A2B Adenosine Receptors Induce IL-19 from BronchialEpithelial Cells, Resulting in TNF-� IncreaseHongyan Zhong, Yuzhi Wu, Luiz Belardinelli, and Dewan Zeng

Department of Drug Research and Pharmacological Sciences, CV Therapeutics, Inc., Palo Alto, California

Adenosine is a signaling nucleoside that has been proposed tocontribute to the pathogenesis of asthma and chronic obstructivepulmonary disease. Previous studies suggest that adenosine mightplay an important role in modulating levels of inflammatory media-tors in the lung. Because airway epithelium is an important cellularsource of inflammatory mediators, the objective of the presentstudy was to determine whether adenosine affects the expressionand release of inflammatory cytokines from human bronchialepithelial cells (HBECs). Among the four subtypes of adenosinereceptors, the A2B receptor was expressed at the highest level.5�-(N-ethylcarboxamido)-adenosine (NECA), a stable analog ofadenosine, increased the release of IL-19 by 4.6- � 1.1-fold. A selec-tive antagonist of the A2B receptor, CVT-6694, attenuated this effectof NECA. The amount of IL-19 released from HBEC was sufficientto activate a human monocytic cell line (THP-1) and increase therelease of TNF-�. Furthermore, TNF-� was found to upregulate A2B

receptor expression in HBECs by 3.1- � 0.3-fold. Hence, these dataindicate that NECA increases the release of IL-19 from HBECs viaactivation of A2B receptors, and IL-19 in turn activates human mono-cytes to release TNF-�, which upregulates A2B receptor expressionin HBECs. The results of this study suggest that there is a novelpathway whereby adenosine can initiate and amplify an inflamma-tory response which might be important in pathogenesis of inflam-matory lung diseases.

Keywords: adenosine; TNF-�; bronchial epithelial cells; IL-19

Airway epithelium is known to play an important role in airwaydefense mechanism via a mucociliary system and as a mechanicalbarrier. Recent studies further indicate that the epithelium isnot only a barrier but can also actively generate a range ofmolecules, including lipid mediators, growth factors, and a vari-ety of cytokines/chemokines, that are important in the inflam-matory and remodeling responses that occur in the lungs (1).In asthma, the epithelium seems more sensitive and respondsabnormally to various stimuli (2). In addition, airway epithelialcells are able to interact with immune and inflammatory cellsvia direct adhesion as well as by releasing mediators includingcytokines (1). Thus, the epithelium is actively involved as regula-tor of airway inflammatory responses important in the pathogen-esis of airway disorders.

Adenosine is a nucleoside that can elicit many physiologicand pathophysiologic responses by activating one or more of itsfour subtypes of G protein–coupled receptors (A1, A2A, A2B, andA3) on target cells. Adenosine has been proposed to contribute

(Received in original form December 21, 2005 and in final form June 1, 2006 )

An abstract (less than 250 words) related to this study has been publishedfor the American Thoracic Society International Conference in 2005: Zhong H,Wu Y, Belardinelli L, Zeng D. Adenosine indirectly activates monocytes by releasingIL-19 from human bronchial epithelial cells. Proc Am Thorac Soc 2005;2:A110.

Correspondence and requests for reprints should be addressed to Hongyan Zhong,Ph.D., CV Therapeutics, Inc., 3172 Porter Drive, Palo Alto, CA 94304. E-mail:hongyan.zhong@cvt.com

Am J Respir Cell Mol Biol Vol 35. pp 587–592, 2006Originally Published in Press as DOI: 10.1165/rcmb.2005-0476OC on June 15, 2006Internet address: www.atsjournals.org

to the pathogenesis of asthma and chronic obstructive pulmonarydisease (COPD) (3). This hypothesis is based on the findingsthat the interstitial concentration of adenosine is elevated in thelungs of individuals with asthma (4) and inhaled adenosinecauses bronchoconstriction in patients with asthma (5). Thebronchoconstrictive effect of adenosine has been attributed tothe activation of lung mast cells (6, 7). In addition to the mastcells (8, 9), adenosine has been found to modulate the functionsof other inflammatory cells such as lymphocytes (10), eosinophils(11, 12), neutrophils (13), and macrophages (14), and lung cellssuch as bronchial smooth muscle cells (15, 16) and fibroblasts (17).The physiologic effects of adenosine on lung epithelial cells havealso been investigated. Adenosine is able to modulate the activityof ion channels (18–20), and induce fibronectin (21) and mucingene expression (22). However, it is unknown whether and howadenosine affects the release of inflammatory cytokines from epi-thelial cells and epithelial cell–inflammatory cell communication.

In this study, we used the primary cultured human bronchialepithelial cells (HBECs) as the model system. The objectives ofthis study were to determine (1) whether adenosine affects therelease of inflammatory cytokines from bronchial epithelial cells,(2) what the potential effects of these cytokines on lung inflam-mation might be, and (3) which adenosine receptor subtype isresponsible for the effect of adenosine.

MATERIALS AND METHODS

Materials

A selective antagonist to the A2B receptor (CVT-6694) was synthesizedby the Department of Bio-Organic Chemistry at CV Therapeutics Inc.(Palo Alto, CA), and was described in our previous publication (15). Allother reagents, such as rolipram, forskolin, 5�-(N-ethylcarboxamido)-adenosine (NECA), and adenosine deaminase (ADA), were purchasedfrom Sigma (St. Louis, MO) unless otherwise stated.

Cell Culture

Primary cultured normal HBECs were obtained from Clonetics (SanDiego, CA) and cultured using bronchial epithelial cell growth medium(Clonetics). HBECs were routinely grown in a humidified incubator with5% CO2 at 37�C. Cells from three different donors and from passages 2–3were used in the following studies. THP-1 cells were purchased fromATCC (Manassas, VA) and cultured according to ATCC’s instructions.

Stimulation of HBECs

HBECs were seeded into 12-well tissue culture plates at a density of1 � 105 cells/well and allowed to adhere overnight and reach � 90%confluence. Cells were washed twice in HEPES-buffered saline, andcultured in bronchial epithelial cell basal medium (Clonetics) containingvarious agonists or antagonists of AdoRs for 1 or 24 h.

RNA Extraction and Real-Time RT-PCR

Total RNA was extracted from HBECs using the Stratagene AbsolutelyRNA RT-PCR Miniprep Kit (Stratagene Corp., La Jolla, CA) followedby DNase treatment to eliminate potential genomic DNA contamina-tion. Real-time RT-RCR for adenosine receptors was performed aspreviously described (15). Specific primers for IL-19 (forward: 5�-AAACAATCTCCCCAAGGTGGAT-3�; reverse: 5�-AGGAAATGCTGTCAAGGTTTGC-3�), GRO-� (forward: 5�-TTTCTTCGTGATGACA

588 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 35 2006

TATCACATGT-3�; reverse: 5�-TCCTCAGCCTCTATCACAGTGG-3�), and GRO-� (forward: 5�-TGCTTGTAGGGCATAATGCCT-3�;reverse: 5�-GGGAAAGAGAAACGCTGCAG-3�) were designed usingPrimer Express 2.0 (Applied Biosystems, Foster City, CA) followingthe recommended guidelines based on sequences from GenBank. Atthe end of the PCR cycle, a dissociation curve was generated to ensurethe amplification of a single product, and the threshold cycle times (Ctvalues) for each gene were determined. Relative mRNA levels werecalculated based on the Ct values, normalized to �-actin in the samesample, and presented as percentages of �-actin mRNA.

Measurement of cAMP Accumulation

Cells were harvested using 0.0025% trypsin and 2 mM EDTA in PBS,washed and resuspended in phenol-free DMEM to a concentration of5 � 105 cells/ml, and then incubated with 1 U/ml of ADA for 30 minat room temperature. Cells were then treated with AdoR agonists,antagonists, and forskolin in the presence of 50 �M of the phosphodies-terase IV inhibitor, rolipram. After incubating for 15 min at 37�C, cellswere lysed and cAMP concentrations were determined using cAMP-Screen Direct System (Applied Biosystems) according to the manufac-turer’s instructions.

cDNA Array Analysis

Expression of inflammatory cytokines was determined using a cDNAarray kit (Cat. No. HS-033; SuperArray, Frederick, MD). The assaywas performed according to manufacturer’s instructions. BiotinylatedcDNA probes were generated from 1–2 �g of total RNA using GEArrayRT-labeling enzyme kit (SuperArray). The labeled cDNA probes werethen hybridized with gene-specific cDNA fragments spotted on nylonmembrane. The signals were detected with the chemiluminescent detec-tion reagents. The relative expression level of each gene was analyzedusing GEArray analysis software (SuperArray).

Measurement of IL-19 and TNF-�

ELISA for IL-19 was developed according to manufacturer’s protocol(R&D Systems, Minneapolis, MN). Plates (96-well) were pre-coatedwith 50 ng/well anti-human IL-19 Ab (R&D Systems). IL-19 was de-tected using 100 ng/ml of biotinylated polyclonal IL-19 Ab (R&D sys-tems), streptavidin-HRP, and TMB (tetramethylbenzidine) chromogen(Biosource, Camarillo, CA). Serial dilutions of recombinant human IL-19 (rhIL-19; PeproTech, Rocky Hill, NJ) was used as standard. Thedetection limit of rhIL-19 was 0.2 ng/ml. The concentrations of TNF-�in the cell medium were determined using ELISA kits obtained fromBiosource according to the manufacturer’s instructions. The minimaldetection level of TNF-� with these kits was 1.7 pg/ml.

Statistical Analysis

Data are presented as mean SEM of at least three separate experi-ments. The statistical analysis was performed using ANOVA followedby Bonferroni test. A P value of 0.05 was considered significant.

RESULTS

Expression of AdoR Subtypes in HBECs

Real-time RT-PCR was performed to quantify the levels oftranscripts for AdoRs. Among the four subtypes, the A2B recep-tor had the highest transcript level (0.49 0.04% of �-actinexpression) (Figure 1). Lower levels of A1 and A2A receptortranscripts were also detected (0.0012 0.0002% and 0.0058 0.0003% of �-actin, respectively), whereas the transcript for A3

receptors was below the detection level. Hence, the rank orderof AdoR mRNA levels was A2B � � A2A � A1 � � A3.

In many cell types, activation of A2A or A2B receptors leadsto increases in cellular cAMP accumulation, whereas activationof A1 or A3 receptors decreases cellular cAMP accumulationcaused by the adenylate cyclase activator, forskolin. To identifythe AdoR subtype(s) that are functionally expressed in HBECs,the effects of a nonselective agonist NECA and several otherselective agonists on cellular cAMP accumulation were deter-

Figure 1. The mRNA levels of AdoR subtypes in HBECs. Total RNA iso-lated from HBECs was subjected to real-time RT-PCR analysis. The rela-tive levels of the AdoR transcripts are presented as percentages of the�-actin transcript. Data shown are averages SEM (n � 6). nd denotesnot detected.

mined. NECA is a stable analog of adenosine, and it activatesall four AdoR subtypes including A2B receptors. As shown inFigure 2A, NECA increased cellular cAMP accumulation in aconcentration-dependent manner, with potency (EC50 value) of8.8 1.3 �M. In contrast, the A2A-selective agonist CGS-21680( 10 �M) did not cause a significant increase in cellular cAMPconcentration. In addition, the A1-selective agonist, CPA (1 �M),and the A3-selective agonist, IB-MECA (1 �M), failed to inhibitthe cellular cAMP accumulation caused by forskolin (10 �M,Figure 2B). Because there is no selective agonist for A2B re-ceptors, the effect of a selective antagonist to A2B receptors,CVT-6694, on NECA-induced cellular cAMP accumulation was

Figure 2. Effects of AdoR agonists and antagonist on cellular cAMPaccumulation in HBECs. (A ) Concentration–response curves of CGS-21680 (CGS, circles) and NECA in the absence (squares) or presence(triangles) of the A2B receptor antagonist CVT-6694 (1 �M). (B ) Lack ofeffect of CPA (1 �M) and IB-MECA (IM, 1 �M) on forskolin (Fsk,10 �M)-induced cellular cAMP accumulation. Data shown are averages

SEM (n � 3). *P 0.05, compared with control; #P 0.05, comparedwith NECA-treated cells in A.

Zhong, Wu, Belardinelli, et al.: Adenosine Induces IL-19 and Activates Monocytes 589

determined. CVT-6694 has a high affinity for the A2B receptor(Ki � 7 nM) and very low affinity for three other AdoR subtypes(Ki values are � 5 �M for A1, A2A, and A3 receptors) (15, 17). Asshown in Figure 2A, CVT-6694 (1 �M) significantly attenuatedNECA-induced cellular cAMP accumulation. Thus, using cellu-lar cAMP concentration as readout for the functional expressionof AdoRs, the results indicate that A2B receptors are functionallyexpressed in HBECs, whereas A1, A2A, or A3 receptors are not.

Effects of NECA on Expression of Inflammatory Cytokines

The effect of NECA on the gene expression of the inflammatorycytokines was determined using a cDNA array containing 85inflammatory genes involved in asthma. Among the 85 genes onthe microarray, GRO-�, GRO-�, and IL-19 genes were increasedabove 2-fold by NECA (data not shown).

To confirm and quantify NECA-induced expression ofGRO-�, GRO-�, and IL-19, gene-specific real-time RT-PCRwas performed on HBECs treated with NECA (10 �M) for1 h. As shown in Figure 3, NECA increased mRNA expressionof IL-19, GRO-�, and GRO-� up to 4.3- 0.9-, 5.6- 0.6-, and4.3- 0.6-fold above control levels, respectively.

Activation of the A2B Receptor Increased the Release ofIL-19 from HBECs

Recent publications suggesting a potential role of IL-19 in lunginflammation prompted us to determine the effect of NECA onIL-19 release. IL-19 concentrations in the culture media fromcells treated with NECA were measured using ELISA. Thelevels of IL-19 in the media from vehicle- and NECA-treatedcells for 24 h were 19.2 6.3 ng/ml and 89.2 20.2 ng/ml,respectively (Figure 4). Hence, NECA (10 �M) caused 4.6- 1.1-fold increase of IL-19 release compared with vehicle-treatedcells. To determine the role of A2B receptors in NECA-inducedIL-19 production, cells were incubated with CVT-6694 (1 �M)together with NECA. The A2B receptor antagonist CVT-6694(1 �M) reduced the NECA-increased IL-19 release by 88.9 0.5%. These results confirmed that NECA-induced IL-19 releaseis mediated by the A2B receptor subtype.

Effects of IL-19 Released from HBEC on Activation of aHuman Monocytic Cell Line (THP-1)

To determine whether the amount of IL-19 released from HBECcan activate inflammatory cells, the effect of IL-19 alone or theconditional medium from NECA-treated HBEC on the releaseof TNF-� from a monocytic cell line (THP-1) was determined.The basal level of TNF-� in the media from vehicle-treated(24 h) THP-1 cells was 7.5 0.5 pg/ml. IL-19 (100 ng/ml) increasedthe concentrations of TNF-� in the media by 4.3- 0.4-fold(Figure 5A). In contrast, GRO-� (100 ng/ml), GRO-� (100 ng/ml),

Figure 3. Effects of NECA onthe mRNA levels of GRO-�,GRO-�, and IL-19determinedusing real-time RT-PCR.HBECs were incubated withNECA (10 �M) for 1 h. Cellsincubated with vehicle wereused as control. The expres-sion levels of the cytokines

were normalized to that of �-actin and are presented as the ratios ofthe expression level in NECA-treated cells versus that in vehicle-treatedcells. The expression levels of GRO-�, GRO-�, and IL-19 in control cellsare 0.020 0.003%, 0.008 0.002%, and 0.168 0.025% of�-actin, respectively. Data shown are averages SEM (n � 3).

Figure 4. Effect of NECAon the release of IL-19 byHBECs. Cells were treatedwith vehicle, NECA in theabsence or presence ofCVT-6694 for 24 h. Mediafrom treated cells werecollected, and the con-centrations of IL-19 weredetermined using ELISA.CVT-6994 alone did not

change the release of IL-19 by HBECs (data not shown). Data shownare the averages SEM (n � 3). *P 0.05, compared with control;#P 0.05, compared with NECA-treated cells.

or NECA (10 �M) alone had no significant effect on TNF-�release from THP-1 cells (Figure 5A). These data demonstratethat IL-19, but not NECA or GRO-�/�, was able to activateTHP-1 cells. Furthermore, the conditional media collected fromHBECs treated with vehicle (control-M) or NECA (NECA-M)for 24 h were used to incubate THP-1 cells for an additional24 h. NECA-M caused a 4.6- 1.0-fold increase in TNF-� releasefrom THP-1 cells compared with control-M. This effect wascompletely abolished by an IL-19–neutralizing Ab (Figure 5B).These results suggest that IL-19 released from NECA-stimulatedHBECs is able to and sufficient to activate THP-1 cells.

Effects of TNF-� on Expression of AdoRs in HBECs

To further explore the interaction between epithelial cells andmonocytes, the effect of TNF-� on AdoR expression in HBECswas determined using real-time RT-PCR. TNF-� caused a sig-nificant time-dependent increase in mRNA expression of A2B

Figure 5. Effects of cytokines and conditional medium from HBECs onTNF-� release from THP-1 cells. (A ) THP-1 cells were incubated withvehicle (control), GRO-� (100 ng/ml), GRO-� (100 ng/ml), IL-19(100 ng/ml), or NECA (10 �M) for 24 h. (B ) Conditional medium col-lected from HBECs that were treated with vehicle (control-M) or NECA(NECA-M) for 24 h was used to incubate THP-1 cells in the absence orpresence of polyclonal IL-19 Ab (1 �g/ml) for an additional 24 h. Theconcentration of TNF-� in media was measured using ELISA. Data shownare the averages SEM (n � 4). *P 0.05, compared with control(A ) or control-M (B ); #P 0.05, compared with NECA-M in B.

590 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 35 2006

receptors (Figure 6). After 24 h incubation with TNF-�, mRNAlevel of A2B receptors was increased to 3.1 0.3 fold abovebasal levels (Figure 6). Incubation of TNF-� for 5 and 24 halso significantly upregulated the mRNA levels of A1 and A2A

receptors up to 2.0- 0.3- and 1.6- 0.1-fold above basal levels,although the expression of both subtypes of receptors was stillvery low (below 0.01% of �-actin mRNA level). The mRNAexpression of the A3 receptors remained below the detectionlevel after TNF-� treatment (Figure 6).

DISCUSSION

The main objective of this study was to determine the effect ofadenosine on the inflammatory responses mediated by epithelialcells. The main findings of our study are: (1) A2B receptorsare the predominant subtype of AdoRs expressed in primaryHBECs; (2) activation of A2B receptors on HBECs increasesthe release of IL-19, which is able to and sufficient to induceTNF-� production from human monocytes; and (3) TNF-� canupregulate the expression of A2B receptors in HBECs. Thesefindings support the hypothesis that adenosine can initiate andamplify inflammatory responses which might be important inpathogenesis of inflammatory lung diseases.

Among the four subtypes of AdoRs, the A2B receptor hasthe lowest affinity for the natural ligand, adenosine, requiringat least micromolar concentrations of adenosine for its activation(23). Hence, under normal physiologic conditions, adenosinelevels might be too low to activate the A2B receptor. However,higher levels of adenosine have been detected during hypoxia,ischemia, and inflammation (24–26). In addition, results of recentstudies (16, 17, 26, 27) suggest the A2B receptor expression canmarkedly increase during inflammation and hypoxia. The molec-ular mechanism of how A2B receptor expression is modulatedduring inflammation is largely unknown. The results of the cur-rent study demonstrate that TNF-�, a well-known inflammatorymarker and mediator, upregulates the expression of A1, A2A,and A2B receptors in HBECs. This is consistent with an earlierreport suggesting that TNF-� regulates the expression of A2A

and A2B receptors in microvascular endothelial cells (28).It should be noted that the expression of A2B receptors are

upregulated in the lung of ragweed-challenged mice (27) andIL-13–transgenic mice (26). More strikingly, increased expres-sion of A2B receptors in the lung of IL-13–transgenic mice hasbeen localized in bronchial epithelium. On the other hand, re-sults of a recent study suggested that the density of A2B receptor,determined using saturation binding assays with an antagonistradioligand, was significantly decreased in the lungs of patients

Figure 6. Effect of TNF-�on the mRNA levels ofAdoR subtypes in HBECs.HBECs were incubatedwith TNF-� (5 ng/ml) for0, 1, 5, and 24 h. Therelative levels of theAdoR transcripts are pre-sented as percentages ofthe �-actin transcript.Data shown are aver-ages SEM (n � 3). nddenotes not detected.*P 0.05, comparedwith 0 h.

with COPD compared with the control group, whereas the ra-dioligand affinity for A2B receptors was not altered (29). How-ever, it is not entirely clear which cells contribute to the decreaseof A2B receptors in COPD, how the expression of A2B receptoris regulated in allergic diseases versus COPD, and the preciserole of A2B receptor in these diseases. Interestingly, in the samestudy, the expression level of the A2B receptor was found to behighest in mast cells and macrophages. Regardless, the authorsconcluded that the A2B receptor might play an important rolein the pathogenesis of COPD. Therefore, future studies to deter-mine the effect of A2B antagonist in COPD would be of interest.

Functional expression of the A2B receptor in airway epithelialcells has been reported. This receptor has been shown to mediateadenosine-modulated ion transport (18, 19); however, the func-tion of A2B receptor in mediating cytokine release is not known.The results of our study showed that activation of A2B receptorsupregulates the expression of proinflammatory cytokines suchas IL-19, GRO-�, and GRO-�. GRO-� (CXCL2) and GRO-�(CXCL3) are CXC chemokines. The functions of these twochemokines are similar to IL-8 (CXCL8), which belongs to thesame chemokine family. GRO-� and GRO-� chemoattract andactivate neutrophils (30), and promote angiogenesis by activatingthe CXCR2 receptor on endothelial cells (31). Because neutro-phils have been implicated in the pathogenesis of inflammatorylung diseases including asthma and COPD, and because CXCchemokine–mediated angiogenesis is associated with idiopathicpulmonary fibrosis (32), CXCR2 is a potential therapeutic targetfor these diseases. A more important finding of the present studyis that NECA increases the release of IL-19 from HBECs.IL-19 belongs to the IL-10 family, which includes IL-10, IL-19,IL-20, IL-22, IL-24, and IL-26 (33). In contrast to the pleiotropicrole of IL-10, IL-19 has been shown to have mainly proinflam-matory roles. IL-19 stimulates T cells to produce IL-4 andIL-13 (34) and alters the balance of Th1/Th2 in favor of Th2.IL-19 also induces production of IL-6, TNF-�, and reactive oxy-gen species from mouse monocytes (35). IL-19 level is elevatedin the serum of patients with asthma compared with healthycontrol subjects, and its transcript is also increased in the lungof an allergen-challenged mouse model (34). Because adenosineis elevated in the asthmatic lung to hundred-micromolar range(4), and at this level it can increase the release of IL-19 fromHBECs (data not shown), it is plausible to postulate that elevatedadenosine in the asthmatic lung may contribute to the elevationof IL-19. Interestingly, recent studies have shown that elevatedadenosine play a key role in the upregulation of IL-13 and severalother cytokines in mouse models of pulmonary diseases (26).

Consistent with earlier reports on the proinflammatory rolesof IL-19, the results of our study demonstrate that IL-19 releasedby bronchial epithelial cells can activate human monocytes. Inthe present study, conditional medium of NECA-stimulatedHBECs was able to activate THP-1 cells. This effect is not dueto the activation of AdoRs on THP-1 cells, because this effect wascompletely abolished by the IL-19–neutralizing Ab and NECAalone did not stimulate TNF-� production from THP-1 cells.Furthermore, the results in this study suggest that elevatedTNF-� could have a positive feedback on the expression of theadenosine receptors on the epithelial cells. Hence, adenosinemay play an important role in the interaction between epithelialcells and inflammatory cells in the airway.

Activation of the A2B receptor increases cAMP, which isgenerally believed to elicit anti-inflammatory responses. How-ever, proinflammatory roles of this receptor have been reported(36). For example, in human mast cells (HMC-1), activation ofA2B receptors increase cAMP (37) and also increase the releaseof IL-4, IL-8, and IL-13 (8). A possible explanation is that A2Breceptor can couple to other cell signaling pathways, such as

Zhong, Wu, Belardinelli, et al.: Adenosine Induces IL-19 and Activates Monocytes 591

Figure 7. Schematic representation of an inflammatory response initi-ated and amplified by adenosine involving HBECs and monocytes. Aden-osine, via activation of the A2B receptors, increases the release of IL-19from HBECs. IL-19, in turn, stimulates human monocytes to releaseTNF-�. TNF-� upregulates A2B receptor expression in HBECs.

Ca/PLC, and MAP kinases (38), and these pathways are likelyto mediate the proinflammatory effect of adenosine or adenosineanalog via activation of A2B receptors. However, it remains tobe established which signaling pathway is responsible for theA2B-mediated upregulation of IL-19. It should be noted that anti-inflammatory roles of the A2B receptor have also been reportedin other in vitro studies (39–41). Therefore, further studies usinganimal models or patients will be helpful to determine the roleof A2B receptors in the lung.

In summary, adenosine, IL-19, and TNF-� are important pro-inflammatory mediators in lung inflammation. These mediatorsmay interact with one another to cause the alleviation or exacer-bation of the disease. The results of the present study demon-strated that the A2B receptor subtype is the predominant AdoRexpressed in HBECs. Activation of this AdoR subtype increasedthe release of IL-19. IL-19 released from these cells, in turn,stimulated the human monocytes to release TNF-�. TNF-� isable to upregulate the expression of A2B receptors in HBECs(Figure 7). Thus, our findings provide a novel mechanism thatcould explain the interactions among adenosine, IL-19, andTNF-� in the initiation, maintenance, and amplification of in-flammatory responses. In addition, our findings suggest that theA2B receptor might be a novel therapeutic target for inflamma-tory lung diseases.

Conflict of Interest Statement : H.Z., Y.W., L.B., and D.Z. are employees of CVTherapeutics, Inc. (CVT) and own stock and stock options in this company. CVThas patented numerous A2B antagonists, and one such antagonist is currentlybeing developed for the treatment of asthma.

References1. Takizawa H. Airway epithelial cells as regulators of airway inflammation

[review]. Int J Mol Med 1998;1:367–378.

2. Holgate ST. Epithelial damage and response. Clin Exp Allergy 2000;30:37–41.

3. Holgate ST, Church MK, Polosa R. Adenosine: a positive modulator ofairway inflammation in asthma. Ann N Y Acad Sci 1991;629:227–236.

4. Driver AG, Kukoly CA, Ali S, Mustafa SJ. Adenosine in bronchoalveolarlavage fluid in asthma. Am Rev Respir Dis 1993;148:91–97.

5. Cushley MJ, Tattersfield AE, Holgate ST. Adenosine-induced broncho-constriction in asthma: antagonism by inhaled theophylline. Am RevRespir Dis 1984;129:380–384.

6. Cushley MJ, Holgate ST. Adenosine-induced bronchoconstriction inasthma: role of mast cell-mediator release. J Allergy Clin Immunol1985;75:272–278.

7. Polosa R, Ng WH, Crimi N, Vancheri C, Holgate ST, Church MK,Mistretta A. Release of mast-cell-derived mediators after endobron-chial adenosine challenge in asthma. Am J Respir Crit Care Med1995;151:624–629.

8. Ryzhov S, Goldstein AE, Matafonov A, Zeng D, Biaggioni I, FeoktistovI. Adenosine-activated mast cells induce IgE synthesis by B lympho-cytes: an A2B-mediated process involving Th2 cytokines IL-4 andIL-13 with implications for asthma. J Immunol 2004;172:7726–7733.

9. Zhong H, Shlykov SG, Molina JG, Sanborn BM, Jacobson MA,Tilley SL, Blackburn MR. Activation of murine lung mast cells by theadenosine A3 receptor. J Immunol 2003;171:338–345.

10. Huang S, Apasov S, Koshiba M, Sitkovsky M. Role of A2a extracellularadenosine receptor-mediated signaling in adenosine-mediated inhibi-tion of T-cell activation and expansion. Blood 1997;90:1600–1610.

11. Walker BA, Jacobson MA, Knight DA, Salvatore CA, Weir T, Zhou D,Bai TR. Adenosine A3 receptor expression and function in eosino-phils. Am J Respir Cell Mol Biol 1997;16:531–537.

12. Young HW, Molina JG, Dimina D, Zhong H, Jacobson M, Chan LN,Chan TS, Lee JJ, Blackburn MR. A3 adenosine receptor signalingcontributes to airway inflammation and mucus production in adenosinedeaminase-deficient mice. J Immunol 2004;173:1380–1389.

13. Cronstein BN. Adenosine regulation of neutrophil function and inhibi-tion of inflammation via adenosine receptors. In: Jacobson KA, JarvisMF, editors. Purinergic approaches in experimental therapeutics.New York: Wiley-Liss; 1997. pp. 285–299.

14. Hasko G, Szabo C, Nemeth ZH, Kvetan V, Pastores SM, Vizi ES. Adeno-sine receptor agonists differentially regulate IL-10, TNF-alpha, andnitric oxide production in RAW 264.7 macrophages and in endotoxe-mic mice. J Immunol 1996;157:4634–4640.

15. Zhong H, Belardinelli L, Maa T, Feoktistov I, Biaggioni I, Zeng D. A2Badenosine receptors increase cytokine release by bronchial smoothmuscle cells. Am J Respir Cell Mol Biol 2004;30:118–125.

16. Feoktistov I, Ryzhov S, Zhong H, Goldstein AE, Matafonov A, Zeng D,Biaggioni I. Hypoxia modulates adenosine receptors in human endo-thelial and smooth muscle cells toward an A2B angiogenic phenotype.Hypertension 2004;44:649–654.

17. Zhong H, Belardinelli L, Maa T, Zeng D. Synergy between A2B adeno-sine receptors and hypoxia in activating human lung fibroblasts. AmJ Respir Cell Mol Biol 2005;32:2–8.

18. Huang P, Lazarowski ER, Tarran R, Milgram SL, Boucher RC, Stutts MJ.From the Cover: Compartmentalized autocrine signaling to cystic fi-brosis transmembrane conductance regulator at the apical membraneof airway epithelial cells. Proc Natl Acad Sci USA 2001;98:14120–14125.

19. Kornerup KN, Page CP, Moffatt JD. Pharmacological characterisationof the adenosine receptor mediating increased ion transport in themouse isolated trachea and the effect of allergen challenge. Br JPharmacol 2005;144:1011–1016.

20. Szkotak AJ, Ng AM, Sawicka J, Baldwin SA, Man SF, Cass CE, Young JD,Duszyk M. Regulation of K(�) current in human airway epithelialcells by exogenous and autocrine adenosine. Am J Physiol Cell Physiol2001;281:C1991–C2002.

21. Roman J, Rivera HN, Roser-Page S, Sitaraman SV, Ritzenthaler JD.Adenosine induces fibronectin expression in lung epithelial cells: impli-cations for airway remodeling. Am J Physiol Lung Cell Mol Physiol2006;290:L317–L325.

22. McNamara N, Gallup M, Khong A, Sucher A, Maltseva I, Fahy JV,Basbaum C. Adenosine up-regulation of the mucin gene, MU2, inasthma. FASEB J 2004;18:1770–1772.

23. Fredholm BB, Irenius E, Kull B, Schulte G. Comparison of the potencyof adenosine as an agonist at human adenosine receptors expressedin Chinese hamster ovary cells. Biochem Pharmacol 2001;61:443–448.

24. Mentzer RM Jr, Rubio R, Berne RM. Release of adenosine by hypoxiccanine lung tissue and its possible role in pulmonary circulation. AmJ Physiol 1975;229:1625–1631.

592 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 35 2006

25. Hagberg H, Andersson P, Lacarewicz J, Jacobson I, Butcher S, SandbergM. Extracellular adenosine, inosine, hypoxanthine, and xanthine inrelation to tissue nucleotides and purines in rat striatum duringtransient ischemia. J Neurochem 1987;49:227–231.

26. Blackburn MR, Lee CG, Young HW, Zhu Z, Chunn JL, Kang MJ,Banerjee SK, Elias JA. Adenosine mediates IL-13-induced inflamma-tion and remodeling in the lung and interacts in an IL-13-adenosineamplification pathway. J Clin Invest 2003;112:332–344.

27. Fan M, Qin W, Mustafa SJ. Characterization of adenosine receptor(s)involved in adenosine-induced bronchoconstriction in an allergicmouse model. Am J Physiol Lung Cell Mol Physiol 2003;284:L1012–L1019.

28. Khoa ND, Montesinos MC, Reiss AB, Delano D, Awadallah N, CronsteinBN. Inflammatory cytokines regulate function and expression of aden-osine A(2A) receptors in human monocytic THP-1 cells. J Immunol2001;167:4026–4032.

29. Varani K, Caramori G, Vincenzi F, Adcock I, Casolari P, Leung E,MacLennan S, Gessi S, Morello S, Barnes PJ, et al. Alteration ofAdenosine Receptors in Patients with Chronic Obstructive PulmonaryDisease. Am J Respir Crit Care Med 2006;173:398–406.

30. Geiser T, Dewald B, Ehrengruber M, Clark-Lewis I, Baggiolini M. Theinterleukin-8-related chemotactic cytokines GRO alpha, GRO beta,and GRO gamma activate human neutrophil and basophil leukocytes.J Biol Chem 1993;268:15419–15424.

31. Romagnani P, Lasagni L, Annunziato F, Serio M, Romagnani S. CXCchemokines: the regulatory link between inflammation and angiogen-esis. Trends Immunol 2004;25:201–209.

32. Strieter RM, Belperio JA, Keane MP. CXC chemokines in angiogenesisrelated to pulmonary fibrosis. Chest 2002;122:298S–301S.

33. Kempuraj D, Frydas S, Kandere K, Madhappan B, Letourneau R, Christ-

odoulou S, Boucher W, Riccioni G, Conti P, Theoharides TC. Interleu-kin-19 (IL-19) network revisited. Int J Immunopathol Pharmacol 2003;16:95–97.

34. Liao SC, Cheng YC, Wang YC, Wang CW, Yang SM, Yu CK, Shieh CC,Cheng KC, Lee MF, Chiang SR, et al. IL-19 induced Th2 cytokinesand was up-regulated in asthma patients. J Immunol 2004;173:6712–6718.

35. Liao YC, Liang WG, Chen FW, Hsu JH, Yang JJ, Chang MS. IL-19induces production of IL-6 and TNF-alpha and results in cell apoptosisthrough TNF-alpha. J Immunol 2002;169:4288–4297.

36. Caruso M, Holgate ST, Polosa R. Adenosine signalling in airways. CurrOpin Pharmacol 2006;6:251–256.

37. Feoktistov I, Biaggioni I. Adenosine A2b receptors evoke interleukin-8secretion in human mast cells: an enprofylline-sensitive mechanismwith implications for asthma. J Clin Invest 1995;96:1979–1986.

38. Fredholm BB, Jzerman API, Jacobson KA, Klotz K-N, Linden J. Interna-tional Union of Pharmacology. XXV. Nomenclature and classificationof adenosine receptors. Pharmacol Rev 2001;53:527–552.

39. Kreckler LM, Wan TC, Ge Z-D, Auchampach JA. Adenosine inhibitstumor necrosis factor-� release from mouse peritoneal macrophagesvia A2A and A2B but not the A3 adenosine receptor. J PharmacolExp Ther 2006;317:172–180.

40. Pelletier S, Dube J, Villeneuve A, Gobeil F Jr, Bernier SG, Battistini B,Guillemette G, Sirois P. Adenosine induces cyclic-AMP formationand inhibits endothelin-1 production/secretion in guinea-pig trachealepithelial cells through A2B adenosine receptors. Br J Pharmacol 2000;129:243–250.

41. Xaus J, Valledor AF, Cardo M, Marques L, Beleta J, Palacios JM,Celada A. Adenosine inhibits macrophage colony-stimulating factor-dependent proliferation of macrophages through the induction ofp27kip-1 expression. J Immunol 1999;163:4140–4149.