Ligand-Activation of the A2a Receptors Inhibits IL-12 Production by Human Monocytes

This information is current as Amrey A. Link, Tomoshige Kino, James A. Worth, Jennifer of October 1, 2021. L. McGuire, Marianna L. Crane, George P. Chrousos, Ronald L. Wilder and Ilia J. Elenkov J Immunol 2000; 164:436-442; ; doi: 10.4049/jimmunol.164.1.436 http://www.jimmunol.org/content/164/1/436 Downloaded from

References This article cites 36 articles, 13 of which you can access for free at: http://www.jimmunol.org/content/164/1/436.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 1, 2021 *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 © 2000 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Ligand-Activation of the Adenosine A2a Receptors Inhibits IL-12 Production by Human Monocytes

Amrey A. Link,* Tomoshige Kino,* James A. Worth,* Jennifer L. McGuire,* Marianna L. Crane,† George P. Chrousos,* Ronald L. Wilder,† and Ilia J. Elenkov1*†

Adenosine (ADO) exerts potent anti-inflammatory and immunosuppressive effects. In this paper we address the possibility that these effects are partly mediated by inhibition of the secretion of IL-12, a proinflammatory cytokine and a major inducer of Th1 -responses. We demonstrate that 5؅-N-ethylcarboxamidoadenosine (NECA), a nonspecific ADO analogue, and 2-p-(2-carbonyl -ethyl)phenylethylamino-5؅-N-ethylcarboxamidoadenosine (CGS-21680), a specific A2a agonist, dose-dependently inhib ited, in whole blood ex vivo and monocyte cultures, the production of human IL-12 induced by LPS and Stapholococcus aureus Cowan strain 1. However, the A1 receptor agonist 2-Chloro-N6-cyclopentyladenosine and the A3 receptor agonists N6-Benzyl- NECA and 1-deoxy-1-[6-[[(3-iodophenyl)methyl]amino]-9H-purin-9-yl]-N-methyl-␤-D-ribofuranuronamide expressed only weak Downloaded from inhibitory effects. On the other hand, NECA and CGS-21680 dose-dependently potentiated the production of IL-10. The differ- ential effect of these drugs on monocyte IL-12 and IL-10 production implies that these effects are mediated by A2a receptor signaling rather than by intracellular toxicity of ADO analogue’s metabolites. Moreover, CGS-21680 inhibited IL-12 production independently of endogenous IL-10 induction, because anti-IL-10 Abs failed to prevent its effect. The selective A2a antagonist 8-(3-Chlorostyryl) prevented the inhibitory effect of CGS-21680 on IL-12 production. The phosphodiesterase inhibitor Ro 20-1724 dose-dependently potentiated the inhibitory effect of CGS-21680 and, furthermore, Rp-cAMPS, a protein kinase A in- http://www.jimmunol.org/ hibitor, reversed the inhibitory effect of CGS-21680, implicating a cAMP/protein kinase A pathway in its action. Thus, ligand activation of A2a receptors simultaneously inhibits IL-12 and stimulates IL-10 production by human monocytes. Through this mechanism, ADO released in excess during inflammatory and ischemic conditions, or tissue injury, may contribute to selective suppression of Th1 responses and cellular immunity. The Journal of Immunology, 2000, 164: 436–442.

ellular and humoral immunity and their effector cells diabetic mice (5) and in type II collagen-induced arthritis in are regulated by APCs, the Th cell subtypes Th1 and DBA/1 mice (6). In addition, stimulation of IL-12 secretion by Th2, and by their secreted soluble messengers (1, 2). Th1 microbial products was shown to be the crucial factor for the pro-

C by guest on October 1, 2021 cells primarily secrete IFN-␥, IL-2, and TNF-␤, which promote liferation and differentiation of pathogenic autoreactive Th1 effec- cellular immunity, whereas Th2 cells secrete a different set of cy- tor cells in experimental allergic encephalomyelitis (7). tokines, primarily IL-4, IL-10, and IL-13, which promote humoral Adenosine (ADO),2 together with ADP and ATP, belongs to a immunity (1, 2). class of endogenous produced by many cells IL-12 is a 75-kDa heterodimeric cytokine composed of p35 and during normal metabolic activity (8–10). These ubiquitous mole- p40 chains; p35 is constitutively expressed in most cells, whereas cules are utilized in selective extracellular signaling (8–10). For p40 is up-regulated in monocytes and macrophages in response to example, ADO appears to be an important endogenous regulator of infection or appropriate activating substances such as LPS (3). coronary blood flow (8). Postganglionic sympathetic nerve termi- IL-12 is a central inducer of Th1 responses and cell-mediated im- nals also release ATP that is rapidly degraded to ADO, which munity. First, it induces IFN-␥ production from NK and T cells, induces vasodilatation mediated by A2 receptors (10). Local ex- which contributes to phagocytic cell activation and inflammation. tracellular ADO levels increase dramatically during ischemia and Second, it favors Th1 cell differentiation and proliferation (3). hypoxia, possibly providing cytoprotection and increased blood IL-12 is essential for the clearance of certain intracellular patho- flow to ischemic tissues (11). Inflammation and tissue injury rep- gens such as Mycobacterium tuberculosis (4). However, excessive resent pathologic states that are also associated with enhanced ex- production of IL-12 may be responsible for the proinflammatory tracelluar ADO concentrations. activities and the tissue damage typical of organ-specific autoim- One of the first clinical observations linking ADO to the im- munity, as seen in insulin-dependent diabetes mellitus in nonobese mune system was made in patients with SCID. These patients lack the enzyme adenosine deaminase, i.e., they are unable to catabo- lize ADO (12). This abnormality results in highly elevated plasma *Developmental Endocrinology Branch, National Institute of Child Health and Hu- man Development, and †Arthritis and Rheumatism Branch, National Institute of Ar- levels of ADO and consequent impairment of cellular immunity, thritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Be- marked by recurrent infections. The immunosuppression observed thesda, MD 20892 Received for publication October 5, 1999. Accepted for publication October 19, 1999. The costs of publication of this article were defrayed in part by the payment of page 2 Abbreviations used in this paper: ADO, adenosine; RA, rheumatoid arthritis; SAC, charges. This article must therefore be hereby marked advertisement in accordance Staphylococcus aureus Cowan strain 1; NECA, 5Ј-N-ethylcarboxamidoadenosine; N 6- with 18 U.S.C. Section 1734 solely to indicate this fact. Benzyl-NECA, N 6-Benzyl-5Ј-N-ethylcarboxamidoadenosine; IB-MECA, 1-deoxy-1-[6- 1 Address correspondence and reprint requests to Dr. Ilia Elenkov, Arthritis and Rheu- [[(3-iodophenyl)methyl]amino]-9H-purin-9-yl]-N-methyl-␤-D-ribofuranuronamide CGS- matism Branch, National Institute of Arthritis and Musculoskeletal and Skin Diseases, 21680, 2-p-(2-carbonylethyl)-phenylethylamino-5Ј-N-ethylcarboxamidoadenosine; Building 10, Room 9N240, National Institutes of Health, 10 Center Drive, MSC 1820, CCPA, 2-Chloro-N6-cyclopentyladenosine; CSC, 8-(3-Chlorostyryl)caffeine; PKA, pro- Bethesda, MD 20892-1820. E-mail address: [email protected] tein kinase A.

Copyright © 2000 by The American Association of Immunologists 0022-1767/00/$02.00 The Journal of Immunology 437 in SCID is explained by the direct lymphotoxicity of ADO catabo- Monocyte elutriation and culture lites and by the strong extracellular ADO-induced inhibition of the Human monocytes were obtained from healthy volunteers of the National TCR-triggered proliferation and T cell effector functions mediated Institutes of Health, Department of Transfusion Medicine, Cell Processing by A2a receptor signaling (13, 14). More recently, it has been Section. Mononuclear cells were isolated on lymphocyte separation me- demonstrated that ADO also exerts diverse anti-inflammatory ef- dium and the monocytes were purified by counterflow centrifugal elutria- ϫ 5 fects, mediated mainly by A2 receptors, including diminished leu- tion as previously described (25). The monocytes were cultured at 5 10 cells/ml. All cultures were set up in RPMI 1640 medium supplemented kocyte accumulation, inhibition of C2 production, and reduction of with 15% FCS, 1% glutamine, and 50 ␮g/ml gentamicin to yield a final the superoxide anion generation (15–17). Therefore, the regulation volume of 1.5 ml. Cytokine production was induced by LPS (1 ␮g/ml), and of ADO receptors is recognized as having significant therapeutic all cultures were incubated in 5% CO2 at 37°C for 18 h. In several exper- potential (9). iments, anti-IL-10 Abs were added at a concentration of 10 ␮g/ml. Ac- cording to the manufacturer’s instruction, this concentration gives a 50% Recent evidence suggests that the anti-inflammatory effects of neutralizing dose in the presence of 5 ng/ml of rhIL-10. and sulfasalzine, currently the most commonly pre- scribed and effective second-line agents for the treatment of rheu- Cytokine assays matoid arthritis (RA), are mediated in part by increases in the extracellular concentrations of ADO (15, 18). Because excessive IL-12 (p40 and p70) and IL-10 were measured using ELISAs employing the multiple-Ab sandwich principle (Quantikine, R&D Systems). These production of IL-12 appears to play an important role in RA, the assays specifically detect human IL-12 p70 (the biologically active het- question arises: does ADO suppress IL-12 production? This may erodimer), p40, and human IL-10, respectively. The IL-12 p70 ELISA explain, at least in part, the beneficial effects of these drugs in RA. specifically recognizes the IL-12 heterodimer without cross-reactivity with This hypothesis arose from previous studies showing that ADO the individual subunits of the dimer. The detection limits of the IL-12 p40 inhibits the production of TNF-␣ (19–22), another proinflamma- and IL-12 p70 high-sensitivity ELISA were 15.0 and 0.5 pg/ml, whereas Downloaded from the IL-10 ELISA detection limit was 2 pg/ml. Plates were read by a mi- tory cytokine implicated in the pathogenesis of RA, while it in- croplate reader (model 550, Bio-Rad, Richmond, CA) and absorbency was creases the secretion of the anti-inflammatory cytokine IL-10 (23, transformed to cytokine concentration (in pg/ml) using a standard curve 24), which suppresses IL-12 and TNF-␣ release. In this paper, we computed by Microplate Manager III Macintosh data analysis software demonstrate that ADO analogues, via activation of A2a receptors (Bio-Rad). on monocytes, cause substantial inhibition of human IL-12 pro- duction and simultaneous stimulation of IL-10 production. Detection of IL-12 p35 and p40 mRNA http://www.jimmunol.org/ Through these mechanisms, increased concentrations of ADO may Human monocytes were cultured at 1.5 ϫ 107 cells, with or without LPS polarize the Th1/Th2 balance toward Th2 dominance and may (1 ␮g/ml), in the presence or absence of 10Ϫ5 or 10Ϫ6 M concentrations of contribute to the immunomodulatory properties of ADO. the A2a agonist CGS-21680 at 37°C for 18 h. Total RNA was isolated by using TRIZOL reagent according to the manufacturer’s instructions. cDNAs were reverse transcribed with TaqMan reverse transcription re- Materials and Methods agents (Applied Biosystems). cDNA solution (5 ␮l) was used to quantitate Drugs and reagents mRNA levels using TaqMan cytokine gene expression plate I (Applied Biosystems)with 65 PCR cycles on the ABI PRISM 7700 sequence detec- LPS from Escherichia coli, serotype K-235 (Sigma, St. Louis, MO), was dis- tion system (Applied Biosystems). Results were analyzed with the se- solved in distilled water. The mixture was sonicated for at least 3 min in a quence detection system software. by guest on October 1, 2021 sonicating bath, and aliquots were stored at Ϫ20°C until they were used. After thawing, appropriate dilutions were made in RPMI 1640 medium. Stapholo- coccus aureus Cowan strain 1 (SAC) was purchased from Calbiochem Statistical analysis (La Jolla, CA). 5Ј-N-ethylcarboxamidoadenosine (NECA), N6-Benzyl-5Ј-N- Ϯ 6 All data are presented as mean SE. Statistical analysis was performed by ethylcarboxamidoadenosine (N -Benzyl-NECA), 2-[p-(2-carbonyl-ethyl)- one-way ANOVA. phenyl-ethylamino]-5Ј-N-ethylcarboxamido-adenosine (CGS-21680) HCL, 2-Chloro-N6-cyclopentyladenosine (CCPA), N6-(3-iodobenzyl)-adenosine-5Ј- N-methyluronamide (IB-MECA), 8-(3-Chlorostyryl)caffeine (CSC), Rp- Results cAMPS triethylamine, 8-(chlorophenylthio)-cAMP, and Ro 20-1724 were The ADO analogue, NECA, suppresses LPS-induced IL-12 and obtained from Research Biochemicals (Natick, MA). TRIZOL reagent was purchased from Life Technologies (Gaithersburg, MD). TaqMan cytokine stimulates IL-10 production gene expression plate I was obtained from Applied Biosystems (Foster City, Because ADO has an extremely short half-life and is rapidly me- CA). Anti-human IL-10 Abs were purchased from R&D Systems (Minneapolis, MN). tabolized, in our experiments we used nonhydrolyzable ADO an- alogues or more metabolically stable analogues that are able to Blood donors mimic the effects of extracellular ADO (8, 9). The addition of Twenty healthy female and male volunteers participated in this study, increasing concentrations of the ADO analogue NECA to the LPS- which was approved by an institutional review board of the National In- stimulated whole blood cultures resulted in a dose-dependent in- stitutes of Health. Volunteers refrained from using any drugs, including hibition of IL-12 p70 production (Fig. 1A). NECA at a concentra- cyclooxygenase inhibitors, or hormones for 1 wk before the study. tion of 10Ϫ5 M suppressed IL-12 production by 95%. However, Whole blood cultures increasing doses of NECA resulted in a dose-dependent increase of LPS-induced IL-10 production (Fig. 1B), showing a substantial Blood was drawn into sodium heparin-containing sterile blood collecting increase at a concentration of 10Ϫ6 M (a 240% increase). tubes (Vacutainer, Becton Dickinson, Lincoln Park, NJ) and diluted 1:5 with RPMI 1640 (supplemented with 1% glutamine and 50 ␮g/ml genta- micin) with no added exogenous serum. The blood (1 ml) was aliquoted in The effect of NECA on the IL-12 and IL-10 production is not 24-well culture plates (Costar, Cambridge, MA). Bacterial LPS was added stimulus-specific to allow a final concentration of 1 ␮g/ml for the induction of cytokines, and the samples were incubated in 5% CO2 at 37°C for 18 h. ADO analogues To determine whether the effects of NECA on IL-12 and IL-10 and receptor agonists were added to the wells 10 min before LPS or SAC production were limited to LPS stimulation, whole blood cultures stimulation. CSC, Rp-cAMPS triethylamine, and the phosphodiesterase in- were stimulated with SAC in the presence of increasing doses of hibitor Ro 20-1724 were added 10 min before the ADO receptor agonists. After incubation, the blood was centrifuged, and the supernatant plasma NECA. As shown in Fig. 2A, NECA strongly inhibited SAC-in- was collected and stored in polypropylene tubes at Ϫ70°C until the sam- duced production of IL-12 p70 in a dose-dependent manner, while ples were measured. it simultaneously enhanced the production of IL-10 (Fig. 2B). 438 ADENOSINE AND IL-12 PRODUCTION Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 1. Effect of the ADO analogue NECA on LPS-induced IL-12 p70 (A) and IL-10 (B) production in human whole blood from six normal FIGURE 2. Effect of the ADO analogue NECA on SAC-induced IL-12 volunteers. Increasing concentrations of NECA were added as indicated. p70 (A) and IL-10 (B) production in human whole blood from four normal Data are expressed as the mean Ϯ SE. Mean LPS-induced IL-12 p70 and volunteers. Increasing concentrations of NECA were added as indicated. Ϯ IL-10 concentrations were 54.2 Ϯ 10.1 and 158.5 Ϯ 51.4 pg/ml, respec- Data are expressed as the mean SE. Mean LPS-induced IL-12 p70 and Ϯ Ϯ tively. Asterisks indicate significant differences at p Ͻ 0.05, compared with IL-10 concentrations were 22.7 6.6 and 98.9 11.26 pg/ml, respec- Ͻ LPS control. tively. Asterisks represent significant differences at p 0.05, compared with LPS control.

ADO receptor agonists inhibit IL-12 and stimulate IL-10 production Monocytes are the main targets for the effect of ADO analogues The ADO analogue NECA is a nonselective agonist of A1, A2, on IL-12 and IL-10 production in human peripheral blood and A3 receptors. To investigate whether the modulation of cyto- Monocytes are the major cellular source of IL-12 and IL-10 pro- kine production is related to the stimulation of specific ADO re- duction in human whole blood (26, 27). To verify that peripheral ceptors, the order of effectiveness of specific ADO receptor ago- blood monocytes were directly affected by ADO analogues, we nists was compared in LPS-stimulated whole blood. As evident stimulated enriched human monocytes with LPS in the presence of from Fig. 3A, the A2a receptor agonist CGS-21680 also exhibited different concentrations of A1, A2, and A3 agonists. Although strong dose-dependent inhibitory activity, whereas the A1 receptor ADO receptor agonists had an inhibitory effect on the IL-12 p40 agonist CCPA and the A3 receptor agonist N6-Benzyl-NECA were production, they potentiated IL-10 production in monocytes (Fig. effective only at high concentrations. The median effective con- 4, A and B). Similar to the findings in whole blood, the A2a re- centration for the A2a receptor agonist CGS-21680 (5 ϫ 10Ϫ8 M) ceptor agonist CGS-21680 expressed the strongest effects. The po- was approximately two orders of magnitude less than that for the tency of two different A3 receptor agonists was also compared. A1 and A3 receptor agonists (10Ϫ6 and 5 ϫ 10Ϫ6 M, respectively). Although IB-MECA induced a slightly stronger effect than N6- The same order of effectiveness of ADO agonists was observed in Benzyl-NECA, the effect of both A3 agonists was still much their enhancing effect of IL-10 production (Fig. 3B). These data weaker than the effect of A2a receptor agonist CGS-21680. Be- suggest that the modulation of LPS-induced IL-12 and IL-10 pro- cause freshly isolated, unprimed human monocytes produce little duction by ADO analogues in human whole blood is primarily IL-12 p70 when stimulated with LPS, the effect of ADO analogues mediated through A2a receptors. on IL-12 p70 could not be reliably evaluated in these experiments. The Journal of Immunology 439 Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 3. Effect of the ADO receptor agonists CCPA (A1), CGS- FIGURE 4. Effect of the ADO receptor agonists CCPA (A1), CGS- 6 I II 21680 (A2a), and N6-Benzyl-NECA (A3) on LPS-induced IL-12 p70 (A) 21680 (A2a), N -Benzyl-NECA (A3 ), and IB-MECA (A3 ) on LPS-in- and IL-10 (B) production in human whole blood from seven normal vol- duced IL-12 p40 (A) and IL-10 (B) production in human monocytes from unteers. Increasing concentrations of ADO receptor agonists were added as six normal volunteers. Increasing concentrations of ADO receptor agonists Ϯ indicated. Data are expressed as the mean Ϯ SE. Mean LPS-induced IL-12 were added as indicated. Data are expressed as the mean SE. Mean Ϯ Ϯ p70 and IL-10 concentrations were 74 Ϯ 15.6 and 298 Ϯ 78.2 pg/ml, concentrations of IL-12 p40 and IL-10 were 1499.4 320.3 and 715.7 Ͻ respectively. Symbols represent significant differences at p Ͻ 0.05, com- 206.2 pg/ml, respectively. Symbols represent significant differences at p ϩ I ء ✦ A2a agonist; and ϩ, A3 agonist). 0.05, compared with LPS control ( , A1 agonist; , A2a agonist; ,A3 ,ء ;pared with LPS control (✦, A1 agonist agonist; and ✤,A3II agonist).

CSC, an A2a-selective antagonist, blocks the effect of A1, A2, and A3 receptor agonists on the production of IL-12 of this drug is at the transcriptional level. On the other hand, the As shown in Fig. 5, the inhibitory effect of the A2a agonist CGS- levels of IL-12 p35 mRNA were not affected by LPS and the A2a 21680 was blocked in the presence of the specific A2a antagonist agonist. CSC. Because the inhibitory effects of the A1 and A3 receptor agonists were observed only at high concentrations, we subse- The A2a receptor agonist CGS-21680 mediates its effect through quently investigated the possibility that these effects may have cAMP/protein kinase A (PKA) pathway resulted from nonspecific action via A2a receptors. As is visible in Stimulation of A2 receptors results in increased cAMP production Fig. 5, the A2a antagonist CSC was able to prevent the inhibitory (8, 9). To analyze the role of cAMP, we antagonized the effect of effect of the A1 and A3 receptor agonists. the A2a agonist CGS-21680 by application of Rp-cAMPS, a di- astereomer of adenosine-3Ј,5Ј cyclic monophosphothiate. Rp- The A2a agonist CGS-21680 inhibits the expression of IL-12 cAMP is known to competitively inhibit the cAMP-induced acti- p40 mRNA, induced by LPS vation of PKA. Monocytes were stimulated by LPS in the presence We examined the levels of IL-12 p35 and p40 mRNA in human of A2a agonist CGS-21680 and, as shown in Fig. 7A, CGS-21680 peripheral monocytes with TaqMan cytokine gene expression plate inhibited IL-12 production by 50%. The presence of Rp-cAMP I. As shown in Fig. 6, A and B, LPS increased the levels of IL-12 together with the A2a agonist CGS-21680 reversed this inhibition p40 mRNA by 10-fold. Addition of the A2a agonist CGS-21680 dose-dependently. Phosphodiesterase is an enzyme involved in the dose-dependently suppressed LPS effect, suggesting that the effect breakdown of cAMP. Thus, inhibition of this enzyme results in an 440 ADENOSINE AND IL-12 PRODUCTION

FIGURE 5. CSC, a selective A2a receptor antagonist (Ant), prevents the effect of A1 (CCPA), A2a (CGS-21680), and A3 (IB-MECA) receptor agonists (Ago) on LPS-induced IL-12 p40 production in human monocytes from six normal volunteers. CSC was added at a concentration of 10Ϫ6 M; Downloaded from A1, A2, and A3 Ago were added at a concentration of 10Ϫ6,10Ϫ7, and 5 ϫ 10Ϫ7 M, respectively. Data are expressed as the mean Ϯ SE. Mean LPS- induced IL-12 p40 was 1792.7 Ϯ 555.04 pg/ml. Asterisks represent sta- tistically significant differences between the effect of A1, A2, and A3 Ago alone and their effect in the presence of the A2a receptor antagonist CSC. .p Ͻ 0.001 ,ءءء p Ͻ 0.01; and ,ءء ;p Ͻ 0.06 ,ء http://www.jimmunol.org/

increase of cAMP levels. As shown in Fig. 7B, the phosphodies- terase type IV inhibitor Ro 20-1724 had a dose-dependent inhib- itory effect on the LPS-induced IL-12 production in the absence of CGS-21680. Moreover, the effect of Ro 20-1724 was additive to the effect of the A2a receptor agonist CGS-21680. Thus, the pres- ence of increasing concentrations of Ro 20-1724 resulted in about by guest on October 1, 2021 a 2-fold potentiation of the inhibition of IL-12 production induced by CGS-21680. To further evaluate the role of cAMP in this mech- anism, we added 8-(chlorophenylthio)-cAMP, a cAMP analogue, to our cultures. This drug dose-dependently inhibited LPS-induced IL-12 production, mimicking the effect of the A2a agonist CGS- 21680 (Fig. 7C). FIGURE 6. Effect of the A2a agonist CGS-21680 on the levels of IL-12 p35 (A) and p40 (B) mRNA in human peripheral monocytes, induced by LPS. The levels of mRNA levels are shown as an n-fold increase compared The effect of A2a receptor agonist CGS-21680 on IL-12 with the levels of baseline condition, LPS(Ϫ). Data are expressed as the production is independent of endogenous IL-10 secretion mean Ϯ SE from three experiments with six measurements. Asterisk in- Previous studies indicate that IL-10 inhibits the secretion of proin- dicates a statistically significant effect of the A2a agonist CGS-21680 com- ϩ Ͻ flammatory cytokines from monocytes (28). These observations pared with LPS( )(p 0.01). prompted us to consider the possibility that the inhibitory effects of A2 receptor agonists on IL-12 release were caused by induction of endogenous IL-10 production. Consequently, we investigated the Discussion effect of adding neutralizing IL-10 Abs to monocyte cultures stim- ADO is an endogenous with immunoregulatory prop- ulated with LPS in the presence of the A2a agonist CGS-21680. erties. The present study suggests that the inhibition of IL-12 pro- The mean LPS-induced IL-10 levels in our experiments were duction by ADO is an additional mechanism that may contribute to 871.2 Ϯ 194.3 pg/ml. Thus, the concentration of endogenously its anti-inflammatory and immunosuppressive effects. In this paper produced IL-10 was ϳ6-fold less than the concentration at which we demonstrate that the ADO analogues NECA and CGS-21680 anti-IL-10 Abs express a 50% neutralizing dose (see Materials and mediate a strong and dose-dependent inhibition of LPS- and SAC- Methods). We concluded that the concentration of neutralizing anti- induced IL-12 production. The order of potency of NECA, CGS- IL-10 Abs used in these experiments was sufficient to eliminate 21680, and A1 and A3 receptor agonists is: NECA ϭ CGS-21680 endogenously produced IL-10. In the presence of 10Ϫ6 M CGS- ϾϾ CCPA Ͼ N6-Benzyl-NECA. Because A2a receptors are char- 21680, the LPS-induced IL-10 levels were 47.1 Ϯ 10.1 pg/ml, acterized by their high-affinity binding of the agonist CGS-21680, whereas in the presence of anti-IL-10 Abs, the addition of this drug whereas A2b receptors display low affinity for NECA (9), this to the cultures resulted in IL-10 production of 41.6 Ϯ 7.0 pg/ml pharmacological profile of the response implies involvement of (data not shown). These data suggest that endogenous production A2a receptors. This is further substantiated by the observation that of IL-10 induced by LPS does not mediate the inhibitory effect of the selective A2a antagonist CSC blocked the inhibitory effect of the A2a receptor agonist on IL-12 production. CGS-21680. Moreover, the inhibitory effect of CGS-21680 was The Journal of Immunology 441

potentiated by a phosphodiesterase inhibitor and prevented by an inhibitor of the type I and II PKA. Monocytes are the main source of IL-12 production in human whole blood (26, 27), and stimula- tion of A2 receptors in many cells, including monocytes, has been associated with increased generation of cAMP production (9, 16). Thus, ligand activation of A2a receptors through stimulation of the cAMP/PKA pathway appears to mediate the inhibition of IL-12 production by human monocytes. This is consistent with previous studies showing that ADO analogues inhibited, also via stimula- tion of A2 receptors, the secretion of the proinflammatory cytokine TNF-␣ by human monocytes (19, 20, 22). In our experiments, the A1 and A3 receptor agonists had an inhibitory effect on IL-12 production, but only at high concentra- tions. Moreover, CSC, a selective A2a receptor antagonist, pre- vented the inhibitory effects of A1 and A3 receptor agonists. Therefore, it appears that the A1 and A3 receptor agonists express weak A2a receptor agonist effects at high concentrations. This is substantiated by previous studies showing that mouse peripheral

lymphocytes (13, 14) and freshly isolated human monocytes ex- Downloaded from press only functionally active A2 receptors (16, 19, 20, 22, 29). In contrast, cultured blood monocyte-derived macrophages and mac- rophage cell lines might express, in addition, functionally active A1 and/or A3 receptors (21, 29). IL-12 is a central inducer of Th1 differentiation and serves as a bridge between innate and specific immunity, whereas IL-10 an- http://www.jimmunol.org/ tagonizes the activities of IL-12 (1–3). The Th1/Th2 pattern is often regarded as a balance between Th1/Th2 cell cytokine activ- ities, but our observations suggest that increased local concentra- tions of ADO may also contribute to Th1/Th2 balance. In addition to IL-12 inhibition, the present study also demonstrates that NECA and CGS-21680 dose-dependently potentiate the production of IL-10 in human whole blood ex vivo. This is in accordance with recent studies showing similar effects of ADO analogues on IL-10

production in vitro by human monocytes and in vivo in endotox- by guest on October 1, 2021 emic mice (23, 24). Interestingly, ADO appears to express iden- tical modulatory effects on the balance of proinflammatory/anti-

inflammatory cytokines such as PGE2 (26), catecholamines (30, 31) and histamine (27). In this process the increase of cAMP seems to be the common intracellular mechanism used by these endog- enous mediators. Thus, conditions related to increased local con- centrations of ADO, through inhibition of IL-12 and potentiation of IL-10 production from monocytes, may simultaneously mediate an inhibition of Th1 responses and a shift toward Th2 dominance. One such condition is ischemia. During ischemia or hypoxia, local ADO concentrations increase to the micromolar range (11), whereas the secretion of TNF-␣ has been implicated in the patho- genesis of ischemia-reperfusion injury (32). Recent evidence in- dicates that ADO attenuates reperfusion injury following ischemia FIGURE 7. Rp-cAMP triethylamine (Rp-cAMP, 100 and 300 ␮M) Ϫ partly through inhibition of TNF-␣ production (33, 34). Our data dose-dependently prevents the inhibition of CGS-21680 A2a Ago (10 7 M) on IL-12 p40 production in human monocytes from six normal volun- suggest that during ischemia-reperfusion injury, increased concen- teers (A). Dose-dependent effect of Ro 20Ϫ1724 (10Ϫ7,10Ϫ6,5ϫ 10Ϫ6, trations of ADO may act to attenuate injury by inhibiting not only and 10Ϫ5 M) and its potentiation of the effect of CGS-21680 (10Ϫ7 M), a TNF-␣ release, but also IL-12 production. In this context, ADO selective A2a receptor agonist (Ago), on IL-12 p40 production in human might be a beneficial component of the normal host defense in monocytes from six normal volunteers (B). The cAMP analogue 8-(chlo- mild ischemic foci. However, in more severe ischemic events, the rophenylthio)-cAMP dose-dependently inhibits LPS-induced IL-12 p70 systemic increase of ADO may have opposite, detrimental effects. production in whole blood from five normal volunteers (C). Data are ex- Thus, major injury, serious burns, and brain injury often lead to pressed as the mean Ϯ SE. Mean LPS-induced IL-12 p70 (Fig. 6C) and severe immunosuppression causing increased susceptibility to in- IL-12 p40 (Fig. 6, A and B) concentrations were 57.6 Ϯ 17.1, 1083.2 Ϯ 339.1, and 3190.6 Ϯ 577.9 pg/ml, respectively. Asterisks in A indicate a fections (31, 35). Although it is still a matter of speculation, during statistically significant difference (p Ͻ 0.05). Asterisks in B indicate sta- major injury massive release of ADO may mediate, through inhi- tistically significant differences of Ro 20-1724/CGS-21680 compared with bition of IL-12 production, part of the severe immunosuppression Ro 20-1724 alone (p Ͻ 0.05). Pluses indicate significant differences of Ro that occurs in these patients. The ADO-induced inhibition of IL-12 20-1724/CGS-21680 compared with CGS-21680 alone (p Ͻ 0.001). As- production may also be relevant to the immunosuppression observed terisks in C indicate statistically significant difference (p Ͻ 0.01). in solid tumors, where hypoxic conditions cause accumulation of high 442 ADENOSINE AND IL-12 PRODUCTION concentrations of extracellular ADO that may contribute to inhibition 15. Cronstein, B. N., D. Naime, and E. Ostad. 1993. The antiinflammatory mecha- of antitumor CTLs and NK activity (36). nism of methotrexate: increased adenosine release at inflamed sites diminishes leukocyte accumulation in an in vivo model of inflammation. J. Clin. Invest. Methotrexate and sulfasalazine are important second-line agents 92:2675. for the treatment of RA. Recent studies have indicated that these 16. Lappin, D., and K. Whaley. 1984. Adenosine receptors on human monocytes drugs enhance extracellular ADO concentrations (15, 18) and that modulate C2 production. Clin. Exp. Immunol. 57:454. ADO, via stimulation of A2 receptors, may be responsible in part 17. Cronstein, B. N., S. B. Kramer, G. Weissman, and R. Hirschhorn. 1983. Aden- osine: a physiological modulator of superoxide anion generation by human neu- for the anti-inflammatory effects of these drugs (15). Our results trophils. J. Exp. Med. 158:1160. suggest that methotrexate, via release of ADO, may inhibit IL-12 18. Marabito, L., M. C. Montesinos, D. M. Schreibman, L. Balter, L. Thompson, production that may explain, at least in part, the beneficial effects R. Resta, G. Carlin, M. A. Huie, and B. N. Cronstein. 1998. Methotrexate and sulfasalazine promote adenosine release by a mechanism that requires ecto-5Ј- of these drugs in RA patients. nucleotidase-mediated conversion of . J. Clin. Invest. 2:295. In summary, we have shown that ADO analogues, via stimula- 19. Bouma, M. G., R. K. Stad, F. A. J. M. van den Wildenberg, and W. A. Buurman. tion of A2a receptors and subsequent increase of cAMP, inhibit the 1994. Differential regulatory effects of adenosine on cytokine release by activated production of IL-12 by activated human monocytes. Through this human monocytes. J. Immunol. 153:4159. 20. Prabhakar, U., D. P. Brooks, D. Lipshlitz, and K. M. Esser. 1995. Inhibition of mechanism, increased concentrations of ADO may selectively sup- LPS-induced TNF-␣ production in human monocytes by adenosine (A2) receptor press Th1 functions and skew the balance toward Th2 dominance. selective agonists. Int. J. Immunopharmacol. 17:221. This may explain, at least in part, the anti-inflammatory properties 21. Sajjadi, F. G., K. Takabayashi, A. C. Foster, R. C. Domingo, and G. S. Firestein. of ADO and supports the rationale for development and use of 1996. Inhibition of TNF-␣ expression by adenosine: role of A3 adenosine recep- tors. J. Immunol. 156:3442. selective A2a receptor agonists or antagonists in Th1- or Th2- 22. Sullivan, G. W., Y. P. Peng, M. Marshall, and J. Linden. 1998. Adenosine inhibits mediated autoimmune processes, respectively. TNF-␣ release from human monocytes by activation of the A2a receptor. FASEB J. 4:A484. Downloaded from 23. Le Moine, O., P. Stordeur, L. Schandene´, A. Marchant, D. de Groote, Acknowledgments M. Goldman, and J. Devie`re. 1996. Adenosine enhances IL-10 secretion by hu- We thank Dr. Michail V. Sitkovsky (Laboratory of Immunology, National man monocytes. J. Immunol. 156:4408. Institute of Allergy and Infectious Diseases) and Dr. Thomas A. Fleisher 24. Hasko´, G., C. Szabo´, Z. H. Ne´meth, V. Kvetan, S. McCarthy Pastores, and E. S. Vizi. 1996. agonists differentially regulate IL-10, (Immunology Service, Clinical Center, National Institutes of Health) for TNF-␣, and nitric oxide production in RAW 264.7 macrophages and in endo- their helpful comments. toxemic mice. J. Immunol. 157:4634. http://www.jimmunol.org/ 25. Czerniecki, B. J., C. Carter, L. Rivoltini, G. K. Koski, H. I. Kim, D. E. Weng, References J. G. Roros, Y. M. Hijazi, S. Xu, S. A. Rosenberg, and P. A. Cohen. 1997. Calcium ionophore-treated peripheral blood monocytes and dendritic cells rap- 1. Fearon, D. T., and R .M. Locksley. 1996. The instructive role of innate immunity idly display characteristics of activated dendritic cells. J. Immunol. 159:3823. in the acquired immune response. Science 272:50. 26. van der Pouw Kraan, T. C., L. C. Boeije, R. J. Smeenk, J. Wijdenes, and 2. Mosmann, T. R., and S. Sad. 1996. The expanding universe of T-cell subsets: L. A. Aarden. 1995. Prostaglandin-E2 is a potent inhibitor of human interleu- Th1, Th2 and more. Immunol. Today 17:138. kin-12 production. J. Exp. Med. 181:775. 3. Trinchieri, G. 1995. Interleukin-12: a proinflammatory cytokine with immuno- 27. Elenkov, I. J., E. Webster, D. E. Papanicolaou, T. A. Fleischer, G. P. Chrousos, regulatory functions that bridge innate resistance and antigen-specific adaptive and R. L. Wilder. 1998. Histamine potently suppresses human IL-12 and stim- immunity. Annu. Rev. Immunol. 13:251. ulates IL-10 production via H2 receptors. J. Immunol. 161:2586. 4. Sher, A., and R. Ahmed. 1995. Immunity to infection: regulation of innate and 28. Mosmann, T. R. 1994. Properties and functions of interleukin-10. Adv. Immunol. adaptive responses to infection: new avenues and angles. Curr. Opin. Immunol. by guest on October 1, 2021 7:471. 56:1. 5. Rabinovitch, A., W. L. Suarez-Pinzon, and O. Sorensen. 1996. Interleukin 12 29. Salmon, J. E., N. Brogle, C. Brownlie, J. C. Edberg, R. P. Kimberley, B. Chen, mRNA expression in islets correletes with ␤-cell destruction in NOD mice. and B. F. Erlanger. 1993. Human mononuclear phagocytes express adenosine A1 J. Autoimmun. 9:645. receptors. J. Immunol. 151:2775. 6. Germann, T., J. Szeliga, H. Hess, S. Storkel, F. J. Podlaski, M. K. Gately, 30. Elenkov, I. J., D. A. Papanicolaou, R. L. Wilder, and G. P. Chrousos. 1996. E. Schmitt, and E. Rude. 1995. Administration of interleukin-12 in combination Modulatory effects of glucocorticoids and catecholamines on human IL-12 and with type II collagen induces severe arthritis in DBA/1 mice. Proc. Natl. Acad. IL-10 production: clinical implications. Proc. Assoc. Am. Physicians 108:374. Sci. USA 92:4823. 7. Segal, B. M., D. M. Klinman, and E. M. Shevach. 1997. Microbial products 31. Woiciechowsky, C., A. Khursu, D. Nestler, B. Eberhardt, C. Platzer, induce autoimmune disease by an IL-12-dependent pathway. J. Immunol. 158: B. Schoening, F. Gloeckner, W. R. Lanksch, H. Volk, and W. Doecke. 1998. 5087. Sympathetic activation triggers systemic interleukin-10 release in immunodepres- 8. Fredholm, B. B., M. P. Abbracchio, G. Burnstock, J. W. Daly, T. K. Harden, sion induced by brain injury. Nat. Med. 4:808. K. A. Jacobsen, P. Leff, and M. Williams. 1994. Nomenclature and classification 32. Coletti, L. M., D. G. Remick, G. D. Burtch, S. L. Kunkel, R. M. Strieter, and of purinoreceptors. Pharmacol. Rev. 46:143. D. A. Campbell. 1990. Role of tumor necrosis factor-␣ in the pathophysiological 9. Olah, M. E., and G. L. Stiles. 1995. Adenosine receptor subtypes: characteriza- alterations after hepatic ischemia/reperfusion injury in the rat. J. Clin. Invest. tion and therapeutic regulation. Annu. Rev. Pharmacol. Toxicol. 35:581. 85:1936. 10. Burnstock, G. 1997. The past, present and future of purine nucleotides as sig- 33. Forman, M. B., C. E. Velasco, and E. K. Jackson. 1993. Adenosine attenuates naling molecules. Neuropharmacology 36:1127. reperfusion injury following regional myocardial ischaemia. Cardiovasc. Res. 11. Van Belle, H., F. Goossens, and J. Wynants. 1987. Formation and release of 27:9. purine catabolites during hypoperfusion, anoxia and ischemia. Am. J. Physiol. 34. Cain, B. S., D. R. Meldrum, C. A. Dinarello, X. Meng, A. Banerjee, and 252:H886. A. H. Harken. 1998. Adenosine reduces cardiac TNF-␣ production and human 12. Giblett, E. R., J. E. Anderson, F. Cohen, B. Pollara, and H. J. Meuwissen. 1972. myocardial injury following ischemia-reperfusion. J. Surg. Res. 76:117. Adenosine deaminase deficiency in two patients with severely impaired cellular immunity. Lancet 2:1067. 35. O’Sullivan, S. T., J. A. Lederer, A. F. Horgan, D. H. L. Chin, J. A. Mannick, and 13. Huang, S., S. Apasov, M. Koshiba, and M. Sitkovsky. 1997. Role of A2a extra- M. L. Rodrick. 1995. Major injury leads to predominance of the T helper-2 cellular adenosine receptor-mediated signalling in adenosine-mediated inhibition lymphocyte phenotype and diminished interleukin-12 production associated with of T cell activation and expansion. Blood 90:1600. decreased resistance to infection. Ann. Surg. 222:482. 14. Koshiba, M., H. Kojima, S. Huang, S. Apasov, and M. Sitkovsky. 1997. Memory 36. Hoskin, D. W., T. Reynolds, and J. Blay. 1994. Adenosine is a possible inhibitor of extracellular adenosine A2A -mediated signaling in murine of killer T-cell activation in the microenvironment of solid tumours. Int. J. Can- T cells. J. Biol. Chem. 272:25881. cer 59:854.