Abstract
BACKGROUND: The functional relevance of HLA-C mismatches in an alloresponse is still much debated, putting into doubt the relevance of matching for this antigen in selection of an allogeneic bone marrow donor. In addition to presenting peptides to T cells, HLA-C also functions as a ligand for killing inhibitory receptors (KIRs) on natural killer (NK) cells. In the current study we provide an in vitro basis to address the question of whether mismatches for this antigen are a risk factor for acute graft-versus-host disease (GVHD). METHODS AND RESULTS: By analysis of cytotoxic and helper T-lymphocyte precursor frequency (CTLp-f and HTLp-f) in 153 pairs, we are able to show that isolated HLA-C mismatches appear less immunogenic than do isolated HLA-A mismatches. Strikingly, the presence of an HLA-C mismatch next to a HLA-DRB or HLA-DQB mismatch leads to a synergistic increase in CTLp-f outcome. Moreover, we are the first to show that absence of a single inhibitory epitope as a result of an HLA-C mismatch can be sufficient to induce NK mediated alloreactivity, that is, kill and proliferate. CONCLUSIONS: We conclude that, in most cases, isolated HLA-C mismatches may be acceptable with respect to T-cell-mediated alloreactivity; however, the presence of a strong helper epitope (DR/DQ mismatch) appears sufficient to overcome the low immunogenicity of HLA-C. HLA-C mismatches that affect KIR epitopes, can induce NK mediated alloreactivity. This suggests that, in HLA-A-, -B-, -DR-, and -DQ-matched patients, NK cells may play a role in the induction and development of acute GVHD.
