Ex vivo expansion of CD56+ cytotoxic cells from human umbilical cord blood

Experimental Hematology 29 (2001) 104–113

Ex vivo expansion of CD56⫹ cytotoxic cells from human umbilical cord blood Reba Condiottia, Yifat Bar Zakaia, Vivian Barakb, and Arnon Naglera,c Departments of aBone Marrow Transplantation and Oncology and the cIsrael National Cord Blood Bank and GenCord Ltd., Hadassah University Hospital, Jerusalem, Israel

b

(Received 15 May 2000; revised 27 August 2000; accepted 31 August 2000)

Objective. The immune-mediated effect of natural killer (NK) and cytotoxic T cells against residual tumor cells previously was shown to prevent relapse and reinduce remission after bone marrow transplantation. Human umbilical cord blood is a rich source of cytotoxic CD56⫹ cells including fetal NK cells (CD16⫺CD56⫹⫹) with high lytic capabilities upon activation with interleukin-2 (IL-2). Cord blood transplantations are reported to be associated with lower risk of graft-vs-host disease, which may jeopardize the graft-vs-leukemia effect. Therefore, our goal was to expand and amplify, ex vivo, cord blood-derived CD56⫹ cell-mediated cytotoxic activity. Materials and Methods. Cord blood-derived CD56⫹ cells were separated using anti-CD56 monoclonal antibody and immunomagnetic beads. The cells were expanded in the presence of irradiated feeder cells and various concentrations of IL-2. Results. Maximal fold expansion (152 ⫾ 29) was achieved on day 22 by culturing the cells in the presence of irradiated autologous lymphocytes. Irradiated murine stromal cells yielded 42 ⫾ fourfold expansion (p ⬍ 0.05). FACS analysis at the peak of expansion revealed that the cells were 96% ⫾ 1% CD56⫹. Interferon-␥ levels significantly decreased throughout the culture period (from 1,034 ⫾ 34 pg/mL to 21 ⫾ 8 pg/mL) as did IL-6 levels (from 11,535 ⫾ 1,452 pg/mL to 323 ⫾ 161 pg/mL) whereas tumor necrosis factor-␣ levels did not change. The expanded cells manifested potent lytic capabilities against K562 and Colo-205 cell lines (70.9% ⫾ 2.0% and 48.2% ⫾ 4.0%, respectively) (n ⫽ 5) (effector-to-target ratio 25:1). Coculturing the expanded NK cells with fresh ALL blasts resulted in 85% ⫾ 1% inhibition of colony growth in methylcellulose (n ⫽ 2). In addition, the CD56⫹ expanded cells induced 44% ⫾ 7.5% apoptosis of K562 target cells (n ⫽ 3). Conclusions. It is possible to effectively expand cord blood-derived CD56⫹ cells, ex vivo, while maintaining their antileukemic capablilities. © 2001 International Society for Experimental Hematology. Published by Elsevier Science Inc. Keywords: Cord blood—Cytotoxic cells—Expansion—Apoptosis—Cytokines

Introduction Human umbilical cord blood transplant (CBT) is associated with a lower risk of acute and chronic graft-vs-host disease (GVHD) [1–5]. The decreased risk of GVHD is considered to be the main advantage of CBT, which recently has been used as an alternative source for allogeneic stem cell transplantation (alloSCT) [6]. One of the reasons for the lower incidence of GVHD may be reduced cytotoxic potential of cord blood (CB)-derived natural killer (NK) and cytotoxic T cells, as well as reduced levels of the Th1 cytokines, which are known to take part in the GVHD mechanistic cascade [7–18]. On the Offprint requests to: Arnon Nagler, M.D., M.Sc., Department of Bone Marrow Transplantation and Israel National Cord Blood Bank and GenCord Ltd., Hadassah University Hospital, Jerusalem, Israel, 91120; E-mail: [email protected]

other hand, cytotoxic T cells and NK cells are the key effector cells mediating the graft-vs-leukemia (GVL) effect, which is used clinically in adoptive cell-mediated immunotherapy to control minimal residual disease and for reinduction of remission in chronic myelogenous leukemia patients who relapse after alloSCT [19–23]. Most clinical and experimental data indicate that the greater the GVL effect after alloSCT, the higher the risk of developing GVHD [24,25]. Reducing the risk of GVHD after alloSCT, either by T-cell depletion or by immunosuppression, is known to lead to an increase in leukemic relapse, which may indicate a decrease in GVL effect [26,27]. Therefore, one of the main concerns in CBT was that the reduced incidence of GVHD observed following CBT would lead to a decrease in the GVL effect and, therefore, an increase in relapse rate. Nevertheless, it has been shown that CB NK and lymphokine activated killer cells are able to lyse

0301-472X/01 $–see front matter. Copyright © 2001 International Society for Experimental Hematology. Published by Elsevier Science Inc. PII S0301-472X(00)0 0 6 1 7 - 2

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

noncultured fresh leukemia blasts, readily respond to interleukin-2 (IL-2) and interleukin-12 (IL-12), and mediate relatively high levels of apoptotic-mediated cytotoxicity against target cell lines [12,13,28–30]. Furthermore, CB is rich in unique NK cell subsets that may possess greater potential proliferative capacities than peripheral blood (PB) NK cells [31,32]. The absolute number of NK cells in CB appears to be similar to adult PB, and these cells may actually have a greater proliferative capacity when exposed to alloantigens or exogenous cytokines [14,28,29,33]. This may suggest that CB has substantial GVL potential and that CB-derived NK cells may be used effectively, if properly amplified, for adoptive cell-mediated immunotherapy and amplification of the GVL effect. In the present study, we attempt to ex vivo expand and amplify CB-derived CD56⫹ cytotoxic cells, taking advantage of their increased proliferative potential. Cells were cultured in the presence of IL-2 and irradiated autologous lymphocytes. Maximum expansion up to 152-fold was observed on day 22. The expanded CD56⫹ cells demonstrated intact cytotoxic capabilities against NK-sensitive and resistant cell lines and fresh leukemic blasts and were able to induce apoptosis of the K562 leukemic cell line.

Methods Cytotoxic cell separation from CB CB was obtained from the umbilical cords of the placentas of normal, full-term, nonstressed newborns of concenting mothers, as previously described [30]. CB was collected in 25 mL of citrate phosphate dextrose adenine (CPDA) by venipuncture of the umbilical cord immediately following delivery of the baby and cutting and clamping of the cord while the placenta remains in utero. The blood (80–100 mL) was diluted 1:2 in phosphate-buffered saline (PBS; Biological Industries, Kibbutz Beit Haemek, Israel) mixed gently, and diluted 1:2 in a 3% gelatin (Sigma Chemical Corp., St. Louis, MO, USA) solution as previously described [33]. After a 20-minute incubation at room temperature, the upper phase was gently layered onto Histopaque-1077 (Sigma Chemical Corp.) for density centrifugation at 1,800 rpm for 20 minutes. The resulting cells were washed two times in PBS and approximately 30 ⫻ 106 cells were removed to be used as autologous irradiated “feeder cells” after completion of the separation (see section on Cytotoxic Cell Expansion). The remaining cells were resuspended in 2 mL of 30% Percoll (Pharmacia, Uppsala, Sweden) and layered on 9 mL of a 40% Percoll solution and centrifuged at 1,800 rpm for 30 minutes to remove small T lymphocytes. The upper phase was washed two times in PBS and resuspended in 1–2 mL of cold RPMI 1640 containing 10% fetal calf serum (FCS; Gibco, Israel) 1 mM L-glutamine, 10 U/mL penicillin, and 0.01 mg/mL streptomycin (complete medium; Biological Industries). Mouse anti-human CD56 monoclonal antibody (mAb; Pharmingen, San Diego, CA, USA) (1 ␮g/1 ⫻ 106 cells) was added to the cell suspension and incubated on ice for 30 minutes. Following three washes with cold complete medium, magnetic beads conjugated to sheep anti-mouse IgG (Dynal, Norway) (one bead/cell) were added to the cells. This cell-bead suspension was incubated for 30 minutes ⫻ 10 rpm at 4⬚C using an end-over-end rotator. The sus-

105

pension was exposed to a magnet (MPC E1; Dynal) for 3 minutes and the cell-bead conjugates were resuspended in 1 mL of complete medium and incubated for 18–24 hours at 37⬚C in a humidified incubator to release the beads from the target cells. These cells were determined by FACS analysis to be 77% ⫾ 2.5% CD56⫹, 15% ⫾ 5% CD3⫹, 3% ⫾ 1% CD19⫹, and 2% ⫾ 1% CD14⫹. Cyotoxic cell separation from PB Cytotoxic cells from PB were isolated as from CB, except the gelatin step was omitted. Buffy coat preparations obtained from consenting normal donors (Hadassah University Hospital Blood Bank) were diluted 1:4 with PBS containing 10 U/mL heparin (CSL, Melbourne, Australia). PB mononuclear cells (MNC) were isolated by Histopaque-1077 (Sigma Chemical Corp.) for density centrifugation at 1,00 rpm for 20 minutes. The remaining procedure for isolation of cytotoxic cells was performed as described earlier. Cytotoxic cell expansion Two types of feeder layers were used to expand cytotoxic cells: (1) irradiated, autologous lymphocytes, and (2) irradiated murine stromal cells. Autologous lymphocytes. CB lymphocytes (following gelatin and Ficoll separation as described earlier) were irradiated (2,000 cGy) and seeded in a 24-well plate with 1 ⫻ 105 separated cytotoxic cells at a ratio of 20:1 in 1 mL of RPMI 1640 (Sigma Chemical Corp.) or Ham’s F12-based NK medium (Gibco, Israel) containing increasing concentrations of recombinant human IL-2 (1–1,000 IU/mL) (Chiron, San Francisco, CA, USA), 10% heat-inactivated human AB serum (Magen David Adom Blood Bank), 1 mM L-glutamine, 10 U/mL penicillin, and 0.01 mg/mL streptomycin (expansion medium; Biological Industries). On day 7, the volume was doubled to 2 mL with fresh expansion medium. Beginning on day 10, the cells were counted every 4 days. In cultures that contained ⬍2 ⫻ 106 cells/mL, half of the expansion medium was removed and replaced with fresh medium. In cultures that contained ⬎2 ⫻ 106 cells/mL, the cells were split 2:1 as needed to maintain cell concentrations below 2–3 ⫻ 106 cells/mL. In preliminary experiments, we assessed fold expansion up to day 30. Analyses of cytotoxicity, phenotype, and apoptotic ability were performed on cells from 26-day cultures. Murine stromal cell line. M210-B4 stromal cells (kindly provided by Dr. J. Miller, University of Minnesota, Minneapolis, MN, USA) were trypsinized and seeded at 2 ⫻ 105 cells/mL in six-well tissue culture plates (Nunclon, Denmark) for 48 hours or until they reached ⵑ80% confluency. The stromal cells were irradiated (6,000 cGy), and the separated cytotoxic cells were seeded over the feeder layer at 2–4 ⫻ 105 cells/well in 3 mL of expansion medium, as previously described [34]. Cytotoxic cells were passaged and counted starting on day 7 as described earlier. Surface antigens Cells 1 ⫻ 106 were stained with 1 ␮g fluorescein isothiocyanate (FITC) or phycoerythrin (PE)-conjugated mAbs for 20 minutes at 4⬚C and then washed in PBS containing 0.02% azide and 1% bovine serum albumin as previously described [35]. The cells were stored at 4⬚C in 0.5 mL of 1% paraformaldehyde until analysis. The following mAbs were used: control mouse immunoglobulin G (IgG) (FITC and PE), CD3 FITC, CD8 FITC, CD14 FITC, CD19 FITC, CD4 PE, and CD56 PE, CD69 FITC (Becton-Dickinson, San Jose, CA, USA), CD25 FITC, and CD71 FITC (Pharmingen). Five thousand events were acquired in list mode with a threshold

106

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

on forward light scatter and a lymphocyte configuration on forward light scatter vs orthogonal light scatter. Compensation of FITC and PE into other than the assigned detectors was obtained by samples stained with CD8 FITC and CD8 PE. Cytotoxicity assay Cytotoxic activity was measured in a standard 4-hour 51Cr (Rotem Industries, Negev, Israel) release assay. Effector cells were expanded cytotoxic cells. A constant number of target cells (K562, ATCC CCL 243 [an NK-sensitive erythroleukemia] or Colo-205, ATCC CCL222 [an NK-resistant colon adenocarcinoma]) was added to serial dilutions of effector cells to obtain different effector-to-target (E:T) ratios. Spontaneous 51Cr release was measured by incubating target cells in the absence of effector cells. Maximum 51Cr release was determined by adding 0.2 mL of 1% Triton X-100 detergent (Sigma Chemical Corp.) to the target cells. Percent lysis was calculated according to the following formula: [cpm (experimental release) ⫺ cpm (spontaneous release)]/[cpm (maximum release) ⫺ cpm (spontaneous release)] ⫻ 100. Apoptosis Expanded CB cytotoxic cells were cocultured with K562 erythroleukemia cells at a 5:1 ratio for 3 hours at 37⬚C in a final volume of 0.5 mL complete medium. Cytotoxic cells subsequently were removed by using anti-CD56 and immunomagnetic beads as described earlier in the section on Cytotoxic Cell Separation from CB. The remaining K562 cells were stained with PE-conjugated antiCD56 to ascertain that the cytotoxic cells were removed. Apoptosis of the target cells was measured as previously described [36]. Briefly, K562 cells (1 ⫻ 106) were centrifuged and the pellet was resuspended in 1 mL of cold ethanol. Following 1-hour incubation, the cells were washed and resuspended in PBS (0.9 mL). RNAse 50 ␮L (10 mg/mL) was added and cells were incubated for 30 minutes at 37⬚C. Propidium iodide (PI) 50 ␮L (0.5 mg/mL) was added and FACS analysis was performed after acquiring 10,000 events. Detection of cytotoxic cell-secreted cytokines Supernatants were collected from expanded cytotoxic cells at various time points and frozen at ⫺80⬚C. Levels of interferon-␥ (IFN-␥), tumor necrosis factor-␣ (TNF-␣), and interleukin-6 (IL6) were determined by enzyme immunoassay test kits (R&D, Mineapolis, MN, USA) as previously described [37]. Leukemic blast colony-forming unit assays PB 60 mL was drawn into a heparinized syringe prior to any form of treatment from consenting patients diagnosed with T-cell acute lymphoid leukemia (T-ALL; FAB classification L1) with 95– 100% blasts present in the PB. The blood was diluted 1:4 in PBS, layered onto Histopaque, and centrifuged for 20 minutes at 1,800 rpm. The cells were washed twice in PBS and stored in 10% DMSO at ⫺80⬚C. The leukemic cells were thawed and assayed in 0.3% methylcellulose cultures. Total cells 2 ⫻ 105 were plated in three aliquots of 0.33-mL methylcellulose (StemCell Technologies, Vancouver, British Columbia, Canada) and 100 ␮L of bladder carcinoma supernatant, as previously described [33]. Blast colonies were counted on day 12 with an inverted microscope and identified by morphology and cytospin preparation. Lysis of leukemic colonies by cytotoxic cells To determine in vitro lysis of leukemic cells by cytotoxic cells, expanded CD56⫹ cytotoxic cells were cocultured with the leukemia

blasts at E:T ratios of 5:1 and 1:1. The cytotoxic cells remained in contact with the leukemic cells for 4 hours and 24 hours in 300 ␮L of complete medium at 37⬚C in a humidified atmosphere following centrifugation (3 minutes ⫻ 500 rpm) as previously described [38]. Following incubation, target cells were seeded in methylcellulose as described earlier, and blast colonies were counted on day 12 and identified by morphology and cytospin preparation. Statistics Results of experimental points from multiple experiments are reported as mean ⫾ SE. Significance levels were determined by twotailed Student’s t-test for differences in means.

Results Optimal conditions for cytotoxic cell expansion Expansion of CB CD56⫹ cells using irradiated autologous lymphocytes vs a murine stromal cell line. CB yielded 1.1 ⫾ 0.2 ⫻ 106 cells following separation (see Materials and Methods) (n ⫽ 10). The purified cells were expanded for up to 30 days in the presence of either irradiated autologous lymphocytes or the murine fibroblast cell line, M2-10B4, and IL-2 (1,000 IU/mL). As shown in Figure 1A, CD56⫹ cells grown in the presence of autologous lymphocytes showed a 56 ⫾ 18-fold expansion on day 10. Significant growth increase continued until day 22, when cells underwent a 152 ⫾ 29-fold expansion (p ⬍ 0.05). Thereafter, cell expansion reached a plateau that was observed until day 30 of culture, which was the final harvest day (n ⫽ 4). CD56⫹ cytotoxic cells expanded in the presence of murine stroma exhibited 12 ⫾ fourfold expansion on day 10, with a significant growth increase reaching 55 ⫾ eightfold expansion on day 18 (p ⬍ 0.05). From day 18 to day 30, expansion remained at a plateau and started to decline, exhibiting a 26 ⫾ sevenfold expansion on day 30 (n ⫽ 3) (Fig. 1A). As shown in Figure 1A, the presence of autologous lymphocytes lead to a significantly greater expansion of CBderived cytotoxic cells than murine stroma feeder cells on days 10, 14, 18, and 22 (p ⬍ 0.05). On days 26 and 30, cytotoxic cell expansion continued to be greater using autologous lymphocytes, although the difference was not significant. Therefore, the remaining experiments were performed using irradiated autologous lymphocytes for the expansion of CB-derived cytotoxic cells. Dose response of CB cytotoxic cells to IL-2. To ascertain the concentration of IL-2 needed for optimal expansion of CB cytotoxic cells, dose-response experiments were performed. As shown in Figure 1B, on days 10 and 14, 100 IU/ mL IL-2 led to a significantly greater number of CD56⫹ cells than 10 IU/mL (13.9 ⫾ 3.1 vs 3.9 ⫾ 1.3 p ⬍ 0.05; and 30.5 ⫾ 3.9 vs 4.9 ⫾ 1.6, p ⬍ 0.01, respectively). On day 18, there was no significant difference in CD56⫹ cell expansion between IL-2 concentrations used. Cytotoxic cell growth on days 22 and 26 revealed a significantly greater expansion with 500 IU/mL than with 100 IU/mL (95.3 ⫾ 20.1 vs 27.8 ⫾

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

107

16.1, p ⬍ 0.05; and 89.5 ⫾ 8.5 vs 11.7 ⫾ 5.7, p ⬍ 0.01, respectively). IL-2 1,000 IU/mL also led to a greater fold expansion than 100 IU/mL on days 22 and 26 (164.7 ⫾ 41.4 vs 27.8 ⫾ 16.1, p ⬍ 0.05; and 105.2 ⫾ 9.2 vs 11.7 ⫾ 5.7, p ⬍ 0.01, respectively) (n ⫽ 3). Therefore, the remaining experiments were performed in the presence of 1,000 IU/mL IL-2. Optimal type of medium for CD56⫹ cytotoxic cell expansion. CB CD56⫹ cells were expanded in the presence of RPMI 1640 or Ham’s F12 NK medium (n ⫽ 7) to establish the type of media best suited for this purpose. As shown in Figure 1C, RPMI 1640 led to a significantly higher fold expansion of CB CD56⫹ cells throughout the culture period beginning on day 10 (44.0 ⫾ 7.9 vs 8.5 ⫾ 2.0, p ⬍ 0.05). The difference remained significant until the last cell count on day 26 (156.3 ⫾ 30.3 vs 55.2 ⫾ 2.0, p ⬍ 0.01). This phenomenon may be due to the difference in components contained in these two media. Ham’s F12 NK medium contains several inorganic salts (e.g., calcium chloride and magnesium chloride) and two amino acids, L-alanine and L-arginine-HCl, which are not found in RPMI 1640.

Figure 1. Optimal conditions for expansion of CD56⫹ cells derived from CB. (A) Isolated CD56⫹ cytotoxic cells were cultured in the presence of irradiated autologous lymphocytes (1:20) ( ) or the murine stroma cell line M210-B4 (2 ⫻ 105 stroma cells/2–4 ⫻ 105 CB cells) ( ) and IL-2 (1,000 IU/mL). On days 10, 14, 18, 22, 26, and 30, cells were counted and fold expansion was determined. *p ⬍ 0.05. (B) Cytotoxic cells were cultured in the presence of irradiated autologous lymphocytes (1:20) and various concentrations of IL-2 (0-1000 IU/mL). On days 10, 14, 18, 22, and 26 of culture, cells were counted and fold expansion was determined. *p ⬍

Surface antigen expression To characterize the subpopulations of cells undergoing expansion, cells were stained with T-cell– and NK-cell–associated surface markers, and FACS analysis was performed. As shown in Figure 2A, 77.0% ⫾ 2.5% of the cells were CD56⫹ prior to expansion; 96% ⫾ 1% stained positively for CD56 following 26 days in culture (n ⫽ 5). Similarly, the percentage of CD56⫹ cells that were CD3⫺ increased significantly from 48.7% ⫾ 2.6% on day 0 to 81% ⫾ 1% on day 26 (n ⫽ 4). There also was an observed increase in the CD56⫹CD8⫹ population (day 0:43 ⫾ 12%, n ⫽ 3; day 26: 53 ⫾ 9%, n ⫽ 4), although the increase was not significant. Throughout the culture period, fluorescence intensity of the CD56⫹ population became increasingly bright (Fig. 2B). Whereas on day 0, 44.0% ⫾ 21.0% of the cells were CD56dim and 23.0 ⫾ 0% were CD56bright, on day 26, the CD56dim population decreased to 8.2 %⫾ 1.9% and the CD56bright subset increased to 91.4% ⫾ 2.1% (n ⫽ 5). Mean fluorescence intensity of the dim and bright population on day 0 was 45.5 ⫾ 2.5 and 1023 ⫾ 605.1, respectively. On day 26, the mean fluorescence intensity of the dim and bright populations was 154.6 ⫾ 33.5 and 2257.6 ⫾ 394.4, respectively (n ⫽ 5). Absolute numbers of the CD56⫹ population on day 0 was 0.68 ⫻ 106 and was divided among 0.48 ⫻ 106 CD56dim cells and 0.20 ⫻ 106 CD56bright cells. On day 26, the absolute numbers of CD56⫹ cells were 252.9 ⫻ 106 cells, in

0.05; **p ⬍ 0.01. (C) CB-derived CD56⫹ cells were cultured in the presence of irradiated autologous lymphocytes (1:20) and IL-2 (1,000 IU/mL) in RPMI 1640 ( ) or Ham’s F12-based NK medium ( ). Fold expansion was calculated by counting the number of cells present on days 10, 14, 18, 22, and 26 of culture. *p ⬍ 0.05; **p ⬍ 0.01.

108

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

which 26.3 ⫻ 106 cells were CD56dim and 226.6 ⫻ 106 cells were CD56bright. The average absolute cell number of CD8⫹CD56⫹ cell on days 0 and 26 was 0.38 ⫻ 106 and 55.7 ⫻ 106 cells, respectively (n ⫽ 3). Expression of activation antigens on the surface of CD56⫹ cells was measured, and 81% ⫾ 5% of the CD56⫹ cells were found to be positive for the leukocyte early activation marker CD69, whereas 15% ⫾ 3% were positive for transferrin receptor (CD71) and only 6% ⫾ 2% of the cells expressed CD25, the IL-2 receptor (n ⫽ 3). Fold expansion of CD56⫹ cells We evaluated absolute cell number and viability of expanded CB cytotoxic cells on days 10, 14, 18, 22, and 26 of

culture (Fig. 3). Absolute cell numbers were 41.9 ⫾ 10.1 ⫻ 106, 85.1 ⫾ 30.7 ⫻ 106, 120.2 ⫾ 35.5 ⫻ 106, 137.4 ⫾ 47.8 ⫻ 106, and 190.6 ⫾ 61.7 ⫻ 106, respectively (n ⫽ 10). Fold expansions were 44.0 ⫾ 7.9, 83.1 ⫾ 14.6, 103.3 ⫾ 14.0, 160.5 ⫾ 29.4, and 156.3 ⫾ 30.3, respectively (n ⫽ 10). In all cases, viability was measured using trypan blue exclusion and found to be ⬎95%. In comparison, we assessed the same parameters in cells derived from PB (Fig. 3). Absolute cell numbers were 26.1 ⫾ 9.4 ⫻ 106, 44.4 ⫾ 11.9 ⫻ 106, 71.6 ⫾ 21.2 ⫻ 106, 89.1 ⫾ 23.1 ⫻ 106, and 130.0 ⫾ 43.6 ⫻ 106 on days 10, 14, 18, 22, and 26, respectively (n ⫽ 5). Viability was assessed on the days that fold expansion was evaluated and found to be ⬎95% by trypan blue exclusion.

Figure 2. FACS analysis of surface antigen expression before and after expansion of CB-derived cytotoxic cells. Prior to culture or on day 26, cells isolated from CB were stained with FITC- or PE-conjugated mAbs (1 ␮g/1 ⫻ 106 cells) or appropriate isotype controls as described in the Methods. Compensation was obtained by samples stained with CD8 FITC and CD8 PE. (A) Expression of surface antigens before ( ) and after (䊏) expansion of CD3⫺ and CD8⫹, subsets of the CD56⫹ population (n ⫽ 5). (B) Representative histogram analysis of the CD56dim and CD56bright populations before (left panel) and after (right panel) expansion.

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

109

(n ⫽ 4). TNF-␣ levels did not change significantly throughout the culture: 247 ⫾ 24 and 238 ⫾ 43 on day 10 and day 26, respectively (n ⫽ 4). Levels of IFN-␥ and IL-6 from CB cultures were significantly higher than those from cultures of PB on day 10 (p ⬍ 0.05), although by day 26 there was no significant difference (Table 1). The levels of TNF-␣ remained stable and were similar between CB- and PB-derived cells.

Figure 3. Cytotoxic cell expansion from CB vs PB. Cytotoxic cells were isolated from CB ( ) and PB ( ) and cultured in the presence of irradiated autologous lymphocytes (1:20) and IL-2 (1,000 IU/mL). On days 10, 14, 18, 22, and 26 of culture, cells were counted and fold expansion was determined. *p ⬍ 0.05.

CD56⫹ cells from CB were shown to have a significantly higher expansion capability than CD56⫹ cells derived from PB on days 14, 18, and 22 of culture (p ⬍ 0.05) (Fig. 3). Cytokine secretion from expanded cytotoxic cells On days 10, 18, and 26 of culture, supernatants were collected and frozen for determination of cytokine levels secreted from expanded cells. As shown in Table 1, IFN-␥ levels were high on day 10 (1,034 ⫾ 34 pg/mL) as compared to fresh CB (3.2 ⫾ 1.1 pg/mL) and gradually decreased to low levels by day 26 (21 ⫾ 8 pg/mL) (n ⫽ 4). The same phenomenon was observed with levels of IL-6 in which there were 11,535 ⫾ 1,452 pg/mL on day 10 as compared to 3.69 ⫾ 1.06 pg/mL in fresh CB and decreased to 323 ⫾ 161 on day 26 (n ⫽ 4). TNF-␣ levels were slightly higher on day 10 than in fresh CB serum (287 ⫾ 58 pg/mL vs 10.1 ⫾ 3.05 pg/mL), although they remained constant throughout the culture period, revealing 215 ⫾ 28 pg/mL on day 26 (n ⫽ 4). In parallel, supernatants were collected from cultures containing cytotoxic cells derived from PB and the cytokine levels were measured. IFN-␥ levels of PB-derived cells were 259 ⫾ 93 pg/mL on day 10 and slightly decreased to 67 ⫾ 38 pg/mL on day 26 (n ⫽ 4). IL-6 levels were 2,548 ⫾ 974 pg/mL and 61 ⫾ 14 on day 10 and day 26, respectively

Cytolytic ability of expanded CD56⫹ cells Target cell lysis. Following expansion of CD56⫹ cells derived from CB and PB, their functional capability was measured in a standard 51Cr release assay using two types of targets: an NK-sensitive erythroleukemia cell line K562, and an NK-resistant colon carcinoma cell line Colo-205. At the end of the 26-day culture period, CD56⫹ cells derived from CB were able to lyse 70.9% ⫾ 2.0% of K562 target cells at an E:T ratio of 25:1 (n ⫽ 5) (Fig. 4A). Cytolysis of Colo205 target cells was 48.2% ⫾ 4.0% (n ⫽ 5) (Fig. 4B) at an E:T ratio of 25:1. In comparison, expanded CD56⫹ cells derived from PB lysed 68.3% ⫾ 15.9% of K562 target cells at an E:T ratio of 25:1 after 26 days in culture (Fig. 4A), whereas cytolysis of Colo-205 was 36% ⫾ 18% at the same E:T ratio (n ⫽ 3) (Fig. 4B). As shown in Figure 4, there was no significant difference between the capability of CD56⫹ cells derived from CB and PB to lyse K562 target cells (CB 70.9% ⫾ 2.0% vs PB 68.3% ⫾ 15.9%), whereas CD56⫹ cells derived from CB had the ability to lyse Colo-205 cells slightly better than those derived from PB (E:T ratio 25:1 CB 48.2% ⫾ 4.0% vs PB 36% ⫾ 18%, NS). Apoptosis. Following 26 days in culture, expanded CB CD56⫹ cells were cocultured with K562 target cells for 3 hours. Following mechanical separation of the two cell populations (see Methods), the percentage of apoptosis in the target cell population was measured by PI staining. CD56⫹ cells caused 44% ⫾ 7.5% of the target cells to undergo apoptosis (n ⫽ 3). A representative histogram FACS analysis of apoptotic events is shown in Figure 5. Time kinetics studies were carried out to determine the length of NK cell/K562 cell coculture needed to lead to target cell apoptosis. Following 20 minutes of coculture, 40.3% of the target cells were apoptotic. The percentage of K562 cells that under-

Table 1. Cytokine secretion from CB- and PB-derived expanded cytotoxic cells IFN-␥

Day 0 Day 10 Day 18 Day 26

IL-6

TNF-␣

CB

PB

CB

PB

CB

PB

3.2 ⫾ 1.1 1033.5 ⫾ 33.5 34.5 ⫾ 7.9 21.0 ⫾ 8.3

300.0 ⫾ 40.0 258.8 ⫾ 93.3 222.0 ⫾ 124.9 67.3 ⫾ 37.5

3.7 ⫾ 1.1 11534.5 ⫾ 1451.6 1443.5 ⫾ 243.8 323.0 ⫾ 161.4

10.0 ⫾ 0.5 2548.0 ⫾ 974.4 466.3 ⫾ 135.3 60.7 ⫾ 13.9

10.1 ⫾ 3.1 286.5 ⫾ 57.9 223.8 ⫾ 20.1 215.0 ⫾ 27.7

175.5 ⫾ 20.3 246.8 ⫾ 24.3 254.3 ⫾ 38.0 237.8 ⫾ 43.2

Values are given in picograms per milliliter.

110

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

Figure 4. Cytotolytic ability of expanded cytotoxic cells derived from CB ( ) and PB ( ). Following expansion of CD56⫹ cells, their functional capability was measured in a standard 4-hour 51Cr release assay using (A) the NK-sensitive erythroleukemia cell line K562 and (B) the NK-resistant colon carcinoma cell line Colo-205.

went apoptosis did not significantly differ after 60 or 180 minutes of contact between effector and target cells (51.3% and 42.6%, respectively). Expanded CD56⫹ cell inhibition of leukemic cell growth. Blasts derived from patients diagnosed with T-ALL (FAB classification ALL-L1) (95–100% blasts in PB) were cocultured for 4 and 24 hours with expanded CB cytotoxic cells at E:T ratios of 5:1 and 1:1. Following incubation, cells were seeded in methylcellulose to measure leukemic colonies as previously described [38,39] and colonies were morphologically identified and counted on day 12. At an E:T ratio of 5:1, expanded CB CD56⫹ cells caused 67% ⫾ 1% inhibition in the growth of leukemic colonies following a 4-hour coculture and 85% ⫾ 1% inhibition after 24 hours of effector/target contact (n ⫽ 2). Conclusions In the present study, we defined the optimal conditions for ex vivo expansion and amplification of CB-derived CD56⫹

Figure 5. Apoptosis of an erythroleukemia cell line by expanded cytotoxic cells derived from CB. Following 26 days in culture, expanded CD56⫹ cells were cocultured with K562 target cells for 3 hours. Following separation of the two cell populations as described in the Methods, the percentage of apoptosis in the target cell population was measured by PI staining and FACS analysis. Representative histograms of percent apoptosis of untreated target cells (control, upper panel) and target cells after contact with expanded NK cells (lower panel) are shown.

cytotoxic cells. We were able to achieve ⬎150-fold expansion using RPMI, irradiated autologous lymphocytes, and 500–1,000 IU/mL IL-2. Maximum expansion was observed on day 26 of culture; however, clinically relevant expansion was already observed on day 18. Moreover, ex vivo expansion of CD56⫹ cytotoxic cells was significantly greater for CB-derived cytotoxic cells than for PB-derived cells. The number of CD56⫹ cytotoxic cells following expansion using the present protocol is 130 ⫻ 106 cells. This is a clinically relevant number if a patient’s weight is 70-100 kg; the number of cytotoxic cells following expansion would be 1-2 ⫻ 106 cells/kg. We previously showed that 0.1-1 ⫻ 106 T cells/kg may be enough to prevent relapse and may even be sufficient to reinduce remission [21,22]. The higher ex vivo expansion potential of CB-derived CD56⫹ cells is in accordance with our and other previous publications indicating high proliferative potential and sensitivity to IL-2 of CD56brightCD16⫺ NK cell subsets, which are the main population of NK cells in CB and fetal tissues

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

while representing a miniscule population of the NK cells in the PB [32,40-42]. Our ability to reach substantial expansion of CD56⫹ cytotoxic cells also may have to do with the fact that cultures were started with relatively purified cells and eliminated putative suppressor cell subsets. Instead of monocytes, which previously were shown to be of importance in expansion of NK cells from PB [43,44], we used irradiated autologous lymphocytes in the present study. Coculture of NK progenitor cells with irradiated PB MNC or lymphocyte conditioned media previously was shown to expand and induce NK cell proliferation [34,42,45]. Subsequently, it was shown that NK cell expansion is regulated by transforming growth factor-␤ (TGF-␤), and it may be independent of accessory cell-derived contact [34]. The other conditions that were presently used for optimizing ex vivo expansion of CB-derived CD56⫹ cytotoxic cells were somewhat different than the conditions that previously were reported for expansion of PB cytotoxic cells. Pierson et al. [46] previously demonstrated that sorted PB NK cells cultured for 18 days with 10% AB serum without accessory cells in a 2:1 DMEM/F12 basal medium expanded 1.7-fold more than when cultured in standard RPMI 1640 basal medium. The difference may be also due to the fact that we used Ham’s F12 and not DMEM/F12 and other culture conditions including irradiated autologous lymphocytes that were the crucial factor for the expansion. In accordance, we observed similar ex vivo expansion potential of PB-derived CD56⫹ cells on murine M2-10B4 stroma and irradiated autologous lymphocytes, whereas CB-derived CD56⫹ cells expanded significantly better with autologous lymphocytes than with irradiated M2-10B4 stroma feeder layer. Irradiated stroma from various sources, including murine stroma such as M2-10B4, AFT024, or allogeneic human stroma, is of vast importance for amplification of NK cell proliferation and differentiation [47–50]. That we were able to replace irradiated heterologous stroma by irradiated autologous lymphocytes is an obvious clinical advantage. As for the IL-2 concentration used, it is in accordance with previously published data [34]. Our culture conditions gave growth advantage to the expanded CD56⫹ cytotoxic cells, as CD56 expression increased while CD3 expression decreased with culture time. The increase in CD56 expression with culture time is in accordance with previous publications indicating that CB is rich in NK progenitor cells, which are immature but readily respond to IL-2 and IL-12 [13,28]. It also may indicate expansion of the CD56bright NK cell subset, which is very sensitive to IL-2, possesses high-affinity IL-2 receptors and has very high proliferative capacity [32,40–42]. The observed shift from CD56dim to CD56bright may be a result of the presence of IL-2 in the culture. These expanded cells expressed also CD16. CD69 is a prominent early activation marker for cytotoxic cells [30,51,52]. In our study, high expression of this antigen was observed following ex vivo expansion, indicating priming and activation of the proliferating CD56⫹

111

cell, which is of importance for amplifying its killing capabilities and cytotoxic functions. Cytokines, especially Th1 cytokines, are likely to be important factors secreted from the irradiated autologous lymphocytes, which induce CD56⫹ cell expansion and activation. In our expanded CB cultures, we observed increased levels of IFN-␥ and TNF-␣ on day 10 as compared to fresh CB, which may be due to the presence of exogenous IL-2. It is known that CB is sensitive to IL-2, and activated cells are known to produce increased levels of IFN-␥ and IL-6 [30,32,41,53,54]. These cytokines are known to activate NK-mediated cytotoxicity and proliferative potential, relative to CB plasma [40,55]. Increased levels of up to 10 known cytokines were previously reported in expanded CB cultures, including IL-2 and granulocyte-macrophage colony-stimulating factor [56]. As we reported previously, these cytokines, especially in combination, are strong stimulatory factors inducing cell proliferation and ex vivo expansion of PB- and CB-derived hematopoietic and immunocompetent cells [57,58]. Importantly, the expanded CB-derived CD56⫹ cells exhibited slightly higher killing capabilities of the NK-resistant cell line Colo-205, whereas comparable cytotoxicity to mature PB adult NK cells was observed against the NK sensitive cell line K562. These differences in killing capabilities are worthy of further investigation. These findings are in accordance with previous publications that CB-derived NK progenitor cells can be effectively activated by IL-2 and other cytokines [13,28,36,59]. Even more importantly, the CB-derived ex vivo expanded CD56⫹ cytotoxic cells demonstrate killing ability against fresh leukemic cells, as we and others previously reported for IL-2–activated mature NK cells [39,60]. The expanded cells were able to induce apoptosis of the K562 erythroleukemia cell line using the PI DNA labeling technique. This technique previously was demonstrated to be a reliable assay to evaluate NK-mediated apoptosis [36]. The level of apoptosis induced by the expanded cells was comparable to those observed in IL-2–activated fresh CB and mature PB NK cells [36,61,62]. Moreover, the kinetics of induction of apoptosis by the ex vivo expanded CBderived CD56⫹ cells was similar to the kinetics observed by Rodella et al. [62] using adult mature PB NK cells. The number of CD56⫹ cytotoxic cells that may be required to mediate significant GVL activity reinducing remission and preventing relapse following CBT depends mainly on the cytolytic activity. That we were able to demonstrate high efficiency of CB-derived CD56⫹ expanded cells in both cytotoxic capabilities against tumor cell lines including NK-resistant tumor cell lines as well as against fresh leukemia cells, and in induction of apoptosis may indicate that the expanded CB-derived cytotoxic cells may be of clinical relevance. Similarly, the fact that we observed significant expansion within 2.5 weeks is advantageous. These findings may have important implications for the emerging

112

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

use of CB as an alternate source of transplantable hematopoietic progenitor cells with lower incidence of GVHD but an inducible potential for GVL. Induction of GVL without increasing the risk for GVHD is the ultimate goal in clinical stem cell transplantation. CB has been previously proven to be a rich source of hematopoietic stem cells for transplantation, and we now show that it also contains a reservoir of CD56⫹ cytotoxic cells that have the potential to be amplified ex vivo for possible use in adoptive cellular immunotherapy for inducing an antileukemic effect in the state of minimal residual disease after CBT.

References 1. Wagner JE, Kernan NA, Steinbuch M, Broxmeyer HE, Gluckman E (1995) Allogeneic sibling umbilical-cord-blood transplantation in children with malignant and non-malignant disease. Lancet 346:214 2. Kurtzberg J, Laughlin M, Graham ML, et al. (1996) Placental blood as a source of hematopoietic stem cells for transplantation into unrelated recipients. N Engl J Med 335:157 3. Wagner JE, Rosenthal J, Sweetman R, et al. (1996) Successful transplantation of HLA-matched and HLA-mismatched umbilical cord blood from unrelated donors: analysis of engraftment and acute graftvs-host disease. Blood 88:795 4. Gluckman E, Rocha V, Boyer-Chammard A, et al. (1997) Outcome of cord-blood transplantation from related and unrelated donors. N Engl J Med 337:373 5. Rubinstein P, Carrier C, Scaradavou A, et al. (1998) Outcomes among 562 recipients of placental-blood transplants from unrelated donors. N Engl J Med 339:1565 6. Rocha V, Wagner JE, Sobocinski KA, et al. (2000) Comparison of graft vs host disease in children transplanted with HLA-identical sibling umbilical cord blood vs HLA-identical sibling bone marrow. N Engl J Med 342:1846 7. Ferrara JLM, Deeg HJ (1991) Graft vs host disease. N Engl J Med 324:667 8. Ghayur T, Seemayer TA, Lapp WS (1987) Kinetics of natural killer cell cytotoxicity during the graft vs host reaction. Transplantation 44:254 9. Rhoades JL, Cibull ML, Thompson JS, et al. (1993) Role of natural killer cells in the pathogenesis of human acute graft vs host disease. Transplantation 56:113 10. Risdon G, Gaddy J, Stehman FB, Broxmeyer HE (1994) Proliferative and cytotoxic responses of human cord blood T lymphocytes following allogeneic stimulation. Cell Immunol 154:14 11. D’arena G, Musto P, Cascavilla N, et al. (1999) Inability of cord blood T lymphocytes to perform Th1-like and Th2-like responses: implications for transplantation. J Hematother Stem Cell Res 8:381 12. Dominguez E, Madrigal JA, Layrisse Z, Cohen SBA (1998) Fetal natural killer cell function is suppressed. Immunology 94:109 13. Gaddy J, Broxmeyer HE (1997) Cord blood CD16⫹CD56⫺ cells with low lytic activity are possible precursors of mature natural killer cells. Cell Immunol 180:132 14. Harris DT, Schumacher MJ, LoCascio J, Booth A, Bard J, Boyse EA (1994) Immunoreactivity of umbilical cord blood and post-partum maternal peripheral blood with regard to HLA-haploidentical transplantation. Bone Marrow Transplant 14:63 15. Webb BJ, Bochan MR, Montel A, Padillla LM, Brahmi Z (1994) The lack of NK cytotoxicity associated with fresh HUCB may be due to the presence of soluble HLA in the serum. Cell Immunol 159:246 16. Cohen SBA, Perez-Cruz I, Fallen P, Gluckman E, Madrigal JA (1999)

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30.

31.

32.

33.

34.

35.

Analysis of the cytokine production by cord and adult blood. Human Immunol 60:331 Sautois B, Fillet G, Beguin Y (1997) Comparative cytokine production by in vitro stimulated mononucleated cells from cord blood and adult blood. Exp Hematol 25:103 Chalmers IMH, Janossy G, Contreras M, Navarrete C (1998) Intracellular cytokine profile of cord and adult blood lymphocytes. Blood 92:11 Hercend T, Takvorian T, Nowill A, et al. (1995) Characterization of natural killer cells with anti-leukemia activity following allogeneic bone marrow transplantation. Blood 67:722 Kolb HJ, Mittermuller J, Clemm C, et al. (1990) Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 76:2462 Naparstek E, Or R, Nagler A, et al. (1995) T-cell-depleted allogeneic bone marrow transplantation for acute leukemia using Campath-1 antibodies and post-transplant administration of donor’s peripheral blood lymphocytes for prevention of relapse. Br J Haematol 89:506 Slavin S, Naparstek E, Nagler A, et al. (1996) Allogeneic cell therapy with donor peripheral blood cells and recombinant human interleukin-2 to treat leukemia relapse after allogeneic bone marrow transplantation. Blood 87:2195 MacKinnon S, Papadopoulos EB, Carabasi MH, et al. (1995) Adoptive immunotherapy evaluating escalating doses of donor leukocytes for relapse of chronic myeloid leukemia after bone marrow transplantation: separation of graft-vs-leukemia responses from graft-vs-host disease. Blood 86:1261 Weiden PL, Flournoy N, Thomas ED, et al. (1979) Antileukemic effect of graft-vs-host disease in human recipients of allogeneic-marrow grafts. N Engl J Med 300:1068 Weiden PL, Sullivan KM, Flournoy N, Storb R, Thomas ED (1981) Antileukemic effect of chronic graft-vs-host disease: contribution to improved survival after allogeneic marrow transplantation. N Engl J Med 304:1529 Apperley JF, Jones L, Hale G, et al. (1986) Bone marrow transplantation for patients with chronic myeloid leukaemia: T-cell depletion with Campath-1 reduces the incidence of graft-vs-host disease but may increase the risk of leukaemic relapse. Bone Marrow Transplant 1:53 Bacigalupo A, Van Lint MT, Occhini D, et al. (1991) Increased risk of leukemia relapse with high-dose cyclosporine A after allogeneic marrow transplantation for acute leukemia. Blood 77:1423 Gaddy J, Risdon G, Broxmeyer HE (1995) Cord blood natural killer cells are functionally and phenotypically immature but readily respond to interleukin-2 and interleukin-12. J Interferon Cytokine Res 15:527 Gardiner CM, O’meara A, Reen DJ (1998) Differential cytotoxicity of cord blood and bone marrow derived natural killer cells. Blood 91:207 Condiotti R, Nagler A (1998) Effect of interleukin-12 on antitumor activity of human umbilical cord blood and bone marrow cytotoxic cells. Exp Hematol 26:571 Cicuttini FM, Martin M, Petrie HT, Boyd AW (1993) A novel population of natural killer progenitor cells isolated from human umbilical cord blood. J Immunol 151:29 Phillips JH, Hori T, Nagler A, Bhat N, Spits H, Lanier LL (1992) Ontogeny of human natural killer (NK) cells: fetal NK cells mediate cytolytic function and express cytoplasmic CD3⑀,␦ proteins. J Exp Med 175:1055 Nagler A, Peacock M, Tantoco M, Lamons D, Okarma TB, Okrongly DA (1994) Red blood cell depletion and enrichment of CD34⫹ hematopoietic progenitor cells from human umbilical cord blood using soybean agglutinin and CD34 immunoselection. Exp Hematol 22:1134 Pierson BA, Gupta K, Hu WS, Miller JS (1996) Human natural killer cell expansion is regulated by thrombospondin-mediated activation of transforming growth factor-␤1 and independent accessory cell-derived contact and soluble factors. Blood 87:180 Nagler A, Condiotti R, Rabinowitz R, Schlesinger M, Nguyen M, Terstappen LWMM (1999) Detection of minimal residual disease (MRD)

R. Condiotti et al./Experimental Hematology 29 (2001) 104–113

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

after bone marrow transplantation (BMT) by multi-parameter flow cytometry (MPFC). Med Oncol 16:177 Umemoto M, Azuma E, Hirayama M, et al. (1997) Two cytotoxic pathways of natural killer cells in human cord blood: implications in cord blood transplantation. Br J Hematol 98:1037 Condiotti R, Slavin S, Barak V, Nagler A (1996) The novel immunomodulator, linomide, stimulates interleukin-2-induced human natural killer (NK) cell and PHA-stimulated T cell proliferation from normal donors. Leuk Res 20:57 Nagler A, Greenberg PL (1990) Bone marrow cell modulation and inhibition of myelopoiesis by large granular lymphocytes and natural killer cells. Int J Cell Cloning 8:171 Nagler A, Greenberg PL, Lanier LL, Phillips JH (1988) The effects of recombinant interleukin 2-activated natural killer cells on autologous peripheral blood hematopoietic progenitors. J Exp Med 168:47 Nagler A, Lanier LL, Cwirla S, Phillips JH (1989) Comparative studies of human FcRIII-positive and negative natural killer cells. J Immunol 143:3183 Nagler A, Lanier LL, Phillips JH (1990) Constitutive expression of high affinity interleukin 2 receptors on human CD16⫺ natural killer cells in vivo. J Exp Med 171:1527 Robertson MJ, Manley TJ, Donahue C, Levine H, Ritz J (1993) Costimulatory signals are required for optimal proliferation of human natural killer cells. J Immunol 150:1705 Miller JS, Oelkers S, Verfaillie C, McGlave P (1992) The role of monocytes in the expansion of human activated natural killer cells. Blood 80:2221 Miller JS, Klingsporn S, Lund J, Perry EH, Verfaillie C, McGlave P (1994) Large scale ex vivo expansion and activation of human natural killer cells for autologous therapy. Bone Marrow Transplant 14:555 Rabinowich H, Sedlmayr P, Herberman RB, Whiteside TL (1991) Increased proliferation, lytic activity and purity of human natural killer cells cocultured with mitogen-activated feeder cells. Cell Immunol 135:454 Pierson BA, McGlave PB, Hu WS, Miller JS (1995) Natural killer cell proliferation is dependent on human serum and markedly increased utilizing an enriched supplemented basal medium. J Hematother 4:149 Miller JS, Alley KA, McGlave PB (1994) Differentiation of natural killer cells from human primitive marrow progenitors in a stroma based long term culture system: identification of a CD34⫺CD7⫹ NK progenitor. Blood 83:2594 Miller JS, McCullar V, Verfaillie CM (1998) Ex vivo culture of CD34⫹/Lin⫺/DR⫺ cells in stroma-derived soluble factors Interleukin-3, and macrophage inflammatory protein-1␣ maintains not only myeloid but also lympoid progenitors in a novel switch culture assay. Blood 91:4516

113

49. Bhatia R, McGlave PB, Miller JS, Wissink S, Lin WN, Verfaillie CM (1997) A clinically suitable ex vivo expansion culture system for LTCIC and CFC using stroma-conditioned medium. Exp Hematol 25:980 50. Thiemann FT, Moore KA, Smogorzewska EM, Lemischka IR, Crooks GM (1998) The murine stromal cell line AFT024 acts specifically on human CD34⫹CD38⫺ progenitors to maintain primitive function and immunophenotype in vitro. Exp Hematol 26:612 51. Testi R, Phillips JH, Lanier LL (1989) T cell activation via Leu-23 (CD69). J Immunol 143:1123 52. Lanier LL, Test R, Bindl J, Phillips JH (1989) Identity of Leu-19 (CD56) leukocyte differentiation antigen and neural cell adhesion molecule. J Exp Med 169:2233 53. Toren A, Novick D, Ackerstein A, Or R, Slavin S, Nagler A (1996) Soluble IL-6 receptors (sIL-6R) in hematological patients receiving immunotherapy with IL-2/IFN-␣ or donor lymphocytes following bone marrow transplantation. Bone Marrow Transplant 18:721 54. Toren A, Barak V, Novick D, Nagler A (1997) Soluble interferon-␥ receptor and interferon-␥ in patients undergoing allogeneic bone marrow transplantation for hematological malignancies. Cytokines Cell Mol Ther 3:153 55. Nagler A, Lanier LL, Phillips JH (1988) The effects of IL-4 on human NK cells: a potent regulator of IL-2 activation and proliferation. J Immunol 141:2349 56. Skea D, Chang NH, Hedge R, et al. (1999) Large ex vivo expansion of human umbilical cord blood CD4⫹ and CD8⫹ T cells. J Hematother 8:129 57. Deutsch VR, Olson TA, Nagler A, Slavin S, Levine RF (1995) The response of cord blood megakaryocyte progenitors to IL-3, IL-6 and aplastic canine serum varies with gestational age. Br J Hematol 89:8 58. Nagler A, Eldor A, Naparstek E, Mumcuoglu M, Slavin S, Deutsch VR (1995) Ex vivo expansion of megakaryocyte precursors by preincubation of marrow allografts with interleukin-3 and granulocyte-macrophage colony-stimulating factor in vitro. Exp Hematol 23:1268 59. Ueno Y, Miyawaki T, Seki H, et al. (1985) Differential effects of recombinant human interferon-␥ and interleukin 2 on natural killer cell activity of peripheral blood in early human development. J Immunol 135:180 60. Keever CA, Abu-Hajir M, Graf W, et al. (1995) Characterization of the alloreactivity and anti-leukemia reactivity of cord blood mononuclear cells. Bone Marrow Transplant 15:407 61. Gardiner CM, Reen DJ (1998) Differential cytokine regulation of natural killer cell-mediated necrotic and apoptotic cytotoxicity. Immunol 93:511 62. Rodella L, Rezzani R, Zauli G, Mariani AR, Rizzoli R, Vitale M (1998) Apoptosis induced by NK cells is modulated by the NK-active cytokines IL-2 and IL-12. Int Immunol 10:719