Donor and host coexpressing KIR ligands promote NK education after allogeneic hematopoietic stem cell transplantation

Abstract

The rate and extent of natural killer (NK)-cell education after hematopoietic cell transplantation correlates with leukemia control. To study the effect of donor and host HLA on NK-cell reconstitution, single killer-cell immunoglobulin-like receptor (KIR)+ NK cells (exhibiting KIR2DL1, KIR2DL2/KIR2DL3, or KIR3DL1 as their sole receptor) were grouped into 4 groups based on the interaction between donor/host HLA and donor inhibitory KIR in 2 cohorts (n = 114 and n = 276, respectively). On days 90 to 180 after transplantation, the absolute number and responsiveness against K562 cells (CD107a or interferon-γ expression) of single-KIR+ NK cells were higher in pairs where donor and host HLA both expressed ligands for donor inhibitory KIRs than in pairs where 1 or both of the donor and recipient HLA lacked at least 1 KIR ligand. NK-cell responsiveness was tuned commensurate with the number of inhibitory receptors from the donor. When both donor and host expressed the 3 major KIR ligands (HLA-C1, HLA-C2, and HLA-Bw4), NK cells expressing 3 inhibitory receptors (KIR2DL1/2DL3/3DL1) reached the maximum responsiveness against K562 cells compared with those NK cells expressing only 1 or 2 inhibitory receptors. When donor and host HLA both expressed all ligands for donor inhibitory KIRs, patients with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) showed the lowest recurrence rate after haploidentical hematopoietic stem cell transplantation (haplo-HSCT). In conclusion, this study demonstrates that when both donors and hosts present all the KIR ligands for donor KIRs, reconstituted NK cells achieve better functional education and contribute to least relapse among patients. This observation study was registered at www.clinicaltrials.gov as #NCT02978274.

Conflict of interest statement

Conflict-of-interest disclosure: The authors declare no competing financial interests.

© 2019 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

KIR2DL2/L3 single+NK cells show higher reactivity than KIR2DL1 single+NK cells in pairs of donor C1C1 or C1C2 and host C1C1. Expression of CD107a and IFN-γ against K562 by KIR2DL2/L3 single+ NK cells and KIR2DL1 single+ NK cells in pairs of donor C1C1 and host C1C1 (n = 14; A-B) and pairs of donor C1C2 and host C1C1 (n = 6; C-D) at days 90 and 180 after transplantation. **P < .01; *P < .05.

Figure 2.

KIR2DL2/L3 single+NK cells show higher reactivity than KIR3DL1 single+NK cells in pairs of donor Bw4C1C1 and host C1C1. Expression of CD107a and IFN-γ against K562 cells by KIR2DL2/L3 single+ NK cells and KIR3DL1 single+ NK cells in pairs of donor Bw4C1C1 and host C1C1 (n = 11; A-B) at days 90 and 180 after transplantation. Meanwhile, the expression of CD107a and IFN-γ against K562 by KIR2DL1 single+, KIR2DL2/L3 single+, and KIR3DL1 single+ NK cells in pairs of donor Bw4C1C2 and host Bw4C1C2 (n = 18; C-D) at days 90 and 180 after transplantation. **P < .01; *P < .05.

Figure 3.

KIR2DL2/L3 single+NK cells in the d-rsKIR group (C1Cx-C1Cx) show higher reactivity compared with other groups (dsKIR and rsKIR). The percentage of KIR2DL2/L3 single+ NK cells at day 90 (A) and day 180 (B) after transplantation in the d-rsKIR group (pairs of donor C1Cx and host C1Cx, n = 96), dsKIR group (pairs of donor C1Cx and host C2C2, n = 7), and rsKIR group (pairs of donor C2C2 and host C1Cx, n = 10). Secretion of IFN-γ (C) of KIR2DL2/L3 single+ NK cells against K562 cells at day 90 after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. MFI expression of DNAM-1 on KIR2DL2/L3 single+ NK cells at day 90 (D) and 180 (E) after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. Heatmap of average NK-cell marker expression on single KIR2DL2/L3+ NK cells from each group at different time points. Sample origin color-coding according to the color key in panel F. In order to adjust all the data to the same order of magnitude, we process the data as follows: the MFI of DNAM-1 and Bcl-2 was divided by 100, the MFI of NKP30, NKP46, and CD122 was divided by 10, and the expression of CD25 was multiplied by 10.

Figure 4.

KIR2DL1 single+NK cells in the d-rsKIR group (C2Cx-C2Cx) showed higher reactivity compared with other groups (nsKIR, dsKIR, and rsKIR). The percentage of KIR2DL1 single+ NK cells at day 90 (A) and day 180 (D) after transplantation in the d-rsKIR (pairs of donor C2Cx and host C2Cx, n = 39), dsKIR (pairs of donor C2Cx and host C1C1, n = 23), rsKIR (pairs of donor C1C1 and host C2Cx, n = 9), and nsKIR groups (pairs of donor C1C1 and host C1C1, n = 43). The expression of CD107a (B,E) and IFN-γ (C,F) of KIR2DL1 single+ NK cells against K562 cells at days 90 and 180 after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. MFI expression of DNAM-1 (G-H) and CD122 (I-J) on KIR2DL1 single+ NK cells at days 90 and 180 after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. Heatmap of average NK-cell marker expression on single KIR2DL1+ NK cells from each group at different time points. Sample origin color-coding according to the color key in panel K. In order to adjust all the data to the same order of magnitude, we processed the data as follows: the MFI of DNAM-1 and Bcl-2 was divided by 100; the MFI of NKP30, NKP46, and CD122 was divided by 10; and the expression of CD25 was multiplied by 10.

Figure 5.

KIR3DL1 single+NK cells in the d-rsKIR group (Bw4Bwx-Bw4Bwx) show higher reactivity compared with other groups (nsKIR, dsKIR, and rsKIR). The percentage of KIR3DL1 single+ NK cells at day 180 (A) after transplantation in the d-rsKIR (pairs of donor Bw4Bwx and host Bw4Bwx, n = 55), dsKIR (pairs of donor Bw4Bwx and host Bw6Bw6, n = 18), rsKIR (pairs of donor Bw6Bw6 and host Bw4Bwx, n = 16), and nsKIR groups (pairs of donor Bw6Bw6 and host Bw6Bw6, n = 25). The expression of CD107a (B) and IFN-γ (C) of KIR3DL1 single+ NK cells against K562 cells at day 180 after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. MFI expression of DNAM-1 (D) and CD122 (E) on KIR3DL1 single+ NK cells at day 180 after transplantation in the d-rsKIR, dsKIR, and rsKIR groups. Heatmap of average NK-cell marker expression on single KIR3DL1+ NK cells from each group at different time points. Sample origin color-coding according to the color key in panel F. In order to adjust all the data to the same order of magnitude, we processed the data as follows: the MFI of DNAM-1 and Bcl-2 was divided by 100; the MFI of NKP30, NKP46, and CD122 was divided by 10; and the expression of CD25 was multiplied by 10.

Figure 6.

Both donor and host coexist all HLAs for donor 3 inhibitory KIRs associated with maximum responsiveness recovery of NK cells and least relapse following haplo-SCT. To allow for the evaluation of inhibitory KIRs and minimize interference from potential class I–recognizing activating KIRs and non-self HLA, we analyzed NK responsiveness from 9 donor-patients pairs of both of donor and host presenting all HLAs (Bw4C1C2) for haplotype A donor KIR presenting KIR3DL1, KIR2DL3, KIR2DL1 at the same time. The hierarchy of expression of CD107a (A) and IFN-γ (B) against K562 cells of NK subsets with expression of 3 (NKG2A−KIR2DL1+KIR2DL3+KIR3DL1+ NK), 2 (NKG2A−KIR2DL1+KIR2DL2/L3+ NK, NKG2A−KIR2DL1+KIR3DL1+ NK, or NKG2A−KIR2DL3+KIR3DL1+ NK), 1 (NKG2A−KIR2DL1+ NK, NKG2A-KIR2DL3+ NK, or NKG2A−KIR3DL1+ NK), or 0 (NKG2A−KIR−) inhibitory KIRs for self-HLA. To evaluate the predictive roles of the interaction of donor and host coexisting HLAs for donor inhibitory KIRs and clinical outcomes, in the second cohort, 276 AML or MDS patients who underwent haplo-SCT were subgrouped into the nsKIR group (n = 156), where both hosts and donors lacked HLA ligands for all 3 donor KIRs (ie, for donor C1C1, C2C2, C1C2, C1C1Bw4, or C2C2Bw4 and host C1C1, C2C2, C1C2, C1C1Bw4, or C2C2Bw4); the d-rsKIR group (n = 31), where donors and hosts encoded all HLA ligands for donor KIRs (ie, donor C1C2Bw4 to host C1C2Bw4); the dsKIR group (n = 33), where donors, but not hosts, encoded all HLA ligands for donor KIR (ie, for donor C1C2Bw4 and host C1C1, C2C2, C1C2, C1C1Bw4, or C2C2Bw4); and the rsKIR group (n = 55), where hosts, but not donors, encoded all HLA ligands for donor KIR (ie, for donor C1C1, C2C2, C1C2, C1C1Bw4, or C2C2Bw4 and host C1C2Bw4). The lowest relapse rate (C) was found in the d-rsKIR group (0%) compared with the rsKIR group (10.0% ± 4.9%, P = .115), dsKIR group (14.9% ± 7.0%, P = .039), or nsKIR group (18% ± 3.5%, P = .022). Therefore, higher LFS (D) and OS (E) were found in the d-rsKIR group compared with the nsKIR group (87.5% ± 5.8% vs 67.2% ± 3.8%, P = .046 and 87.5% ± 5.8% vs 68.1% ± 4.0%, P = .066).