CELLULAR
IMMUNOLOGY
125326-336
(1990)
Interleukin-1 cvand Tumor Necrosis Factor-a Protect Cells against Natural Killer Cell-Mediated Cytotoxicity and Natural Killer Cytotoxic Factor’ ZVI REITER AND MENACHEM
RUBINSTEIN~
Department ofMolecular Genetics and Virology, The Weizmann Institute of Science, Rehovot, 76100, Israel Received July 12, 1989; accepted October 5, I989 The protective effectsof interferons (IFNs) against NK cell-mediated cytotoxicity (NK-CMC) is well established. We report here that both recombinant tumor necrosis factor-a (TNF-a) and recombinant interleukin- I (Y(IL- 1a) can also protect some adherent target cells (e.g., the amniotic cells WISH and the cervical epithelial carcinoma cells HeLa-229) from NK-CMC in a dosedependent manner. Like in the case of IFNs, the level of conjugate formation between target and effector cells (nonadherent peripheral blood lymphocytes) is not affected by pretreatment of the target cells with either TNF-ol or IL-la. However, while the main effect of IFNs is to reduce the ability of target cells to stimulate the release of NK cytotoxic factor (NKCF) from effector cells, TNF-a and IL- 1a do not affect this processbut rather reduce the target cell sensitivity to the lytic effect of NKCF. Therefore TNF-ol and IL 1a induce resistance to NK-CMC by a mechanism that differs from the one attributed to IFNs. The protective effect of TNF-a and ILo 1990 Academic PIES, IX. 1LYis not mediated by the induction of EN-&/IL-6.
INTRODUCTION Interleukin- 1 (IL- I), a macrophage-derived cytokine and tumor necrosis factor(Y(TNF-a), a macrophage-derived cytotoxin, share some regulatory activities with interferons (IFNs). These cytokines directly inhibit the growth of some tumor cells (1, 2), activate or promote the differentiation of cytotoxic T lymphocytes (1, 2) and enhance the cytocidal activities of monocytes (3-6). Furthermore, it was reported recently that both IL- 1 and TNF augment the cytotoxic activity of natural killer (NK) cells in a manner similar to that of IFNs ( 1,7). The sensitivity of various cells to NK cell-mediated cytotoxicity (NK-CMC) is significantly inhibited by pretreatment of the target cells with IFNs (8-25). Although the mechanism of this protective effect is not yet understood, it was found in the murine system that IFN lowered the ability of target cells to induce the releaseof NK cytotoxic factor (NKCF) upon binding to NK cells (26). In this study, we demonstrate that certain cells acquire resistance to NK-CMC and to NKCF by pretreatment with either TNF-(U or IL- 1-a by a mechanism that differs ’ This work was supported by a grant from InterYeda Ltd., Nes-Ziona, Israel. ’ M.R. has the Maurice and Edna Weiss Chair of Interferon Research.
0008-8749/90 $3.00 Copyright All riglm
0 1990 by Academic Press, Inc. of reproduction in any form resewed.
IL-la
AND
TNF-a
PROTECT
CELLS FROM
NK-CMC
AND
NKCF
327
from that of IFNs. This protective effect is not mediated via the induction of IFN-/3,/ IL-6 by these cytokines. MATERIALS AND METHODS Cells NK efictor cells. Mononuclear cells were isolated from freshly collected peripheral blood of healthy donors by density gradient sedimentation on Ficoll-Paque (27). Adherent cells were removed by adsorption to nylon wool (28). The nonadherent cells were used as effector cells. Target and stimulator cells. WISH cells, a human amniotic cell line (ATCC CCL 25) and HeLa-229, a human cervical epitheloid carcinoma cell line (ATCC CLL 2.1) were used as target cells. K-562 cells, a human erythromyeloid leukemia cell line (ATCC CCL 243) and U-937 cells, a human monocytic cell line (ATCC CRL 1593) were used as stimulator cells for induction of the release of NKCF by NK effector cells. All the cells were grown and assayedat 37°C in a humidified 5% CO2 incubator. Reagents and Media Homogeneous Escherichia coli rIFN-LuC (sp activ 10’ units/mg), homogeneous rIFN-y from recombinant Chinese hamster ovary (CHO) cells (sp activ 5 X 10’ units/ mg) and homogenous rIFN-&/IL-6 (CHO) (sp activ 2 X lo6 HGF units/mg) were from Interpharm Laboratories, Ness-Ziona, Israel. E. coli rTNF-a (sp activ 2 X 10’ units/mg) was a gift from Dr. D. Wallach. E. coli rIL- la (sp activ 5 X lo6 units/mg) was a gift from Hoffmann-La Roche Inc. (NJ). Ficoll-Paque was from Pharmacia, Sweden. [5’Cr]-sodium chromate (sp activ 15 to 40 Ci/mmol) was from the Nuclear Research Center, Negev, Israel. RPMI- 1640 medium, minimal essential medium (MEM) and fetal calf serum (FCS) were from GIBCO. Effector cells as well as K-562, U-937, and HeLa-229 cells were cultured in RPMI1640 medium supplemented with FCS (lo%), glutamine (2 mM), penicillin (100 units/ml) and streptomycin (100 pg/ml), (RPMI-10%). WISH cells were cultured in MEM supplemented similarly (MEM- 10%). Treatment qfNK Efector and Target Cells with Cytokines Nonadherent peripheral blood mononuclear cells (5 X 106/ml) were cultured in medium (RPMI- 10%) containing IFN-(YC ( 1000 units/ml, 14- 18 hr, 37°C) prior to use as effector cells. WISH and HeLa-229 cells were seeded in 96-well flat-bottom plates (4 X lo4 cells/well, 100 ~1 medium) and incubated for 24 hr at 37°C. Various cytokines were then added (50 ~1) in different concentrations and incubation was continued for 12- 14 hr. “Cr ReleaseAssay of NK-CMC This assaywas performed with some modification as described (29). Briefly, HeLa229 and WISH cells (4 X lo4 cells/well, 100 ~1) were seededin 96-well flat-bottom plates and incubated for 24 hr at 37°C. [5’Cr]-sodium chromate (1 &i/well) in medium alone (RPMI- 10%or MEM- lo%, 100 ~1,respectively) or in medium containing the indicated cytokines, was then added and incubation continued for 12- 14 hr. The
328
REITER
AND
RUBINSTEIN
plates were then rinsed three times, various amounts of IFN-activated effector cells in RPM1 10%(200 ~1)were added, incubation was continued for 5-6 hr at 37”C, and aliquots of the supernatants were then collected for counting released radioactivity which was defined as experimental cpm. Total cpm was obtained by adding 1% Triton X- 100 to wells with target cells. Spontaneous release of “Cr (7- 12% and 8- 14% in HeLa-229 and WISH cells, respectively) was determined in wells containing target cells only. Data are expressed as percentage of “Cr-specific release according to the formula: % specific release = 100 X
experimental cpm - spontaneous cpm total cpm - spontaneous cpm
Each data point represents an average of 5-8 experiments, each done in triplicates with effector cells from a different donor. Conjugate Formation Assay Target cells (HeLa-229 and WISH) were seeded in 96-well flat-bottom plates ( 1OOO/wellin 50 ~1 RPMI- 10% or MEM- lo%, respectively) and incubated for 24 hr at 37°C. Effector cells (2 X 103/well in 50 ~1RPMI- 10% or MEM- 10%) were added, spun (15g, 5 min), and incubated for various periods. Following incubation, trypan blue (l%, 50 ~1) was added followed by glutaraldehyde (3%, 50 ~1). The number of conjugates was determined by counting under a light microscope. Each sample was assayedin duplicates. Data are expressed as percentage of binding according to the formula: % binding =
conjugated effector cells x 100 total effector cells
Preparation of Crude NKCF IFN-activated effector cells (5 X 106/ml) were cocultured with stimulator cells (5 104/ml) in RPMI-1640 medium supplemented with 1% FCS for 24 hr at 37°C in a humidified atmosphere of 5% COZ. Preliminary experiments indicated that these conditions gave maximal NKCF activity. Following incubation, the supernatant was harvested and concentrated five- to eightfold by ultrafiltration on an Amicon YM-5 ultrafiltration membrane. The concentrated supernatants were passed through a 0.45-pm filter, and either used immediately or kept frozen at -70°C until used. Concentrated supernatants from cultures of either stimulator cells or effector cells were used as control. X
Assay of NKCF Two different assayswere used for the measurement of NKCF activity: (a) 5’Cr release assay of NKCF. Either HeLa-229 or WISH cells (4 X lo4 cells/ well, 100 ~1 RPMI-1% or MEM- I%, respectively) were seededin 96-well plates and incubated in the presence of “0-sodium chromate (1 &i/well) for 24 hr at 37°C. Various concentrations of the different cytokines in 100 ~1medium were added and incubation continued for 12- 14 hr. The plates were then rinsed three times, various dilutions of crude NKCF (in 100 ~1medium) were added and incubation continued
IL-la
AND
TNF-a
PROTECT
CELLS FROM
NK-CMC
AND
NKCF
329
for an additional period of 20-24 hr at 37°C. Aliquots of the supernatants were then collected for radioactivity measurement. Total cpm, spontaneous cpm and “0-specific releasewere calculated as described in “Cr releaseassayof NK-CMC. (b) Neutral Red uptake assay.Either HeLa-229 or WISH cells (4 X lo4 cells/well, 100 ~1)were seededin 96-well plates and incubated for 24 hr at 37°C. Various concentrations of the different cytokines in 100 ~1medium were added and incubation continued for 12- 14 hr. The plates were then washed three times, various dilutions of crude NKCF (in 100 ~1RPMI- 1% or MEM- l%, respectively) were added, and incubation continued for an additional period of 44 hr. Viability of the cells was determined by incubation with neutral red (100 &well) for 2 hr, extracting the neutral red that was taken up by the cells with Sorenson’s citrate buffer-ethanol mixture ( 150 &well), and quantitating calorimetrically at 570 nm by a micro ELISA autoreader (30). The extent of lysis by NKCF was calculated according to the formula: % Viability =
Dye uptake by cells treated with NKCF x 1o. Dye uptake by untreated cells
Each data point represents an average of a triplicate. Statistics “Cr release NKCF, and conjugate formation assayswere expressedas the arithmetic mean f standard error of mean. Comparisons were carried out by student’s t test. RESULTS IL-la and TNF-a Protect Target Cells from NK-CMC Preincubation of HeLa-229 and WISH cells with either IL- 1CYor TNF-a rendered them resistant to NK-CMC. As a positive control, the protective effect of IFN-(UC and IIN-7 was demonstrated under the same conditions. The effect of all tested cytokines was significant (P < 0.05) and dose dependent. Moreover this protective effect was observed in all tested effector to target ratios (Figs. 1, 2). When these cells were pretreated with IL-6 no such protective effect was observed (data not shown). In contrast to the above results with adherent target cells, very little or no protective effect was observed with K-562 or U-937 target cells when pretreated with either IL-la or TNF-(u, whereas significant protection was obtained when these cells were pretreated with either IFN-arc or IFN--/ (data not shown). Lack of Eflect of IL-l (Yand TNF--a or Conjugate Formation betweenNK Efector and Target Cells Either HeLa-229 or WISH target cells were preincubated with various concentrations of cytokines (II+&&, IFN--y 10-1000 units/ml; IL-6 0.01-10 PM; IL- 1CY5500 units/ml, TNF-a lo-200 units/ml) for 12 hr. NK effector cells were then added and incubation continued for 30-240 min. No differences in the level of conjugate formation were observed with the various cytokines as compared with control cells. A typical result following 1 hr incubation with effector cells is shown in Table 1.
330
REITER AND RUBINSTEIN
IL-la
40 -
I\
20 30.k z
IFN- d:
1
0
20
40
60
80
100
0
I
20
I
L
40
I
I
60
I
I
80
I
I
100
CYTOKINES CONCENTRATION (units/ml)
FIG. 1. The protective effect of IL-la and TNF-a against NK-CMC on HeLa-229 cells. Wr-labeled HeLa-229 cells were pretreated with various concentrations of IL- 101or TNF-a for 12- 14 hr. The protective effect of IFN-aC and IFN-7 is given as a positive control. The cells were washed and incubated with NK effector cells at different E:T rations (0, 40: 1; W,20: 1; A, 10:1). The curves represent an arithmetic mean f standard error of experiments with NK effector cells from four different healthy donors.
Lack of Eflect of IL-l (Yand TNF-LXon the Ability of Target Cells to Stimulate NKCF Generation HeLa-229 and WISH cells were pretreated with IL- 1(Y(500 units/ml); TNF-a (200 units/ml); IFN-(YC ( 1000 units/ml); IFN-7 ( 1000 units/ml) or IL-6 ( 1 PM) for 12 hr, washed, and NK effector cells (E:T-50: 1) were added. After 24 hr of coculturing, aliquots of the supernatants were collected, concentrated, and assayed by the 5’Cr release assay for NKCF (Fig. 3) and by the neutral red uptake assay(Fig. 4). In both assaysit was shown that while IFN-aC and IFN-y reduced significantly (P < 0.05) the level of NKCF generation, the other cytokines (IL- 1cy,TNF+, and IL-6) did not affect the ability of HeLa-229 and WISH cells to induce NKCF generation. IL-l (Yand TNF-a Protect Target Cells from NKCF Preincubation of either HeLa-229 or WISH cells with either IL- la (500 units/ml) or TNF-(Y (200 units/ml) reduced significantly (P < 0.05) the sensitivity of the cells to the cytotoxic effect of NKCF, as determined both by “Cr releaseassay(Fig. 5) and by the neutral red uptake assay (Fig. 6). A significant (P < 0.05) protective effect against NKCF was observed with either IL- 1(Yor TNF-a! only after 6-8 hr of incubation with these cytokines and was maximal after lo- 12 hr. In contrast, pretreatment
IL-la AND TNF-(U PROTECT CELLS FROM NK-CMC AND NKCF
331
IFNq
20
ih,,
:$g
10
’
I
)
0
I
1
20
.
I
I
40
I
60
I
B
I
60
CYTOKINES
I
I
100
I
,
0
CONCENTRATION
I I I I 1 1 I I 20 40 60 60
I 100
(units/ml)
FIG. 2. The protective effect of IL-la and TNF-LUagainst NK-CMC on WISH cells. For details seelegend to Fig. 1.
of these cells as well as nonadherent cells (K-562, U-937) with either IFN-(YC, IFN--/ or IL-6 did not exert protection against NKCF activity (data not shown). DISCUSSION NK cells may have a roIe in immune surveillance against the development of at least certain types of spontaneous tumors, as well as against the metastatic spread of TABLE 1 Lack of Effect of IL-n and TNF-ol on Conjugate Formation Between NK Effector and Target Cells Conjugate formation (%)” HeLa-229 No treatment IL-la (500 units/ml) TNF-a (200 units/ml) IFN-olC (1000 units/ml) IFN--/ (1000 units/ml) IFN-&/IL-6 (1 PM)
6.8 + 1.3 6.2 f 1.6 7.6 -t 0.9 7.1 ?z 1.2 6.4 -c 1.3 6.9 + 1.1
WISH 7.1 * 8.0 * 7.6 i 7.9 f 6.8 k 6.7 f
1.8 1.6 1.2 1.1 1.4 1.3
’ The percentage conjugate formation represents an arithmetic mean f standard error of experiments with NK effector cells from three different donors.
REITER AND RUBINSTEIN
1:2
1 :4
1:2 1:8 CRUDE NKCF (DILUTIONS)
114
1:8
FIG. 3. Lack of effect of IL-la, TNF-LU,and IM-&/IL-6 on the ability of target cells to stimulate NKCF generation (“0 release assay). Either HeLa-229 or WISH cells were pretreated for 12 hr with various cytokines. Effector cells were added, the supematants were collected after 24 hr and assayedby the “Cr release test on the same target cells (HeLa-229-A or WISH-B, respectively). Cytokines were: IL-la (500 units/ml, n ); TNF-ol(200 units/ml, A); IFN-&/IL-6 (1 PM, +); Medium alone, (0). IEN-& (1000 units/ ml, 0) and IFN--y (1000 units/ml, A) were used as positive controls. The curves represent an arithmetic mean (standard error was f 1.O-2.6%) of experiments with NKCF from three different healthy donors.
tumors (3 1, 32). Interferons reduced the susceptibility of target cells to NK-CMC (S25). Other cytokines were tested for this activity as well. Crude preparations of the contact inhibitory factor (CIF) devoid of IFN activity reduced the susceptibility of some tumor cells to NK-CMC (33). Crude IL-2 preparations were reported to induce a significant decreasein the susceptibility of YAC cells to murine NK-CMC (34,35). Since highly purified IL-2 preparations did not exhibit a similar effect, it was con-
t
B.
T
50 -L 1 :2
114
1:2 1:8 CRUDE NKCF (DILUTIONS)
1:4
FIG. 4. Lack of effect of IL- 1(Y,TNF-(Y and IFN-&/IL-6 on the ability of target cells to stinnulate NKCF generation (neutral red assay). NKCF was induced by cytokine-treated target cells (see Fig. 3) and was measured in mixed-culture supematants by the neutral red uptake assay on HeLa-229 (A) and WISH (B) cells. Cytokines were: IL- 101(500units/ml, q ); TNF-ol(200 units/ml, m); and IFN-/&/IL-6 ( 1 PM, o~C(1000 units/ml, @I)and IFN--/ (1000 units/ml, n ) were both used as positive controls. Medium alone, (0). The histograms represent an arithmetic mean f standard error of experiments with NKCF from three different healthy donors.
IL-la
AND
TNF-o( PROTECT
CELLS FROM
NK-CMC
1:2 114 1:8 CRUDE NKCF (DILUTIONS)
1:2
AND
1A
333
NKCF
1:8
FIG. 5. IL- 101and TNF-cu protect target cells from killing by NKCF (5’Cr release assay). Either HeLa229 (A) or WISH (B) cells were preincubated for 12 hr with IL-la (500 units/ml, n ); TNF-cu (200 units/ ml, A); IFN-(uC (1000 units/ml, 0); IFN-y (1000 units/ml, a); IFN-&/IL-6 (I p&f, +), or medium alone (0). The cultures were washed, various dilutions of crude NKCF preparations were added, and percentage ofcytotoxicity was determined. Each curve represents an arithmetic mean (standard error was + 0.7-2.1%) of two experiments using separate NKCF preparations from four healthy donors.
eluded that other cytokines present in the crude IL-2 preparations were probably responsible for the induced resistance to lysis. There is however another study where IL-2 was found to reduce the sensitivity of primary hairy leukemia cells to activated NK cells (36). In the present study we identified both IL-l (Yand TNF-a as the agents that reduce the susceptibility of target cells to NK-CMC at very low concentrations.
1:2
1 14
I :8 1:2 CRUDE NKCF (DILUTIONS)
114
1:8
FIG. 6. IL-la and TNF-(U protect target cells from killing by NKCF(neutral red assay). Either HeLa-229 (A) or WISH (B) cells were preincubated for 12 hr with IL-la (500 units/ml, EI); TNF-(U (200 units/ml, Q; IFN-& (1000 units/ml, q ); IIN--, (1000 units/ml, n ); IFN-&/IL-6 (1 PM, ) or medium alone (Cl). The cultures were washed, various dilutions of crude NKCF preparations were added, and percentage of cytotoxicity wasdetermined. The histograms represent an arithmetic mean f standard error oftwo experiments using separate preparations from four healthy donors.
334
REITER
AND
RUBINSTEIN
This effect was located at the level of target cell lysis by NKCF and not at the initial step of conjugate formation between NK effector and target cell. Since TNF and IL1 can protect cells from lysis by TNF (37, 38), their ability to confer resistance to NKCF is not unexpected. Nevertheless, NKCF differs considerably from TNF in various aspects.For example we found that unlike TNF, crude NKCF kills cells even in the absence of cycloheximide. Moreover, we found also that crude NKCF does not confer protection against TNF or NKCF. Immunological and other differences between NKCF and TNF were reported in the literature. NKCF differed from TNF and lymphotoxin with respect to target cell specificity and was not neutralized by antibodies to the other cytotoxins (39-42). Although TNF was produced by NK cells (43) it was not involved in NK-CMC (42). Finally liposomes made from acidic phospholipids were lysed by NKCF and not by TNF (44). TNF and IFNs have some common activities. Both have antiviral activity (45, 46) antiproliferative activity, both induce the expression of MHC antigens (47,48), increase ADCC and reduce c-myc expression (49, 50). In most casesit is not known whether these activities are exerted directly by TNF or via the induction of IFN by TNF. Some studies suggestthat TNF can promote the production of IFN-r and IFN0 (5 1) and therefore protection of target cells by TNF-induced IFN could explain our results. However, in the present study the mode of action of TNF and IL- 1 was found to be quite different from that of IFNs. Both TNF and IL- 1 directly protected target cells from the killing effect of NKCF while IFN-a and EN-7 did not. ENS and not IL-l or TNF reduced the ability of target cells to induce NKCF. Finally, TNF and IL-l exerted a protective effect against NK-CMC in adherent cells only, while IFNs induced their protective effect both in adherent and in nonadherent cells. Therefore it is concluded that the protective effect of TNF and IL- 1 is not mediated via IFNs. Recently, it was shown that IL-l can also induce the production of IFN&/IL-6 (52,53). We have examined the possibility that IFN-p,/IL-6 is involved in the protective effect of IL- 1 and TNF against NK-CMC. In fact we observed that IFN-&/IL-6 does not protect target cells either against NK-CMC or against NKCF. Therefore the protective effect of TNF and IL-l is either direct or exerted through a mediator different from EN-&/IL-6. K-562 target cells which were treated with highly purified human IL-1 for 1 hr were found to produce more conjugates with large granular lymphocytes (LGL) and became more sensitive to NK-CMC than untreated cells (54). Differences in the response of target cells to IL- 1 between that study and ours may be attributed not only to time and assayconditions but also to the different IL- 1 preparation (recombinant IL- 1(Yin our case vs highly purified IL- 1 preparation containing IL- 1cuand IL- 1P at ratio of 1 to 9, respectively (Ref. 1) in the previously published case).Normal vascular endothelial cells which were treated with IL-l or TNF were recently found to be unaffected in their susceptibility to the cytotoxic activity of LAK cells (55). Such contrasting results may also be attributed to the use of different target and effector cells. In conclusion, we demonstrate that two major immunomodulators IL- 1 and TNF can protect cells from NK-CMC. Unlike the case of the protective effect of IFNs, these cytokines reduced also the sensitivity of target cells to NKCF. Therefore, it is suggestedthat the protective effect of TNF and IL-1 against NK-CMC is mediated via mechanisms which are different from those of IFNs.
IL- 10~AND TNF-ol PROTECT CELLS FROM NK-CMC AND NKCF
335
ACKNOWLEDGMENT The technical assistanceof Mrs. R. Eisenstadt is greatly acknowledged.
REFERENCES 1. Oppenheim, J. J., Kovacs, E. J., Matsushima, K., and Durum, S. K., Zmmunol. Today 7,45, 1986. 2. Wallach, D., In “Interferon 7” (I. Gresser, Ed.), pp. 89-124. Academic Press,London, 1986. 3. Feinman, R., Henriksen-De Stefano, D., Tsujimoto, M., and Vilcek, J., J. Zmmunol. 138,635, 1987. 4. Fischer, D. G., and Rubinstein, M., Zmmunobiology72, 110, 1986. 5. Onozaki, K., Matsushima, K., Kleinerman, E. S., Saito, T., and Oppenheim, J. J., J. Zmmunol. 135, 314, 1985. 6. Perussia, B., Dayton, E. T., Fanning, V., Thiagarajan, P., Hoxie, J., and Trinchieri, G., J. Exp. Med. 158,2058, 1983.
7. Ostensen, M. E., Thicle, D. L., and Lipsky, P. E., J. Zmmunol. 138,4185, 1987. 8. Trinchieri, G., and Santoli, D., J. Exp. Med. 148, 1314, 1978. 9. Greenberg, A. H., Miller, V., Jablonski, T., and Pohajdak, B., J. Zmmunol. 132,2 129, 1984. 10. Ball, E. D., Nichols, K. E., Pettengill, 0. S., Sorenson, G. D., and Fanger, M. W., Cancer Zmmunol. Immunotherapy 22,211, 1986. 11. Bukowski, J. F., and Welsh, R. M., J. Zmmunol. 135,3537, 1985. 12. Gronberg, A., Kiessling, R., Masucci, G., Guevara, L. A., Eriksson, E., and Klein, G., Znf. J. Cancer 32,609, 1983.
13. Gronberg, A., Kiessling, R., and Fiers, W., Cell. Zmmunol. 95, 195, 1985. 14. Hansson, M., Kiesshng, R., Andersson, B., and Welsh, R. M., J. Zmmunol. 125,2225, 1980. 15. Heiskala, M., Zmmunology60, 167, 1987. 16. Laskay, T., and Kiessling, R., Natl. Immunity Cell Growth Regult. 5,2 I 1, 1986. 17. Moore, M., White, W. J., and Potter, M. R., Znt. J. Cancer 25,565, 1980. 18. Trinchiery, G., Granato, D., and Perussia, B., J. Zmmunol. 126,335, 1981. 19. Wallach, D., J. Inter&on Rex 2,329, 1982. 20. Wallach, D., CeNZmmunol. 75,390, 1983. 21. Welsh, R. M., Karre, K., Hansson, M., Kunkel, L. A., and Kiessling, R. W., J. Zmmunol. 126, 219, 1981. 22. Yeoman, H., and Robins, R. A., Zmmunology63,291,
1988.
23. Bukowski, J. F., and Welsh, R. M., In “lnterferons as Cell Growth Inhibitors and Antitumor Factors.” (R. M. Friedman, T. Merigan, and T. Sreevalsan, Eds.), pp. 175-184. A. R. Liss, New York, 1986. 24. Piontek, G. E., Ljunggren, H. G., Petersson, M., Gronberg, A., Kiessling, R., and Karre, K., In “The Biology of the Interferon System. 1984.” (K. Kirchner and H. Schellekens, Eds.), pp. 247-253. Elsevier, New York, 1985. 25. Piontek, G. E., Taniguchi, K., Ljunggren, H. G., Gronberg, A., Kiessling, R., Klein, G., and Karre, K., J. Zmmunol. 135,4281, 1985. 26. Wright, S. C., and Bonavida, B., J. Zmmunol. 130,2965, 1983. 27. Boyum, A., Stand. J. Clin. Lab. Invest. 2l(Suppl. 97), 77, 1968. 28. Leung, K. H., and Koren, H. S., J. Zmmunol. 129, 1742, 1982. 29. Welsh, R. M., J. Exp. Med. 148, 163, 1978. 30. Finter, N. B., J. Gen. Viral. 5,4 19, 1969. 3 1. Woodruff, M. F. A., Biochim. Biophys. Acta 865,43, 1986. 32. Hanna, N., Biochim. Biophys. Acta 780,2 13, 1985. 33. Nabi, Z. F., Zucker-Franklin, D., Lipkin, G., and Rosenberg, M., Cancer 58, 1461, 1986. 34. Saxena, R. K., Zmmunol. Left. 15, 105, 1987. 35. Saxena, R. K., Saxena, Q. B., and Adler, W. H., J. Zmmunol. 141,1782, 1988. 36. Lanat, N., Aghai, E., Kother, A., Kinarty, A., Sobel, E., Gruener, N., and Froom, P., Blood 71, 114 1, 1988.
37. Holtman, H., and Wallach, D., J. Zmmunol. 139, 1161, 1987. 38. Wallach, D., J. Zmmunol. 132,2464, 1984. 39. Ortaldo, J. R., Ransom, J. R., Sayers,T. J., and Herberman, R. B., J. Zmmunol. 137,2857, 1986. 40. Wright, S. C., and Bonavida, B., J. Zmmunol. 138, 1791, 1987. 41. Koren, H. S., Zmmunol. Todays, 69, 1987.
336
REITER AND RUBINSTEIN
42. Bialas, T., Kolitz, J., Levi, E., Polivka, A., Oez, S., Miller, G., and Welle, K., Cancer Res. 48, 891, 1988. 43. Cuturi, M. C., Murphy, M., Costa-Giomi, M. P., Weinmann, R., Perussia, B., and Trichieri, G., J. Exp. Med. 165,1581, 1987. 44. Roozemond, R. C., Urli, D. C., Jansen, J., and Bonaida, B., J. Immunol. 142, 1209, 1989. 45. Arakawa, T., Hsu, Y. R., Toth, E., and Stebbing, N., J. Interferon Rex 7, 103, 1987. 46. Mestan, J., Digel, W., Mittnacht, S., Millen, H., Blohm, D., Moller, A., Jacobsen, H., and Kirchner, H., Nature (London) 323,816, 1986. 47. Revel, M., In “Antiviral Drugs and Interferon: The Molecular Basis oftheir Activity” (Y. Becker, Ed.), pp. 357-433. Martinus Nijhoff, Notwell, MA, 1984. 48. Pfizenmaier, K., Scheurich, P., Schuter, C., Kronke, M., J. Immunol. 138,975, 1987. 49. Kimchi, A., Yarden, A., and Resnitzky, D., In “Interferons as Cell Growth Inhibitors and Antitumor Factors” (R. M. Friedman, T. Merigan, and T. Sreevalsan, Eds.), pp. 391-402. A. R. Liss, New York, 1986. 50. Yarden, A., and Kimchi, A., Science 234, 1419, 1986, 5 1. Kohasa, M., Henriksen-De Stefano, D., May, L. T., Vilcek, J., and Sehgal, P. B., Cell45,659, 1986. 52. Kohasa, M., May, L. T., Tamm, I., Vilcek, J., and Sehgal, P. B., Molec. Cell Biol. 7,273, 1987. 53. Le, J., and Vilcek, J., Lab. Invest. 56,235, 1987. 54. Herman, J., Dinarello, C. A., Kew, M. C., and Rabson, A. R., J. Immunol. 135,2882, 1985. 55. Rankonen, R., Ristimaki, A., and Hayry, P., Eur. J. Immunol. l&1839,1988.
IMMUNOLOGY
125326-336
(1990)
Interleukin-1 cvand Tumor Necrosis Factor-a Protect Cells against Natural Killer Cell-Mediated Cytotoxicity and Natural Killer Cytotoxic Factor’ ZVI REITER AND MENACHEM
RUBINSTEIN~
Department ofMolecular Genetics and Virology, The Weizmann Institute of Science, Rehovot, 76100, Israel Received July 12, 1989; accepted October 5, I989 The protective effectsof interferons (IFNs) against NK cell-mediated cytotoxicity (NK-CMC) is well established. We report here that both recombinant tumor necrosis factor-a (TNF-a) and recombinant interleukin- I (Y(IL- 1a) can also protect some adherent target cells (e.g., the amniotic cells WISH and the cervical epithelial carcinoma cells HeLa-229) from NK-CMC in a dosedependent manner. Like in the case of IFNs, the level of conjugate formation between target and effector cells (nonadherent peripheral blood lymphocytes) is not affected by pretreatment of the target cells with either TNF-ol or IL-la. However, while the main effect of IFNs is to reduce the ability of target cells to stimulate the release of NK cytotoxic factor (NKCF) from effector cells, TNF-a and IL- 1a do not affect this processbut rather reduce the target cell sensitivity to the lytic effect of NKCF. Therefore TNF-ol and IL 1a induce resistance to NK-CMC by a mechanism that differs from the one attributed to IFNs. The protective effect of TNF-a and ILo 1990 Academic PIES, IX. 1LYis not mediated by the induction of EN-&/IL-6.
INTRODUCTION Interleukin- 1 (IL- I), a macrophage-derived cytokine and tumor necrosis factor(Y(TNF-a), a macrophage-derived cytotoxin, share some regulatory activities with interferons (IFNs). These cytokines directly inhibit the growth of some tumor cells (1, 2), activate or promote the differentiation of cytotoxic T lymphocytes (1, 2) and enhance the cytocidal activities of monocytes (3-6). Furthermore, it was reported recently that both IL- 1 and TNF augment the cytotoxic activity of natural killer (NK) cells in a manner similar to that of IFNs ( 1,7). The sensitivity of various cells to NK cell-mediated cytotoxicity (NK-CMC) is significantly inhibited by pretreatment of the target cells with IFNs (8-25). Although the mechanism of this protective effect is not yet understood, it was found in the murine system that IFN lowered the ability of target cells to induce the releaseof NK cytotoxic factor (NKCF) upon binding to NK cells (26). In this study, we demonstrate that certain cells acquire resistance to NK-CMC and to NKCF by pretreatment with either TNF-(U or IL- 1-a by a mechanism that differs ’ This work was supported by a grant from InterYeda Ltd., Nes-Ziona, Israel. ’ M.R. has the Maurice and Edna Weiss Chair of Interferon Research.
0008-8749/90 $3.00 Copyright All riglm
0 1990 by Academic Press, Inc. of reproduction in any form resewed.
IL-la
AND
TNF-a
PROTECT
CELLS FROM
NK-CMC
AND
NKCF
327
from that of IFNs. This protective effect is not mediated via the induction of IFN-/3,/ IL-6 by these cytokines. MATERIALS AND METHODS Cells NK efictor cells. Mononuclear cells were isolated from freshly collected peripheral blood of healthy donors by density gradient sedimentation on Ficoll-Paque (27). Adherent cells were removed by adsorption to nylon wool (28). The nonadherent cells were used as effector cells. Target and stimulator cells. WISH cells, a human amniotic cell line (ATCC CCL 25) and HeLa-229, a human cervical epitheloid carcinoma cell line (ATCC CLL 2.1) were used as target cells. K-562 cells, a human erythromyeloid leukemia cell line (ATCC CCL 243) and U-937 cells, a human monocytic cell line (ATCC CRL 1593) were used as stimulator cells for induction of the release of NKCF by NK effector cells. All the cells were grown and assayedat 37°C in a humidified 5% CO2 incubator. Reagents and Media Homogeneous Escherichia coli rIFN-LuC (sp activ 10’ units/mg), homogeneous rIFN-y from recombinant Chinese hamster ovary (CHO) cells (sp activ 5 X 10’ units/ mg) and homogenous rIFN-&/IL-6 (CHO) (sp activ 2 X lo6 HGF units/mg) were from Interpharm Laboratories, Ness-Ziona, Israel. E. coli rTNF-a (sp activ 2 X 10’ units/mg) was a gift from Dr. D. Wallach. E. coli rIL- la (sp activ 5 X lo6 units/mg) was a gift from Hoffmann-La Roche Inc. (NJ). Ficoll-Paque was from Pharmacia, Sweden. [5’Cr]-sodium chromate (sp activ 15 to 40 Ci/mmol) was from the Nuclear Research Center, Negev, Israel. RPMI- 1640 medium, minimal essential medium (MEM) and fetal calf serum (FCS) were from GIBCO. Effector cells as well as K-562, U-937, and HeLa-229 cells were cultured in RPMI1640 medium supplemented with FCS (lo%), glutamine (2 mM), penicillin (100 units/ml) and streptomycin (100 pg/ml), (RPMI-10%). WISH cells were cultured in MEM supplemented similarly (MEM- 10%). Treatment qfNK Efector and Target Cells with Cytokines Nonadherent peripheral blood mononuclear cells (5 X 106/ml) were cultured in medium (RPMI- 10%) containing IFN-(YC ( 1000 units/ml, 14- 18 hr, 37°C) prior to use as effector cells. WISH and HeLa-229 cells were seeded in 96-well flat-bottom plates (4 X lo4 cells/well, 100 ~1 medium) and incubated for 24 hr at 37°C. Various cytokines were then added (50 ~1) in different concentrations and incubation was continued for 12- 14 hr. “Cr ReleaseAssay of NK-CMC This assaywas performed with some modification as described (29). Briefly, HeLa229 and WISH cells (4 X lo4 cells/well, 100 ~1) were seededin 96-well flat-bottom plates and incubated for 24 hr at 37°C. [5’Cr]-sodium chromate (1 &i/well) in medium alone (RPMI- 10%or MEM- lo%, 100 ~1,respectively) or in medium containing the indicated cytokines, was then added and incubation continued for 12- 14 hr. The
328
REITER
AND
RUBINSTEIN
plates were then rinsed three times, various amounts of IFN-activated effector cells in RPM1 10%(200 ~1)were added, incubation was continued for 5-6 hr at 37”C, and aliquots of the supernatants were then collected for counting released radioactivity which was defined as experimental cpm. Total cpm was obtained by adding 1% Triton X- 100 to wells with target cells. Spontaneous release of “Cr (7- 12% and 8- 14% in HeLa-229 and WISH cells, respectively) was determined in wells containing target cells only. Data are expressed as percentage of “Cr-specific release according to the formula: % specific release = 100 X
experimental cpm - spontaneous cpm total cpm - spontaneous cpm
Each data point represents an average of 5-8 experiments, each done in triplicates with effector cells from a different donor. Conjugate Formation Assay Target cells (HeLa-229 and WISH) were seeded in 96-well flat-bottom plates ( 1OOO/wellin 50 ~1 RPMI- 10% or MEM- lo%, respectively) and incubated for 24 hr at 37°C. Effector cells (2 X 103/well in 50 ~1RPMI- 10% or MEM- 10%) were added, spun (15g, 5 min), and incubated for various periods. Following incubation, trypan blue (l%, 50 ~1) was added followed by glutaraldehyde (3%, 50 ~1). The number of conjugates was determined by counting under a light microscope. Each sample was assayedin duplicates. Data are expressed as percentage of binding according to the formula: % binding =
conjugated effector cells x 100 total effector cells
Preparation of Crude NKCF IFN-activated effector cells (5 X 106/ml) were cocultured with stimulator cells (5 104/ml) in RPMI-1640 medium supplemented with 1% FCS for 24 hr at 37°C in a humidified atmosphere of 5% COZ. Preliminary experiments indicated that these conditions gave maximal NKCF activity. Following incubation, the supernatant was harvested and concentrated five- to eightfold by ultrafiltration on an Amicon YM-5 ultrafiltration membrane. The concentrated supernatants were passed through a 0.45-pm filter, and either used immediately or kept frozen at -70°C until used. Concentrated supernatants from cultures of either stimulator cells or effector cells were used as control. X
Assay of NKCF Two different assayswere used for the measurement of NKCF activity: (a) 5’Cr release assay of NKCF. Either HeLa-229 or WISH cells (4 X lo4 cells/ well, 100 ~1 RPMI-1% or MEM- I%, respectively) were seededin 96-well plates and incubated in the presence of “0-sodium chromate (1 &i/well) for 24 hr at 37°C. Various concentrations of the different cytokines in 100 ~1medium were added and incubation continued for 12- 14 hr. The plates were then rinsed three times, various dilutions of crude NKCF (in 100 ~1medium) were added and incubation continued
IL-la
AND
TNF-a
PROTECT
CELLS FROM
NK-CMC
AND
NKCF
329
for an additional period of 20-24 hr at 37°C. Aliquots of the supernatants were then collected for radioactivity measurement. Total cpm, spontaneous cpm and “0-specific releasewere calculated as described in “Cr releaseassayof NK-CMC. (b) Neutral Red uptake assay.Either HeLa-229 or WISH cells (4 X lo4 cells/well, 100 ~1)were seededin 96-well plates and incubated for 24 hr at 37°C. Various concentrations of the different cytokines in 100 ~1medium were added and incubation continued for 12- 14 hr. The plates were then washed three times, various dilutions of crude NKCF (in 100 ~1RPMI- 1% or MEM- l%, respectively) were added, and incubation continued for an additional period of 44 hr. Viability of the cells was determined by incubation with neutral red (100 &well) for 2 hr, extracting the neutral red that was taken up by the cells with Sorenson’s citrate buffer-ethanol mixture ( 150 &well), and quantitating calorimetrically at 570 nm by a micro ELISA autoreader (30). The extent of lysis by NKCF was calculated according to the formula: % Viability =
Dye uptake by cells treated with NKCF x 1o. Dye uptake by untreated cells
Each data point represents an average of a triplicate. Statistics “Cr release NKCF, and conjugate formation assayswere expressedas the arithmetic mean f standard error of mean. Comparisons were carried out by student’s t test. RESULTS IL-la and TNF-a Protect Target Cells from NK-CMC Preincubation of HeLa-229 and WISH cells with either IL- 1CYor TNF-a rendered them resistant to NK-CMC. As a positive control, the protective effect of IFN-(UC and IIN-7 was demonstrated under the same conditions. The effect of all tested cytokines was significant (P < 0.05) and dose dependent. Moreover this protective effect was observed in all tested effector to target ratios (Figs. 1, 2). When these cells were pretreated with IL-6 no such protective effect was observed (data not shown). In contrast to the above results with adherent target cells, very little or no protective effect was observed with K-562 or U-937 target cells when pretreated with either IL-la or TNF-(u, whereas significant protection was obtained when these cells were pretreated with either IFN-arc or IFN--/ (data not shown). Lack of Eflect of IL-l (Yand TNF--a or Conjugate Formation betweenNK Efector and Target Cells Either HeLa-229 or WISH target cells were preincubated with various concentrations of cytokines (II+&&, IFN--y 10-1000 units/ml; IL-6 0.01-10 PM; IL- 1CY5500 units/ml, TNF-a lo-200 units/ml) for 12 hr. NK effector cells were then added and incubation continued for 30-240 min. No differences in the level of conjugate formation were observed with the various cytokines as compared with control cells. A typical result following 1 hr incubation with effector cells is shown in Table 1.
330
REITER AND RUBINSTEIN
IL-la
40 -
I\
20 30.k z
IFN- d:
1
0
20
40
60
80
100
0
I
20
I
L
40
I
I
60
I
I
80
I
I
100
CYTOKINES CONCENTRATION (units/ml)
FIG. 1. The protective effect of IL-la and TNF-a against NK-CMC on HeLa-229 cells. Wr-labeled HeLa-229 cells were pretreated with various concentrations of IL- 101or TNF-a for 12- 14 hr. The protective effect of IFN-aC and IFN-7 is given as a positive control. The cells were washed and incubated with NK effector cells at different E:T rations (0, 40: 1; W,20: 1; A, 10:1). The curves represent an arithmetic mean f standard error of experiments with NK effector cells from four different healthy donors.
Lack of Eflect of IL-l (Yand TNF-LXon the Ability of Target Cells to Stimulate NKCF Generation HeLa-229 and WISH cells were pretreated with IL- 1(Y(500 units/ml); TNF-a (200 units/ml); IFN-(YC ( 1000 units/ml); IFN-7 ( 1000 units/ml) or IL-6 ( 1 PM) for 12 hr, washed, and NK effector cells (E:T-50: 1) were added. After 24 hr of coculturing, aliquots of the supernatants were collected, concentrated, and assayed by the 5’Cr release assay for NKCF (Fig. 3) and by the neutral red uptake assay(Fig. 4). In both assaysit was shown that while IFN-aC and IFN-y reduced significantly (P < 0.05) the level of NKCF generation, the other cytokines (IL- 1cy,TNF+, and IL-6) did not affect the ability of HeLa-229 and WISH cells to induce NKCF generation. IL-l (Yand TNF-a Protect Target Cells from NKCF Preincubation of either HeLa-229 or WISH cells with either IL- la (500 units/ml) or TNF-(Y (200 units/ml) reduced significantly (P < 0.05) the sensitivity of the cells to the cytotoxic effect of NKCF, as determined both by “Cr releaseassay(Fig. 5) and by the neutral red uptake assay (Fig. 6). A significant (P < 0.05) protective effect against NKCF was observed with either IL- 1(Yor TNF-a! only after 6-8 hr of incubation with these cytokines and was maximal after lo- 12 hr. In contrast, pretreatment
IL-la AND TNF-(U PROTECT CELLS FROM NK-CMC AND NKCF
331
IFNq
20
ih,,
:$g
10
’
I
)
0
I
1
20
.
I
I
40
I
60
I
B
I
60
CYTOKINES
I
I
100
I
,
0
CONCENTRATION
I I I I 1 1 I I 20 40 60 60
I 100
(units/ml)
FIG. 2. The protective effect of IL-la and TNF-LUagainst NK-CMC on WISH cells. For details seelegend to Fig. 1.
of these cells as well as nonadherent cells (K-562, U-937) with either IFN-(YC, IFN--/ or IL-6 did not exert protection against NKCF activity (data not shown). DISCUSSION NK cells may have a roIe in immune surveillance against the development of at least certain types of spontaneous tumors, as well as against the metastatic spread of TABLE 1 Lack of Effect of IL-n and TNF-ol on Conjugate Formation Between NK Effector and Target Cells Conjugate formation (%)” HeLa-229 No treatment IL-la (500 units/ml) TNF-a (200 units/ml) IFN-olC (1000 units/ml) IFN--/ (1000 units/ml) IFN-&/IL-6 (1 PM)
6.8 + 1.3 6.2 f 1.6 7.6 -t 0.9 7.1 ?z 1.2 6.4 -c 1.3 6.9 + 1.1
WISH 7.1 * 8.0 * 7.6 i 7.9 f 6.8 k 6.7 f
1.8 1.6 1.2 1.1 1.4 1.3
’ The percentage conjugate formation represents an arithmetic mean f standard error of experiments with NK effector cells from three different donors.
REITER AND RUBINSTEIN
1:2
1 :4
1:2 1:8 CRUDE NKCF (DILUTIONS)
114
1:8
FIG. 3. Lack of effect of IL-la, TNF-LU,and IM-&/IL-6 on the ability of target cells to stimulate NKCF generation (“0 release assay). Either HeLa-229 or WISH cells were pretreated for 12 hr with various cytokines. Effector cells were added, the supematants were collected after 24 hr and assayedby the “Cr release test on the same target cells (HeLa-229-A or WISH-B, respectively). Cytokines were: IL-la (500 units/ml, n ); TNF-ol(200 units/ml, A); IFN-&/IL-6 (1 PM, +); Medium alone, (0). IEN-& (1000 units/ ml, 0) and IFN--y (1000 units/ml, A) were used as positive controls. The curves represent an arithmetic mean (standard error was f 1.O-2.6%) of experiments with NKCF from three different healthy donors.
tumors (3 1, 32). Interferons reduced the susceptibility of target cells to NK-CMC (S25). Other cytokines were tested for this activity as well. Crude preparations of the contact inhibitory factor (CIF) devoid of IFN activity reduced the susceptibility of some tumor cells to NK-CMC (33). Crude IL-2 preparations were reported to induce a significant decreasein the susceptibility of YAC cells to murine NK-CMC (34,35). Since highly purified IL-2 preparations did not exhibit a similar effect, it was con-
t
B.
T
50 -L 1 :2
114
1:2 1:8 CRUDE NKCF (DILUTIONS)
1:4
FIG. 4. Lack of effect of IL- 1(Y,TNF-(Y and IFN-&/IL-6 on the ability of target cells to stinnulate NKCF generation (neutral red assay). NKCF was induced by cytokine-treated target cells (see Fig. 3) and was measured in mixed-culture supematants by the neutral red uptake assay on HeLa-229 (A) and WISH (B) cells. Cytokines were: IL- 101(500units/ml, q ); TNF-ol(200 units/ml, m); and IFN-/&/IL-6 ( 1 PM, o~C(1000 units/ml, @I)and IFN--/ (1000 units/ml, n ) were both used as positive controls. Medium alone, (0). The histograms represent an arithmetic mean f standard error of experiments with NKCF from three different healthy donors.
IL-la
AND
TNF-o( PROTECT
CELLS FROM
NK-CMC
1:2 114 1:8 CRUDE NKCF (DILUTIONS)
1:2
AND
1A
333
NKCF
1:8
FIG. 5. IL- 101and TNF-cu protect target cells from killing by NKCF (5’Cr release assay). Either HeLa229 (A) or WISH (B) cells were preincubated for 12 hr with IL-la (500 units/ml, n ); TNF-cu (200 units/ ml, A); IFN-(uC (1000 units/ml, 0); IFN-y (1000 units/ml, a); IFN-&/IL-6 (I p&f, +), or medium alone (0). The cultures were washed, various dilutions of crude NKCF preparations were added, and percentage ofcytotoxicity was determined. Each curve represents an arithmetic mean (standard error was + 0.7-2.1%) of two experiments using separate NKCF preparations from four healthy donors.
eluded that other cytokines present in the crude IL-2 preparations were probably responsible for the induced resistance to lysis. There is however another study where IL-2 was found to reduce the sensitivity of primary hairy leukemia cells to activated NK cells (36). In the present study we identified both IL-l (Yand TNF-a as the agents that reduce the susceptibility of target cells to NK-CMC at very low concentrations.
1:2
1 14
I :8 1:2 CRUDE NKCF (DILUTIONS)
114
1:8
FIG. 6. IL-la and TNF-(U protect target cells from killing by NKCF(neutral red assay). Either HeLa-229 (A) or WISH (B) cells were preincubated for 12 hr with IL-la (500 units/ml, EI); TNF-(U (200 units/ml, Q; IFN-& (1000 units/ml, q ); IIN--, (1000 units/ml, n ); IFN-&/IL-6 (1 PM, ) or medium alone (Cl). The cultures were washed, various dilutions of crude NKCF preparations were added, and percentage of cytotoxicity wasdetermined. The histograms represent an arithmetic mean f standard error oftwo experiments using separate preparations from four healthy donors.
334
REITER
AND
RUBINSTEIN
This effect was located at the level of target cell lysis by NKCF and not at the initial step of conjugate formation between NK effector and target cell. Since TNF and IL1 can protect cells from lysis by TNF (37, 38), their ability to confer resistance to NKCF is not unexpected. Nevertheless, NKCF differs considerably from TNF in various aspects.For example we found that unlike TNF, crude NKCF kills cells even in the absence of cycloheximide. Moreover, we found also that crude NKCF does not confer protection against TNF or NKCF. Immunological and other differences between NKCF and TNF were reported in the literature. NKCF differed from TNF and lymphotoxin with respect to target cell specificity and was not neutralized by antibodies to the other cytotoxins (39-42). Although TNF was produced by NK cells (43) it was not involved in NK-CMC (42). Finally liposomes made from acidic phospholipids were lysed by NKCF and not by TNF (44). TNF and IFNs have some common activities. Both have antiviral activity (45, 46) antiproliferative activity, both induce the expression of MHC antigens (47,48), increase ADCC and reduce c-myc expression (49, 50). In most casesit is not known whether these activities are exerted directly by TNF or via the induction of IFN by TNF. Some studies suggestthat TNF can promote the production of IFN-r and IFN0 (5 1) and therefore protection of target cells by TNF-induced IFN could explain our results. However, in the present study the mode of action of TNF and IL- 1 was found to be quite different from that of IFNs. Both TNF and IL- 1 directly protected target cells from the killing effect of NKCF while IFN-a and EN-7 did not. ENS and not IL-l or TNF reduced the ability of target cells to induce NKCF. Finally, TNF and IL-l exerted a protective effect against NK-CMC in adherent cells only, while IFNs induced their protective effect both in adherent and in nonadherent cells. Therefore it is concluded that the protective effect of TNF and IL- 1 is not mediated via IFNs. Recently, it was shown that IL-l can also induce the production of IFN&/IL-6 (52,53). We have examined the possibility that IFN-p,/IL-6 is involved in the protective effect of IL- 1 and TNF against NK-CMC. In fact we observed that IFN-&/IL-6 does not protect target cells either against NK-CMC or against NKCF. Therefore the protective effect of TNF and IL-l is either direct or exerted through a mediator different from EN-&/IL-6. K-562 target cells which were treated with highly purified human IL-1 for 1 hr were found to produce more conjugates with large granular lymphocytes (LGL) and became more sensitive to NK-CMC than untreated cells (54). Differences in the response of target cells to IL- 1 between that study and ours may be attributed not only to time and assayconditions but also to the different IL- 1 preparation (recombinant IL- 1(Yin our case vs highly purified IL- 1 preparation containing IL- 1cuand IL- 1P at ratio of 1 to 9, respectively (Ref. 1) in the previously published case).Normal vascular endothelial cells which were treated with IL-l or TNF were recently found to be unaffected in their susceptibility to the cytotoxic activity of LAK cells (55). Such contrasting results may also be attributed to the use of different target and effector cells. In conclusion, we demonstrate that two major immunomodulators IL- 1 and TNF can protect cells from NK-CMC. Unlike the case of the protective effect of IFNs, these cytokines reduced also the sensitivity of target cells to NKCF. Therefore, it is suggestedthat the protective effect of TNF and IL-1 against NK-CMC is mediated via mechanisms which are different from those of IFNs.
IL- 10~AND TNF-ol PROTECT CELLS FROM NK-CMC AND NKCF
335
ACKNOWLEDGMENT The technical assistanceof Mrs. R. Eisenstadt is greatly acknowledged.
REFERENCES 1. Oppenheim, J. J., Kovacs, E. J., Matsushima, K., and Durum, S. K., Zmmunol. Today 7,45, 1986. 2. Wallach, D., In “Interferon 7” (I. Gresser, Ed.), pp. 89-124. Academic Press,London, 1986. 3. Feinman, R., Henriksen-De Stefano, D., Tsujimoto, M., and Vilcek, J., J. Zmmunol. 138,635, 1987. 4. Fischer, D. G., and Rubinstein, M., Zmmunobiology72, 110, 1986. 5. Onozaki, K., Matsushima, K., Kleinerman, E. S., Saito, T., and Oppenheim, J. J., J. Zmmunol. 135, 314, 1985. 6. Perussia, B., Dayton, E. T., Fanning, V., Thiagarajan, P., Hoxie, J., and Trinchieri, G., J. Exp. Med. 158,2058, 1983.
7. Ostensen, M. E., Thicle, D. L., and Lipsky, P. E., J. Zmmunol. 138,4185, 1987. 8. Trinchieri, G., and Santoli, D., J. Exp. Med. 148, 1314, 1978. 9. Greenberg, A. H., Miller, V., Jablonski, T., and Pohajdak, B., J. Zmmunol. 132,2 129, 1984. 10. Ball, E. D., Nichols, K. E., Pettengill, 0. S., Sorenson, G. D., and Fanger, M. W., Cancer Zmmunol. Immunotherapy 22,211, 1986. 11. Bukowski, J. F., and Welsh, R. M., J. Zmmunol. 135,3537, 1985. 12. Gronberg, A., Kiessling, R., Masucci, G., Guevara, L. A., Eriksson, E., and Klein, G., Znf. J. Cancer 32,609, 1983.
13. Gronberg, A., Kiessling, R., and Fiers, W., Cell. Zmmunol. 95, 195, 1985. 14. Hansson, M., Kiesshng, R., Andersson, B., and Welsh, R. M., J. Zmmunol. 125,2225, 1980. 15. Heiskala, M., Zmmunology60, 167, 1987. 16. Laskay, T., and Kiessling, R., Natl. Immunity Cell Growth Regult. 5,2 I 1, 1986. 17. Moore, M., White, W. J., and Potter, M. R., Znt. J. Cancer 25,565, 1980. 18. Trinchiery, G., Granato, D., and Perussia, B., J. Zmmunol. 126,335, 1981. 19. Wallach, D., J. Inter&on Rex 2,329, 1982. 20. Wallach, D., CeNZmmunol. 75,390, 1983. 21. Welsh, R. M., Karre, K., Hansson, M., Kunkel, L. A., and Kiessling, R. W., J. Zmmunol. 126, 219, 1981. 22. Yeoman, H., and Robins, R. A., Zmmunology63,291,
1988.
23. Bukowski, J. F., and Welsh, R. M., In “lnterferons as Cell Growth Inhibitors and Antitumor Factors.” (R. M. Friedman, T. Merigan, and T. Sreevalsan, Eds.), pp. 175-184. A. R. Liss, New York, 1986. 24. Piontek, G. E., Ljunggren, H. G., Petersson, M., Gronberg, A., Kiessling, R., and Karre, K., In “The Biology of the Interferon System. 1984.” (K. Kirchner and H. Schellekens, Eds.), pp. 247-253. Elsevier, New York, 1985. 25. Piontek, G. E., Taniguchi, K., Ljunggren, H. G., Gronberg, A., Kiessling, R., Klein, G., and Karre, K., J. Zmmunol. 135,4281, 1985. 26. Wright, S. C., and Bonavida, B., J. Zmmunol. 130,2965, 1983. 27. Boyum, A., Stand. J. Clin. Lab. Invest. 2l(Suppl. 97), 77, 1968. 28. Leung, K. H., and Koren, H. S., J. Zmmunol. 129, 1742, 1982. 29. Welsh, R. M., J. Exp. Med. 148, 163, 1978. 30. Finter, N. B., J. Gen. Viral. 5,4 19, 1969. 3 1. Woodruff, M. F. A., Biochim. Biophys. Acta 865,43, 1986. 32. Hanna, N., Biochim. Biophys. Acta 780,2 13, 1985. 33. Nabi, Z. F., Zucker-Franklin, D., Lipkin, G., and Rosenberg, M., Cancer 58, 1461, 1986. 34. Saxena, R. K., Zmmunol. Left. 15, 105, 1987. 35. Saxena, R. K., Saxena, Q. B., and Adler, W. H., J. Zmmunol. 141,1782, 1988. 36. Lanat, N., Aghai, E., Kother, A., Kinarty, A., Sobel, E., Gruener, N., and Froom, P., Blood 71, 114 1, 1988.
37. Holtman, H., and Wallach, D., J. Zmmunol. 139, 1161, 1987. 38. Wallach, D., J. Zmmunol. 132,2464, 1984. 39. Ortaldo, J. R., Ransom, J. R., Sayers,T. J., and Herberman, R. B., J. Zmmunol. 137,2857, 1986. 40. Wright, S. C., and Bonavida, B., J. Zmmunol. 138, 1791, 1987. 41. Koren, H. S., Zmmunol. Todays, 69, 1987.
336
REITER AND RUBINSTEIN
42. Bialas, T., Kolitz, J., Levi, E., Polivka, A., Oez, S., Miller, G., and Welle, K., Cancer Res. 48, 891, 1988. 43. Cuturi, M. C., Murphy, M., Costa-Giomi, M. P., Weinmann, R., Perussia, B., and Trichieri, G., J. Exp. Med. 165,1581, 1987. 44. Roozemond, R. C., Urli, D. C., Jansen, J., and Bonaida, B., J. Immunol. 142, 1209, 1989. 45. Arakawa, T., Hsu, Y. R., Toth, E., and Stebbing, N., J. Interferon Rex 7, 103, 1987. 46. Mestan, J., Digel, W., Mittnacht, S., Millen, H., Blohm, D., Moller, A., Jacobsen, H., and Kirchner, H., Nature (London) 323,816, 1986. 47. Revel, M., In “Antiviral Drugs and Interferon: The Molecular Basis oftheir Activity” (Y. Becker, Ed.), pp. 357-433. Martinus Nijhoff, Notwell, MA, 1984. 48. Pfizenmaier, K., Scheurich, P., Schuter, C., Kronke, M., J. Immunol. 138,975, 1987. 49. Kimchi, A., Yarden, A., and Resnitzky, D., In “Interferons as Cell Growth Inhibitors and Antitumor Factors” (R. M. Friedman, T. Merigan, and T. Sreevalsan, Eds.), pp. 391-402. A. R. Liss, New York, 1986. 50. Yarden, A., and Kimchi, A., Science 234, 1419, 1986, 5 1. Kohasa, M., Henriksen-De Stefano, D., May, L. T., Vilcek, J., and Sehgal, P. B., Cell45,659, 1986. 52. Kohasa, M., May, L. T., Tamm, I., Vilcek, J., and Sehgal, P. B., Molec. Cell Biol. 7,273, 1987. 53. Le, J., and Vilcek, J., Lab. Invest. 56,235, 1987. 54. Herman, J., Dinarello, C. A., Kew, M. C., and Rabson, A. R., J. Immunol. 135,2882, 1985. 55. Rankonen, R., Ristimaki, A., and Hayry, P., Eur. J. Immunol. l&1839,1988.
