Targeting B cell leukemia with highly specific allogeneic T cells with a public recognition motif

The possibility that allogeneic T cells may be targeted to leukemia has important therapeutic implications. As most tumor antigens represent self-proteins, high-avidity tumor-specific T cells are largely deleted from the repertoire of the patient. In contrast, T cells from major histocompatibility complex (MHC)-mismatched donors provide naïve repertoires wherein such cells have not been systematically eliminated. Yet, evidence for peptide degeneracy or poly-specificity warrants caution in the use of foreign human leukocyte antigen (HLA) or peptide complexes as therapeutic targets. Here, we cocultured HLA-A(*)0201-negative T cells with autologous dendritic cells engineered to present HLA-A(*)0201 complexed with a peptide from the B cell antigen CD20 (CD20p). HLA-A(*)0201/CD20p pentamer-reactive CD8(+) T cells were readily obtained from all donors. The polyclonal cells showed exquisite peptide and MHC specificity, and efficiently killed HLA-A(*)0201-positive B cells, including primary chronic lymphocytic leukemia cells. The T cell receptor (TCR) sequences displayed a novel type of conservation, with extensive homology in the TCR β chain complementarity-determining region 3 and in J, but not V, region. This is surprising, as the donors were HLA disparate and their TCR repertoires are expected to show little overlap. The results demonstrate the first public recognition motif for an allogeneic HLA/peptide complex. The allo-restricted T cells or TCRs could provide graft-versus-leukemia in the absence of graft-versus-host disease.