- Email: [email protected]
Assessment of physiologic natural killer cell cytotoxicity in vitro
Human Immunology 72 (2011) 1007-1012
Contents lists available at SciVerse ScienceDirect
Assessment of physiologic natural killer cell cytotoxicity in vitro Heidi Duske a, Andreas Sputtek a, Thomas Binder a, Nicolaus Kr×ger b, Sonja Schrepfer c, Thomas Eiermann a,* a
Transfusion Medicine, HLA-Laboratory, University Hospital Hamburg-Eppendorf, Hamburg, Germany Clinic for Stem Cell Transplantation, University Hospital Hamburg-Eppendorf, Hamburg, Germany c Cardiovascular Surgery, Transplant and Stem cell Immunobiology-Laboratory, University Hospital Hamburg-Eppendorf, Hamburg, Germany b
A R T I C L E
I N F O
Article history: Received 26 November 2010 Accepted 4 August 2011 Available online 10 August 2011
Keywords: Natural killer cells Leukemia K562 51 Cr release
A B S T R A C T
Here, we describe an improved 51chromium release assay (CRA) to compare donor natural killer (NK) cell activity. To validate the assay, we analyzed sample preparation, incubation, and cryopreservation of NK cells. The effector-to-target ratio was corrected for the percentage of NK cells. A logarithmic curve was fitted to the data of the CRA for calculation of the maximum activity. The specific lysis was standardized to a reference sample and normalized to the mean specific lysis of the reference. We found that a longer time span involved with both the addition and the removal of DMSO increased the recovery of NK cell activity. Freezing and thawing reduced the cytotoxicity of NK cells but sustained the relative differences that were seen between freshly prepared NK cells. In contrast, medium incubation of thawed cells markedly increased the cytotoxic potential but also deranged these relative differences. Those were widely equalized when cells were stimulated with IL-2. In conclusion, we established a standardized assay with cryopreserved peripheral blood mononuclear cells as an appropriate tool for investigation of individual physiologic NK cell activity. This assay may help to predict donor NK cell activity in vivo, to reconcile conflicting data about NK cells obtained in transplantation studies. 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
1. Introduction The majority of natural killer (NK) cells in the peripheral blood display cytotoxic properties that are crucial in the early defense against virally infected and mutated or stressed cells [1,2]. NK cells are regulated by integrated signaling through a repertoire of germ line– encoded activating and human leukocyte antigen (HLA) class I–specific inhibitory receptors [3,4]. Ljunggren and Karre first published, in the “missing self” hypothesis, that HLA class I downregulation renders infected or mutated cells susceptible to NK cell killing [5]. It is now well established that, during maturation, NK cells require binding between inhibitory receptors and their respective HLA class I molecules to achieve full responsiveness, a process referred to as “licensing” or “education” [6 – 8]. The most important educating receptors belong to the highly diverse killer immunoglobulin-like receptor (KIR) family: KIR2DL2/3 and KIR2DL1, which are specific for HLA-Cw molecules of the C1 (Asp at position 80) and C2 (Lys at position 80) groups, respectively [9,10], and KIR3DL1, which is specific for HLA-Bw4 epitopes present in one-third of HLA-B, and some HLA-A, alleles [11,12]. Furthermore, recent data have revealed an antagonizing role of
* Corresponding author. E-mail address: [email protected] (T.H. Eiermann).
the activating KIR2DS1 in NK cell education only in the absence of its inhibitory counterpart KIR2DL1 [13,14]. KIRs are inherited in haplotype patterns and segregate independently from HLA class I molecules located on different chromosomes. Haplotypes are distinguished as type A, coding for a set of seven distinct KIR genes, including one single activating KIR (aKIR), KIR2DS4. The B haplotypes are diverse and characterized by the presence of at least two and up to four aKIRs [15]. The strength of the cytotoxic response is likely defined through the combination of educating receptors and corresponding ligands. Whether and how the expression of noneducating receptors shape the NK cell response upon ligand binding and affect overall NK cell activity remains elusive. Segregating independently, both HLA class I and KIR genotypes affect the development of the individual functional cytotoxic phenotype mature NK cells display. In the last few years, extensive studies have revealed an important role for NK cells in the outcome of hematopoietic stem cell transplantation (HSCT) in leukemia patients. Being the first lymphocytic cell population to arise in the patient’s peripheral blood after transplantation [16], NK cells display a beneficial graftversus-leukemia (GvL) effect by targeting residual malignant cells in the patient. However, which combination of donor KIR type and recipient HLA class I background has the most promising GvL effect is still discussed. Conflicting data from clinical studies on transplantation outcome regarding KIR-ligand mismatch [17–19] and
0198-8859/11/$32.00 - see front matter 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.humimm.2011.08.006
1008
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
KIR genotype [20 –22] seem to be generated through different transplantation setups [23,24]. Several groups have elucidated the role of different KIR-ligand interactions in NK cell (allo-) reactivity by investigating the cytotoxic function of single NK cell subsets that express no or single KIR receptors [12–14,25– 28]. However, NK cells that arise in patients after bone marrow transplantation will express every possible receptor combination provided by the genotype. Thus, the actual strength of the GvL effect will depend on the overall NK cell population and not individual subpopulations. Our aim in this study was to design an assay for examining the properties of NK cells, derived from numerous individuals, in their physiologic state that is created during the developmental licensing. This should provide a better understanding of the influence that the genetic configuration of KIR, HLA class I, and other regulating molecules have on the overall cytotoxic potential of NK cells. That, in turn may advance the prediction of the strength of the GvL effect in the selection of the most promising stem cell donor for leukemia patients. We investigated the influence of cryopreservation, cell collection and separation as well as stimulation and incubation on the cytotoxic activity of NK cells, and designed an appropriate assay setup.
protocol, the cells were supplemented with 20% DMSO in 10 50-l steps every 30 seconds, extending the addition process to 5 minutes in total. All tubes were immediately placed into a cryo 1⬚C/min freezing container (“Mr. Frosty,” Nalgene) and cooled to ⫺80⬚C overnight before transfer to the gaseous phase of liquid nitrogen. For thawing, the cryotubes were directly transferred from the liquid nitrogen tank to a 37⬚C water bath and completely thawed by continuous shaking. Thereafter cells were immediately cooled on ice and transferred to 14-ml tubes placed in a rack on a cooled thermal pack and washed (all centrifugations at 250 g for 5 minutes at 4⬚C). In the fast removal of cryoprotectant (FRCP) protocol, 10 ml of precooled culture medium was slowly added under continuous shaking within 1 minute, followed by 2 washing steps with 10 ml culture medium added all at once. The slow removal of cryoprotectant (SRCP) protocol was performed according to the following timetable: wash 1: 4 ⫻ 250 l culture medium, 8 ⫻ 1 ml cold culture medium, every 30 seconds (total 6 minutes); wash 2: 10 ⫻ 1 ml culture medium every 20 seconds (total 3.5 minutes); and wash 3: 10 ml cold culture medium at once (total 30 seconds). After washing, the cells were resuspended in an appropriate volume of culture medium to obtain the desired E:T ratio and were directly used for the 51Cr-release assay (CRA). 2.3. HLA typing
2. Subjects and methods 2.1. Cell lines and sample collection The HLA class I– deficient myeloid cell line K562 was cultivated in RPMI 1640 supplemented with 10% fetal calf serum culture medium at 37⬚C in 5% CO2. The identity of the cell line was confirmed by HLA typing (HLA-A*11:01, -B*18:01, -DRB1*03:01). K562 cells were used as the NK cell target in all cytotoxic assays. To exclude reactions of T cells with K562 cells, we affirmed the predicted HLA class I deficiency in K562 cells using FACS analysis (data not shown). Peripheral blood mononuclear cells (PBMCs) were collected from the buffycoats or whole blood of healthy donors by density centrifugation (FicollHypaque, GE Healthcare), (28 HLA-C1/C1 and seven not HLA typed). The cells were either used immediately or cryopreserved (as discussed in next section). Frozen PBMCs were thawed immediately before use or incubated either in culture medium without supplements for different time spans or in culture medium supplemented with 100 U/ml interleukin (IL)–2 (eBioscience) for 16 hours at 37⬚C in 5% CO2 atmosphere at a density of 1 ⫻ 106 cells/ml. NK cell separation was performed by depletion of Non-NK cells using the MACS isolation kit (Order no. 130-092657, Miltenyi Biotec). Freshly isolated PBMC samples were incubated in culture medium for 2–3 hours before NK cell separation. Frozen PBMC samples were thawed, resuspended in culture medium containing 100 U/ml IL-2, and incubated overnight at 37⬚C and 5% CO2. The next day, NK cells were separated and resuspended in culture medium containing 100 U/ml IL-2. Membrane integrity was tested by trypan blue exclusion referred to percentage of viable cells. 2.2. Cell storage and preparation For freezing, freshly prepared PBMCs were diluted to a concentration of 40 ⫻ 106 cells/ml in RPMI-1640 supplemented with 50% fetal calf serum. Cell suspension aliquots (500 l) were transferred to 2-ml cryotubes (Greiner Bio-One Frickenhausen, Germany) and cooled on ice. Two timelines for addition and removal of cryoprotectant, respectively, have been tested: In the fast addition of cryoprotectant (FACP) protocol, an equal volume of cold 20% dimethyl sulfoxide (DMSO, purity ⱖ99.7%; Order no. 1.09678.0100 Merck KGaA, Darmstadt, Germany) was added to the cells dropwise within 30 seconds. In the slow addition of cryoprotectant (SACP)
Genomic DNA was isolated from 5 ml of donor buffycoat using the NucleoSpin Blood XL kit (Macherey-Nagel, Dueren, Germany). The donor HLA-A, -B, and -C loci were typed using the reverse SSO line blot assay (Dynal Reli SSO, Invitrogen) 2.4. Cytolytic assays Cytolytic activity of NK cells was measured in a standard 4-hour CRA. One million K562 target cells were stained for 2 hours with 80 Ci 51Cr at 37⬚C, 5% CO2. PBMCs were added to 5000 target cells at triplicate effector to target (E:T) ratios of 10:1, 20:1, 40:1, and 80:1 in a final volume of 200 l in U-bottom 96-well plates. After 4-hour incubation at 37⬚C, 5% CO2, a 100-l quantity of supernatant was transferred to fresh tubes and the counts per minute (cpm.) measured in a gamma counter (PerkinElmer/Wallac). Cytolytic activity was calculated according to the following formula: specific lysis [ % ] ⫽
cpm test ⫺ cpm spontaneous release cpm max release ⫺ cpm spontaneous release
⫻ 100
To compare the cytotoxicity of the NK cells from different donors, the percentage of NK cells in each sample was used for the calculation of the actual effector-to-target ratio. A logarithmic curve was fitted to the data for calculation of the maximum cytolytic activity mediated by each donor cell sample (y ⫽ a ln(x) ⫹ b). Frozen/ thawed PBMCs from the same donor were used as a reference sample in every test. Each sample has been tested in two or three independent experiments. 2.5. Flow-cytometric assay and antibodies All FACS analyses were performed with 5 ⫻ 105 cells using a FACScan cytometer (Becton Dickinson). Data were analyzed with the CellQuestPro software. PBMCs were stained with the mouse mAb anti-human CD3-FITC clone OKT3 (IgG2a; eBioscience) and either anti-human CD56-PE clone B159 (IgG1; BD Biosciences) or the isotype control mouse IgG1-PE (eBioscience) for 30 minutes at 4⬚C. The samples were washed once and kept on ice in the dark until the measurement. To ensure HLA class I negativity, K562 cells were stained with anti-human MHC-ABC-PE clone DX17 (IgG1; BD) and anti-human MHC-DR ⫹ DP ⫹ DQ-PE clone CR3143 (IgG2a; Abcam).
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
1009
2.6. Statistical analysis Significance was tested with the two-tailed t test for paired samples. The level of significance was specified as p ⬍ 0.05 (*), ⬍ 0.01 (**), ⬍ 0.001 (***) or ⬎ 0.05 (NS). 3. Results 3.1. Effect of addition and removal of cryoprotectant Cryopreservation of cell samples considerably facilitates the collection and standardized testing of cells, especially when a large number of donors is needed to address the question. Because NK cells are especially sensitive to freezing and thawing, we optimized the protocols for best recovery of cell function. By comparing the cytolytic activity of PBMC samples treated according to the “fast” or “slow” addition/removal protocols, we found that a longer time span involved with both the addition and the removal of DMSO increased the recovery of NK cell activity (Figs. 1A, 1B; p ⫽ 0.0006 and p ⫽ 0.003). The membrane integrity of the cells was not affected by slow or fast addition of DMSO but slightly decreased after fast removal compared with slow removal of DMSO (Figs. 1C, 1D; NS and p ⫽ 0.01). These findings indicate that a determination of viability alone is not sufficient for deciding if NK cells are still responsive to a stimulus. A careful cryopreservation treatment, in terms of slowly adding and removing of DMSO, is necessary for NK cells to maintain the highest cell function. For all of the following experiments, the slow addition and removal protocols were used.
Fig. 2. Extrapolation of data allows a comparison of the cytotoxicity of various donor PBMC samples. The E:T ratio was corrected for the percentage of NK cells of PBMC and a logarithmic extrapolation curve was applied to each sample (dashed lines). A reference PBMC sample was included in each 51chromium release assay (CRA). To correct for interassay variations, the specific lysis was standardized to this reference sample and normalized to the mean specific lysis of the reference sample (29% ⫾ 7%) obtained in 15 experiments. The percentages of specific lysis of all samples were compared at the extrapolated E:T ratio 15:1, where all curves had reached the plateau phase.
3.2. Extrapolating data: Putting NK cell percentages into perspective One problem with comparing the cytotoxic activities of NK cells from various donors is the individually changing NK cell proportion, which ranges from 1% to 10% of PBMCs. To overcome this difficulty, we developed a strategy that takes into account NK cell percentage in the analysis of data obtained in CRAs. First, the E:T ratios were corrected for the actual number of NK cells in each well as determined by cell counting and FACS analysis. A logarithmic regression curve was fitted to the values obtained for the specific lysis at different E:T ratios. Using this curve, the values for the specific lysis of K-562 were normalized to an actual E:T ratio of 15:1. This ratio was chosen because it was likely that the majority of tested samples would have reached the plateau (e.g., maximum) phase of lysis (Fig. 2). Second, in each assay, a reference sample was included as a standard set to 100% to eliminate interassay variations. Finally, the results were normalized to the mean specific lysis obtained for the reference sample (29% ⫾7%) in 15 separate experiments. 3.3. Individual cytotoxic activity of NK cells is not altered by cryopreservation We next examined the impact of cryopreservation on the effector properties of NK cells derived from different individuals. PBMCs from 9 healthy donors were prepared and either immediately used in CRA or cryopreserved. One week later, the frozen samples were thawed and immediately tested by CRA. As expected, the cytotoxic activity was significantly reduced after freez-/thawing the PBMC samples (mean recovery 25% ⫾ 2%). However, the relative cytolytic potential of the individual donor NK cells was not affected (Fig. 3). Thus, cryopreservation seems to be an appropriate tool for investigation of individual cytotoxic NK cell properties. 3.4. Incubation of thawed PBMCs helps to restore NK cell cytotoxicity but shifts individual differences
Fig. 1. Slow addition and slow removal of DMSO helps to preserve cytotoxic activity of NK cells. Effect of slow and fast DMSO addition and removal protocols for freshly prepared peripheral blood mononuclear cells (PBMCs) on cell viability and recovery of NK cell cytotoxicity was determined by trypan blue exclusion and standard 51 chromium release assay (CRA), respectively. (A, C) Samples were frozen according to either the slow or fast addition of DMSO protocol and were thawed using slow removal protocols. (B, D) All samples were frozen according to the slow addition of DMSO protocol and were thawed using either the slow or fast removal of DMSO protocol.
To determine whether the reduced NK cell activity of frozen samples could be restored, we tested thawed PBMC samples that were incubated in culture medium for 0, 3, 16, or 24 hours in a CRA. The measured cytotoxic activity markedly increased with incubation time; however, as Fig. 4 shows, this increase was not consistent among the six donors. Although a 3-hour incubation could either increase or decrease the basal activity (factor 3.0 – 0.8), the increase
1010
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
Fig. 3. Cryopreservation uniformly reduces the cytotoxicity of NK cells but does not change their capacity for specific lysis. Cytotoxicity of fresh and cryopreserved PBMCs from seven randomly selected healthy donors was measured in a standard 51 chromium release assay (CRA) with K562 target cells. Mean recovery of NK cell cytotoxicity after cryopreservation was 25% ⫾ 2%. All values were standardized and normalized to an E:T-ratio of 15:1 as shown in Fig. 2.
after 24 hours was more pronounced the lower the basal activity of the individual sample (factor 1.0 –11.2). Thus, incubation of thawed PBMCs shifted the donor specific activity of NK cells relatively to each other and therefore cannot be used to compare the original individual cytotoxic NK cell activity. 3.5. Individual NK cell activity can be measured with unstimulated PBMCs The NK cell cytolytic abilities are sensitive to environmental signals like cytokines, antibodies or cell– cell contacts. In vivo, cytotoxic CD56dim NK cells reside in the peripheral blood together with T- and B-lymphocytes, Natural Killer T cells and monocytes. Thus, we investigated if cytolytic activity of NK cells can be specifically measured without separating them from PBMC. We also examined the influence of non-NK cell subsets in PBMC samples by comparing the cytolytic activity of PBMCs and separated NK cells. The cytotoxic activity of freshly prepared PBMCs, either from buffycoat or whole-blood samples, and subsequently separated NK cells was compared in a standard chromium-release
Fig. 4. Incubation of thawed PBMCs restores the cytotoxicity of NK cells nonuniformly. The cytotoxicity of cryopreserved PBMCs from seven randomly selected healthy donors was measured in a standard CRA with K562 target cells after incubation in culture medium for 0, 3, 16, or 24 hours, respectively. Within 24 hours of incubation, the NK cell activity increased in a wide range between 1.0- and 11.2-fold. All values were standardized and normalized to an E:T-ratio of 15:1.
Fig. 5. Cytotoxic activity of PBMC and separated NK cell samples is correlated to each other. (A) Percentage of CD3⫺CD56⫹ NK cells was determined by FACS analysis before and after NK cell separation and is shown as percentage of lymphocytes. (B) Fresh, PBMCs derived from the buffycoat (solid bars) or whole blood (ruled bars) of six random healthy donors and subsequently separated NK cells were measured in a standard 51chromium release assay (CRA) with K562 target cells. All cytolytic values were standardized and normalized to an E:T-ratio of 15:1 as shown in Fig. 2.
assay (Fig. 5). The percentage of NK cells in the PBMC samples ranged between 8% and 40% of lymphocytes and could be enhanced up to 87–95% by negative selection of CD3⫺CD56⫹ NK cells (Fig. 5A). Despite the broad range of NK cell percentages in the PBMC samples, there was a clear trend toward correlation between the cytotoxic activity of whole PBMC and separated NK cells, with a slightly lower activity of PBMC, thus preserving the relative difference of the cytotoxic potential between the individuals. The difference between the cytotoxic activity displayed by PBMC or NK was slightly higher when samples were derived from whole blood (factor 1.3 ⫾ 0.1; ruled bars in Fig. 5B) rather than buffycoat (factor 1.1 ⫾ 0.05; solid bars in Fig. 5B). The data show that the established normalization renders either fresh PBMCs or separated NK cells equally valid samples for the determination of individual NK cell activity. To test the suitability of NK cells separated from thawed PBMC samples, we added IL-2 to the cell suspension, as NK cells separated from frozen PBMCs essentially need interleukin supplementation for survival. Cell samples from 10 healthy donors with homozygous HLA-C1 were tested in a CRA. As expected, the mean specific lysis was significantly higher for IL-2–stimulated PBMCs than for native PBMCs (mean 78% ⫾ 8% vs 13% ⫾ 7%; p ⬍ 10⫺4), whereas the difference between stimulated PBMCs and separated NK cells was comparatively low (78% ⫾ 8% vs 84% ⫾ 8%; Fig. 6). The differential cytolytic potential of native PBMCs was widely equalized after IL-2 stimulation, suggesting that the preactivation overrides the physiologic NK cell state. No significant difference was found between the diverse lytic activity of unstimulated fresh or stimulated thawed PBMCs and
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
Fig. 6. IL-2 stimulation overrides physiologic cytotoxic potential of NK cells. Differential cytotoxic potential of cryopreserved samples from 10 randomly selected healthy donors was compared (native PBMC, IL-2–stimulated PBMC, or separated NK cells stimulated with IL-2). Error bars indicate SEM of two or three repeated measurements. Cytolytic activity of cells was determined in a standardized 51chromium release assay (CRA). Values have been normalized, as shown in Fig. 2.
subsequently separated NK cells, implying that no other cell population, such as T or Natural Killer T cells, in the PMBC mix had an effect on the specific lysis of K562. This finding was affirmed through a significant correlation between the specific lysis and the percentage of CD56⫹CD3- NK cells in thawed unstimulated PBMC samples of 28 healthy donors (r ⫽ 0.77, p ⬍ 10⫺4; Fig. 7). All donors were homozygous for HLA-C1 to exclude licensing differences of NK cells and thereby minimize side effects on cytotoxic potential. No correlation was seen for any of the other cell subsets (data not shown). Taken together, the results indicate that, in PBMC samples, only NK cells contribute to the cytotoxic activity toward K562 cells, making these samples an adequate material for investigating NK cell–mediated cytotoxicity. 4. Discussion The data on NK cell cytotoxicity and GvL properties that can be obtained from clinical studies is limited because of the huge impact of different transplantation settings on NK cell alloreactivity, such as immune suppression conditions or T-cell count. Thus, obtaining valid summary information about the role that receptor–ligand interactions play in NK cell activity is difficult. An investigation of donor-derived NK cell properties dependent on different genotypes or expression patterns requires standardized assays. Cryopreservation is an important tool for the investigation of cell properties, as it simplifies the logistic challenges in collecting blood samples and reduces interassay variability. The cryopreservation of PBMCs is a well-known and frequently used technique. However, freezing and thawing can have a significant impact on their function [29,30]. NK cells are especially sensitive to treatments such as centrifugation, change in environment and media, or lack of nutrients, all of which may directly affect their cytotoxic potential. DMSO is a widely used cryoprotectant for different types of cells, including lymphocytes [31]. Our results regarding the slow and fast addition and removal protocols for the cryoprotectant DMSO can be understood in the view of accepted cryobiological knowledge [32]. When a solution with a high osmolarity (e.g., because it contains DMSO) is added to an isotonic cell suspension, the osmotic equilibrium is disturbed. As water can move out of the cells faster then DMSO can move in, this results in an initial shrinkage of the cells. When the cryoprotectant (and water along with the cryoprotectant) permeates through the cell membrane into the cell with
1011
time, the cells increase in volume again. However, too much of such shrinkage may damage the cells lethally or sub-lethally before the actual freezing takes place. The extent of the shrinkage and the rate of the change of the cell volume depend on both the permeability of the cells to water and to the cryoprotectant, a phenomenon that itself is also temperature dependent. The final equilibrium cell volume depends on the concentration of the impermeant solutes in the solution (e.g., electrolytes, proteins) and equals the normal volume only if the concentration of the impermeant solutes is isotonic. For this reason, it makes sense to add penetrating cryoprotectants (e.g., DMSO) stepwise and slowly, which allows the cells to adapt to the changing osmolarity. By contrast, extremely slow additions (especially at higher temperatures) involve the danger of toxic cell damage, something that cannot be ruled out for DMSO, although the literature is not unequivocal regarding this [33,34]. We speculate that the correct answer strongly depends on the cell type concerned. In any case, our “slow addition of the cryoprotectant protocol” turned out to be a suitable one. When a permeating cryoprotectant such as DMSO needs to be removed, the same problem (but reciprocal) shows up. When the DMSO-loaded cells after thawing are exposed to an isotonic medium for the removal of the cryoprotectant, the osmotic equilibrium is disturbed again. As water is able to enter into the cells much more quickly than the DMSO is able to get out, the cells swell above their normal volume, and shrink again as the cryoprotectant (accompanied by water) moves out. The cells will return to their physiologic volume only if nonpenetrating solutes have neither been gained nor lost during this process and the freezing and thawing itself. Our results show that, under our experimental conditions, PBMC are better preserved when a slow stepwise dilution process is used. It is accepted cryobiologic knowledge that too much swelling is more damaging than too much shrinkage, and that therefore the removal of the cryoprotectant is more critical then the addition. It is known that cryopreservation reduces NK cell activity [35]. We showed that the time span involved in adding or removing DMSO directly effects NK cell activity without disrupting the cells’ basic physiologic properties, but no data are available on the possible differential influence on NK cells with different genotypic and phenotypic properties. We measured a strong but comparable reduction in NK cell cytotoxicity for seven different donors, indicating that cryopreserved cell samples are an adequate tool for comparative measurement of physiologic NK cell function. Because incubation of thawed PBMC samples unevenly increased the cytotoxic activity of NK cells, we exclude incubation as a possibility to restore cytotoxic activity when the individual differences should be compared. The reason for this unexpected observation is not clear. It might be that DMSO stimu-
Fig. 7. Significant correlation between cytolytic activity of PBMCs and percentage of NK cells. Percentage of CD3⫺CD56⫹ NK cells in PBMC samples was determined by FACS analysis and compared with their cytotoxic activity determined in a standardized 51chromium release assay (CRA) without normalization for NK cell count. Only HLA class I identical donors were included to avoid interference from licensing effects caused by differences in HLA background (n ⫽ 28, r ⫽ 0.77).
1012
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
lates NK cells similar to IL-2; however, this requires further experimentation. The cytotoxic activity of freshly prepared PBMC and subsequently separated NK cell samples showed a clear correlation, indicating that PBMCs and separated NK cells can be equally used as a tool for investigation of individual NK cell activity. However, NK cells separated from thawed PBMCs are not able to survive without cytokine supplementation. It has been shown that cultivation of NK cells alters cytokine production and receptor expression [36,37]. De Rham et al. reported the induction of KIR expression on previously KIR-negative NK cells by IL-2 and IL-15 in vitro. They showed an up-regulation of the natural cytotoxicity receptors (NCRs) NKp30 and NKp44, activating receptors that play a crucial role in NK cell activation [38]. In addition, IL-2 was observed to induce an increase in KIR2DL1/2-positive and KIR2DL3-positive NK cells [39]. In line with these observations, we found that the cytotoxic activity of NK cells was not only increased upon IL-2 stimulation, but the differential responsiveness of several donor NK cells was equalized, indicating that IL-2 cannot be used for the cultivation of NK cells when physiologic NK cell function is being investigated. Thus, thawed PBMC samples rather then separated NK cells are an appropriate tool for investigation of cytotoxic NK cell activity, when large groups of donors are needed. Our extrapolation method allowed us to correct for different NK cell numbers from different donors, thus excluding the possibility of false correlations in comparative NK cell studies. In conclusion, our protocol was shown to be suitable for the investigation of individual NK cell cytotoxicity, which we hope will predict the in vivo situation. This method may help elucidate the interacting role of activating and inhibitory receptors and their specific ligands on NK cell education, as well as NK cell alloreactivity in clinical settings. Acknowledgments We thank Rudi Wank (Munich) for the gift of the K562 cells and for helping us to establish the 51Cr-release assay in our laboratory. We also thank Cathrin Schwarz for support in daily work. This work was supported by Erich and Gertrud Roggenbuck-Stiftung. References [1] Parham P. Innate immunity: The unsung heroes. Nature 2003;423:20. [2] Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. Functions of natural killer cells. Nat Immunol 2008;9:503–10. [3] Raulet DH, Held W. Natural killer cell receptors: The offs and ons of NK cell recognition. Cell 1995;82:697–700. [4] Middleton D, Gonzelez F. The extensive polymorphism of KIR genes. Immunology 2010;129:8 –19. [5] Ljunggren HG, KÅrre K. In search of the “missing self”: MHC molecules and NK cell recognition. Immunol Today 1990;11:237– 44. [6] Anfossi N, AndrÊ P, Guia S, Falk CS, Roetynck S, Stewart CA, et al. Human NK cell education by inhibitory receptors for MHC class I. Immunity 2006;25:331– 42. [7] Kim S, Poursine-Laurent J, Truscott SM, Lybarger L, Song Y-J, Yang L, et al. Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature 2005;436:709 –13. [8] Fernandez NC, Treiner E, Vance RE, Jamieson AM, Lemieux S, Raulet DH. A subset of natural killer cells achieves self-tolerance without expressing inhibitory receptors specific for self-MHC molecules. Blood 2005;105:4416 –23. [9] Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, et al. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity 1995;2:439 – 49. [10] Falk CS, Steinle A, Schendel DJ. Expression of HLA-C molecules confers target cell resistance to some non-major histocompatibility complex-restricted T cells in a manner analogous to allospecific natural killer cells. J Exp Med 1995;182:1005–18. [11] Moretta L, Moretta A. Killer immunoglobulin-like receptors. Curr Opin Immunol 2004;16:626 –33. [12] Stern M, Ruggeri L, Capanni M, Mancusi A, Velardi A. Human leukocyte antigens A23, A24, and A32 but not A25 are ligands for KIR3DL1. Blood 2008;112: 708 –10. [13] Fauriat C, Ivarsson MA, Ljunggren H-G, Malmberg K-J, MichaÌlsson J. Education of human natural killer cells by activating killer cell immunoglobulin-like receptors. Blood 2010;115:1166 –74.
[14] Cognet C, Farnarier C, Gauthier L, Frassati C, AndrÊ P, MagÊrus-Chatinet A, et al. Expression of the HLA-C2-specific activating killer-cell Ig-like receptor KIR2DS1 on NK and T cells. Clin Immunol 2010;135:26 –32. [15] Uhrberg M, Parham P, Wernet P. Definition of gene content for nine common group B haplotypes of the Caucasoid population: KIR haplotypes contain between seven and eleven KIR genes. Immunogenetics 2002;54:221–9. [16] Niederwieser D, Gastl G, Rumpold H, Marth C, Kraft D, Huber C. Rapid reappearance of large granular lymphocytes (LGL) with concomitant reconstitution of natural killer (NK) activity after human bone marrow transplantation (BMT). Br J Haematol 1987;65:301–5. [17] Bignon J-D, Gagne K. KIR matching in hematopoietic stem cell transplantation. Curr Opin Immunol 2005;17:553–9. [18] Pende D, Marcenaro S, Falco M, Martini S, Bernardo ME, Montagna D, et al. Anti-leukemia activity of alloreactive NK cells in KIR ligand-mismatched haploidentical HSCT for pediatric patients: Evaluation of the functional role of activating KIR and redefinition of inhibitory KIR specificity. Blood 2009;113:3119 –29. [19] Feuchtinger T, Pfeiffer M, Pfaffle A, Teltschik H-M, Wernet D, Schumm M, et al. Cytolytic activity of NK cell clones against acute childhood precursor-B-cell leukaemia is influenced by HLA class I expression on blasts and the differential KIR phenotype of NK clones. Bone Marrow Transplant 2009;43:875– 81. [20] Kr×ger N, Binder T, Zabelina T, Wolschke C, Schieder H, Renges H, et al. Low number of donor activating killer immunoglobulin-like receptors (KIR) genes but not KIR-ligand mismatch prevents relapse and improves disease-free survival in leukemia patients after in vivo T-cell depleted unrelated stem cell transplantation. Transplantation 2006;82:1024 –30. [21] McQueen KL, Dorighi KM, Guethlein LA, Wong R, Sanjanwala B, Parham P. Donor-recipient combinations of group A and B KIR haplotypes and HLA class I ligand affect the outcome of HLA-matched, sibling donor hematopoietic cell transplantation. Hum Immunol 2007;68:309 –23. [22] Cooley S, Weisdorf DJ, Guethlein LA, Klein JP, Wang T, Le CT, et al. Donor selection for natural killer cell receptor genes leads to superior survival after unrelated transplantation for acute myelogenous leukemia. Blood 2010;116:2411–9. [23] Witt CS. The influence of NK alloreactivity on matched unrelated donor and HLA identical sibling haematopoietic stem cell transplantation. Curr Opin Immunol 2009;21:531–7. [24] Cooley S, McCullar V, Wangen R, Bergemann TL, Spellman S, Weisdorf DJ, et al. KIR reconstitution is altered by T cells in the graft and correlates with clinical outcomes after unrelated donor transplantation. Blood 2005;106:4370 – 6. [25] David G, Morvan M, Gagne K, Kerdudou N, Willem C, Devys A, et al. Discrimination between the main activating and inhibitory killer cell immunoglobulinlike receptor positive natural killer cell subsets using newly characterized monoclonal antibodies. Immunology 2009;128:172– 84. [26] Della Chiesa M, Romeo E, Falco M, Balsamo M, Augugliaro R, Moretta L, et al. Evidence that the KIR2DS5 gene codes for a surface receptor triggering natural killer cell function. Eur J Immunol 2008;38:2284 –9. [27] Gabriel IH, Sergeant R, Szydlo R, Apperley JF, DeLavallade H, Alsuliman A, et al. Interaction between KIR3DS1 and HLA-Bw4 predicts for progression-free survival after autologous stem cell transplantation in patients with multiple myeloma. Blood 2010;116:2033–9. [28] Verheyden S, Ferrone S, Mulder A, Claas FH, Schots R, De Moerloose B, et al. Role of the inhibitory KIR ligand HLA-Bw4 and HLA-C expression levels in the recognition of leukemic cells by natural killer cells. Cancer Immunol Immunother 2009;58:855– 65. [29] Disis ML, dela Rosa C, Goodell V, Kuan L-Y, Chang JCC, Kuus-Reichel K, et al. Maximizing the retention of antigen specific lymphocyte function after cryopreservation. J Immunol Methods 2006;308:13– 8. [30] Betensky RA, Connick E, Devers J, Landay AL, Nokta M, Plaeger S, et al. Shipment impairs lymphocyte proliferative responses to microbial antigens. Clin Diagn Lab Immunol 2000;7:759 – 63. [31] Sputtek A, K×rber C. Cryopreservation of red blood cells, platelets, lymphocytes, and stem cells. In: Fuller BJ, Grout BWW, editors: Clinical Applications of Cryobiology. Boca Raton, FL, CRC Press 1991, pp 95–147. [32] Pegg DE. Principles of Cryopreservation. In: Day JG, Stacey GN, editors: Cryopreservation and Freeze-Drying Protocols. Totowa, NJ, Humana Press 2007, pp 39 –57. [33] Rowley SD, Anderson GL. Effect of DMSO exposure without cryopreservation on hematopoietic progenitor cells. Bone Marrow Transplant 1993;11:389 –93. [34] Fahy G. Cryoprotectant toxicity neutralization. Cryobiology 2010;60:45–53. [35] MartÎ F, Miralles A, PeirÔ M, Amill B, de Dalmases C, PiÒol G, et al. Differential effect of cryopreservation on natural killer cell and lymphokine-activated killer cell activities. Transfusion 1993;33:651–5. [36] Berg M, Lundqvist A, McCoy P, Samsel L, Fan Y, Tawab A, et al. Clinical-grade ex vivo-expanded human natural killer cells up-regulate activating receptors and death receptor ligands and have enhanced cytolytic activity against tumor cells. Cytotherapy 2009;11:341–55. [37] de Rham C, Ferrari-Lacraz S, Jendly S, Schneiter G, Dayer J-M, Villard J. The proinflammatory cytokines IL-2, IL-15 and IL-21 modulate the repertoire of mature human natural killer cell receptors. Arthritis Res Ther 2007;9:R125. [38] Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, et al. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu Rev Immunol 2001;19:197–223. [39] Chrul S, Polakowska E, Szadkowska A, Bodalski J. Influence of interleukin IL-2 and IL-12 ⫹ IL-18 on surface expression of immunoglobulin-like receptors KIR2DL1, KIR2DL2, and KIR3DL2 in natural killer cells. Mediators Inflamm 2006;2006:46957.
Contents lists available at SciVerse ScienceDirect
Assessment of physiologic natural killer cell cytotoxicity in vitro Heidi Duske a, Andreas Sputtek a, Thomas Binder a, Nicolaus Kr×ger b, Sonja Schrepfer c, Thomas Eiermann a,* a
Transfusion Medicine, HLA-Laboratory, University Hospital Hamburg-Eppendorf, Hamburg, Germany Clinic for Stem Cell Transplantation, University Hospital Hamburg-Eppendorf, Hamburg, Germany c Cardiovascular Surgery, Transplant and Stem cell Immunobiology-Laboratory, University Hospital Hamburg-Eppendorf, Hamburg, Germany b
A R T I C L E
I N F O
Article history: Received 26 November 2010 Accepted 4 August 2011 Available online 10 August 2011
Keywords: Natural killer cells Leukemia K562 51 Cr release
A B S T R A C T
Here, we describe an improved 51chromium release assay (CRA) to compare donor natural killer (NK) cell activity. To validate the assay, we analyzed sample preparation, incubation, and cryopreservation of NK cells. The effector-to-target ratio was corrected for the percentage of NK cells. A logarithmic curve was fitted to the data of the CRA for calculation of the maximum activity. The specific lysis was standardized to a reference sample and normalized to the mean specific lysis of the reference. We found that a longer time span involved with both the addition and the removal of DMSO increased the recovery of NK cell activity. Freezing and thawing reduced the cytotoxicity of NK cells but sustained the relative differences that were seen between freshly prepared NK cells. In contrast, medium incubation of thawed cells markedly increased the cytotoxic potential but also deranged these relative differences. Those were widely equalized when cells were stimulated with IL-2. In conclusion, we established a standardized assay with cryopreserved peripheral blood mononuclear cells as an appropriate tool for investigation of individual physiologic NK cell activity. This assay may help to predict donor NK cell activity in vivo, to reconcile conflicting data about NK cells obtained in transplantation studies. 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
1. Introduction The majority of natural killer (NK) cells in the peripheral blood display cytotoxic properties that are crucial in the early defense against virally infected and mutated or stressed cells [1,2]. NK cells are regulated by integrated signaling through a repertoire of germ line– encoded activating and human leukocyte antigen (HLA) class I–specific inhibitory receptors [3,4]. Ljunggren and Karre first published, in the “missing self” hypothesis, that HLA class I downregulation renders infected or mutated cells susceptible to NK cell killing [5]. It is now well established that, during maturation, NK cells require binding between inhibitory receptors and their respective HLA class I molecules to achieve full responsiveness, a process referred to as “licensing” or “education” [6 – 8]. The most important educating receptors belong to the highly diverse killer immunoglobulin-like receptor (KIR) family: KIR2DL2/3 and KIR2DL1, which are specific for HLA-Cw molecules of the C1 (Asp at position 80) and C2 (Lys at position 80) groups, respectively [9,10], and KIR3DL1, which is specific for HLA-Bw4 epitopes present in one-third of HLA-B, and some HLA-A, alleles [11,12]. Furthermore, recent data have revealed an antagonizing role of
* Corresponding author. E-mail address: [email protected] (T.H. Eiermann).
the activating KIR2DS1 in NK cell education only in the absence of its inhibitory counterpart KIR2DL1 [13,14]. KIRs are inherited in haplotype patterns and segregate independently from HLA class I molecules located on different chromosomes. Haplotypes are distinguished as type A, coding for a set of seven distinct KIR genes, including one single activating KIR (aKIR), KIR2DS4. The B haplotypes are diverse and characterized by the presence of at least two and up to four aKIRs [15]. The strength of the cytotoxic response is likely defined through the combination of educating receptors and corresponding ligands. Whether and how the expression of noneducating receptors shape the NK cell response upon ligand binding and affect overall NK cell activity remains elusive. Segregating independently, both HLA class I and KIR genotypes affect the development of the individual functional cytotoxic phenotype mature NK cells display. In the last few years, extensive studies have revealed an important role for NK cells in the outcome of hematopoietic stem cell transplantation (HSCT) in leukemia patients. Being the first lymphocytic cell population to arise in the patient’s peripheral blood after transplantation [16], NK cells display a beneficial graftversus-leukemia (GvL) effect by targeting residual malignant cells in the patient. However, which combination of donor KIR type and recipient HLA class I background has the most promising GvL effect is still discussed. Conflicting data from clinical studies on transplantation outcome regarding KIR-ligand mismatch [17–19] and
0198-8859/11/$32.00 - see front matter 䉷 2011 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.humimm.2011.08.006
1008
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
KIR genotype [20 –22] seem to be generated through different transplantation setups [23,24]. Several groups have elucidated the role of different KIR-ligand interactions in NK cell (allo-) reactivity by investigating the cytotoxic function of single NK cell subsets that express no or single KIR receptors [12–14,25– 28]. However, NK cells that arise in patients after bone marrow transplantation will express every possible receptor combination provided by the genotype. Thus, the actual strength of the GvL effect will depend on the overall NK cell population and not individual subpopulations. Our aim in this study was to design an assay for examining the properties of NK cells, derived from numerous individuals, in their physiologic state that is created during the developmental licensing. This should provide a better understanding of the influence that the genetic configuration of KIR, HLA class I, and other regulating molecules have on the overall cytotoxic potential of NK cells. That, in turn may advance the prediction of the strength of the GvL effect in the selection of the most promising stem cell donor for leukemia patients. We investigated the influence of cryopreservation, cell collection and separation as well as stimulation and incubation on the cytotoxic activity of NK cells, and designed an appropriate assay setup.
protocol, the cells were supplemented with 20% DMSO in 10 50-l steps every 30 seconds, extending the addition process to 5 minutes in total. All tubes were immediately placed into a cryo 1⬚C/min freezing container (“Mr. Frosty,” Nalgene) and cooled to ⫺80⬚C overnight before transfer to the gaseous phase of liquid nitrogen. For thawing, the cryotubes were directly transferred from the liquid nitrogen tank to a 37⬚C water bath and completely thawed by continuous shaking. Thereafter cells were immediately cooled on ice and transferred to 14-ml tubes placed in a rack on a cooled thermal pack and washed (all centrifugations at 250 g for 5 minutes at 4⬚C). In the fast removal of cryoprotectant (FRCP) protocol, 10 ml of precooled culture medium was slowly added under continuous shaking within 1 minute, followed by 2 washing steps with 10 ml culture medium added all at once. The slow removal of cryoprotectant (SRCP) protocol was performed according to the following timetable: wash 1: 4 ⫻ 250 l culture medium, 8 ⫻ 1 ml cold culture medium, every 30 seconds (total 6 minutes); wash 2: 10 ⫻ 1 ml culture medium every 20 seconds (total 3.5 minutes); and wash 3: 10 ml cold culture medium at once (total 30 seconds). After washing, the cells were resuspended in an appropriate volume of culture medium to obtain the desired E:T ratio and were directly used for the 51Cr-release assay (CRA). 2.3. HLA typing
2. Subjects and methods 2.1. Cell lines and sample collection The HLA class I– deficient myeloid cell line K562 was cultivated in RPMI 1640 supplemented with 10% fetal calf serum culture medium at 37⬚C in 5% CO2. The identity of the cell line was confirmed by HLA typing (HLA-A*11:01, -B*18:01, -DRB1*03:01). K562 cells were used as the NK cell target in all cytotoxic assays. To exclude reactions of T cells with K562 cells, we affirmed the predicted HLA class I deficiency in K562 cells using FACS analysis (data not shown). Peripheral blood mononuclear cells (PBMCs) were collected from the buffycoats or whole blood of healthy donors by density centrifugation (FicollHypaque, GE Healthcare), (28 HLA-C1/C1 and seven not HLA typed). The cells were either used immediately or cryopreserved (as discussed in next section). Frozen PBMCs were thawed immediately before use or incubated either in culture medium without supplements for different time spans or in culture medium supplemented with 100 U/ml interleukin (IL)–2 (eBioscience) for 16 hours at 37⬚C in 5% CO2 atmosphere at a density of 1 ⫻ 106 cells/ml. NK cell separation was performed by depletion of Non-NK cells using the MACS isolation kit (Order no. 130-092657, Miltenyi Biotec). Freshly isolated PBMC samples were incubated in culture medium for 2–3 hours before NK cell separation. Frozen PBMC samples were thawed, resuspended in culture medium containing 100 U/ml IL-2, and incubated overnight at 37⬚C and 5% CO2. The next day, NK cells were separated and resuspended in culture medium containing 100 U/ml IL-2. Membrane integrity was tested by trypan blue exclusion referred to percentage of viable cells. 2.2. Cell storage and preparation For freezing, freshly prepared PBMCs were diluted to a concentration of 40 ⫻ 106 cells/ml in RPMI-1640 supplemented with 50% fetal calf serum. Cell suspension aliquots (500 l) were transferred to 2-ml cryotubes (Greiner Bio-One Frickenhausen, Germany) and cooled on ice. Two timelines for addition and removal of cryoprotectant, respectively, have been tested: In the fast addition of cryoprotectant (FACP) protocol, an equal volume of cold 20% dimethyl sulfoxide (DMSO, purity ⱖ99.7%; Order no. 1.09678.0100 Merck KGaA, Darmstadt, Germany) was added to the cells dropwise within 30 seconds. In the slow addition of cryoprotectant (SACP)
Genomic DNA was isolated from 5 ml of donor buffycoat using the NucleoSpin Blood XL kit (Macherey-Nagel, Dueren, Germany). The donor HLA-A, -B, and -C loci were typed using the reverse SSO line blot assay (Dynal Reli SSO, Invitrogen) 2.4. Cytolytic assays Cytolytic activity of NK cells was measured in a standard 4-hour CRA. One million K562 target cells were stained for 2 hours with 80 Ci 51Cr at 37⬚C, 5% CO2. PBMCs were added to 5000 target cells at triplicate effector to target (E:T) ratios of 10:1, 20:1, 40:1, and 80:1 in a final volume of 200 l in U-bottom 96-well plates. After 4-hour incubation at 37⬚C, 5% CO2, a 100-l quantity of supernatant was transferred to fresh tubes and the counts per minute (cpm.) measured in a gamma counter (PerkinElmer/Wallac). Cytolytic activity was calculated according to the following formula: specific lysis [ % ] ⫽
cpm test ⫺ cpm spontaneous release cpm max release ⫺ cpm spontaneous release
⫻ 100
To compare the cytotoxicity of the NK cells from different donors, the percentage of NK cells in each sample was used for the calculation of the actual effector-to-target ratio. A logarithmic curve was fitted to the data for calculation of the maximum cytolytic activity mediated by each donor cell sample (y ⫽ a ln(x) ⫹ b). Frozen/ thawed PBMCs from the same donor were used as a reference sample in every test. Each sample has been tested in two or three independent experiments. 2.5. Flow-cytometric assay and antibodies All FACS analyses were performed with 5 ⫻ 105 cells using a FACScan cytometer (Becton Dickinson). Data were analyzed with the CellQuestPro software. PBMCs were stained with the mouse mAb anti-human CD3-FITC clone OKT3 (IgG2a; eBioscience) and either anti-human CD56-PE clone B159 (IgG1; BD Biosciences) or the isotype control mouse IgG1-PE (eBioscience) for 30 minutes at 4⬚C. The samples were washed once and kept on ice in the dark until the measurement. To ensure HLA class I negativity, K562 cells were stained with anti-human MHC-ABC-PE clone DX17 (IgG1; BD) and anti-human MHC-DR ⫹ DP ⫹ DQ-PE clone CR3143 (IgG2a; Abcam).
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
1009
2.6. Statistical analysis Significance was tested with the two-tailed t test for paired samples. The level of significance was specified as p ⬍ 0.05 (*), ⬍ 0.01 (**), ⬍ 0.001 (***) or ⬎ 0.05 (NS). 3. Results 3.1. Effect of addition and removal of cryoprotectant Cryopreservation of cell samples considerably facilitates the collection and standardized testing of cells, especially when a large number of donors is needed to address the question. Because NK cells are especially sensitive to freezing and thawing, we optimized the protocols for best recovery of cell function. By comparing the cytolytic activity of PBMC samples treated according to the “fast” or “slow” addition/removal protocols, we found that a longer time span involved with both the addition and the removal of DMSO increased the recovery of NK cell activity (Figs. 1A, 1B; p ⫽ 0.0006 and p ⫽ 0.003). The membrane integrity of the cells was not affected by slow or fast addition of DMSO but slightly decreased after fast removal compared with slow removal of DMSO (Figs. 1C, 1D; NS and p ⫽ 0.01). These findings indicate that a determination of viability alone is not sufficient for deciding if NK cells are still responsive to a stimulus. A careful cryopreservation treatment, in terms of slowly adding and removing of DMSO, is necessary for NK cells to maintain the highest cell function. For all of the following experiments, the slow addition and removal protocols were used.
Fig. 2. Extrapolation of data allows a comparison of the cytotoxicity of various donor PBMC samples. The E:T ratio was corrected for the percentage of NK cells of PBMC and a logarithmic extrapolation curve was applied to each sample (dashed lines). A reference PBMC sample was included in each 51chromium release assay (CRA). To correct for interassay variations, the specific lysis was standardized to this reference sample and normalized to the mean specific lysis of the reference sample (29% ⫾ 7%) obtained in 15 experiments. The percentages of specific lysis of all samples were compared at the extrapolated E:T ratio 15:1, where all curves had reached the plateau phase.
3.2. Extrapolating data: Putting NK cell percentages into perspective One problem with comparing the cytotoxic activities of NK cells from various donors is the individually changing NK cell proportion, which ranges from 1% to 10% of PBMCs. To overcome this difficulty, we developed a strategy that takes into account NK cell percentage in the analysis of data obtained in CRAs. First, the E:T ratios were corrected for the actual number of NK cells in each well as determined by cell counting and FACS analysis. A logarithmic regression curve was fitted to the values obtained for the specific lysis at different E:T ratios. Using this curve, the values for the specific lysis of K-562 were normalized to an actual E:T ratio of 15:1. This ratio was chosen because it was likely that the majority of tested samples would have reached the plateau (e.g., maximum) phase of lysis (Fig. 2). Second, in each assay, a reference sample was included as a standard set to 100% to eliminate interassay variations. Finally, the results were normalized to the mean specific lysis obtained for the reference sample (29% ⫾7%) in 15 separate experiments. 3.3. Individual cytotoxic activity of NK cells is not altered by cryopreservation We next examined the impact of cryopreservation on the effector properties of NK cells derived from different individuals. PBMCs from 9 healthy donors were prepared and either immediately used in CRA or cryopreserved. One week later, the frozen samples were thawed and immediately tested by CRA. As expected, the cytotoxic activity was significantly reduced after freez-/thawing the PBMC samples (mean recovery 25% ⫾ 2%). However, the relative cytolytic potential of the individual donor NK cells was not affected (Fig. 3). Thus, cryopreservation seems to be an appropriate tool for investigation of individual cytotoxic NK cell properties. 3.4. Incubation of thawed PBMCs helps to restore NK cell cytotoxicity but shifts individual differences
Fig. 1. Slow addition and slow removal of DMSO helps to preserve cytotoxic activity of NK cells. Effect of slow and fast DMSO addition and removal protocols for freshly prepared peripheral blood mononuclear cells (PBMCs) on cell viability and recovery of NK cell cytotoxicity was determined by trypan blue exclusion and standard 51 chromium release assay (CRA), respectively. (A, C) Samples were frozen according to either the slow or fast addition of DMSO protocol and were thawed using slow removal protocols. (B, D) All samples were frozen according to the slow addition of DMSO protocol and were thawed using either the slow or fast removal of DMSO protocol.
To determine whether the reduced NK cell activity of frozen samples could be restored, we tested thawed PBMC samples that were incubated in culture medium for 0, 3, 16, or 24 hours in a CRA. The measured cytotoxic activity markedly increased with incubation time; however, as Fig. 4 shows, this increase was not consistent among the six donors. Although a 3-hour incubation could either increase or decrease the basal activity (factor 3.0 – 0.8), the increase
1010
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
Fig. 3. Cryopreservation uniformly reduces the cytotoxicity of NK cells but does not change their capacity for specific lysis. Cytotoxicity of fresh and cryopreserved PBMCs from seven randomly selected healthy donors was measured in a standard 51 chromium release assay (CRA) with K562 target cells. Mean recovery of NK cell cytotoxicity after cryopreservation was 25% ⫾ 2%. All values were standardized and normalized to an E:T-ratio of 15:1 as shown in Fig. 2.
after 24 hours was more pronounced the lower the basal activity of the individual sample (factor 1.0 –11.2). Thus, incubation of thawed PBMCs shifted the donor specific activity of NK cells relatively to each other and therefore cannot be used to compare the original individual cytotoxic NK cell activity. 3.5. Individual NK cell activity can be measured with unstimulated PBMCs The NK cell cytolytic abilities are sensitive to environmental signals like cytokines, antibodies or cell– cell contacts. In vivo, cytotoxic CD56dim NK cells reside in the peripheral blood together with T- and B-lymphocytes, Natural Killer T cells and monocytes. Thus, we investigated if cytolytic activity of NK cells can be specifically measured without separating them from PBMC. We also examined the influence of non-NK cell subsets in PBMC samples by comparing the cytolytic activity of PBMCs and separated NK cells. The cytotoxic activity of freshly prepared PBMCs, either from buffycoat or whole-blood samples, and subsequently separated NK cells was compared in a standard chromium-release
Fig. 4. Incubation of thawed PBMCs restores the cytotoxicity of NK cells nonuniformly. The cytotoxicity of cryopreserved PBMCs from seven randomly selected healthy donors was measured in a standard CRA with K562 target cells after incubation in culture medium for 0, 3, 16, or 24 hours, respectively. Within 24 hours of incubation, the NK cell activity increased in a wide range between 1.0- and 11.2-fold. All values were standardized and normalized to an E:T-ratio of 15:1.
Fig. 5. Cytotoxic activity of PBMC and separated NK cell samples is correlated to each other. (A) Percentage of CD3⫺CD56⫹ NK cells was determined by FACS analysis before and after NK cell separation and is shown as percentage of lymphocytes. (B) Fresh, PBMCs derived from the buffycoat (solid bars) or whole blood (ruled bars) of six random healthy donors and subsequently separated NK cells were measured in a standard 51chromium release assay (CRA) with K562 target cells. All cytolytic values were standardized and normalized to an E:T-ratio of 15:1 as shown in Fig. 2.
assay (Fig. 5). The percentage of NK cells in the PBMC samples ranged between 8% and 40% of lymphocytes and could be enhanced up to 87–95% by negative selection of CD3⫺CD56⫹ NK cells (Fig. 5A). Despite the broad range of NK cell percentages in the PBMC samples, there was a clear trend toward correlation between the cytotoxic activity of whole PBMC and separated NK cells, with a slightly lower activity of PBMC, thus preserving the relative difference of the cytotoxic potential between the individuals. The difference between the cytotoxic activity displayed by PBMC or NK was slightly higher when samples were derived from whole blood (factor 1.3 ⫾ 0.1; ruled bars in Fig. 5B) rather than buffycoat (factor 1.1 ⫾ 0.05; solid bars in Fig. 5B). The data show that the established normalization renders either fresh PBMCs or separated NK cells equally valid samples for the determination of individual NK cell activity. To test the suitability of NK cells separated from thawed PBMC samples, we added IL-2 to the cell suspension, as NK cells separated from frozen PBMCs essentially need interleukin supplementation for survival. Cell samples from 10 healthy donors with homozygous HLA-C1 were tested in a CRA. As expected, the mean specific lysis was significantly higher for IL-2–stimulated PBMCs than for native PBMCs (mean 78% ⫾ 8% vs 13% ⫾ 7%; p ⬍ 10⫺4), whereas the difference between stimulated PBMCs and separated NK cells was comparatively low (78% ⫾ 8% vs 84% ⫾ 8%; Fig. 6). The differential cytolytic potential of native PBMCs was widely equalized after IL-2 stimulation, suggesting that the preactivation overrides the physiologic NK cell state. No significant difference was found between the diverse lytic activity of unstimulated fresh or stimulated thawed PBMCs and
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
Fig. 6. IL-2 stimulation overrides physiologic cytotoxic potential of NK cells. Differential cytotoxic potential of cryopreserved samples from 10 randomly selected healthy donors was compared (native PBMC, IL-2–stimulated PBMC, or separated NK cells stimulated with IL-2). Error bars indicate SEM of two or three repeated measurements. Cytolytic activity of cells was determined in a standardized 51chromium release assay (CRA). Values have been normalized, as shown in Fig. 2.
subsequently separated NK cells, implying that no other cell population, such as T or Natural Killer T cells, in the PMBC mix had an effect on the specific lysis of K562. This finding was affirmed through a significant correlation between the specific lysis and the percentage of CD56⫹CD3- NK cells in thawed unstimulated PBMC samples of 28 healthy donors (r ⫽ 0.77, p ⬍ 10⫺4; Fig. 7). All donors were homozygous for HLA-C1 to exclude licensing differences of NK cells and thereby minimize side effects on cytotoxic potential. No correlation was seen for any of the other cell subsets (data not shown). Taken together, the results indicate that, in PBMC samples, only NK cells contribute to the cytotoxic activity toward K562 cells, making these samples an adequate material for investigating NK cell–mediated cytotoxicity. 4. Discussion The data on NK cell cytotoxicity and GvL properties that can be obtained from clinical studies is limited because of the huge impact of different transplantation settings on NK cell alloreactivity, such as immune suppression conditions or T-cell count. Thus, obtaining valid summary information about the role that receptor–ligand interactions play in NK cell activity is difficult. An investigation of donor-derived NK cell properties dependent on different genotypes or expression patterns requires standardized assays. Cryopreservation is an important tool for the investigation of cell properties, as it simplifies the logistic challenges in collecting blood samples and reduces interassay variability. The cryopreservation of PBMCs is a well-known and frequently used technique. However, freezing and thawing can have a significant impact on their function [29,30]. NK cells are especially sensitive to treatments such as centrifugation, change in environment and media, or lack of nutrients, all of which may directly affect their cytotoxic potential. DMSO is a widely used cryoprotectant for different types of cells, including lymphocytes [31]. Our results regarding the slow and fast addition and removal protocols for the cryoprotectant DMSO can be understood in the view of accepted cryobiological knowledge [32]. When a solution with a high osmolarity (e.g., because it contains DMSO) is added to an isotonic cell suspension, the osmotic equilibrium is disturbed. As water can move out of the cells faster then DMSO can move in, this results in an initial shrinkage of the cells. When the cryoprotectant (and water along with the cryoprotectant) permeates through the cell membrane into the cell with
1011
time, the cells increase in volume again. However, too much of such shrinkage may damage the cells lethally or sub-lethally before the actual freezing takes place. The extent of the shrinkage and the rate of the change of the cell volume depend on both the permeability of the cells to water and to the cryoprotectant, a phenomenon that itself is also temperature dependent. The final equilibrium cell volume depends on the concentration of the impermeant solutes in the solution (e.g., electrolytes, proteins) and equals the normal volume only if the concentration of the impermeant solutes is isotonic. For this reason, it makes sense to add penetrating cryoprotectants (e.g., DMSO) stepwise and slowly, which allows the cells to adapt to the changing osmolarity. By contrast, extremely slow additions (especially at higher temperatures) involve the danger of toxic cell damage, something that cannot be ruled out for DMSO, although the literature is not unequivocal regarding this [33,34]. We speculate that the correct answer strongly depends on the cell type concerned. In any case, our “slow addition of the cryoprotectant protocol” turned out to be a suitable one. When a permeating cryoprotectant such as DMSO needs to be removed, the same problem (but reciprocal) shows up. When the DMSO-loaded cells after thawing are exposed to an isotonic medium for the removal of the cryoprotectant, the osmotic equilibrium is disturbed again. As water is able to enter into the cells much more quickly than the DMSO is able to get out, the cells swell above their normal volume, and shrink again as the cryoprotectant (accompanied by water) moves out. The cells will return to their physiologic volume only if nonpenetrating solutes have neither been gained nor lost during this process and the freezing and thawing itself. Our results show that, under our experimental conditions, PBMC are better preserved when a slow stepwise dilution process is used. It is accepted cryobiologic knowledge that too much swelling is more damaging than too much shrinkage, and that therefore the removal of the cryoprotectant is more critical then the addition. It is known that cryopreservation reduces NK cell activity [35]. We showed that the time span involved in adding or removing DMSO directly effects NK cell activity without disrupting the cells’ basic physiologic properties, but no data are available on the possible differential influence on NK cells with different genotypic and phenotypic properties. We measured a strong but comparable reduction in NK cell cytotoxicity for seven different donors, indicating that cryopreserved cell samples are an adequate tool for comparative measurement of physiologic NK cell function. Because incubation of thawed PBMC samples unevenly increased the cytotoxic activity of NK cells, we exclude incubation as a possibility to restore cytotoxic activity when the individual differences should be compared. The reason for this unexpected observation is not clear. It might be that DMSO stimu-
Fig. 7. Significant correlation between cytolytic activity of PBMCs and percentage of NK cells. Percentage of CD3⫺CD56⫹ NK cells in PBMC samples was determined by FACS analysis and compared with their cytotoxic activity determined in a standardized 51chromium release assay (CRA) without normalization for NK cell count. Only HLA class I identical donors were included to avoid interference from licensing effects caused by differences in HLA background (n ⫽ 28, r ⫽ 0.77).
1012
H. Duske et al. / Human Immunology 72 (2011) 1007-1012
lates NK cells similar to IL-2; however, this requires further experimentation. The cytotoxic activity of freshly prepared PBMC and subsequently separated NK cell samples showed a clear correlation, indicating that PBMCs and separated NK cells can be equally used as a tool for investigation of individual NK cell activity. However, NK cells separated from thawed PBMCs are not able to survive without cytokine supplementation. It has been shown that cultivation of NK cells alters cytokine production and receptor expression [36,37]. De Rham et al. reported the induction of KIR expression on previously KIR-negative NK cells by IL-2 and IL-15 in vitro. They showed an up-regulation of the natural cytotoxicity receptors (NCRs) NKp30 and NKp44, activating receptors that play a crucial role in NK cell activation [38]. In addition, IL-2 was observed to induce an increase in KIR2DL1/2-positive and KIR2DL3-positive NK cells [39]. In line with these observations, we found that the cytotoxic activity of NK cells was not only increased upon IL-2 stimulation, but the differential responsiveness of several donor NK cells was equalized, indicating that IL-2 cannot be used for the cultivation of NK cells when physiologic NK cell function is being investigated. Thus, thawed PBMC samples rather then separated NK cells are an appropriate tool for investigation of cytotoxic NK cell activity, when large groups of donors are needed. Our extrapolation method allowed us to correct for different NK cell numbers from different donors, thus excluding the possibility of false correlations in comparative NK cell studies. In conclusion, our protocol was shown to be suitable for the investigation of individual NK cell cytotoxicity, which we hope will predict the in vivo situation. This method may help elucidate the interacting role of activating and inhibitory receptors and their specific ligands on NK cell education, as well as NK cell alloreactivity in clinical settings. Acknowledgments We thank Rudi Wank (Munich) for the gift of the K562 cells and for helping us to establish the 51Cr-release assay in our laboratory. We also thank Cathrin Schwarz for support in daily work. This work was supported by Erich and Gertrud Roggenbuck-Stiftung. References [1] Parham P. Innate immunity: The unsung heroes. Nature 2003;423:20. [2] Vivier E, Tomasello E, Baratin M, Walzer T, Ugolini S. Functions of natural killer cells. Nat Immunol 2008;9:503–10. [3] Raulet DH, Held W. Natural killer cell receptors: The offs and ons of NK cell recognition. Cell 1995;82:697–700. [4] Middleton D, Gonzelez F. The extensive polymorphism of KIR genes. Immunology 2010;129:8 –19. [5] Ljunggren HG, KÅrre K. In search of the “missing self”: MHC molecules and NK cell recognition. Immunol Today 1990;11:237– 44. [6] Anfossi N, AndrÊ P, Guia S, Falk CS, Roetynck S, Stewart CA, et al. Human NK cell education by inhibitory receptors for MHC class I. Immunity 2006;25:331– 42. [7] Kim S, Poursine-Laurent J, Truscott SM, Lybarger L, Song Y-J, Yang L, et al. Licensing of natural killer cells by host major histocompatibility complex class I molecules. Nature 2005;436:709 –13. [8] Fernandez NC, Treiner E, Vance RE, Jamieson AM, Lemieux S, Raulet DH. A subset of natural killer cells achieves self-tolerance without expressing inhibitory receptors specific for self-MHC molecules. Blood 2005;105:4416 –23. [9] Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, Vitale M, et al. Molecular clones of the p58 NK cell receptor reveal immunoglobulin-related molecules with diversity in both the extra- and intracellular domains. Immunity 1995;2:439 – 49. [10] Falk CS, Steinle A, Schendel DJ. Expression of HLA-C molecules confers target cell resistance to some non-major histocompatibility complex-restricted T cells in a manner analogous to allospecific natural killer cells. J Exp Med 1995;182:1005–18. [11] Moretta L, Moretta A. Killer immunoglobulin-like receptors. Curr Opin Immunol 2004;16:626 –33. [12] Stern M, Ruggeri L, Capanni M, Mancusi A, Velardi A. Human leukocyte antigens A23, A24, and A32 but not A25 are ligands for KIR3DL1. Blood 2008;112: 708 –10. [13] Fauriat C, Ivarsson MA, Ljunggren H-G, Malmberg K-J, MichaÌlsson J. Education of human natural killer cells by activating killer cell immunoglobulin-like receptors. Blood 2010;115:1166 –74.
[14] Cognet C, Farnarier C, Gauthier L, Frassati C, AndrÊ P, MagÊrus-Chatinet A, et al. Expression of the HLA-C2-specific activating killer-cell Ig-like receptor KIR2DS1 on NK and T cells. Clin Immunol 2010;135:26 –32. [15] Uhrberg M, Parham P, Wernet P. Definition of gene content for nine common group B haplotypes of the Caucasoid population: KIR haplotypes contain between seven and eleven KIR genes. Immunogenetics 2002;54:221–9. [16] Niederwieser D, Gastl G, Rumpold H, Marth C, Kraft D, Huber C. Rapid reappearance of large granular lymphocytes (LGL) with concomitant reconstitution of natural killer (NK) activity after human bone marrow transplantation (BMT). Br J Haematol 1987;65:301–5. [17] Bignon J-D, Gagne K. KIR matching in hematopoietic stem cell transplantation. Curr Opin Immunol 2005;17:553–9. [18] Pende D, Marcenaro S, Falco M, Martini S, Bernardo ME, Montagna D, et al. Anti-leukemia activity of alloreactive NK cells in KIR ligand-mismatched haploidentical HSCT for pediatric patients: Evaluation of the functional role of activating KIR and redefinition of inhibitory KIR specificity. Blood 2009;113:3119 –29. [19] Feuchtinger T, Pfeiffer M, Pfaffle A, Teltschik H-M, Wernet D, Schumm M, et al. Cytolytic activity of NK cell clones against acute childhood precursor-B-cell leukaemia is influenced by HLA class I expression on blasts and the differential KIR phenotype of NK clones. Bone Marrow Transplant 2009;43:875– 81. [20] Kr×ger N, Binder T, Zabelina T, Wolschke C, Schieder H, Renges H, et al. Low number of donor activating killer immunoglobulin-like receptors (KIR) genes but not KIR-ligand mismatch prevents relapse and improves disease-free survival in leukemia patients after in vivo T-cell depleted unrelated stem cell transplantation. Transplantation 2006;82:1024 –30. [21] McQueen KL, Dorighi KM, Guethlein LA, Wong R, Sanjanwala B, Parham P. Donor-recipient combinations of group A and B KIR haplotypes and HLA class I ligand affect the outcome of HLA-matched, sibling donor hematopoietic cell transplantation. Hum Immunol 2007;68:309 –23. [22] Cooley S, Weisdorf DJ, Guethlein LA, Klein JP, Wang T, Le CT, et al. Donor selection for natural killer cell receptor genes leads to superior survival after unrelated transplantation for acute myelogenous leukemia. Blood 2010;116:2411–9. [23] Witt CS. The influence of NK alloreactivity on matched unrelated donor and HLA identical sibling haematopoietic stem cell transplantation. Curr Opin Immunol 2009;21:531–7. [24] Cooley S, McCullar V, Wangen R, Bergemann TL, Spellman S, Weisdorf DJ, et al. KIR reconstitution is altered by T cells in the graft and correlates with clinical outcomes after unrelated donor transplantation. Blood 2005;106:4370 – 6. [25] David G, Morvan M, Gagne K, Kerdudou N, Willem C, Devys A, et al. Discrimination between the main activating and inhibitory killer cell immunoglobulinlike receptor positive natural killer cell subsets using newly characterized monoclonal antibodies. Immunology 2009;128:172– 84. [26] Della Chiesa M, Romeo E, Falco M, Balsamo M, Augugliaro R, Moretta L, et al. Evidence that the KIR2DS5 gene codes for a surface receptor triggering natural killer cell function. Eur J Immunol 2008;38:2284 –9. [27] Gabriel IH, Sergeant R, Szydlo R, Apperley JF, DeLavallade H, Alsuliman A, et al. Interaction between KIR3DS1 and HLA-Bw4 predicts for progression-free survival after autologous stem cell transplantation in patients with multiple myeloma. Blood 2010;116:2033–9. [28] Verheyden S, Ferrone S, Mulder A, Claas FH, Schots R, De Moerloose B, et al. Role of the inhibitory KIR ligand HLA-Bw4 and HLA-C expression levels in the recognition of leukemic cells by natural killer cells. Cancer Immunol Immunother 2009;58:855– 65. [29] Disis ML, dela Rosa C, Goodell V, Kuan L-Y, Chang JCC, Kuus-Reichel K, et al. Maximizing the retention of antigen specific lymphocyte function after cryopreservation. J Immunol Methods 2006;308:13– 8. [30] Betensky RA, Connick E, Devers J, Landay AL, Nokta M, Plaeger S, et al. Shipment impairs lymphocyte proliferative responses to microbial antigens. Clin Diagn Lab Immunol 2000;7:759 – 63. [31] Sputtek A, K×rber C. Cryopreservation of red blood cells, platelets, lymphocytes, and stem cells. In: Fuller BJ, Grout BWW, editors: Clinical Applications of Cryobiology. Boca Raton, FL, CRC Press 1991, pp 95–147. [32] Pegg DE. Principles of Cryopreservation. In: Day JG, Stacey GN, editors: Cryopreservation and Freeze-Drying Protocols. Totowa, NJ, Humana Press 2007, pp 39 –57. [33] Rowley SD, Anderson GL. Effect of DMSO exposure without cryopreservation on hematopoietic progenitor cells. Bone Marrow Transplant 1993;11:389 –93. [34] Fahy G. Cryoprotectant toxicity neutralization. Cryobiology 2010;60:45–53. [35] MartÎ F, Miralles A, PeirÔ M, Amill B, de Dalmases C, PiÒol G, et al. Differential effect of cryopreservation on natural killer cell and lymphokine-activated killer cell activities. Transfusion 1993;33:651–5. [36] Berg M, Lundqvist A, McCoy P, Samsel L, Fan Y, Tawab A, et al. Clinical-grade ex vivo-expanded human natural killer cells up-regulate activating receptors and death receptor ligands and have enhanced cytolytic activity against tumor cells. Cytotherapy 2009;11:341–55. [37] de Rham C, Ferrari-Lacraz S, Jendly S, Schneiter G, Dayer J-M, Villard J. The proinflammatory cytokines IL-2, IL-15 and IL-21 modulate the repertoire of mature human natural killer cell receptors. Arthritis Res Ther 2007;9:R125. [38] Moretta A, Bottino C, Vitale M, Pende D, Cantoni C, Mingari MC, et al. Activating receptors and coreceptors involved in human natural killer cell-mediated cytolysis. Annu Rev Immunol 2001;19:197–223. [39] Chrul S, Polakowska E, Szadkowska A, Bodalski J. Influence of interleukin IL-2 and IL-12 ⫹ IL-18 on surface expression of immunoglobulin-like receptors KIR2DL1, KIR2DL2, and KIR3DL2 in natural killer cells. Mediators Inflamm 2006;2006:46957.