Skip to main content
. 2023 Mar 2;14:1028162. doi: 10.3389/fimmu.2023.1028162

Figure 1.

Figure 1

Schematic representation of the six main categories of models used for NK alloreactivity prediction and summary table of the characteristics of each category. Adapted from Dhuyser et al. (16). (A) Ligand–ligand model confronts the MHC of the donor with the MHC of the recipient: KIR genotyping is unknown and NK alloreactivity of the donor toward host cells is expected when the recipient lacks MHC class I ligand present in the donor. (B) Receptor–ligand model considers the KIR of the donor and the MHC of the recipient: if at least one KIR gene expressed in the donor does not recognize any of the MHC molecules of the recipient (“missing-ligand”), the NK cells of the donor will increase their cytotoxic activity. (C) Educational models consider the MHC class I molecules of the donor and recipient and the KIR typing of the donor. It should reflect the “education” process required for NK cells to become competent. (D) The KIR haplotypes of the donor: the B/x of the donor and particularly those carrying Cen-B/B are expected to be more alloreactive toward the cell of the recipients, as they carry mostly activating KIR genes. (E) KIR matching models represent the number of aKIR and/or iKIR gene present in the donor but absent in the recipient and vice versa. (F) KIR polymorphism leads to KIR molecules with relevant biological differences.