CELLULAR IMMUNOLOGY 75, 390-395 (1983)

Interferon-Induced Resistance to the Killing by NK Cells: A Preferential Effect of IFN-7

DAVID WALLACH

Department of Virology, The Weizmann Institute of Science, Rehovot, Israel

Received August 25, 1982; accepted November 18, 1982

HeLa cells show a decrease of susceptibility to the killing by natural killer (NK) cells when treated with IFN-CX, 0, or y. The concentrations at which preparations of EN-(Y or j3 induce the resistance to killing are those which also induce resistance of HeLa cells to infection by vesicular stomatitis virus (VSV). Stimulation of the killing activity of NK cells is also induced at that same range of concentrations of IFN-(Y and 8. In contrast with preparations of IFN-7, induction of the resistance to lolling occurs at IFN concentrations which have only marginal stimulatory effect on the activity of NK cells and have no antiviral effect against VSV. IFN--/, produced with cloned 1FN-y cDNA, is as effective as lymphocyte-produced IFN in inducing the resistance to natural killing. The potent effect of IFN-y on the target cells is, therefore, not due to the function of lymphokines which might contaminate lymphocyte-produced preparations of IFN- y, but a genuine property of the IFN itself.

INTRODUCTION

Natural killer (NK) cells are believed to be involved in the specific elimination of virus-infected cells or of cells showing uncontrolled growth (l-3). Yet, the nature of selectivity in the action of these killer cells is far from being clearly understood. In fact, under in vitro conditions, NK cells appear to act quite nonspecifically, not only against transformed and virus-infected cells, but against a variety of normal cells as well (1). Interferon (IFN) increases natural killing activity and this has been taken as a further indication of the defensive role of NK cells, particularly in viral diseases (3). However, this IFN effect does not impose higher selectivity on the function of the NK cells. On the contrary, activity of NK cells against poorly susceptible targets is increased even more strongly by IFN than the activity against the more susceptible ones (4). If the activity of IFN-stimulated NK cells would have indeed been exerted nonspecifically under the in vivo situation as well, then one would have expected that formation of IFN, or its injection into the body and the consequent activation of NK cells, would have led to extensive damage to cells like thymocytes or fibroblasts which are, in vitro, quite effectively killed by NK cells (5-7). However, damage to normal tissues is rarely induced by IFN (8), and, therefore, one has to assume the existence of some protective mechanisms which limit the in vivo activity of NK cells.

Several agents have indeed been found to inhibit natural killing; these include certain cells which seem to be able to suppress the natural killing (5, 9, lo), as well as some humoral factors. Prostaglandins, for example, can inhibit the activity of the NK cells (11) and so also would other agents which can increase the production of

390

0008-8749/83 $3.00 Copyright 0 1983 by Academic Press. Inc. All rights of reproduction in any form reserved.

SHORT COMMUNICATIONS 391

cyclic AMP in these cells (12). In addition, it appears that IFN itself plays a dual role in regulating natural killing. Besides its stimulatory effect on the NK cells, it can also decrease the killing of certain types of cells by making them less susceptible to the NK cells’ effect (3, 7, 13). One of the factors which has to be taken into account in evaluating the response of natural killing to IFN formation is the ability of cells to produce various types of IFNs which differ in structure and possibly also in function. IFN--/, which is formed by leukocytes in response to stimulation by agents such as lectins and antigens, has a very different structure from that of the “virus-induced” types of interferon-a and p (the type I IFNs) and there are several indications that it affects cell functions differently (for a review, see (14)). We have recently shown that preparations of IF%-, induce resistance to natural killing (“anti-NK” effect) in SV-80 human fibroblastic cells at much lower concentrations than those inducing an antiviral state in these cells, while IFN-(Y or /3 induced the anti-NK and antiviral effects at a similar range of concentrations. The maximal protection obtained with IFN--r was also significantly higher than that induced by the type I IFNs (15, 16). We present here further evidence for the preferential induction of anti-NK effect by IFN-7. We show that the resistance of HeLa cells to killing is induced by preparations of IFN-7 below the concentrations effective in stimulating the NK cells and show evidence that the higher effectiveness of IFN-y preparations in inducing the resistance to killing does not reflect the function of lymphokines which might contaminate these IFN preparations.

MATERIALS AND METHODS

Interferons. IFN-p was induced in cultured human foreskin fibroblasts by combined treatment with poly(rI):(rC), cycloheximide, and actinomycin D (17), and purified to about 10’ U/mg protein on a carboxymethyl Sephadex column. A preparation of IFN-(r (subtype &) was a generous gift of Dr. Menachem Rubinstein (18). IFN-y was induced in the mononuclear fraction of the peripheral blood leukocytes (isolated on Ficoll-Hypaque) by incubating the cells for 2 days with concanavalin A (30 pg/ ml) and phorbol myristate acetate (5 rig/ml). It was partially purified to about lo6 U/mg protein by adsorption and elution from controlled pore glass (19). A sample of IFN-y, secreted by COS-7 monkey cells which had been transfected with the recombinant IFN--y plasmid pSV-y-69 (20) was a generous gift from Drs. P. Gray and D. V. Goeddel (Genentech Inc.).

Cells. HeLa cells were grown in Dulbecco’s modified Eagle’s medium in the pres- ence of 10% fetal calf serum. Human peripheral blood leukocytes were fractionated on Ficoll-Hypaque. The mononuclear cell fraction was incubated with IFN in tissue culture dishes with RPM1 1640 medium containing 20% heat-inactivated fetal calf serum at a cell concentration of 3 - lo6 cells/ml.

Measurements of natural killer cell activity. HeLa cells were seeded in microwells at a concentration of 10000 cells/well. They were incubated for 18 hr in the presence of “Cr (2 &i/ml) and IFN, at various concentrations, and then rinsed four times before incubation with the NK cells. Incubation of the leukocytes with IFN was also for 18 hr. They were then rinsed twice and applied on the target cells in a volume of 250 ~1. The plates were then spun for 5 min at 15g. Samples of the medium were collected from the microtiter wells 3 hr later and the extent of killing was calculated as described by others (13).

392 SHORT COMMUNICATIONS

Quantitation of the antiviral activity of the IFNs. HeLa cells were treated for 18 hr with IFNs, at various concentrations, then infected with VSV (Indiana strain, 5 PFU/cell), and incubated further for 20 hr. The suppression of VSV replication was determined by immunoassay of the virus proteins as described elsewhere (( 17), Wal- lath et al., submitted for publication).

RESULTS

To compare the effectiveness of the various IFNs in stimulating NK activity and in protecting target cells from the killing, we have treated, separately, samples of peripheral blood mononuclear cells and of cultured HeLa cells with IFN, at various concentrations, and then measured the extent of killing. The effect of each type of IFN on the resistance of the HeLa cells to the killing is compared in Fig. 1 (open circles) to its effect on the activity of the NK cells (solid circles).

The patterns of response to IFN-CY and /3 were similar. Protection of the fibroblasts against killing could be induced only at IFN concentrations which were strongly effective in stimulating the activity of the NK cells. In contrast, IFN-y induced the resistance of the target cells to killing, at concentrations which had only a marginal stimulatory effect on the activity of the NK cells.

0

1000

600

600

400

200

0 01 IO loo 1000

IFN hlatwe concentmtiwd

FG 1. Titration of IFN effects on NK activity and on the resistance of target cells to the killing. Samples of peripheral blood mononuclear cells and of HeL.a cells were pretreated with IFN and then the killing activity, at a leukocyte to target ratio of 2&l, was measured as described under Materials and Methods. The stimulatory effect of EN on natural killing (0) was measured by incubating HeLa cells which have not been treated with IFN, with leukocytes pretreated with IFNs at various concentrations. The protective effect against killing (0) was quantitated by incubating HeLa cells pretreated with IFN at various concen- trations, with leukocytes treated with IFN-8 (500 U/ml). The protective and stimulatory effects of the IFNs on the killing are expressed as the percentage of killing in their absence (2.5% of the target cells when neither the HeLa cells nor the leukocytes were treated by IFN and 15% of the target cells when the leukocytes were pretreated with 500 U/ml IFN-8).

SHORT COMMUNICATIONS 393

To relate the antiviral capacity of the IFNs to their effectiveness in inducing the anti-NK effect, we have titrated the two effects simultaneously in separate cultures of the HeLa cells. We have used for that purpose an immunoassay of the proteins of VSV which allows an accurate and detailed titration of the inhibitory effect of IFN on the replication of this virus (( 17), Wallach et al., submitted for publication). IFN- y induced resistance to natural killing at concentrations significantly lower than those inducing the antiviral state, while with EN-/3 (Fig. 2) and IFN-(Y (not shown), the induction of resistance to killing occurred at IFN concentrations which also effectively reduced virus growth.

Since preparations of EN-7 contain various lymphokines and these have multiple effects on cell functions, it could be suspected that the strong protective effect of these preparations against natural killing, which is shown even at concentrations at which there is no antiviral effect, reflects the function of a protein which is not IFN. This appears unlikely since we have found that the protective activity against natural killer copurifies with the antiviral activity of EN-~ preparations (15, 16). To further rule out the possibility that contaminating lymphokines contribute to the effect of IFN- y on the resistance to natural killing, we have tested the effects of a preparation of EN-y, free of other lymphokines, produced with the cloned cDNA of human IFN- y (20). As shown in Fig. 2, the effectiveness of this EN-7 preparation in inducing the resistance to natural killing was comparable to that of the partially purified IFN- y. At IFN concentrations inducing reduction of VSV protein synthesis by 50% (de- fined as 1 antiviral unit/ml), killing of the HeLa cells by NK cells was reduced in the experiment described in Fig. 2 by 13% with II?@, by 42% with partially purified II+&-/, and by 39% with the “cloned” IFN-y.

DISCUSSION

None of the three types of IFNs are devoid of the ability to stimulate the activity of NK cells against HeLa cells or to induce resistance of the HeLa cells to killing. There are, however, clear quantitative differences between the efficiencies by which

FIG. 2. Quantitative relation between the ability of the IFNs to induce resistance of the HeLa cells to natural killing and their antiviral effect in those same cells. Samples of II+3 (A), partially purified IFN- y (B), and IFN--, produced by cells transfected with cloned IFN-7 cDNA (20) (C), were serially diluted and samples of each dilution were applied in parallel on two cultures of HeLa cells seeded in microwells. On one, the induction of the antiviral state was measured by quantitating the yield of VW proteins (0) and on the other, the resistance to natural killing was quantitated (0). Natural killing was measured by incubating the target cells with leukocytes pretreated with 500 U/ml IIN+ at a leukocytes to target ratio of 40:l. Of the HeLa cells, 27% were killed by the NK cells in target cell cultures which had not been pretreated by IFN and this extent of killing is presented as 100% of activity. Antiviral activity of the IFNs is presented as the viral protein yield comparatively to that obtained in cultures which have not been pretreated with IFN.

394 SHORT COMMUNICATIONS

these two effects are induced with the various IFN preparations. IFIQY and p induce

the resistance to killing only at IFN concentrations which are also highly effective in stimulating the NK cells, while preparations of IFN-y induce significxnt resistance to the killing at concentrations in which they show only a slight stimulatory effect on the function of the NK cells. Reduction of cell killing, by the protective effect of IFN might, therefore, be induced more effectively with preparations of IFN-7 than with the type I IFNs.

Since partially purified preparations of EN-y contain various lymphokines, the difference between effects of such preparations and of other IFNs might not necessarily be due to differing modes of action of the IFNs themselves, but might rather reflect effects of lymphokines on natural killing. An effect of lymphokines on NK cell activity has indeed been demonstrated in a study showing that, in addition to IFN, interleukin 2 can augment natural killing (21). It is possible that there are also lymphokines which can have an inhibitory effect on the activity of NK cells and that these lym- phokines contribute to the lower ratio between NK stimulation and anti-NK effect, observed with preparations of IFN-7.

In this study, we show, however, that at least part of the difference between the effect of IFNq and of the type I IFN on natural killing is not due to effects of lymphokines, but rather reflects a genuine difference in the mode of action of the IFNs themselves. We show that an IFN preparation free of lymphokines, produced with cloned IFN-y cDNA, is as effective in inducing the anti-NK effect as partially purified IFN-7 preparations and is significantly more effective, in doing so, than IFN-p.

To study the nature of the effective resistance to killing exhibited by IFN-y-treated cells, we have previously examined the effect of this IFN on the ability of fibroblast monolayers to deplete NK cells from suspensions of leukocytes (15, 16). To our surprise, we found that following incubation with IFN-y-treated fibroblasts, NK cells were not more, but significantly less capable of exerting natural killing activity, as compared to NK cells which had been incubated, in parallel, with fibroblasts not treated with IFN-7. We have, therefore, suggested that the resistance of IFN-y-treated fibroblasts to the killing by NK cells does not reflect decreased binding of the killer cells to their targets but rather some inhibitory effect of the IFN-treated fibroblasts on the function of the NK cells ( 15, 16).

The establishment of the anti-NK effect, like that of all other known IFN effects, can be prevented by inhibitors of protein synthesis (13). We therefore might assume that similarly to the antiviral effect (22) the anti-NK effect of IFN reflects the function of some cellular proteins whose synthesis is specifically stimulated by the IFN. We may further speculate that the proteins responsible for the anti-NK effect are induced by IFN-7 at IFN concentrations significantly lower than those inducing the proteins involved in the resistance to infection by VSV. Recently, we have found that induction of certain proteins can indeed be detected in cells treated by much lower concentra- tions of IFN--y than those inhibiting VSV replication; we found that the synthesis of the major histocompatibility proteins HLA-A, B, and C and &microglobulin and of the mRNAs coding for these proteins, can be induced by IFN--/ at concentrations hundreds to thousands of times lower than those necessary for inhibition of VSV growth, while with EN-a and /3 these proteins are induced only at concentrations which produce virus inhibition. Another EN-induced protein-the enzyme (2-5’) oligo(A)synthetase, which is probably involved in the antiviral state, is increased by

SHORT COMMUNICATIONS 395

IFN-y only at much higher IFN concentrations and its increase correlates to the antiviral effect (23). Further characterization of those proteins induced specifically at low concentrations of IFN--, might give a clue to the biochemical nature of the anti- NK effect.

ACKNOWLEDGMENTS

I am grateful to Dr. M. Revel for his support and interest in this study. I thank Drs. P. Gray and D. V. Goeddel for the generous gift of IFN--( produced with cloned cDNA and Ms. S. Budilovsky for her skillful assistance.

REFERENCES

1. Herberman, R. B., Djeu, J. Y., Kay, H. D., Ortaldo, J. R., Riccardi, C. R., Bonnard, G. D., Holden, H. T., Fagnani, R., Santoni, A., and Puccetti, P., Zmmunol. Rev. 44, 43, 1979.

2. Kiessling, R., and Wigzell, H., Zmmunol. Rev. 44, 165, 1979. 3. Welsh, R. M., Antiviral Rex 1, 5, 1981. 4. Wigzell, H., In “The Biology of the Interferon System,” pp. 211-215. Elsevier/North-Holland, Am-

sterdam, 198 I. 5. Parkman, R., and Rosen, F., J. Exp. Med. 144, 1520, 1976. 6. Saksela, E., Timonen, T., Ranki, A., and Hayry, P., Zmmunol. Rev. 44, 71, 1979. 7. Hansson, M., Kiessling, R., Anderson, B., and Welsh, R. M., J. Zmmunol. 125, 2225, 1980. 8. Gresser I., In “The Biology of the Interferon System,” pp. 141-147. Elsevier/North-Holland, Amster-

dam, 1981. 9. Savary, C. A., and Lo&ova, E., J. Zmmunol. 120, 239, 1978.

10. Hochman, P. S., and Codkowicz, G., J. Zmmunol. 123,968, 1979. 11. Droller, M. J., Perlman, P., and Schneider, M. I., Cell. Zmmunol. 39, 154, 1978. 12. Roder, J. C., and Klein, M., J. Zmmunol. 123, 2785, 1979. 13. Trinchieri, G., and Santoli, D., J. Exp. Med. 147, 1314, 1978. 14. Epstein, L. B., In “Interferon 3” (I. Gresser, Ed.) pp. 13-44. Academic Press, New York, 1981. 15. Wallach, D., J. Interferon Res. 2, 329, 1982. 16. Wallach, D., In “Human lymphokines, Biological Immune Response Modifiers” (A. Khan and N. 0.

Hill, Eds.), pp. 435-449. Academic Press, New York, 1982. 17. Weissenhach, J., Zeevi, M., Landau, T., and Revel, M., Eur. J. Biochem. 98, I, 1979. 18. Rubinstein, M., Levy, W. P., Moschera, J. A., Lai, C. Y., Hershberg, R. D., Bartlett, R. T., and Pestka,

S., Arch. Biochem. Biophys. 210, 307, 1981. 19. Yip, Y. K., Pang, R. H. L., Urban, C., and Vilcek, J., Proc. Nut. Acud. Sci. USA 78, 1601, 198 1. 20. Gray, P. W., et al., Nature (London) 295, 503, 1982. 21. Henney, C. S., Kuribayashi, K., Kern, D. E. and Gillis, S., Nature (London) 291, 335, 1981. 22. Revel, M., In “Interferon 1” (I. Gresser, Ed.) pp. 101-163. Academic Press, New York, 1979. 23. Wallach, D., Fellous, M., and Revel, M., Nature (London) 299, 833, 1982.