Epstein-Barr virus (EBV) is an ubiquitous human γ-herpesvirus and is associated with several lymphoid and epithelial malignancies, like Burkitt's lymphoma (BL), nasopharyngeal carcinoma (NPC), Hodgkin's disease (HD), B- and T-cell non-Hodgkin's lymphoma and gastric carcinomas. EBV is transcriptionally active in the malignant cells of these cancers, leading to abundant expression of a restricted set of EBV encoded proteins. These gene expression patterns can be divided into 3 different latency patterns. Latency II type is characterised by the expression of Epstein-Barr nuclear antigen 1 (EBNA1), latent membrane protein 1 (LMP1) and latent membrane protein (LMP2). This latency pattern is found in HD and NPC and these type n malignancies characteristically occur in apparently immunocompetent individuals. Of the three EBV proteins expressed in HD and NPC, EBNA1 is not processed. The glycine-alanine repeat (GAr) sequences within this protein inhibit proteasomal processing and thus this protein is not an option for immunotherapy. Thus LMP1 and LMP2 are considered to be the only suitable targets for immunotherapy.
In recent years CD8+ cytotoxic T-cell (CTL) responses to multiple LMP2 epitopes have been described, but only a limited number of epitopes have been described for LMP1. Little is known about the CD4+T-cell responses directed against LMPl and LMP2. In the present study, the ex vivo T-cell response directed against LMPl was analysed using interferon-γ-(IFN-γ)- based enzyme linked immunospot (ELISPOT) assays in a large panel of healthy EBV carriers of diverse ethnic origin and NPC patients. By comparing the frequencies of T cells specific for overlapping peptides spanning LMP1, 18 novel CD8 and CD4 restricted LMP1 T-cell epitopes were identified. Of the 7 new T-cell epitopes with defined HLA restrictions, 5 were HLA A2-restricted and one is restricted through both HLA-B57 and HLA-B58 while another epitope showed potential CD4+ T-cell reactivity. More importantly, extensive sequence analysis of LMP1 revealed that the majority of the T-cell epitopes are highly conserved in EBV isolates from Caucasian, Papua New Guinean, African and Southeast Asian populations.
In the next phase of this study, the efficacy of delivering LMP1 epitopes as a therapeutic vaccine was examined to test whether this strategy would potentially provide a long-term benefit to EBV-associated HD and NPC patients. Hence a polyepitope vaccine comprising 6 HLA A2-restricted LMP1 epitopes expressed was developed using a recombinant vaccinia virus vector. Human cells infected with this recombinant polyepitope construct were efficiently recognised by LMP1-specific CTL lines from HLA A2 healthy individuals. Furthermore, immunisation of HLA A/Kb mice with this polyepitope vaccine consistently generated strong LMP1 specific CTL responses to 5 of the 6 epitopes which were readily detected by both ex vivo and in vitro assays. More importantly, this polyepitope vaccine successfully reversed the outgrowth of LMP1 expressing tumours in HLA A2/Kb mice.
Furthermore, another polyepitope vaccine was made to provide coverage over a broad range of HLA types with particular emphasis on HLA types present in NPC endemic regions (HLA A2, Al 1, A24, B27, B40) and both LMP1 and LMP2 epitopes were included. This polyepitope was inserted into a clinical grade replication-incompetent adenovirus vector. Immunisation with this adenovirus vector expressing the LMP polyepitope was capable of inducing multiple independent MHC-restricted CTL responses. These epitopes were not only efficiently processed endogenously by the human cells but also recalled memory CTL responses specific for LMP antigens in healthy virus carriers. In vivo studies showed that this polyepitope vaccine was capable of inducing a primary T-cell response which was shown to be therapeutic in a tumour challenge model by reversing LMP1-expressing tumours in HLA-A2/Kb mice.
The final phase of this study was aimed to profile LMP memory response in primary (newly diagnosed and relapsed) and long-term HD patients. The HD patients were all serologically positive but importantly in about half of the patients the malignant Hodgkin-Reed-Stemberg (H-RS) cells were LMP positive. The results show that the memory response was significantly weak or undetectable in primary HD patients when compared to long-term HD patients although the latter group were not as strong as that seen in healthy immune individuals. These results indicate that HD patients with active disease suffer a reduced LMP-specific CTL memory response. Importantly, this reduction of LMP-specific T-cell response was independent of the LMP status of the H-RS cells. The activation protocol includes a number of novel aspects designed to minimise the period required to achieve significant levels of lysis in bulk cultures from patients.
The results of these studies are important in terms of developing therapeutic strategies against HD and NPC. On the one hand, the adenovirus-polyepitope might be used as a therapeutic vaccine to raise the CTL response to LMP1 and LMP2 epitopes. On the other hand, the in vitro activation and expansion of these memory responses to these proteins offers the potential of their use in autologous adoptive transfer. In this application, the adenovirus-polyepitope would be used to activate autologous LMP1- and 2-specific bulk cultures and to adoptively transfer them back into HD and NPC patients. Both of these protocols might be used as either stand-alone therapy or in concert.