Prostaglandin E2 and Tumor Necrosis Factor α Cooperate to Activate Human Dendritic Cells: Synergistic Activation of Interleukin 12 Production

The medium used in this study was RPMI 1640 supplemented with 1% heat-inactivated (30 min, 56°C) pooled human AB serum, 50 U/ml penicillin, 50 μg/ml streptomycin, 2 mM l-glutamine, 10 mM Hepes, 0.1 mM nonessential amino acids, 1 mM pyruvate, and 5 × 10⁻⁵ M 2-mercaptoethanol (all from Boehringer Ingelheim Bioproducts, Vienna, Austria). Human albumin (for intravenous use; Octapharma, Vienna, Austria) was added to a final concentration of 2 mg/ml (= complete medium). Recombinant human GM-CSF (Leucomax; 1.11 × 10⁷ U/mg) was from Novartis (Basel, Switzerland). Recombinant human IL-4 (2 × 10⁷ U/mg) was supplied by the Schering-Plough Research Institute (Kenilworth, NJ). Recombinant human TNF-α (10⁷ U/mg) was purchased from Genzyme (Cambridge, MA). PGE2 was purchased from Sigma Chemical Company (St. Louis, MO) and from Calbiochem-Novabiochem International (San Diego, CA). Forskolin, N-2-O-dibutyryl-cAMP (db-cAMP), LPS (from Salmonella abortus equi), and fluoresceinated Dextran (FITC-DX) were all from Sigma Chemical Co.

DCs were generated from PBMCs as described (16, 17). In brief, PBMCs were isolated from leukocyte-enriched buffy coats by standard density gradient centrifugation on Ficoll-Paque (Pharmacia, Uppsala, Sweden), resuspended in complete medium and 5.0 × 10⁷ PBMCs were allowed to adhere in 75-cm² cell culture flasks (45 min 37°C). Nonadherent cells were removed and adherent cells were cultured in 10 ml of complete medium containing 1,000 U/ml of each GM-CSF and IL-4. DCs were developed under essentially endotoxin-free conditions as indicated by the absence of spontaneous TNF-α production (<5 pg/ml of TNF-α/5 × 10⁵ DC). On day 2, 5 ml of fresh medium containing 1,000 U/ml of GM-CSF and IL-4 were added. After 5 d of culture, cells were harvested, washed extensively, and recultured in cytokine-containing medium at 3 × 10⁵ cells/ml with or without inflammatory mediators. After 48 h, supernatants were harvested for IL-12 ELISA assays and the cells were analyzed for surface Ag expression by flow cytometry.

To determine surface Ag expression, cells (10⁵ DC in 50 μl) were labeled with primary mAb in complete medium followed by FITC-conjugated F(ab′)₂ fragments of goat anti–mouse Ig (Dako, Glostrup, Denmark). The following mAbs were used: VIM-13 (IgM, anti-CD14, a gift of Dr. W. Knapp, Vienna, Austria), G46-2.6 (IgG1, anti-HLA-ABC), L243 (IgG2a, anti-HLA-DR), HB-15a (IgG2b, anti-CD83), 84H10 (IgG1, anti-CD54), AIDC58 (IgG2a, anti-CD58), BB1 (IgM, anti-CD80), BU63 (IgG1, anti-CD86), 5C3 (IgG1, anti-CD40), 1G10 (IgG1, anti-CD43), G44-26 (IgG2b, anti-CD44). Washes were in HBSS containing 0.2% albumin. After the last wash, the cells were stored in HBSS containing 0.2% albumin and 2% formaldehyde.

The endocytic activity of DCs was measured as described previously (12). FITC-DX was used to measure mannose receptor– mediated endocytosis. Cells (10⁵) were incubated with FITC-DX (0.5 mg/ml) for 30 min at 37°C (control at 0°C) and then washed extensively. The samples were analyzed on a FACScan^® (Becton Dickinson, Mountain View, CA). Data were analyzed and presented using CellQuest^® software from Becton Dickinson.

Day-5 DCs (3 × 10⁵ cells/ml) were incubated for 48 h with various stimuli and IL-12 was measured in culture supernatants by ELISA using a commercially available kit (Genzyme) that detects both IL-12 p40 and the bioactive IL-12 p70 heterodimer consisting of p40 and p35. IL-12 was quantitated using a microtiter plate reader.

DCs were either untreated or treated with TNF-α, PGE2, or PGE2 plus TNF-α for 48 h. Irradiated (3,000 rad) DCs (1 × 10³ or 5 × 10⁴) were then cultured with a constant number of 2 × 10⁵ allogeneic T cells in 96-well flat-bottomed tissue culture plates in medium containing 5% pooled human serum. T cell proliferation was measured as [³H]thymidine incorporation (1 μCi/well; 6.7 Ci/mmol, New England Nuclear, Boston, MA). Cells were pulsed during the last 16 h of a 5-d culture period, harvested onto glass fiber filters using a Skatron cell harvester (Skatron Instruments, Lier, Norway), and analyzed in a liquid scintillation counter. Results are expressed as the mean cpm of triplicate wells ± SD.

Day-5 DCs were exposed to graded doses of PGE2, and IL-12 was determined in culture supernatants after 48 h. Fig. 1,A demonstrates that PGE2 stimulated IL-12 production in a dose-dependent fashion. However, the total amount of IL-12 induced by PGE2 was relatively low. PGE2-dependent IL-12 production by DCs occurred in the absence of endogenous TNF-α production (data not shown). TNF-α has been described to enhance IL-12 production in the context of LPS stimulation (18). Therefore, we tested the effects of combinations of PGE2 and TNF-α on IL-12 production by DCs. In the presence of TNF-α, PGE2 dose dependently stimulated high level IL-12 production in DCs (Fig. 1,C). Conversely, the level of IL-12 production also depended on the dose of TNF-α (Fig. 1,D). The effects of PGE2 and TNF-α were synergistic since TNF-α alone completely failed to induce IL-12 production by DCs (Fig. 1,C). In contrast, PGE2 inhibited the LPS-induced IL-12 production by DCs (Fig. 1 B) confirming previous observations (8). PGE2 from two different commercial sources were used with consistent results.

IL-12 production by DCs is always associated with DC activation which can be induced by treatment with LPS (14), BCG (13), a monocyte-conditioned medium (19, 20), or by cross-linking of the CD40 Ag (21). DC activation, which has also been referred to as DC maturation, is characterized by the neo-expression of the CD83 Ag (22), by the upregulation of molecules related to Ag presentation (12, 22), and by the downmodulation of endocytic activity (12, 23). Therefore, we then tested the effects of PGE2 and TNF-α on the process of DC maturation.

The combination of PGE2 and TNF-α turned out to be potent in inducing CD83 expression in DCs (Fig. 2,A). Addition of PGE2 (1 μM) along with TNF-α (1,000 U/ml) induced CD83 expression in 75% of the cells within 24 h (data not shown) and in almost all cells within 48 h (Fig. 2,A). Addition of either substance alone also induced CD83 expression, although to a smaller extent (Fig. 2,A). Moreover, treatment with PGE2 plus TNF-α efficiently upregulated the expression of MHC class I and II molecules, adhesion molecules (CD54, CD58), and costimulatory molecules (CD80, CD86) in DCs (Fig. 3). The expression of CD40 and CD44 was also enhanced. Inspection of these cells by phase contrast microscopy revealed a pronounced dendritic morphology with numerous large veils, a feature of mature DCs (data not shown; reference 10). Conversely, the ability to capture soluble Ag (DX) by mannose receptor–mediated endocytosis was almost completely downregulated in DCs cultured in the presence of PGE2 and TNF-α (Fig. 2 B). Again, either substance alone also downmodulated Ag uptake, although less efficiently.

Most of the effects of PGE2 are mediated by the intracellular second messenger cAMP (24). Therefore, PGE2 effects can be mimicked by nonphysiologic modulators of cAMP such as forskolin and db-cAMP. To investigate the mechanism of the stimulatory effects of PGE2 on DCs, we tested the effects of forskolin or db-cAMP either alone or in the presence of TNF-α on IL-12 synthesis and CD83 expression in human DCs. Fig. 4,A demonstrates that db-cAMP (100 μM) stimulated low level IL-12 production and synergized with TNF-α to induce substantial IL-12 production by DCs. CD83 expression was also induced by db-cAMP (Fig. 4,B). Forskolin at 10 μM failed to induce IL-12 production (Fig. 4,A). Higher concentrations of forskolin could not be tested due to the high toxicity of this compound. However, in the presence of TNF-α forskolin at 10 μM induced measurable IL-12 production (Fig. 4,A). Forskolin also induced CD83 expression (Fig. 4 B).

High levels of adhesion molecules expressed by mature DCs (Fig. 3) contribute to the T cell stimulatory potential of DCs by facilitating the formation of clusters with T cells (11). DCs treated with PGE2 and TNF-α were most potent in cluster formation with allogeneic T cells (data not shown). Either substance alone also enhanced cluster formation, although at a lower level.

The formation of clusters serves to establish close contacts between DCs and T cells, thereby favoring TCR engagement and T cell activation (11). Fig. 5 shows that DCs treated with PGE2 and TNF-α also exhibited the highest capacity to stimulate allogeneic T cell proliferation. Treatment of DCs with either substance alone also enhanced T cell proliferation, but to a smaller extent (Fig. 5).

PGE2 has been demonstrated to be a potent inhibitor of the LPS-induced human IL-12 production (8). In the present work, we demonstrate that PGE2 in the absence of LPS is capable of stimulating IL-12 production by human DCs. PGE2 per se was a weak stimulus (Fig. 1,A), but was synergized with TNF-α to induce the production of high levels of IL-12 by DC (Fig. 1, C and D). This is consistent with previous reports that demonstrated that PGs are relatively weak inflammatory stimuli, but potentiate inflammatory reactions in the presence of other mediators (6, 7). Moreover, we show that PGE2 contributes to DC maturation (11) which is characterized by the downmodulation of mannose receptor–mediated endocytosis (Fig. 2,B), by the upregulation of molecules that are important for T cell activation (Fig. 3), as well as by the neo-expression of the CD83 Ag (Fig. 2,A). DC maturation can also be induced by monocyte-conditioned medium (19, 20). Our findings suggest that PGE2, which is released by activated monocytes, contributes to the DC maturation function of monocyte-conditioned medium. PGE2-treated DCs also exhibited an increased accessory cell potential which was evident as an enhanced capacity to form clusters with allogeneic T cells (data not shown) and to stimulate allogeneic T cell proliferation (Fig. 5).

Our data suggest the involvement of a cAMP signaling pathway in DC maturation since addition of db-cAMP and forskolin, which are known to increase intracellular levels of cAMP (24), mimicked the effects of PGE2. Similarly, in human monocytes, cAMP synergized with TNF-α to upregulate the synthesis of IL-1β (25), suggesting that the TNF-α signal transduction pathway that leads to intracellular accumulation of ceramide (26) and the PG pathway via elevation of the intracellular second messenger cAMP (24), cooperate to induce cytokine production by APCs (reference 25 and this work).

DCs obviously produce IL-12 at different anatomical sites and different states of their development. DCs residing in peripheral tissues are immature. Tissue DCs are actively endocytosing Ag. DCs cultured with GM-CSF and IL-4 exhibit high endocytic capacity and appear to correspond to these tissue DCs (11). At this state, PGE2 has stimulatory effects and strongly enhances IL-12 synthesis in the presence of TNF-α (Fig. 1, C and D). The capability of immature DCs to produce large amounts of IL-12 in response to inflammatory conditions reflects the contribution of DCs to innate resistance and primarily functions to activate NK cells (27). Accumulation of inflammatory mediators then prompts DCs to migrate to the T cell areas of secondary lymphoid organs where they present Ags to recirculating T cells. During contact with Ag-specific T cells, IL-12 produced by DCs after triggering of their CD40 antigen by T cell CD40 ligand (28, 29) favors the generation of Th1 type T cell responses. At this level, PGE2 appears to inhibit T cell activation and to selectively suppress the production of Th1 cytokines (30–32). Taken together, PGE2 appears to exert both stimulatory and inhibitory effects on the development of Th1 type T cell responses depending on the site of PGE2 action.

The combination of PGE2 and TNF-α induced full DC activation (Figs. 2 and 3), which included IL-12 production by DCs (Fig. 1). This mechanism of activation, which is independent of infectious Ags, may play a role in autoimmune diseases. Mononuclear cells derived from inflamed rheumatoid arthritis joints exhibited enhanced production of both PG (33) and TNF-α (34), which can cooperate to increase IL-12 production (Fig. 1). IL-12 production is indeed increased in synovial mononuclear cells from arthritis patients (35) and in tubular epithelial cells of mice with lupus nephritis (36). Moreover, IL-12 is known to accelerate the development of autoimmune diseases (37). Furthermore, the induction of DC maturation by PGE2 and TNF-α in the inflammatory lesions may result in continuous presentation of self Ag by fully activated DCs in vivo, and may thereby support the development of autoimmunity.

In conclusion, our data indicate that PGE2 must also be considered an important activator of human DCs. The beneficial effect of nonsteroidal antiinflammatory drugs in inflammatory diseases such as rheumatoid arthritis may relate to the inhibition of PGE2-mediated DC activation.

We gratefully acknowledge the excellent technical assistance of Christine Papesh. The authors are grateful to Dr. Nikolaus Romani for careful reading of the manuscript and to Dr. G. Utermann for kind support.

This work was supported by grant P11758MED of the Austrian Science Fund (to Martin Thurnher).