Immunobiol., vol. 182, pp. 44-55 (1990)

Institute of Immunology, Philipps University, Marburg, Germany

Enhancement of Tumor Necrosis Factor-a Gene Expression by Low Doses of Prostaglandin E2 and Cyclic GMPI:. JIANG-HONG GONG, HARALD RENZ, HANS SPRENGER, MARIANNE NAIN, and DIETHARD GEMSA Received October 12, 1990 . Accepted in Revised Form October 23, 1990

Abstract Macrophage-derived PGE 2 is usually considered to be a down-regulator of TNF-a production. However, we recently demonstrated that PGE z may display dual activities in that low concentrations stimulated whereas higher doses suppressed TNF-a synthesis in resident peritoneal macrophages. To examine the underlying molecular mechanisms, we studied TNF-a gene expression in rat peritoneal macrophages and the murine PU5-1.8 macrophage line. In both macrophage types, PGE 2 enhanced TNF-a gene transcription and production at an optimal concentration of 1 ng/m!. Furthermore, evidence was obtained that PGE2 may stimulate TNF-a mRNA accumulation via a rise of the intracellular messenger cGMP. Both, exogenously added as well as endogenously, by sodium nitroprusside generated cGMP were found to enhance TNF-a gene expression and production. These findings lend further support to the concept that cGMP may represent one of the positive signals for TNF-a synthesis.

Introduction The cytokine tumor necrosis factor-a (TNF-a), produced primarily by macrophages (1-4), displays a broad spectrum of biological activities which reach beyond its originally described oncolytic properties. TNF-a represents an important co-factor of macrophage activation (5-7), is an endogenous pyrogen (8), stimulates prostaglandin E2 (PGE 2) synthesis (9), mediates septic shock syndromes and induces cachexia (10). Synthesis of TNF-a may be controlled on several levels (4, 11-15), and a particular relevance has been attributed to the macrophage product PGE 2 which has been shown to down-regulate TNF-a synthesis in an autocrine manner (16). However, suppression of TNF-a synthesis may only be one facet of PGE/s effect, since we recently found that PGE 2 may as well up-regulate TNF-a production (17). Upon detailed analysis, it became apparent that only higher PGE 2 concentrations inhibited TNF-a production in macrophages, whereas lower concentrations were stimulatory. Since it has been demonstrated in other studies that high doses of PGE 2 reduce TNF-a ,:. This work was supported by the Deutsche Forschungsgemeinschaft (Ge 354/5-2, Ge 354/ 7-1 and KI238/1-1).

Up-Regulation of TNF-a Gene Expression by cGMP . 45

synthesis at the transcriptional and post-transcriptional level (16), it was of particular interest to perform comparative investigations with low PGE z doses. In this paper, we demonstrate that low doses of PGE z are capable of stimulating TNF-a gene expression in a dose-dependent manner and, furthermore, evidence is provided that cGMP may represent the responsible intracellular mediator.

Materials and Methods Animals Lewis rats of both sexes were obtained from the Zentrale Versuchstieranstalt, Hannover, Germany. All rats were free from pathogenic vital and bacterial infections. They were used for experiments, when they were 6 to 8 wk old and weighed 150 g.

Macrophage culture for TNF-a release Non-elicited, resident macrophages were harvested from Lewis rats by peritoneal lavage with pyrogen-free, physiologic saline. After washing, peritoneal cells were suspended in RPMI 1640 medium supplemented with L-glutamine (2 mM), penicillin (100 U/ml), streptomycin (100 ug/ml), 2-ME (10 mM), 50 mM HEPES buffer, and 5 % (v/v) heat-inactivated (56°C, 30 min), pyrogen-free fetal calf serum. Peritoneal macrophages (1 x 106/ml) were seeded into 24-well culture plates (Falcon 3003, Becton Dickinson, NJ, USA) and were allowed to adhere for 1 h at 37°C in a 5 % CO r 95 % air atmosphere. After 3 washings with culture medium, the non-adherent cells (approximately 50 %) were removed and the remaining monolayer (0.5 x 106 /ml) consisted of > 95 % macrophages as determined by phagocytosis of carbon particles and staining for nonspecific esterase (18, 19). After this pre-incubation step to purify macrophages, further experiments employing stimuli of TNF-a synthesis were performed in culture medium without serum addition. Because of the limited availability of resident peritoneal macrophages for functional and, in particular, molecular studies, these cells were partially substituted by the murine macrophage cell line PU5-l.8. Although PU5-l.8 macrophages were less responsive to PGE 2 and cGMP (see Results), they were sufficiently sensitive to study regulation of TNF-a synthesis, which is in line with another report using a different macrophage-like cell line (20). Mycoplasma-free PU5-l.8 macrophages were propagated in vitro and were used for experiments at a concentration of 0.5 x 106/ml under experimental conditions identical to peritoneal macrophages.

Determination of TNF-a TNF-a content of macrophage culture supernatants was assayed by determining the cytotoxicity against TNF-a-sensitive L929 cells (17, 21). Briefly, L929 cells (6 x 104 /0.1 ml) were seeded in 96-well microtiter plates, and grown overnight at 37°C in a 5 % CO r 95 % air atmosphere to establish a dense monolayer. In order to enhance sensitivity of L929 cells to TNF-a, actinomycin D (1 f,lg/ml) was added together with 0.1 ml supernatants of macrophage cultures. Viability of cells was measured after 18 h of incubation by staining for 1 h with 20 f,l1! 0.2 ml of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 5 mg/ml), followed by removal of culture supernatants, and lysis of cells with 0.1 ml isopropanol containing 40 mM HCI (22). The photometric measurement was performed at As60 in a microplate-autoreader MR 600 (Dynatech, Denkendorf, Germany). TNF-a content in a sample, expressed in ng/ml, was calculated by comparison to a calibration curve established with either recombinant murine TNF-a (kindly provided by Dr. G. R. ADOLF, ErnstBoehringer-Institute, Vienna, Austria) or recombinant human TNF-a (kindly provided by BASF/Knoll AG, Ludwigshafen, Germany). Specificity of TNF-a determinations in culture

46 . JIANG-HoNG GONG et al. supernatants was corroborated by two means: 1) parallel use of L929 cells rendered TNF-a insensitive by prolonged cultivation in TNF-a-containing medium and, 2) neutralization of TNF-a by highly specific rabbit anti murine TNF-a antibodies (kindly provided by Dr. D. MANNEL, German Cancer Research Center, Heidelberg, Germany). Preparation of total cellular RNA Adherence-purified macrophages (2 x 10 7120 ml) were cultured on 100-mm Petri dishes in serum-free medium with or without stimulants as indicated under Results. As the end of incubation, the reaction was stopped by washing with ice-cold phosphate-buffered saline. Because of the limited supply of cells, particularly of resident peritoneal macrophages, total RNA was isolated by a mini preparation method (23). Briefly, 2 x 107 macrophages were lysed in a 6M guanidine hydrochloride solution (pH 7.6). Using a 25-gauge needle, the DNA was sheared 5 times. After the RNA was precipitated with 2.5 volume of ethanol, 0.2 M NaCI and 0.01 M dithiothreitol, the DNA and protein contamination were removed by treatment with 2M lithium chloride solution and chloroform/phenol. Analysis of TNF-a gene expression The mRNA accumulation was analyzed by Northern and dot blot method (24). For Northern blot analysis, 20 [lg of total RNA were applied per lane, denatured with glyoxal and dimethylsulfoxide, separated on 1 % agarose gel and blotted onto nitrocellulose filter. For dot blot analysis, 2 [lg of total RNA were denatured with formaldehyde and blotted on nitrocellulose filter directly (25). The blots were pre-hybridized for 16 h at 60 DC in 20 ml of a mixture consisting of 1 M NaCl, 1 % SDS and were subsequently hybridized for 24 h at 60 DC in 15 ml of a mixture of 1 M NaCl, 10 % dextransulfate, 1 % SDS, 100 [lg/ml denatured salmon sperm DNA and 0.2 [lg of a nick-translated 3zP-labelled TNF-a cDNA probe (spec. activity 4 x 108 cpm/ug DNA) (26) under gentle shaking. The blots were then washed 3 times for 10 min at room temperature in 2 x SSC (sodium chloride, 3 M, sodium citrate, 0.3 M, pH 7.0) and 0.2 % SDS, and 3 times for 20 min in 0.1 x SSC and 0.1 % SDS at 50 DC. Following hybridization, filters were exposed at -70 DC to Kodak XAR film using intensifying screens. Densitometric scans of the resulting autoradiographs were area-integrated in order to quantitate the relative mRNA levels. cDNAprobes The mouse TNF-a cDNA probe was a 1.6 Kb Pst I fragment of a pAT 153 plasmid, which was a kind gift from Dr. W. FIERS (Biogent, Ghent, Belgium). The ~-actin cDNA probe was a 0.6-0.7 Kb Sal I1Eco RI fragment of a PBR 322 plasmid, and was donated by Dr. GALWITZ, Giittingen, Germany. Reagents PGE z (kindly donated by Dr. J. PIKE, Upjohn Company, Kalamazoo, MI, USA) was dissolved in ethanol, and this solvent did not exceed 0.05 % (v/v) in the incubation medium. cAMP and cGMP, sodium nitroprusside, polymyxin B, and cycloheximide (all obtained from Sigma, Munich, Germany) were dissolved in culture medium.

Results PGErinduced TNF-a release Addition of PGE z was capable of dose-dependently inducing TNF-a release in resident peritoneal macrophages and the PUS-1.S macrophage cell line (Fig. 1). A rather narrow concentration range between 0.1 to 10 ng/ml was found to be active, with a maximal stimulatory effect of PGE z at 1 ng/

Up-Regulation of TNF-a Gene Expression by cGMP . 47 Perit.MIa

-:: 200


-c, c

~, u.











«I Q) Q)


Figure 1. Effect of various concentrations of PGE 2 on TNF-a release. Resident peritoneal macrophages (0.5 x 106Iml) of Lewis rats or murine PUS-1.S macrophages (0.5 x 1061m I) were incubated with increasing concentrations of PGE2 . After 20 h of incubation, TNF-a content in the culture supernatant was determined by bioassay. Values are representative of five independent experiments and are the mean ± SD of four identically prepared wells.

ml. PGE z concentrations> 10 ng/ml were suppressive. Peritoneal macrophages were more responsive than PUS-1.8 macrophages, the former showing a IS-fold and the latter a 6-fold increase of TNF-a production

Perit. Mia PU5-1.8 M0 C(



e ~

• • •

•• • •

5 4


Z I-


..,. .... ..

3 2






~.~/~,r---~--~,r---~,­ 0 .1 10 100


PGE2 lng/mIl

Figure 2. Dot blot analysis of PGErinduced TNF-a gene expression. Resident peritoneal macrophages or PUS-1.S macrophages (2 x 10 7 /20 ml) were incubated with indicated PGE 2 concentrations. After 12 h of incubation, RNA was prepared and used for dot blot analysis as described in Materials and Methods. Autoradiographs (upper panel) were scanned by densitometry, and relative levels of TNF-a mRNA are represented (lower panel). Shown is a representative result of four independently performed experiments.

48 .

et al.


above control when stimulated with 1 ng/ml PGE 2 . In this and the following experiments, a possible TNF-a production and gene expression by contaminating bacterial lipopolysaccharide could be excluded by addition of S f,lg/ml LPS-binding polymyxin B (data not shown).

PGErinduced TNF-a gene expression The ability of PGE 2 to stimulate TNF-a release from macrophages could have been due to an altered gene expression. Therefore, TNF-a gene expression was studied in peritoneal and PUS-1.8 macrophages after treatment with various doses of PGE 2 . As shown in Figure 2, PGE 2 dosedependently enhanced TNF-a mRNA accumulation in both types of macrophages, again with a maximum at a concentration of 1 ng/ml. PGE 2 had stronger effects on resident peritoneal than on PUS-l.S macrophages, a finding which closely paralleled the different amounts of TNF-a protein that were released from both types of macrophages (Fig. 1). PGE 2 concentrations at 100 ng/ml did not reduce TNF-a mRNA below the level of







Time of Incubation (h) 8








.. • .. ~



• •


~ 0




• 0





Time of Incubation (h)

Figure 3. Kinetics of PGErinduced TNF-a gene expression. PUS-lo8 macrophages (2 x 107 1 20 ml) were incubated with PGE 2 (1 ng/ml). After indicated time intervals, cells were harvested, RNA was prepared and a Northern blot analysis was performed. Autoradiographs (upper panel) were scanned by densitometry, and relative levels of TNF-a mRNA are represented (lower panel). Shown is a representative result of three independently performed experiments.

Up-Regulation of TNF-a Gene Expression by cGMP . 49 80

o cGMP

E .....

Sodium nitroprusside






Z I-





C\l Q) Q)




'jIEI;;u;i=:;a-====;;;;;;IjI C on t r 0 I 024



Time of Incubation (h)

Figure 4. Time course of TNF-a release from peritoneal macrophages treated with cGMP or sodium nitroprusside. Resident peritoneal macrophages (0.5 x 106/ml) were either incubated with cGMP (10-3M), sodium nitroprusside (10-3M) or without additions (control). At indicated time intervals, the culture supernatant was harvested and TNF-a content was determined. Values are representative of five independent experiments and are the mean ± SD of four identically prepared wells.

unstimulated controls which was different from secretion of the bioactive TNF-a protein (Fig. 1) and indicates that higher PGE 2 concentrations are suppressive at the posttranscriptionallevel.

Kinetics of PGErinduced TNF-a gene expression Previously, we had shown that PGE z (1 ng/ml) stimulated TNF-a production with a slow onset between 2-4 h, a strong increase until 10 hand a continuous release until 24 h (17). This particular TNF-a production pattern appeared to be a reflection of the kinetics of TNF-a mRNA accumulation as shown in Figure 3. The initially delayed action of PGE 2 on TNF-a gene expression and TNF-a production suggested that intermediate messengers may be involved.

The role of cGMP It has been well established that PGE 2 efficiently stimulates adenylyl cyclase which leads to a rise of intracellular cAMP that has generally been associated with suppressive effects on leukocytes (27-30). However, we recently found that PGEz, only when added at low concentrations, produced a selective increase of cGMP which suggested that cGMP may represent the intracellular messenger to stimulate TNF-a synthesis (17). Figure 4 demonstrates that exogenous cGMP was indeed capable of rapidly inducing TNF-a release. Furthermore, sodium nitroprusside, an efficient

50 .


et al.

stimulus of guanylyl cyclase and a fast generator of intracellular cGMP (31), was a similarly active agent for TNF-a production. To determine whether cGMP would also stimulate TNF-a gene expression, exogenous cGMP was added to peritoneal macrophages and a kinetic study was performed by Northern blot analysis. Figure 5 shows that cGMP started to induce TNF-a mRNA accumulation within 30 min, with a maximum after 2 h. The small size variation of TNF-a mRNA is an apparent technical artifact and was not representative of other experiments. As a control, ~-actin mRNA levels were studied in parallel and exhibited no increase. The stimulatory effect of cGMP was further substantiated, although indirectly, by the use of cGMP-generating sodium nitroprusside (Fig. 6). Addition of sodium nitroprusside also induced an accumulation of TNF-a mRNA within 1 h, followed by a return to normal levels after 2 to 3 h. When compared to exogenous cGMP (Fig. 5), the peak of TNF-a mRNA accumulation was reached earlier by sodium nitroprusside but the


Enhancement of tumor necrosis factor-alpha gene expression by low doses of prostaglandin E2 and cyclic GMP.

Macrophage-derived PGE2 is usually considered to be a down-regulator of TNF-alpha production. However, we recently demonstrated that PGE2 may display ...
922KB Sizes 0 Downloads 0 Views