Molecular Regulation of Tumor Necrosis Factor-a! - The Journal of ...

1 downloads 0 Views 2MB Size Report
Nicholas R. FerreriSQ, Timothy Sarrl, Philip W. AskenaseS, and Nancy H. Ruddlel. From the $Section of ..... were exposed at -70 "C to XAR-. 5 x-ray film (Kodak).
Vol. 267,, No. 13, Issue of May 6, pp. 94434449.1992 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMICITRY

Q 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Molecular Regulation of Tumor Necrosis Factor-a! and Lymphotoxin Production in T Cells INHIBITION BY PROSTAGLANDIN E2* (Received for publication, August 22, 1991)

Nicholas R. FerreriSQ,Timothy Sarrl, Philip W. AskenaseS, and Nancy H.Ruddlel From the $Section of Allergy and Clinical Immunology, Department of Medicine, and the llDepartment of Epidemiology and Public Health, Yale University School of Medicine, New Haven. Connecticut 06510

The effects of prostaglandin Ea (PGEa) and other eicosanoids were determinedon tumor necrosis factora (TNF-a) and lymphotoxin (LT; TNF-@)production by murine T h l , antigen-specific major histocompatibility complex class II-restricted T cell clones. T cells activated withimmobilized anti-CD3 or with soluble concanavalin A (ConA) produced approximately 1001,000 units/ml TNF-a/LT as determined by cytotoxicity against the WEHI-164 murine fibrosarcoma cell line. TNF-a/LT biological activity wasinduced rapidly with significant production evident approximately4 h after stimulation witheither anti-CD3 or ConA. PGEz inhibited the anti-CD3- and ConA-induced production of TNF-a/LT bioactivity in a dose-dependent manner. Incubation with 1% lo-’ M PGEz inhibited production of TNF-a/LT biological activity from anti-CD-activated F1.28, or 450A.1 T cell clones by approximately 60%. Incubation with 1X 10” M PGEa resulted in90% inhibition of biological activity. PGEa also inhibited both TNF-a and LT mRNA accumulation by more than 90%.These results suggest that thereduced production of cytotoxic activityinthe presence of PGEz was caused by the inhibition of both TNF-a and LT. No inhibition of T cell TNF-a or LT production was observed when anti-CD3- or ConA-activated cells were incubated with PGDa, PGFap, S-hydroxyeicosatetraenoic acid or leukotriene Cr. Nuclear run-on experiments indicated that the PGEa-mediated decrease of TNF-a and LT mRNA accumulation was caused, in part, by an inhibitory effect on the transcription of these genes. This is the first report of PGEa-mediated inhibition of TNF-a in T cells and PGEa-mediated inhibition of LT production in any cell type.

TNF-p) are genetically and functionally related, pleiotropic cytokines that participate in several types of immune reactions (1, 2). They compete for the same receptors (3), which are expressed on a variety of different cell types, and mediate a wide range of immunomodulatory activities including cytolytic/cytostatic effects on certain tumor cell lines (4), induction of MHC molecules (5),and augmentation of cellular proliferation and differentiation (6-8). TNF-a and LT are the only cytokines whose genes are known to map within the MHC (9-13). Molecular regulation of TNF-a and LTdiffers despite the similarities of these two cytokines regarding biological activity, receptor binding characteristics, and genetic organization. Both TNF-a and LT are produced by activated, but not resting, T h l CD4+, and CD8+ T cells (14, 15). However, only TNF-a but not LT is produced by macrophages activated by LPS or other stimuli(16, 17). TNF-a and LT areinduced in T cells after stimulationwith antigen presented in thecontext of class I or class I1 MHC molecules (18) or after activation by staphylococcal enterotoxin (19), T cell mitogens, and phorbo1 esters (20, 21). The kinetics of TNF-a and LT mRNA accumulation andprotein production differs (22,23), and their baseline mRNA accumulation differs as well (21, 23). There have been no studies describing the effects of eicosanoids on production of TNF-a in cells, such as T cells, which do not simultaneously produce PGE2. Furthermore, the effects of PGE2 on production of the related cytokine LT have not been analyzed in anycell type. In T cells, regulation of TNFa and LT production by PGE2 can be studied without the interference of endogenously produced prostanoids because T cells do not metabolize arachidonic acid via the cyclooxygenase pathway. PGEz is induced from macrophages by several types of Tumor necrosis factor-a (TNF-a)’ and lymphotoxin (LT; stimuli and accounts for most of the prostanoid production by cells of the immune system. Several immunosuppressive * This work wassupported by National Institutes of Health Grants effects have been observed for PGE2. For example, PGEz CA-16885 (to N. H. R.), AI-12211 and AI-26689 (to P. W. A.), and by a PMA Foundation research starter grant (to N. R. F.). The costs inhibits IL-1, IL-2, and granulocyte-macrophage colony-stimof publication of this article were defrayed in part by the payment of ulating factor production (24, 25), T cell cytotoxicity (26), page charges. This article must therefore be hereby marked “adver- natural killer and antibody-dependent cytotoxicity (27-29), tisement” in accordance with 18U.S.C. Section 1734 solelyto indicate and theresponsiveness of T cells to mitogens (30). However, this fact. PGE, can also increase the expression of interferon-y recep$ To whom correspondence and reprint requests should be sent: Dept. of Pharmacology, New York Medical College, Valhalla, NY tors on humanCD8+ lymphocytes and IgG2a Fc receptors on WEHI-3 cells (31, 32). In addition, PGE2 potentiates the 10595. Tel: 914-993-4115;Fax: 914-347-4956. ‘The abbreviations used are: TNF-a, tumor necrosis factor-a; immunoglobulin isotype switching effects of IL-4 on B cells MOPS, 4-morpholinepropanesulfonic acid; LT, lymphotoxin (tumor (33). Thus, the effects of PGE2 on immune function are not necrosis factor+); MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl- limited to suppressive activities and may vary according to tetrazolium bromide; MHC, major histocompatibility complex; LPS, lipopolysaccharide;PG, prostaglandin; IL, interleukin;ConA, concan- the cell types involved and interactions with other cytokines. avalin A; LTC,, leukotriene C,; 5-HETE, 5-hydroxyeicosatetraenoic Previously, PGE2 was shown to inhibit LPS-induced TNF-a acid; mAb, monoclonal antibody(ies); CTL, cytotoxic T lympho- gene transcription incomplete Freund’s adjuvant-elicited murine peritoneal macrophages (34) and 12-O-tetradecanoylcYte(s).

9443

9444

PGEz Inhibits

T Cell TNF-a and LT Production

phorbol-13-acetate-induced TNF-a gene expression in human HL-60 promyelocytic leukemia cells (35). In thepresent study we have shown that low doses of PGE2 inhibitthe production of TNF-a and LT from murine T h l T cell clones stimulated with immobilized anti-CD3 monoclonal antibody or soluble ConA. This is the first report to demonstrate inhibition by PGE2 of both TNF-a and LT in T cells and may partially explain the phenomenon of prostaglandin-mediated inhibition of T cell cytotoxicity.

ical School, Dallas) was excised from the BamHI and PstI sites of the vector pGEM4 (Promega Biotech, Madison, WI). A 0.71-kilobase KpnIlHincII fragment of the murine LT cDNA was used as the LT probe as described previously (39). Total RNA Zsohtion-Total cellular RNA was isolated by lysing the cells in guanidine isothiocyanate sodium citrate buffer and extracting RNA with phenol/chloroform as described previously (40). RNA samples were stored at -70 "C as ethanol precipitates. Northern Blot Analysis-10 pg of total RNA was electrophoresed in a 1%agarose/formaldehyde gel in 1X MOPS as therunning buffer. RNA was then transferred to a nylon membrane (Genescreen, Du Pont-New England Nuclear) and hybridized to random-primed 32PMATERIALS AND METHODS labeled probes (42) in a buffer containing 50% formamide, 10% Mice-Female SJL mice5-7 weeks old were obtained from the dextran sulfate, 0.2% polyvinylpyrrolidone, 0.2% Ficoll, 0.2% bovine Jackson Laboratory, Bar Harbor, ME. Mice were rested a t least 2-3 serum albumin, 1.0 M NaC1, 1.0% sodium dodecyl sulfate, 0.05 M Tris, weeks before use and were kept in filter-topped microisolator cages pH 7.5, and 0.1% sodium pyrophosphate at 42 "C for 18 h. Posthythat were changed in a laminarflow hood. bridization washes of the membrane were carried out in 2 X SSC, Reagents-Porcine myelin basic protein was obtained from Calbi- 1.0% sodium dodecyl sulfate at 65 "C for 1 h and in 0.1 X SSC for ochem. Concanavalin A (ConA), 3-[4,5-dimethylthiazol-2-yI]-2,5-di-25 "Cfor 1h. Then theprobed blots were exposed a t -70 "Cto XARphenyltetrazolium bromide (MTT), PGEz, PGF,,, PGD,, 5-hydrox- 5 x-ray film (Kodak). yeicosatetraenoic acid (5-HETE), and leukotriene C4 (LTC,) were Nuclear Run-on Transcription Assays-The transcriptional activpurchased from Sigma.A rat monoclonal antibody (mAb) specific for ity of the TNF-a and LT genes was determined in vitro by nuclear the CD3e chain (a gift from Dr. Kim Bottomly, Yale University, New run-on analysis (43). Briefly, F1.28 cells were added at a density of 1 Haven, CT), obtained as culture supernatant (36), was affinity puri- x lo6 cells/ml to tissue culture flasks coated with 100 ng/ml affinitytied using goat anti-rat IgG coupled to Sepharose 4B (Zymed Labo- purified rat anti-CD3 mAb. Cells were incubated for 4 h in the absence ratories, Inc., South San Francisco, CA). Recombinant human TNF- or presence ofPGE,. To isolate nuclei, cells (40-60 X lo6) were a was a kind gift from Drs. Y. Devries and J. Huang, Du Pont, washed twice in ice-cold phosphate-buffered saline and then resusGlenolden, PA. Recombinant murine TNF-a was a kind gift of Dr. pended in 10 ml of lysis buffer containing 10 mM Tris-HC1, pH 7.4, Regina Turetskaya, and recombinant LT was from Genentech, Inc., 2 mM MgCl,, 3 mM CaC12, 3 p~ dithiothreitol, 0.3 M sucrose, and South SanFrancisco). Guanidine isothiocyanate was purchased from 0.5% Triton X-100. The nuclei were incubated immediately after Fluka Chemie Ag (9470 Buchs, Switzerland). 3ZP-LabeledUTP was isolation for 30 min at 27 "C in transcription buffer containing 20 purchased from Amersham Corp. mM Tris-HC1, pH 8.3, 100 mMKC1,4.5mMMgCl,,0.4mM concenCell Lines-The myelin basic protein-specific (F1.28) (from Dr. tration each of ATP, CTP, and GTP,2 mM dithiothreitol, 20% (v/v) Robert Clark, University of Connecticut, Farmington) and ovalbu- glycerol, and 200pCiof [a-32P]UTP(>760 Ci/mmol; Amersham). min-specific (450A.1) murine CD4+ MHC class 11-restricted T h l T The reaction was stopped by adding 40 pg/ml DNase I, 10 mM Triscell clones (15) were maintained in Click's medium containing 10% HC1, pH 8.4, 100 mM NaCl, 1 mM EDTA, and 150 units/ml RNasin fetal calf serum (HyClone Laboratories, Logan, UT), 2 mM L-gluta- for 15 min a t 27 "C, and nuclear RNAwas isolated by phenol/ mine, penicillin (100 units/ml), streptomycin (100 pg/ml) (GIBCO), chloroform extractions. The extracted, 32P-labeledRNA was hybridand rat spleen cell-derived T cell growth factor added at a final ized to linearized plasmid DNA (immobilized on nylon membranes) concentration of 2% (v/v). T cell clones were stimulated every 4th that contained insertsfor TNF-a, LT,or murine cytoskeletal (3-actin. day with T cell growth factor, and given antigen, irradiatedsyngeneic spleen cells, and T cell growth factor once every other week. The cells RESULTS were used for experiments 9 days after theywere last fed with antigen and presenting cells. A clone of the murine fibrosarcoma cell line, Kinetics of Anti-CD3-induced TNF-a/LT Production by T WEHI-164 (37), which is sensitive to TNF-a and LT, was maintained Cell Clone F1.28--In preliminary experiments we determined in RPMI 1640 (GIBCO), and supplemented with 10% fetalcalf serum that 100 ng/ml immobilized purified anti-CD3 mAb induced (GIBCO), 2 mM L-glutamine, penicillin, and streptomycin (GIBCO). Anti-CD3 ana' ConA Activation of F1.28-Tissue culture plates(24 maximal production of TNF-a/LT biological activity from well) or flasks (T75) were coated overnight at 4 'C with 100 ng/ml the T cell clones F1.28 and 450A.1. This was calculated as affinity-purified rat anti-CD3. The plates orflasks were washed twice cytotoxicity against the WEHI-164 murine fibrosarcoma cell with sterile phosphate-buffered saline, and F1.28 cells were added at line. Neither this bioassay nor the available monoclonal ana density of 1 X lo6 cells/ml. Cells were exposed to the immobilized tibody for TNF-a which we developed can distinguish between anti-CD3 antibody or to soluble ConA (5 pg/ml), in the absence or the two cytokines (41). Therefore, the results are expressed presence of eicosanoids for various times, after which the supernatants were harvested, serially diluted, and tested for TNF-a/LT as TNF-a/LT activity. The production of TNF-a and LT is cytotoxicity using the WEHI-164 indicator cell line. The T cells were dependent on cellular activation because unstimulated T cells harvested for Northern blot analysis of steady-state mRNA levels or did not produce any biologically active TNF-a/LT (Fig. 1). nuclear run-on experiments to assess the effects of eicosanoids on The kinetics of TNF-a/LT production by F1.28 cells stimuTNF-a and LTgene transcription. lated with immobilized anti-CD3 was determined at several Bioassay for TNF-a and LT-TNF-a and LT bioactivity was time points. Thus, cells that were stimulated via their T cell measured using the WEHI-164 fibrosarcoma cell line as described previously (37). This cell line is sensitive and specific for cytotoxic receptor with anti-CD3 antibody produced significant activity in F1.28 supernatants thatcan be inhibited by a monoclonal amounts of TNF-a/LT 4 h and8 h after stimulation(Fig. 1). antibody (TN3.19.12) that recognizes both TNF-a and LT (37, 41). Interestingly, there was a second phase of production after Serial dilutions of cell-free supernatants containing TNF-a and LT the plateau at 4-8 h because cytotoxic activity accumulated were incubated with 5 X lo3 WEHI-64 cells in 96-well microtiter to a maximum by 24 h. plates for 48 h at 37 "C in 5% COZ. At the end of the incubation Inhibition of Anti-CD3-induced TNF-a/LT Production by period the viability of the cells was determined by assessing their metabolic capacity (38). MTT was added at 5 mg/ml, and plates were PGE2-It has been demonstrated previously that PGEz can incubated for 4 hat 37 "C in 5% COZ. Acidified isopropyl alcohol (100 inhibit LPS-induced TNF-a production by peritoneal macpl of 0.04 N HC1 in isopropyl alcohol) was added to each well, and the rophages (34). Because the murine T h l T cell clone F1.28 can plates were read using a test wavelength of 570 nm and a reference be induced to release TNF-a and LT (15), we tested the wavelength of 630 nm. One unit of bioactivity is defined as the effects of PGEz andother eicosanoids on anti-CD3-induction supernatant dilution a t which 50% cytotoxicity was observed. Recom- of TNF-a and LT production. In addition, similar experibinant TNF-a, LT, or a supernatant from a T cell clone known to ments were carried out with the ovalbumin-specific MHC produce these cytokines was included as a control. cDNA Probes-A 1.4-kilobase murine TNF-a cDNA probe (a kind class 11-restricted T h l T cell clone 450A.1. Addition of PGE2 gift from Dr. Bruce Beutler, University of Texas Southwestern Med- but not PGF2,, PGD2, or leukotriene C4 inhibited anti-CD3

PGE2 Inhibits T Cell TNF-a and LT Production induction of TNF-a/LT cytotoxicity against WEHI-164 fibrosarcoma cells (Figs. 2 and 3). The effects of PGE, were apparent at concentrations as low as 1 nM (Fig. 3), a dose that is physiologically relevant with respect to concentrations observed at inflammatory sites. These resultswere not caused by a direct cytotoxic effect of PGE, on the WEHI-164 target 100

r

cells alone

anti-CD3

"D- 4 h r

+

3 .0 ._ x

-B

?-

24hr

"t

U

.

0

5

8hr

U

.

-

60

-

40

I

s

8

20

0 I

10

100

1000

10000

100000

lldilution of supernatant

FIG. 1. Kinetics of anti-CD3-induced TNF-a/LT production T cell clone F1.28. F1.28 T cells were incubated at 37 "Cin 5% COz forthe indicated times inthe absence or presence of immobilized anti-CD3-purified mAb. Supernatants were harvested and assayed for TNF-a/LT bioactivity using the WEHI-164 fibrosarcoma indicator cell line as indicated under "Materials and Methods." These results are representative of three similar experiments.

by

-0-

cell* alone anti-CD3

& anli-CDB r P G E , -7M

4 anti-CD3 + PGF,,-

10

100 10000 1000

7M

100000

lldilution of Supernatant

FIG. 2. Inhibition of anti-CD3-induced TNF-a/LT production by PGE,. F1.28 T cells were stimulated with immobilized antiCD3 mAb in the absence or presence of either PGEz or PGF2, (10" M). After 24 h, supernatants were harvested, and TNF-a/LT bioactivity was determined. This figure is representative of five similar experiments.

cell line inasmuch as theaddition of PGE2 alone did not alter the ability of WEHI-164 cells to metabolize MTT. Inaddition, PGE2 was not toxic to the T cell clones because more than 95% of the cells excluded trypan blue after treatment with PGE,. Furthermore, transcription of the 8-actin gene in these cells was not affected by PGE, (see below, Fig. 7). Finally, PGE, did not interfere with the cytotoxic effects of recombinant human TNF-a added directly to WEHI-164 cells. Thus, PGE, appears to prevent the production of TNF-a/LT by a direct effect on the T cell. PGE2 Inhibits TNF-a/LT Production but Does Not Alter the Kinetics of Production-We determined the effects of PGE2 on anti-CD3-induced production of TNF-a/LT from F1.28 T cells at various times after stimulationto rule out the possibility that the inhibitory effects of PGE, were merely caused by altered kinetics of TNF-a/LT production. As shown in Figs. 1and 4, production of TNF-a/LT was relatively rapid after exposure to anti-CD3, with near maximal biological activity by approximately 4-8 h. A second phase of TNF-a/ LT bioactivity was also observed between 8 and 24 h. The effects of PGE2 were independent of the rate or extent of TNF-a/LT production at any time point because addition of PGE2 completely inhibited the anti-CD3-induced production of TNF-a/LT at each of the time points measured (Fig. 4). This suggests that the inhibition exerted by PGE2 occurred rapidly and was sustained throughout the incubation period. In addition, the amount of TNF-a/LT produced in the presence of PGE, never approached the levels measured in the absence of PGE,. This further suggests that the inhibitory effects of PGE2 were not caused by a delayed production of TNF-a/LT. PGE,-mediated Inhibition of Mitogen-induced TNF-a/LT Production-Several different signals induce the production of TNF-a/LT from T cells. Thus, we also studied the effects of PGEz on TNF-a/LT production by 450A.1 T cells stimulated with the mitogen, ConA. Stimulation of450A.1 cells with ConA induced TNF-a/LT production that was similar, with respect to magnitude and kinetics, to that observed for anti-CD3-stimulated 450A.1 and F1.28 T cells. The levels of biologically active TNF-a/LT ranged from 100 to 1,000 units/ ml for 450A.1 and F1.28 T cells stimulated with either antiCD3 or ConA. The ConA-induced TNF-a/LT production was inhibitable by PGEz but not byPGF,,,PGD,, or the C5-

100 7

FIG.3. PGEa inhibition of antiCD3-induced TNF-a/LT production b y T cell clone 460A.1: dose response. 450A.1 T cells were stimulated with immobilized anti-CD3 mAb in the absence or presence of various doses of several eicosanoids. Supernatants were harvested at the end of the 24-h incubation period, and TNF-a/LT bioactivity was determined. The data are a representative figurefrom three separate experiments.

9445

100 _I

cells alone anli.CD3 PGEP -10M PGEZ -9M PGEZ -8M PGEP -7M

1

10

100 1000 10000 100000 IldllUtlon of aupernaard

cells abne anIiCO3 PGDP -10M PGDP -9M PGOP BM PGDZ -7M

I

10

100 1000 10000 1ldlluIlOn01 aupernalard

cells abne anti-CD3 PGFP -1oM PGFP -9M PGFP BM PGFP .7M

100 1000 10000 lldllullon 01 Supernatant

10

cells abne anli-CDJ LTCI -1OM LTCI -9M LTC4 -8M LTC4 -7u

100 1000 10000 1ldllutlnn of supernatant

10

PGEz Inhibits T Cell TNF-a and LT Production

9446

lipoxygenase metabolites LTC, (Fig. 5) and 5-HETE (data not shown). Thus, theinhibitory effects of PGE2were demonstrable, to a similar extent, in cells that were activated by ConA or by anti-CD3. In addition, both ConA- and anti-CD3activated cells were susceptible to inhibition of TNF-a/LT production only by PGE2, but notby other eicosanoids. Time-dependentInhibitory Effectof PGE2on ConA-induced TNF-a/LT Production-The induction of TNF-a/LT production by anti-CD3 or ConA had two apparent phases. The first was rapid and plateaued by approximately 8 h. The second proceeded more slowly between8 and 24 h. Thus, we determined how long after the initial stimulation of F1.28 cells PGE2 could be added and still inhibit TNF-a/LT production. As described above, PGE2 inhibited the ConA-induced TNF-a/LT production when added concurrently with ConA (addition at time zero). In addition, PGE2 was just as effective when added 1, 2, 4, and 6 h after the initial stimulation with ConA because greater than 97% inhibition was achieved at each of these time points (Table I). However, only 50%inhibition was observed whenPGE2was added 20 h after the addition of ConA. Thus, PGE2-mediated inhibition was greater than 95% when added up to 6 h after stimulation because both the initial and later phases of TNF-a/LT production were inhibited. However, whenPGE2was added 20 h after stimulation, there was no longerinhibition of the initial

F

phase and only partial inhibition of the later phase of TNFa/LT production. These data also suggest a possible posttranscriptional inhibitory effect forPGE2. Inhibition of F1.28 TNF-a and LT mRNA Accumulation by PGE2-Northern blot analyses were performed to determine whether the inhibitory effect of PGE2 on T cell TNF-a/LT production was causedby a decrease in mRNA accumulation. These experiments were also done to assess whether PGE2 could inhibit the production of both TNF-a and LT inasmuch asthe bioassaydoes not discriminate between these two cytokines. As noted above, the WEHI-164 cell line is sensitive to killing by both TNF-a andLT, andantibodies that distinguish the two cytokines are not available. Thus, Northern blot analyses were performed to determine whether both TNF-a and LTmRNA were inhibited. As shown in Fig. 6, A and B, ConA induced the production of both TNF-a and LT mRNA in F1.28 cells, and PGE2 inhibited the accumulation of mRNA for both cytokines. Similar results were obtained when cells were activated with anti-CD3. These data suggest that the PGE2-mediateddecrease of biologically active cytokines, that are cytotoxic to the WEHI-164 fibrosarcoma cell line, is caused by inhibition of both TNF-a and LT. Furthermore, the decrease in TNF-a and LT production results from a decrease in mRNA accumulation for these cytokines. Effects of PGE, on TNF-a and LT Gene TramcriptionNuclear run-on experiments were performed to determine whether the PGE2-mediatedinhibition of mRNA accumulation reflected an inhibition of transcription of the TNF-aand LT genes. F1.28 cells wereincubated in the absence or presence of anti-CD3 mAb for 4 h after which the supernatants were harvested to assess biological activity while nuclei were isolated and labeled with [321UTPfor in vitro transcription analyses. As shown in Table 11, constitutive transcription of the TNF-a gene in the absence of any stimulation was high. It was approximately 12 times greater than for the LT gene in F1.28 T cells. However, this constitutive transcription did not result in thesecretion of biologically active TNF-a or LT. Incubation ofF1.28 cells with anti-CD3 induced approximately an 8- and 12-fold increase in the transcription of the TNF-a and LT genes, respectively (Table 11). Transcription of the &actin genewasminimallyaffected. Thus, the increases in transcription rates for TNF-a and LT are similar after activation with anti-CD3. PGE2 inhibited the anti-CD3-

2-3mo 0

5

10

15

20

Hours

FIG. 4. PGE2 inhibits TNF-a/LT productionbut does not affect the kinetics of production. F1.28 T cells were stimulated with immobilized anti-CD3 mAb in the absence or presence of PGE, (10" M). Supernatants were harvested at theindicated times. Similar results were obtained in two other experiments.

cells abne Con A 5 w m l PGEZ -10 M PGE2 -9 M PGEZ -8 M PGEP -7 M

FIG.5. PGE2 inhibition of ConAinduced TNF-a/LT production by T cell clone 450A.1: dose response. 450A.1T cells were stimulated with soluble ConA (5 pg/ml) in the absence or presence of various eicosanoids. Supernatants were harvested at the end of the 24-h incubation period, and TNF-a/LT cytotoxicity was measured using the WEHI-164 indicator cell line as described above. Representative figure from four separate experiments.

1

10

100 100000 10000 1000

-e-

cellsabne Con A 5 p g m l

-f-

PGF2 -10M

-c

PGFZ -9M FGF2 B M PGFZ .7M

1

lldllutbn of wpsrnaant

-c

"-c

10

10000 1000 100

10

100

100000

cells a b e ConAlwrd P G D Z -10 M PGM-9 M PGDZB M PGDZ -7 M

i

# 1

10

100

1000

lldllutbn of wpernatant

10000 100000

1

10000 1000

lldllutbn of supernatant

100000

PGE,Inhibits T Cell TNF-a and LT Production

9447

TABLE I Time-dependent inhibitory effectof PGE, on TNF-a and LT production from activatedT cells F1.28 cells were incubated with ConA (5 pg/ml) for 24 h. PC& (1 X M) wasadded to the cells at the indicated times after the addition of ConA, and the cells were cultured until a total time of 24 h had elapsed. Supernatants were harvested and assayed forTNF-a/ LT bioactivity. Time Of PGEz addition

Unita of TNF-a/LT

% Inhibition

4 0 10 20 20 20 320

>98 98 97 97 97 50

none

h

0 1 2 4 6 20

P

anti.CD3

.. *

PGE,(lO'M)

1

A

S

e

0

9