Abstract

The hematopoietic stem cell transplant (HCT) donor KIR genotype has been correlated with disease-free survival in patients with acute myelogenous leukemia. The Killer Cell Immunoglobulin-like Receptor (KIR) gene family encodes highly homologous pairs of activating and inhibiting receptors, 2DL1–2DS1; 2DL2/3–2DS2; and 3DL1–3DS1. Inhibitory members are known to regulate NK cell function through interactions with HLA Class I antigens. The role of activating KIRs and their ligand specificity is, however, not well defined. The activating receptor, KIR2DS1, is known to bind the HLA-Cw C2 group antigens and we have recently demonstrated a role for this receptor in NK cell allorecognition. In contrast, KIR2DS2 does not bind HLA-Cw C1 group antigens, and a functional role of this receptor even in NK allorecognition has not been established. We now demonstrate, that presence of the activating KIR2DS2 gene in NK donors homozygous for the HLA-KIR ligand group C2 is associated with significant alloreactivity against C1 homozygous target cells (polyclonal NK cells, p=0.006; NK clones, p=0.001). This alloreactivity is mediated by “missing self” on the target and is dominated by “lack of C2 group on target”. The “missing C2” effect was absent, however, in C2 homozygous donors lacking 2DS2 (p=0.99). Only very rare cytotoxic NK clones expressing GL183 (2DL2/3, 2DS2) and with alloreactivity against C1 targets could be generated in vitro from 2DS2-positive, C2 homozygous donors. A majority of these rare GL183-positive clones did not demonstrate inhibitory function against the HLA class I deficient 721.221 transfected with Cw3 (C1-group), and GL183 cross-linking of the clones resulted in increased cytokine production. Thus, KIR2DS2 is an activating receptor in NK clones from C2 homozygous donors, but does not appear to recognize C1 ligand. We next investigated 2DS2 function in donors heterozygous for the C groups (i.e. C1/C2). Analysis of NK cell function in a 2DS2-positive, C1/C2 donor revealed a “missing HLA-KIR ligand” effect for the C2 group. Cytotoxicity by IL2-propagated, polyclonal NK cells and NK clones revealed allocytotoxicity against targets lacking the C2 group (p<0.001). In addition, a repertoire analysis on 138 NK clones generated from this donor revealed a marked increase in the number of EB6 (KIR2DL1/S1)-expressing NK clones (95%) compared to both the fresh (10%–50%) and the IL-2-expanded polyclonal NK repertoire (12%–60%). Additionally, all EB6-expressing clones from this donor were inhibited by the C2 ligand. Subsequent studies in freshly isolated NK cells following activating receptor cross-linking (NKp46, NKG2C, and CD16) or by alloantigen activation demonstrated that the functioning subset of NK cells in this donor predominantly expressed the EB6 receptor. Other inhibitory receptors (e.g. NKG2A, KIR3DL1, and KIR3DL2) did not contribute significantly to the functional subset of NK cells. Presence of 2DS2 in this donor was therefore associated with a “skewing” of the NK repertoire towards EB6 positivity, and dominated by functional NK cells that were inhibited by the “self” C2 ligand. Collectively, these studies provide the first evidence that activating KIR can influence the NK cell repertoire. Furthermore, our studies would indicate that presence of activating KIRs in HCT donors for recipients homozygous HLA-KIR ligands might induce post-transplantation graft versus host NK alloreactivity.

Author notes

Disclosure: No relevant conflicts of interest to declare.