Eur. J. Immunol. 1990. 20: 2439-2443

Anders Waageo, Geir SlupphaugO and Refaat ShalabyO Institute of Cancer Research, University of Trondheimo, Center for Molecular Biologyo, UNIGEN, Trondheim

Glucocorticoids inhibit IL 6 production

2439

Glucocorticoids inhibit the production of IL 6 from monocytes, endothelial cells and fibroblasts* We have examined the effect of dexamethasone (DM) and cortisol on the production of interleukin (IL)6 from the murine macrophage cell line RAW 264.9, human monocytes, human endothelial cells and the human fibroblast cell line FS4. In RAW 264.9 cells DM in the concentration range lop9M to M inhibited the lipopolysaccharide (LPS)-induced production of IL 6 by 10% to 90%. Cortisol had a similar effect, but was about 25 times less potent than DM. Also, when M of DM was added to the cultures after addition of LPS, it completely inhibited the residual 24-h production of IL6. Corresponding to the effect on I L 6 production, DM (lop6M) reduced the mRNA levels for IL 6 in the RAW 264.9 cells. The glucocorticoid analogue RU 486 competes with DM and cortisol for the glucocorticoid receptor and reversed the inhibitory effect of DM, demonstrating that DM exerts its effect via the glucocorticoid receptor. DM also had an inhibitory effect on LPS-stimulated IL 6 production in freshly isolated human monocytes, and on IL 1-stimulated I L 6 production in human endothelial cells and FS4 fibroblasts. These results demonstrate that DM via a receptormediated mechanism inhibits IL 6 production at the transcriptional level, and this may contribute to the anti-inflammatory and immunosuppressive effect of glucocorticoids.

1 Introduction The pleiotropic mediator I L 6 (previously called B cell stimulatory factor 2, IFN42, 26-kDa protein and hepatocyte-stimulating factor) is produced by a number of cells including monocytes [1], fibroblasts [2], endothelial cells [3] and T lymphocytes [4]. The regulation of the expression of IL6 has been studied in many of these cells. For instance different stimuli like LPS [1, 51, IL 1 [5], IFN-y [5], HIV [6] and adhesion to plastic surfaces [5] induce transcription of the gene for IL6 in monocytes. In endothelial cells [3] and fibroblasts [2],TNF, IL 1and LPS can induce the production of IL 6, and inT lymphocytes [4],TNFand IL 1are stimuli of IL 6 production. While a number of positive stimuli have been identified, the negative regulation of I L 6 expression has not been studied to the same extent. However, it is well known that glucocorticoids have an inhibitory effect on the production or the cytokines TNF and IL 1. In monocytes, dexamethasone (DM) decreases the levels of mRNA for TNF [7] and IL1 [8], and the release of the respective proteins [7-101. The present study was undertaken to investigate the effect of steroids on the production of IL6. We demonstrate herein that the murine M a cell line RAW 264.9persistently produces IL6 upon stimulation by bacterial LPS, and that DM by a receptor-mediated mechanism reduces the mRNA levels for I L 6 and production of the protein. [I 84951

* This study was supported by the Norwegian Research Council for Science and Humanities.

Correspondence: Anders Waage, Institute of Cancer Research, University of Trondheim, N-7006 Trondheim, Norway Abbreviations: DM: Dexamethasone HEC: Human endothelial cells 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990

Furthermore, DM inhibits the production of IL 6 from human endothelial cells (HEC) and fibroblasts stimulated with IL 1.

2 Materials and methods 2.1 Reagents Escherichiu coli (strain 0 2 6 :B6)-derived LPS was purchased from Sigma (St. Louis, MO), dissolved in PBS and stored at -20 "C until use. Steroids (Steraloids Inc. ,Wilton, NH)were dissolved at a concentration of lo-' M ethanol, and diluted further in RPMI 1640 to appropriate concentrations. Human rTNF (generously provided by Genentech Inc., South San Francisco, CA) had a sp. act. of 4 x lo7 U/mg as determined by a cytotoxicity assay [11]. Human rIL1P [12] was a gift from Dr. A. Shaw, Glaxo Institute for Molecular Biology, Geneva, Switzerland, and had a sp. act. of 7.6 x lo7 U/mg as determined in the thymocyte proliferation assay. An antiserum to human rIL6 (generously provided by Dr. W. Fiers, University of Gent, Belgium) with neutralizing capacity of approximately 400 ng/ml antiserum, and an mAb to murine IL6 (SN from the 6B4 hybridoma generously supplied by Dr. Van Snick, Ludwig Institute, Brussels, Belgium) were used to confirm the specificity of the IL6 activity in test SN.

2.2 Cells The murine M@ cell line RAW 264.9 was maintained in complete medium consisting of RPMI 1640(Gibco, Paisley, Scotland) with 10% FCS (Sigma), 2 mmolA L-glutamine and 30 pg/ml of gentamycin. For measurement of IL6 production the cells were trypsinized (0.25% trypsin-0.02% EDTA), washed once, and 200 pl cell suspension was seeded in each well (5 x 104 cells/well) in a 96-well microtiter plate (Costar, Cambridge, MA). 0014-2980/90/1111-2439$3.50+.25/0

2440

Eur. J. Immunol. 1990. 20: 2439-2443

A. Waage, G. Slupphaug and R. Shalaby

HEC were obtained by collagenase (Sigma) treatment of human umbilical cord veins as described [13]. The identity of HEC was confirmed as detailed in a previous publication [3]. The cells were seeded at a uniform density (1.5 X lo5 cells/well) in each well in 24-well plates (Costar). Diploid FS4 fibroblasts (kindly provided by Dr. J.Vilcek, New York School of Medicine, New York) were maintained in complete medium. On the day of the experiment the cells were trypsinized, and 200 p1 of the cell suspension was seeded in each well (lo4 cells/well) in a 96-well microtiter plate.

and murine cells, confirming the specificityof the activities. In control experiments we discovered that concentrations of DM > lop9M reduced the sensitivity of the B9 cells to IL6 in a dose-dependent manner. This problem was resolved by cloning of the B9 cells and growth of a subclone which was resistant to the effect of DM. Samples in which carry-over of DM might interfere with the IL6 measurements were assayed with this subclone. SD in the triplicate measurements was < 10%.

Cells were incubated with the IL6-inducing agent for 24 h unless otherwise indicated. Experiments with RAW 264.9 cells, FS4 cells and human monocytes were performed in triplicate or quadruplicate cultures in 6-mm wells. Experiments with HEC were carried out in duplicate 12-mmwells. At the end of each experiment the plates were centrifuged, and SN were harvested and pooled, and stored at -20°C until they were assayed for IL6 activity.

3 Results

2.3 Northern blot analysis For Northern blot experiments, 0.3 x lo6 cells/ml were seeded in T-75 cm2 flasks (Costar). After 3 days, when the density of the cells was appropriate, LPS and DM were added and the cells were incubated for a further 8 h. The flasks were then placed on ice, the cells washed twice, scraped off using a rubber policeman, gently pelleted by centrifugation and resuspended in preheated (55 "C) lysis buffer (1% SDS, 10 mM EDTA). After vortexing, sodium acetate was added to a final concentration of 0.3 M (pH 5 ) , and the lysates were extracted once with preheated (55 "C) phenolkhloroform (2/1). Following ethanol precipitation and drying, the samples were dissolved in RNAse-free water and quantified spectrophotometrically. Samples were electrophoresed in agarose-formaldehyde gels and transferred to nylon filters (Hybond N; Amersham Int., Amersham, GB) according to Fourney et al. [14]. Standardization was done with respect to 28 S and 18S RNA as well as by control hybridizations with glyceraldehyde-3-phosphate dehydrogenase cDNA. Hybridizations were performed with 32P-labeledcDNA for murine I L 6 [15] (kindly provided by Dr. J. Van Snick) according to the Amersham Hybond protocol and filters were exposed to HyperfilmMP (Amersham) at -70°C.

2.4 Assay for IL6 IL6 in the SN was measured in a bioassay based on the IL 6-dependent hybridoma cell line B9 (kindly provided by Dr. L. Aarden, University of Amsterdam) as previously described [16]. In short, triplicate test samples were fivefold serially diluted in 96-well microtiter plates (100 pl/well), and 100 pl of a suspension of B9 cells (5 x lo3cells/well)was added using medium consisting of RPMI 1640 (Gibco), 5% FCS (Gibco), 2 mmol/l L-glutamine, 30 pg/ml gentamycin and 50 pmolA 2-ME. Dilutions of human rIL6 were included as a standard [17] (provided by Dr. L. Aarden). After 72 h, cell growth was measured by adding a tetrazolium salt, as described [18].The lower detection limit of the assay was about 0.5 pg of rIL6/ml.The antibodies to human and rnurine IL 6 completely neutralized the growth-stimulatory effect of the culture SN from, respectively, human

3.1 IL 6 production by RAW 264.9 We first established the kinetics of IL6 production from RAW264.9 cells by measuring the IL6 concentrations in the SN at various time points up to 26 h after addition of LPS (Fig. 1). When cells in exponential growth were seeded, IL6 was detected in the SN 4 h after addition of LPS. The IL6 concentration in the SN gradually increased during at least 26 h. There was variation in the total production of I L 6 from experiment to experiment; however, the kinetics of the production was similar. To examine the requirement for LPS for the sustained production of IL 6, we replaced the medium/LPS after 7 h in some experiments (Fig. 2). In the wells receiving fresh medium without LPS, the IL 6 production rapidly decreased, whereas the wells receiving fresh medium/LPS continued to produce IL6 at the same rate after a short intermittent decrease. The results demonstrate that persistent IL 6 production by the RAW 264.9 cells is dependent on the continual presence of LPS.

3.2 Effect of glucocorticoids We next examined the effect of adding cortisol and the glucocorticoid analogue DM to the cultures together with

cd d

1000-

500 -

0-

Eur. J. Immunol. 1990. 20: 2439-2443

Glucocorticoids inhibit I L 6 production

2441

LPS. Both steroids at concentrations ranging from lop5to M inhibited the production of IL6 (Fig. 3). However, DM was about 25 times more potent than cortisol, as estimated by the concentration giving 50% reduction of IL 6 production. DM at a concentration of was also added to the cultures at various time intervals after LPS addition (Fig. 4). These results demonstrate that DM added to the cultures at any time point almost completely inhibited the residual 26-h production of IL6. Glucocorticoids may be toxic to lymphocytes; however, they did not affect the viability of the monocytes, fibro-

-jj

1000

-

800

-

600

-

k

a

hours Figure4. Effect of adding DM after addition of LPS. LPS (30 pg/rnl) was added to all the cultures at time 0 and DM M) was added after LPS at the time indicated along the abscissa. All SN were harvested after 26 h. Control cultures were incubated with LPS for 26 h and DM was added immediately before harvesting. IL 6 concentrations in these cultures were taken as 100%. Each bar represents mean of triplicate determinations of pooled SN. Results from one experiments are presented. Similar results were obtained in four experiments.

blasts or endothelial cells as estimated by trypan blue exclusion. The inhibition of IL6 synthesis can thus not be explained by cell death. Figure 2. Requirement for presence of LPS for continual production of IL6. All cultures received medium and LPS (30 pg/ml) at the beginning of the experiment. After 7 h the medium was removed and the cultures were washed once. Half of the cultures received LPS (A),whereas the other half received medium only (W). SN from triplicate cultures were harvested at the time indicated. Experimental conditions and data presentation are as described in the legend t o Fig. 1. Similar results were obtained in three experiments.

1 2

2a cd d

I5O0-

3.3 Northern blot analysis

To study the mechanism of inhibition we measured the levels of mRNA for I L 6 in unstimulated cultures and in cultures stimulated with LPS and LPSDM. In preliminary experiments we found that 8 h of LPS stimulation regularly produced a high mRNA level for I L 6 and therefore this time point was used in further experiments. Cultures incubated with LPSDM yielded no or only a faint signal for IL 6 mRNA whereas cultures incubated with medium only had a signal which was higher, but still much lower than that for LPS-stimulated cultures (Fig. 5a). mRNA for glyceraldehyde-3-phosphate dehydrogenase was used as internal control and demonstrates that the effect of DM was specific for IL6 mRNA (Fig. 5b). These experiments indicate that I L 6 production is inhibited at the level of gene transcription.

1°00-

500

-

3.4 Effect of glucocorticoid antagonist

The synthetic steroid analogue RU 486 competes with DM for binding t o the cytosolic receptor for glucocorticoids 0[19].We conducted five experiments with combinations of -‘to 29 18 17 26 25 A4 DM and RU486, and in all of them the glucocorticoid log concentration, M antagonist reversed the effect of DM. Fig. 6 shows that increasing concentrations of RU 486 block the effect of Figure 3. Dose-dependent suppression of I L 6 production by glucocorticoids. Various concentrations of DM (W) or cortisol (A) DM, consistent with a competitive mechanism of receptor were added together with LPS (30 pglml), and SN were harvested binding for the two steroids.The effect of RU 486 alone was after 24 h. Experimental conditions and data presentation are as insignificantly different from control. These results demondescribed in the legend to Fig. 1. Similar results were obtained in strate that the effect of DM on the I L 6 production is six experiments. mediated by specific binding of DM to its receptor.

3.5 HEC, FS4 fibroblasts and human monocytes Cells were cultured as described in Sect. 2.2 and stimulated with IL 1which is a potent stimulator of I L 6 production by both HEC [3] and FS4 fibroblasts [2].The effect of DM and cortisol on HEC, and DM and RU486 on FS4 fibroblasts M reduced the I L 6 was examined (Table 1). DM at production by 53% and 74% from, respectively, HEC and FS4 fibroblasts, and also cortisol had an inhibitory effect on HEC (42%). RU486 antagonized the effect of DM on FS4 fibroblasts. The results demonstrate that glucocorticoids

2

3

1

2

3

Kb

- 2.4 -1.4

'

11.4 6.6 7.6 0.7

17.4 4.5 8.6 0.1

19.1 10.4 ND 0.1

88.1 11.5

50.8 12.2

35.3 14.1

60.2 1.3

39.3 2.1

ND 1.9

53 f 17 42

74f 14 28

We also examined freshly isolated human monocytes and found that DM and cortisol inhibited the LPS-stimulated IL6 production in a similar way as in the murine M@ cell line (data not shown).

4 Discussion The kinetics of I L 6 production from the murine M@ cell line RAW264.9 was characterized by a constant rate of release of bioactive IL6 between 4 and 26 h after addition of LPS and this production was dependent on the continual presence of LPS in the medium. When lop6M of DM was added to the cultures within this time, the residual 26 h production of I L 6 was almost completely inhibited. DM can thus abrogate an already ongoing production of IL6.

300

a 6 =!

+

Inhibition (% 1

inhibit the IL 1-induced production of IL6 from HEC and fibroblasts, and at least the effect on FS4 fibroblasts is mediated via the glucocorticoid receptor.

Figure 5. (A) Effect of LPS and DM on IL 6 gene transcription. Total cellular RNAwas extracted from RAW 264.9 cells which were incubated for 8 h in medium only (lane l ) , in medium plus LPS (30 pg/ml, lane 2) and in medium plus LPS/DM M; lane 3). Filters were hybridized with murine I L 6 cDNA, and the duration of autoradiography was 10 days. Sp. act. of the I L 6 probe was 2.7 X lo9 cpm/pg. (B) Control hybridization of the same filter with human glyceraldehyde-3-phosphatedehydrogenase cDNA. The duration of autoradiography was 2 days. Sp. act. of the probe was 2.4 x loy cpm/pg. Total RNA loaded per lane was 35 pg (both A and B).

M

HEC"' IL 1pb' IL 10 DMb) ILlp+cortisolb) None

IL 6 (ng/ml) Exp. 1 Exp. 2 Exp. 3

a) FS4 fibroblasts and freshly isolated HEC were incubated for 24 h with various agents as indicated. The IL6 concentration was measured in pooled SN from duplicate (HEC) and triplicate cultures. b) For HEC experiments. IL 1s dose was 13 pg/ml, for fibroblast experiments, I L l P dose was 500pg/ml. In all cases DM. cortisol and RU 486 concentrations were M.

-4.4

\

Agents added

+

-9.5 - 7.5

400

Table 1. Effect of glucocorticoids on IL6 production by HEC and FS4 fibroblasts

FS4 fibroblasts") IL 1p IL 1fi DM IL l p + DM fRU486") None

(B)

(A) 1

Eur. J. Immunol. 1990. 20: 2439-2443

A. Waage, G. Slupphaug and R. Shalaby

2442

200

100

-10

-9

-8

-7

-6

-5

-4

log concentration, M

Figure 6. Reversal of DM effect by RU486. DM M) and various concentrations of RU 486, as indicated along the abscissa, were added to the cultures together with LPS (30 pglml). SN were harvested after 24 h. Experimental conditions and data presentation are as described in legend to Fig. 1. Similar results were obtained in five experiments.

The receptor-mediated effect of DM and the inhibitory effect on IL6 at the level of gene transcription, as demonstrated in the present study, indicate that D M exerts its effect on the RAW264.9 cells according to a well established glucocorticoid mechanism, i.e. binding to the cytoplasmatic glucocorticoid receptor, translocation of the receptor-ligand complex to the nucleus and an effect of this complex on the gene transcription [20]. This mechanism of action is further supported by the demonstration of a glucocorticoid-responsiveelement located in the IL 6 gene [21]. This mechanism appears to be unrelated t o the IL 6-inducing stimulus as DM also inhibited IL 1-induced IL 6 production in FS4 fibroblasts and HEC. Man is a rather steroid-resistant species as compared with mouse [22], thus it is important to study the glucocorticoid effect on cells of human origin in addition to the murine cell

Eur. J. Imrnunol. 1990. 20: 2439-2443

line. Our study shows that the inhibitory effect of DM o n IL 6 production also holds for human monocytes as well as for HEC and fibroblasts, and indicates that DM is profoundly involved in the regulation of IL6 production in the organism.The glucocorticoids suppressed the production of IL6 by murine and human MWmonocytes in a dosedependent manner, and DM was about 25 times more potent than cortisol. This potency ratio is in accordance with the effect of DM and cortisol on the TNF production [9], and with results from other systems [23]. On background of the diversity of biological effects demonstrated for IL6 [24], the inhibitory effect of steroids on the IL 6 production has important implications. IL 6 promotes growth and differentiation of B cells [25], plasma cells [26] and Tcells [27]. Furthermore, IL 6 induces the production of acute-phase proteins by hepatocytes [28] and induces fever in the rabbit [29]. High levels of IL6 can be detected in the circulation and in the cerebrospinal fluid in patients with meningococcal septic shock [30] and meningitis [31], respectively. Clearly, the marked suppressive effect on IL 6 production from various cell types and stimuli provides an additional mechanism by which glucocorticoids exert their immunosuppressive and anti-inflammatory effects. In conclusion, this study demonstrates that DM via binding to the glucocorticoid receptor inhibits the IL 6 production at the transcriptional level, and that the inhibitory effect is operative under a variety of experimental conditions including LPS-stimulated monocytes and IL 1-stimulated HEC and fibroblasts. The results provide additional evidence for the negative regulatory effect of the glucocorticoid hormones on the cytokine system [7-10, 32, 331. We thank M . S~rensenand B. St@rdalfor technical assistance, and D. Moholdt for typing the manuscript. Received April 26, 1990; in revised form July 9, 1990.

5 References Bauer, J . , Canter, U., Geiger, T., Jacobshagen. U., Hirano,T., Matsuda,T., Kishimoto,T., Andus,T., Acs, G., Gerok,W. and Ciliberto. G., Blood 1988. 72: 1134. Kohase, M., Henriksen-De Stefano, D., May, L. T.,Vilcek, J. and Sehgal, P. B., Cell 1986. 45: 659. Shalaby, M. R., Waage, A. and Espevik, T., Cell. Immunol. 1989. 121:372. Espevik, T., Waage, A , , Faxvaag, A. and Shalaby, R., Cell. Immunol. 1990. 126: 47. Navarro. S., Debili, N., Bernaudin, J.-E,Vainchenker, W. and Doly, J., J. Immunol. 1989. 142: 4339. Nakajima, K., Martinez-Maza. O., Hirano,T., Nisharnian, P., Salazar-Gonzalez, J. F. and Kishirnoto, T., J. Immunol. 1989. 142: 531.

Glucocorticoids inhibit IL 6 production

2443

7 Beutler, B., Krochin, N.. Milsark, 1. W.. Luedke, C. and Cerami. A., Science 1986. 232: 977. 8 Lew,W.. Oppenheim, J. J. and Matsushima, K.. J. Immunol. 1988. 140: 1895. 9 Waage, A. and Bakke, O., Immunology 1988. 63: 299. 10 Snyder, D. S. and Unanue, E. R., J. Immunol. 1982. 129: 1803. 11 Kramer, S. M. and Carver, M. E., J. Immunol. Merhotfs 1986. 93: 210. 12 Wingfield, P., Payton, M. ,Tavernier, J., Bornes, M . , Shaw, A , , Rose, K., Simora, G., Demczuk, S.,Williamson, K. and Dayer. J.-M., Eur. J. Biochem. 1986. 160: 491. 13 Wall, R.T., Harker, L. A., Quadrucchi, L. J. and Striker. G. E.. J. Cell. Physiol. 1978. 96: 203. 14 Fourney, R. M., Miyakoshi, J., Day 111, R. S., and Paterson. M. C., Focus (Gibco BRL) 1988. 10: 5. 15 Van Snick, J., Cayphas, S., Szikora, J.-P.. Renauld. J.-C.,Van Roost, E., Boon,T. and Simpson, R. J., Eur. J. Immunol. 1988. 18: 193. 16 Aarden, L., Landsdorp, P. and De Groot. E., Lymphokine~ 1985. 10: 175. 17 Brakenhoff, J. P. J . , De Groot, E . R., Evers, R. F., Pannckoek. H. and Aarden, L. A , , J. Immunol. 1987. 139: 4116. 18 Mosmann, T., J. Immunol. Methods 1983. 65: 55. 19 Moguilewsky, M. and Philibert, D., J. Steroid Biochem. 1984. 20: 271. 20 Gustafsson, J.-a Carstedt-Duke. J., Poellinger. L., Okret, S., Wikstrom, A.-C., Bronnegbrd, M., Gillner, M., Dong, Y.. Fuxe, K., Cintra, A , , Harfstrand, A. and Agnati. L.. Endocr. Rev. 1987. 8: 185. 21 Tanabe, C., Akira. S., Kamiya,T.,Wong, G. G.. Hiran0.T. and Kishimoto, T., J. Immunol. 1988. 141: 3875. 22 Clarnan, H. N., Moorehead, J. W. and Brenner, W. H.. J. Lab. Clin. Med. 1971. 18: 499. 23 Munck, A. and Leung, K., in Pasqualini, (Ed.), Keceprors and Mechanisms of Action of Steroid Hormones. Marcel Dekkcr. New York 1977, p. 311. 24 Kishimoto, T.. Blood 1989. 74: 1. 25 Hirano,T.,Yasukawa, K., Haranda, H..Taga,T.. Watanabe.Y.. Matsuda, T., Kashiwamura, S., Nakajima, K.. Koyama. K.. Iwamatsu, A., Tsunasawa, S., Sakiyama. F.. Matsui, H.. Takahara-Y.,Taniguchi,T. and Kishimoto,T.. Nature 1986.324: 73. 26 Kawano, M., Hirano,T., Matsuda,T. ,Taga,T., Horii.Y., Iwato. K., Asaoku, H.. Tang, B.,Tanabe, 0.,Tanaka. H., Kuramoto. A. and Kishimoto, T., Nature 1988. 332: 83. 27 Uyttenhove, C.. Coulie, P. G. and Van Snick, J., J. Exp. Med. 1988. 167: 1417. 28 Caddie, J., Richards, C., Harnish, D., Landsdorp, P. and Baurnann, H., Proc. Natl. Acad. Sci. USA 1987. 84: 7251. 29 Helle, M., Brakenhoff, J. P. J., De Groot, E. R. and Aarden. L. A., Eur. J. Immunol. 1988. 18: 957. 30 Waage, A , , Brandtzaeg, P., Halstensen, A., Kicrulf. P. and Espevik,T., J. Exp. Med. 1989. 169: 333. 31 Waage, A., Halstensen, A., Shalaby, R., Brandtzaeg, P . Kierulf, P. and Espevik,T., J. Exp. Med. 1989. 170: 1859. 32 Culpepper, J. A. and Lee, F., J. Immunol. 1985. 135: 3191. 33 Arya, S. K.,Wong-Staal, F. and Gallo, R. C., J. Immunol. 1984. 133: 273.