Reevaluating HLA-A2-restricted Lassa epitopes in human Lassa fever survivors

Abstract

Many factors contribute to the selection of epitope-specific T cells. Often, candidate epitopes are generated through in silico prediction of various biological processes, mainly peptide generation and MHC binding, that are then tested in relevant biological assays. This is especially true of epitopes derived from BSL-4 pathogens or where relevant patient samples are difficult to obtain. Two previous studies identified CD8+ T cell epitopes from the Lassa virus glycoprotein through in silico prediction, experimental MHC binding assays, and epitope generation in HLA-A2 transgenic mice. Using samples from ten HLA-A\*02:01 Lassa fever survivors, we tested whether these previously described epitope-specific CD8+ T cells were present and their relation to the broader Lassa virus-specific T cell response. Using overnight stimulation assays, we detected robust LASV-specific responses to the glycoprotein and nucleoprotein, but only one of the three epitopes (GP60–68)shown to be present and protective in mice made a substantial contribution to the overall LASV-specific response. Using a more sensitive proliferation assay, we detected the remaining two epitopes in some individuals at a very low frequency. Overall, this study shows the limitations of epitope discovery through in silico prediction, MHC binding and transgenic mouse models and highlights the complex nature of T cell selection during natural infection of humans.