A novel mechanism of immunosuppression mediated by ethanol

CELLULAR

IMMUNOLOGY

102,1-9 ( 1986)

A Novel Mechanism of lmmunosuppression DAVIDR.

Mediated by Ethanol’

KAPLAN*

Institute of Pathology, Case WesternReserve University School of Medicine, Cleveland, Ohio 44106 Received February 24,1986; acceptedApril 24, I986 Ethanol in concentrations equivalent to levels achieved by the ingestion of moderate amounts ofalcoholic beverageshas been shown to inhibit human T-lymphocyte proliferation after activation in vitro by mitogens, phorbol myristic acetate, or anti-CD3 monoclonal antibodies. As determined by probit analysis, this inhibition was monophasic, suggestingthat ethanol affected a single limiting component of T-cell proliferation. In experiments designed to test the effect of ethanol on various aspects of proliferation, it was demonstrated that ethanol did not inhibit interleukin 2 (IL-2) production or the acquisition of IL-2 receptors. However, the capacity of exogenously supplied IL-2 to stimulate proliferation of T cells that had previously acquired IL-2 receptors was suppressedby ethanol in a dose-dependent manner, and this suppression was monophasic. Consequently, ethanol was able to inhibit T-lymphocyte proliferation when added several days after the initial stimulation. This mechanism appears to be unique among immunosuppressive agents thus far studied; cyclosporin A and corticosteroids inhibit IL-2 production and are required at the initiation of activation for maximal effect. Synergistic inhibition of T-cell proliferation was seen with ethanol plus cyclosporin A: the level of inhibition with 250 rig/ml cyclosporin A alone was equivalent to the level seen with 62 rig/ml cyclosporin A plus 5 ti (24 mg%) ethanol. 0 1986 Academic PESS,hc.

INTRODUCTION Suppression of the immune response is a therapeutically important technique for the treatment of autoimmune diseasesand transplantation rejection. Several drugs mediate immunosuppression by different mechanisms of action, including corticosteroids which inhibit interleukin 2 (IL-2) production ( 1, 2), Fc receptor expression (2), mononuclear cell circulation (3), and monocyte differentiation (4), cyclosporin A which inhibits IL-2 production ( 1, 5), and azathioprine which suppressesmitosis (6). Because of the significant but different side effects of these drugs, a rationale for multidrug immunosuppressive therapy has been developed based on reduced doses of each of the drugs in the presence of the others. In this light the search for agents that suppressthe immune response via novel mechanisms or with different attendant side effectscontinues to be important. Ethanol is a powerful immunosuppressive agent. In vitro effectsof ethanol include profound inhibition of lectin and antigen-stimulated T-lymphocyte proliferation ’ This work was supported in part by USPHS Grants CA36 189 and A122505. 2 To whom correspondence should be addressed at the Institute of Pathology, Case Western Reserve University, 2085 Adelbert Road, Cleveland, Ohio 44106.

0008-8749186$3.00 Copyright 0 1986 by Academic Press,Inc. All rights of reproduction in any form reserved.

2

DAVID

R. KAPLAN

(7, 8), suppression of bone marrow colony formation (7, 9), and alteration in the glycosylation of immunoglobulin molecules (10). These effects have been observed at concentrations of ethanol that are not cytotoxic and that are equivalent to serum levels achieved by ingestion of moderate to large quantities of alcoholic beverages. Although ethanol is immunosuppressive, the mechanism of action of ethanol on lymphocytes has not been elucidated. In this manuscript the ethanol-mediated inhibition of T-lymphocyte proliferation is investigated and a novel mechanism of action is described. MATERIALS AND METHODS Materials. Ethanol purchased from AAPER (Louisville, KY.) at 190 proof was used at various concentrations in culture medium. This source of ethanol was free of methanol, benzene, amyl alcohol, isobutyl alcohol, and other contaminants. A concentration of 80 mM ethanol was equivalent to 375 mg/dl measured by specific enzymatic spectrophotometry (Dr. Clive Hamlin, University Hospitals of Cleveland, Cleveland, Ohio). Phytohemagglutinin-P (PHA)3 was purchased from Difco Laboratories (Detroit, Mich.), Concanavalin A (Con A) from Miles Laboratories (Naperville, Ill.), phorbol myristic acetate (PMA) from Sigma Chemical Co. (St. Louis, MO.), and OKT3, a mouse monoclonal antibody specific for the CD3 (formerly known as T3) antigen on the surface of T cells, from Ortho Diagnostics (Raratan, N.J.). Cyclosporin A and cyclosporin H were kindly provided by Dr. J. Bore1 (Sandoz, Basel, Switzerland). These drugs were dissolved in dimethyl sulfoxide at 1 mgfml. Proliferation assay. Peripheral blood mononuclear cells were isolated from healthy volunteers by Ficoll-Hypaque (Sigma) discontinuous gradient centrifugation. These cells were immediately suspended in culture medium: RPM1 1640 (M. A. Bioproducts, Walkersville, Md.) with 10% fetal bovine serum, penicillin, streptomycin, glutamine, and Hepes. For stimulation 5 X lo4 cells were placed in four replicate wells of 96-well flat-bottom plates with 0.125% PHA, 10 g/ml Con A, 1 rig/ml OKT3, or 20 r&ml PMA and various concentrations of ethanol. After 3 days of incubation at 37°C in 7% COZ the cultures were pulsed with 1 PCi of tritiated thymidine and incubated for an additional 4 hr for mitogen stimulation and 18 hr for anti-CD3 and PMA stimulation. The contents of the wells were collected on glass fiber filters with an automated harvester (Mini-Mash, M. A. Bioproducts, Walkersville, Md.). Radioactivity was measured after immersion in a toluene scintillation fluid on a Beckman 380 1 scintillation counter. The quench of each sample was measured and compared to a standardized quench curve in order to calculate the efficiency of counting. This efficiency measurement was used with cpm to derive sample dpm. Mean dpm was used to calculate the percentage inhibition; standard deviations were usually lessthan 10% of the mean and always lessthan 20%. Interleukin 2 production and assay. One million peripheral blood mononuclear cells were cultured in 1 ml of complete medium with 10 pg/ml Con A. After 16-20 hr of incubation at 37°C in 7% CO2 the supematants were collected and filtered through 0.2-pm-size filters and then assayedfor the presence of IL-2 determined in a biological assay using PHA-stimulated peripheral blood mononuclear cells cultured 3 Abbreviations used: Con A, concanavalin A; PHA, phytohemagglutinin; tate; PBMNC, peripheral blood mononuclear cells.

PMA, phorbol myristic ace-

IMMUNOSUPPRESSION

BY ETHANOL

3

for lo- 15 days ( 11). Twenty-forty thousand blasts were plated in four replicate wells of a flat-bottom 96-well microtiter plate. Test culture supernatants were added in varying concentrations with 50 mA4 LYmethyl mannoside (Sigma). This concentration of cymethyl mannoside was shown not to interfere with IL-2 activity while completely inhibiting stimulation by 10 pg/ml Con A (data not shown). After 1 day of incubation at 37°C the cultures were pulsed with 1 PCi of tritiated thymidine and incubated for an additional 18 hr. The contents of the well were collected on glass fiber filters with an automated harvester and processedas described above. This assay gave parallel straight lines in probit analysis ( 12) for MLA 144 supernatant (a source of IL-2, Ref. (13)), recombinant IL-2 (Genzyme, Boston, Mass.), and affinity-purified IL-2 (kindly provided by Dr. Gary Thurman, Frederick, Md.). T-cell clones. Establishment and maintenance of the human cytotoxic T-lymphocyte clones V 1, A4, and T5 have been previously described ( 14, 15). Cell su~facej7~o~esce~ce. IL-2 receptor expression was determined by staining cells with anti-Tat monoclonal antibody kindly provided by Dr. Thomas Waldmann (National Institutes of Health, Bethesda, Md.). The procedure has been previously described (14). Briefly, lo6 cells were incubated for 30 min at 4°C with a saturating amount of antibody (0.02 ml of 1:500 ascites) in phosphate-buffered saline containing 0.1% azide and 0.5% bovine serum albumin. After three washesa second incubation for 30 min at 4°C was made with fluoresceinated rabbit anti-mouse IgG (Miles Laboratories) in the same medium. After this second incubation the cells were washed, and then 10,000 cells were analyzed for log relative surface fluorescence by flow cytometry (EPICS V, Coulter, Hialeah, Fla.). RESULTS PBMNC from 17 healthy volunteers were isolated and stimulated with PHA for 3 days in the presence of various concentrations of ethanol. On the third day tritiated thymidine incorporation was measured during a 4-hr pulse. Ethanol mediated a dosedependent inhibition of thymidine incorporation; furthermore, probit analysis of this inhibition resulted in a single straight line which suggestedthat ethanol affected a single component of proliferation (representative results shown in Fig. IA). The same pattern of inhibition of PHA stimulation by ethanol was seen for the PBMNC from all 17 volunteers. Moreover, the concentration of ethanol that resulted in 50% inhibition of PHA stimulation was fairly constant in these studies; mean and standard deviation of the ethanol concentration required to inhibit 50% of the lectin-stimulated thymidine incorporation for the 17 volunteers was 26 + 8 mA4 (range, 14-43 mM). Similar results were obtained with PBMNC stimulated with another T-lymphocyte mitogen, Con A (representative results shown in Fig. 1B). Since T cells can be specifically activated via surface CD3 (formerly known as T3) which forms part of the T-cell antigen receptor (16- 18), ethanol effects on anti-CD3 monoclonal antibody stimulation of PBMNC were also assessed,and monophasic inhibition was again observed (representative results shown in Fig. 1C). Furthermore, the ethanolmediated decreasein thymidine incorporation by anti-CD3-stimulated PBMNC corresponded to a dose-dependent inhibition of blast number: in one experiment the number of blasts in control cultures was 101 blasts/ml, in the presence of 40 mM ethanol, 23 blasts/ml, and in the presence of 80 Methanol, 6 blasts/ml. Because proliferation of T cells is known to require the production and secretion of IL-2, the acquisition of IL-2 surface receptors, and the interaction of this lympho-

DAVID R. KAPLAN

FIG. I. Effect of ethanol on T-lymphocyte proliferation. Peripheral blood mononuclear ce.lls from healthy volunteers were stimulated with (A) a I :800 dilution of PHA, (B) IO &ml Con A, or (C) 2 mg/ml OKT3. Log, dilutions of ethanol (I log2dilution = 80 mM) were added and tritiated thymidine incorporation during a 4-hr pulse was measured on the third day. The control responseswithout ethanol (mean dpm + SD)were 159.491 f 14,177(A), 141,841 k 13,812 (B), and 292,019 rt 24,5l0(C).Theresultsshownare representative of several experiments: I7 for PHA stimulation, 2 for Con A stimulation, and 8 for antiCD3 stimulation.

kine with the receptor ( 14, 19), studies were initiated to determine the mechanism of inhibition mediated by ethanol. The results in Table 1 demonstrate that ethanol, even in high concentrations, does not significantly suppress IL-2 secretion. The sensitivity of this assay was demonstrated by incubating the cells with cyclosporin A which almost completely inhibited IL-2 production. Cyclosporin H, a nonimmunosuppressive congener of cyclosporin A, had no effect. Similarly ethanol had little effect on IL2 surface receptor expression (Table 2). There was no significant decreasein cellular viability in those cultures containing up to 80 mM ethanol (data not shown). Conversely, the activity of IL-2 to promote proliferation ( 19) of T cells through interaction with specific receptors was markedly inhibited by ethanol. Ten days after PBMNC were stimulated with either PHA or anti-CD3 antibodies, the cells were cultured for 24 hr in the presence of 2 U/ml exogenous IL-2 and various concentrations of ethanol. Tritiated thymidine incorporation during the last 4 hr of this culture period was inhibited by ethanol in a dose-dependent manner, and probit analysis demonstrated that this inhibition was also monophasic (Fig. 2). Furthermore, IL-2driven

IMMUNOSUPPRESSION

5

BY ETHANOL

TABLE 1 Effect of Ethanol on the Production of Interleukin 2

Inhibitor

Concentration of inhibitor

-

-

Ethanol Ethanol Ethanol Ethanol Ethanol Cyclosporin A Cyclosporin H

1omM 2omM 4omM 80 mM 16OmM 1000 @ml 1000 rig/ml

IL-2 production (dpm f SD) 13,680 III 1040 14,166+ 1113 12,897 -t 490 11,381 + 155 12,300 k 728 10,746 +- 184 2,063 f 207 12,274 +- 465

Note. One million peripheral blood mononuclear cells were cultured in 1 ml complete medium in the presence of Con A with various concentrations of ethanol. After 20 hr incubation supematants were collected, filtered, and assayed at a 10% concentration for IL-2 production using lectin-activated indicator cells. 0.05 M LYmethyl mannoside was included in these assaysto inhibit stimulation by residual Con A. The background incorporation of thymidine in the absence of exogenous IL-2 was 1472 dpm.

proliferation of human cytotoxic T-cell clones V 1 and A4 ( 14, 15) 5 days after stimulation was similarly inhibited by ethanol (Fig. 2), showing that the ethanol-mediated depression of proliferation was not dependent on interactions between T cells and other cells. Six other cloned T-cell lines gave similar results (data not shown). It is interesting to note that the long-term T-cell clones were approximately four times more sensitive to ethanol than mitogen-stimulated PBMNC cultured for 10 days. Although the explanation for this difference has not been determined, dose-dependent inhibition of IL-2-driven proliferation occurred in both instances. Thus, it appears that the suppression of T-cell proliferation by ethyl alcohol can be accounted for by an effect on the ability of IL-2 to drive proliferation in cells that have previously acquired IL-2 receptors. Moreover, it is clear that ethanol added after the initial stimulus can inhibit an ongoing T-lymphocyte proliferation.

TABLE 2 Effect of Ethanol on Surface Expression of Interleukin 2 Receptors Concentration of ethanol (mm

Percentage Tat+ cells

Mean fluorescence intensity (channel number of positive cells)

0 10 20 40 80 160

47 59 53 54 48 57

66 72 66 66 68 67

Note. One million peripheral blood mononuclear cells were cultured in 1 ml complete medium in the presence of Con A with various concentrations of ethanol. After 3 days incubation the cells were assayed for surface expression of IL-2 receptors by staining with the monoclonal anti-Tat antibody and a fluoresceinated goat anti-mouse immunoglobulin. Stained cells were analyzed for fluorescenceby flow cytometry.

DAVID R. KAPLAN

160 mM I

I 2

I

I

I

3

4

LOG2 DILUTION

I

I 5

I

6

OF ETHANOL

FIG. 2. Effect of ethanol on interleukin 2 activity. Log2 dilutions of ethanol (1 log* dilution = 160 mM) were added to IL-2 receptor positive cells in the presence of 2 U/ml IL-2 (2% MLA 144 supematant). PBMNC previously cultured for 10 days after mitogen stimulation (m) were plated in microtiter plates, pulsed with tritiated thymidine, and harvested 18 hr later. The response in the absence of ethanol was 12,647 dpm + 1330. The results shown are representative of four separate experiments. Long-term cytotoxic T cell clones A4 (A) and Vl (0) were stimulated with anti-CD3 and 4 days later were plated in microtiter plates. After culture overnight the wells were pulsed with tritiated thymidine and harvested 4 hr later. The response in the absence of ethanol was 96,68 1 dpm f 3744 for clone V 1 and 20,700 dpm 5 998 for clone A4. These results are representative of five separate experiments, and similar results were obtained with six other clones.

PMA alone has been shown to induce proliferation of human T lymphocytes from PBMNC without any detectable increase in the level of IL-2 mRNA (20) or IL-2 secretion (2 1). Since it appears that PMA stimulates an IL-24ndependent proliferation of human T cells, it was interesting to determine the effect of ethanol on this activity. Figure 3 shows that ethanol also inhibited PMA-induced proliferation in a monophasic way. Thus, ethanol might act on a pathway common to both IL-2 and PMA activity.

80 mn

, 1 1 1 1 2 3 4 LOG2 DILUTION OF ETHANOL

J

5

FIG. 3. Effect of ethanol on phorbol ester-stimulated proliferation. As in Fig. 1 except peripheral blood mononuclear cells were stimulated with 20 n&ml phorbol myristic acetate. The control response without ethanol (mean dpm + SD) was 86,276 + 7,5 19. The results shown are representative of five separate experiments.

IMMUNOSUPPRESSION

7

BY ETHANOL

TABLE 3 Enhanced Inhibition of T-Cell Proliferation with Cyclosporine A and Ethanol

Concentration of CsA” b-cdml)

Concentration of ethanol h‘w

500 500 500 250 250 250 125 125 125 125 62 62 62 62 31 31 31 31

5 10 20 5 10 5 10 5 10 20 5 10 20 5 10 20

Tritiated thymidine incorporation (dpm + SD) 283304 f 2411942 186480 + 27432 f 63374 + 6666k 1198+ 11744Ok 21705 f 2104k 160767 f 68866 2 37394 f 1505 + 198567 f 124055 f 69259 f 3815+ 229094 f 163904 f 106738 f 8805 +

6867 7369 5870 7986 10979 2109 619 13315 11745 764 2554 3382 1630 800 9224 4731 11038 1351 11331 11809 5203 2389

Percentage inhibition (&SD)

15+3 34+-2 90+3 78 + 4 98+ 1 (81)b 100 f 0 (85) 59+5 92 + 4 (65) 99 + 0 (73) 43+ 1 76 f 1(52) 87 + 1 (62) 99 + 0 (94) 3Ok3 56 f 2 (40) 76 + 4 (54) 99 It 0 (93) 19+4 42+4(31) 62 k 2 (47) 97 f 1 (92)

Note. One hundred thousand PBMNC per microtiter well were stimulated with anti-CD3 antibodies in the presence of various concentrations of CsA and ethanol. Tritiated thymidine incorporation was measured on the third day after an 18-hr pulse. ’ CsA was dissolved in dimethyl sulfoxide. b Expected inhibition calculated by assuming no interaction between the two inhibitors is shown in parentheses.Analysis of variance shows significant synergy of inhibition between ethanol and cyclosporin A:P
Since the mechanism of ethanol-mediated suppression of T-cell proliferation is different from inhibition via cyclosporin A which depressesthe production of IL-2 and is required during the initiation of stimulation (1, 5), it was interesting to seeif ethanol could enhance the inhibition seenwith cyclosporin A. Table 3 shows that the inhibition of anti-CD3-stimulated T-cell proliferation was significantly enhanced by the presence of ethanol in addition to cyclosporin A and that this enhancement appeared to be synergistic. For instance, the percentage inhibition seen with 62 rig/ml CsA and 5 Methanol was 56% whereas the expected inhibition assuming no interaction between the two inhibitors was 40%. It is also interesting to note that the level of response seen in this case was equivalent to that seen with 250 rig/ml CsA alone (59%). DISCUSSION The mechanism of ethanol-mediated immunosuppression has been investigated with human T lymphocytes, and effects on T-cell proliferation have been described.

8

DAVID R. KAPLAN

The doses that demonstrated immunosuppression were equivalent to serum levels easily achieved in individuals: 10 mM = 48 mg%. Becauseethanol suppressesproliferation in a monophasic fashion, it appearsthat a specific, ethanol-sensitive reaction is limiting. For lectin-stimulated proliferation the mechanism of ethanol inhibition has been defined in terms of IL-2 and its receptor. The ability of exogenously supplied IL-2 to stimulate proliferation of T cells with preformed IL-2 receptors was specifically suppressedin a dose-related manner. There was no inhibition of cellular viability, IL-2 production, or IL-2 receptor expression in the presence of immunosuppressive concentrations of ethanol. The concentration of ethanol required to inhibit IL-2-driven proliferation varied according to t-he proliferating cell population. Long-term T-cell clones were four times more sensitive to ethanol than mitogen-stimulated PBMNC cultured for 10 days. Moreover, activation of PBMNC by anti-CD3 or mitogen showed intermediate sensitivity to ethanol. The reason for this differential responsiveness has not been determined; however, it should be noted that in all instances, including inhibition of PMA-induced stimulation of PBMNC, the inhibition is dose-dependent and probit analysis revealed parallel lines of inhibition. The molecular explanation for ethanol-mediated inhibition has not been determined; however, there are several possibilities. Ethanol could act on soluble IL2 to prevent its binding to the IL2 receptor. Since there are high-affinity and low-affinity IL-2 receptors and only the high-affinity receptors are involved in IL-2 signal transduction (22) it is also possible that ethanol either inactivates the high-affinity receptor or mediated the transformation of a high-affinity receptor to a low-affinity receptor. Finally, it is conceivable that ethanol affects an event subsequent to the receptorligand interaction. It has been recently shown that the interaction of IL-2 with its specific receptor induces a redistribution of protein kinase C from the cytosol to the plasma membrane (23). PMA has also been associated with a similar redistribution and activation of protein kinase C (24) and its action on this enzyme has been implicated in cell surface signal transduction (25). Since PMA stimulates cellular proliferation in a IL-2Gndependent fashion (20,2 l), it appears that IL-2 and PMA may independently activate the same pathway leading to cellular proliferation. Inhibition by ethanol of both IL-Zdriven and PMA-driven T-cell proliferation suggeststhat ethanol might act at a step common to both pathways. Many substances suppress T-lymphocyte proliferation. Corticosteriods (1, 2), cyclosporin A ( 1,5), and 1,25-dihydroxyvitamin D (26) inhibit IL-2 production; azathioprine nonspecifically inhibits cellular proliferation by acting on DNA replication (6); prostaglandin E2 suppressesIL-2 production and transferrin receptor expression (27); oubain binds to the (Na+-K+)ATPase and thereby nonspecifically inhibits both IL-2 production and IL-2 action on IL-2 receptor positive cells (28); and quinacrine (29) and wheat germ agglutinin (30) suppress the expression of IL-2 receptor. Thus, the specific mechanism of action of ethanol on T-cell proliferation appears to be unique. Furthermore, both cyclosporin A and corticosteroids inhibit T-cell function by affecting the initial stage of T-cell activation, and consequently these drugs must be present early in the immune response to achieve optimal immunosuppression. Ethanol, on the other hand, inhibits T-cell proliferation in a latter stage, and thus it represents an agent with different immunosuppressive characteristics. In this study the addition of ethanol markedly decreasedthe dose of cyclosporin A required to achieve a given level of inhibition of T-lymphocyte proliferation in vitro,

IMMUNOSUPPRESSION

BY ETHANOL

9

and the immunosuppressive effects of the two drugs together were synergistic. It appears that ethanol represents a class of agent that inhibits T-cell function in vitro by a novel mechanism which synergizes with the inhibitory action of cyclosporin A. A drug with the inhibitory characteristics of ethanol, but without the marked attendant toxicity, might be useful as an additional therapeutic agent for the treatment of undesirable immune responses. ACKNOWLEDGMENTS I thank Dr. K. Smachlo, Dr. M. Lamm, Dr. A. Stavitsky, Dr. A. TartakolT, Dr. N. Smith, and Dr. J. Nedrud for critical reading of the manuscript, Ms. D. Gould for technical assistance,and Ms. J. Nagy for help in preparing the manuscript.

REFERENCES 1. Larsson, E., J. Immunol. 124,2828, 1980. 2. Gillis, S., Crabtree, G., and Smith, K., J. Immunol. 123, 1642, 1979. 3. Cupps, T., Edgar, L., Thomas, C., and Fauci, A., J. lmmunol. 132,170, 1984. 4. Rinehart, J., Wuest, D., and Ackerman, G., J. Immunol. 129, 1436, 1982. 5. Bunjes, D., Hardt, C., Rollinghoff, M., and Wagner, H., Eur. J. Immunol. 11,657, 198 1. 6. Bach, J. F., In “Frontiers of Biology,” Vol. 41. North-Holland, Amsterdam, 1975. 7. Tisman, G., andHerbert, V., J. Clin. Invest. 52,1410, 1973. 8. Roselle, G., and Mendenhall, C., J. Clin. Lab. Immunol. 9,33, 1982. 9. Meagher, R., Sieber, R., and Spivak, J., N. Engl. J. Med. 307,845, 1982. 10. Eaton, L., and Ingram, L., Alcohol Clin. Lab. Rex 6,459, 1982. 11. Ruscetti, R., Morgan, D., and Gallo, R., J. Immunol. 119, 13 1, 1977. 12. Gillis, S., Ferm, M., Ott, W., and Smith, K., J. Immunol. 120,2027, 1978. 13. Rabin, H., Hopkins, R., Ruscetti, F., Neubauer, R., Brown, R., and Kawakami, T., J. Immunol. 127, 1852,198l. 14. Kaplan, D., Braciale, V., and Braciale, T., J. Immunol. 133, 1966, 1984. 15. Kaplan, D., Griffith, R., Braciale, V., and B&ale, T., Cell. Immunol. 88, 193, 1984. 16. Van Wauwe, J., DeMey, J., and Goossens,J., J. Immunol. 124,2708, 1980. 17. Chang, T., Kung, P., Gingas, S., and Goldstein, G., Proc. Natl. Acad. Sci. USA 78, 1805, 1981. 18. Tax, W., Willems, H., Reekers, P., Capel, P., and Koene, R., Nature (London) 304,445, 1983. 19. Cantrell, D., and Smith, K., J. Exp. Med. 158, 1895, 1983. 20. Hirano, T., Fugimoto, K., Teranishi, T., Nishino, N., Onoue, K., Maeda, S., and Shimada, K., J. Immunol. 132,2165,1984. 21. Stadler, B., Dougherty, S., Farrar, J., and Oppenheini, J., J. Immunol. 127, 1936, 1981. 22. Robb, R., Greene, W., and Rusk, C., J. Exp. Med. 160, 1126, 1984. 23. Farrar, W., and Anderson, W., Nature (London) 315,233, 1985. 24. Kraft, A. S., and Anderson, W. B., Nature (London) 301,62 1, 1983. 25. Nishizuka, Y., Nature (London) 308,693, 1984. 26. Rigby, W., Stacy, T., and Fanger, M., J. Clin. Invest. 74, 1451, 1984. 27. Chouaib, S., Welte, K., Mertelsmann, R., and DuPont, B., J. Immunol. 135, 1172, 1985. 28. Stoeck, M., NortolT, H., and Resch, K., J. Immunol. 131, 1433, 1983. 29. Namiuchi, S., Kumagai, S., Imura, H., Suginoshita, T., Hattori, T., and Hirata, F., J. Immunol. 132, 1456,1984. 30. Reed, J., Robb, R., Greene, W., and Nowell, P., J. Immunol. 134,314, 1985.