Adenosine Inhibits Macrophage Colony-Stimulating Factor

Home , Adenylyl cyclase

Adenosine Inhibits Macrophage Colony-Stimulating Factor-Dependent Proliferation of Macrophages Through the Induction of p27 kip-1 Expression This information is current as of September 28, 2021. Jordi Xaus, Annabel F. Valledor, Marina Cardó, Laura Marquès, Jorge Beleta, José M. Palacios and Antonio Celada J Immunol 1999; 163:4140-4149; ; http://www.jimmunol.org/content/163/8/4140 Downloaded from

References This article cites 71 articles, 25 of which you can access for free at: http://www.jimmunol.org/content/163/8/4140.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 September 28, 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 © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Adenosine Inhibits Macrophage Colony-Stimulating Factor-Dependent Proliferation of Macrophages Through the Induction of p27kip-1 Expression1

Jordi Xaus,2* Annabel F. Valledor,2* Marina Cardo´,* Laura Marque`s,* Jorge Beleta,† Jose´M. Palacios,† and Antonio Celada3*

Adenosine is produced during inﬂammation and modulates different functional activities in macrophages. In murine bone mar- ␮ row-derived macrophages, adenosine inhibits M-CSF-dependent proliferation with an IC50 of 45 M. Only speciﬁc agonists that ؅ 6 can activate A2B adenosine receptors such as 5 -N-ethylcarboxamidoadenosine, but not those active on A1 (N -(R)-phenylisopro- ؅ 6 pyladenosine), A2A ([p-(2-carbonylethyl)phenylethylamino]-5 -N-ethylcarboxamidoadenosine), or A3 (N -(3-iodobenzyl)adenosine-5؅-N-methyluronamide) receptors, induce the generation of cAMP and modulate macrophage proliferation. This suggests that Downloaded from adenosine regulates macrophage proliferation by interacting with the A2B receptor and subsequently inducing the production of ␮ ␮ cAMP. In fact, both 8-Br-cAMP (IC50 85 M) and forskolin (IC50 7 M) inhibit macrophage proliferation. Moreover, the inhibition of adenylyl cyclase and protein kinase A blocks the inhibitory effect of adenosine and its analogues on macrophage

proliferation. Adenosine causes an arrest of macrophages at the G1 phase of the cell cycle without altering the activation of the extracellular-regulated protein kinase pathway. The treatment of macrophages with adenosine induces the expression of p27kip-1, kip-1 aG1 cyclin-dependent kinase inhibitor, in a protein kinase A-dependent way. Moreover, the involvement of p27 in the adenosine inhibition of macrophage proliferation was conﬁrmed using macrophages from mice with a disrupted p27kip-1 gene. http://www.jimmunol.org/

These results demonstrate that adenosine inhibits macrophage proliferation through a mechanism that involves binding to A2B adenosine receptor, the generation of cAMP, and the induction of p27kip-1 expression. The Journal of Immunology, 1999, 163: 4140–4149.

acrophages are generated through a process of differ- nal-related kinase (ERK)4 kinase pathway in macrophages (15). entiation known as myelopoiesis. In the bone marrow, The activation of ERK-1/2 is required for macrophage prolifera- M pluripotent stem cells differentiate into monocytes in tion in response to M-CSF.5 Active ERKs phosphorylate and reg-

the presence of M-CSF and other cytokines. Monocytes leave the ulate several cellular proteins (16), including other protein kinases, by guest on September 28, 2021

bone marrow and, by circulating through blood vessels, reach dif- cytoskeletal components, phospholipase A2, and nuclear transcrip- ferent tissues where they terminally differentiate into macrophages tion factors, such as Elk1/TCF and c-Jun, which regulate the ex- and perform their specialized functions (1–3). Tissue macrophages pression of immediate early genes (17, 18). are able to proliferate thanks to the autocrine production of M-CSF Mitogen-activated protein kinase phosphatase-1 (MKP-1) is a (4). Macrophages require M-CSF for proliferation, differentiation, member of a family of dual-speciﬁcity phosphatases (19, 20) that and survival (5). M-CSF is the main growth factor for macro- dephosphorylate both phosphotyrosine and phosphothreonine res- phages and also the only one speciﬁc for these cells. After inter- idues on target proteins. MKP-1 dephosphorylates and inactivates acting with the tyrosine kinase receptor c-fms, M-CSF triggers the ERK-1 and -2 both in vitro and in vivo, suggesting that this phos- activation of several signal transducing molecules in macrophages phatase has a critical effect on maintaining the balance between (5–8), such as some protein kinases of the Src family (9), the ERK phosphorylation and dephosphorylation (21). The overex- transcription factors Stat-1, Stat-3, and Stat-5 (10, 11), protein ki- pression of MKP-1 inhibits ERK-regulated reporter gene expres-

nase C (12), and phosphatidylinositol 3-kinase (13, 14). M-CSF sion, the synthesis of DNA induced by Ras and G1-specific gene also activates the Raf/mitogen-activated protein/extracellular sig- transcription, and entry of fibroblasts to the S phase in response to mitogenic stimuli (22, 23). The induction of the expression of MKP-1 is a mechanism used by the cell to control and attenuate *Departamento de Fisiologia (Biologia del Macro´fag), Facultat de Biologia and Fun- proliferative signaling pathways. dacioÁugust Pi i Sunyer, Campus de Bellvitge, Universitat de Barcelona, Barcelona, Spain; and †Laboratorios Almirall Prodesfarma SA, Research Center, Barcelona, All these signaling pathways allow macrophages to enter the Spain cell cycle in response to M-CSF. Passage through the cell cycle is Received for publication February 24, 1999. Accepted for publication July 28, 1999. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 4 Abbreviations used in this paper: ERK, extracellular signal-related kinase; NECA, 5Ј-N-ethylcarboxamidoadenosine; R-PIA, N6-(R)-phenylisopropyladenosine; DPCPX, 1 This work was supported by grants from Comisioń Interministerial de Ciencia y 1,3-dipropyl-8-cyclopentylxanthine; CGS 21680, 2-[p-(2-carbonyl-ethyl)phenyl- Tecnologia (SAF98-102 and PM 98/0200 to A.C.). J.X. and A.F.V. are recipients of ethylamino]-5Ј - N - ethylcarboxamido adenosine; IB - MECA, N6 - (3 - iodobenzyl) fellowships from the Comissio´ Interdepartamental de Recerca i Innovacio´ adenosine-5Ј-N-methyluronamide; BMDM, bone marrow-derived macrophages; Tecnolo´gica. PKA, protein kinase A; MKP, mitogen-activated protein kinase phosphatase; cdk, 2 J.X. and A.F.V. contributed equally to this work. cyclin-dependent kinse; cki, cdk inhibitor; CADO, 2-choloradenosine; DAPI, 4,6-diamidino-2-phenylindole. 3 Address correspondence and reprint requests to Dr. Antonio Celada, Departamento de Fisiologia, Facultat de Biologia, Av. Diagonal 645, 08028 Barcelona, Spain. E- 5 A. F. Valledor, J. Xaus, L. Marque`s, and A. Celada. M-CSF induces the expression mail address: [email protected] of MKP-1 through a PKC-dependent mechanism. Submitted for publication.

Copyright © 1999 by The American Association of Immunologists 0022-1767/99/$02.00 The Journal of Immunology 4141 regulated through the action of a family of protein kinase com- MO). 2-[p-(2-Carbonylethyl)phenylethylamino]-5Ј-N-ethylcarboxamido- plexes. Each complex is composed of at least a catalytic subunit, adenosine (CGS 21680) and N6-(3-iodobenzyl)adenosine-5Ј-N-methylu- a cyclin-dependent kinase (cdk), and its essential activating part- ronamide (IB-MECA) were purchased from Research Biochemicals (Natick, MA). 8-Br-cAMP and forskolin were obtained from Fluka Bio- ner, which is a cyclin (reviewed in Refs. 24 and 25). These com- chemika (Buchs, Switzerland). [3H]Thymidine was obtained from Amer- plexes are activated at various checkpoints after specific intervals sham (Buckinghamshire, U.K.). 4,6-Diamidino-2-phenylindole (DAPI), during the cell cycle, but they can also be modulated by exogenous SQ 22536, and KT 5720 were purchased from Calbiochem (La Jolla, CA). factors. The cdks are inhibited when associated with a group of All the other products were of the best grade available and were purchased from Sigma. Deionized water further purified with a Millipore Milli-Q proteins known as cdk inhibitors (cki) (26). The ckis best described system (Bedford, MA) was used. are p16Ink-4, which belongs to the INK 4 family, p21waf-1, and p27Kip-1, the latter belonging to the CIP/KIP family of cdk inhib- Cell culture itors (27–29). The presence of growth factors is only required dur- BMDM were isolated from 6-wk-old BALB/c mice (Charles River Labo- ing the G1 phase. Once the cell reaches the restriction point that ratories, Wilmington, MA) as previously described (48). The cells were appears late during the G1 phase and at the beginning of the S cultured in plastic tissue culture dishes (150 mm) in 40 ml DMEM con- phase, the growth factors are no longer necessary and the cell is taining 20% FBS and 30% L cell-conditioned media as a source of M-CSF. The cells were incubated at 37°C in a humidified 5% CO atmosphere. committed to complete the cell cycle (30, 31). This may explain 2 After 7 days of culture, an homogeneous population of adherent macro- why several regulatory steps of the cell cycle, such as the p53 phages was obtained. To render the cells quiescent, when the macrophages checkpoint, pRb phosphorylation, and the activity of the cdk in- were 80% confluent they were deprived of L cell-conditioned medium for hibitors, take place at this point (32–35). 14–16 h before the experiment. BMDM from p27kip-1 knockout mice were

Adenosine is a purine nucleoside produced and secreted to the isolated in the same conditions. These mice were kindly donated by Dr. J. Downloaded from Roberts from Howard Hughes Medical Institute (Seattle, WA) (49). extracellular media by cells during normal intracellular ATP me- tabolization and degradation. Nevertheless, in stress situations like Antibodies ischemia or hypoxia, massive ATP degradation increases local Surface expression of the M-CSF receptor (c-fms) was analyzed by using adenosine concentration to micromolar values (36, 37). In these afﬁnity-puriﬁed rabbit Abs anti-mouse c-fms (Upstate Biotechnology, Lake situations, adenosine modulates several physiological functions, Placid, NY). Fluorescein-conjugated rat anti-rabbit IgG Ab from Sigma

acting mainly as an endogenous antiinflammatory agent (reviewed was used as a secondary Ab. To block Fc receptors, we used an anti-CD16/ http://www.jimmunol.org/ kip-1 in Ref. 38). CD32 Ab (PharMingen, San Diego, CA). For the analysis of p27 expression by Western blotting, we used a monoclonal anti-mouse p27kip-1 Most functional activities of extracellular adenosine are medi- Ab (PharMingen). The rabbit anti-mouse MKP-1 Ab was purchased from ated through binding to specific surface receptors. However, it has Santa Cruz Biotechnology (Santa Cruz, CA). For ERK mobility shift as- been reported that adenosine needs to be internalized to induce says, we used an anti-ERK-1/2 Ab, which was a kind gift of Dr. M. J. some functions such as NO production and inhibition of LPS- Weber (University of Virginia School of Medicine, Charlottesville, VA). induced TNF-␣ expression in some models (39). So far, four dif- Peroxidase-conjugated anti-mouse and anti-rabbit IgGs (Cappel, Turnhout, Belgium) were used as secondary Abs. Primary Abs against mouse ␤-actin ferent adenosine receptors have been described and called A1, were purchased from Sigma. A2A,A2B, and A3 depending on their structural, functional, and pharmacological characteristics (40). Recently, all four adenosine Plasmids and constructs by guest on September 28, 2021 receptor subtypes have been cloned from several species (41–43). The pMH117 plasmid corresponds to the mouse p21waf-1 full-length cDNA All belong to the G-protein-coupled receptor superfamily. The cloned in pEx-lox and was kindly provided by Dr. Massague (Sloan Ket- type and density of adenosine receptors present on the cell surface tering Institute, Howard Hughes Medical Institute, New York, NY). The pET-3d plasmid corresponds to the D cyclin cDNA cloned in pET-12 as are characteristic of each cell type. 1 described (50). The pCMJ3/cdk-4 plasmid contains the mouse cdk-4 full- The antiinflammatory role of adenosine has been associated to length cDNA cloned in pBluescript KS as described (51). The probe for the its effects on neutrophil activity. Additionally, adenosine may also 18S rRNA was obtained as described (52). play an important role in the attenuation of macrophage activity, as it modulates several functions of macrophages, such as the regu- Proliferation assay lation of nitrite production (39, 44), the inhibition of LPS-induced Cell proliferation was measured as previously described (53, 54) with mi- TNF-␣ expression (39, 45), and the induction of IL-6 (46) and nor modifications. The cells were deprived of M-CSF for 18 h and then 105 IL-10 production (39, 47). BMDM were incubated for 24 h in 24-well plates (3424 MARK II; Costar, Cambridge, MA) in 1 ml of complete medium in the presence or absence We have found that adenosine inhibits M-CSF-dependent pro- of the indicated adenosine analogues or derivatives. After this period of liferation of murine bone marrow-derived macrophages (BMDM). time, the medium was removed and replaced with 0.5 ml of media con- 3 ␮ To do this, adenosine interacts with the A2B receptor at the cell taining [ H]thymidine (1 Ci/ml). After two additional h of incubation at surface and induces a subsequent increase of cAMP levels. Treat- 37°C, the medium was removed and the cells were fixed in ice-cold 70% ment of macrophages with adenosine does not inhibit the activa- methanol. After three washes in ice-cold 10% TCA, the cells were solu- bilized in 1% SDS, 0.3 N NaOH. Radioactivity was counted by liquid tion of the ERK pathway. Instead, adenosine induces the expres- scintillation using a 1500 Packard Tri-Carb scintillation counter (Meriden, sion of p27kip-1 in a protein kinase A (PKA)-dependent pathway, CT). Each experiment was performed three times, and the results were expressed as the mean Ϯ SD. thus causing the growth arrest at the G1 phase of the cell cycle without inducing apoptosis. These results reveal the molecular Determination of cAMP mechanism involved in the adenosine-mediated inhibition of the M-CSF-dependent proliferation of macrophages and remark the The production of cAMP was measured using a standard procedure. 6 relevance of this nucleoside as an immunosuppressor of macro- Briefly, 10 macrophages were cultured in 24-well plates in complete media. The cells were stimulated with the indicated adenosine agonists for 15 phage activity and proliferation. min. Extraction of cAMP from the cells was conducted using a liquid phase extraction method. Ice-cold ethanol was added to the cell suspension to a Materials and Methods final concentration of 65% (v/v) ethanol. After being allowed to settle, the Reagents supernatants were transferred to test tubes and centrifuged at 2000 ϫ g for 15 min at 4°C. The supernatants were transferred to new tubes and dried Adenosine, 2-chloroadenosine (CADO), 5Ј-N-ethylcarboxamidoadenosine using a speed-vac system (Bio-Rad, Hercules, CA). The dried extracts were (NECA), N6-(R)-phenylisopropyladenosine (R-PIA), and 1,3-dipropyl-8- dissolved in assay buffer, and the amount of cAMP was analyzed using a cyclopentylxanthine (DPCPX) were obtained from Sigma (St. Louis, nonacetylation cAMP enzyme immunoassay system (Amersham). Each 4142 ADENOSINE BLOCKS MACROPHAGE PROLIFERATION THROUGH p27

sample was analyzed in triplicate, and the results were represented as the Determination of ERK activity by in-gel-kinase assay mean Ϯ SD. First, 50 ␮g of total protein were separated by 12.5% SDS-PAGE in the Determination of c-fms cell-surface expression presence of 0.1 mg/ml of myelin basic protein (Sigma) copolymerized in the gel. After electrophoresis, SDS was removed by washing the gel with Cell-surface staining was conducted using specific Abs and cytofluoromet- 6 two changes of 20% 2-propanol in 50 mM Tris-HCl (pH 8.0) for1hat ric analysis. After treatment with adenosine or its analogues for 24 h, 10 room temperature. The gel was then incubated with 50 mM Tris-HCl (pH cells were harvested and washed in cold PBS. After fixing with 2% para- 8.0) containing 5 mM 2-ME (buffer A) for1hatroom temperature. The formaldehyde during 30 min at 4°C, the cells were resuspended in 50 ml proteins were denatured by incubating the gel with two changes of 6 M PBS containing 5% FBS and then incubated at 4°C for 15 min with 1 guanidine-HCl for1hatroom temperature and then renatured by incu- ␮ 6 g/10 cells of anti-CD16/CD32 mAb to block Fc receptors. Then, the bating with five changes of buffer A containing 0.04% Tween 20 for 16 h cells were incubated for1hatroom temperature with murine c-fms-spe- at 4°C. To perform the phosphorylation assay, the gel was first equilibrated ␮ 6 cific Ab (1 g/10 cells). The cells were then washed by centrifugation in 40 mM HEPES-NaOH (pH 7.4) containing 2 ␮M DTT, 0.1 mM EGTA, through a FBS cushion. Finally, cells were incubated with FITC-conju- ␮ 15 mM MgCl2, 300 M sodium orthovanadate for 30 min at 25°C and then gated anti-mouse IgG Ab for1hat4°C. Stained cell suspensions were incubated for1hinthesame solution containing 50 ␮M ATP and 100 ␮Ci analyzed using an Epics XL flow cytometer (Coulter, Hialeah, FL). FITC [␥-32P]ATP (ICN Pharmaceuticals, Costa Mesa, CA). The reaction was excitation was obtained using a 488-nm Argon laser lamp, and its fluores- stopped by washing the gel with 5% TCA containing 10 mM sodium py- cence was collected with a 525-nm band-pass filter. The parameters used rophosphate to inhibit phosphatase activity. The gel was dried, exposed to to select cell populations for analysis were forward and side light scatter. x-ray films (Kodak, Rochester, NY), and quantitated with a Bio-Rad mo- As a control for specificity, we used a nonrelated Ab. lecular analyst system. Analysis of DNA content with DAPI Northern blot analysis 6 A total of 10 cells previously subjected to a specific treatment were re- Total cellular RNA (20 ␮g), extracted with the acid guanidinium thiocy- Downloaded from suspended and fixed in ice-cold 70% ethanol. The cells were then washed anate-phenol-chloroform method (57), were separated in 1% agarose with in PBS, resuspended in 0.2 ml of a solution containing 150 mM NaCl, 80 5mM3-[N-morpholino]propanesulfonic acid (pH 7.0)/1 M formaldehyde mM HCl, and 0.1% Triton X-100, and incubated at 4°C for 10 min. Af- buffer. The RNA was transferred overnight to a GeneScreen nitrocellulose terward, 1 ml of a solution containing 180 mM Na2HPO4, 90 mM citric membrane (Life Science Products, Boston, MA) and fixed by UV irradi- ␮ acid, and 2 g/ml DAPI, pH 7.4, was added to each sample. After incu- ation (150 mJ). For p21waf-1 mRNA detection, we obtained the full-length bating the cells at 4°C for 24 h, their fluorescence was measured with an cDNA of p21waf-1 by digesting pMH117 with EcoRI/HindIII and used it as Epics Elite flow cytometer (Coulter). For this analysis, we used an UV a probe. The D cyclin probe was prepared by digesting the pET-3d con-

1 http://www.jimmunol.org/ laser with an excitation beam of 25 mW at 333–364 nm, and fluorescence struct with Bgl2/EcoRV. PCMJ3/cdk-4 digestion with EcoRI allowed us to was collected with a 525-nm band-pass filter. Cell doublets were gated out obtain a cdk-4 probe. To check for differences in RNA loading, we ana- by comparing the pulse area vs the pulse width. A total of 12,000 cells were lyzed the expression of the 18S rRNA transcript. All probes were labeled counted for each histogram, and cell cycle distributions were analyzed with with [␣-32P]dCTP (ICN Pharmaceuticals) with the oligolabeling kit the Multicycle program (Phoenix Flow Systems, San Diego, CA). method (Pharmacia Biotech, Uppsala, Sweden). After incubating the membranes for 18 h at 65°C in hybridization solution (20% formamide, 5ϫ Chromatin fragmentation assay ϫ Denhart’s, 5 SSC, 10 mM EDTA, 1% SDS, 25 mM Na2HPO4,25mM 6 32 Fragmentation of DNA due to internucleosomal cleavage was determined NaH2PO4, 0.2 mg/ml salmon sperm DNA, and 10 cpm/ml of P-labeled as described previously (55). Briefly, 3 ϫ 106 cells were harvested and probe), they were exposed to Kodak X-AR films. The bands of interest washed in ice-cold PBS. The cells were lysed in 0.5 ml of lysis buffer (50 were quantified with a molecular analyst system (Bio-Rad).

mM Tris-HCl, 10 mM EDTA, 1% SDS, pH 8.0) for 16 h at 4°C, and the by guest on September 28, 2021 lysates were centrifuged (15,000 ϫ g) to separate high m.w. DNA (pellet) Results from cleaved low m.w. DNA (supernatant). The DNA supernatants were phenol-extracted twice and precipitated. The pellets were resuspended in Adenosine induces a serie of functional alterations in macro- Tris-EDTA buffer containing 250 ␮g/ml RNase (Boehringer Mannheim, phages. We used macrophages obtained from bone marrow cul- Mannheim, Germany). The samples were heated at 65°C for 10 min and tures, because they represent an homogeneous population of mac- subjected to electrophoresis in a 2% agarose gel containing ethidium rophages that require M-CSF to proliferate and survive. Under the bromide. effect of M-CSF, macrophages proliferate in a dose-dependent Protein extraction and Western blot analysis manner (Fig. 1A). The addition of 100 ␮M adenosine inhibited Cells were washed twice in cold PBS and lysed on ice with lysis solution macrophage proliferation almost completely. The adenosine-in- (1% Triton X-100, 10% glycerol, 50 mM HEPES, pH 7.5, 150 mM NaCl, duced inhibition of M-CSF-dependent proliferation was a dose- ␮ ␮ protease inhibitors). Then, 100 M sodium orthovanadate was added to dependent process with an IC50 of 45 M (Fig. 1B). inhibit the activity of tyrosine phosphatases when necessary. The protein Some responses to adenosine are mediated through its direct concentration of the samples was determined by the Bio-Rad protein assay. The proteins from the cell lysates (100–150 ␮g) were boiled at 95°C in interaction with target molecules inside the macrophages, whereas Laemmli SDS-loading buffer, separated in 12% SDS-PAGE, and electro- others are caused by its recognition by specific cell-surface recep- transferred to nitrocellulose membranes (Hybond-ECL; Amersham). The tors. Adenosine internalization is mediated through mechanisms of membranes were blocked for at least1hatroom temperature in TBS-T nucleoside transport. Recently, four types of adenosine receptors containing 5% nonfat dry milk and then incubated with TBS-T containing have been characterized and they have been classified according to the primary Ab. For p27kip-1 and ␤-actin immunoblotting, incubation was performed for1hatroom temperature. After three washes of 15 min each their biochemical properties. These receptors are seven transmem- in TBS-T, the membranes were incubated with peroxidase-conjugated anti- brane-domain proteins coupled to G proteins (58). Recently, using mouse or anti-rabbit IgG Abs (Cappel) for 1 h. After three washes of 15 specific Abs and radiolabeled ligands we have identified the pres- min with TBS-T, enhanced chemiluminescence detection was performed ence of A and A , but not A or A , adenosine receptors in (Amersham) and the membranes were exposed to x-ray films (Amersham). 2B 3 1 2A Quantification of the blot was conducted by densitometric analysis. bone marrow macrophages (59). To determine the type of adenosine receptor that mediates the inhibition of macrophage prolifer- Determination of the ERK phosphorylation state by mobility ation, we measured the activity of adenosine analogues that at shift assay small concentrations are recognized specifically by each type of This assay was performed as described for the Western blot analysis with receptors. However, at higher concentrations these analogues slight modifications (56). Proteins from cell lysates (50–100 ␮g) were could also affect other adenosine receptors. NECA (60) and CADO subjected to 7.5% SDS-PAGE to allow efficient separation of the phos- (61) are synthetic adenosine analogues recognized by all types of phorylated and dephosphorylated forms of ERK. The incubation of the membranes with anti-ERK-1/2 primary Ab or peroxidase-conjugated anti- adenosine receptors. NECA and CADO inhibited M-CSF-depen- Ͻ ␮ mouse IgG Ab were done sequentially in TBS-T for 1 h at room dent proliferation with an IC50 10 M (Fig. 2A). In contrast, temperature. confirming our previous results, other analogues such as R-PIA, The Journal of Immunology 4143 Downloaded from http://www.jimmunol.org/

FIGURE 1. Adenosine inhibits M-CSF-dependent proliferation of BMDM. A, BMDM were obtained after 7 days of culture in the presence FIGURE 2. Adenosine inhibits M-CSF-dependent proliferation through 5 of M-CSF. A total of 10 macrophages were incubated in 24-well plates in activating the A2B adenosine receptors. A, The effects of synthetic adeno- the presence of the indicated amounts of M-CSF either alone (Control) or sine analogues on macrophage proliferation were analyzed by [3H]thymi- Ϫ with 10 4 M adenosine. Proliferation was determined as indicated in Ma- dine incorporation. A total of 105 macrophages grown in the presence of terials and Methods. B, Adenosine inhibits macrophage proliferation in a 1000 U/ml of M-CSF were treated with NECA, CADO, R-PIA, CGS -At this concen ,ء .dose-dependent manner. A total of 105 macrophages were incubated in 21680, and IB-MECA at the indicated concentrations

24-well plates in the presence of 1000 U/ml of M-CSF and the indicated tration, R-PIA binds to A2B receptors. B, The A2B adenosine receptor an- amounts of adenosine. Control cells were incubated with M-CSF alone. tagonist DPCPX blocks the inhibitory effect of NECA on macrophage pro- by guest on September 28, 2021 Each determination was made in triplicate, and the values represented cor- liferation. A total of 105 cells cultured in 24-well plates were treated with respond to the mean Ϯ SD of one representative of four independent NECA at the indicated concentrations in the presence or absence of 10Ϫ6 experiments. or 10Ϫ8 M DPCPX. Control cells from A and B were incubated in the presence of M-CSF alone. Each determination was made in triplicate, and the values represented correspond to the mean Ϯ SD of one representative of three independent experiments. specific for A1 receptor (62), IB-MECA, specific for A3 (63) and CGS21680, recognized by A2A receptors (64), showed little or no effect as inhibitors of macrophage proliferation at those concen- proliferation. Treatment of macrophages with 10Ϫ5 M CGS 21680, trations at which these compounds are specific for their respective 10Ϫ6 M R-PIA, or 10Ϫ5 M IB-MECA did not induce a detectable receptors, which are at nanomolar range (Fig. 2A); however, at increase of intracellular cAMP levels. To date, there are no specific Ͼ ␮ higher concentrations ( 10 M) they could also bind to A2B re- agonists for A2B receptors. The results, obtained with adenosine, ceptors and thus inhibit macrophage proliferation. The order of CADO, or NECA, three nonspecific agonists, indicate that the ef- potency of adenosine analogues as inhibitors was: NECA Ͼ fect of adenosine is mediated through their interaction with the Ն Ͼ Ͼ Ͼ CADO R-PIA adenosine IB-MECA CGS21680. There- A2B adenosine receptor, because specific agonists for the other fore, these results suggested that adenosine inhibits M-CSF-depen- adenosine receptor subtypes did not have any effect on cAMP pro- dent proliferation by interacting with the A2B receptor. These data duction at those concentrations specific for their receptors, while at were confirmed by the fact that the adenosine antagonist DPCPX, higher concentrations they could bind to A2B receptors and induce which inhibits A2B receptors (65), blocked the inhibitory effect of cAMP production, thus inhibiting macrophage proliferation. This NECA on the proliferation of macrophages in a dose-dependent is the case of R-PIA, an inhibitor of A1 receptors (data not shown). manner (Fig. 2B). Furthermore, 10Ϫ6 M DPCPX inhibited completely the cAMP in- The mechanism mediated by adenosine to inhibit macrophage crease induced by 10Ϫ5 M NECA (data not shown). Although we proliferation was further investigated. As adenosine receptors are cannot exclude the presence of other adenosine receptors, all our coupled to adenylyl cyclase (66, 67), we stimulated macrophages results suggest that the A2B adenosine receptor is the main respon- with different agonists and measured the intracellular production of sible for the cAMP increases found in response to adenosine cAMP. The treatment of BMDM with 10Ϫ5 M NECA induced a in BMDM. marked increase in the intracellular levels of cAMP (Fig. 3A). The natural ligand adenosine induced a lower macrophage re- Similar results were obtained when macrophages were treated with sponse than NECA, probably due to the surface expression of ecto- 5 ϫ 10Ϫ5 M adenosine or 10Ϫ5 M CADO, although in these cases adenosine deaminase, a molecule that degrades adenosine before it the increase was lower than that mediated by NECA, which was in can be metabolized (68). Therefore, we used NECA in the follow- accordance with their relative potency as inhibitors of macrophage ing experiments. The NECA-induced production of intracellular 4144 ADENOSINE BLOCKS MACROPHAGE PROLIFERATION THROUGH p27 Downloaded from FIGURE 3. Induction of cAMP production by adenosine agonists. A, A total of 106 BMDM were stimulated for 15 min with 5 ϫ 10Ϫ5 M adenosine (Ado), 10Ϫ5 M NECA, 10Ϫ5 M CADO, 10Ϫ5 M IB-MECA, 10Ϫ5 M CGS 21680, or 10Ϫ6 M R-PIA. The cultures were liquid phase extracted (see Materials and Methods), and the cAMP content was measured with an enzyme immunoassay system (Amersham). B, The production of cAMP in BMDM stimulated with NECA is dose-dependent. The cells were treated for 15 min with the indicated concentrations of NECA. C, Time-course of NECA-induced cAMP production in macrophages. BMDM were treated with 10Ϫ5 M NECA for the indicated periods of times. Each sample was analyzed in triplicate, and the data are represented as the mean Ϯ SD of triplicate determinations of two independent experiments. http://www.jimmunol.org/

cAMP in BMDM was both dose- and time-dependent (Fig. 3, B So far we have shown that the adenosine-induced inhibition of ␮ and C), with an EC50 of 5 M similar to that observed in previous M-CSF-dependent proliferation is mediated through the A2B re- reports (69). ceptors, and that the interaction of adenosine with these receptors by guest on September 28, 2021

FIGURE 4. cAMP inhibits M-CSF-dependent proliferation of macrophages. Macrophages were treated with 8-Br-cAMP (A) or forskolin (B)atthe indicated concentrations, and thymidine incorporation was measured as described in Materials and Methods. Control cells were grown in the presence of 1000 U/ml of M-CSF with no other treatment. C, NECA inhibits macrophage proliferation through an adenylyl cyclase- and PKA-dependent pathway. A total of 105 macrophages grown in the presence of the indicated concentrations of M-CSF (Control) were treated with 10Ϫ5 M NECA alone or combined with either 10Ϫ4 M SQ 22536 or 10Ϫ7 M KT 5720. D, Dose-dependent effect of PKA and adenylyl cyclase inhibitors on NECA-induced inhibition of macrophage proliferation. Macrophages were treated with 1000 U/ml of M-CSF alone or combined with 10Ϫ5 M NECA and the indicated concentrations of SQ 22536 or KT 5720. Each determination was made in triplicate, and the values represented correspond to the mean Ϯ SD of one representative of three independent experiments. The Journal of Immunology 4145

FIGURE 5. Adenosine analogues do not modify c-fms surface expression in macrophages. A, A total of 106 macrophages were deprived of Downloaded from M-CSF for 24 h or treated with 5 ϫ 10Ϫ5 M adenosine in the presence of M-CSF. The control cells were grown in the presence of M-CSF with no additional treatment. The c-fms surface expression was analyzed by ﬂow cytometry using mAbs. B, Quantiﬁcation of c-fms surface expression after M-CSF starvation or treatment with 5 ϫ 10Ϫ5 M adenosine (Ado), 10Ϫ5 M NECA, 10Ϫ5 M forskolin (Forsk), or 10Ϫ4 M 8-Br-cAMP for 24 h. Data are represented as the mean Ϯ SD of two experiments. http://www.jimmunol.org/

induces the production of cAMP. We were also interested in studying the role of cAMP in the inhibition of M-CSF-dependent proliferation of macrophages. 8-Br-cAMP, a cell membrane-perme- able and nonmetabolizable cAMP analogue, inhibited macrophage proliferation in response to 1200 U/ml of M-CSF. The IC50 for this inhibition was 85 ␮M (Fig. 4A). Treatment of macrophages with by guest on September 28, 2021 forskolin, a drug that directly activates adenylyl cyclase and in- FIGURE 6. Adenosine causes a growth arrest of macrophages at the G /G phase of the cell cycle and does not induce apoptosis. A, A total of duces the generation of cAMP, also inhibited M-CSF-dependent 0 1 106 macrophages grown in the presence of M-CSF were treated or not with macrophage proliferation in a dose-dependent manner (Fig. 4B). Ϫ 5 ϫ 10 5 M adenosine for 24 h and their DNA content was measured by This suggested that the production of cAMP was sufficient to in- flow cytometry. Cell cycle distribution was analyzed using Immuno-4 soft- hibit macrophage proliferation. To confirm that the activation of ware (Coulter). The apoptotic cells (Apop) appear as a subdiploid fraction. adenylyl cyclase and the production of cAMP was responsible for B, A total of 3 ϫ 106 BMDM were treated with 5 ϫ 10Ϫ5 M adenosine, the inhibitory effect of NECA on macrophage proliferation, we 10Ϫ5 M NECA, or 10Ϫ5 M forskolin for 24 h, and chromatin fragmentation analyzed the effect of the inhibitor of adenylyl cyclase SQ 22536. was analyzed by electrophoresis in an agarose gel. The positive control This drug blocked the inhibitory effect on macrophage prolifera- cells for apoptosis correspond to macrophages treated with 5 ␮g/ml acti- tion induced by NECA (Fig. 4, C and D), thus indicating that nomycin D. Negative control cells correspond to macrophages grown in the NECA inhibits M-CSF-dependent proliferation through activating presence of M-CSF with no additional treatment. Both analyses were per- the adenylyl cyclase and the subsequent production of cAMP. formed twice with identical results. Moreover, treatment of macrophages with KT 5720, a PKA inhibitor, also blocked the anti-proliferative effect of NECA (Fig. 4, C and D). This confirms that NECA inhibits macrophage prolifera- (Fig. 6). In response to adenosine, macrophages appeared to be tion through a PKA-dependent pathway. distributed homogeneously (88% of total cells) in a peak corre-

We next determined the molecular mechanism used by adeno- sponding to the G1 phase of the cell cycle (Fig. 6A). We did not sine and its analogues to inhibit macrophage proliferation. First, find any subdiploid peak corresponding to apoptotic cells. Besides, we studied the effect of adenosine on the modulation of M-CSF cells treated with adenosine, NECA, or forskolin did not show receptors. An 18-h starvation of M-CSF induced a 4- to 5-fold DNA fragmentation in comparison to cells in which apoptosis had increase in the level of expression of M-CSF receptors compared been induced by treatment with actinomycin D (70) (Fig. 6B). with that observed in cells growing in the presence of M-CSF These results indicated that adenosine-mediated inhibition of pro- (control) (Fig. 5). In contrast, no modification in the number of liferation was not due to a massive induction of apoptosis. Instead, M-CSF receptors was detected in macrophages incubated for 24 h the cell cycle stop induced by adenosine explains the inhibition of with adenosine, NECA, forskolin, or 8-Br-cAMP (Fig. 5). There- proliferation. Therefore, we were interested in studying the mech- fore, the adenosine-induced inhibition of M-CSF-dependent pro- anisms used by adenosine to stop the cell cycle at the G1 phase. liferation was not due to an altered expression of M-CSF receptors. In an initial approach, we first analyzed the effects of adenosine Bone marrow macrophage cultures grown in the presence of on the activation of the ERK pathway. The activation of ERK-1/2 M-CSF are not cell cycle-synchronized and showed a random dis- is required for M-CSF-dependent proliferation of macrophages (5). tribution, with 51% of cells in G0/G1, 30% in S, and 17% in G2/M Phosphorylation of ERK-2 was analyzed by a mobility shift assay. 4146 ADENOSINE BLOCKS MACROPHAGE PROLIFERATION THROUGH p27 Downloaded from

FIGURE 8. Effect of adenosine agonists on the cell cycle machinery. A, The effect of adenosine and related agents on the cell cycle machinery was

analyzed by Northern blotting. Macrophages were treated for 24 h with http://www.jimmunol.org/ FIGURE 7. An increase in the intracellular levels of cAMP does not inhibit Ϫ 10 5 M of each of the indicated adenosine analogues and related agents in the M-CSF-induced activation of the ERK pathway. A, Cell extracts were the presence of M-CSF. Then, 20 ␮g of total RNA from each sample was obtained after stimulation of quiescent macrophages with or without M-CSF blotted for the expression of p21waf-1,D cyclin, and cdk-4 kinase. In this for 10 min in the presence or absence of 5 ϫ 10Ϫ5 M adenosine, 10Ϫ5 M 1 figure, we show the expression and quantification of one representative of NECA, 10Ϫ5 M forskolin, or 10Ϫ4 M 8-Br-cAMP. The induction of ERK-2 three independent experiments. phosphorylation by M-CSF was assessed with a mobility shift assay. Then, 80 ␮g of total protein were loaded per lane. The positions of phosphorylated and dephosphorylated forms of ERK-2 are indicated with arrows. B, cAMP does not modify the kinetics of the M-CSF-induced ERK activation. Quiescent cell cycle, we studied the expression of G cdks. The expression of Ϫ5 1 macrophages were either left untreated or preincubated with 10 M NECA the mRNA for the components of the cyclin D/cdk-4 (Fig. 8) or by guest on September 28, 2021 for 15 min and then stimulated with M-CSF (1200 U/ml) for the indicated cyclin E/cdk-2 (data not shown) complex were not modified by times. ERK activity was analyzed with an in-gel-kinase assay. treatment with adenosine and its analogues. Moreover, the analysis of the immunoprecipitates and blotting of the cyclin D/cdk-4 complex formation showed no differences between control and NECA- ERK-2 phosphorylation can be used as an indicator of ERK acti- treated cells (data not shown).We also analyzed the expression of vation in macrophages, because there is a close correlation be- cdk inhibitors. The expression of p21waf-1, a dual inhibitor of cdk tween the phosphorylation state of ERK-1/2 and their activity in an required for the passage through the cell cycle (28), was not mod- in-gel-kinase assay (data not shown). M-CSF induced ERK-2 ified by adenosine or analogues (Fig. 8). phosphorylation in bone marrow macrophages (Fig. 7A). No dif- In contrast, adenosine, NECA, forskolin, and 8-Br-cAMP in- ferences in ERK-2 phosphorylation were observed when M-CSF- duced a 3- to 8-fold increase of p27kip-1 protein levels (Fig. 9). kip-1 treated macrophages were incubated in the presence of adenosine, p27 is another G1 cki of the CIP/KIP family that binds to NECA, or 8-Br-cAMP (Fig. 7A). We also analyzed the effect of cyclin/cdk complexes and inhibits their activity (29). NECA in- NECA on the kinetics of ERK activation by in-gel-kinase assays duced the expression of p27kip-1 in a time- and dose-dependent using myelin basic protein as a substrate for ERK-dependent phos- manner (Fig. 9, B and C). The induction of p27kip-1 by NECA was phorilation (Fig. 7B). The addition of M-CSF to quiescent macro- observed after6hoftreatment and it was maintained during the phages induced the rapid activation of ERK proteins. ERK acti- whole time-course of 24 h. Besides, the expression of p27kip-1 vation peaked after 5 min and then the activity of these kinases induced by NECA was inhibited by the treatment with either QS decreased progressively to basal levels. Treatment of macrophages 22536 or KT 5720 (Fig. 9D); therefore, the expression of p27kip-1 with 10Ϫ5 M NECA did not modify the kinetics of the M-CSF- depended on the production of cAMP by activated adenylyl cy- induced ERK activation. We also analyzed the expression of clase and on the activation of the PKA pathway. Thus, the induc- MKP-1, a dual specificity phosphatase responsible for the inacti- tion of p27kip-1 expression mediated by adenosine and its ana- vation of ERK-1/2. According to the results on ERK phosphory- logues may be responsible for the adenosine-induced arrest of lation and activity, treatment of macrophages with NECA, forsko- macrophages at the G1 phase of the cell cycle. lin, or 8-Br-cAMP did not modify the mRNA nor protein To confirm this hypothesis, we analyzed the inhibitory effect of expression of MKP-1 (data not shown). These results indicated NECA on macrophage proliferation using macrophages from mice that adenosine and related drugs did not affect the activation/de- with the p27kip-1 gene disrupted. In these macrophages, in contrast activation of the ERK pathway in macrophages. with that observed in macrophages from normal mice, NECA, but We next determined the effects of adenosine on different ele- not IFN-␥, did not inhibit cell proliferation (Fig. 10), demonstrat- ments involved in cell cycle control. Because adenosine and its ing that the expression of p27kip-1 is necessary for the inhibition by analogues cause an arrest of macrophages at the G1 phase of the NECA of the M-CSF-dependent proliferation of macrophages. The Journal of Immunology 4147 Downloaded from

FIGURE 9. Adenosine induces the expression of p27kip-1. A, The induction of p27kip-1 by adenosine and related agents was measured by Western blotting after treating the macrophages for 24 h with 10Ϫ5 M of each of the indicated adenosine analogues and related agents in the presence of M-CSF. The blots were quantiﬁed by densitometry, and the results are indicated as fold-induction of p27kip-1 expression compared with the level of expression in control cells grown in the presence of M-CSF with no additional treatment. B, The induction of p27kip-1 by NECA is dose-dependent. BMDM were treated for 24 h with the indicated concentrations of NECA. C, The induction of p27kip-1 by NECA is time-dependent. BMDM were treated with 10Ϫ5 M NECA at the indicated http://www.jimmunol.org/ times. D, NECA-induced p27kip-1 expression is adenylyl cyclase- and PKA-dependent. Macrophages were treated or not with 10Ϫ5 M NECA alone or in combination with 10Ϫ4 M SQ 22536 or 10Ϫ7 M KT 5720. Then, 100 ␮g of total protein were loaded per lane and p27 expression was analyzed as indicated in Materials and Methods. The expression of ␤-actin was used as a control of sample loading and transfer efﬁciency. Each experiment was performed twice.

Discussion macrophages, inhibits M-CSF-induced proliferation and also down-modulates the expression of c-fms (74). However, neither Most experiments on proliferation and cell cycling have been con- adenosine nor cAMP-increasing agents modulated the expression ducted using transformed cell lines. In this report, we have used pri-

of c-fms, thus indicating that growth arrest was not due to a re- by guest on September 28, 2021 mary cultures of BMDM, which is an homogeneous population that duction in M-CSF recognition at the cell surface. responds to physiological proliferative or activating stimuli (54). One of the earliest events in the signal transduction of M-CSF is We have found that adenosine blocks M-CSF-dependent prolif- the activation of ERK-1/2 (15, 75). The inhibition of ERK activa- eration of macrophages. The effect of adenosine seems to be me- tion by using PD98059, a specific inhibitor of the mitogen-acti- diated through the engagement of the A adenosine receptor. This 2B vated protein/ERK kinase, blocked macrophage proliferation in is supported by the use of specific agonists and antagonists for the 5 different types of adenosine receptors. NECA, an agonist for the response to M-CSF. Recently, several reports have shown that the four types of adenosine receptors so far described, induces a stron- capability of Ras to activate Raf-1 was impaired in cells treated ger inhibition of M-CSF-dependent proliferation than the agonists with cAMP-elevating agents, leading to a loss in the capability to activate ERKs (76). Therefore, the activation of PKA may prevent specific for A1,A3, and A2A adenosine receptors. The NECA- induced increase of the intracellular levels of cAMP is likely me- growth factor-mediated cell division by interfering with the activation of ERK-1/2. However, cAMP enhances the M-CSF-induced diated by A2B receptors, because DPCPX, an inhibitor of this type of receptors, inhibits the production of cAMP in response to ERK activity in macrophages (77). In our experiments, neither NECA. This confirms our previous observations about the expres- adenosine nor cAMP altered the phosphorylation state or the ac- sion of adenosine receptors in macrophages (59). By using binding tivation kinetics of ERK-2 in response to M-CSF. Besides, the assays with radiolabeled NECA in competition with different ago- expression of the phosphatase MKP-1 was not modified by cAMP- nists and immunoblotting with specific Abs, we have identified the increasing agents. Therefore, cAMP does not seem to inhibit macrophage proliferation by reducing the capability of M-CSF to ac- presence of A2B and A3 adenosine receptors on the cell surface of tivate the ERK pathway. macrophages, whereas A1 and A2A receptors are poorly represented in these cells. Macrophages incubated with cAMP are blocked at the G1 phase The inhibition of M-CSF-dependent proliferation by adenosine despite synthesizing normal amounts of cyclin D1 and cdk-4. The may be caused by an increase in the production of cAMP. Both phosphorylation of cdk4 at threonine 172 is necessary for cdk-4 cAMP and PKA activators inhibit the proliferation of macro- activation. The cdk-activating kinase precipitated from cAMP- phages. Treatment of BMDM with adenosine arrests them in the treated cells was as active as that obtained from nontreated pro-

G1 phase of the cell cycle, as it has been already demonstrated with liferating cells. This suggested the presence of an inhibitory activ- cAMP analogues (71–73). Moreover, the inhibition of adenylyl ity present in cell lysates of cAMP-treated cells. Recently p27kip-1, cyclase and PKA blocks the antiproliferative effect of NECA in an inhibitor of cdk-4, has been identified (72, 78). We have studied macrophages. the effects of adenosine analogues on the expression of two cdk-4 The first step in the induction of macrophage proliferation in inhibitors, p21waf-1 and p27 kip-1. We detected p27kip-1 expression response to M-CSF is the interaction of this growth factor with its only in those cells treated with cAMP-increasing agents, which specific cell-surface receptor, c-fms. IFN-␥, the major activator of may account for the cAMP-mediated arrest of the cell cycle at the 4148 ADENOSINE BLOCKS MACROPHAGE PROLIFERATION THROUGH p27

ERK1/2 Abs. We also thank Dr. J. Massague´(Sloan Kettering Institute, Howard Hughes Medical Institute, New York, NY) for the pMH117 plasmid and Dr. J. Roberts (Howard Hughes Medical Institute, Seattle, WA) for the the p27 kip-1 knockout mice. We especially thank Dr. Gabriel Gil (Institut Municipal d’Investigaciones Biome´diques, Barcelona, Spain) for his help with the p27 kip-1 knockout mice. We also thank Martı´n Cullell- Young for the revision of the manuscript.

References 1. Abbas, A. K., A. H. Lichtman and J. S. Pober. 1991. Cells and tissues of the immune system. In Cellular and Molecular Immunology. W. B. Saunders, Phil- adelphia, pp. 13–15. 2. Celada, A. 1994. Ce´lulas del sistema inmunitario: granulocitos y fagocitos mono- nucleares. In Immunologı´a Ba´sica. A. Celada, ed. Editorial LABOR S.A., Barce- lona, Spain, pp. 61–94. 3. Valledor, A. F., F. E. Borras, M. Cullell-Young, and A. Celada. 1998. Transcrip- tion factors that regulate monocyte/macrophage differentiation. J. Leukocyte Biol. 63: 405. 4. Nathan, C. F. 1987. Secretory products of macrophages. J. Clin. Invest. 79:319. 5. Stanley, E. R., K. L. Berg, D. B. Einstein, P. S. W. Lee, F. J. Pixley, Y. Wang, and Y. G. Yeung. 1997. Biology and action of colony-stimulating factor 1. Mol. Reprod. Dev. 46:4. Downloaded from 6. Vairo, G., and J. A. Hamilton. 1991. Signalling through CSF receptors. Immunol. Today 12:361. 7. Hamilton, J. A. 1997. CSF-1 signal transduction. J. Leukocyte Biol. 62:145.

8. Roussel, M. F. 1997. Regulation of cell cycle entry and G1 progression by CSF-1. Mol. Reprod. Dev. 46:11. 9. Courtneidege, S. A., R. Dhand, D. Pilat, G. M. Twarmley, M. D. Waterﬁeld, and M. F. Roussel. 1993. Activation of Src family kinases by colony stimulating factor-1, and their association with its receptor. EMBO J. 12:943.

10. Novak, U., A. G. Harpur, L. Paradiso, V. Kanagasundaram, A. Jaworowski, http://www.jimmunol.org/ A. F. Wilks, and J. A. Hamilton. 1995. CSF-1 induced STAT1 and STAT3 activation is accompanied by phosphorylation of Tyk2 in macrophages and Tyk2 and JAK1 in fibroblasts. Blood 86:2948. FIGURE 10. NECA did not inhibit M-CSF-dependent proliferation of 11. Novak, U., A. Mui, A. Miyajima, and L. Paradiso. 1996. Formation of STAT5 BMDM from p27kip-1 knockout mice. BMDM were obtained after 7 days containing DNA binding complexes in response to colony stimulating factor 1 of culture in the presence of M-CSF. A total of 105 macrophages from and platelet-derived growth factor. J. Biol. Chem. 271:26947. control (A) and p27kip-1 knockout (B) mice were incubated in 24-well plates 12. Katayama, N., M. Nishikawa, N. Mirami, and S. Shirakawa. 1989. Putative in- Ϫ5 volvement of protein kinase C in proliferation of human myeloid progenitor cells. in the presence of the indicated amounts of M-CSF alone or with 10 M Blood 73:123. NECA or 300 U/ml IFN-␥. Proliferation was determined as indicated in 13. Yusoff, P., J. A. Hamilton, R. D. Nolan, and W. A. Phillips. 1994. Hematopoietic Materials and Methods. Each determination was made in triplicate, and the colony stimulating factors CSF-1 and GM-CSF increase phosphatidylinositol by guest on September 28, 2021 values represented correspond to the mean Ϯ SD of one representative of 3-kinase activity in murine bone marrow-derived macrophages. Growth Factors 10:181. two independent experiments. 14. Roche, S., M. Koel, and S. A. Coutneidge. 1994. The phosphatidylinositol 3-kinase ␣ is required for DNA synthesis induced by some, but not all, growth factors. Proc. Natl. Acad. Sci. USA 91:9185. 15. Bu¨scher, D., R. A. Hipskind, S. Krautwald, T. Reimann, and M. Baccarini. 1995. G1 phase. Moreover, cAMP-increasing agents induce the expres- kip-1 Ras-dependent and -independent pathways target the mitogen-activated protein sion of p27 in an adenylyl cyclase- and PKA-dependent path- kinase network in macrophages. Mol. Cell Biol. 15:466. way. The involvement of p27 in the adenosine-induced inhibition 16. Seger, R., and E. G. Krebs. 1995. The MAPK signaling cascade. FASEB J. 9:726. of the M-CSF-dependent proliferation was confirmed using mac- 17. Treisman, R. 1996. Regulation of transcription by MAP kinase cascades. Curr. kip-1 Opin. Cell Biol. 8:205. rophages from mice with the p27 gene disrupted (49), which 18. Pelech, S. L. 1993. Networking with protein kinases. Curr Biol. 3:513. are unresponsive to the inhibitory effect of NECA on macrophage 19. Charles, C. H., H. Sun, L. F. Lau, and N. K. Tonks. 1993. The growth factor- proliferation. inducible immediate-early gene 3CH134 encodes a protein-tyrosine-phosphatase. Proc. Natl. Acad. Sci. USA 90:5292. The data presented in this report allow us to suggest a model for 20. Keyse, S. M., and E. A. Emslie. 1992. Oxidative stress and heat shock induce a adenosine-mediated inhibition of macrophage proliferation. Aden- human gene encoding a protein-tyrosine phosphatase. Nature 359:644. 21. Sun, H., C. H. Charles, L. F. Lau, and N. K. Tonks. 1993. MKP-1 (3CH134), an osine binds to A2B receptors on the surface of macrophages and immediate early gene product, is a dual specificity phosphatase that dephospho- subsequently induces the production of cAMP by activating ad- rylates MAP kinase in vivo. Cell 75:487. enylyl cyclase. As a consequence of the cAMP increase, PKA is 22. Sun, H., N. K. Tonks, and D. Bar-Sagi. 1994. Inhibition of Ras-induced DNA kip-1 synthesis by expression of the phosphatase MKP-1. Science 266:285. activated and the expression of p27 is induced, thus blocking 23. Brondello, J. M., F. R. McKenzie, H. Sun, N. K. Tonks, and J. Pouysse´gur. 1995. the activity of G1 cyclin/cdk complexes. As a result, macrophages Constitutive MAPK phosphatase (MKP-1) expression blocks G1 specific gene cannot progress through the G phase of the cell cycle and their transcription and S-phase entry in fibroblasts. Oncogene 10:1895. 1 24. Sherr, C. J. 1994. G phase progression: cyclin on cue. Cell 79:551. proliferation is blocked even in the presence of M-CSF. 1 25. Draetta, G. F. 1994. Mamalian G1 cyclins. Curr. Opin. Cell Biol. 6:842. These results could have clinical relevance. In fact, the inhibi- 26. Sherr, J. C., and J. M. Roberts. 1995. Inhibitors of mammalian G1 cyclin-depen- tion of macrophage proliferation could be part of the immunosup- dent kinases. Genes Dev. 9:1149. 27. Serrano, M., G. J. Hannon, and D. Beach. 1993. A new regulatory motif in cell pressive effect of adenosine. The inhibition of macrophage prolif- cycle control causing specific inhibition of cyclin D/CDK4. Nature 366:704. eration and activation could result in the modulation and the 28. Xiong, Y., G. J. Hannon, H. Zhang, D. Casso, R. Kobayashi, and D. Beach. 1993. resolution of the inflammatory process. p21 is a universal inhibitor of cyclin kinases. Nature 366:710. 29. Toyoshima, H., and T. Hunter. 1994. p27, a novel inhibitor of G1 cyclin/cdk protein kinase activity, is related to p21. Cell 78:67. Acknowledgments 30. Sherr, C. J., and E. R. Stanley. 1990. Colony stimulating factor-1, in Peptide Growth Factors and Their Receptors. Springer-Verlag, Heidelberg, Germany. We acknowledge the help received from Jaume Comas and Rosario Gonza´- 31. Matsushime, H., M. F. Roussel, R. A. Ashmun, and C. J. Sherr. 1991. Colony- lez in charge of the flow cytometry facility of the Serveis Cientı´fico Teć- stimulating factor 1 regulates novel cyclins during the G1 phase of the cell cycle. Cell 65:701. nics de la Universitat de Barcelona (Barcelona, Spain). We thank Dr. Mi- 32. Cooper, J. A., and P. Whyte. 1989. RB and the cell cycle: entrance or exit? Cell chael Weber (University of Virginia, Charlottesville, VA) for the anti- 58:1009. The Journal of Immunology 4149

33. Vogelstein, B., and K. W. Kinzler. 1992. p53 function and dysfunction. Cell 57. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by acid 70:523. guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156. 34. Cobrinik, D., S. F. Dowdy, P. W. Hinds, S. Mittnacht, and R. A. Weinberg. 1992. 58. North, R. A., and E. A. Barnard. 1997. Nucleotide receptors. Curr. Opin. Neu- The retinoblastoma protein and the regulation of cell cycling. Trends Biochem. robiol. 7:346. Sci. 17:312. 59. Xaus, J., M. Mirabet, J. Lloberas, C. Soler, C. Lluis, R. Franco, and A. Celada. 35. Murray, A. 1994. Cell cycle checkpoints. Curr. Opin. Cell Biol. 6:872. ␥ 1999. IFN- up-regulates the A2B adenosine receptor expression in macrophages: 36. Van Belle, H., F. Goossens, and J. Wynants. 1987. Formation and release of a mechanism of macrophage deactivation. J. Immunol. 162:3607. purine catabolites during hypoperfusion, anoxia and ischemia. Am. J. Physiol. 60. Bruns, R. F., G. H. Lu, and G. H. Pugsley. 1986. Characterization of the A2 252:H886. adenosine receptor labelled by [3H]-NECA in rat strial membranes. Mol. Phar- 37. Matherne, G. P., J. P. Headrick, S. D. Coleman, and R. M. Berne. 1990. Inter- macol. 29:331. sticial transudate purines in normoxic and hipoxic immature and mature rabbit 61. Wu, P. H., and J. W. Phillis. 1982. Adenosine receptors in rat brain membranes: hearts. Pediatr. Res. 28:348. 3 38. Cronstein, B. N. 1994. Adenosine, an endogenous anti-inflammatory agent. characterization of high affinity binding of [ H]-2-chloroadenosine. Int. J. Bio- J. Appl. Physiol. 76:5. chem. 14:399. 39. Hasko´, G., C. Szabo´, Z. H. Ne´meth, V. Kvetan, S. M. Pastores, and E. S. Vizi. 62. Trivedi, B. K., A. J. Bridges, and R. F. Bruns. 1990. Structure-activity relation- 1996. Adenosine receptor agonist differentially regulate IL-10, TNF-␣ and nitric- ships of adenosine A1 and A2 receptors. In Adenosine and Adenosine Receptors. oxide production in Raw-264.7 macrophages and in endotoxemic mice.J.Im- M. Williams, ed. Humana, Clifton, NJ, pp. 57. munol. 157:4634. 63. Gallo-Rodriguez C., X. D. Ji, N. Melman, B. D. Siegman, L. H. Sanders, 40. Dalziel, H. H., and D. P. Westfall. 1994. Receptors for adenine nucleotides and J. Orlina, B. Fisher, Q. L. Pu, M. E. Olah, P. J. M. Van Galen, G. L. Stiles, and nucleosides: subclassification, distribution, and molecular characterization. Phar- K. A. Jacobson. 1994. Structure-activity relationships of N6-benzyladenosine-5Ј- macol. Rev. 46:449. uronamides as A3-selective adenosine agonists. J. Med. Chem. 37:636. 41. Libert, F., M. Parmentier, A. Lefort, C. Dinsart, J. Van Sande, C. Maenhaut, 64. Jarvis, M. F., R. Schulz, A. J. Hutchinson, U. H. Do, M. A. Sills, and 3 M. J. Simons, J. E. Dumont, and G. Vassart. 1989. Selective amplification and clon- M. Williams. 1989. [ H]-CGS21680C, a selective A2 adenosine receptor agonist ing of four new members of the G protein-coupled receptor family. Science 24:570. directly labels A2 receptors in rat brain. J. Pharmacol. Exp. Ther. 251:888. 42. Stehle, J. H., S. A. Rivkees, J. J. Lee, D. R. Weaver, J. D. Deeds, and 65. Haleen, S. J., R. P. Seffen, and H. W. Hamilton. 1987. PD 116,948, a highly Downloaded from S. M. Reppert. 1992. Molecular cloning and expression of the cDNA for a novel selective A1 adenosine receptor antagonist. Life Sci. 40:555. A2-adenosine receptor subtype. Mol. Endocrinol. 6:384. 66. Van Calker, D., M. Muller, and B. Hamprecht. 1979. Adenosine regulates, via 43. Zhou, Q., C. Li, M. E. Olah, R. A. Johnson, G. L. Stiles, and O. Civelli. 1992. two different types of receptors, the accumulation of cyclic AMP in cultured brain Molecular cloning and characterization of an adenosine receptor: the A3 adeno- cell. J. Neurochem. 33:999. sine receptor. Proc. Natl. Acad. Sci. USA 89:7432. 67. Gurden, M. F., J. Coates, F. Ellis, B. Evans, M. Foster, E. Hornby, I. Kennedy, 44. Mullet, D., R. H. Fertel, D. Kniss, and G. W. Cox. 1997. An increase in intra- ␥ D. P. Martin, P. Strong, C. J. Vardey, and A. Wheeldon. 1993. Functional char- cellular cyclic AMP modulates nitric oxide production in IFN- -treated macro- acterization of three adenosine receptor types. Br. J. Pharmacol. 109:693. phages. J. Immunol. 158:897. 68. Ciruela, F., C. Saura, E. I. Canela, J. Malloll, C. Lluis, and R. Franco. 1996.

45. Sajjadi, F. G., K. Takabayashi, A. C. Foster, R. C. Domingo, and G. S. Firestein. http://www.jimmunol.org/ ␣ Adenosine deaminase affects ligand-induced signaling by interacting with aden- 1996. Inhibition of TNF- expression by adenosine: role of A3 adenosine receptors. J. Immunol. 156:3435. osine receptors. FEBS Lett. 380:219. 46. Ritchie, P. K., B. L. Spangelo, D. K. Krzymowski, T. B. Rossiter, E. Kurth, and 69. Fredholm, B. B., M. P. Abbrachio, G. Burnstock, J. W. Daly, T. K. Harden, A. M. Judd. 1997. Adenosine increases interleukin 6 release and decreases tumor K. A. Jacobson, P. Leff, and M. Williams. 1994. Nomenclature and classification necrosis factor release from rat adrenal zona glomerulosa cells, ovarian cells, of purinoreceptors. Pharmacol. Rev. 46:143. anterior pituitary cells, and peritoneal macrophages. Cytokine 9:187. 70. Cotter, T. G. 1992. Induction of apoptosis in cells of the immune system by 47. Le Moine, O., P. Stordeur, L. Schandene´, A. Marchant, D. de Groote, cytotoxic stimuli. Semin. Immunol. 4:399. M. Goldman, and J. Devie`re. 1996. Adenosine enhances IL-10 secretion by hu- 71. Rock, C. O., J. L. Cleveland, and S. Jackowski. 1992. Macrophage growth arrest man monocytes. J. Immunol. 156:4408. by cyclic AMP defines a distinct checkpoint in the mid-G1 stage of the cell cycle 48. Celada, A., P. W. Gray, E. Rinderknecht, and R. D. Schreiber. 1984. Evidence for and overrides constitutive c-myc expression. Mol. Cell Biol. 12:2351. a ␥-interferon receptor that regulates macrophage tumoricidal activity. J. Exp. 72. Kato, J., M. Matsuoka, K. Polyak, J. Massague, and C. J. Sherr. 1994. Cyclic kip1 Med. 160:55. AMP-induced G1 phase arrest mediated by an inhibitor (p27 ) of cyclin-de- by guest on September 28, 2021 49. Cheng, M., P. Olivier, J. A. Diehl, M. Fero, M. F. Roussel, J. M. Roberts, and pendent kinase 4 activation. Cell 79:487. Cip1 Kip1 C. J. Sherr. 1999. The p21 and p27 CDK “inhibitors” are essential ac- 73. Vairo, G., S. Argyrious, A. M. Bordun, G. Whitty, and J. A. Hamilton. 1990. tivators of cyclin D-dependent kinases in murine fibroblasts. EMBO J. 18:1571. Inhibition of the signaling pathways for macrophage proliferation by cyclic AMP: 50. 49. Xiong, Y., T. Connolly, B. Futcher, and D. Beach. 1991. Human D-type lack of effect on early responses to colony stimulating factor-1. J. Biol. Chem. cyclin. Cell 65:691. 265:2692. 51. Matsushime, H., M. E. Ewen, D. K. Strom, J. Y. Kato, S. K. Hanks, M. F. Roussel, 74. Baccarini, M., S. P. Dello, D. Buscher, A. Bartocci, and E. R. Stanley. 1992. and C. J. Sherr. 1992. Identification and properties of an atypical catalytic subunit IFN-␥/lipopolysaccharide activation of macrophages is associated with protein (p34PSK-J3/cdk4) for mammalian D type G1 cyclins. Cell 71:323. kinase C-dependent down-modulation of the colony-stimulating factor-1 recep- 52. Torczynski, P., A. P. Bollon, and M. Fuke. 1983. The complete nucleotide se- tor. J. Immunol. 149:2656. quence of the rat 18S ribosomal RNA gene and comparison with the respective yeast and frog genes. Nucleic Acids Res. 11:4879. 75. Jaworowski, A., E. Christy, P. Yusoff, R. Byrne, and J. A. Hamilton. 1996. 53. Celada, A., and R. A. Maki. 1992. Transforming growth factor-␤ enhances the M- Differences in the kinetics of activation of protein kinases and ERK-1 in CSF- and CSF and GM-CSF-stimulated proliferation of macrophages. J. Immunol. 148:1102. LPS-stimulated macrophages. Biochem J. 320:1011. 54. Celada, A., F. E. Borras, C. Soler, J. Lloberas, M. Klemsz, C. van Beveren, 76. Burgering, B. M. T., and J. L. Bos. 1995. Regulation of Ras-mediated signalling: S. McKercher, and R. A. Maki. 1996. The transcription factor PU.1 is involved more than one way to skin a cat. Trends Biochem. Sci. 20:18. in macrophage proliferation. J. Exp. Med. 184:61. 77. Wilson, N. J., A. Jaworowski., A. C. Ward, and J. A. Hamilton. 1998. cAMP 55. Hogquist, K. A., M. A. Nett, E. R. Unanue, and D. D. Chaplin. 1991. Interleukin enhances CSF-1-induced ERK activity and c-fos mRNA expression via a MEK- 1 is processed and released during apoptosis. Proc. Natl. Acad. Sci. USA 88:8485. dependent and Ras-independent mechanism in macrophages. Biochem. Biophys. 56. Samuels, M. L., M. J. Weber, J. M. Bishop, and M. McMahon. 1993. Conditional Res. Commun. 244:475. transformation of cells and rapid activation of the mitogen-activated protein ki- 78. Polyak, K., M. Lee, H. Erdjument-Bromage, A. Koff, J. M. Roberts, P. Tempst, nase cascade by an estradiol-dependent human Raf-1 protein kinase. Mol. Cell and J. Massague´. 1994. Cloning of p27kip1, a cyclin dependent kinase inhibitor Biol. 13:6241. and a potential mediator of extracellular antimitogenic signals. Cell 78:59.