share one of the most common MHC Class I alleles in the U. S. population, we predict that cross-reactive T cells are common components of human immune responses. The variability in the magnitude and specificity of each cross-reactive T cell response is dependent on each individual's unique history of infection and th,eir unique TCR repertoire, and such responses likely represent one of many factors that could explain the individual variability in disease severity associated with EBV and many other human viral infections.



Rights and Permissions

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