Inhibition of Interferon-Mediated Induction of Indoleamine 2,3

Proc. Natt Acad. Sci. USA Vol. 78, No. 12, pp. 7327-7330, December 1981 Biochemistry

Inhibition of interferon-mediated induction of indoleamine 2,3-dioxygenase in mouse lung by inhibitors of prostaglandin biosynthesis (lung slices/cyclooxygenase/phospholipase AO) SHIGETOSHI SAYAMA*, RYOTARO YOSHIDA*, TOHRU OKUt, JIRO IMANISHIt, TSUNATARO KISHIDAt, AND OSAMU HAYAISHI* *Department of Medical Chemistry, Kyoto University Faculty of Medicine, Kyoto 606, Japan; and tDepartment of Microbiology, Kyoto Prefectural University of Medicine, Kyoto 602, Japan Contributed by Osamu Hayaishi, August 10, 1981

ABSTRACT Inhibitors of fatty acid cyclooxygenase such as I'lnergie Atomique, France, and purified by column chro- indomethacin (0.1 mM), phenylbutazone (0.3 mM), and aspirin (1 matography on Dowex 50W-X2 (H' form; 0.5 x 1.0 cm) as de- mM) were found to suppress almost completely the interferon- scribed (2). Catalase was a product of Boehringer Mannheim mediated induction ofindoleamine 2,3-dioxygenase in mouse lung and was dialyzed to remove a thymol preservative. Methylene slices. However, phenacetin, an anti-inflammatory agent devoid blue and ascorbic acid were obtained from Wako Pure Chem- ofcyclooxygenase inhibitory activity, and sodium salicylate, a weak icals (Osaka, Japan). Aspirin, phenylbutazone, indomethacin, inhibitor ofcyclooxygenase, were much less active under identical dexamethasone, betamethasone, and 17(3estradiol were pur- conditions. Glucocorticoids including dexamethasone, betametha- chased from Sigma. Sodium salicylate, testosterone, and phen- sone, and cortisone, all of which are inhibitors of phospholipase acetin were from Nakarai Chemicals (Kyoto, A2, blocked induction of the dioxygenase by interferon in the na- Japan), and aldos- nomolar range, whereas other steroid hormones, such as aldosterone was from Merck (Darmstadt, Federal Republic of terone, testosterone, and 171-estradiol, were all but ineffective. Germany). Cortisone acetate was a product of Banyu Pharma- These results suggest that the enzymes phospholipase A2 and fatty ceutical (Tokyo). Collagenase (type III) was purchased from acid cyclooxygenase, both of which are essential for the biosyn- Worthington and trypsin was from Difco. Mouse L-cell inter- thesis of prostaglandins, play an important role in the induction feron (specific activity, 2.96 X 10r' units/mg of protein) was of indoleamine 2,3-dioxygenase by interferon. purified as described (3). All other chemicals were of reagent grade. Indoleamine 2,3-dioxygenase [IDOase; indoleamine:oxygen Animals. Specific-pathogen-free male slc:ICR mice weighing 2,3-oxidoreductase (decyclizing)], a heme-containing enzyme, 31 ± 2 g were purchased from the Shizuoka Agricultural Co- catalyzes oxidative cleavage ofthe indole ring ofvarious indole- operative Association for Laboratory Animals (Shizuoka, Japan), amine derivatives such as tryptophan, 5-hydroxytryptophan, and raised under specific-pathogen-free conditions at 25 ± 20C tryptamine, and serotonin, with the incorporation oftwo atoms and -50% humidity in the Institute of Laboratory Animals, of oxygen into the pyrrole moiety (1). The enzyme is ubiqui- Kyoto University. tously distributed in various organs ofmammals except for the Incubation of Tissues, Enzyme Assay, and Viability of Dis- liver (2). We previously reported that IDOase is markedly in- persed Lung Cells. The preparation of lung slices, incubation duced by mouse interferon in mouse lung slices (3). conditions, and enzyme assays were carried out as described Recently, considerable attention has been directed to the in- (3) except that the medium did not contain fetal calfserum. After terferon-associated changes in the level ofprostaglandins (4, 5). incubation, lung slices were totally digested with 0.1% trypsin/ Pottathil et aL (6) have shown that establishment of the inter- 0.2% collagenase and viability of the dispersed lung cells was feron-mediated antiviral state is suppressed by nonsteroidal estimated by the trypan blue exclusion technique (8). Detailed anti-inflammatory agents that inhibit fatty acid cyclooxygenase. methods for the tissue digestion will be described elsewhere. In addition, Chandrabose et aL (7) showed that cells of the in- Protein Determination. Protein concentration was deter- terferon-resistant R3 clone of L 1210 mouse leukemia cells lack mined by the method of Lowry et aL (9) using bovine serum fatty acid cyclooxygenase activity. To elucidate the precise albumin as standard. mechanism of the IDOase induction in the mouse lung by interferon, effects of various anti-inflammatory agents on the RESULTS IDOase induction were examined. The present study shows When mouse lung slices were incubated with interferon (104 that nonsteroidal anti-inflammatory agents that possess fatty units/ml), the specific activity of IDOase6in the supernatant acid cyclooxygenase inhibitory activity and steroid hormones fraction of the slices increased almost linearly for 50 hr. then that inhibit phospholipase A2 suppress the induction ofIDOase started to decrease, and reached a normal value after --7 days by mouse interferon. as reported previously (3) (Fig. 1). When indomethacin (10 or 30 AiM) was added simultaneously with interferon (104 units/ ml) to the mouse lung slices, the time course of the IDOase MATERIALS AND METHODS induction was essentially identical but the initial rate, as well Chemicals. L-[ring-2-'4C]Tryptophan (30-35 Ci/mol; 1 Ci as the extent ofinduction, decreased in a dose dependent man- = 3.7 X 10'° becquerels) was purchased from Commisariat a ner. In the absence of interferon, slight but reproducible en-

The publication costs ofthis article were defrayed in part by page charge Abbreviations: IDOase, indoleamine 2,3-dioxygenase [indole- payment. This article must therefore be hereby marked "advertise- amine:oxygen 2,3-oxidoreductase(decyclizing)]; IC50, concentration re- ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. quired for 50% inhibition. 7327 Downloaded by guest on September 25, 2021 7328 Biochemistry: Sayama et aL Proc. Nad Acad. Sci. USA 78 (1981)

100 *d 04 c0~

U bo < 50

U~~~~~

2 4 6 Time, days FIG. 1. Time course of induction of IDOase by interferon with or without indomethacin. Indomethacin was added simultaneously with lo1-6 10-5 10-4 10-3 interferon (104 units/ml) at zero time and mouse lung slices were incubated. Samples were harvested at the indicated times and the spe- Inhibitor, M cific activity of IDOase was determined. Results represent means of duplicate experiments. o, Untreated; 9, interferon alone; A, interferon/ FIG. 2. Effects of nonsteroidal anti-inflammatory agents on in- 10 indomethacin; A, interferon/30 indomethacin. duction of IDOase by interferon. Anti-inflammatory agents were ttM /iM added simultaneously with interferon (104 units/ml) to the culture medium at zero time, and incubation was carried out for 30 hr. Enzyme activity is expressed as nmol of product per hr/mg of protein. Results zyme induction was observed as reported previously (3). The represent mean ± SEM of quadruplicate experiments. Enzyme activ- induction in the absence of interferon was inhibited by indo- ity of interferon-treated samples minus enzyme activity of untreated methacin to the same extent. The IDOase activity was deter- samples in the absence of inhibitors represents 100% activity. o, As- mined at 30 hr of incubation in the following experiments be- pirin; *, sodium salicylate; A, indomethacin; A, phenylbutazone; a, cause this incubation time seemed most suitable to show clearly phenacetin. the effects of various agents on the induction of IDOase by interferon. pholipase A2 reported by other investigators (14), although the When inhibitors offatty acid cyclooxygenase such as aspirin, effective concentrations were three orders ofmagnitude higher indomethacin, and phenylbutazone were added simultaneously in the latter case. with interferon (10 units/ml), these agents blocked the induction ofIDOase by interferon in a dose-dependent manner (Fig. 2). In contrast, sodium salicylate, a weak inhibitor offatty acid DISCUSSION cyclooxygenase, and phenacetin, an anti-inflammatory agent Interferon has been reported to be involved in various biological devoid of cyclooxygenase inhibitory activity, were much less events including establishment ofthe antiviral state (15), phago- effective at 1 mM. The total number of cells in the dispersed cytosis (16), cell-mediated immunity (17), and tumorigenesis cell suspensions from lung slices was almost the same in the (18). Numerous studies have been carried out concerning the presence and absence of inhibitors during the incubation pe- mechanisms of these interferon actions. Four enzymes (2'- riod, and the viability ofcells after 30 hrofincubation was >90% 5')oligoadenylate synthetase (19), protein kinase (20), phospho- in all cases (data not shown). The concentrations of inhibitors diesterase (21), and IDOase (3), have been reported to date to required for 50% inhibition (IC50) ofIDOase induction showed be induced by interferon. The first two have been suggested fairly good correlation with those of fatty acid cyclooxygenase as mediators in the establishment of the antiviral state by in- reported by other investigators (10-12) (Table 1). terferon (22). However, Wood and Hovanessian reported re- To ascertain that this inhibition ofenzyme induction by non- cently that (2'-5')oligoadenylate increased in embryonal carci- steroidal anti-inflammatory agents is due to inhibition of the noma cells without consequent development of the antiviral enzyme related to prostaglandin biosynthesis, the effects ofglu- state (23). Thus, the mechanism through which interferon man- cocorticoids, inhibitors of phospholipase A2, on the induction ifests its biological functions remains obscure. of IDOase by interferon were investigated. When glucocorti- Recently, Pottathil et aL (6) reported that certain nonsteroidal coids such as dexamethasone, betamethasone, and cortisone, anti-inflammatory agents, such as indomethacin, aspirin, and all ofwhich are potent anti-inflammatory drugs, were added at oxyphenylbutazone, suppressed the establishment of the anti- low concentration (1 nM) to the medium containing the mouse viral state by interferon using mouse L cells and suggested that lung slices, the induction ofIDOase by interferon was inhibited fatty acid cyclooxygenase might be an essential component in -50% (Fig. 3). By contrast, aldosterone, a mineralocorticoid, the optimal induction of the antiviral state by interferon. In- showed a suppressive effect at considerably higher concentra- terferon-mediated induction of (2'-5')oligoadenylate synthe- tions (>100 nM), probably due to its weak glucocorticoid ac- tase, however, was hardly affected by inhibitors of fatty acid tivity (13). Testosterone and 17,8-estradiol, both of which are cyclooxygenase (Y. Sokawa, personal communication). In the sex hormones, were inactive even at 1 mM. These findings in- present study, the IDOase induction by interferon was abol- dicate that, of steroid hormones, only glucocorticoids are ef- ished by inhibitors ofthe cyclooxygenase activity. The order of fective in suppressing the induction of IDOase by interferon. inhibitory potency of these agents for the induction of IDOase The order ofinhibitory potency ofthese agents for the IDOase by interferon was essentially the same as that for prostaglandin induction by interferon was essentially the same as that forphos- biosynthesis, although their IC50 values for the induction of Downloaded by guest on September 25, 2021 Biochemistry: Sayama et aL Proc. NatL Acad. Sci. USA 78 (1981) 7329

Table 1. Comparison of effects of anti-inflammatory agents on induction* of IDOase, fatty acid cyclooxygenase, and phospholipase A2 IC50, PzM IDOase Fatty acid Phospho- Agent induction* cyclooxygenase lipase A2 Ref. Indomethacin 12 0.16-3.60 - 10, 11 Phenylbutazone 83 7.23 10 Aspirin 120 36.7-150 - 10, 11 Phenacetin >1000 - Sodium salicylate >1000 750 - 12 Dexamethasone 0.001 - 3.6 14 Betamethasone 0.0006 - 3.6 14 Cortisone 0.02 - 120.7 14 Aldosterone 0.28 170-Estradiol >1000 - Testosterone >1000 * Calculated from Figs. 2 and 3.

IDOase by interferon were about one order ofmagnitude higher Therefore, it seems to be difficult to investigate the effects of than those for fatty acid cyclooxygenase (Table 1). glucocorticoids on interferon action by using antiviral activity Since indomethacin and related compounds are well known as the assay. In contrast, the assay procedure for IDOase activity to have biological activities other than inhibition ofprostaglan- is simple and not affected by anti-inflammatory drugs (data not din biosynthesis, such as inhibition of cyclicAMP-dependent shown). Thus, IDOase, the physicochemical properties of protein kinase (24) and of 11-lipoxygenase (25), it may be dif- which are well known (28), is a possible tool for study of the ficult to conclude that prostaglandins are related to the induc- mechanism of interferon action. tion ofIDOase by interferon on the basis ofthe inhibitory phe- The interferon response of virus-infected animals that had nomenon by these agents alone. Another enzyme involved in lost their ability to respond to interferon inducers can be re- prostaglandin biosynthesis is phospholipase A2, the activity of stored to normal levels when inducers are administered with which is inhibited by a newly synthesized protein induced by certain prostaglandins (29). Moreover, Mundy and Yoneda (30) glucocorticoids (26). As shown in Fig. 3, glucocorticoids blocked reported recently that exogenous prostaglandin E1 restored the the induction of IDOase by interferon at the nanomolar range production of osteoclast-activating factor in cultures treated but the reported IC50 values for phospholipase A2 are in the with indomethacin. Figs. 1-3 strongly suggest the possible par- micromolar range (Table 1). This effect may be simply due to ticipation of the prostaglandins in the induction of IDOase by the difference in the systems used since the latter experiments interferon. However, preliminary experiments in our labora- were carried out with perfused guinea pig lungs (13) or may im- tory indicate that exogenous arachidonic acid, prostaglandins ply that the induction of IDOase by interferon is suppressed E2, F2,, D2, 12 (sodium salt), and 6-keto Fin; and thromboxane through other glucocorticoid actions in addition to the inhibi- B2 (up to 1 mM) do not restore the IDOase induction suppressed tory activity on phospholipase A2. by inhibitors of prostaglandin biosynthesis. However, a possi- On the other hand, glucocorticoids are not only potent anti- bility still remains that continuous release of unstable prosta- inflammatory drugs but also agents enhancing viral growth (27). glandins such as G2 and H2, thromboxane A2, or other as yet unidentified metabolites of arachidonic acid might be respon- sible for the IDOase induction.

We are indebted to Dr. K. Takai in our laboratory for critical reading 100 ofthis manuscript. This work was supported in part by research grants from the Sakamoto Foundation, Nippon Shinyaku Co., Ltd., and the 0 Intractable Diseases Division, Public Health Bureau, Ministry of Health and Welfare, Japan and by grants-in-aid for scientific research and cancer research from the Ministry ofEducation, Science, and Cul- ad ture, Japan. Ex 50 6- 1. Hayaishi, 0. (1975)J. Biochem. (Tokyo) 79, 13-21. 2. Yoshida, R., Nukiwa, T., Watanabe, Y., Fujiwara, M., Hirata, F. & Hayaishi, 0. (1980) Arch. Biochim. Biophys. 203, 343-351. 3. Yoshida, R., Imanishi, J., Oku, T., Kishida, T. & Hayaishi, 0. (1981) Proc. Natl. Acad. Sci. USA 78, 129-132. 4. Yaron, M. & Yaron, I. (1977) Nature (London) 267, 457-459. 5. Fitzpatrick, F. A. & Stringfellow, D. A. (1980)1. Immunol 125, 0 lo-11 lo-10 10-9 1o-, 10-7 431-437. Hormone, M 6. Pottathil, R., Chandrabose, K. A., Cuatrecasas, P. & Lang, D. J. (1980) Proc. Natl Acad. Sci. USA 77, 5437-5440. FIG. 3. Effects of steroid hormones on induction of IDOase by in- 7. Chandrabose, K. A., Cuatrecasas, P., Pottathil, R. & Lang, D. terferon. Conditions were as described in the legend to Fig. 2. Results J. (1981) Science 212, 329-331. represent means of duplicate experiments. o, Dexamethasone; 0, be- 8. Phillips, H. J. (1973) in Tissue Culture Methods and Applica- tamethasone; A, cortisone; *, aldosterone; o, 17,8-estradiol; *, tions, eds. Kruse, P. F. & Patterson, M. K. (Academic, New testosterone. York), pp. 406-408. Downloaded by guest on September 25, 2021 7330 Biochemistry: Sayama et aL Proc. Natl. Acad. Sci. USA 78 (1981)

9. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. 21. Schmidt, A., Zilberstein, A., Shulman, L., Federman, P., Ber- (1951) J. Biol. Chem. 193, 265-275. issi, H. & Revel, M. (1978) FEBS Lett. 95, 257-264. 10. Flower, R., Gryglewski, R., Herbaczynska-Cedro, K. & Vane, J. 22. Kimchi, A., Shulman, L., Schmidt, A., Chernajovsky, Y., Fra- R. (1972) Nature (London) New Biol. 238, 104-106. din, A. & Revel, M. (1979) Proc. Nat. Acad. Sci. USA 76, 11. Lee, R. E. (1974) Prostaglandins 5, 63-68. 3208-3212. 12. Flower, R. J. (1974) Pharmacol. Rev. 26, 33-67. 23. Wood, J. N. & Hovanessian, A. G. (1979) Nature (London) 282, 13. Fieser, L. F. (1959) in Steroids, eds. Fieser, L. F. & Fieser, M. 74-76. (Van Nostrand-Reinhold, New York), pp. 600-726. 24. Kantor, H. S. & Hampton, M. (1978) Nature (London) 276, 14. Flower, R. (1978) in Advances in Prostaglandin and Thrombox- 841-842. ane eds. G. & G. (Raven, 25. Siegel, M. I., McConnell, R. T., Porter, N. A., Selph, J. L., Research, Galli, C., Galli, Porcellati, Truax, J. F., Vinegar, R. & Cuatrecasas, P. (1980) Biochem. Bio- New York), Vol. 3, pp. 105-112. phys. Res. Commun. 92, 688-695. 15. Isaacs, A. & Lindenmann, J. (1957) Proc. R. Soc. London Ser. B 26. Flower, R. J. & Blackwell, G. J. (1979) Nature (London) 278, 147, 258-273. 456-459. 16. Donahoe, R. M. & Huang, K. Y. (1976) Infect. Immun. 13, 27. Parks, W. P., Scolnick, E. M. & Kozikowski, E. H. (1974) Sci- 1250-1257. ence 184, 158-160. 17. Herberman, R. R., Ortaldo, J. R. & Bonnard, G. D. (1979) Na- 28. Shimizu, T., Nomiyama, S., Hirata, F. & Hayaishi, 0. (1978)J. ture (London) 277, 221-223. Biol Chem. 253, 4700-4706. 18. Gresser, I., Maury, C. & Brouty-Boy6, D. (1972) Nature (Lon- 29. Stringfellow, D. A. (1978) Science 201, 376-378. don) 239, 167-168. 30. Mundy, G. R. & Yoneda, T. (1980) in Biochemical Characteriza- 19. Kerr, I. M., Brown, R. E. & Hovanessian, A. G. (1977) Nature tion of Lymphokines, eds. de Weck, A. L., Kristensen, F. & (London) 268, 540-542. Landy, M. (Academic, New York), pp. 123-127. 20. Roberts, W. K., Clemens, M. J. & Kerr, I. M. (1976) Proc. Nati Acad. Sci. USA 73, 3136-3140. Downloaded by guest on September 25, 2021