words)Human herpesviruses are antigenically rich agents that induce strong CD8+T cell responses in primary infection yet persist for life, continually challenging T cell memory through recurrent lytic replication and potentially influencing the spectrum of antigen-specific responses. Here we describe the first lytic proteome-wide analysis of CD8+ T cell responses to the Epstein-Barr gamma1-herpesvirus (EBV), and the first such proteome-wide analysis of primary versus memory CD8 responses to any human herpesvirus. Primary effector preparations were generated directly from activated CD8+ T cells in the blood of infectious mononucleosis (IM) patients by in vitro mitogenic expansion. For memory preparations, EBV-specific cells in the blood of long-term virus carriers were first re-stimulated in vitro by autologous dendritic cells loaded with a lysate of lytically-infected cells, then expanded as for IM cells. Preparations from 7 donors of each type were screened against each of 70 EBV lytic cycle proteins in combination with the donor's individual HLA class I alleles. Multiple reactivities against immediate early (IE), early (E) and late (L) lytic cycle proteins, including many hitherto unrecognised targets, were detected in both contexts. Interestingly however, the two donor cohorts showed a different balance between IE, E and L reactivities. Primary responses targeted IE and a small group of E proteins preferentially, seemingly in line with their better presentation on the infected cell surface before later-expressed viral evasins take full hold. By contrast, target choice equilibrates in virus carriage with responses to key IE and E antigens still present but with responses to a select subset of L proteins now often prominent. We infer that, for EBV at least, long-term virus carriage with its low level virus replication and lytic antigen release is associated with a re-shaping of the virus-specific response.3 Author Summary (156 words) Epstein-Barr virus (EBV) is a herpesvirus and an important human pathogen that normally persists for life and induces strong CD8+T cell responses. Primary EBV infection can cause mononucleosis (IM) disease and EBV is associated with malignant lymphocytes cancers and epithelial cells. Here for the first time, we described proteome-wide analysis of CD8+ T cell responses to EBV lytic antigens, and compare the CD8 T cell responses between primary and memory CD8 responses to EBV. The EBV primary CD8 T cells responses targeted IE and a small group of E proteins preferentially. By contrast, the memory CD8 T cells responses from EBV persistent infection are dominated by L proteins. This study can help to understand how human CD8 T cells respond evolves with the long-term virus infection. For EBV at least, we infer that long-term EBV virus carriage with its low level virus replication and lytic antigen release is associated with a re-shaping of the virus-specific response.4