Expression Profiling Defines ATP as a Key Regulator of Human Dendritic Cell Functions

This information is current as Nathalie Bles, Michael Horckmans, Anne Lefort, Frédéric of September 30, 2021. Libert, Pascale Macours, Hakim El Housni, Frédéric Marteau, Jean-Marie Boeynaems and Didier Communi J Immunol 2007; 179:3550-3558; ; doi: 10.4049/jimmunol.179.6.3550

http://www.jimmunol.org/content/179/6/3550 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2008/03/12/179.6.3550.DC1 Material http://www.jimmunol.org/ References This article cites 36 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/179/6/3550.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 30, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Gene Expression Profiling Defines ATP as a Key Regulator of Human Dendritic Cell Functions1

Nathalie Bles,2* Michael Horckmans,2* Anne Lefort,* Fre´de´ric Libert,* Pascale Macours,† Hakim El Housni,‡ Fre´de´ric Marteau,* Jean-Marie Boeynaems,*† and Didier Communi3*

Extracellular ATP and PGE2 are two cAMP-elevating agents inducing semimaturation of human -derived dendritic cells (MoDCs). We have extensively compared the gene expression profiles induced by adenosine 5؅-O-(3-thiotriphosphate) (ATP␥S) ␥ and PGE2 in human MoDCs using microarray technology. At6hofstimulation, ATP S initiated an impressive expression profile compared with that of PGE2 (1125 compared with 133 genes, respectively) but after 24 h the number of genes regulated by ␥ ATP SorPGE2 was more comparable. Many target genes involved in inflammation have been identified and validated by ␥ quantitative RT-PCR experiments. We have then focused on novel ATP S and PGE2 target genes in MoDCs including CSF-1,

MCP-4/CCL13 , vascular endothelial -A, and neuropilin-1. ATP␥S strongly down-regulated CSF-1 re- Downloaded from ceptor mRNA and CSF-1 secretion, which are involved in monocyte and dendritic cell (DC) differentiation. Additionally, ATP␥S down-regulated several involved in monocyte and DC migration including CCL2/MCP-1, CCL3/MIP-1␣, CCL4/ MIP-1␤, CCL8/MCP-2, and CCL13/MCP-4. Interestingly, vascular endothelial growth factor A, a major angiogenic factor dis- ␥ playing immunosuppressive properties, was secreted by MoDCs in response to ATP S, ATP, or PGE2, alone or in synergy with ␥ LPS. Finally, flow cytometry experiments have demonstrated that ATP S, ATP, and PGE2 down-regulate neuropilin-1, a receptor playing inter alia an important role in the activation of T lymphocytes by DCs. Our data give an extensive overview of the genes http://www.jimmunol.org/ ␥ regulated by ATP S and PGE2 in MoDCs and an important insight into the therapeutic potential of ATP- and PGE2-treated human DCs. The Journal of Immunology, 2007, 179: 3550–3558.

aturation of dendritic cells (DCs)4 in response to LPS, regulation of costimulatory molecules (2–4) and regulates various CD40 ligand, or proinflammatory is reflected chemokines and chemokine receptors (5, 6). ATP also regulates the M by a loss of endocytosis, the surface expression of stable action of LPS and other maturating agents on human DCs by inhib- MHC-peptide complexes and costimulatory molecules (CD80, iting the production of proinflammatory cytokines such as IL-12, IL- CD86), the production of cytokines like IL-12, and a shift in the ex- 1␤, TNF-␣, and IL-6 and by potentiating anti-inflammatory IL-10 by guest on September 30, 2021 pression of chemokines and their receptors allowing DC migration to (2, 7). These features of ATP-treated DCs are compatible with semi- lymphoid organs (1). In human monocyte-derived DCs (MoDCs), mature DCs whose potential involvement in central and peripheral ATP, which is released inter alia from necrotic cells, induces the up- tolerance has been previously discussed (8–10). This profile of action

of ATP is similar to that of cAMP-elevating agents such as PGE2 and is indeed associated with an increase in cAMP, presumably mediated

*Institute of Interdisciplinary Research, Interdisciplinaire en Biologie Humaine et by the P2Y11 receptor (4). ATP, via inhibition of IL-12 and potenti- Mole´culaire, Universite´Libre de Bruxelles, Brussels, Belgium; †Department of Med- ation of IL-10, will thus impair the initiation of a Th1 response and ical Chemistry, Erasme Hospital, Universite´Libre de Bruxelles, Brussels, Belgium; and ‡Department of Genetics, Erasme Hospital, Universite´Libre de Bruxelles, Brus- favor a Th2 response or tolerance (2, 7). Recently, we reported the sels, Belgium critical role of ATP-mediated signal transduction in triggering two Received for publication October 31, 2006. Accepted for publication July 6, 2007. targets (thrombospondin-1 (TSP-1) and IDO) involved in im- The costs of publication of this article were defrayed in part by the payment of page munosuppression, suggesting a potential role of extracellular nucleo- charges. This article must therefore be hereby marked advertisement in accordance tides in immune tolerance (11). with 18 U.S.C. Section 1734 solely to indicate this fact. In the present study, we have extensively compared the target 1 This work was supported by an Action de Recherche Concerte´e of the Communaute´ genes of adenosine 5Ј-O-(3-thiotriphosphate) (ATP␥S) and PGE Franc¸aise de Belgique, by the Belgian Programme on Interuniversity Poles of Attrac- 2 tion initiated by the Belgian State, Prime Minister’s Office, Federal Service for Sci- in human MoDCs using a combination of microarray technology, ence, Technology and Culture, by grants of the Fonds de la Recherche Scientifique quantitative RT-PCR experiments, ELISAs, and flow cytometry Me´dicale, the Fonds Emile DEFAY, and the LifeSciHealth programme of the Euro- pean Community (Grant LSHB-2003-503337). N. B., M. H., and F. M. were sup- analysis. This study is the first one to provide gene expression ␥ ported by the Fonds National de la Recherche Scientifique/Fonds pour la Recherche profiles of ATP S and PGE2 in MoDCs and gives an overview of dans l’Industrie et dans l’Agriculture, Belgium. D. C. and F. L. are Research Asso- the potential actions of ATP and PGE on human DCs. ciate of the Fonds National de la Recherche Scientifique (FNRS). 2 2 N. B. and M. H. contributed equally to the work. Materials and Methods 3 Address correspondence and reprint requests to Dr. Didier Communi, Institute of Reagents Interdisciplinary Research (IRIBHM), Universite´Libre de Bruxelles, Building C (5th floor), Campus Erasme, 808 Route de Lennik, Brussels, Belgium. E-mail address: ATP, ATP␥S, PGE , forskolin, and LPS were obtained from Sigma-Aldrich. [email protected] 2 4 Abbreviations used in this paper: DC, dendritic cell; ATP␥S, adenosine 5Ј-O-(3- Preparation of MoDCs thiotriphosphate); MoDC, monocyte-derived DC; TSP-1, thrombospondin-1; NRP-1, neuropilin-1; qRT-PCR, quantitative RT-PCR; VEGF, vascular endothelial growth PBMCs were isolated from leukocyte-enriched buffy coats of healthy volun- factor. teer donors by standard density gradient centrifugation using Lymphoprep so- lution from Nycomed. PBMCs (2.5 ϫ 108) were allowed to adhere for1hand 2 Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 30 min at 37°C at 5% CO2 in air in 75-cm cell culture flasks. Nonadherent www.jimmunol.org The Journal of Immunology 3551

␥ cells were removed and adherent cells were cultured in 15 ml of RPMI 1640 Table I. Number of genes regulated by ATP S and/or PGE2 at 2, 6, medium supplemented with 800 U/ml GM-CSF and 500 U/ml IL-4. GM-CSF 12, or 24 h in human MoDCsa and IL-4 were also added a second time 2 days after the adhesion step. Five days after the adhesion step, the purity of each cell preparation was evaluated 2h 6h 12h 24h using flow cytometry by analyzing the expression of two markers of DCs, HLA-DR and CD1a. Moreover, the absence of , lymphocytes, and ATP␥S 534 1125 595 413 mature DCs was always checked by staining cell preparation using CD14, PGE2 42 133 117 307 CD3, and CD83 markers, respectively. For our experiments, we have only ATP␥S/PGE 38 119 90 164 used cell preparations of HLA-DRϩCD1aϩ immature DCs displaying at least 2 95% purity. Cells were then plated at 106 cells/ml in 24 multiwells in complete a Evaluated by using microarray technology. medium. Agents were then added for different periods of time. Flow cytometry analysis Cells were labeled with FITC-conjugated anti-human CD83 and PE-con- is more resistant to degradation by ectonucleotidases, was used jugated anti-human CD1a, HLA-DR, CD14, CD3, and neuropilin-1 instead of ATP to avoid additional gene regulations due to its ϫ 5 ␮ (NRP-1) Abs (BD Pharmingen). Cells (2 10 ) were incubated in 100 l degradation products such as ADP and adenosine. PGE2 displays of PBS with 0.1% sodium azide for 30 min in the dark at 4°C, washed with effects similar to those of ATP on DC maturation and is also able 1 ml of PBS, and analyzed on a Cytomics FC 500 flow cytometry system to activate the cAMP pathway in MoDCs (14, 15). Immature (Beckman Coulter). Data were analyzed using CXP cytometry software; 6 ␮ the number of events was at least 10,000. We have checked the purity MoDCs (10 /ml) were stimulated for 2, 6, 12, or 24 h with 100 M ␥ (Ͼ95%) and immaturity of our preparations of DCs by FACS analysis, ATP S or 500 nM PGE2. Microarray experiments have been per- identifying CD1aϩHLA-DRϩCD83ϪCD14ϪCD3Ϫ cells. formed using total RNA extracted from the stimulated and un- Downloaded from RNA isolation and microarray analysis stimulated DCs obtained from two independent donors. After am- plification, the RNAs were labeled to hybridize arrays containing Immature DCs (106 cells/ml) were stimulated by ATP␥S (100 ␮M) or PGE 2 ϳ40,000 human specific oligonucleotides. We first performed a (500 nM) for 2, 6, 12, or 24 h in complete RPMI 1640 medium. RNA was isolated using TRIzol reagent (Invitrogen Life Technologies) and an RNeasy selection of sequences regulated at least 2-fold for at least for one kit column (Qiagen). RNA was reverse transcribed using oligo(dT) primers agent and one time point (see Table S1 in the supplementary da- and ArrayScript (Ambion; Applied Biosystems), labeled and hybridized as ta).5 Among these 3,512 regulated sequences, 1,896 oligonucleo- previously described (12) to Human Exonic Evidence Based Oligonucleotide tides were regulated at least 2-fold by ATP␥S and only 97 oligo- http://www.jimmunol.org/ (HEEBO) arrays containing on average 44,544 human 70-mer oligonucleo- tides (Stanford University, Palo Alto, CA). Data were obtained for four time nucleotides were regulated at least 2-fold by PGE2. From these points using RNA from two independent donors. experiments, it thus appeared that ATP␥S was able to induce a very wide and early gene expression profile in MoDCs. As shown Quantitative RT-PCR (qRT-PCR) experiments in Table I, there was a significant disproportion between the num- ␥ For each target gene, primers were selected using Primer Express 2.0 soft- ber of genes regulated by ATP S and by PGE2 at a short time of ware (PCR product size: 100–150 bp; primer size: 20–25 bp; Tm: 58°C to stimulation (2 h and 6 h). At 24 h of stimulation, the number of 60°C). Several control genes were tested for their stability in our system ␥ (YWHAZ, B2M, RPL13A, and SDHA) (13). Two of these control genes genes commonly regulated by ATP S and PGE2 was the highest

(B2M and SDHA) were selected after analysis using the geNorm program. (Table I). by guest on September 30, 2021 RT-PCR amplification mixtures (25 ␮l) contained 2 ng of template cDNA, We then identified 57 regulated genes playing a role in inflam- Power SYBR Green PCR Master Mix (12.5 ␮l) (Applied Biosystems), and mation, including chemokines, , CD markers, recep- 200 nM forward and reverse primer. Reactions were run on a 7500 Fast tors, and glycoproteins. Among these genes, 26 sequences were Real-Time PCR System (Applied Biosystems). The cycling conditions ␥ were 10 min for polymerase activation at 95°C and 40 cycles at 95°C for regulated at least 2-fold by ATP S only (Table II) and 31 se- ␥ 15 s and 60°C for 60 s. Mean Ϯ SD values were obtained for each gene quences were regulated at least 2-fold by both ATP S and PGE2 using qBase software. Each assay was performed in duplicate for two in- (Table III). dependent donors. ␥ ELISA Common ATP S and PGE2 target genes in human MoDCs Immature DCs were stimulated by different agents for 24 h at 106 cells/ml First, our microarray data were consistent with the effects of ␥ ␥ in 24 multiwells. DC supernatants were collected and CSF-1, vascular ATP S and PGE2 on DC maturation (2–4, 6, 14, 15). ATP S and endothelial growth factor (VEGF)-A, and CCL13 were measured by PGE2 both regulated genes encoding chemokine receptors such as ELISA using commercially available kits from R&D Systems. CXCR4, CCR5, and CCR1 (Table III). Moreover, ATP␥S up-reg- Kynurenine measurements ulated CD83, which is a surface marker specifically up-regulated upon DC maturation (3, 16). Many genes encoding chemokines DCs were stimulated with IFN-␥ at 100 U/ml alone or in combination with ␥ ␮ ␮ (e.g., CCL2, CCL3, CCL4, and CCL22), glycoproteins (e.g., TSP- ATP Sat100 MorPGE2 at 5 M for 24 h. Cells were then washed and resuspended in red phenol-free complete medium supplemented with 300 1), IL-7 and IL-15 receptors (Table III), and many other known or ␮M L-tryptophan. After 5 h, supernatants were collected and kynurenine unknown genes (see supplemental data) were regulated in a similar concentration was quantified by HPLC. Culture supernatants (400 ␮l) were way by ATP␥S and PGE and more particularly at 24 h of ␮ 2 extracted with 80 l of 10% trichloroacetic acid, the precipitate was re- stimulation. moved by centrifugation, and the supernatant was diluted in the initial mobile phase composed of deionized water, 5% (v/v) methanol, 1% (v/v) ␥ acetic acid, and 5 mM hexane sulfonic acid. Samples were injected onto an ATP S regulates a large panel of genes not regulated by PGE2 Atlantis dC18 reverse phase column (4.6 ϫ 300 mm, 3 ␮m; Waters) and in MoDCs eluted with a linear gradient of methanol (5–40% over 35 min) at 1 ml/min. The large expression profile of ATP␥S included genes encoding Absorbance was measured at 370 nm and compared against a standard curve of L-kynurenine. chemokines (CCL7 and CCL24), interleukins (IL-1A and IL-16), CD markers (e.g., CD55, CD69, and CD72) and receptors (eryth- Results ropoietin and CSF-1 receptors) (Table II), and a large number of ␥ Comparison of ATP S and PGE2 gene expression profiles other known or unknown genes (Table II and Table S1 in the in MoDCs online supplemental material). First, we obtained the gene expression profiles of ATP␥S and ␥ 5 PGE2 in MoDCs by using microarray technology. ATP S, which The online version of this article contains supplemental material. 3552 GENE PROFILING DEFINES ATP AS A KEY REGULATOR OF DCs

Table II. List of genes regulated by ATP␥S only and involved in inflammationa

Ratios

␥ ␥ ␥ ␥ ATP S ATP S ATP S ATP S PGE2 PGE2 PGE2 PGE2 (2 h) (6 h) (12 h) (24 h) (2 h) (6 h) (12 h) (24 h)

Name Symbol Unigen P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2

Chemokines/chemokine receptors Chemokine (C-C motif) CCL24 Hs.247838 0.93 0.85 1.60 0.86 0.52 0.52 0.43 0.45 1.16 1.07 1.88 1.05 1.03 0.87 0.75 0.60 ligand 24 Chemokine (C-C motif) CCL7 Hs.251526 0.43 0.52 0.30 0.28 0.63 0.71 1.11 0.99 1.10 0.92 1.05 0.72 1.01 0.86 1.15 1.04 ligand 7

Interleukins/ receptors IL-15 Hs.168132 1.03 1.61 1.12 1.09 2.83 2.23 6.03 2.53 0.89 1.30 0.98 1.13 1.22 1.45 3.81 1.90 Interleukin 1 receptor, type II IL-1R2 Hs.25333 0.68 0.74 0.41 0.34 0.19 0.29 0.23 0.46 0.79 0.84 0.86 0.75 0.77 0.66 0.53 0.78 IL-16 Hs.459095 0.41 0.68 0.17 0.36 0.25 0.61 0.63 0.80 0.71 1.02 0.77 0.95 1.07 0.84 0.74 0.92 Interleukin 1, ␣ IL-1A Hs.1722 0.32 0.43 1.17 0.72 1.18 1.33 1.45 1.06 1.11 1.16 1.36 1.05 1.10 1.60 0.90 0.92

CD markers CD72 molecule CD72 Hs.116481 5.77 2.18 6.72 3.23 2.33 1.39 2.14 1.52 1.56 1.12 1.25 0.96 0.99 0.98 1.98 1.34 CD55 molecule CD55 Hs.527653 2.86 2.71 7.48 2.95 3.54 2.37 0.98 1.84 1.20 1.12 1.40 1.04 1.05 1.25 0.62 1.30 CD69 molecule CD69 Hs.208854 2.80 2.25 2.58 1.31 0.70 1.15 0.60 0.93 1.94 1.09 1.97 1.09 0.92 1.07 0.80 0.96 CD109 molecule CD109 Hs.399891 1.98 1.93 3.26 4.58 2.56 3.43 0.92 2.19 0.89 0.97 1.04 1.12 1.04 1.10 0.68 1.17 Downloaded from CD8a molecule CD8A Hs.85258 1.62 1.59 2.49 2.03 1.34 1.39 1.22 1.14 1.21 1.23 1.22 1.32 1.08 1.22 1.01 1.02 CD83 molecule CD83 Hs.484703 1.58 1.55 3.55 1.57 3.33 2.15 2.02 2.23 0.95 0.90 1.46 0.87 1.69 1.31 1.11 1.47 CD33 molecule CD33 Hs.83731 1.14 0.96 0.52 0.43 0.22 0.31 0.44 0.44 1.07 0.91 0.95 0.89 0.85 0.69 0.79 0.60 CD40 molecule CD40 Hs.472860 0.88 0.96 0.51 0.48 1.81 0.73 1.26 0.73 1.07 1.04 1.17 0.68 1.52 0.75 0.71 0.76

Receptors Erythropoietin receptor EPOR Hs.631624 4.23 4.70 1.17 1.13 0.99 0.79 0.93 0.84 1.10 1.14 0.96 0.91 0.99 0.82 1.17 0.85 Colony stimulating factor 1 CSF1R Hs.483829 1.16 0.88 0.37 0.45 0.31 0.60 1.01 1.54 1.08 0.95 0.82 1.05 1.00 1.04 1.81 1.97 receptor http://www.jimmunol.org/ Toll-like receptor 1 TLR1 Hs.575090 0.17 0.48 0.13 0.28 0.59 0.82 0.49 0.87 0.66 0.80 1.03 0.93 0.85 1.05 0.85 1.02

Enzymes Matrix metallopeptidase 9 MMP-9 Hs.297413 2.54 1.87 3.04 1.78 1.84 1.23 1.08 0.77 1.42 1.10 1.36 0.81 1.56 1.01 0.73 0.82 Superoxide dismutase 2, SOD2 Hs.487046 2.24 2.02 3.80 2.67 2.67 2.73 2.52 1.52 1.71 1.39 1.47 1.30 1.32 1.85 1.95 1.81 mitochondrial Tryptophanyl-tRNA synthetase WARS Hs.497599 1.31 1.93 2.10 2.11 1.10 1.44 0.55 0.81 0.84 1.22 1.02 1.15 0.95 1.10 0.56 0.79

Others Thyrotropin-releasing hormone TRH Hs.182231 13.20 13.71 18.77 9.48 9.45 9.06 5.66 6.63 1.00 1.08 1.01 1.02 0.99 1.11 1.08 0.97 Forkhead box O3A FOXO3A Hs.220950 3.28 2.19 4.12 4.60 1.55 1.33 1.32 1.55 1.48 1.41 1.34 1.61 1.05 1.26 1.53 1.73

Complement component 3 C3 Hs.529053 2.80 1.93 2.32 3.01 1.19 1.44 0.59 0.53 1.68 1.20 1.00 1.01 1.03 0.67 1.38 0.54 by guest on September 30, 2021 Peroxiredoxin 2 PRDX2 Hs.631612 2.26 1.64 2.49 2.43 2.06 2.12 1.69 1.68 1.31 1.08 1.10 1.18 1.24 1.14 1.92 1.14 Neutrophil cytosolic factor 2 NCF2 Hs.587558 0.98 1.35 0.24 0.40 0.33 0.74 0.72 0.77 1.16 1.16 1.03 0.89 0.92 0.90 0.83 0.92 Macrophage expressed gene 1 MPEG1 Hs.643518 0.67 0.82 0.43 0.39 0.21 0.27 0.35 0.45 0.75 0.85 0.72 0.87 0.95 1.73 0.75 1.96

a Ratios were obtained for two independent preparations (P1 and P2) of MoDCs.

Validation of target genes using quantitative down-regulations were confirmed in our microarray experiments

PCR experiments and were also observed in response to PGE2 (Table III). Addition- We have validated the regulation of several promising target ally, we have observed down-regulation of genes encoding CCL4/ ␤ ␥ genes displaying a link with the immune system using quanti- MIP-1 and CCL8/MCP-2 chemokines in response to ATP S and tative PCR experiments. SYBR Green experiments have been PGE2 (Table III). performed for CCL2, CCL3, CCL4, CCL13, CCR5, CD36, Our microarray and quantitative PCR data additionally re- ␥ NRP-1, THBS-1, VEGF-A, INDO, CSF-1, and CSF-1R. The vealed the significant down-regulation of CCL13 by ATP S and primer sequences for these 12 genes are listed in Table IV. The to a lesser extent by PGE2 (Table III), which was confirmed by data have been obtained on two independent preparations of quantitative PCR (Fig. 1). CCL2, CCL3, CCL4, CCL8, and DCs and there was a good correlation with our microarray data CCL13 are all recruiters of monocytes and immature DCs. ␥ (Fig. 1). Quantitative PCR experiments confirmed that VEGF-A, Their concomitant down-regulation in response to ATP S could CCL13, CSF-1 and NRP-1 were effectively regulated by both be correlated with the reduced capacity of adenine nucleotide- ␥ treated MoDCs to recruit monocytes and DCs (5). We have ATP S and PGE2 at least at one stimulation time point. Even if ATP␥S expression profile was very large, we decided to observed by ELISA using DC supernatants that ATP, PGE2, ␥ focus our attention on four genes commonly regulated by ATP␥S and more strongly ATP S inhibit CCL13 release by MoDCs both in basal and LPS-stimulated conditions (Fig. 2). We have and PGE2 such as CSF-1, VEGF-A, CCL13/MCP-4 chemokine, and NRP-1 because their regulation was not yet reported, unex- also shown that forskolin inhibited CCL13 release both in the pected, and promising. Some particular genes regulated by ATP␥S absence or the presence of LPS at 100 ng/ml (Fig. 2). such as FOXO3A transcription factor and SOD2 (Table II) have ␥ also retained our attention. ATP S strongly down-regulates CSF-1 and CSF-1 receptor ␥ ATP S is a potent negative signal for the secretion of several An interesting observation was the down-regulation of CSF-1 and chemokines by MoDCs the gene encoding its receptor CSF-1R in response to ATP␥S (Ta- ␥ We have previously shown that ATP S down-regulated the secre- bles II and III). PGE2 was able to down-regulate CSF-1 cytokine tion of chemokines such as CCL2 and CCL3 by MoDCs (5). These but not its receptor (Table II). These regulations were first a Immunology of Journal The Table III. List of genes regulated by both ATP␥S and PGE2 and involved in inflammation

Ratios

ATP␥S(2h) ATP␥S(6h) ATP␥S (12 h) ATP␥S (24 h) PGE2 (2 h) PGE2 (6 h) PGE2 (12 h) PGE2 (24 h)

Name Symbol Unigen P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2 P1 P2

Chemokines/chemokine receptors Chemokine (C-X-C motif) receptor 4 CXCR4 Hs.421986 6.41 7.68 12.57 12.76 8.44 7.91 5.58 6.50 2.90 2.39 4.84 4.92 5.55 5.55 5.42 4.70 Chemokine (C-C motif) receptor 7 CCR7 Hs.370036 1.43 1.40 3.70 2.29 9.53 6.61 16.30 10.77 0.93 1.00 1.32 1.01 2.86 1.93 6.71 3.27 Chemokine (C-X-C motif) ligand 7 CXCL7 Hs.2164 1.27 1.26 3.29 19.77 3.26 16.98 2.10 10.66 1.42 1.18 2.98 11.12 3.87 15.41 8.39 9.22 Chemokine (C-C motif) ligand 22 CCL22 Hs.534347 0.88 1.09 1.30 2.98 3.61 4.20 4.96 2.71 1.16 1.39 1.36 1.58 3.26 1.94 2.99 2.13 Chemokine (C-C motif) ligand 13 CCL13 Hs.414629 0.80 0.91 0.59 0.60 0.24 0.46 0.42 1.07 1.04 1.07 0.78 0.92 0.50 0.89 0.85 1.30 Chemokine (C-C motif) ligand 4 CCL4 Hs.75703 0.36 0.80 0.47 0.30 0.47 0.40 0.12 0.52 1.01 0.55 0.63 0.29 0.35 0.50 0.18 0.33 Chemokine (C-C motif) receptor 5 CCR5 Hs.450802 0.34 0.60 0.12 0.27 0.29 0.77 0.46 0.67 0.73 0.90 0.44 0.53 0.53 0.58 0.35 0.68 Chemokine (C-C motif) ligand 3 CCL3 Hs.514107 0.26 0.36 0.46 0.29 0.41 0.39 0.10 0.34 0.82 0.31 0.62 0.26 0.48 0.35 0.20 0.23 Chemokine (C-C motif) receptor 1 CCR1 Hs.301921 0.13 0.44 0.06 0.25 0.26 0.80 0.34 0.73 0.60 0.72 0.52 0.67 0.43 0.64 0.32 0.64 Chemokine (C-C motif) ligand 2 CCL2 Hs.303649 0.06 0.16 0.05 0.11 0.15 0.34 0.57 0.38 0.48 0.67 0.30 0.38 0.58 0.63 0.41 0.57 Chemokine (C-C motif) ligand 8 CCL8 Hs.271387 0.15 0.53 0.14 0.19 0.47 1.21 0.63 0.85 0.53 0.48 0.45 0.33 0.67 1.59 0.69 1.39

Interleukins/interleukin receptors Interleukin 15 receptor, ␣ IL-15RA Hs.524117 3.06 2.55 2.28 2.36 4.14 2.25 3.68 2.04 2.69 2.10 1.72 1.80 2.35 1.86 3.07 2.13 IL-10 Hs.193717 1.77 3.08 2.19 1.58 0.84 1.33 1.58 2.32 1.27 1.73 1.84 1.79 1.26 1.86 1.69 2.34 receptor IL-7R Hs.591742 0.95 1.13 5.35 7.06 4.35 10.79 4.42 16.48 1.87 2.36 2.62 4.23 3.18 8.17 2.55 7.01 IL-8 Hs.624 0.90 2.60 3.69 3.06 1.91 5.95 2.33 3.15 1.14 1.26 2.04 1.49 1.02 4.20 2.05 2.13

CD markers CD300E molecule CD300E Hs.158954 2.33 1.53 7.80 4.88 2.46 1.88 1.05 1.27 1.39 0.99 2.37 3.21 1.46 1.89 1.10 1.35 CD86 molecule CD86 Hs.171182 1.47 1.40 1.57 2.14 1.94 2.57 1.71 2.46 1.41 1.16 1.44 1.88 1.32 1.94 1.56 2.00

Receptors Neuropilin-1 NRP-1 Hs.131704 0.55 1.02 0.14 0.23 0.26 0.40 0.26 0.63 0.83 1.03 0.69 0.68 0.62 0.70 0.33 0.90

Enzymes Indoleamine-pyrrole 2.3 dioxygenase INDO Hs.840 2.59 1.37 15.88 6.84 18.52 11.01 9.51 14.47 1.90 1.06 4.71 3.30 6.04 5.70 9.38 8.36 Matrix metallopeptidase 19 MMP-19 Hs.591033 3.40 5.10 4.69 5.64 2.99 4.37 1.25 2.04 1.95 1.36 1.58 1.55 1.67 1.57 0.86 1.70

Glycoproteins Amphiregulin AREG Hs.270833 123.04 125.02 81.23 74.96 8.17 28.66 2.68 8.44 7.82 7.87 7.08 8.50 1.49 4.85 2.01 2.11 Thrombospondin-1 THBS-1 Hs.164226 32.29 28.58 37.56 124.67 14.99 37.87 8.47 55.87 9.69 7.52 13.82 42.58 6.97 44.72 13.45 86.88

Cytokines Vascular endothelial growth factor VEGF Hs.73793 3.36 4.68 5.99 7.13 3.77 4.56 1.02 3.03 1.03 1.07 1.41 1.47 1.00 1.80 0.77 2.02 Colony stimulating factor-1 (macrophage) CSF-1 Hs.591402 0.85 1.04 0.75 0.78 0.70 0.92 0.31 0.66 0.92 0.86 0.86 0.79 0.93 0.68 0.39 0.61 TNF Hs.241570 0.28 0.19 0.60 0.34 0.58 0.51 0.34 0.60 0.53 0.36 0.63 0.34 0.92 0.65 0.38 0.48

Others Epiregulin EREG Hs.115263 23.52 16.13 45.22 27.19 5.25 5.51 1.63 10.17 2.31 2.08 2.66 4.98 1.02 7.23 2.11 16.78 B cell translocation gene 1, anti-proliferative BTG1 Hs.255935 8.88 8.33 11.58 13.56 5.76 4.84 6.25 5.59 2.41 1.96 2.89 3.79 2.27 4.25 3.78 4.57 Cystatin F (leukocystatin) CST7 Hs.143212 5.05 3.66 8.53 7.93 8.99 8.54 4.85 9.88 2.73 1.63 3.53 3.60 6.86 8.29 6.57 8.12 Transforming growth factor, ␣ TGF-A Hs.170009 3.06 2.36 6.54 14.01 6.23 11.03 1.98 7.70 1.91 1.62 3.71 5.10 3.18 5.26 2.22 4.71 regulatory factor-4 IRF-4 Hs.401013 2.52 7.72 3.14 4.11 4.19 5.30 3.41 5.09 1.15 1.32 1.60 1.47 1.56 1.55 1.56 2.29

TNF receptor-associated 1 TRAP-1 Hs.30345 1.22 1.01 0.70 0.74 0.48 0.49 0.72 0.59 0.87 1.10 1.08 0.88 0.76 0.53 0.50 0.50 3553

a Ratios were obtained for two independent preparations (P1 and P2) of MoDCs.

Downloaded from from Downloaded http://www.jimmunol.org/ by guest on September 30, 2021 30, September on guest by 3554 GENE PROFILING DEFINES ATP AS A KEY REGULATOR OF DCs

Table IV. List of the specific primers used for the SYBR Green experiments

Gene Forward Primer Reverse Primer

THBS1 5Ј-GCTGGTGGTAGACTAGGGTTGTTT-3Ј 5Ј-CCAGAAGGTGCAATACCAGCAT-3Ј VEGF-A 5Ј-TGCTGTCTTGGGTGCATTG-3Ј 5Ј-TGATTCTGCCCTCCTCCTTCT-3Ј NRP-1 5Ј-GTGACCACTGGAAGGAAGGG-3Ј 5Ј-CAGCAATCCCACCAAGGTTT-3Ј CSF-1 5Ј-GATAACACCCCCAATGCCATC-3Ј 5Ј-CAGGCCTTGTCATGCTCTTCA-3Ј CD36 5Ј-GACAACACAGTCTCTTTCCTGCAG-3Ј 5Ј-GCCACAGCCAGATTGAGAACT-3Ј CSF-1R 5Ј-GAGCGGACTATACCAATCTGCC-3Ј 5Ј-AGCAGGTCAGGTGCTCACTAGAG-3Ј CCL2 5Ј-CACCAATAGGAAGATCTCAGTGCA-3Ј 5Ј-TGGCCACAATGGTCTTGAAG-3Ј CCL3 5Ј-CCAGTTCTCTGCATCACTTGCT-3Ј 5Ј-CTGCTCGTCTCAAAGTAGTCAGCTA-3Ј CCL4 5Ј-TCTCCTCATGCTAGTAGCTGCCTT-3Ј 5Ј-GCTTCCTCGCAGTGTAAGAAAAG-3Ј CCR5 5Ј-ATGACGCACTGCTGCATCAA-3Ј 5Ј-GAAGCGTTTGGCAATGTGCT-3Ј CCL13 5Ј-ACATGAAAGTCTCTGCAGTGCTTC-3Ј 5Ј-AGTAGATGGGACGTTGAGTGCAT-3Ј INDO 5Ј-CCATATTGATGAAGAAGTGGGCT-3Ј 5Ј-GATCAGGCAGATGTTTAGCAATGA-3Ј FOXO3A 5Ј-GCAAAGCAGACCCTCAAACTG-3Ј 5Ј-GCGTGGGATTCACAAAGGTG-3Ј SOD2 5Ј-ACCTCAGCCCTAACGGTGGT-3Ј 5Ј-CAGCCGTCAGCTTCTCCTTAAA-3Ј B2Ma 5Ј-TGCTGTCTCCATGTTTGATGTATCT-3Ј 5Ј-TCTCTGCTCCCCACCTCTAAGT-3Ј SDHAa 5Ј-TGGGAACAAGAGGGCATCTG-3Ј 5Ј-CCACCACTGCATCAAATTCATG-3Ј

a Control gene. Downloaded from confirmed using quantitative PCR experiments (Fig. 1). To regulate IDO in DCs both in vitro (20) and in vivo (21). We have quantify CSF-1 release, we have performed ELISAs using su- observed a stronger INDO up-regulation in MoDCs in response to ␥ ␥ pernatants of DCs treated for 24 h with ATP S, ATP, or PGE2. ATP S than in response to PGE2 in microarray (Table III). Ad- ␥ We have observed that ATP S, ATP, and PGE2 induce an in- ditionally, several genes potentially linked to IDO up-regulation hibition of CSF-1 release in the absence or the presence of LPS and tryptophan metabolism such as genes encoding the superoxide http://www.jimmunol.org/ (100 ng/ml) (Fig. 3). dismutase SOD2 and the FOXO3a transcription factor were up- regulated in response to ATP␥S (Table II). Up-regulation of ␥ MoDCs release VEGF-A in response to ATP S, ATP and PGE2 FOXO3a and SOD2 has been related to a potential alternative In our microarray and qRT-PCR experiments, VEGF was a gene pathway that up-regulates IDO in response to CTLA4 indepen- ␥ ␥ up-regulated by ATP S and PGE2 (Table III and Fig. 1). We have dently of IFN- increase (22). We observed that SOD2 and used an ELISA kit specific for the most common form of VEGF FOXO3A mRNAs were up-regulated at6hinresponse to ATP␥S called VEGF-A. VEGF-A protein levels have been quantified in but not in response to PGE2 or forskolin by quantitative PCR anal- supernatants of MoDCs treated during 24 h by ATP, ATP␥S, and ysis (Table V).

␥ ␥ by guest on September 30, 2021 PGE2 in the presence or the absence of LPS (Fig. 4). ATP, ATP S, We then compared the effect of ATP S and PGE2 on the gen- and PGE2 were able to induce a moderate secretion of VEGF-A. eration of kynurenine derivatives in the absence or the presence of Whereas LPS induced only a weak VEGF-A secretion, it was able IFN-␥ (100 U/ml). As shown in Fig. 6, ATP␥S was able to po- ␥ to significantly potentiate the production of VEGF-A induced by tentiate IFN- effect as previously described (11) whereas PGE2 ␥ Ͼ ATP, ATP S, and PGE2. VEGF-A concentration was 2 ng/ml in displayed a weak effect alone but was not able to potentiate an the supernatant of DCs treated with a combination of 100 ng/ml IFN-␥ response. These data suggest that ATP␥S could also stim- ␥ ␮ LPS and ATP S (100 M) or PGE2 (500 nM). No VEGF-C pro- ulate IDO activity in MoDCs through an alternative IDO induction duction was detected using the same DC supernatants (data not pathway as described for CTLA4 (22). shown). Discussion ATP␥S and PGE down-regulate NRP-1 expression on MoDCs 2 The present study combines gene profiling, qRT-PCR, ELISA, and One of the down-regulations observed in response to ATP␥S was flow cytometry experiments to identify and compare target genes ␥ that of NRP-1 (Table III). In addition to its known functions in of ATP S and PGE2 in human MoDCs. Several papers have pre- axon guidance (17) and angiogenesis (18) as a semaphorin and viously described the effects of ATP on the expression of DC VEGF coreceptor, NRP-1 also plays a key role in the initiation of maturation markers and some chemokines and chemokine recep- the primary immune response by regulating interactions between tors (2–6). ATP also regulates the action of LPS on human DCs by DCs and T cells (19). We have confirmed down-regulation of inhibiting the production of proinflammatory cytokines like IL-12, NRP-1 mRNA in response to ATP␥S and more weakly in response IL-1␤, IL-6, and TNF-␣ and by potentiating anti-inflammatory to PGE2 by quantitative PCR experiments (Fig. 1). Flow cytometry IL-10 (2, 4, 7). These effects of ATP are reproduced by other experiments have been performed using PE-conjugated anti- cAMP-elevating agents such as PGE2 (14, 15) and are most prob- human NRP-1 Ab. We have observed a strong down-regulation of ably mediated by the P2Y11 receptor (4). PGE2 effects on DC NRP-1 expression at the membrane of MoDCs in response to maturation are likely to result from activation of the EP2/EP4 re- ATP␥S (Fig. 5). Weaker effects were observed in response to ATP ceptors (14). Many of the effects of ATP␥S on MoDCs, such as its and PGE2 (Fig. 5). effect on CD83 expression and TSP-1 release, are reproduced by forskolin or dibutyryl-cAMP, demonstrating their action through ␥ ATP S regulated expression of genes related to the tryptophan cAMP elevation (4, 5, 11, 14). metabolism and IDO activation It appeared necessary and useful to follow and compare the ␥ ␥ We have previously reported that ATP and ATP S significantly expression of all of the target genes of ATP S and PGE2 simul- potentiate the activity of IDO, a negative regulator of T lympho- taneously in MoDCs. Microarray technology was a good option for cyte proliferation, and kynurenine production from tryptophan ini- performing this study and revealed an impressive expression pro- ␥ ␥ tiated by IFN- in human DCs (11). PGE2 was reported to up- file of ATP S compared with that of PGE2, especially at a short The Journal of Immunology 3555 Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021

␥ FIGURE 1. Quantitative RT-PCR data obtained for 12 genes regulated by ATP S and/or PGE2 and involved in inflammation. Ratios were obtained for ␥ Ϯ ATP S(—) and PGE2 (- - -) at 2, 6, 12, and 24 h for two independent preparation of MoDCs (means SEM) using SYBR Green technology. mRNA ␥ expression in ATP S- or PGE2-treated cells and untreated cells has been normalized for each gene and each time point using two housekeeping genes (B2M ␥ and SDHA). Ratios were calculated comparing normalized expression of each gene in ATP S- or PGE2- treated DCs to its normalized expression in untreated DCs.

␥ time of stimulation. At 24 h of stimulation, ATP S and PGE2 We then performed a validation of microarray analysis by using regulated a significant pool of common genes including regula- qRT-PCR experiments for 12 target genes that displayed a link tions compatible with their effect on DC maturation. The compar- with the immune system. The expression profile of ATP␥S was ison between the genes regulated by ATP␥S and by PGE2 iden- very large and provided a large series of novel target genes that tified the regulation of many ATP target genes that could not will be studied in the future. We decided to focus our attention on depend on cAMP increase. The large and early expression profile unexpected and promising target genes: four genes regulated by ␥ ␥ of ATP S results most likely from the activation of P2Y11 receptor ATP S and more weakly by PGE2 such as NRP-1, VEGF-A, and other purinergic receptors coupled to calcium-dependent path- CCL13, and CSF-1 and some particular genes regulated more spe- ␥ ways (e.g., P2Y2 and P2X7) (23, 24). The use of forskolin allows cifically by ATP S such as FOXO3A transcription factor and su- us to confirm the involvement of the cAMP pathway in the regu- peroxide dismutase SOD2. ␥ lation of genes regulated by ATP S and PGE2, such as CCL13,as We observed CSF-1 and CSF-1R down-regulation in response we had previously shown for other target genes including CCL2 to ATP␥S in our microarray and qRT-PCR experiments. CSF-1 is and CCL3 (5) as well as TSP-1 (11). known to modulate the development and immune function of DCs 3556 GENE PROFILING DEFINES ATP AS A KEY REGULATOR OF DCs

␥ FIGURE 4. Effects of ATP, ATP S ,and PGE2 on VEGF-A release by human MoDCs. DCs were stimulated by ATP (300 ␮M), ATP␥S (100 ␥ FIGURE 2. Effects of ATP, ATP S, PGE2, and forskolin on CCL13/ ␮ M), or PGE2 (500 nM) in the absence or the presence of LPS (100 ng/ml) MCP-4 release by human MoDCs. DCs were stimulated by ATP (300 for 24 h. Supernatants of treated DCs were collected for ELISA measure- ␮ ␥ ␮ ␮ Downloaded from M), ATP S (100 M), PGE2 (500 nM), or forskolin (FK; 10 M) in the ments of human VEGF-A. Results are expressed as picograms per 106 absence or presence of LPS (100 ng/ml) for 24 h. Supernatants of treated cells/ml and represent the mean Ϯ SEM of three independent experiments. p Ͻ 0.001). Student’s ,ءءء ;p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء) .DCs were collected for ELISA measurements of human CCL13/MCP-4. CONT, Control 6 Results are expressed as picograms per 10 cells/ml and represent the t test were performed using GraphPad Prism. p Ͻ ,ء) .mean Ϯ SEM of three independent experiments. CONT, Control p Ͻ 0.001). Student’s t test were performed using ,ءءء ;p Ͻ 0.01 ,ءء ;0.05 GraphPad Prism.

lated with the reduced capacity of adenine nucleotide-treated DCs http://www.jimmunol.org/ to attract monocytes and immature DCs (5). Inhibition of CSF-1 and CCL13 release in response to ATP could reduce monocyte and but also the survival, proliferation, and differentiation of mononu- DC differentiation and recruitment at the site of inflammation. clear phagocytes (25). The transcription of the CSF-1 receptor We have also shown that ATP, ATP␥S, and PGE induced a CSF-1R is inactive in precursors of DCs and up-regulated in DCs 2 VEGF-A secretion by MoDCs that was strongly potentiated in during differentiation (26). Reduction of CSF-1 secretion and combination with LPS. Released VEGF-A concentrations were CSF-1R expression on DCs has been associated with a loss of sufficient to activate known VEGF receptors. DCs that matured in proliferative response as well as a loss of their phagocytic and the presence of anti-inflammatory molecules such as PGE , IL-10, adhesive properties (27). 2 and calcitriol have been previously reported to secrete VEGF-A by guest on September 30, 2021 Besides the strong down-regulation of the CCL2, CCL3, CCL4, ␥ and CCL8 chemokines by ATP S and PGE2 that was observed our ␥ in microoarray data, ATP S, ATP, and PGE2 were able to inhibit CCL13 release from MoDCs. CCL13 and these other chemokines are all involved in the recruitment of monocytes, immature DCs, NKs, and activated lymphocytes (28, 29). The down-regulation of ␥ CCL13 release in response to ATP S and PGE2 could be corre-

␥ ␥ FIGURE 3. Effects of ATP, ATP S, and PGE2 on CSF-1 release by FIGURE 5. Effects of ATP, ATP S, and PGE2 on NRP-1 expression on human MoDCs. DCs were stimulated by ATP (300 ␮M), ATP␥S (100 human MoDC cells. DCs were stimulated by ATP (300 ␮M), ATP␥S (100 ␮ ␮ M), or PGE2 (500 nM) in the absence or the presence of LPS (100 ng/ml) M), or PGE2 (500 nM) for 24 h. NRP-1 expression was analyzed by flow for 24 h. Supernatants of treated DCs were collected for ELISA measure- cytometry analysis using an anti-human PE-NRP-1. The flow cytometry ments of human CSF-1. Results are expressed as picograms per 106 data in A were obtained in an experiment representative of five independent cells/ml and represent the mean Ϯ SEM of five independent experiments. experiments. In B is displayed the mean of fluorescence Ϯ S.D. obtained from p Ͻ ,ءء ;p Ͻ 0.05 ,ء) .p Ͻ 0.001). Student’s t these five independent experiments. CONT, Control ,ءءء ;p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء) .CONT, Control .p Ͻ 0.001). Student’s t test were performed using GraphPad Prism ,ءءء ;test were performed using GraphPad Prism. 0.01 The Journal of Immunology 3557

Table V. Quantitative RT-PCR data obtained for FOXO3A and SOD2 of the FOXO3A transcription factor and superoxide dismutase ␥ a in response to ATP S, PGE2, or forskolin SOD2 genes in response to ATP␥S. Quantitative PCR experiments have confirmed their regulation in response to ATP␥S, whereas Mean SEM PGE2 and forskolin had no significant effect. FOXO3A and SOD2 FOXO3A are associated with an alternative pathway of IDO activation de- ATP␥S 5.59 1.02 scribed in response to CTLA4 (22). It has been proposed that

PGE2 1.38 0.04 CTLA4-Ig up-regulates IDO through an IFN-␥-dependent path- Forskolin 0.71 0.21 way and an IFN-␥-independent pathway involving FOXO3A and SOD2 up-regulations coupled with peroxynitrite down-regulation SOD2 ␥ ATP␥S 2.05 0.23 (22). FOXO3A and SOD2 up-regulation could explain why ATP S ␥ PGE2 1.38 0.34 but not PGE2 potentiated IFN- -mediated up-regulation of IDO Forskolin 1.49 0.10 and kynurenine production. As there has been a proposal to explain a ␣ Mean Ϯ SEM of ratios were obtained for two independent preparations of the synergy between PGE2 and TNF- on IDO activity (20), our MoDCs using SYBR Green technology. MoDCs were stimulated for 6 h with ATP␥S experiments suggest a synergy between the signaling pathways ␮ ␮ (100 M), PGE2 (500 nM), or forskolin (10 M). induced by the activation of ATP and IFN-␥ receptors. ␥ The identification and study of ATP S and PGE2 target genes (30). This regulation may provide a link between DCs and angio- extend the scope and give a large overview of ATP and PGE2 genesis, but VEGF-A secreted by DCs could also be able to inhibit effects on human MoDCs. Apart from its effect on DC maturation

␥ Downloaded from T cell development and may contribute to tumor-induced immune markers, ATP S regulated a lot of interesting target genes in suppression (31). Interestingly, we have previously reported that MoDCs. First, ATP might be considered as a strong negative sig- the most up-regulated gene in the expression profile induced by nal for chemokine secretion by DCs. ATP is also able to down- ␥ regulate the expression of membrane receptors involved in primary ATP S and PGE2 is the gene THBS1 encoding TSP-1 (11), a pro- tein that displays antiangiogenic properties but is also able to immune response and DC functions such as NRP-1 and CSF-1R. down-regulate CD4ϩ T cell proliferation and behave as an auto- Additionally, this is the first time that VEGF-A secretion by human crine inhibitor of IL-12 release by DCs (32). Finally, another DCs in response to extracellular nucleotides has been described. http://www.jimmunol.org/ ␥ High concentrations of TSP-1 released by DCs in response to ATP ATP S and PGE2 target gene, NRP-1, is expressed on human DCs and resting T cells. NRP-1 was described as a critical membrane suggest a negative global effect of ATP-treated DCs on angiogen- protein in the interaction between DCs and T lymphocytes and in esis, but TSP-1 and VEGF-A release by DCs could lead to immu- the initiation of DC-induced proliferation of resting T cells (19). nosuppression through their negative paracrine effect on T cell These data support the notion that the down-regulation of NRP-1 proliferation (33). ␥ The large expression profile of ATP in human DCs reflects the expression on MoDCs in response to ATP, ATP S, and PGE2 could also contribute to their immunosuppressive effects. complexity of its action on human DCs and in the immune system The tolerogenic properties of ATP-treated DCs was also sup- in general. Several studies have described both proinflammatory by guest on September 30, 2021 ported inter alia by the up-regulation of TSP-1 and IDO in MoDCs and anti-inflammatory actions of ATP. Swennen et al. (34) re- as well as the secretion of kynurenine derivatives (11). IDO is an cently reported that ATP inhibits the release of the proinflamma- ␣ enzyme involved in tryptophan metabolism and considered as a tory cytokine TNF- and stimulates the release of the anti- inflammatory cytokine IL-10. ATP also stimulates IL-1 release negative regulator of T lymphocyte proliferation. PGE2 was shown to induce IDO activity in DCs by both in vitro (20) and in vivo (21) through the P2X7 receptor, a low affinity ATP receptor expressed studies. We have shown that ATP␥S up-regulated IDO mRNA on DCs (24). The balance between proinflammatory and anti- inflammatory actions of ATP depends on the immune cell types more strongly than PGE2 by microarray and qRT-PCR experi- ments. It was interesting to observe the additional up-regulations involved but could also be a matter of the concentration and/or the location of released ATP. At the level of DCs ATP could exert a proinflammatory action when it is released at a low concentration, and a massive release of ATP through cell lysis could lead to an anti-inflammatory or tolerogenic signal through the activation of

its low affinity receptor P2Y11. It has been described, for example, that low concentrations of ATP and ADP induce DC migration

through P2Y2 and P2Y1 activation, respectively, whereas high

concentrations of ATP inhibit DC migration through the P2Y11 receptor (35). The synergy between LPS and ATP on VEGF re- lease and between IFN-␥ and ATP on IDO activity highlights pos- sible cross-talks between their distinct signaling pathways and de- fines ATP as a key cosignal on human DCs. The action of released ATP will thus also depend on the identity and the concentration of the different factors present at the site of inflammation. We have previously reported that ATP could act as an immu- nosuppressive agent through a direct negative effect on cytokine FIGURE 6. ATP␥S but not PGE potentiates IFN-␥ action on kynure- ϩ 2 release from T CD4 (36). In this report we have shown that, at the nine production. DCs were either untreated (CONT, Control) or treated ␥ ␮ ␮ level of DCs, besides its negative effect on chemokine release ATP with ATP S (100 M) or PGE2 (5 M) alone or in combination with 100 U/ml IFN-␥ for 24 h in complete medium. DCs were then washed and down-regulates the expression of major receptors involved in DC incubated for five additional hours in red phenol-free RPMI 1640 supple- differentiation and function (CSF-1R and NRP-1) and stimulates mented with 300 ␮M L-tryptophan. Kynurenine levels were determined in the expression of displaying immunosuppressive proper- each supernatant by HPLC. Data (mean Ϯ range) were obtained in dupli- ties. Even if proinflammatory actions of ATP have been previously

cate and are representative of three independent experiments. described, it is important to consider ATP and also PGE2 as agents 3558 GENE PROFILING DEFINES ATP AS A KEY REGULATOR OF DCs that might inhibit or reduce inflammation in specific conditions. 17. Bagri, A., and M. Tessier-Lavigne. 2002. Neuropilins as Semaphorin receptors: Our data suggest that ATP derivatives specific of a cAMP-coupled in vivo functions in neuronal cell migration and axon guidance. Adv. Exp. Med. Biol. 515: 13–31. P2Y11 receptor could be used as therapeutic agents to regulate 18. Miao, H. Q., and M. Klagsbrun. 2000. Neuropilin is a mediator of angiogenesis. major DC functions and leukocyte recruitment. Cancer Metastasis Rev. 19: 29–37. 19. Tordjman, R., Y. Lepelletier, V. Lemarchandel, M. Cambot, P. Gaulard, O. Hermine, and P. H. Romeo. 2002. A neuronal receptor, neuropilin-I, is es- Acknowledgments sential for the initiation of the primary immune response. Nat. Immunol. 3: We thank F. Bulte´for technical assistance. We thank Dr. B. Dessars for 477–482. helpful discussions. 20. Braun, D., R. S. Longman, and M. L. Albert. 2005. A two-step induction of indoleamine 2,3 dioxygenase (IDO) activity during dendritic-cell maturation. Disclosures Blood 106: 2375–2381. 21. Bergwelt-Baildon, M. S., A. Popov, T. Saric, J. Chemnitz, S. Classen, The authors have no financial conflict of interest. M. S. Stoffel, F. Fiore, U. Roth, M. Beyer, S. Debey, et al. 2006. CD25 and indoleamine 2,3-dioxygenase are up-regulated by prostaglandin E2 and expressed References by tumor-associated dendritic cells in vivo: additional mechanisms of T-cell in- hibition. Blood 108: 228–237. 1. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. T. Liu, B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. 22. Fallarino, F., R. Bianchi, C. Orabona, C. Vacca, M. L. Belladonna, M. C. Fioretti, Rev. Immunol. 18: 767–811. D. V. Serreze, U. Grohmann, and P. Puccetti. 2004. CTLA-4-Ig activates fork- 2. la Sala, A., D. Ferrari, S. Corinti, A. Cavani, F. Di Virgilio, and G. Girolomoni. head transcription factors and protects dendritic cells from oxidative stress in 2001. Extracellular ATP induces a distorted maturation of dendritic cells and nonobese diabetic mice. J. Exp. Med. 200: 1051–1062. inhibits their capacity to initiate Th1 responses. J. Immunol. 166: 1611–1617. 23. Ferrari, D., A. la Sala, P. Chiozzi, A. Morelli, S. Falzoni, G. Girolomoni, 3. Schnurr, M., F. Then, P. Galambos, C. Scholz, B. Siegmund, S. Endres, and M. Idzko, S. Dichmann, J. Norgauer, and F. Di Virgilio. 2000. The P2 purinergic A. Eigler. 2000. Extracellular ATP and TNF-␣ synergize in the activation receptors of human dendritic cells: identification and coupling to cytokine re- and maturation of human dendritic cells. J. Immunol. 165: 4704–4709. lease. FASEB J. 14: 2466–2476. Downloaded from 4. Wilkin, F., X. Duhant, C. Bruyns, N. Suarez-Huerta, J. M. Boeynaems, and 24. Ferrari, D., C. Pizzirani, E. Adinolfi, R. M. Lemoli, A. Curti, M. Idzko, B. Robaye. 2001. The P2Y(11) receptor mediates the ATP-induced maturation of E. Panther, and F. Di Virgilio. 2006. The P2X7 receptor: a key player in IL-1 human monocyte-derived dendritic cells. J. Immunol. 166: 7172–7177. processing and release. J. Immunol. 176: 3877–3883. 5. Horckmans, M., B. Marcet, F. Marteau, F. Bulte, A. Maho, M. Parmentier, 25. Chitu, V., and E. R. Stanley. 2006. Colony-stimulating factor-1 in immunity and J. M. Boeynaems, and D. Communi. 2006. Extracellular adenine nucleotides inflammation. Curr. Opin. Immunol. 18: 39–48. inhibit the release of major monocyte recruiters by human monocyte-derived 26. MacDonald, K. P. A., V. Rowe, H. M. Bofinger, R. Thomas, T. Sasmono, dendritic cells. FEBS Lett. 580: 747–754. D. A. Hume, and G. R. Hill. 2005. The colony-stimulating factor 1 receptor is

6. la Sala, A., S. Sebastiani, D. Ferrari, F. Di Virgilio, M. Idzko, J. Norgauer, and expressed on dendritic cells during differentiation and regulates their expansion. http://www.jimmunol.org/ G. Girolomoni. 2002. Dendritic cells exposed to extracellular adenosine triphos- J. Immunol. 175: 1399–1405. phate acquire the migratory properties of mature cells and show a reduced ca- 27. Takashima, A., and T. Kitajima. 1998. T cell-mediated terminal maturation of pacity to attract type 1 T lymphocytes. Blood 99: 1715–1722. dendritic cells, a critical transition into fully potent antigen presenting cells. 7. Wilkin, F., P. Stordeur, M. Goldman, J. M. Boeynaems, and B. Robaye. 2002. Pathol. Biol. 46: 53–60. Extracellular adenine nucleotides modulate cytokine production by human mono- 28. Sozzani, S., W. Luini, A. Borsatti, N. Polentarutti, D. Zhou, L. Piemonti, G. cyte-derived dendritic cells: dual effect on IL-12 and stimulation of IL-10. Eur. D’Amico, C. A. Power, T. N. Wells, M. Gobbi, et al. 1997. Receptor expression J. Immunol. 32: 2409–2417. and responsiveness of human dendritic cells to a defined set of CC and CXC 8. Lutz, M. B., and G. Schuler. 2002. Immature, semi-mature and fully mature chemokines. J. Immunol. 159: 1993–2000. dendritic cells: which signals induce tolerance or immunity? Trends Immunol. 23: 29. Stellato, C., P. Collins, P. D. Ponath, D. Soler, W. Newman, G. La Rosa, H. Li, 445–449. J. White, L. M. Schwiebert, C. Bickel, et al. 1997. Production of the novel C-C 9. Moser, M. 2003. Dendritic cells in immunity and tolerance: do they display chemokine MCP-4 by airway cells and comparison of its biological activity to

opposite functions? Immunity 19: 5–8. other C-C chemokines. J. Clin. Invest. 99: 926–936. by guest on September 30, 2021 10. Shortman, K., and W. R. Heath. 2001. Immunity or tolerance? That is the ques- 30. Riboldi, E., T. Musso, E. Moroni, C. Urbinati, S. Bernasconi, M. Rusnati, tion for dendritic cells. Nat. Immunol. 2: 988–989. L. Adorini, M. Presta, and S. Sozzani. 2005. Cutting edge: proangiogenic prop- 11. Marteau, F., N. S. Gonzalez, D. Communi, M. Goldman, J. M. Boeynaems, and erties of alternatively activated dendritic cells. J. Immunol. 175: 2788–2792. D. Communi. 2005. Thrombospondin-1 and indoleamine 2,3-dioxygenase are 31. Ohm, J. E., D. I. Gabrilovich, G. D. Sempowski, E. Kisseleva, K. S. Parman, major targets of extracellular ATP in human dendritic cells. Blood 106: S. Nadaf, and D. P. Carbone. 2003. VEGF inhibits T-cell development and may 3860–3866. contribute to tumor-induced immune suppression. Blood 101: 4878–4886. 12. Puskas, L. G., A. Zvara, L. Hackler, Jr., and P. Van Hummelen. 2002. RNA amplification results in reproducible microarray data with slight ratio bias. Bio- 32. Doyen, V., M. Rubio, D. Braun, T. Nakajima, J. Abe, H. Saito, G. Delespesse, Techniques 32: 1330–1334, 1336, 1338, 1340. and M. Sarfati. 2003. Thrombospondin 1 is an autocrine negative regulator of 13. Vandesompele, J., K. De Preter, F. Pattyn, B. Poppe, N. Van Roy, A. De Paepe, human dendritic cell activation. J. Exp. Med. 198: 1277–1283. and F. Speleman. 2002. Accurate normalization of real-time quantitative RT-PCR 33. Takahashi, A., K. Kono, F. Ichihara, H. Sugai, H. Fujii, and Y. Matsumoto. 2004. data by geometric averaging of multiple internal control genes. Genome Biol. 3: Vascular endothelial growth factor inhibits maturation of dendritic cells induced RESEARCH0034. by lipopolysaccharide, but not by proinflammatory cytokines. Cancer Immunol. 14. Kubo, S., H. K. Takahashi, M. Takei, H. Iwagaki, T. Yoshino, N. Tanaka, Immunother. 53: 543–550. S. Mori, and M. Nishibori. 2004. E-prostanoid (EP)2/EP4 receptor-dependent 34. Swennen, E. L., A. Bast, and P. C. Dagnelie. 2005. Immunoregulatory effects of maturation of human monocyte-derived dendritic cells and induction of helper T2 adenosine 5Ј-triphosphate on cytokine release from stimulated whole blood. Eur. polarization. J. Pharmacol. Exp. Ther. 309: 1213–1220. J. Immunol. 35: 852–858. 15. Rieser, C., G. Bock, H. Klocker, G. Bartsch, and M. Thurnher. 1997. Prostaglan- 35. Schnurr, M., T. Toy, P. Stoitzner, P. Cameron, A. Shin, T. Beecroft, I. D. Davis, ␣ din E2 and tumor necrosis factor cooperate to activate human dendritic cells: J. Cebon, and E. Maraskovsky. 2003. ATP gradients inhibit the migratory ca- synergistic activation of production. J. Exp. Med. 186: 1603–1608. pacity of specific human dendritic cell types: implications for P2Y(11) receptor 16. Wille, S., A. Szekeres, O. Majdic, E. Prager, G. Staffler, J. Stockl, D. Kunthalert, signaling. Blood 102: 613–620. E. E. Prieschl, T. Baumruker, H. Burtscher, et al. 2001. Characterization of 36. Duhant, X., L. Schandene, C. Bruyns, N. S. Gonzalez, M. Goldman, CDw92 as a member of the choline transporter-like protein family regulated J. M. Boeynaems, and D. Communi. 2002. Extracellular adenine nucleotides specifically on dendritic cells. J. Immunol. 167: 5795–5804. inhibit the activation of human CD4ϩ T lymphocytes. J. Immunol. 169: 15–21.