0021-972X/91/7302-0296$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society

Vol. 73, No. 2 Printed in U.S.A.

Glucocorticoids Inhibit Lipopolysaccharide-Induced Production of Tumor Necrosis Factor-a by Human Fetal Kupffer Cells* WILLIAM H. KUTTEH, WILLIAM E. RAINEY, AND BRUCE R. CARR Division of Reproductive Endocrinology and the Cecil H. and Ida Green Center for Reproductive Biology Sciences, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas, Texas 75235

fetal Kupffer cells. The inhibitory effects of glucocorticoids appeared to be specific, since estrogen, progesterone, and testosterone had no effect on LPS stimulation of TNFa production. None of the steroids tested altered basal production or enhanced the LPS-stimulated production of TNFa by fetal Kupffer cells. The inhibition by glucocorticoids could be reversed by the addition of RU 486, indicating that this effect was mediated by the glucocorticoid receptor. These results demonstrate that human fetal macrophages demonstrate mature macrophage function in early gestation; they can be activated to produce TNFa by a well characterized modulator of cellular function (LPS) and suppressed by glucocorticoids. (J Clin Endocrinol Metab 73: 296301,1991)

ABSTRACT. Inflammatory mediators, such as interleukin-1/? (IL-1/?) and tumor necrosis factor-a (TNFa) are secreted by fixed tissue macrophages and exhibit local autocrine and paracrine effects as well as distant endocrine effects. Human fetal Kupffer cells, the fixed tissue macrophages of the liver, may play a role as modulators of immune and endocrine function in early embryonic and fetal development. In the present study we isolated human fetal Kupffer cells to greater than 90% purity and prepared short term cultures to investigate the effect of glucocorticoids on the secretion of the cytokine TNFa. Fetal Kupffer cells secreted TNFa and IL-1/3 after culture with bacterial lipopolysaccharide (LPS), indicating that these cells express mature macrophage function. Cortisol and dexamethasone dramatically suppressed the LPS-stimulated secretion of TNFa by

G

adult, these cells function in the clearance of biologically harmful materials and in the modification of immune function by the production of the cytokines IL-1/? and TNFa (7-9). The human fetal liver, as a fraction of body mass, is large compared with that of the adult. The liver represents approximately 10% of the total fetal weight at the start of the second trimester (10). Studies of the functional capabilities of Kupffer cells from the human fetus have been limited because of the absence of a method for their isolation. Recently, the purification and characterization of human fetal Kupffer cells was described (11). TNFa is a macrophage product secreted in response to endotoxin and other stimuli that indicate host infection (12). TNFa exhibits diverse activities, ranging from local mediation of cellular or tissue homeostasis to systemic events resulting in alterations of physiological processes (13). These activities depend upon the concentration of TNFa at the cellular, tissue, or systemic level and can be modified by glucocorticoids in animal systems (14, 15). We recently reported that human fetal Kupffer cells are able to phagocytose 0.8-jiim latex beads and respond to bacterial lipopolysaccharide (LPS) with the produc-

LUCOCORTICOIDS exhibit a wide range of effects on cells, including those of the immune system. These hormones enter the target cell primarily by diffusion, interact with intracellular receptors, and then mediate their effects by regulation of specific genes or their products (1, 2). Glucocorticoids are known to have immunosuppressive and antiinflammatory effects on macrophage differentiation and function. In animal and adult human models, glucocorticoids have been shown to alter the macrophage production of cytokines, such as tumor necrosis factor-a (TNFa) and interleukin-1/3 (IL-1/3) (3). The largest population of resident tissue macrophages in the adult is found in the liver (4). These hepatic macrophages (Kupffer cells) play major roles in antigen processing, secretion of cytokines, mediation of immune reactions, maintenance of hepatocyte function, and metabolism of lipoproteins and cholesterol (5, 6). In the Received August 31,1990. Address requests for reprints to: William H. Kutteh, M.D., Ph.D., Department of Obstetrics and Gynecology-J6.114, 5323 Harry Hines Boulevard, University of Texas Southwestern Medical Center, Dallas, Texas 75235-9032. * This work was supported by NIH Research Grants HD-07190 and HD-1784. WHK is the recipient of the AFS/Ortho Distinguished Fellowship in Reproduction.

296

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 February 2016. at 14:35 For personal use only. No other uses without permission. . All rights reserved.

CORTISOL BLOCKS FETAL KUPFFER CELLS TNFa tion of TNFa and IL-1/3, demonstrating mature macrophage function (11). Because we believe that these cells play important roles in the regulation of endocrine and immune function during fetal development, we sought to determine the influence of glucocorticoids and other steroids on the production of TNFa by fetal Kupffer cells stimulated with LPS in vitro. We report here that both dexamethasone and cortisol had profound inhibitory effects on the production of LPS-stimulated TNF« by cultured human fetal Kupffer cells. This effect was blocked by RU 486, indicating that glucocorticoid inhibition was mediated by the glucocorticoid receptor. The glucocorticoid inhibitory effects appeared to be specific, since estrogen, progesterone, and testosterone had no effect on the production of TNFa. These results demonstrate the human fetal Kupffer cells are active in early gestation and suggest that these cells may play a role in the response of the fetus to infection or other immunological challenges. Materials and Methods Human fetal tissues Fetal tissues were obtained at the time of therapeutic termination at 13-19 weeks gestational age, as determined by fetal foot length. Tissues were obtained in accordance with the Donors Anatomical Gift Act of the state of Texas. A consent form and protocol approved by the Institutional Review Board of the University of Texas Southwestern Medical Center at Dallas were used. Kupffer cell isolation and cell culture We have previously described in detail the methods used for the purification and characterization of human fetal Kupffer cells (11). Briefly, fresh tissue was rinsed extensively, puncture perfused in calcium-free buffer, minced into 1-cm3 pieces, then dissociated into single cells between layers of fine nylon mesh in 25 mM HEPES buffer with 0.05% collagenase type IV (Sigma, St. Louis, MO). The cells were centrifuged at 100 x g for 3 min to pellet the hepatocytes. The supernatant, enriched for Kupffer cells, was centrifuged at 375 X g for 10 min, the pellet was resuspended in medium, and the cells were overlayed onto Ficoll-Hypaque (1.090 g/cc; Whitaker Bioproducts, Walkersville, MD). After centrifugation at 350 X g for 40 min, the buffy interface was collected, rinsed, and resuspended in RPMI1640 with 10% fetal calf serum and antibiotics (Gibco, Grand Island, NY). Kupffer cells were plated at a concentration of 2 X 106 cells/well onto 24-well tissue culture plates (Costar, Cambridge, MA) and incubated overnight at 37 C in 95% air and 5% CO2. After the overnight incubation, nonadherent cells were aspirated, wells were rinsed, and 1.0 mL complete medium was added. We determined that the remaining adherent cells numbered 1 x 106 by directly counting the cells after trypsin treatment and detachment. Routinely, 90-95% of the adherent cells were identified as macrophages using a panel of three

297

monoclonal antibodies to human macrophages or monocytes (11). Adherent cells were further characterized by staining for a-naphthyl acetate esterase (Sigma), one of the most reliable markers for the identification of macrophages (7). To determine the phagocytic capacity of Kupffer cells, the adherent cells were incubated with 0.8-Aim latex polystyrene beads (Sigma) for 30 min, then washed extensively before examination by direct microscopy. Cells were considered positive if 10 or more latex beads were visualized intracytoplasmically. Phagocytosis of particles greater than 0.1 pm has been described as an important criterion to distinguish Kupffer cells from other sinusoidal cells of the liver (16). Each experimental point was performed in triplicate. Each experiment was repeated at least four times. In certain experiments, LPS from E. coli 026:B6 (Sigma) was added at 0.5-10 jig/mL. All media, calf serum, and solutions used in these experiments were tested and found to be negative for the presence of bacterial LPS, as determined by the Limulus amebocyte lysate assay (Sigma) using a modification capable of detecting picograms of endotoxin contamination (17). Cell viability was assayed by trypan blue (0.5%) exclusion (Flow Laboratories, McLean, VA). Other materials Dexamethasone, cortisol, testosterone, progesterone, and 17a-estradiol were obtained as a purified powder from Sigma. RU 486 (mifepristone) was graciously provided by Dr. Etienne E. Baulieu (Bicetre, France). Stock solutions were prepared in 95% ethanol, aliquoted, and stored at —20 C until use. Immunofluorescence Monoclonal antibodies against human macrophages or monocytes used to identify Kupffer cells were: EMB-11 (Dakopatts, Glostrup, Denmark), 3C10 (American Type Culture Collection, Rockville, MD), anti-Leu-Ml (CD15), and antiLeu-M5 (CD11C; Becton-Dickinson, Mountainview, CA). Purified mouse immunoglobulin G (IgG; Becton-Dickinson) was used as a negative control in these studies, using an indirect immunofluorescence technique. After incubation with the primary antibody for 30 min, slides were washed, then incubated with the fluorescein-labeled F(ab') 2 fragment of antimouse IgG (heavy and light chains) absorbed with human, bovine, and horse Ig (American Qualex International, LaMarata, CA) for 30 min (18). Slides of cells prepared on a Cytotech centrifuge (Miles Laboratories, Inc., Elkhart, IN) or cells grown on microscope slide chambers (Flaskette, Miles Laboratories, Inc., Naperville, IL) were examined after immunofluorescence staining using a Leitz-Orthoplan microscope (Rockleigh, NJ) with a super high pressure mercury lamp. Mononuclear reagents were incubated at various dilutions with mononuclear cells from adult peripheral blood, as described above, to determine ideal conditions and to serve as a positive control. Assay for TNFa TNF« is a well characterized product of macrophages (7,12, 13). Commercially obtained enzyme-linked immunosorbent assay (ELISA) kits for TNFa with a sensitivity of 5 pg/mL (Research and Diagnostic Systems, Minneapolis, MN) were

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 February 2016. at 14:35 For personal use only. No other uses without permission. . All rights reserved.

KUTTEH, RAINEY, AND CARR

298

JCE&M«1991 Vol 73 • No 1

used to assay Kupffer cell supernatants. The intraassay coefficient of variation was 8.0%, and the interassay coefficient of variation was less than 8.0%. We have previously reported that the immunodetectable TNFa was completely bioactive using an 18-h cytotoxicity assay on mouse fibroblast cell line L929 (American Type Culture Collection, Rockville, MD) (11, 19). Recombinant human sequence TNFa (>2 X 107 U/mg in the L929 cytotoxicity assay; Research and Diagnostic Systems) was used to standardize measurement of TNFa activity. All data presented here were obtained using the ELISA for TNFa:. 0.1

Statistics Statistical analysis of the data was performed using analysis of variance with Newman-Keuls multiple comparison of the t test. The Academic Computing Services of the University of Texas Southwestern Medical Center at Dallas were used.

Results The final preparations of Kupffer cells were greater than 95% viable, as accessed by trypan blue exclusion, and 90-95% of the adherent cells were judged to be macrophages based on immunofluorescence staining with EMB-11, 3C10, and anti-Leu-M5. We believe that most of the adherent cells are macrophages, because less than 1% of the adherent cells were positive when stained with a panel of monoclonal reactive with T-cell determinants (11). Less than 0.5% of the cells were recognized by anti-Leu-Ml, a monoclonal antibody characterized to react with human monocytes (11). The adherent cells were positive for esterase using a-naphthyl acetate as substrate under conditions that detect activity primarily in monocytes, macrophages, and histiocytes (7). Typically, 95% of the adherent cells could be identified using this marker. In addition, 90-95% of the adherent cells were able to phagocytize 0.8-jum latex beads within 30 min, a characteristic of Kupffer cells but not other sinusoidal cells of the liver (16). Furthermore, approximately 1% of the cells present at 13-19 weeks gestation could be identified as macrophages on frozen sections and touch preparations of intact fetal liver based on staining with a-naphthyl acetate esterase and mouse monoclonal antibodies to macrophages/monocytes (11). Our calculations indicate that approximately 50% of the Kupffer cells were recovered during isolation. Purified fetal Kupffer cells were incubated for 24 h with 0.1-20 /xg LPS/mL with or without dexamethasone (1 /zmol/L) to determine if dexamethasone would inhibit the production of TNFa (Fig. 1). Quantities of released TNFa were noted to increase with 0.1 jug/mL LPS, with half-maximal stimulation at 0.5 ixg/mL LPS and a plateau response between 5-20 ng LPS/106 cells. Dexamethasone had a profound inhibitory effect on the production of TNFa by LPS-stimulated Kupffer cells that was significant (P < 0.001) at all concentrations of LPS tested

0.5 1.0

2.5 5.0 10.0 20.0

Hg LPS added/ml FIG. 1. Effect of dexamethasone (Dex) on the dose response of TNFa production by LPS-treated Kupffer cells. After overnight culture, adherent Kupffer cells (1 X 106) were rinsed, then treated with the indicated amounts of LPS for 24 h in the presence (O) or absence (•) of dexamethasone (1 ^ m ol/L). The quantities of TNFa released into culture supernatants were measured by ELISA. Figure 1 represents one of four similar experiments, and the results are expressed as the mean ± SD. The differences between cultures with dexamethasone compared to those without dexamethasone were significant (P < 0.001) at all doses of LPS except 0.1

except 0.1 /zg/mL. To determine if the inhibitory effect of dexamethasone was shared by other steroids, adherent Kupffer cells were incubated with estrogen, progesterone, testosterone, cortisol, or dexamethasone in the absence or presence of a maximally stimulating concentration of LPS (10 fig/mL; Fig. 2). After 24 h, the quantities of TNFa released into culture supernatants were measured. Hormones were tested at various concentrations, ranging from 10 jumol/ L to 10 pmol/L. Figure 2 is one experiment representative of five similar experiments using hormone concentrations of 1 jimol/L. None of the steroid hormones tested altered the low baseline production of TNFa by human fetal Kupffer cells. Dexamethasone and cortisol dramatically inhibited LPS-stimulated TNFa levels (P