Human CD8+ CD57- T(EMRA) cells: Too young to be called "old"

End-stage differentiation of antigen-specific T-cells may precede loss of immune responses against e.g. viral infections after allogeneic stem cell transplantation (SCT). Antigen-specific CD8+ T-cells detected by HLA/peptide multimers largely comprise CD45RA-/CCR7- effector memory (T(EM)) and CD45RA+/CCR7- T(EMRA) subsets. A majority of terminally differentiated T-cells is considered to be part of the heterogeneous T(EMRA) subset. The senescence marker CD57 has been functionally described in memory T-cells mainly composed of central memory (T(CM)) and T(EM) cells. However, its role specifically in T(EMRA) cells remained undefined. Here, we investigated the relevance of CD57 to separate human CD8+ T(EMRA) cells into functionally distinct subsets. CD57- CD8+ T(EMRA) cells isolated from healthy donors had considerably longer telomeres and showed significantly more BrdU uptake and IFN-γ release upon stimulation compared to the CD57+ counterpart. Cytomegalovirus (CMV) specific T-cells isolated from patients after allogeneic SCT were purified into CD57+ and CD57- T(EMRA) subsets. CMV specific CD57- T(EMRA) cells had longer telomeres and a considerably higher CMV peptide sensitivity in BrdU uptake and IFN-γ release assays compared to CD57+ T(EMRA) cells. In contrast, CD57+ and CD57- T(EMRA) cells showed comparable peptide specific cytotoxicity. Finally, CD57- CD8+ T(EMRA) cells partially changed phenotypically into T(EM) cells and gained CD57 expression, while CD57+ CD8+ T(EMRA) cells hardly changed phenotypically and showed considerable cell death after in vitro stimulation. To the best of our knowledge, these data show for the first time that CD57 separates CD8+ T(EMRA) cells into a terminally differentiated CD57+ population and a so far functionally undescribed “young” CD57- T(EMRA) subset with high proliferative capacity and differentiation plasticity.