Prostaglandin D2 affects the differentiation and functions of human dendritic cells: impact on the T cell response



The local environment in which dendritic cells (DC) differentiate is important for the acquisition of their immunostimulatory properties. Since prostaglandin D2 (PGD2), a major prostanoid produced during inflammatory reactions, is involved in the control of immune responses, its effect on the differentiation and functions of human monocyte-derived dendritic cells (MDDC) was studied. We show that DC differentiated in the presence of PGD2 (PG/DC) have an unusual phenotype, with modifications in the expression of molecules involved in antigen (Ag) capture and presentation, leading to higher endocytic and Ag-processing activities. However, under conditions that necessitated Ag processing and presentation, PG/DC have an impaired ability to stimulate naive T cells, whereas superAg-pulsed DC efficiently promote their proliferation. Upon lipopolysaccharide or TNF-α/IL-1β stimulation, PG/DC phenotypically mature but produce abnormal amounts of immunoregulatory cytokines (decreased IL-12p70/IL-10 ratio). Moreover, mature PG/DC fail to up-regulate the chemokine receptor CCR7 and show an impaired migration towards its ligand CCL19. Finally, PG/DC favor the differentiation of naive T cells toward Th2 cells, an effect dependent on IL-10 and inducible costimulator ligand expression by DC. Most of the herein described effects of PGD2 on MDDC can be reproduced, usually with a higher efficacy, with a selective D prostanoid receptor (DP)1, but not DP2, agonist. Taken as a whole, these results demonstrate that PGD2 impacts DC differentiation and functions, and extend the concept that it exerts important roles in immunity.


Monocyte-derived dendritic cells


D prostanoid receptor


DC-specific ICAM-3 grabbing nonintegrin


Mannose receptor


Inducible costimulator ligand


Staphylococcal enterotoxin Ag


Dermatophagoides pteronyssinus


Peroxisome proliferator-activated receptor-γ


Dendritic cells (DC), the most potent Ag-presenting cells, play key roles in orchestrating cellular and humoral immune responses to self and foreign Ag 1, 2. Activation of naive T cells and their polarization into effector cells require interactions with mature DC in the secondary lymphoid organs. Immature DC are located in the peripheral tissue where they encounter incoming Ag. Upon activation by inflammatory factors or microbial components, they undergo a complex process of maturation, enabling their migration towards the regional lymph nodes through a modification of adhesion molecule and chemokine receptor expression 3. Moreover, DC maturation is associated with high surface expression of MHC and costimulatory molecules, and with secretion of immunomodulatory cytokines and chemokines. These factors drive the polarization by DC of naive T cells activation into Th1- or Th2-producing cells, or alternatively into regulatory T cells.

Allergic diseases occur as a consequence of excessive development of Th2 cells 4. This pathological process implicates the participation of DC, as previously shown in mice 5 and by ex vivo approaches in allergic patients. Indeed, DC are attracted to the airways at times of allergen challenge 6, and the response of DC from atopic patients to allergens is modified in comparison with that of nonatopic subjects 7. In addition to the genetic background, it is likely that the tissue environment also determines the capacity of DC to induce Th2 responses consecutively to allergen exposure. For instance, among the main mediators released early during the allergic reaction, histamine and PGD2 are important for DC functions 811.

The activities of PGD2 are effected through two plasma membrane receptors, the D prostanoid receptor (DP)1 (coupled with Gαs-type G protein) 12 and DP2 (coupled with Gαi-type G protein) 13. We have reported that PGD2 inhibits the migration of maturing DC via DP1 9, 14, 15 whereas it increases the chemokinetic activity of DC precursors through DP2 11. Moreover, PGD2 profoundly modulates the immunoregulatory capacity of DC via DP1 11, 16. Along with their action on DP, PGD2 and/or PGD2 metabolites can also act directly on intracellular targets. In particular, PGD2 and 15-deoxy-δ(12,14)PGJ2 have been shown to activate peroxisome proliferator-activated receptor-γ (PPAR-γ) 17, a nuclear hormone receptor known to be expressed in human monocytes 18 and to affect many cellular functions, including DC differentiation 19 and maturation 18.

The ability of DC to activate and to polarize naive T cells depends on environmental instructions acting not only during DC maturation but also during their differentiation. For instance, many inflammatory mediators such as IFN-α 2022 have been shown to differentiate DC progenitors into Th1-inducing cells (herein named DC1), whereas IL-3 or PGE2 favor DC2 23, 24. In addition, vitamin D3 and TGF-β induce the differentiation of DC able to induce the development of IL-10-producing regulatory T cells 25, 26.

Because PGD2 regulates DC precursor recruitment and affects DC maturation 11, we sought to determine whether it might act on DC differentiation and functions. We also studied the involvement of both DP1 and DP2 in these settings. Using human monocyte-derived DC (MDDC), we showed that DC differentiated in the presence of PGD2 (PG/DC) exhibit a distorded phenotype and display increased endocytic and Ag-processing activities but reduced capacities to activate Ag-specific naive T cells. Upon activation, PG/DC phenotypically mature, although they failed to up-regulate CCR7 expression, and produce different amounts of immunoregulatory cytokines, compared to controls. We also show that the Th-driving capacity of PG/DC shifts to a Th2 profile. Finally, we suggest that most of the effects exerted by PGD2 on DC functions are mediated by DP1, but not DP2.


Immature DC differentiated with PGD2 display an unusual phenotype

First, we determined the effects of exogenous PGD2 on the phenotype of immature DC, generated in vitro from monocytes in the presence of GM-CSF and IL-4. Having previously established that freshly isolated monocytes express mRNA and protein for both DP1 and DP2 11, BW245C (a potent DP1 agonist) and DK-PGD2 (a potent DP2 agonist) were also used in our settings.

As shown in Fig. 1A, B, compared to the control (DMSO/DC), PG/DC expressed a higher amount of mannose receptor (MR) whereas BW/DC increased the expression of both MR and DC-specific ICAM-3-grabbing nonintegrin (DC-SIGN). This led us to question whether PGD2 treatment affected the MR-mediated endocytosis of immature DC. Compared with DMSO/DC, the FITC-dextran uptake was significantly enhanced in PG/DC and BW/DC, but not in DK/DC (Fig. 1A). On the other hand, the phagocytic activity, as assessed by FITC-labeled Staphylococcus aureus internalization, was not modified (not shown). PGD2 treatment also affected the synthesis of molecules involved in Ag presentation. Indeed, PGD2 increased HLA-DR expression whereas it abrogated that of the typical DC markers CD1a (Fig. 1A, B) and CD1c (not shown). It is noteworthy that BW245C, but not DK-PGD2, reproduced these effects, although with a different intensity.

Figure 1.

Phenotypic analysis of immature MDDC differentiated in the presence of PGD2, DP agonists or forskolin. On day 5, the expression of DC-SIGN, MR, HLA-DR, CD1a, CD80, CD86 and ICOS-L and the uptake of FITC-dextran were assessed by flow cytometry. (A) Results are expressed as means ± SEM (n=12) of ΔMFI; *p<0.05 and **p<0.01 (compared with DMSO/DC). (B) Shown is a representative experiment (out of eight) showing the effect of BW245C (bold line), PGD2 (thin line) in the upper part and, in the lower part, forskolin (Fsk/DC, thin line) compared with vehicle (shaded line) on the phenotype of immature DC. Isotype control (dotted line) is also shown. (C) Effect of PG on the mRNA levels of the costimulatory molecules B7-H1, B7-DC and OX40L, as assessed by RT-PCR. One representative experiment out of six is depicted.

The impact of PGD2 treatment on the expression of costimulatory molecules on immature MDDC was also studied. Treatment with PGD2 increased CD86 expression whereas that of CD80, inducible costimulator ligand (ICOS-L) (Fig. 1A, B), CD40 and CD54 (data not shown) was not significantly modulated. BW245C enhanced the membrane expression of CD80 and ICOS-L (Fig. 1A, B) whereas it was inactive on that of CD40, CD54 (not shown) and CD86 (Fig. 1A). In contrast, DK-PGD2 did not impact the synthesis of these markers. The expression of mRNA encoding for other costimulatory molecules was next analyzed by RT-PCR. Whereas PGD2 only enhanced the expression of OX40L mRNA, BW245C augmented the mRNA levels of B7-H1, OX40L and B7-DC (Fig. 1C; p<0.05 as assessed by image analysis, not shown). Taken together, although their effects differed in some cases, PGD2 and BW245C, but not DK-PGD2, strongly impact the phenotype of immature DC.

DP1 is known to activate the adenylate cyclase system with a rapid increase of intracellular cyclic AMP (cAMP) 27. To further confirm the role of DP1 in the observed effects exerted by BW245C, and because specific potent antagonists are still lacking, Rp-8Br-cAMP, a selective cAMP antagonist was used. Rp-8Br-cAMP reversed the effects of BW245C on DC-SIGN, MR, HLA-DR, CD1a and ICOS-L expression on immature DC (by 46%, 74%, 59%, 66% and 82 %, respectively). In addition, forskolin, a cAMP activator, reproduced the effects of BW245C on DC phenotype (Fig. 1B). Thus, the effects of BW245C are likely to be dependent on intracellular cAMP elevation triggered by DP1 engagement.

Our data also suggest that some of the effects exerted by PGD2 are mediated by (a) DP1-independent pathway(s). To study the potential involvement of PPAR-γ, the potent antagonist GW662 was used. Our data revealed that it did not reverse the effects of PGD2 on DC phenotype (not shown).

DC differentiated in the presence of PGD2 phenotypically mature but failed to migrate upon CCR7 activation

The ability of DC to phenotypically mature was then studied after stimulation with LPS or a cocktail of pro-inflammatory cytokines (TNF-α/IL-1β). Compared to unstimulated cells, MDDC maturation was associated with a markedly increased expression of HLA-DR, CD83, CCR7 (Fig. 2A, B), and of CD40, CD54, CD80 and CD86 (not shown), whereas the levels of CD1a and ICOS-L (not shown) were down-regulated. DC differentiated in the presence of PG analogs displayed a similar phenotype (compared to DMSO/DC) after either mode of activation (not shown), although BW/DC expressed higher amounts of CD83 and HLA-DR after LPS or TNF-α/IL-1β (Fig. 2A, B) stimulation, a phenomenon reversed by Rp-8Br-cAMP (not shown). Most notably, cells differentiated in the presence of PGD2 and BW245C limited the LPS- and TNF-α/IL-1β-induced expression of CCR7, a key chemokine receptor involved in DC migration. As shown in Fig. 2C, mature PG/DC, and particularly BW/DC (but not DK/DC), had a reduced capacity to migrate towards the CCR7 ligand CCL19, compared to DMSO/DC (p<0.05). Taken as a whole, mature DC differentiated with PGD2 almost normally express molecules involved in T cell activation, but have a reduced ability to migrate in response to CCL19, an effect mimicked by the DP1 agonist BW245C.

Figure 2.

Ability of mature DC, differentiated in the presence of PGD2 or DP agonists, to phenotypically mature and migrate. (A) After 5 days, MDDC were activated with LPS or TNF-α/IL-1β, and 2 days afterwards the phenotype was determined by FACS analysis. A representative experiment is shown; DMSO/DC (shaded line), PG/DC (thin line), BW/DC (bold line), unstimulated DMSO/DC (hatched line). (B) Data are expressed as mean ΔMFI ± SEM (n=8–12); *p<0.05 compared with activated DMSO/DC. (C) Impairment of chemotactic responses of PG/DC and BW/DC to CCL19. Results represent the mean number of emigrated cells/high-power field (hpf) ± SEM (n=3); *p<0.05 compared with activated DMSO/DC.

Mature PG/DC have a modified ability to produce cytokines and chemokines

The capacity of differentiated DC to secrete typical pro-Th1- and Th2-type cytokines and chemokines upon activation was next tested. Compared to unstimulated cells (2.1±1.1 and 5.9±1.3 pg/ml, respectively), LPS and, to a lesser extent, TNF-α/IL-1β induced the production of IL-12p70 and IL-10 by DMSO/DC (Fig. 3A). Interestingly, PG/DC and BW/DC produced lower amounts of IL-12p70 upon LPS or TNF-α/IL-1β stimulation, whereas the production of IL-10 by BW/DC, and at a lower level by PG/DC, was enhanced (Fig. 3A). Addition of Rp-8Br-cAMP during the differentiation step significantly reversed (by approximately 40–50%) the effects of BW245C on the secretion of IL-10 and IL-12p70 by mature DC whereas forskolin mimicked the action of DP1 agonist (Fig. 3A). On the other hand, GW9662 slightly decreased (by approximately 25–30%) the effects of PGD2 on the TNF-α/IL-1β-induced secretion of IL-10 and IL-12p70.

Figure 3.

Ability of mature DC, differentiated in the presence of PGD2, DP agonists or forskolin, to secrete cytokines (A) and chemokines (B). Immature MDDC were stimulated with LPS (left panel) or TNF-α/IL-1β (right panel). Data are the means ± SEM (n=12); *p<0.05 and **p<0.01 compared with stimulated DMSO/DC.

When the synthesis of the canonical Th1-type (CXCL10) and Th2-type (CCL1) chemokines was assessed, LPS and, to a lesser extent, TNF-α/IL-1β also significantly increased their production compared to unstimulated DMSO/DC (47±15.2 and 245±60.5 pg/ml, respectively; Fig. 3B). Mature PG/DC (except after LPS stimulation) and particularly BW/DC had a defected ability to produce both CXCL10 and CCL1, whereas DK/DC produced comparable levels relative to DMSO/DC.

In summary, differentiation of DC in the presence of PGD2 strongly impacts the secretion of IL-12p70 and IL-10 by mature DC, a phenomenon in part due to DP1 and PPAR-γ. On the other hand, PGD2 (except upon LPS stimulation) and particularly BW245C down-regulate the secretion of both Th1- and Th2-type chemokines by mature DC.

PG/DC have a reduced ability to activate naive T cells after Ag capture and processing

The capacity of PG/DC to activate naive T cells was next investigated. As shown in Fig 4A, sensitization with the superAg staphylococcal enterotoxin (SEA) induced a significant [3H]thymidine incorporation in autologous naive Th cells whatever DC populations used. We next measured T cell activation induced by DC stimulated with the allergen extract from Dermatophagoides pteronyssinus (Dpt), a process that requires Ag capture through MR 28, processing and presentation. Compared to unpulsed cells, DC pulsed with Dpt induced an increased T cell proliferation (Fig. 4B). In contrast, PG/DC and particularly BW/DC had a reduced ability to promote Dpt-induced T cell proliferation (p<0.05). We next investigated the possibility that this effect may be due to a reduced ability of DC to process Ag. For this purpose, DQ-ovalbumin (DQ-OVA), a self-quenched Ag conjugate, was used. After endocytosis through MR and proteolysis, released fluorescent DQ-OVA-derived peptides 29 were quantified by flow cytometry. Fig. 4C shows that treatment with PGD2 or BW245C increased the processing capacities of DC, suggesting that the reduced allergen-specific T cell proliferation is rather due to alteration of Ag presentation.

Figure 4.

Analysis of the Ag processing and of T cell stimulatory function in differentiated DC. The proliferation of autologous naive Th cells induced by SEA- (A) and Dpt-sensitized DC (B) differentiated with PGD2 or DP agonists was quantified at day 4 (SEA) or 5 (Dpt). The results are expressed as means ± SEM (n=6); *p<0.05 compared with Ag-sensitized DMSO/DC. (C) Effects of PGD2 and BW245C on the processing of DQ-OVA by MDDC. DMSO/DC (shaded line), PG/DC (dotted line) and BW/DC (bold line) were incubated for 30 min or 1 h with DQ-OVA at 37°C and analyzed by flow cytometry. As a negative control, DC were incubated at 4°C with DQ-OVA (hatched line). One representative experiment out of four is shown.

PG/DC favor the differentiation of Th2-cytokine-secreting T cells

We then studied whether PG/DC or BW/DC could modulate the polarization of SEA-activated autologous CD45RA+CD4+ T cells. After restimulation with anti-CD3 plus anti-CD28 mAb, MDDC sensitized with SEA induced the differentiation of naive CD4+ T cells to IL-4- and IFN-γ-producing cells, whereas the number of IL-10+ cells was low (Fig. 5A). PG/DC and BW/DC (but not DK/DC) strongly reduced the percentage of IFN-γ-producing cells compared with DMSO/DC, whereas they markedly increased that of IL-4+ cells. Although not significant, BW/DC tended to increase the percentage of IL-10-producing CD4+ cells. This result was confirmed by quantifying the cytokine concentrations in the supernatants of T cells (Fig. 5B). Compared to DMSO/DC, the production of IL-4 and IL-10 was significantly increased by SEA-pulsed PG/DC and BW/DC whereas that of IFN-γ was reduced (p<0.05). Of note, DC differentiated with forskolin also favored the differentiation of IL-4- and IL-10-producing cells (not shown).

Figure 5.

Promotion of Th2 response by SEA-pulsed PG/DC and BW/DC. (A) Naive T cells were co-cultured with unstimulated or SEA-sensitized DC and, after restimulation for 6 h with anti-CD3 + anti-CD28 mAb, the percentage of CD4+ cells producing IFN-γ, IL-4 and IL-10 was quantified by flow cytometry. Shown is a representative experiment out of six. (B) Quantification of secreted cytokines was performed by ELISA in supernatants from 24-h activated T cells obtained in the same conditions. Data shown are means ± SEM (n=6); *p<0.05 compared with SEA-pulsed DMSO/DC.

IL-10 and ICOS-L are involved in the Th2-biased effect induced by PG/DC

Among DC-derived factors, IL-10 and ICOS-L play pivotal functions in the differentiation of Th2 and regulatory T cells 25, 30. Although their effects on the frequency of IFN-γ-producing T cells were modest, neutralizing anti-IL-10 or anti-ICOS-L Ab strongly decreased the percentage of IL-4+ cells when PG/DC or BW/DC were used to stimulate T cells (Fig. 6). Finally, although the anti-IL-10 Ab reversed the effect of PG/DC and BW/DC on the percentage of IL-10+ cells, anti-ICOS-L was ineffective. Of note, whatever cytokine analyzed, no additive effect was observed when both Ab were added simultaneously to DC cultures (not shown). Altogether, these data suggest that IL-10 and ICOS-L are involved in the Th2-biased response induced by PG/DC and BW/DC.

Figure 6.

Role of IL-10 and ICOS-L in the promotion of Th2 response by PG/DC and BW/DC. Neutralizing anti-IL-10 and anti-ICOS-L Ab or the isotype control were added during the co-culture. The percentage of CD4+/cytokine+ cells in one representative experiment out of three is represented.


DC differentiation and acquisition of immunostimulatory capacity by bone marrow-derived precursors take place in peripheral tissues under the control of several factors such as cytokines and lipid mediators. Among them, PGE2 appears to play a key role in modulating DC development and functions (for review, 31). However, depending on the stimulus and according to the PGE2 receptors recruited, the effects of PGE2 on the immunostimulatory functions of differentiated DC are diverse and sometimes opposite 31. PGD2 represents another major cyclooxygenase metabolite produced in vivo during inflammation, particularly during allergic reactions 32. Along with its dual role in the genesis and in the control of inflammatory reactions, PGD2 is thought to be an important component of the immune system. Recently, we showed that PGD2 favors the migration of DC precursors (monocytes) 11 and strongly modulates the maturation process of differentiated DC 10, 11. In this study, we show that PGD2 acts as a potent modulator of DC differentiation and functions.

Based on the cell morphology and the expression of CD14 and CD16, PGD2 did not affect the changes associated with monocyte differentiation into DC. No difference in the yield of immature DC (CD11c+) was found, suggesting that PGD2 (used at 1 µM in our assays) induces neither apoptosis nor necrosis during the differentiation process. However, major differences were found between DMSO/DC and PG/DC. First, a marked increase in the expression of MR, a C-type lectin involved in Ag capture and internalization 29, 33, was observed in PG/DC and was associated with an increased MR-mediated endocytosis, as assessed by dextran uptake. This observation may be of importance in the context of allergic diseases. Indeed, we have recently shown that MR is involved in the internalization of the major Dpt allergen and that its expression is higher on DC isolated from allergic patients, compared to healthy donors 28. It is therefore possible that PGD2 may be involved in such an increased uptake of mannosylated Ag.

Second, the expression of molecules involved in Ag presentation was strongly modified in PG/DC. In particular, PGD2 abrogates the expression of CD1a and CD1c, two important molecules involved in the presentation of (glyco)lipid Ag to T cells 34. This regulatory pathway may blunt the activation of CD1-restricted T cells, whose role in some pathologies, including asthma and cancer, has recently been highlighted 35, 36. The expression of HLA-DR and of the costimulatory molecules OX40L and, to a lesser extent, ICOS-L was also enhanced in PG/DC. Interaction of these molecules with their counter-receptors present on T cells are important for both the development of regulatory T (ICOS-L, B7-H1) 30, 37 or Th2 cells (ICOS-L and OX40L) 38, 39 and for the control of effector T cells. Although Ag capture and expression of MHC class II were increased, PG/DC have a lower ability to activate naive T cells under conditions where Ag processing is required. This effect is, however, neither due to a reduced ability to process Ag (Fig. 4C) nor to a defect in costimulatory molecule expression (Fig. 2). Moreover, the functions of MHC class II or costimulatory molecules appear to be preserved in PG/DC as shown by SEA-induced T cell proliferation (Fig. 4A) or by allogeneic mixed lymphocyte reaction (not shown). Although speculative, PGD2 probably affects in DC the processes that target peptide fragments into MHC class II compartments or that translocate MHC class II-peptide complexes to the cell surface. Among possible candidates is the DC-specific lysosome-associated membrane glycoprotein DC-LAMP, a lysosomal molecule recently shown to be targeted by cAMP-elevating drugs 40.

Upon stimulation with LPS or TNF-α/IL-1β, PG/DC mature almost normally, as assessed by cell surface expression of costimulatory molecules. In contrast, PG/DC failed to up-regulate CCR7, a chemokine receptor involved in DC migration from inflamed tissues to the afferent lymph nodes 41. This reduced expression affects the migratory response of mature PG/DC to CCL19. This result is in line with the concept that PGD2 plays a negative role in the emigration of DC 9, 11, 14, 15. Concerning the production of soluble factors, PGD2 reduces the production of both type-1 and type-2 chemokines by mature DC upon TNF-α/IL-1β (but not LPS) stimulation. This phenomenon may represent a feed-back mechanism to regulate the recruitment of T cells towards inflammatory sites. Taken together, our data suggest that PGD2 might hold back the development of acquired immunity, although the in vivo pertinence of this mechanism remains to be determined.

In addition, differentiation of DC precursors in the presence of PGD2 strongly modifies their capacities to produce immunoregulatory cytokines. Indeed, PG/DC produced lower amounts of IL-12p70 and larger amounts of IL-10, a cytokine pattern that might hamper an efficient Th1 response. To address this issue, we evaluated the effect of SEA-sensitized DC to polarize autologous naive T cells. SEA-pulsed DC generated in the presence of PGD2 reduced the frequency of IFN-γ-producing T cells whilst they enhance that of IL-4-producing T cells. Moreover, PG/DC did not favor regulatory T cell generation, as shown by the lack of IL-10 and CD25 expression on T cells and of inhibition of activated T cell proliferation (not shown). These data suggest that PG/DC induce the generation of Th2 rather than of regulatory T cells. This could be due to the decreased IL12p70/IL-10 ratio and/or to differential expression of costimulatory molecules on maturing DC. Indeed, experiments using neutralizing Ab indicate that DC-derived IL-10 and ICOS-L are partially involved in the observed Th2 response induced by PG/DC.

The use of specific agonists enabled us to study the respective contribution of DP1 and DP2 in the herein described effects exerted by PGD2. Our data show that DP2, although expressed in monocytes 11, is not recruited by PGD2 to modify DC differentiation and functions. This result is consistent with data showing a rapid down-regulation of DP2 expression during monocyte differentiation 11. Therefore, its contribution may be restricted to the recruitment of DC progenitors into inflamed tissues. On the other hand, our data demonstrate that DP1 (and the cAMP pathway that it triggers upon engagement) is implicated for most (Ag capture, cytokine production, ability to polarize the immune response), but not all (expression of some surface markers by immature DC) effects induced by PGD2. This suggests, for the latter criteria, the involvement of DP1-independent mechanism(s). Among them, we postulated that PPAR-γ activation may be of importance. However, when we used a selective antagonist, our data showed that it plays a minor role in DC differentiation and functions, thus implying other targets.

In conclusion, our data show that PGD2 affects the differentiation of DC and modulates their biological functions. PGD2 induces an unusual phenotype on immature DC, modulates their maturation program and favors the differentiation of Th2 cells. Since micromolar concentrations of PGD2 are detected in certain noninflamed tissues 42 such as the spleen and the bone marrow, it may be the case that PGD2 modulates DC functions under steady-state conditions. Moreover, the enhanced production of PGD2 in peripheral sites during inflammation, particularly during the allergic reaction 32 or tumor formation 43, highlights its role on DC functions in pathological conditions and suggests that PGD2 may be an important component in the control of the immune response in these diseases.

Materials and methods

Reagents and Ab

PGD2, DK-PGD2 and BW245C were purchased from Cayman Chemicals (Ann Arbor, MI), and LPS (type 055B5), Rp-8Br-cAMP and SEA from Sigma-Aldrich (St. Louis, MI). GW9662 was a gift from Dr. P. Brown (GlaxoSmithKline, Research Triangle Park, NC). The allergen extract from Dpt (Stallergènes, Paris, France) was detoxified by passage through a column of Detoxigel (Pierce, Rockford, IL). Human recombinant GM-CSF was purchased from Peprotech (Rocky Hill, NJ), and TNF-α, IL-1β and IL-4 were from R&D Systems (Abingdon, UK). The following FITC-conjugated mouse mAb: anti-CD1a, anti-CD14, anti-CD16, anti-CD40, anti-CD80, anti-DC-SIGN and anti-HLA-DR were obtained from Becton-Dickinson, and anti-CD1c mAb from Miltenyi-Biotech (Bergisch Gladbach, Germany)). The PE-conjugated mouse anti-CD11c, anti-CD54, anti-CD80, anti-CD86 and anti-MR mAb as well as the corresponding isotype controls were purchased from Becton-Dickinson, whereas the anti-CD83 and anti-CCR7 mAb were from Coulter (Miami, FL) and R&D Systems, respectively. A PE-conjugated and a purified anti h-ICOS-L mAb (clone HIL131) with a neutralizing activity was a generous gift from Dr. R. Kroczek (Berlin, Germany).

Preparation of human MDDC and naive T cells

Blood monocytes (95% CD14+ cells) and naive Th cells (95% of CD4+CD45RA+ cells) were purified by positive and negative selection, respectively, over a MACS column (Milteny Biotech). Monocytes were then differentiated into DC by standard procedures 18. Briefly, monocytes were cultivated at 106 cells/ml for 7 days in RPMI 1640 with 10% heat-inactivated FCS (Invitrogen, Paisley, UK) containing 10 ng/ml IL-4 and 25 ng/ml GM-CSF. PGD2, BW245C, DK-PGD2 (final concentration 1 µM), forskolin (5 µM) or DMSO (used as a control) were added from the initiation of the culture until day 5. In some cases, GW9662 (1 µM, a PPAR-γ antagonist) or Rp-8Br-cAMPS (25 µM, a selective cAMP antagonist) was added with PGD2 or BW245C, respectively. At day 5, control MDDC (DMSO/DC, 99% pure) showed phenotypic characteristics (CD14, CD68, CD83low, CD1ahigh and HLA-DRhigh) of immature DC. Of note, compared to controls, PGD2, BW245C and DK-PGD2 impacted neither the yield, nor the morphological characteristics, nor the level of CD14 and CD16 (both were nearly undetectable) and CD11c (not shown) in immature DC.

MDDC stimulation

At day 5, DC were stimulated with LPS (1 μg/ml) or with TNF-α and IL-1β (10 and 5 ng/ml, respectively). After 48 h, supernatants were collected for cytokine analysis and MDDC were mechanically detached for FACS analysis.

Flow cytometric analysis

Immature and mature MDDC were collected at day 5 and 7, respectively, in PBS containing 2 mM EDTA at 4°C and labeled as previously described 11. After washing, cells were analyzed on a FACScalibur flow cytometer (Becton Dickinson). Results are expressed as the mean fluorescence intensity (MFI) obtained with specific Ab minus the value obtained with the isotype control (ΔMFI). For analysis of micropinocytosis and Ag processing, DC were resuspended in RPMI plus 5% FCS and incubated for 30 min and 1 h in the presence of 0.1 mg/ml FITC-conjugated dextran (Sigma-Aldrich) or DQ-OVA (Molecular Probes, Leiden, The Nederlands). Cells were incubated for 1 h at 4°C as a negative control.

Chemotaxis assays

Chemotaxis assays were performed using 48-well Boyden microchambers as previously described 11. Briefly, LPS-activated MDDC collected at day 7 were applied to the upper wells of the chamber. In the lower chamber, CCL19 (MIP-3β, 200 ng/ml) or medium alone was added. Each assay was performed in quadruplicate and the results are expressed as the mean ± SD of migrated cells per four fields.

Cytokine assay

The concentrations of cytokines and chemokines in the culture supernatants were determined by ELISA as described by the manufacturers: R&D Systems for the determination of IL-6, TNF-α, CCL1 (I-309), CCL5 (RANTES), CCL22 (MDC) and CXCL10 (IP-10), or Diaclone (Besançon, France) for IL-4, IL-10, IL-12p70 and IFN-γ.

Evaluation of mRNA expression by RT-PCR

Total RNA were obtained using the Trizol reagent (Life Technologies, Grand Island, NY) and reverse-transcribed (Life Technologies). mRNA expression for B7-H1, B7-DC, CD80, CD86, ICOS-L, OX40L, IL-10 and IL-12p40 was evaluated by a two-step semi-quantitative RT-PCR using 6-glyceraldehyde-phosphate dehydrogenase (GAPDH) mRNA as reference. The primer sequences as well as the numbers of cycles have previously been reported 11. After gel electrophoresis and staining with Gel Star (FMC, Rockland, ME), the intensity of each band was measured with Gel Analyst system (Claravision, Orsay, France).

Modulation of naive Th cell priming by DC differentiated in the presence of PG

Immature MDDC differentiated with DMSO alone or with PGD2, BW245C and DK-PGD2 (5×103 cells/well) were pulsed or not with Dpt (1 IRU/ml) as mentioned above. After 6 h incubation, MDDC were washed and 1×105 autologous naive T cells (ratio 1/20) were added to 1 ml of complete medium. T cell proliferation was evaluated at day 5 by incorporation of [3H]thymidine (0.5 µCi/well) for 12 h.

In addition, MDDC were pretreated with SEA at 2 μg/ml in the same conditions and naive T cells were added after 6 h incubation. T cell proliferation was analyzed at day 4 as reported above. T cell polarization was evaluated after restimulation at day 7 with autologous SEA-pulsed MDDC. At day 14, T cells were activated by anti-CD3 plus anti-CD28 mouse mAb for 6 or 24 h. At 6 h, the percentage of CD4+ cells expressing IFN-γ, IL-4 and IL-10 was determined by secretion assays (Miltenyi Biotech). Cytokine release was also quantified after 24 h. In some cases, neutralizing anti-IL-10 (R&D Systems) and anti-ICOS-L Ab (5 and 20 µg/ml, respectively) were added during the co-culture to define the role of both molecules in the T cell polarization.

Statistical analysis

Data from flow cytometry analysis are expressed as the difference of the MFI obtained with either the specific Ab or the isotype control. Results are expressed as means ± SEM. Statistical analysis was performed using of the Wilcoxon test for paired data. p<0.05 was considered as significant.


  1. 1


  2. 2


  3. 3


  4. 4


  5. 5


  6. 6