High-avidity PR1-specific CD8+ lymphocytes contribute to the maintenance of remission in patients with myeloid leukemia.

摘要

Our lab has identified a leukemia-associated antigen called PR1, which is recognized by T cells that specifically kill leukemia. These PR1-specific CD8+ cytotoxic-T-lymphocytes (PR1-CTL) can be found in HLA-A*0201+ healthy individuals and patients that go into remission after treatment with interferon-alpha or bone marrow transplantation, but cannot be detected in newly diagnosed leukemia patients and patients that are refractory to treatment. We have also determined that the T cell receptor (TCR) avidity of ex-vivo expanded PR1-CTL is inversely proportional to the concentration of PR1 used to elicit them and directly proportional to both tetramer fluorescence intensity and T cell effector function. We hypothesize that patients with cytogenetic remission mediated in part by T cells should have more high-avidity PR1-CTL than patients that are refractory to treatment. Furthermore, differences in effector function of high and low avidity PR1-CTL may be due to differences in the TCR structure.; To determine whether PR1-CTL contribute to maintaining remission in interferon-alpha-sensitive leukemia patients, we studied the TCR avidity, surface phenotype and effector function. Patients that remain in remission have interferon-gamma secreting, high avidity PR1-CTL in their peripheral blood that are predominantly CD45RA+CCR7+CD28+CD57-, a phenotype attributed to naive T cells. A fraction of PR1-CTL expresses the CD8+CD45RA-CCR7+CD28+CD57- phenotype of central memory T cells. Similar results were observed in patients vaccinated with the PR1 peptide. The high frequencies of a self-antigen specific T cell population that exhibit a recall response suggest that the PR1-CTL are not antigen-inexperienced despite expressing a naive phenotype and that the central memory PR1-CTL are a self-renewing population whose progeny may re-express CD45RA. Patients whose high avidity PR1-CTL lose function suffer relapse of leukemia. Finally, to determine whether observed differences in the effector function of PR1-CTL that correlated with TCR avidity were due differences in TCR structure, we established high and low avidity PR1-CTL clones, sequenced their TCR and compared the structures by homology modeling. We determined that the differences in the avidity of these two populations of PR1-CTL clones may be due to differences in structure of the CDR2 regions that interact with the HLA-A*0201 molecule and influence TCR affinities.