2396
Xiang-Yu Zhao et al.
DOI: 10.1002/eji.201445057
Eur. J. Immunol. 2015. 45: 2396–2408
Recipient expression of ligands for donor inhibitory KIRs enhances NK-cell function to control leukemic relapse after haploidentical transplantation Xiang-Yu Zhao1 , Ying-Jun Chang1 , Xiao-Su Zhao1 , Lan-Ping Xu1 , Xiao-Hui Zhang1 , Kai-Yan Liu1 , Dan Li1 and Xiao-Jun Huang1,2 1
Peking University People’s Hospital, Peking University Institute of Hematology, Beijing Key Laboratory of Hematopoietic Stem Cell Transplantation, Beijing, China 2 Peking-Tsinghua Center for Life Sciences, Beijing, China Natural killer (NK) cells that express self-HLA-specific receptors (where HLA is human leukocyte antigen) are “licensed” and more readily activated than unlicensed cells; therefore, NK-cell licensing could influence the antileukemia effects of NK cells following haploidentical stem cell transplantation (haplo-SCT). In this study, we compared the functionality of reconstituting NK cells, based on CD107α expression and interferon-γsecretion, in a cohort of 29 patients that expressed (n = 8) or lacked (n = 21) class I human leukocyte antigens for donor inhibitory killer cell immunoglobulin-like receptors (KIRs) following T-cell-replete haplo-SCT. We also addressed whether recipient expression of class I ligands for donor inhibitory KIRs could predict relapse occurrence in another cohort of 188 patients. A longitudinal analysis indicated that patients presenting class I for all donor inhibitory KIRs showed more capable functional NK effector cells when tested against class I negative K562 cells and primary leukemic cells within 3 months of transplantation. The lowest 7-year relapse incidence was observed when donor KIRs were ligated by recipient class I (n = 60) compared with donor–host partnerships where donor KIR+ cells were ligated by donor, but not recipient class I (n = 86, p = 0.026) or KIRs that were ligated by neither donor nor recipient class I (n = 42, p = 0.043). This study suggests that haplo-SCT recipients presenting class I for donor inhibitory KIRs promote NK-cell licensing, leading to decreased relapse rates.
Keywords: KIR r Haploidentica
r
HSCT
r
NK cells
r
T-cell replete
r
Leukemia
Additional supporting information may be found in the online version of this article at the publisher’s web-site
Introduction Natural killer (NK) cells, a major cell type of the innate immune system, express surface receptors that mediate potent effector functions, including cytolytic activity and cytokine release, and play a central role in inflammation and immunoregulation
Correspondence: Dr. Xiao-Jun Huang e-mail: [email protected] C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[1]. Alloreactive NK cells have antileukemia effects in the context of human leukocyte antigen (HLA) matched or mismatched hematopoietic stem cell transplantation (HSCT) without leading to graft versus host disease (GVHD) [2–6]. NK cells are the most rapidly reconstituted immune cells following transplantation [7–9]. During maturation, NK cells require the recognition of self-HLA class I molecules to acquire full functionality, a phenomenon referred to as “licensing” or “education” [10–12]. Conflicting clinical results have been reported based on in vitro models
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
of NK-cell alloreactivity involving HSCT outcomes with or without T-cell depletion in leukemia patients [6]. In HLA-matched, T-cell-depleted sibling transplantations, Yu et al. demonstrated that “unlicensed” NK cells with inhibitory killer cell immunoglobulin-like receptors (KIRs) for nonself-HLA are functional and patients lacking class I ligands for donor inhibitory KIRs showed improved posttransplant outcomes [12]. These findings were contested by Bjorklund et al., who demonstrated that unlicensed NK cells remained hyporesponsive following T-cellreplete HLA-matched sibling HSCT; consequently, the presence of inhibitory KIRs for nonself-HLA class I ligands had no effect on relapse frequency [11]. Miller et al. reported that licensed NK cells (for which ligands were present in the donor and recipient) were more functional than unlicensed NK cells following unrelated HSCT [13]. Therefore, the antileukemia effects of NK cells post transplantation are determined by the functional recovery of licensed and unlicensed NK cells in vivo, and may be influenced by the presence of T cells in the allograft. These findings nevertheless imply that NK-cell function against leukemic cells is an important predictor of outcomes following HLA-matched HSCT. In haploidentical HSCT, where donor and recipient HLA are mismatched for up to five alleles, it is not known whether the HLA of donor, recipient or both determine NK-cell function and the outcome of HSCT. Recently, we successfully established a novel haploidentical stem cell transplantation (haplo-SCT) protocol that includes antithymocyte globulin treatment followed by unmanipulated blood and marrow graft infusion without in vitro T-cell depletion [14–16]. As described in our original report of T-cell-replete haplo-SCT, we observed that patients lacking class I ligands for donor KIR ligand or donor-inhibitory or donor-activating KIR gene had higher relapse rate—a result that directly contradicts the findings of the Perugia group [17–19]. The results suggests that the presence of class I ligands for the donor-activating or donor-inhibitory KIR gene in the recipient might confer some protection against leukemic relapse in T-cell-replete haploSCT. In addition to the conflicts between reported clinical correlations, it remains unclear when and where NK cells are exposed to class I HLA molecules to acquire functionality [20]. Murine studies suggest that this may be a dynamic process wherein NK cells could adjust to the class I environment of the recipient following adoptive transfer or HSCT [21, 22]. In HLA-matched HSCT, all NK cells would be licensed in recipients presenting all ligands for donor-inhibitory KIR receptors, however, in haploidentical transplantation, it is less clear whether and how NK cells would be licensed in the absence of KIR ligands in the donor or recipient. We postulated that, in haplo-SCT, recipient presentation of all KIR ligands for donor-inhibitory KIRs might promote the functional education of NK cells. Our data show that recipients expressing KIR ligands for all donor-inhibitory KIRs promote the early functional reconstitution of NK cells and are associated with decreased relapse, suggesting that the recipient HLA determines NK licensing post T-cell-replete haploSCT. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Clinical immunology
Results Patient characteristics Patient characteristics in the first cohort are described in Table 1. All patients showed engraftment and complete donor chimerism following transplantation. In the first cohort, 23 patients survived without leukemia, 3 patients relapsed, and 3 patients died of transplant-related complications. All patients showed CMV reactivation post transplantation. Eight patients experienced at least two instances of CMV reactivation. Only one patient suffered from posttransplantation CMV disease (CMV uveitis). The median time of CMV reactivation post transplantation was 29 days (range, 18– 47 days). The durations of initial CMV clearance post transplantation were 18 days (range, 5–16 days). There were 19 and 10 patients who had grade 0–I aGVHD and II–IV aGVHD, respectively. Moreover, there were 15 patients with mild-to-moderate (six patients) or severe (nine patients) cGVHD. Patient characteristics in the second cohort are described in Table 2, 108 patients survived without leukemia, 44 patients relapsed, and 36 patients died due to transplant-related complications following HSCT. The median follow-up time for the cohort was 7.0 years (range, 3.8– 11.4 years). Three models of KIR and HLA interaction were considered (Fig. 1). Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (i) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIRs; (ii) nonself-KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR; and (iii) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIRs.
KIR reconstitution is modulated by the recipient HLA environment To identify the influences of donor and recipient HLA environments on NK-cell education, we monitored the phenotypes of donor-derived reconstituted NK cells at 15, 30, 60, 100, and 180 days posttransplantation. The percentages of KIR-expressing donor NK cells were significantly higher in donors encoding one or two KIR ligands compared to those encoding three KIR ligands (p = 0.01, Fig. 2A). Following transplantation, however, the percentages of KIR-expressing donor-derived NK cells was similar to pre-transplant values and did not differ among recipients of donor cells encoding one, two or three KIR ligands (Fig. 2A). The percentages of KIR-expressing donor-derived NK cells were significantly higher in recipients encoding one KIR ligand compared to those encoding three KIR ligands at days 30 and 60 post transplantation (p = 0.040 and 0.050, respectively) and to those encoding two KIR ligands at day 60 posttransplantation (p = 0.013, Fig. 2B), suggesting that the posttransplantation reconstitution kinetics of KIR expression in donor-derived NK cells could be influenced by recipient HLA. Overall, the percentage of KIR-expressing donor-derived www.eji-journal.eu
2397
C1C2BW4 C1 C1BW4 C1C2BW4 C1C2BW4 C1BW4 C1 C1C2BW4 C1 C1C2BW4 C1
C1 C1BW4
C1C2BW4 C1BW4 C1C2BW4 C1C2BW4 C1C2BW4 C1C2BW4 C1BW4 C1 C1 C1 C1BW4 C1C2 C1BW4 C1BW4 C1C2BW4 C1BW4
1 2 3 4 5 6 7 8 9 10 11
12 13
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
KIR-AA KIR-BX KIR-AA KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-BX KIR-AA KIR-BX
KIR-AA KIR-AA
KIR-BX KIR-BX KIR-BX KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-BX KIR-BX KIR-AA
Donor KIR gene
C1C2BW4 C1 C1 C2BW4 C1C2 C1BW4 C1 C1 C1 C1C2 C1BW4 C1C2BW4 C1BW4 C1BW4 C1C2BW4 C1BW4
C1C2BW4 C1BW4
C1C2BW4 C1 C1BW4 C1C2BW4 C2 C1BW4 C1 C1C2BW4 C1 C1C2BW4 C1
Recipients KIR ligand
No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes No Yes
No Yes
No Yes Yes No Yes Yes Yes No Yes No Yes
Missing self
No Yes Yes Yes Yes Yes Yes No No No No No No No No No
No No
No No No No Yes No No No No No No
Missing ligand
44 14 28 43 41 12 30 15 25 18 32 40 21 17 41 18
30 25
25 54 16 14 33 23 12 18 25 20 25
Age
CML AML-M2 ETO(+) Ph+ALL AML-M2 AML-M2 AML-M5 MDS-RAEB T-ALL AML-M2 AML-M6 AML-M5 Ph+B-ALL MDS-RAEB AML-M4EO B-ALL CML B-ALL AML-M2
AML-M2 MDS-RAEB Pre-B-ALL AML-M2 MDS-RAEB B-ALL B-ALL ALL-L2 ALL B-ALL AML-M2
Disease
CR2 CR2 CR1 CR1 IPSS 1.5 CR1 CR1 CR1 CR1 CR2 IPSS 2.0 CR1, FLT3(+) CR1 CP2 Ph(+),CR1 CR1
CR1, FLT3(+) IPSS 1.5 CR1 CR1 IPSS 1.5 Ph(+),CR1 CR2 CR2 CR1 CR1 relapse post auto-HSCT CP2 CR1
Risk
3 3 2 2 3 2 3 2 3 3 1 3 2 3 3 2
3 2
3 1 3 3 3 3 3 3 3 2 1
Number of HLA loci mismatch (A/B/DR)
III I I II IV 0 II I II II 0 III 0 0 0 0
II II
II II II 0 I I II 0 0 II 0
Acute GVHD
Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y TRM at 11M Relapse at 7M TRM at 8M Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Relapse at 9M Alive at 2Y Alive at 2Y Alive at 2Y
Alive at 2Y Alive at 2Y
Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y TRM at 9M Alive at 2Y Relapase at 10M Alive at 2Y Alive at 2Y Alive at 2Y
Outcomes
Xiang-Yu Zhao et al.
MDS-RAEB: myelodysplastic syndrome refractory anaemia with an excess of blast, CML: chronic myeloid leukaemia, CP2: chronic phase 2, C1 HLA-C group1, C2 HLA-C group2, BW4 HLA-BW4, TRM: transplant-related mortality, HLA: histocompatibility.
Donor KIR ligand
Transplant Number
Table 1. Patients characteristics for Cohort 1
2398 Eur. J. Immunol. 2015. 45: 2396–2408
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
Clinical immunology
Table 2. Patients characteristics in Cohort 2
Factor
Ds-KIR partnership (n = 42)
Ns-KIR partnership (n = 86)
Rs-KIR partnership (n = 60)
Median age at transplantation, years (range) Male, n (%) Diagnosis, n (%) AML ALL MDS Myeloid disease n (%) High risk disease, n (%) HLA matching, n (%) 3/6 4/6 5/6 6/6 Median infused TNC (BM), ×109 /L Median infused TNC (PB), ×109 /L Median infused CD34+ (BM), ×106 /L Median infused CD34+ (PB), ×106 /L Median infused CD3+ (BM), ×108 /L Median infused CD3+ (PB), ×108 /L Median follow up among survivors, years (range) engraftment WBC, median, days (range) PLT, median, days (range)
23.5 (10–42) 26 (61.9%)
24 (2–50) 44 (51.2%)
27 (12-49) 36 (60.0%)
26 (61.9%) 14 (35.7%) 1 (2.4%) 27 (64.3%) 24 (57.1%)
61 (70.9%) 19 (22.1%) 6 (7.0%) 67 (77.9%) 54 (62.8%)
34 (57.6%) 18 (30.5%) 7 (11.9%) 42 (70%) 38 (63.3%)
18 (42.9%) 17 (40.5%) 7 (16.7%) 0 3.80 (1.37–8.05) 3.55 (1.70–7.16) 0.59 (0.11–3.1) 1.24 (0.29–2.9) 0.17 (0.04–0.36) 1.41 (0.06–2.33) 8.27 (3.84–10.68)
30 (34.9%) 36 (41.9%) 19 (22.1%) 1 (1.2%) 3.74 (0.24–8.16) 3.45 (0.91–11.70) 0.69 (0.06–3.39) 1.2 (0.22–6.45) 0.18 (0.03–2.16) 1.37 (0.25–5.78) 5.77 (3.90–9.72)
28 (46.7%) 21 (35.0%) 11 (18.3%) 0 3.78 (1.5–5.90) 3.79 (2.0–7.01) 0.92 (0.22–2.94) 1.67 (0.33–4.07) 0.2 (0.07–2.13) 1.44 (1.02–4.13) 6.60 (3.87–11.39)
12.5 (10–22) 15.5 (10–151)
14 (10–26) 16 (9–80)
12 (10–25) 15.5 (8–156)
Three models of KIR and HLA interaction were considered. Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (i) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIR; (ii) non-self KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR, and (iii) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIR. MDS-RAEB: myelodysplastic syndrome refractory anaemia with an excess of blast, TNC: total nuclear cell, WBC: white blood cell, PLT: platelet.
NK cells gradually increased following transplantation, although they had not reached donor pretransplant levels by day 180 (p = 0.029). By days 100 and 180 posttransplantation, the percentages of KIR-expressing NK cells in donor-derived NK cells were equivalent among all recipients irrespective of KIR ligands. We observed a significant correlation between frequency of KIR-expressing NK cells in donor NK cells and in donor-derived NK cells at day 180 posttransplantation (p = 0.027, Fig. 2C).
The expansion of NKG2C+ NK cells is modulated early by the recipient HLA environment All recipients in this study experienced posttransplant CMV reactivation, thus, we evaluated NKG2C expression on NK cells posttransplantation. NKG2C has been previously demonstrated to correlate with CMV reactivation, memory NK formation [23, 24], and terminal maturation of NK cells [25–27], hence, we hypothesized that CMV reactivation would be help to promote the NK-cell reconstitution post T-cell-replete haplo-SCT. We observed increased expansion of NKG2C+ NK cells within 180 days posttransplantation. Donor-derived NKG2C+ NK cells were consistently lower in recipients until day 180 posttransplan C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tation compared with donors (p < 0.001 at days 15 and 30, p = 0.017 at day 60, p = 0.079 at day 100, and p = 0.737 at day 180; Fig. 2D). The percentages of NKG2C-expressing NK cells in donorderived NK cells were significantly higher in recipients encoding one or two KIR ligand compared to those encoding three KIR ligands by days 60 and 100 posttransplantation (p = 0.064 and 0.064, and p = 0.077 and 0.077, respectively, Fig. 2D). Combined, recipients who lacked KIR ligands for donor inhibitory KIRs (dsKIR plus ns-KIR partnerships) showed greater NKG2C+ NK-cell expansion by days 60 and 100 posttransplantation (p = 0.042 and p = 0.042, respectively, Fig. 2D), suggesting that the posttransplantation expansion of NKG2C+ NK cells in donor-derived NK cells could be influenced by recipient HLA. However, the reconstitution of KIR+ NKG2C+ NK cells was comparable between ds-KIR plus ns-KIR partnerships and rs-KIR partnership (Fig. 2E).
NK-cell differentiation correlates with posttransplant licensing but is not affected by recipient or donor HLA CD57+ NK cells have been shown to be highly mature and could represent terminally differentiated NK cells: KIR+ CD57+ and KIR− CD57− phenotypes represent terminally differentiated NK www.eji-journal.eu
2399
2400
Xiang-Yu Zhao et al.
Eur. J. Immunol. 2015. 45: 2396–2408
Figure 1. Donor KIR and recipient HLA interaction pattern in the setting of haploidentical transplantation. Three models of KIR and HLA interaction were considered in this study. Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (A) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIRs; (B) nonself KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR; and (C) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIRs.
cells and NK cells at early developmental stages, respectively [28]. The percentage of donor-derived CD57-expressing NK cells and KIR-expressing NK cells were significantly positively correlated until day 180 posttransplantation (Fig. 3A–E), and the percentage of donor-derived CD57+ NK cells increased gradually during the first 180 days posttransplantation in all recipients. The ratio of CD57+ to CD57− NK cells increased gradually, with donor-derived NK cells in recipients eventually reaching the level of the donors by day 180 posttransplantation (p = 0.068, Fig. 3F). An inverse correlation between the ratio of KIR− CD57− and KIR+ CD57+ NK cells was observed in donors and recipients before and after transplantation up to day 180 posttransplantation (p < 0.001). No differences in the CD57+ to CD57− ratios of donor-derived total NK cells, NKG2C+ NK cells, or KIR+ NK cells were detected among recipients posttransplantation with different recipient and donor KIR ligands (data not shown). Therefore, these data suggest that NK-cell differentiation occurs in parallel with NK-cell education but is not influenced by the recipient or donor HLA environment.
Recipient HLA modulates NK-cell functionality early in the posttransplantation period To characterize the impact of the recipient or donor KIR ligand number on NK-cell functional reconstitution, we monitored the recovery of NK-cell cytotoxicity (indicated by CD107α expression) and cytokine secretion (indicated by IFN-γ production) at days 30, 100, and 180 posttransplantation. The percentage of CD107α expression on donor-derived NK cells was similar among accepting donor cells with one, two, or three KIR ligands by 180 days C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
posttransplantation (Fig. 4A). However, the percentage of CD107α expression on donor-derived NK cells was significantly higher in recipients with three KIR ligands compared with those with only one or two KIR ligands at days 30, 100, and 180 posttransplantation (p = 0.002 and 0.004 at day 30, p = 0.050 and 0.058 at day 100, p = 0.110 and 0.017 at day 180, respectively, Fig. 4B). In agreement, the cytotoxicity of donor-derived NK cells against K562 cells was significantly higher in rs-KIR group compared with those in ds-KIR group plus ns-KIR group at days 30, 100, and 180 (p = 0.001, 0.028 and 0.021, respectively, Fig. 4C). We also measured the percentage of CD107α expression on donor-derived KIR+ NK cells at days 100 and 180, which was found to be higher in rs-KIR group at days 100 and 180 compared with those in dsKIR group and ns KIR group (p = 0.032 and 0.079, respectively, Fig. 4D). Additionally, ds-KIR had lower CD107α expression on donor-derived NK cells within the 180 day posttransplantation period (n = 7) compared with rs-KIR or ns-KIR (n = 22, p = 0.020 at day 30, p = 0.185 at day 100, and p = 0.083 at day 180). As expected, NK cells from rs-KIR (n = 8) demonstrated the highest NK-cell cytotoxicity (versus ns-KIR, n = 14 and versus dsKIR, n = 7, p = 0.001 at day 30, p = 0.082 at day 100, and p = 0.092 at day 180, Fig. 4E). In addition, we tested the activity of reconstituted NK cells at day 30 posttransplantation against primary leukemia cells with the same KIR-ligand typing and same disease category (myeloid leukemia or lymphoid leukemia) as the respective patients, and the total or KIR+ rs-KIR NK cells remained more capable effectors compared with ns-KIR cells or ds-KIR cells (p = 0.090 for total NK cells and 0.021 for KIR+ NK cells, respectively, Fig. 4F, Supporting Information Table 1). Similar trends were observed for the recovery of IFN-γ secretion by donor-derived NK cells. The percentage of IFN-γ expression
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
Clinical immunology
Figure 2. KIR reconstitution and NKG2C+ NK-cell expansion is modulated by the recipient HLA environment. NK cells were isolated from donors and recipients and KIR expression was measured by staining and flow cytometry. (A–C) The percentage of KIR-expressing NK cells were compared among groups, defined by the number of KIR ligands in (A) donor or (B) recipient cells within donor-derived NK cells at different time points posttransplantation, or in (A) donors (D) or (B) recipients (R) before transplantation. (C) The percentage of KIR+ NK cells in donor-derived NK cells at 180 days positively correlated with the percentage of KIR+ NK cells in donor cells. (D) The percentage of NKG2C-expressing NK cells were also compared among groups, defined by the number of KIR ligands in recipients within donor-derived NK cells, at different time points posttransplantation. (E) The percentage of KIR+ NKG2C+ NK cells on donor-derived NK cells from patients who either lacked (ds-KIR plus ns-KIR partnerships) or presented (rs-KIR partnership) KIR ligands for donor inhibitory KIRs, at different time points posttransplantation, and compared with that in their donors (donors) was measured by flow cytometry. (F) The phenotypic analyses and gating strategies used in these experiments. Each symbol represents an individual donor or recipient sample and data are shown as mean ± SEM. NK cells were determined to be KIR+ if they expressed at least one of the following KIRs: CD158a/h (KIR2DL1, KIR2DS1), CD158b/j (KIR2DL2, KIR2DL3, KIR2DS2), or CD158e (KIR3DL1). Statistical significance determined by Wilcoxon rank-sum test; * p < 0.05.
on donor-derived NK cell was significantly higher in rs-KIR group by day 30 compared to ds-KIR plus ns-KIR group (Fig. 4G, p = 0.025), as well as in rs-KIR plus ns-KIR group by days 30 and 100 compared to ds-KIR group (Fig. 4H, p = 0.032 and 0.047). Therefore, NK cells from rs-KIR (n = 8) demonstrated the highest NK cell IFN-γ secretion at day 30 posttransplantation (versus ns-KIR, n = 14 and versus ds-KIR, n = 7, p = 0.028, Fig. 4I). The recovery of CD56bright , CD56dim , CD3+ T cells, CD4+ T cells, CD8+ T cells, and CD4+ CD25high CD127−/low regulatory T (Treg) cells were comparable among the three groups (data not shown). Furthermore, when we examined the percentage of NKp30 expressing on NK cells, donor-derived NK cells in ds-KIR and ns-KIR had higher percentages of NKp30 expressing on NK cells compared with donor-derived NK cells from rs-KIR (data not shown), however the former subset showed reduced percent C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
age of CD107α and IFN-γ expression on NK cells. Therefore, the enhanced responsiveness of donor-derived NK cells in recipients of rs-KIR was not due to the expression of activating or coactivating receptors, suggesting that these differences are caused by intrinsic differences downstream of these receptors.
Donor-recipient rs-KIR pairings are associated with better clinical outcomes We analyzed data collected prior to November 1, 2013. There were no significant differences in recipient age, diagnosis, pretransplantation risk category, HLA incompatibility, or dose of CD3+ T cells and CD34+ cells administered among the three groups www.eji-journal.eu
2401
2402
Xiang-Yu Zhao et al.
Eur. J. Immunol. 2015. 45: 2396–2408
Figure 3. NK-cell differentiation is correlated with education posttransplantation. CD57 expression on donor-derived NK cells was evaluated by surface staining and flow cytometry. (A–E) The percentage of CD57-expressing NK cells was positively correlated with KIR expression on donorderived NK cells at days (A) 15, (B) 30, (C) 60, (D) 100, and (E) 180 posttransplantation. Each symbol represents an individual a donor–recipient pair. (F) The ratios of CD57+ to CD57– NK cells during reconstitution were compared between patients and donors at different times points posttransplantation using the Wilcoxon signed-rank test. Bars indicate the mean ± SEM. * p
Xiang-Yu Zhao et al.
DOI: 10.1002/eji.201445057
Eur. J. Immunol. 2015. 45: 2396–2408
Recipient expression of ligands for donor inhibitory KIRs enhances NK-cell function to control leukemic relapse after haploidentical transplantation Xiang-Yu Zhao1 , Ying-Jun Chang1 , Xiao-Su Zhao1 , Lan-Ping Xu1 , Xiao-Hui Zhang1 , Kai-Yan Liu1 , Dan Li1 and Xiao-Jun Huang1,2 1
Peking University People’s Hospital, Peking University Institute of Hematology, Beijing Key Laboratory of Hematopoietic Stem Cell Transplantation, Beijing, China 2 Peking-Tsinghua Center for Life Sciences, Beijing, China Natural killer (NK) cells that express self-HLA-specific receptors (where HLA is human leukocyte antigen) are “licensed” and more readily activated than unlicensed cells; therefore, NK-cell licensing could influence the antileukemia effects of NK cells following haploidentical stem cell transplantation (haplo-SCT). In this study, we compared the functionality of reconstituting NK cells, based on CD107α expression and interferon-γsecretion, in a cohort of 29 patients that expressed (n = 8) or lacked (n = 21) class I human leukocyte antigens for donor inhibitory killer cell immunoglobulin-like receptors (KIRs) following T-cell-replete haplo-SCT. We also addressed whether recipient expression of class I ligands for donor inhibitory KIRs could predict relapse occurrence in another cohort of 188 patients. A longitudinal analysis indicated that patients presenting class I for all donor inhibitory KIRs showed more capable functional NK effector cells when tested against class I negative K562 cells and primary leukemic cells within 3 months of transplantation. The lowest 7-year relapse incidence was observed when donor KIRs were ligated by recipient class I (n = 60) compared with donor–host partnerships where donor KIR+ cells were ligated by donor, but not recipient class I (n = 86, p = 0.026) or KIRs that were ligated by neither donor nor recipient class I (n = 42, p = 0.043). This study suggests that haplo-SCT recipients presenting class I for donor inhibitory KIRs promote NK-cell licensing, leading to decreased relapse rates.
Keywords: KIR r Haploidentica
r
HSCT
r
NK cells
r
T-cell replete
r
Leukemia
Additional supporting information may be found in the online version of this article at the publisher’s web-site
Introduction Natural killer (NK) cells, a major cell type of the innate immune system, express surface receptors that mediate potent effector functions, including cytolytic activity and cytokine release, and play a central role in inflammation and immunoregulation
Correspondence: Dr. Xiao-Jun Huang e-mail: [email protected] C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[1]. Alloreactive NK cells have antileukemia effects in the context of human leukocyte antigen (HLA) matched or mismatched hematopoietic stem cell transplantation (HSCT) without leading to graft versus host disease (GVHD) [2–6]. NK cells are the most rapidly reconstituted immune cells following transplantation [7–9]. During maturation, NK cells require the recognition of self-HLA class I molecules to acquire full functionality, a phenomenon referred to as “licensing” or “education” [10–12]. Conflicting clinical results have been reported based on in vitro models
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
of NK-cell alloreactivity involving HSCT outcomes with or without T-cell depletion in leukemia patients [6]. In HLA-matched, T-cell-depleted sibling transplantations, Yu et al. demonstrated that “unlicensed” NK cells with inhibitory killer cell immunoglobulin-like receptors (KIRs) for nonself-HLA are functional and patients lacking class I ligands for donor inhibitory KIRs showed improved posttransplant outcomes [12]. These findings were contested by Bjorklund et al., who demonstrated that unlicensed NK cells remained hyporesponsive following T-cellreplete HLA-matched sibling HSCT; consequently, the presence of inhibitory KIRs for nonself-HLA class I ligands had no effect on relapse frequency [11]. Miller et al. reported that licensed NK cells (for which ligands were present in the donor and recipient) were more functional than unlicensed NK cells following unrelated HSCT [13]. Therefore, the antileukemia effects of NK cells post transplantation are determined by the functional recovery of licensed and unlicensed NK cells in vivo, and may be influenced by the presence of T cells in the allograft. These findings nevertheless imply that NK-cell function against leukemic cells is an important predictor of outcomes following HLA-matched HSCT. In haploidentical HSCT, where donor and recipient HLA are mismatched for up to five alleles, it is not known whether the HLA of donor, recipient or both determine NK-cell function and the outcome of HSCT. Recently, we successfully established a novel haploidentical stem cell transplantation (haplo-SCT) protocol that includes antithymocyte globulin treatment followed by unmanipulated blood and marrow graft infusion without in vitro T-cell depletion [14–16]. As described in our original report of T-cell-replete haplo-SCT, we observed that patients lacking class I ligands for donor KIR ligand or donor-inhibitory or donor-activating KIR gene had higher relapse rate—a result that directly contradicts the findings of the Perugia group [17–19]. The results suggests that the presence of class I ligands for the donor-activating or donor-inhibitory KIR gene in the recipient might confer some protection against leukemic relapse in T-cell-replete haploSCT. In addition to the conflicts between reported clinical correlations, it remains unclear when and where NK cells are exposed to class I HLA molecules to acquire functionality [20]. Murine studies suggest that this may be a dynamic process wherein NK cells could adjust to the class I environment of the recipient following adoptive transfer or HSCT [21, 22]. In HLA-matched HSCT, all NK cells would be licensed in recipients presenting all ligands for donor-inhibitory KIR receptors, however, in haploidentical transplantation, it is less clear whether and how NK cells would be licensed in the absence of KIR ligands in the donor or recipient. We postulated that, in haplo-SCT, recipient presentation of all KIR ligands for donor-inhibitory KIRs might promote the functional education of NK cells. Our data show that recipients expressing KIR ligands for all donor-inhibitory KIRs promote the early functional reconstitution of NK cells and are associated with decreased relapse, suggesting that the recipient HLA determines NK licensing post T-cell-replete haploSCT. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Clinical immunology
Results Patient characteristics Patient characteristics in the first cohort are described in Table 1. All patients showed engraftment and complete donor chimerism following transplantation. In the first cohort, 23 patients survived without leukemia, 3 patients relapsed, and 3 patients died of transplant-related complications. All patients showed CMV reactivation post transplantation. Eight patients experienced at least two instances of CMV reactivation. Only one patient suffered from posttransplantation CMV disease (CMV uveitis). The median time of CMV reactivation post transplantation was 29 days (range, 18– 47 days). The durations of initial CMV clearance post transplantation were 18 days (range, 5–16 days). There were 19 and 10 patients who had grade 0–I aGVHD and II–IV aGVHD, respectively. Moreover, there were 15 patients with mild-to-moderate (six patients) or severe (nine patients) cGVHD. Patient characteristics in the second cohort are described in Table 2, 108 patients survived without leukemia, 44 patients relapsed, and 36 patients died due to transplant-related complications following HSCT. The median follow-up time for the cohort was 7.0 years (range, 3.8– 11.4 years). Three models of KIR and HLA interaction were considered (Fig. 1). Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (i) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIRs; (ii) nonself-KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR; and (iii) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIRs.
KIR reconstitution is modulated by the recipient HLA environment To identify the influences of donor and recipient HLA environments on NK-cell education, we monitored the phenotypes of donor-derived reconstituted NK cells at 15, 30, 60, 100, and 180 days posttransplantation. The percentages of KIR-expressing donor NK cells were significantly higher in donors encoding one or two KIR ligands compared to those encoding three KIR ligands (p = 0.01, Fig. 2A). Following transplantation, however, the percentages of KIR-expressing donor-derived NK cells was similar to pre-transplant values and did not differ among recipients of donor cells encoding one, two or three KIR ligands (Fig. 2A). The percentages of KIR-expressing donor-derived NK cells were significantly higher in recipients encoding one KIR ligand compared to those encoding three KIR ligands at days 30 and 60 post transplantation (p = 0.040 and 0.050, respectively) and to those encoding two KIR ligands at day 60 posttransplantation (p = 0.013, Fig. 2B), suggesting that the posttransplantation reconstitution kinetics of KIR expression in donor-derived NK cells could be influenced by recipient HLA. Overall, the percentage of KIR-expressing donor-derived www.eji-journal.eu
2397
C1C2BW4 C1 C1BW4 C1C2BW4 C1C2BW4 C1BW4 C1 C1C2BW4 C1 C1C2BW4 C1
C1 C1BW4
C1C2BW4 C1BW4 C1C2BW4 C1C2BW4 C1C2BW4 C1C2BW4 C1BW4 C1 C1 C1 C1BW4 C1C2 C1BW4 C1BW4 C1C2BW4 C1BW4
1 2 3 4 5 6 7 8 9 10 11
12 13
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
KIR-AA KIR-BX KIR-AA KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-BX KIR-AA KIR-BX
KIR-AA KIR-AA
KIR-BX KIR-BX KIR-BX KIR-AA KIR-AA KIR-BX KIR-AA KIR-AA KIR-BX KIR-BX KIR-AA
Donor KIR gene
C1C2BW4 C1 C1 C2BW4 C1C2 C1BW4 C1 C1 C1 C1C2 C1BW4 C1C2BW4 C1BW4 C1BW4 C1C2BW4 C1BW4
C1C2BW4 C1BW4
C1C2BW4 C1 C1BW4 C1C2BW4 C2 C1BW4 C1 C1C2BW4 C1 C1C2BW4 C1
Recipients KIR ligand
No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes No Yes
No Yes
No Yes Yes No Yes Yes Yes No Yes No Yes
Missing self
No Yes Yes Yes Yes Yes Yes No No No No No No No No No
No No
No No No No Yes No No No No No No
Missing ligand
44 14 28 43 41 12 30 15 25 18 32 40 21 17 41 18
30 25
25 54 16 14 33 23 12 18 25 20 25
Age
CML AML-M2 ETO(+) Ph+ALL AML-M2 AML-M2 AML-M5 MDS-RAEB T-ALL AML-M2 AML-M6 AML-M5 Ph+B-ALL MDS-RAEB AML-M4EO B-ALL CML B-ALL AML-M2
AML-M2 MDS-RAEB Pre-B-ALL AML-M2 MDS-RAEB B-ALL B-ALL ALL-L2 ALL B-ALL AML-M2
Disease
CR2 CR2 CR1 CR1 IPSS 1.5 CR1 CR1 CR1 CR1 CR2 IPSS 2.0 CR1, FLT3(+) CR1 CP2 Ph(+),CR1 CR1
CR1, FLT3(+) IPSS 1.5 CR1 CR1 IPSS 1.5 Ph(+),CR1 CR2 CR2 CR1 CR1 relapse post auto-HSCT CP2 CR1
Risk
3 3 2 2 3 2 3 2 3 3 1 3 2 3 3 2
3 2
3 1 3 3 3 3 3 3 3 2 1
Number of HLA loci mismatch (A/B/DR)
III I I II IV 0 II I II II 0 III 0 0 0 0
II II
II II II 0 I I II 0 0 II 0
Acute GVHD
Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y TRM at 11M Relapse at 7M TRM at 8M Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Relapse at 9M Alive at 2Y Alive at 2Y Alive at 2Y
Alive at 2Y Alive at 2Y
Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y Alive at 2Y TRM at 9M Alive at 2Y Relapase at 10M Alive at 2Y Alive at 2Y Alive at 2Y
Outcomes
Xiang-Yu Zhao et al.
MDS-RAEB: myelodysplastic syndrome refractory anaemia with an excess of blast, CML: chronic myeloid leukaemia, CP2: chronic phase 2, C1 HLA-C group1, C2 HLA-C group2, BW4 HLA-BW4, TRM: transplant-related mortality, HLA: histocompatibility.
Donor KIR ligand
Transplant Number
Table 1. Patients characteristics for Cohort 1
2398 Eur. J. Immunol. 2015. 45: 2396–2408
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
Clinical immunology
Table 2. Patients characteristics in Cohort 2
Factor
Ds-KIR partnership (n = 42)
Ns-KIR partnership (n = 86)
Rs-KIR partnership (n = 60)
Median age at transplantation, years (range) Male, n (%) Diagnosis, n (%) AML ALL MDS Myeloid disease n (%) High risk disease, n (%) HLA matching, n (%) 3/6 4/6 5/6 6/6 Median infused TNC (BM), ×109 /L Median infused TNC (PB), ×109 /L Median infused CD34+ (BM), ×106 /L Median infused CD34+ (PB), ×106 /L Median infused CD3+ (BM), ×108 /L Median infused CD3+ (PB), ×108 /L Median follow up among survivors, years (range) engraftment WBC, median, days (range) PLT, median, days (range)
23.5 (10–42) 26 (61.9%)
24 (2–50) 44 (51.2%)
27 (12-49) 36 (60.0%)
26 (61.9%) 14 (35.7%) 1 (2.4%) 27 (64.3%) 24 (57.1%)
61 (70.9%) 19 (22.1%) 6 (7.0%) 67 (77.9%) 54 (62.8%)
34 (57.6%) 18 (30.5%) 7 (11.9%) 42 (70%) 38 (63.3%)
18 (42.9%) 17 (40.5%) 7 (16.7%) 0 3.80 (1.37–8.05) 3.55 (1.70–7.16) 0.59 (0.11–3.1) 1.24 (0.29–2.9) 0.17 (0.04–0.36) 1.41 (0.06–2.33) 8.27 (3.84–10.68)
30 (34.9%) 36 (41.9%) 19 (22.1%) 1 (1.2%) 3.74 (0.24–8.16) 3.45 (0.91–11.70) 0.69 (0.06–3.39) 1.2 (0.22–6.45) 0.18 (0.03–2.16) 1.37 (0.25–5.78) 5.77 (3.90–9.72)
28 (46.7%) 21 (35.0%) 11 (18.3%) 0 3.78 (1.5–5.90) 3.79 (2.0–7.01) 0.92 (0.22–2.94) 1.67 (0.33–4.07) 0.2 (0.07–2.13) 1.44 (1.02–4.13) 6.60 (3.87–11.39)
12.5 (10–22) 15.5 (10–151)
14 (10–26) 16 (9–80)
12 (10–25) 15.5 (8–156)
Three models of KIR and HLA interaction were considered. Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (i) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIR; (ii) non-self KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR, and (iii) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIR. MDS-RAEB: myelodysplastic syndrome refractory anaemia with an excess of blast, TNC: total nuclear cell, WBC: white blood cell, PLT: platelet.
NK cells gradually increased following transplantation, although they had not reached donor pretransplant levels by day 180 (p = 0.029). By days 100 and 180 posttransplantation, the percentages of KIR-expressing NK cells in donor-derived NK cells were equivalent among all recipients irrespective of KIR ligands. We observed a significant correlation between frequency of KIR-expressing NK cells in donor NK cells and in donor-derived NK cells at day 180 posttransplantation (p = 0.027, Fig. 2C).
The expansion of NKG2C+ NK cells is modulated early by the recipient HLA environment All recipients in this study experienced posttransplant CMV reactivation, thus, we evaluated NKG2C expression on NK cells posttransplantation. NKG2C has been previously demonstrated to correlate with CMV reactivation, memory NK formation [23, 24], and terminal maturation of NK cells [25–27], hence, we hypothesized that CMV reactivation would be help to promote the NK-cell reconstitution post T-cell-replete haplo-SCT. We observed increased expansion of NKG2C+ NK cells within 180 days posttransplantation. Donor-derived NKG2C+ NK cells were consistently lower in recipients until day 180 posttransplan C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tation compared with donors (p < 0.001 at days 15 and 30, p = 0.017 at day 60, p = 0.079 at day 100, and p = 0.737 at day 180; Fig. 2D). The percentages of NKG2C-expressing NK cells in donorderived NK cells were significantly higher in recipients encoding one or two KIR ligand compared to those encoding three KIR ligands by days 60 and 100 posttransplantation (p = 0.064 and 0.064, and p = 0.077 and 0.077, respectively, Fig. 2D). Combined, recipients who lacked KIR ligands for donor inhibitory KIRs (dsKIR plus ns-KIR partnerships) showed greater NKG2C+ NK-cell expansion by days 60 and 100 posttransplantation (p = 0.042 and p = 0.042, respectively, Fig. 2D), suggesting that the posttransplantation expansion of NKG2C+ NK cells in donor-derived NK cells could be influenced by recipient HLA. However, the reconstitution of KIR+ NKG2C+ NK cells was comparable between ds-KIR plus ns-KIR partnerships and rs-KIR partnership (Fig. 2E).
NK-cell differentiation correlates with posttransplant licensing but is not affected by recipient or donor HLA CD57+ NK cells have been shown to be highly mature and could represent terminally differentiated NK cells: KIR+ CD57+ and KIR− CD57− phenotypes represent terminally differentiated NK www.eji-journal.eu
2399
2400
Xiang-Yu Zhao et al.
Eur. J. Immunol. 2015. 45: 2396–2408
Figure 1. Donor KIR and recipient HLA interaction pattern in the setting of haploidentical transplantation. Three models of KIR and HLA interaction were considered in this study. Donor–recipient partnerships were defined by the presence or absence of ligands for donor KIR: (A) donor “self” KIR (ds-KIR), where donors, but not recipients, encoded HLA ligands for all KIRs; (B) nonself KIR (ns-KIR), where both recipients and donors lacked HLA ligands for at least one donor KIR; and (C) recipient “self” KIR (rs-KIR), where the recipient encodes all ligands for donor KIRs.
cells and NK cells at early developmental stages, respectively [28]. The percentage of donor-derived CD57-expressing NK cells and KIR-expressing NK cells were significantly positively correlated until day 180 posttransplantation (Fig. 3A–E), and the percentage of donor-derived CD57+ NK cells increased gradually during the first 180 days posttransplantation in all recipients. The ratio of CD57+ to CD57− NK cells increased gradually, with donor-derived NK cells in recipients eventually reaching the level of the donors by day 180 posttransplantation (p = 0.068, Fig. 3F). An inverse correlation between the ratio of KIR− CD57− and KIR+ CD57+ NK cells was observed in donors and recipients before and after transplantation up to day 180 posttransplantation (p < 0.001). No differences in the CD57+ to CD57− ratios of donor-derived total NK cells, NKG2C+ NK cells, or KIR+ NK cells were detected among recipients posttransplantation with different recipient and donor KIR ligands (data not shown). Therefore, these data suggest that NK-cell differentiation occurs in parallel with NK-cell education but is not influenced by the recipient or donor HLA environment.
Recipient HLA modulates NK-cell functionality early in the posttransplantation period To characterize the impact of the recipient or donor KIR ligand number on NK-cell functional reconstitution, we monitored the recovery of NK-cell cytotoxicity (indicated by CD107α expression) and cytokine secretion (indicated by IFN-γ production) at days 30, 100, and 180 posttransplantation. The percentage of CD107α expression on donor-derived NK cells was similar among accepting donor cells with one, two, or three KIR ligands by 180 days C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
posttransplantation (Fig. 4A). However, the percentage of CD107α expression on donor-derived NK cells was significantly higher in recipients with three KIR ligands compared with those with only one or two KIR ligands at days 30, 100, and 180 posttransplantation (p = 0.002 and 0.004 at day 30, p = 0.050 and 0.058 at day 100, p = 0.110 and 0.017 at day 180, respectively, Fig. 4B). In agreement, the cytotoxicity of donor-derived NK cells against K562 cells was significantly higher in rs-KIR group compared with those in ds-KIR group plus ns-KIR group at days 30, 100, and 180 (p = 0.001, 0.028 and 0.021, respectively, Fig. 4C). We also measured the percentage of CD107α expression on donor-derived KIR+ NK cells at days 100 and 180, which was found to be higher in rs-KIR group at days 100 and 180 compared with those in dsKIR group and ns KIR group (p = 0.032 and 0.079, respectively, Fig. 4D). Additionally, ds-KIR had lower CD107α expression on donor-derived NK cells within the 180 day posttransplantation period (n = 7) compared with rs-KIR or ns-KIR (n = 22, p = 0.020 at day 30, p = 0.185 at day 100, and p = 0.083 at day 180). As expected, NK cells from rs-KIR (n = 8) demonstrated the highest NK-cell cytotoxicity (versus ns-KIR, n = 14 and versus dsKIR, n = 7, p = 0.001 at day 30, p = 0.082 at day 100, and p = 0.092 at day 180, Fig. 4E). In addition, we tested the activity of reconstituted NK cells at day 30 posttransplantation against primary leukemia cells with the same KIR-ligand typing and same disease category (myeloid leukemia or lymphoid leukemia) as the respective patients, and the total or KIR+ rs-KIR NK cells remained more capable effectors compared with ns-KIR cells or ds-KIR cells (p = 0.090 for total NK cells and 0.021 for KIR+ NK cells, respectively, Fig. 4F, Supporting Information Table 1). Similar trends were observed for the recovery of IFN-γ secretion by donor-derived NK cells. The percentage of IFN-γ expression
www.eji-journal.eu
Eur. J. Immunol. 2015. 45: 2396–2408
Clinical immunology
Figure 2. KIR reconstitution and NKG2C+ NK-cell expansion is modulated by the recipient HLA environment. NK cells were isolated from donors and recipients and KIR expression was measured by staining and flow cytometry. (A–C) The percentage of KIR-expressing NK cells were compared among groups, defined by the number of KIR ligands in (A) donor or (B) recipient cells within donor-derived NK cells at different time points posttransplantation, or in (A) donors (D) or (B) recipients (R) before transplantation. (C) The percentage of KIR+ NK cells in donor-derived NK cells at 180 days positively correlated with the percentage of KIR+ NK cells in donor cells. (D) The percentage of NKG2C-expressing NK cells were also compared among groups, defined by the number of KIR ligands in recipients within donor-derived NK cells, at different time points posttransplantation. (E) The percentage of KIR+ NKG2C+ NK cells on donor-derived NK cells from patients who either lacked (ds-KIR plus ns-KIR partnerships) or presented (rs-KIR partnership) KIR ligands for donor inhibitory KIRs, at different time points posttransplantation, and compared with that in their donors (donors) was measured by flow cytometry. (F) The phenotypic analyses and gating strategies used in these experiments. Each symbol represents an individual donor or recipient sample and data are shown as mean ± SEM. NK cells were determined to be KIR+ if they expressed at least one of the following KIRs: CD158a/h (KIR2DL1, KIR2DS1), CD158b/j (KIR2DL2, KIR2DL3, KIR2DS2), or CD158e (KIR3DL1). Statistical significance determined by Wilcoxon rank-sum test; * p < 0.05.
on donor-derived NK cell was significantly higher in rs-KIR group by day 30 compared to ds-KIR plus ns-KIR group (Fig. 4G, p = 0.025), as well as in rs-KIR plus ns-KIR group by days 30 and 100 compared to ds-KIR group (Fig. 4H, p = 0.032 and 0.047). Therefore, NK cells from rs-KIR (n = 8) demonstrated the highest NK cell IFN-γ secretion at day 30 posttransplantation (versus ns-KIR, n = 14 and versus ds-KIR, n = 7, p = 0.028, Fig. 4I). The recovery of CD56bright , CD56dim , CD3+ T cells, CD4+ T cells, CD8+ T cells, and CD4+ CD25high CD127−/low regulatory T (Treg) cells were comparable among the three groups (data not shown). Furthermore, when we examined the percentage of NKp30 expressing on NK cells, donor-derived NK cells in ds-KIR and ns-KIR had higher percentages of NKp30 expressing on NK cells compared with donor-derived NK cells from rs-KIR (data not shown), however the former subset showed reduced percent C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
age of CD107α and IFN-γ expression on NK cells. Therefore, the enhanced responsiveness of donor-derived NK cells in recipients of rs-KIR was not due to the expression of activating or coactivating receptors, suggesting that these differences are caused by intrinsic differences downstream of these receptors.
Donor-recipient rs-KIR pairings are associated with better clinical outcomes We analyzed data collected prior to November 1, 2013. There were no significant differences in recipient age, diagnosis, pretransplantation risk category, HLA incompatibility, or dose of CD3+ T cells and CD34+ cells administered among the three groups www.eji-journal.eu
2401
2402
Xiang-Yu Zhao et al.
Eur. J. Immunol. 2015. 45: 2396–2408
Figure 3. NK-cell differentiation is correlated with education posttransplantation. CD57 expression on donor-derived NK cells was evaluated by surface staining and flow cytometry. (A–E) The percentage of CD57-expressing NK cells was positively correlated with KIR expression on donorderived NK cells at days (A) 15, (B) 30, (C) 60, (D) 100, and (E) 180 posttransplantation. Each symbol represents an individual a donor–recipient pair. (F) The ratios of CD57+ to CD57– NK cells during reconstitution were compared between patients and donors at different times points posttransplantation using the Wilcoxon signed-rank test. Bars indicate the mean ± SEM. * p
