Journal of Immunological Methods, 136 (1991) 1-9 © 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0022-1759/91/$03.50 ADONIS 002217599100051F JIM05796
Immunomagnetic isolation of N K and LAK cells Bjorn N a u m e 1, Unni Nonstad 1, Bjorg Steinkjer 1, Steinar F u n d e r u d 2, Erlend Smeland 2 and Terje Espevik 1 I Institute of Cancer Research, University of Trondheim, N-7006 Trondheim, Norway, and 2 Laboratory for Immunology, Institute of Cancer Research, The Norwegian Radium Hospital, Montebello, N-0310 Oslo 3, Norway (Received 20 July 1990, revised received 10 September 1990, accepted 17 September 1990)
The present study describes the irnmunomagnetic isolation of human natural killer (NK) and lymphokine activated killer (LAK) cells. Antibodies against CD56 and sheep anti-mouse IgG-coated magnetic monodisperse particles (Dynabeads M-450) were used for the positive isolation of CD56 + cells from unstimulated mononuclear cells (PBMC). A highly enriched population of CD56 + cells (~< 3% contaminating cells) was obtained with this method. The cellular yield of CD56 + cells was high (5.3% of the unseparated PBMC). The CD56 + cells remained unactivated after separation and preserved their functional characteristics, as measured by cytotoxic activity against the NK sensitive K562 cells. Incubating the CD56 + cells with IL-2 resulted in high LAK activity, as measured by cytotoxic activity against Daudi cells. Large numbers of functionally active CD56 + cells were obtained from IL-2 stimulated lymphocytes using anti-CD56 coated Dynabeads 450. A further enrichment of effector cells with LAK activity was accomplished by depleting the CD56 + cells for T-cells by anti-CD3 coated Dynabeads M450. The immunomagnetic isolation technique described was easy to perform, did not require expensive equipment and yielded NK and LAK cells of satisfactory purity. Key words: Immunomagnetic isolation; CD56 + cell; Natural killer cell; lymphokine-activated killer cell; Monoclonal antibody, CD56; Positive separation
Introduction
Natural killer (NK) cells are a type of lymphocyte found within the large granular lymphocyte Correspondence to: B. Naume, Institute of Cancer Research, University of Trondheim, N-7006 Trondheim, Norway. Abbrevations: NK cells, natural killer cells; LAK cells, lymphokine-activated killer cells; PBMC, peripheral blood mononuclear cells; PBL, peripheral blood lymphocytes; IL-2, interleukin-2; MoAb, monoclonal antibody; FACS, fluorescence activated cell sorter; PE, phycoerythrin; FITC, fluorescein isothiocyanate; HBSS, Hanks' balanced salt solution; BSA, bovine serum albumin; FCS, fetal calf serum; PBS, phosphate buffered saline; SAM, sheep anti-mouse; GAM, goat anti-mouse.
population (Timonen et al., 1979, 1981). At present the NK cell phenotype may be defined as CD3 negative ( - ) , CD16 positive (+), CD56 + or CD3-, CD16 , CD56 + lymphocytes (Perussia et al., 1983; Lanier et al., 1986; Nagler et al., 1989). NK cell activity is increased by stimulation with interleukin-2 (IL-2), and IL-2 stimulated NK cells can kill NK resistant tumor cells, a phenomenon which is defined as lymphokine activated killer (LAK) cell activity (Henney et al., 1981; Grimm et al., 1982; Trinchieri et al., 1984; Phillips and Lanier, 1986). It is of great importance, when studying NK cell function, to obtain a highly enriched population of these cells. The most commonly used tech-
niques for isolation of NK cells are: Percoll density gradient centrifugation (Timonen et al., 1980, 1982), erythrocyte (E) rosetting (Timonen et al., 1982; Perussia et al., 1983), complement depletion of other lymphocyte subsets (Hercend et al., 1982; London et al., 1985), panning techniques (Ortaldo et al., 1986; Garcia-Peharrubia et al., 1989) and fluorescence-activated cell sorting (FACS) (Phillips and Lainier, 1986; Ostensen et al., 1987). The FACS technique is time consuming, requires expensive equipment and the cell yield is limited. Panning techniques may also give a low cell yield when used for the isolation of small lymphocyte subsets. Negative selection with complement is time consuming and involves many steps. Percoll gradient separation is easy to perform, but does not give an optimal purity of NK cells. Another method which has proved to be efficient for T lymphocyte depletion or providing populations of B and T lymphocytes in high purity is immunomagnetic isolation. The isolation procedure is based on the use of magnetic monodisperse particles (Dynabeads M-450) either coupled directly to a monoclonal antibody (MoAb) of IgM class (Gaudernack et al., 1986; Funderud et al., 1990) or firstly attached to an anti-mouse IgG prior to adhering the MoAb (IgG) and the MoAb specific cell (Lea et al., 1985, 1986; Vartdal et al., 1987). The purpose of this study was to adapt the immunomagnetic isolation technique for the enrichment of NK and LAK cells. By means of positive selection, using an anti-CD56 MoAb, the present method provides a highly enriched population of CD56 + cells with high NK and LAK activities.
Materials and methods
Monoclonal antibodies Unconjugated anti-Leu-19 (CD56), phycoerythtin (PE)-conjugated anti-Leu-19, anti-Leu-4 (CD3) and anti-Leu-llc (CD16) (all 50 tlg/ml), fluorescein isothiocyanate (FITC)-conjugated antiLeu-12 (CD19) and anti-Leu-M3 (CD14) (25 /xg/ml) were purchased from Becton Dickinson (Mountain View, CA, U.S.A.). Unconjugated antiLeu-19, used in the separation procedure, was dialysed against phosphate-buffered saline (PBS)
to eliminate azide. Supernatant from the OKT3 hybridoma (CRL 8001, ATCC, Rockville, MD, U.S.A.) was used as a source of anti-CD3 MoAb. The HKB1 MoAb (IgM) produced in the Laboratory for Immunology, The Norwegian Radium Hospital, recognizes the gene products of all three major histocompatibility complex class II loci (DP, DQ, DR) (Holte et al., 1989). My7 MoAb (CD33) was purchased from Coulter Immunology (Hileah, FL, U.S.A.). Magnetic beads Dynabeads M-450 coated with sheep antimouse (SAM) IgG (M-450) (prod. no. 110.02), Dynal MPC-6, a neodynium magnet particle separator (Prod. nr. 120.02) was supplied by DYNAL, Oslo, Norway. Cell lines OKT3 hybridoma (ATCC CRL 8001) was cultured in DMEM (Gibco, Glascow, U.K.) containing 600 /~g/ml glutamine and 1 mM sodium pyruvate with 10% fetal calf serum (FCS) (Gibco). The NK-sensitive K562 cell line and the NK-resistant Daudi cell line were generously provided by Drs. Anthony Rayner and Brett Gemlo (Department of Surgery, University of California Medical School, San Francisco, CA, U.S.A.). These cell lines were cultured in RPMI 1640 (Gibco) containing 40 /~g/ml garamycin and 100 /~g/ml glutamine (complete RPMI) with 10% FCS. Immunornagnetic isolation of CD56 + cells (indirect technique) Bully coats were obtained from The Blood Bank, The University Hospital of Trondheim. All of the subsequent steps described were performed at 4°C, if not otherwise specified. Mononuclear cells (PBMC) were isolated by Lymphoprep (Nycomed, Oslo, Norway) centrifugation (Boyum, 1976) and washed four times with Hanks' balanced salt solution (HBSS) (Gibco). The PBMC were diluted in complete RPMI with 5% A + human serum (referred to as complete medium (CM)) to a concentration of 1-2 X 10 ~ cells/ml. Thereafter 0.08 ~tg anti-CD56/1 x 10 6 PBMC was added and the cells incubated for 1 h. The PBMC were then washed twice in HBSS and diluted in CM and incubated with 7.5 × 105 Dynabeads/106
PBMC for 30 rain in 10 ml tubes (Nunc, Roskilde, Denmark) with gentle rotation. Cells attached to the Dynabeads were recovered with the MPC-6 magnet and washed five times in CM. Thereafter the cell/bead suspension was incubated for 16-20 h at 37°C in 25 or 75 cm 2 culture flasks (Costar, Cambridge, MA, U.S.A.) so that most of the Dynabeads detached from the CD56 + isolated cells. Using the magnet, the cells were separated from the Dynabeads and were stained directly for flow cytometry analysis, whilst others were tested in cytotoxicity assays (together with unfractionated PBMC and CD56 depleted PBMC). A portion of the CD56 + isolated cells was incubated with 103 U / m l recombinant IL-2 (3 × 10 6 U / m g protein, Cetus Corp. Emeryville, CA, U.S.A.), at a concentration of 2 x 10 6 cells/ml for 2-3 days.
Immunomagnetic isolation of CD56 + LAK cells by anti-CD56 coated Dynabeads (direct technique) Anti-CD56 MoAb was added to the Dynabeads, at a concentration of 0.42 /~g MoAb to 1 mg Dynabeads, and incubated for 24 h on a rotator. The suspension was washed four times in CM using the magnet. PBMC depleted of monocytes by 90 min plastic adherence (PBL), were incubated for 3 days with 103 U / m l IL-2. CD56 + cells were isolated by adding 7.5 x 105 of the anti-CD56 coated Dynabeads to 1 × 10 6 PBL and separated as described. The isolated cells were analyzed for cytotoxicity against Daudi cells. Depletion of CD3 + cells from the CD56 + LAK cells 10 /~1 anti-CD3 MoAb from the OKT3 hybridoma (approximately 10 /~g M o A b / m l supernatant) were added to 1 mg Dynabeads and prepared as described for the anti-CD56 conjugated Dynabeads. CD56 + LAK cells were depleted of CD3 + cells by incubating 2 × 10 6 anti-CD3 coated Dynabeads with 1 × 10 6 CD56 + cells and separated as described. The CD3 depleted cells were also analyzed for cytotoxic activity against Daudi cells. Negative selection of NK cells PBL were incubated with HKBl-coated Dynabeads (directly adsorbed to unconjugated Dynabeads) and anti-CD3 coated Dynabeads, 7.5 ×
105 and 4.0 × 10 6 beads respectively, to 1 × 10 6 PBL for 30 rain in CM as described. The cells attached to the Dynabeads were removed and the unattached cells were analyzed by immunofluorescence microscopy and flow cytometry.
Immunofluorescence microscopy and flow cytometry Cells were stained with the appropriate antibody by incubating 1 × 106 cells with 0.5 /~g of PE- or 0.25 ~tg FITC-conjugated MoAb for 30 min. Control samples were stained with 0.13 /zg FITC-conjugated goat anti-mouse IgG (Becton Dickinson) or 0.20 /~g PE-streptavidin (Vector Laboratories, CA, U.S.A.) diluted in PBS with 0.1% BSA (Sigma, St.Louis, MO, U.S.A.). The cells were then washed twice in PBS/BSA before being fixed in 2% formaldehyde (Merck, Darmstadt, F.R.G.) in PBS for 20 rain. After one wash in PBS, 5000 cells were analyzed using a FACScan flow cytometer (Becton Dickinson). Two color fluorescence analysis was accomplished by incubating the cells with 0.5 /~g unconjugated antiLeu-4 MoAb, followed by three washes in PBS/BSA before 0.13 /zg FITC-conjugated goat anti-mouse IgG was added. After 30 min the cells were washed twice in PBS/BSA, and stained with 0.5/zg PE-conjugated anti-Leu-19 for 30 rain. After washing and fixation, the cells were analyzed by flow cytometry. The stained cells were also examined by immunofluorescence microscopy (Wild Leitz Orthoplan, Heerbrugg, Switzerland). Cytotoxicity assays The isolated cell populations were analyzed for cytotoxic activity against K562 and Daudi cells. About 3 x 10 6 target cells were labeled with 300 ~Ci Na~lCrO4 (Amersham, Buckinghamshire, U.K.) for 90 min at 37°C, followed by two washes in RPMI 1640 and a third wash in CM. 100 ~1 of target cells (105 cells/ml) were added to various concentrations of effector cells in triplicate in 96 well round bottomed microtiter plates (Costar). After 4 hours of incubation at 37°C, the supernatants (100 jal/well) were harvested, and radioactivity determined in a gamma counter (LKB Wallac 1270, Wallac Oy., Turku, Finland). Percent specific lysis was calculated as follows: % specific lysis
A-B C - B × 100
where A represents counts per minute (cpm) from test supernatants; B represents cpm from supernatants of target cells incubated alone (spontaneous release); and C represents cpm after lysis of target cells with 0.25% SDS (Sigma) (maximum release). Results are presented as the mean _+ SD of triplicate cultures. Spontaneous release did not exceed 20% of maximum release. Calculations The amounts of different cell types were recorded in replicate experiments and the results are presented as mean (SD).
CD56 E 40.2% =:
----4
''''t
...............
B
CD3 F 32.4%
C
CD19
CD56 78.5%
I ""'
......
P
........
t
.......
CD3 13.7%
.J ..I LU
o LL
o U.I
m
~ , ,
Results
G
9.9%.I1
CD19
2.8%
CD14 I H
CD14
/I
0.3%
z
Selection of CD56 + cells from unstimulated P B M C We first purified N K cells by depleting PBL for monocytes, B cells and T cells by using HKB1 and anti-CD3 coated Dynabeads as described in the materials and methods section. This resulted in an enrichment of CD16 + cells (71% (3), n = 3). However, the numbers of contaminating cells were too high (data not shown) and this approach resulted in a high consumption of Dynabeads. Positive selection of CD56 + cells was next performed by adding anti-CD56 coated Dynabeads to PBMC. This resulted in a low cellular yield ( < 1%). Thus, an indirect technique was tried by incubating PBMC with anti-CD56 MoAb before adding SAM coated Dynabeads. This method gave a satisfactory cellular yield for further analysis as described below. The purity of the CD56 + cell population was analyzed with monoclonal antibodies specific for different subsets of mononuclear cells. As shown in Fig. 1, the CD56 + selected cells were almost completely depleted of monocytes (CD14 +) and B lymphocytes ( C D ] 9 +) and the numbers of T cells (CD3 +) were markedly decreased. Based on five experiments, we calculated that there were 10% (6) CD19 + cells in the PBMC population, and 1% (1) in the CD56 + selected population. The amount of monocytes was 29% (4) before and 0.5% (0.5) after separation. The amount of CD3 + cells decreased from 42% (17) (n = 6) prior to isolation to 19% (8) (n = 6) in the CD56 + selected population. These results were verified by immunofluorescence mi-
D
28.3%
L/'/I
. . . . . . . . . . . . . . . .
FLUORESCENCE
I
........
I
........
INTENSITY (log)
Fig. 1. Flow cytometric analysis of PBMC (A, B, C, D) and the CD56 + isolated cells immunomagnetically separated from PBMC (E, F, G, H). The two different cell populations were stained with anti-Leu-19 (CD56) (A, E), anti-Leu-4 (CD3) (B,F) (both PE-conjugated), anti-Leu-12 (CD19) (C, G) and anti-Leu-M3 (CD14) (D, H) (both FITC-conjugated), respectively. The samples were analyzed on a FACScan flow cytometer. The dotted lines represent samples stained with PE-streptavidin or FITC-conjugated SAM IgG only (G and H, respectively). The analysis presented in A, B and C are gated on light scatter to eliminate the monocytes. The numbers in the histograms represent the percentage of cells positively stained with the respective MoAbs. The percentage levels of positive cells were estimated by setting limits and subtracting the percentage of cells defined as positive in the control sample from the positive cells in the MoAb treated sample. In panel E the intersection point between the CD56 and the control histogram was used for defining the positively and negatively stained cells. Similar data were obtained in at least four experiments.
O9 _1 ._1 ILl tO
CD56 91.3%
LI.
0
: .......>.../.., - , \ ................ \
W o3
Z
-'-7'~,,,,,
10 °
I
191 1
7'
t aJ~,,, I
102
jr,; ''''i
~ J''''''
10"4
FLUORESCENCE INTENSITY (log) Fig. 2. Flow cytometric analysis of the CD56 + isolated cells stained with PE-conjugated anti-Leu-19 (CD56) immediately after the immunomagnetic separation ( ...... ) and after 3 days incubation with IL-2, 103 U / m l ( ). The control ( . . . . . . ) represents IL-2-stimulated cells stained with PE-streptavidin only. The samples were analyzed on a FACScan flow cytometer. The percentage in the histogram represents the a m o u n t of CD56 ÷ cells after IL-2 stimulation. The levels of CD56 + cells registered immediately after separation reached 75.0%. Similar data were obtained in four experiments.
croscopy (data not shown). The presence of T lymphocytes in the CD56 + population was consistent with the fact that a fraction of the CD56 + cells are CD3 + (Lanier et al., 1986). To determine whether or not some of the CD56 + isolated cells were CD3 +, C D 5 6 - , we analyzed the isolated cells by two color immunofluorescence with anti-CD3 and anti-CD56. These data showed that only 0.5% (0.9) ( n = 4 ) of the isolated cells were CD3 +, CD56 +. In none of the experiments did the contaminating monocytes, B cells and C D 5 6 - , CD3 + T cells exceed 3%. The PBMC contained 31% ( l l ) (n = 7) CD56 + cells. Immunomagnetic isolation of these cells resulted in a reduced immunofluorescence intensity when stained with anti-CD56 (Fig. 1). This made an exact quantification of CD56 + cells in the positively selected population difficult. However, we provisionally recorded 75% (9) (n = 7) CD56 + cells. These data were verified by immunofluorescence microscopy (data not shown). To overcome the problem of low CD56 expression, we stimulated the CD56 + selected cells with IL-2 for 3 days before the amount of CD56 + cells was measured by flow cytometry. As shown in Fig. 2, the
addition of IL-2 resulted in an increased expression of the CD56 antigen which made quantification of the amount of CD56 + cells more accurate. The isolation procedure thus gave a population consisting of 92% (3) (n = 4) CD56 + cells as determined by flow cytometry. By immunofluorescence microscopy (counting 400 cells) > 94% of the positively selected cells were CD56 +. The addition of IL-2 for 3 days did not increase the total number of cells in the CD56 + population (data not shown), indicating no selective proliferation of CD56 + cells. Furthermore, the numbers of CD3 + cells remained unaltered after IL-2 stimulation (data not shown). The method described gave a cellular yield of 5.3% (1.3) (n = 8) of the PBMC. Consequently, approximately 1.4-2.3 × 1 0 7 CD56 + cells could be obtained from a representative buffy coat containing approximately 3.5 × 108 mononuclear cells. The viability of the CD56 selected cells was found to be > 91% as determined by trypan blue exclusion.
• CD56+ cells • PBMC
6O
40
20
I 1.57:1
I 3.13:1
_ _ l 6.25:1
I 12.5:1
I 25:1
EFFECTOR : TARGET CELL RATIO
Fig. 3. Cytotoxic activity against K562 cells mediated by PBMC (A A), CD56 + cells (O O) and CD56 depleted PBMC ( I i ) immediately after immunomagnetic isolation as described in the materials and methods section. Results are presented as mean_+SD of triplicate determinations. Similar data were obtained in three experiments.
60
• CD56+ cells (+IL-2) • PBMC (+IL-2)
t
• CD56 depleted PBMC (+IL-2) 50
.[
o CD56+ ceils ( I L - 2 )
~
4o
N K activity of the CD56 + selected cells. After incubation with IL-2 for 2 - 3 days, the CD56 + cells expressed high cytotoxic activity against Daudi cells compared to the PBMC and the CD56 depleted population (Fig. 4). These results suggest that high L A K activity can be induced in the CD56 + population.
co (D ~ 3O
% 2O
lC
1.57:1
3.13:1
6.25:1
12.5:1
25:1
EFFECTOR : TARGET CELL RATIO
Fig. 4. Cytotoxic activity against Daudi cells, mediated by CD56 + cells in the absence of IL-2 ( o O), and CD56 + cells ( e o), PBMC (A A) and CD56 depleted PBMC ( I II) after incubation with IL-2 (103 U / m l ) for 3 days. Results are presented as mean+_SD of triplicate determinations. Similar data were obtained in three experiments.
Functional properties of CD56 + cells The functional properties of the CD56 + cells were tested against the N K sensitive K562 cell line and the N K resistant Daudi cell line. Fig. 3 demonstrates that the capacity of the CD56 + enriched cells to kill the K562 cells was high compared with unfractionated PBMC and PBMC depleted with respect to CD56 + cells. This result is consistent with an N K cell enrichment in the CD56 + population. There was no significant killing of Daudi cells by the CD56 + population (Fig. 4) or the PBMC population (data not shown). To test whether the separation procedure altered N K activity, we incubated PBL with either Dynabeads alone or firstly with anti-CD56 M o A b and secondly with the Dynabeads. After 20 h the Dynabeads were separated from the PBL, which then were tested for cytotoxicity against K562. These manipulations of the effector cells did not change their ability to kill K562 cells compared to unmanipulated PBL (data not shown). Thus, the separation technique did not interfere with the
Isolation of CD56 + L A K cells from IL-2 stimulated P B L and depletion of CD3 + cells from the CD56 + L A K cells Immunomagnetic isolation, using anti-CD56 coated Dynabeads (direct technique), was also performed on PBL incubated for 3 days with 103 U / m l IL-2. In contrast to unstimulated PBMC treated with anti-CD56 coated Dynabeads, this approach resulted in a high cellular yield, 5.4% (3.1) (n = 8), of CD56 + selected cells. The CD56 ~ L A K cells were further depleted for CD3 + cells by immunomagnetic isolation using anti-CD3 coated Dynabeads. The different cell populations
60 • CD56+ LAK celts • CD56 depleted PBL (LAK~ • CD3 depleted Cb56+ LAK cells
4o co _o u_ co
0
0.6:1
1.2:1
I 2.5:1
I 5:1
I 10:1
EFFECTOR : TARGET CELL RATIO
Fig. 5. Cytotoxic activity against Daudi cells mediated by immunomagnetically isolated cells from IL-2 stimulated PBL (103 U / m l for 3 days). The effector ceils tested were CD56 + LAK cells (O O), CD3 depleted CD56 + LAK ceils (A --A) and CD56 depleted PBL (LAK) (m -I). Results are presented as mean+_SD of triplicate determinations. Similar data were obtained in two experiments.
were examined for LAK activity against Daudi cells. The data show that the CD56 + cells, prior to CD3 depletion, expressed a much higher cytotoxic activity than the CD56 depleted LAK cells and that depletion of CD3 + cells resulted in an additional increase in LAK activity (Fig. 5).
Discussion
In this paper we have described an immunomagnetic isolation technique which resulted in a satisfactory amount of highly enriched and functionally active N K / L A K cells. We found that the isolated cells contained 92% CD56 + cells and no more than 3% monocytes, B cells and CD56-, CD3 + T cells. Granulocyte contamination (as determined by immunofluorescence staining with MoAb My7 (CD33)) was negligible ( < 0.5%, data not shown). Approximately 5% of the CD56 isolated cells were negative by flow cytometry for CD56, CD3, CD19 or CD14. T cells and B cells are strictly defined by the CD3 and CD19 markers, respectively (Kung et al., 1979; Dorken et al., 1989). CD14 is known to be expressed on approximately 90% of the monocytes (Dimitriu-Bona et al., 1983) and this indicates that some C D 1 4 monocytes could be present in the isolated population. However, we have observed that monocytes efficiently engulf the Dynabeads at 37°C and consequently will be removed together with the beads after overnight incubation (data not shown). Any significant monocyte contamination in the CD56 + population is therefore unlikely. The immunomagnetic isolation procedure resulted in reduced CD56 expression (Fig. 1), which could be due to CD56 antigen shedding a n d / o r low turnover and renewal of the antigen when being manipulated with the Dynabeads (Lea et al., 1986; Funderud et al., 1990). The expression of CD56 was increased in the presence of IL-2 (Fig. 2), which is in accordance with Lainier et al. (1989). This CD56 upregulation was not due to proliferation of CD56 + cells and is consistent with earlier reports that have shown no increase in total cell numbers after IL-2 incubation of lymphocytes or N K cells for 3 days (Talmadge et al., 1986; Yamada et al., 1987). The possibility exists that a small percentage of the isolated cells will not
upregulate their CD56 antigen after the addition of IL-2, which then would be detected as CD56-, CD3 , C D 1 9 - and C D 1 4 - cells. Conclusively, the isolated cell population appears to contain ~< 3% contaminating cells, as registered by flow cytometry. Despite the reduction in CD56 antigen expression immediately after the immunomagnetic isolation, the enriched N K cells showed high cytotoxicity against K562. Furthermore, the isolation procedure did not stimulate the N K cells to kill the N K resistant Daudi cells. The addition of IL-2 resulted in high LAK activity by the CD56 + cells. The raised N K and LAK activity found in the CD56 + population compared to the unseparated or CD56 depleted cell populations is in accordance with the described enrichment of N K cells. The immunomagnetic isolation technique, using anti-CD56 MoAb, did not activate or impair the functional abilities of the N K cells. This result also agrees with earlier studies demonstrating that CD56 antibodies do not alter the functional abilities of N K cells (Lanier et al., 1986), and the previously described immunomagnetic techniques for B and T cell isolation showing no functional alterations (Gaudernack et al., 1986; Funderud et al., 1990). We therefore conclude that the immunomagnetically isolated CD56 + cells have preserved their functional characteristics after the separation procedure. The N K / L A K cells in the CD56 + population was further purified by depletion of T cells with anti-CD3 coated beads. This resulted in increased LAK activity (Fig. 5) which is in accordance with an activated CD56 +, C D 3 - N K cell as the predominant LAK cell (Phillips and Lanier, 1986). To further increase the N K cell purity, any additional T cells, B cells and monocytes can be removed with Dynabeads coated with a n t i - C D 4 / anti-CD8 (Lea et al., 1985), anti-CD19 (Funderud et al., 1990) and ID5 (monocyte specific) (Kaplan and Gaudernack, 1982), antibodies respectively. In conclusion, we have successfully adapted the immunomagnetic separation technique for the isolation of CD56 + cells. This technique should be of value when it is intended to undertake functional studies on isolated N K / L A K cells. Because of its simplicity, the method is preferable to previously published methods.
Acknowledgements We thank Dagmar Moholdt for assistance in preparing the manuscript. This work was supported by the Norwegian Cancer Society, Norsk Hydro and Dynal, Norway.
References Boyum, A. (1976) Isolation of lymphocytes, granulocytes, and macrophages. Scand. J. Immunol. 5 (suppl. 5), 10. Dimitriu-Bona, A., Burmester, G.R., Waters, S.J. and Winchester, R,J. (1983) Human mononuclear phagocyte differentiation antigens. I. Patterns of antigenic expression on the surface of human monocytes and macrophages defined by monoclonal antibodies. J. Immunol. 130, 145. D~Srken, B., MiSller, P., Pezzutto, A., Schwartz-Albiez, R. and Moldenhauer, G. (1989) B-cell antigens: CD19. In: W. Knapp, B. D~Srken, W.R. Gilks, E.P. Rieber, R.E. Schmidt, H. Stein and A.E.G.Kr. Von dem Borne (Eds.), Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press, Oxford. Funderud, S., Erikstein, B., Asheim, H.C., Nustad, K., Stokke, T., Kiil Blomhoff, H., Holte, H. and Smeland, E.B. (1990) Functional properties of CD19+B lymphocytes positively selected from buffy coats by immunomagnetic separation. Eur. J. Immunol. 20, 201. Garcia-Peharrubia, P., Koster, F.T., Kelley, R.O., McDowell, T.D. and Bankhurst, A.D. (1989) Antibacterial activity of human natural killer cells. J. Exp. Med. 169, 99. Gaudernack, G., Leivestad, T., Ugelstad, J. and Thorsby, E. (1986) Isolation of pure functionally active CD8+T cells. Positive selection with monoclonal antibodies directly conjugated to monosized magnetic microspheres. J. Immunol. Methods 90, 179. Grimm, E.A., Mazumder, A., Zhang, H.Z. and Rosenberg, S.A. (1982) Lymphokine-activated killer cell phenomenon. Lysis of natural killer-resistant fresh solid tumor cells by interleukin 2-activated autologous human peripheral blood lymphocytes. J. Exp. Med. 155, 1823. Henney, C.S., Kuribayashi, K., Kern, D.E. and Gillis, S. (1981) lnterleukin-2 augments natural killer cell activity. Nature 291, 335. Hercend, T., Meuer, S., Reinherz, E.L. Schlossman, S.F. and Ritz, J. (1982) Generation of a cloned NK cell line derived from the 'null cell' fraction of human peripheral blood. J. lmmunol. 129, 1299. Holte, H., Kiil Blomhoff, H., Beiske, K., Funderud, S., Torjesen, P., Gaudernack, G., Stokke, T. and Smeland, E.B. (1989) Intracellular events associated with inhibition of B cell activation by monoclonal antibodies to HLA class II antigens. Eur. J. Immunol. 19, 1221. Kan, E.A.R., Wang, C.Y., Wang, L.C. and Evans, R.L. (1983) Non-covalently bonded subunits of 22 and 28 kd are rapidly
internalized by T cells reacted with anti-leu-4 antibody. J. Immunol. 131,536. Kaplan, G. and Gaudernack, G. (1982) In vitro differentiation of human monocytes. Differences in monocyte phenotypes induced by cultivation on glass or on collagen. J. Exp. Med. 156, 1101. Kung, P.C., Goldstein, G., Reinherz, E.L. and Schlossman, S.F. (1979) Monoclonal antibodies defining distinctive human T cell surface antigens. Science 206, 347. Lanier, L.L., Le, A.M., Civin, C.I., Loken, M.R. and Phillips, J.H. (1986) The relationship of CDI6 (Leu-ll) and Leu-19 (NKH-1) antigen expression on human peripheral blood NK cells and cytotoxic T lymphocytes. J. Immunol. 136, 4480. Lanier, L.L., Testi, R., Bindl, J. and Phillips, J.H. (1989) Identity of Leu-19 (CD56) leukocyte differentiation antigen and neural cell adhesion molecule. J. Exp. Med. 169, 2233. Lea, T., Vartdal, F., Davies, C. and Ugelstad, J. (1985) Magnetic monosized polymer particles for fast and specific fractination of human mononuclear cells, Scand. J. Immunol. 22, 207. Lea, T., Smeland, E., Funderud, S., Vartdal, F., Davies, C.. Beiske, K. and Ugelstad, J. (1986) Characterization of human mononuclear cells and positive selection with immunomagnetic particles. Scand. J. Immunol. 23, 509. London, L., Perussia, B. and Trinchieri, G. (1985) Induction of proliferation in vitro of resting human natural killer cells: expression of surface activation antigens. J. lmmunol. 134, 718. Nagler, A., Lanier, L.L., Cwirla, S. and Phillips, J.H. (1989) Comparative studies of human FcRIll-positive and negative natural killer cells. J. lmmunol. 143, 3183. Ortaldo, J.R., Mason, A. and Overton, R. (1986) Lymphokineactivated killer cells. Analysis of progenitors and effectors. J. Exp. Med. 164, 1193. Ostensen, M.E., Thiele, D.L. and Lipsky, P.E. (1987) Tumor necrosis factor-c~ enhances cytolytic activity of human natural killer cells. J. Immunol. 138. 4185. Perussia, B., Starr, S., Abraham~ S., Fanning, V. and Trinchieri, G. (1983) Human natural killer cells analyzed by B73.1, a monoclonal antibody blocking Fc receptor functions. I: characterization of the lymphocyte subset reactive with B73.1. J. Immunol. 130, 2133. Phillips, 3.H. and Lanier, L.L. (1986) Dissection of the lymphokine-activated killer phenomenon. Relative contribution of peripheral blood natural killer cells and T lymphocytes to cytolysis. J. Exp. Med. 164, 814. Talmadge, J.E., Wiltrout, R.H., Counts, D.F., Herberman, R.B., McDonald, T. and Ortaldo, J.R. (1986) Proliferation of human peripheral blood lymphocytes induced by recombinant human interleukin 2: Contribution of large granular lymphocytes and T lymphocytes. Cell. lmmunol. 102, 261. Timonen, T. and Saksela, E. (1980) Isolation of human NK cells by density gradient centrifugation. J. Immunol. Methods 36, 285. Timonen, T., Saksela, E., Ranki A. and H~yry, P. (1979) Fractionation, morphological and functional characteriza-
tion of effector cells responsible for human natural killer activity against cell-line targets. Cell. Immunol. 48, 133. Timonen, T., Ortaldo, J.R. and Herberman, R.B. (1981) Characteristics of human large granular lymphocytes and relationship to natural killer and K cells. J. Exp. Med. 153, 569. Timonen, T., Reynolds, C.W., Ortaldo, J.R. and Herberman, R.B. (1982) Isolation of human and rat natural killer cells. J. Immunol. Methods 51, 269. Trinchieri, G., Matsumoto-Kobayashi, M., Clark, S.C., Seehra, J., London, L. and Perussia, B. (1984) Response of resting human peripheral blood natural killer cells to interleukin 2. J. Exp. Med. 160, 1147.
Vartdal, F., Kvalheim, G., Lea, T.E,, Bosnes, V., Gaudernack, G., Ugelstad, J. and Albrechtsem D. (1987) Depletion of T lymphocytes from human bone marrow. Transplantation 43, 366. Yamada, S., Ruscetti, F.E., Overton, W.R., Herberman, R.B., Birchenall-Sparks, M.C. and Ortaldo, J.R. (1987) Regulation of human large granular lymphocyte and T cell growth and function by recombinant interleukin 2: Induction of interleukin 2 receptor and promotion of growth of cells with enhanced cytotoxicity. J. Leukocyte Biol. 41,505.
Immunomagnetic isolation of N K and LAK cells Bjorn N a u m e 1, Unni Nonstad 1, Bjorg Steinkjer 1, Steinar F u n d e r u d 2, Erlend Smeland 2 and Terje Espevik 1 I Institute of Cancer Research, University of Trondheim, N-7006 Trondheim, Norway, and 2 Laboratory for Immunology, Institute of Cancer Research, The Norwegian Radium Hospital, Montebello, N-0310 Oslo 3, Norway (Received 20 July 1990, revised received 10 September 1990, accepted 17 September 1990)
The present study describes the irnmunomagnetic isolation of human natural killer (NK) and lymphokine activated killer (LAK) cells. Antibodies against CD56 and sheep anti-mouse IgG-coated magnetic monodisperse particles (Dynabeads M-450) were used for the positive isolation of CD56 + cells from unstimulated mononuclear cells (PBMC). A highly enriched population of CD56 + cells (~< 3% contaminating cells) was obtained with this method. The cellular yield of CD56 + cells was high (5.3% of the unseparated PBMC). The CD56 + cells remained unactivated after separation and preserved their functional characteristics, as measured by cytotoxic activity against the NK sensitive K562 cells. Incubating the CD56 + cells with IL-2 resulted in high LAK activity, as measured by cytotoxic activity against Daudi cells. Large numbers of functionally active CD56 + cells were obtained from IL-2 stimulated lymphocytes using anti-CD56 coated Dynabeads 450. A further enrichment of effector cells with LAK activity was accomplished by depleting the CD56 + cells for T-cells by anti-CD3 coated Dynabeads M450. The immunomagnetic isolation technique described was easy to perform, did not require expensive equipment and yielded NK and LAK cells of satisfactory purity. Key words: Immunomagnetic isolation; CD56 + cell; Natural killer cell; lymphokine-activated killer cell; Monoclonal antibody, CD56; Positive separation
Introduction
Natural killer (NK) cells are a type of lymphocyte found within the large granular lymphocyte Correspondence to: B. Naume, Institute of Cancer Research, University of Trondheim, N-7006 Trondheim, Norway. Abbrevations: NK cells, natural killer cells; LAK cells, lymphokine-activated killer cells; PBMC, peripheral blood mononuclear cells; PBL, peripheral blood lymphocytes; IL-2, interleukin-2; MoAb, monoclonal antibody; FACS, fluorescence activated cell sorter; PE, phycoerythrin; FITC, fluorescein isothiocyanate; HBSS, Hanks' balanced salt solution; BSA, bovine serum albumin; FCS, fetal calf serum; PBS, phosphate buffered saline; SAM, sheep anti-mouse; GAM, goat anti-mouse.
population (Timonen et al., 1979, 1981). At present the NK cell phenotype may be defined as CD3 negative ( - ) , CD16 positive (+), CD56 + or CD3-, CD16 , CD56 + lymphocytes (Perussia et al., 1983; Lanier et al., 1986; Nagler et al., 1989). NK cell activity is increased by stimulation with interleukin-2 (IL-2), and IL-2 stimulated NK cells can kill NK resistant tumor cells, a phenomenon which is defined as lymphokine activated killer (LAK) cell activity (Henney et al., 1981; Grimm et al., 1982; Trinchieri et al., 1984; Phillips and Lanier, 1986). It is of great importance, when studying NK cell function, to obtain a highly enriched population of these cells. The most commonly used tech-
niques for isolation of NK cells are: Percoll density gradient centrifugation (Timonen et al., 1980, 1982), erythrocyte (E) rosetting (Timonen et al., 1982; Perussia et al., 1983), complement depletion of other lymphocyte subsets (Hercend et al., 1982; London et al., 1985), panning techniques (Ortaldo et al., 1986; Garcia-Peharrubia et al., 1989) and fluorescence-activated cell sorting (FACS) (Phillips and Lainier, 1986; Ostensen et al., 1987). The FACS technique is time consuming, requires expensive equipment and the cell yield is limited. Panning techniques may also give a low cell yield when used for the isolation of small lymphocyte subsets. Negative selection with complement is time consuming and involves many steps. Percoll gradient separation is easy to perform, but does not give an optimal purity of NK cells. Another method which has proved to be efficient for T lymphocyte depletion or providing populations of B and T lymphocytes in high purity is immunomagnetic isolation. The isolation procedure is based on the use of magnetic monodisperse particles (Dynabeads M-450) either coupled directly to a monoclonal antibody (MoAb) of IgM class (Gaudernack et al., 1986; Funderud et al., 1990) or firstly attached to an anti-mouse IgG prior to adhering the MoAb (IgG) and the MoAb specific cell (Lea et al., 1985, 1986; Vartdal et al., 1987). The purpose of this study was to adapt the immunomagnetic isolation technique for the enrichment of NK and LAK cells. By means of positive selection, using an anti-CD56 MoAb, the present method provides a highly enriched population of CD56 + cells with high NK and LAK activities.
Materials and methods
Monoclonal antibodies Unconjugated anti-Leu-19 (CD56), phycoerythtin (PE)-conjugated anti-Leu-19, anti-Leu-4 (CD3) and anti-Leu-llc (CD16) (all 50 tlg/ml), fluorescein isothiocyanate (FITC)-conjugated antiLeu-12 (CD19) and anti-Leu-M3 (CD14) (25 /xg/ml) were purchased from Becton Dickinson (Mountain View, CA, U.S.A.). Unconjugated antiLeu-19, used in the separation procedure, was dialysed against phosphate-buffered saline (PBS)
to eliminate azide. Supernatant from the OKT3 hybridoma (CRL 8001, ATCC, Rockville, MD, U.S.A.) was used as a source of anti-CD3 MoAb. The HKB1 MoAb (IgM) produced in the Laboratory for Immunology, The Norwegian Radium Hospital, recognizes the gene products of all three major histocompatibility complex class II loci (DP, DQ, DR) (Holte et al., 1989). My7 MoAb (CD33) was purchased from Coulter Immunology (Hileah, FL, U.S.A.). Magnetic beads Dynabeads M-450 coated with sheep antimouse (SAM) IgG (M-450) (prod. no. 110.02), Dynal MPC-6, a neodynium magnet particle separator (Prod. nr. 120.02) was supplied by DYNAL, Oslo, Norway. Cell lines OKT3 hybridoma (ATCC CRL 8001) was cultured in DMEM (Gibco, Glascow, U.K.) containing 600 /~g/ml glutamine and 1 mM sodium pyruvate with 10% fetal calf serum (FCS) (Gibco). The NK-sensitive K562 cell line and the NK-resistant Daudi cell line were generously provided by Drs. Anthony Rayner and Brett Gemlo (Department of Surgery, University of California Medical School, San Francisco, CA, U.S.A.). These cell lines were cultured in RPMI 1640 (Gibco) containing 40 /~g/ml garamycin and 100 /~g/ml glutamine (complete RPMI) with 10% FCS. Immunornagnetic isolation of CD56 + cells (indirect technique) Bully coats were obtained from The Blood Bank, The University Hospital of Trondheim. All of the subsequent steps described were performed at 4°C, if not otherwise specified. Mononuclear cells (PBMC) were isolated by Lymphoprep (Nycomed, Oslo, Norway) centrifugation (Boyum, 1976) and washed four times with Hanks' balanced salt solution (HBSS) (Gibco). The PBMC were diluted in complete RPMI with 5% A + human serum (referred to as complete medium (CM)) to a concentration of 1-2 X 10 ~ cells/ml. Thereafter 0.08 ~tg anti-CD56/1 x 10 6 PBMC was added and the cells incubated for 1 h. The PBMC were then washed twice in HBSS and diluted in CM and incubated with 7.5 × 105 Dynabeads/106
PBMC for 30 rain in 10 ml tubes (Nunc, Roskilde, Denmark) with gentle rotation. Cells attached to the Dynabeads were recovered with the MPC-6 magnet and washed five times in CM. Thereafter the cell/bead suspension was incubated for 16-20 h at 37°C in 25 or 75 cm 2 culture flasks (Costar, Cambridge, MA, U.S.A.) so that most of the Dynabeads detached from the CD56 + isolated cells. Using the magnet, the cells were separated from the Dynabeads and were stained directly for flow cytometry analysis, whilst others were tested in cytotoxicity assays (together with unfractionated PBMC and CD56 depleted PBMC). A portion of the CD56 + isolated cells was incubated with 103 U / m l recombinant IL-2 (3 × 10 6 U / m g protein, Cetus Corp. Emeryville, CA, U.S.A.), at a concentration of 2 x 10 6 cells/ml for 2-3 days.
Immunomagnetic isolation of CD56 + LAK cells by anti-CD56 coated Dynabeads (direct technique) Anti-CD56 MoAb was added to the Dynabeads, at a concentration of 0.42 /~g MoAb to 1 mg Dynabeads, and incubated for 24 h on a rotator. The suspension was washed four times in CM using the magnet. PBMC depleted of monocytes by 90 min plastic adherence (PBL), were incubated for 3 days with 103 U / m l IL-2. CD56 + cells were isolated by adding 7.5 x 105 of the anti-CD56 coated Dynabeads to 1 × 10 6 PBL and separated as described. The isolated cells were analyzed for cytotoxicity against Daudi cells. Depletion of CD3 + cells from the CD56 + LAK cells 10 /~1 anti-CD3 MoAb from the OKT3 hybridoma (approximately 10 /~g M o A b / m l supernatant) were added to 1 mg Dynabeads and prepared as described for the anti-CD56 conjugated Dynabeads. CD56 + LAK cells were depleted of CD3 + cells by incubating 2 × 10 6 anti-CD3 coated Dynabeads with 1 × 10 6 CD56 + cells and separated as described. The CD3 depleted cells were also analyzed for cytotoxic activity against Daudi cells. Negative selection of NK cells PBL were incubated with HKBl-coated Dynabeads (directly adsorbed to unconjugated Dynabeads) and anti-CD3 coated Dynabeads, 7.5 ×
105 and 4.0 × 10 6 beads respectively, to 1 × 10 6 PBL for 30 rain in CM as described. The cells attached to the Dynabeads were removed and the unattached cells were analyzed by immunofluorescence microscopy and flow cytometry.
Immunofluorescence microscopy and flow cytometry Cells were stained with the appropriate antibody by incubating 1 × 106 cells with 0.5 /~g of PE- or 0.25 ~tg FITC-conjugated MoAb for 30 min. Control samples were stained with 0.13 /zg FITC-conjugated goat anti-mouse IgG (Becton Dickinson) or 0.20 /~g PE-streptavidin (Vector Laboratories, CA, U.S.A.) diluted in PBS with 0.1% BSA (Sigma, St.Louis, MO, U.S.A.). The cells were then washed twice in PBS/BSA before being fixed in 2% formaldehyde (Merck, Darmstadt, F.R.G.) in PBS for 20 rain. After one wash in PBS, 5000 cells were analyzed using a FACScan flow cytometer (Becton Dickinson). Two color fluorescence analysis was accomplished by incubating the cells with 0.5 /~g unconjugated antiLeu-4 MoAb, followed by three washes in PBS/BSA before 0.13 /zg FITC-conjugated goat anti-mouse IgG was added. After 30 min the cells were washed twice in PBS/BSA, and stained with 0.5/zg PE-conjugated anti-Leu-19 for 30 rain. After washing and fixation, the cells were analyzed by flow cytometry. The stained cells were also examined by immunofluorescence microscopy (Wild Leitz Orthoplan, Heerbrugg, Switzerland). Cytotoxicity assays The isolated cell populations were analyzed for cytotoxic activity against K562 and Daudi cells. About 3 x 10 6 target cells were labeled with 300 ~Ci Na~lCrO4 (Amersham, Buckinghamshire, U.K.) for 90 min at 37°C, followed by two washes in RPMI 1640 and a third wash in CM. 100 ~1 of target cells (105 cells/ml) were added to various concentrations of effector cells in triplicate in 96 well round bottomed microtiter plates (Costar). After 4 hours of incubation at 37°C, the supernatants (100 jal/well) were harvested, and radioactivity determined in a gamma counter (LKB Wallac 1270, Wallac Oy., Turku, Finland). Percent specific lysis was calculated as follows: % specific lysis
A-B C - B × 100
where A represents counts per minute (cpm) from test supernatants; B represents cpm from supernatants of target cells incubated alone (spontaneous release); and C represents cpm after lysis of target cells with 0.25% SDS (Sigma) (maximum release). Results are presented as the mean _+ SD of triplicate cultures. Spontaneous release did not exceed 20% of maximum release. Calculations The amounts of different cell types were recorded in replicate experiments and the results are presented as mean (SD).
CD56 E 40.2% =:
----4
''''t
...............
B
CD3 F 32.4%
C
CD19
CD56 78.5%
I ""'
......
P
........
t
.......
CD3 13.7%
.J ..I LU
o LL
o U.I
m
~ , ,
Results
G
9.9%.I1
CD19
2.8%
CD14 I H
CD14
/I
0.3%
z
Selection of CD56 + cells from unstimulated P B M C We first purified N K cells by depleting PBL for monocytes, B cells and T cells by using HKB1 and anti-CD3 coated Dynabeads as described in the materials and methods section. This resulted in an enrichment of CD16 + cells (71% (3), n = 3). However, the numbers of contaminating cells were too high (data not shown) and this approach resulted in a high consumption of Dynabeads. Positive selection of CD56 + cells was next performed by adding anti-CD56 coated Dynabeads to PBMC. This resulted in a low cellular yield ( < 1%). Thus, an indirect technique was tried by incubating PBMC with anti-CD56 MoAb before adding SAM coated Dynabeads. This method gave a satisfactory cellular yield for further analysis as described below. The purity of the CD56 + cell population was analyzed with monoclonal antibodies specific for different subsets of mononuclear cells. As shown in Fig. 1, the CD56 + selected cells were almost completely depleted of monocytes (CD14 +) and B lymphocytes ( C D ] 9 +) and the numbers of T cells (CD3 +) were markedly decreased. Based on five experiments, we calculated that there were 10% (6) CD19 + cells in the PBMC population, and 1% (1) in the CD56 + selected population. The amount of monocytes was 29% (4) before and 0.5% (0.5) after separation. The amount of CD3 + cells decreased from 42% (17) (n = 6) prior to isolation to 19% (8) (n = 6) in the CD56 + selected population. These results were verified by immunofluorescence mi-
D
28.3%
L/'/I
. . . . . . . . . . . . . . . .
FLUORESCENCE
I
........
I
........
INTENSITY (log)
Fig. 1. Flow cytometric analysis of PBMC (A, B, C, D) and the CD56 + isolated cells immunomagnetically separated from PBMC (E, F, G, H). The two different cell populations were stained with anti-Leu-19 (CD56) (A, E), anti-Leu-4 (CD3) (B,F) (both PE-conjugated), anti-Leu-12 (CD19) (C, G) and anti-Leu-M3 (CD14) (D, H) (both FITC-conjugated), respectively. The samples were analyzed on a FACScan flow cytometer. The dotted lines represent samples stained with PE-streptavidin or FITC-conjugated SAM IgG only (G and H, respectively). The analysis presented in A, B and C are gated on light scatter to eliminate the monocytes. The numbers in the histograms represent the percentage of cells positively stained with the respective MoAbs. The percentage levels of positive cells were estimated by setting limits and subtracting the percentage of cells defined as positive in the control sample from the positive cells in the MoAb treated sample. In panel E the intersection point between the CD56 and the control histogram was used for defining the positively and negatively stained cells. Similar data were obtained in at least four experiments.
O9 _1 ._1 ILl tO
CD56 91.3%
LI.
0
: .......>.../.., - , \ ................ \
W o3
Z
-'-7'~,,,,,
10 °
I
191 1
7'
t aJ~,,, I
102
jr,; ''''i
~ J''''''
10"4
FLUORESCENCE INTENSITY (log) Fig. 2. Flow cytometric analysis of the CD56 + isolated cells stained with PE-conjugated anti-Leu-19 (CD56) immediately after the immunomagnetic separation ( ...... ) and after 3 days incubation with IL-2, 103 U / m l ( ). The control ( . . . . . . ) represents IL-2-stimulated cells stained with PE-streptavidin only. The samples were analyzed on a FACScan flow cytometer. The percentage in the histogram represents the a m o u n t of CD56 ÷ cells after IL-2 stimulation. The levels of CD56 + cells registered immediately after separation reached 75.0%. Similar data were obtained in four experiments.
croscopy (data not shown). The presence of T lymphocytes in the CD56 + population was consistent with the fact that a fraction of the CD56 + cells are CD3 + (Lanier et al., 1986). To determine whether or not some of the CD56 + isolated cells were CD3 +, C D 5 6 - , we analyzed the isolated cells by two color immunofluorescence with anti-CD3 and anti-CD56. These data showed that only 0.5% (0.9) ( n = 4 ) of the isolated cells were CD3 +, CD56 +. In none of the experiments did the contaminating monocytes, B cells and C D 5 6 - , CD3 + T cells exceed 3%. The PBMC contained 31% ( l l ) (n = 7) CD56 + cells. Immunomagnetic isolation of these cells resulted in a reduced immunofluorescence intensity when stained with anti-CD56 (Fig. 1). This made an exact quantification of CD56 + cells in the positively selected population difficult. However, we provisionally recorded 75% (9) (n = 7) CD56 + cells. These data were verified by immunofluorescence microscopy (data not shown). To overcome the problem of low CD56 expression, we stimulated the CD56 + selected cells with IL-2 for 3 days before the amount of CD56 + cells was measured by flow cytometry. As shown in Fig. 2, the
addition of IL-2 resulted in an increased expression of the CD56 antigen which made quantification of the amount of CD56 + cells more accurate. The isolation procedure thus gave a population consisting of 92% (3) (n = 4) CD56 + cells as determined by flow cytometry. By immunofluorescence microscopy (counting 400 cells) > 94% of the positively selected cells were CD56 +. The addition of IL-2 for 3 days did not increase the total number of cells in the CD56 + population (data not shown), indicating no selective proliferation of CD56 + cells. Furthermore, the numbers of CD3 + cells remained unaltered after IL-2 stimulation (data not shown). The method described gave a cellular yield of 5.3% (1.3) (n = 8) of the PBMC. Consequently, approximately 1.4-2.3 × 1 0 7 CD56 + cells could be obtained from a representative buffy coat containing approximately 3.5 × 108 mononuclear cells. The viability of the CD56 selected cells was found to be > 91% as determined by trypan blue exclusion.
• CD56+ cells • PBMC
6O
40
20
I 1.57:1
I 3.13:1
_ _ l 6.25:1
I 12.5:1
I 25:1
EFFECTOR : TARGET CELL RATIO
Fig. 3. Cytotoxic activity against K562 cells mediated by PBMC (A A), CD56 + cells (O O) and CD56 depleted PBMC ( I i ) immediately after immunomagnetic isolation as described in the materials and methods section. Results are presented as mean_+SD of triplicate determinations. Similar data were obtained in three experiments.
60
• CD56+ cells (+IL-2) • PBMC (+IL-2)
t
• CD56 depleted PBMC (+IL-2) 50
.[
o CD56+ ceils ( I L - 2 )
~
4o
N K activity of the CD56 + selected cells. After incubation with IL-2 for 2 - 3 days, the CD56 + cells expressed high cytotoxic activity against Daudi cells compared to the PBMC and the CD56 depleted population (Fig. 4). These results suggest that high L A K activity can be induced in the CD56 + population.
co (D ~ 3O
% 2O
lC
1.57:1
3.13:1
6.25:1
12.5:1
25:1
EFFECTOR : TARGET CELL RATIO
Fig. 4. Cytotoxic activity against Daudi cells, mediated by CD56 + cells in the absence of IL-2 ( o O), and CD56 + cells ( e o), PBMC (A A) and CD56 depleted PBMC ( I II) after incubation with IL-2 (103 U / m l ) for 3 days. Results are presented as mean+_SD of triplicate determinations. Similar data were obtained in three experiments.
Functional properties of CD56 + cells The functional properties of the CD56 + cells were tested against the N K sensitive K562 cell line and the N K resistant Daudi cell line. Fig. 3 demonstrates that the capacity of the CD56 + enriched cells to kill the K562 cells was high compared with unfractionated PBMC and PBMC depleted with respect to CD56 + cells. This result is consistent with an N K cell enrichment in the CD56 + population. There was no significant killing of Daudi cells by the CD56 + population (Fig. 4) or the PBMC population (data not shown). To test whether the separation procedure altered N K activity, we incubated PBL with either Dynabeads alone or firstly with anti-CD56 M o A b and secondly with the Dynabeads. After 20 h the Dynabeads were separated from the PBL, which then were tested for cytotoxicity against K562. These manipulations of the effector cells did not change their ability to kill K562 cells compared to unmanipulated PBL (data not shown). Thus, the separation technique did not interfere with the
Isolation of CD56 + L A K cells from IL-2 stimulated P B L and depletion of CD3 + cells from the CD56 + L A K cells Immunomagnetic isolation, using anti-CD56 coated Dynabeads (direct technique), was also performed on PBL incubated for 3 days with 103 U / m l IL-2. In contrast to unstimulated PBMC treated with anti-CD56 coated Dynabeads, this approach resulted in a high cellular yield, 5.4% (3.1) (n = 8), of CD56 + selected cells. The CD56 ~ L A K cells were further depleted for CD3 + cells by immunomagnetic isolation using anti-CD3 coated Dynabeads. The different cell populations
60 • CD56+ LAK celts • CD56 depleted PBL (LAK~ • CD3 depleted Cb56+ LAK cells
4o co _o u_ co
0
0.6:1
1.2:1
I 2.5:1
I 5:1
I 10:1
EFFECTOR : TARGET CELL RATIO
Fig. 5. Cytotoxic activity against Daudi cells mediated by immunomagnetically isolated cells from IL-2 stimulated PBL (103 U / m l for 3 days). The effector ceils tested were CD56 + LAK cells (O O), CD3 depleted CD56 + LAK ceils (A --A) and CD56 depleted PBL (LAK) (m -I). Results are presented as mean+_SD of triplicate determinations. Similar data were obtained in two experiments.
were examined for LAK activity against Daudi cells. The data show that the CD56 + cells, prior to CD3 depletion, expressed a much higher cytotoxic activity than the CD56 depleted LAK cells and that depletion of CD3 + cells resulted in an additional increase in LAK activity (Fig. 5).
Discussion
In this paper we have described an immunomagnetic isolation technique which resulted in a satisfactory amount of highly enriched and functionally active N K / L A K cells. We found that the isolated cells contained 92% CD56 + cells and no more than 3% monocytes, B cells and CD56-, CD3 + T cells. Granulocyte contamination (as determined by immunofluorescence staining with MoAb My7 (CD33)) was negligible ( < 0.5%, data not shown). Approximately 5% of the CD56 isolated cells were negative by flow cytometry for CD56, CD3, CD19 or CD14. T cells and B cells are strictly defined by the CD3 and CD19 markers, respectively (Kung et al., 1979; Dorken et al., 1989). CD14 is known to be expressed on approximately 90% of the monocytes (Dimitriu-Bona et al., 1983) and this indicates that some C D 1 4 monocytes could be present in the isolated population. However, we have observed that monocytes efficiently engulf the Dynabeads at 37°C and consequently will be removed together with the beads after overnight incubation (data not shown). Any significant monocyte contamination in the CD56 + population is therefore unlikely. The immunomagnetic isolation procedure resulted in reduced CD56 expression (Fig. 1), which could be due to CD56 antigen shedding a n d / o r low turnover and renewal of the antigen when being manipulated with the Dynabeads (Lea et al., 1986; Funderud et al., 1990). The expression of CD56 was increased in the presence of IL-2 (Fig. 2), which is in accordance with Lainier et al. (1989). This CD56 upregulation was not due to proliferation of CD56 + cells and is consistent with earlier reports that have shown no increase in total cell numbers after IL-2 incubation of lymphocytes or N K cells for 3 days (Talmadge et al., 1986; Yamada et al., 1987). The possibility exists that a small percentage of the isolated cells will not
upregulate their CD56 antigen after the addition of IL-2, which then would be detected as CD56-, CD3 , C D 1 9 - and C D 1 4 - cells. Conclusively, the isolated cell population appears to contain ~< 3% contaminating cells, as registered by flow cytometry. Despite the reduction in CD56 antigen expression immediately after the immunomagnetic isolation, the enriched N K cells showed high cytotoxicity against K562. Furthermore, the isolation procedure did not stimulate the N K cells to kill the N K resistant Daudi cells. The addition of IL-2 resulted in high LAK activity by the CD56 + cells. The raised N K and LAK activity found in the CD56 + population compared to the unseparated or CD56 depleted cell populations is in accordance with the described enrichment of N K cells. The immunomagnetic isolation technique, using anti-CD56 MoAb, did not activate or impair the functional abilities of the N K cells. This result also agrees with earlier studies demonstrating that CD56 antibodies do not alter the functional abilities of N K cells (Lanier et al., 1986), and the previously described immunomagnetic techniques for B and T cell isolation showing no functional alterations (Gaudernack et al., 1986; Funderud et al., 1990). We therefore conclude that the immunomagnetically isolated CD56 + cells have preserved their functional characteristics after the separation procedure. The N K / L A K cells in the CD56 + population was further purified by depletion of T cells with anti-CD3 coated beads. This resulted in increased LAK activity (Fig. 5) which is in accordance with an activated CD56 +, C D 3 - N K cell as the predominant LAK cell (Phillips and Lanier, 1986). To further increase the N K cell purity, any additional T cells, B cells and monocytes can be removed with Dynabeads coated with a n t i - C D 4 / anti-CD8 (Lea et al., 1985), anti-CD19 (Funderud et al., 1990) and ID5 (monocyte specific) (Kaplan and Gaudernack, 1982), antibodies respectively. In conclusion, we have successfully adapted the immunomagnetic separation technique for the isolation of CD56 + cells. This technique should be of value when it is intended to undertake functional studies on isolated N K / L A K cells. Because of its simplicity, the method is preferable to previously published methods.
Acknowledgements We thank Dagmar Moholdt for assistance in preparing the manuscript. This work was supported by the Norwegian Cancer Society, Norsk Hydro and Dynal, Norway.
References Boyum, A. (1976) Isolation of lymphocytes, granulocytes, and macrophages. Scand. J. Immunol. 5 (suppl. 5), 10. Dimitriu-Bona, A., Burmester, G.R., Waters, S.J. and Winchester, R,J. (1983) Human mononuclear phagocyte differentiation antigens. I. Patterns of antigenic expression on the surface of human monocytes and macrophages defined by monoclonal antibodies. J. Immunol. 130, 145. D~Srken, B., MiSller, P., Pezzutto, A., Schwartz-Albiez, R. and Moldenhauer, G. (1989) B-cell antigens: CD19. In: W. Knapp, B. D~Srken, W.R. Gilks, E.P. Rieber, R.E. Schmidt, H. Stein and A.E.G.Kr. Von dem Borne (Eds.), Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press, Oxford. Funderud, S., Erikstein, B., Asheim, H.C., Nustad, K., Stokke, T., Kiil Blomhoff, H., Holte, H. and Smeland, E.B. (1990) Functional properties of CD19+B lymphocytes positively selected from buffy coats by immunomagnetic separation. Eur. J. Immunol. 20, 201. Garcia-Peharrubia, P., Koster, F.T., Kelley, R.O., McDowell, T.D. and Bankhurst, A.D. (1989) Antibacterial activity of human natural killer cells. J. Exp. Med. 169, 99. Gaudernack, G., Leivestad, T., Ugelstad, J. and Thorsby, E. (1986) Isolation of pure functionally active CD8+T cells. Positive selection with monoclonal antibodies directly conjugated to monosized magnetic microspheres. J. Immunol. Methods 90, 179. Grimm, E.A., Mazumder, A., Zhang, H.Z. and Rosenberg, S.A. (1982) Lymphokine-activated killer cell phenomenon. Lysis of natural killer-resistant fresh solid tumor cells by interleukin 2-activated autologous human peripheral blood lymphocytes. J. Exp. Med. 155, 1823. Henney, C.S., Kuribayashi, K., Kern, D.E. and Gillis, S. (1981) lnterleukin-2 augments natural killer cell activity. Nature 291, 335. Hercend, T., Meuer, S., Reinherz, E.L. Schlossman, S.F. and Ritz, J. (1982) Generation of a cloned NK cell line derived from the 'null cell' fraction of human peripheral blood. J. lmmunol. 129, 1299. Holte, H., Kiil Blomhoff, H., Beiske, K., Funderud, S., Torjesen, P., Gaudernack, G., Stokke, T. and Smeland, E.B. (1989) Intracellular events associated with inhibition of B cell activation by monoclonal antibodies to HLA class II antigens. Eur. J. Immunol. 19, 1221. Kan, E.A.R., Wang, C.Y., Wang, L.C. and Evans, R.L. (1983) Non-covalently bonded subunits of 22 and 28 kd are rapidly
internalized by T cells reacted with anti-leu-4 antibody. J. Immunol. 131,536. Kaplan, G. and Gaudernack, G. (1982) In vitro differentiation of human monocytes. Differences in monocyte phenotypes induced by cultivation on glass or on collagen. J. Exp. Med. 156, 1101. Kung, P.C., Goldstein, G., Reinherz, E.L. and Schlossman, S.F. (1979) Monoclonal antibodies defining distinctive human T cell surface antigens. Science 206, 347. Lanier, L.L., Le, A.M., Civin, C.I., Loken, M.R. and Phillips, J.H. (1986) The relationship of CDI6 (Leu-ll) and Leu-19 (NKH-1) antigen expression on human peripheral blood NK cells and cytotoxic T lymphocytes. J. Immunol. 136, 4480. Lanier, L.L., Testi, R., Bindl, J. and Phillips, J.H. (1989) Identity of Leu-19 (CD56) leukocyte differentiation antigen and neural cell adhesion molecule. J. Exp. Med. 169, 2233. Lea, T., Vartdal, F., Davies, C. and Ugelstad, J. (1985) Magnetic monosized polymer particles for fast and specific fractination of human mononuclear cells, Scand. J. Immunol. 22, 207. Lea, T., Smeland, E., Funderud, S., Vartdal, F., Davies, C.. Beiske, K. and Ugelstad, J. (1986) Characterization of human mononuclear cells and positive selection with immunomagnetic particles. Scand. J. Immunol. 23, 509. London, L., Perussia, B. and Trinchieri, G. (1985) Induction of proliferation in vitro of resting human natural killer cells: expression of surface activation antigens. J. lmmunol. 134, 718. Nagler, A., Lanier, L.L., Cwirla, S. and Phillips, J.H. (1989) Comparative studies of human FcRIll-positive and negative natural killer cells. J. lmmunol. 143, 3183. Ortaldo, J.R., Mason, A. and Overton, R. (1986) Lymphokineactivated killer cells. Analysis of progenitors and effectors. J. Exp. Med. 164, 1193. Ostensen, M.E., Thiele, D.L. and Lipsky, P.E. (1987) Tumor necrosis factor-c~ enhances cytolytic activity of human natural killer cells. J. Immunol. 138. 4185. Perussia, B., Starr, S., Abraham~ S., Fanning, V. and Trinchieri, G. (1983) Human natural killer cells analyzed by B73.1, a monoclonal antibody blocking Fc receptor functions. I: characterization of the lymphocyte subset reactive with B73.1. J. Immunol. 130, 2133. Phillips, 3.H. and Lanier, L.L. (1986) Dissection of the lymphokine-activated killer phenomenon. Relative contribution of peripheral blood natural killer cells and T lymphocytes to cytolysis. J. Exp. Med. 164, 814. Talmadge, J.E., Wiltrout, R.H., Counts, D.F., Herberman, R.B., McDonald, T. and Ortaldo, J.R. (1986) Proliferation of human peripheral blood lymphocytes induced by recombinant human interleukin 2: Contribution of large granular lymphocytes and T lymphocytes. Cell. lmmunol. 102, 261. Timonen, T. and Saksela, E. (1980) Isolation of human NK cells by density gradient centrifugation. J. Immunol. Methods 36, 285. Timonen, T., Saksela, E., Ranki A. and H~yry, P. (1979) Fractionation, morphological and functional characteriza-
tion of effector cells responsible for human natural killer activity against cell-line targets. Cell. Immunol. 48, 133. Timonen, T., Ortaldo, J.R. and Herberman, R.B. (1981) Characteristics of human large granular lymphocytes and relationship to natural killer and K cells. J. Exp. Med. 153, 569. Timonen, T., Reynolds, C.W., Ortaldo, J.R. and Herberman, R.B. (1982) Isolation of human and rat natural killer cells. J. Immunol. Methods 51, 269. Trinchieri, G., Matsumoto-Kobayashi, M., Clark, S.C., Seehra, J., London, L. and Perussia, B. (1984) Response of resting human peripheral blood natural killer cells to interleukin 2. J. Exp. Med. 160, 1147.
Vartdal, F., Kvalheim, G., Lea, T.E,, Bosnes, V., Gaudernack, G., Ugelstad, J. and Albrechtsem D. (1987) Depletion of T lymphocytes from human bone marrow. Transplantation 43, 366. Yamada, S., Ruscetti, F.E., Overton, W.R., Herberman, R.B., Birchenall-Sparks, M.C. and Ortaldo, J.R. (1987) Regulation of human large granular lymphocyte and T cell growth and function by recombinant interleukin 2: Induction of interleukin 2 receptor and promotion of growth of cells with enhanced cytotoxicity. J. Leukocyte Biol. 41,505.
