Pharmacological Characterization of AdenosineReceptors on Isolated Human Bronchi

Luigino Calzetta1,2, Domenico Spina2, Mario Cazzola1,3, Clive P. Page2, Francesco Facciolo4,Erino A. Rendina5, and Maria G. Matera6

1Department of Internal Medicine, University of Rome “Tor Vergata,” Rome, Italy; 2The Sackler Institute of Pulmonary Pharmacology,

Institute of Pharmaceutical Science, King’s College London, London, United Kingdom; 3Pulmonary Rehabilitation Group, Istituto di Ricovero e Cura

a Carattere Scientifico, San Raffaele Pisana, Rome, Italy; 4Thoracic Surgery, Regina Elena National Cancer Institute, Rome, Italy; 5Division of ThoracicSurgery, University of Rome “La Sapienza,” Sant’Andrea Hospital, Rome, Italy; and 6Department of Experimental Medicine, Second University of

Naples, Naples, Italy

Adenosine induces airways obstruction in subjects with asthma,but the receptor subtype responsible remains unknown. Theobjectives of this study were to determine the pharmacologicalprofile of adenosine receptor subtypes mediating contraction andto investigate the mechanism in normal and passively sensitizedhuman airway tissues. Contraction of bronchial rings isolated fromresected lung tissue of patients with lung carcinomawasmeasuredin response to nonselective adenosine receptor agonists, 5-AMPand 5’-(N-Ethylcarboxamido)adenosine, and A1 receptor agonist,N6-cyclopentyladenosine, in the absence and presence of selectiveadenosine receptor antagonists. Pharmacological antagonists,chemical ablation of airway sensory nerves using capsaicin, andpassive sensitization of tissue with serum from subjects with atopyand asthmawas used to investigate themechanism of contraction.Human bronchial tissue contracted in a concentration-dependentmanner to adenosine agonists that showed a rank order of activ-ity of A1. A2B.. A2A ¼ A3. Themaximum contractile response toN6-cyclopentyladenosine (231.0 6 23.8 mg) was significantly re-duced in tissues chemically treated with capsaicin to desensitizesensory nerves (desensitized: 101.6 6 15.2 mg; P , 0.05). Passivesensitization significantly augmented the contraction induced byadenosine A1 receptor activation (sensitized: 389.76 52.8 mg ver-sus nonsensitized; P , 0.05), which was linked to the release ofleukotrienes, and not histamine (MK571: 25.561.7mg; epinastine260.0 6 22.2 mg versus control; P , 0.05). This study providesevidence for a role for adenosine A1 receptors in eliciting humanairway smooth muscle constriction, which, in part, is mediated bythe action of capsaicin sensitive sensory nerves.

Keywords: adenosine; N6-cyclopentyladenosine; human bronchi; air-

way smooth muscle contraction

Adenosine is a purine that plays a pivotal role in many biologicalprocesses, such as energy generation and protein metabolism. Itis also considered to be a proinflammatory mediator involved inthe pathogenesis of asthma and other inflammatory disorders.Thus, adenosine induces airway obstruction in allergic animal

models and in patients with inflammatory airway diseases, suchas asthma and chronic obstructive pulmonary diseases, but not inhealthy individuals (1–3). Adenosine receptor subtypes havebeen identified on many cell types involved in airways inflam-mation (2, 4, 5). The biological response to adenosine is medi-ated by four G-coupled receptors. Adenosine A1 and A3

receptors are coupled to Gi/0, whereas adenosine A2A andA2B receptors are coupled to Gs, with the latter resulting inthe activation of phospholipase C, possibly via Gq/11 (6). Fur-thermore, adenosine A1 and A3 receptors can activate both K1

channels and Ca12 channels, and for adenosine A2B receptors,cAMP-independent intracellular pathways have also been de-scribed (7).

Adenosine may exert either deleterious or protective roles inthe lung. In mammals, adenosine can be released as a result ofhypoxia, lung injury, or chronic inflammation (8, 9). On theother hand, some anti-inflammatory effects of adenosine andprotection of tissues have been described in the lung, heart,and brain (10–12). It is also recognized that more than oneadenosine receptor type can be expressed in a single cell type,and this may sometimes result in complex or atypical responseswhen using adenosine as the agonist. For example, the receptorsubtype involved in adenosine-induced airways obstruction andinflammatory responses vary between species (2). Thus,adenosine-induced airways obstruction is mediated by adeno-sine A1 and A2B receptors in rat and mouse, by adenosine A3

receptors in rat, guinea-pig, and mouse (13–16), and by adeno-sine A1 receptors in the guinea pig and rabbit (13–17). How-ever, which receptor subtype is responsible for airwaysobstruction in patients with asthma remains unknown, althoughadenosine A1 receptors have been implicated in mediating hu-man airway smooth muscle contraction in vitro (2, 3), and anup-regulation of adenosine A1 receptors has been demonstratedon airway smooth muscle obtained from biopsies from subjectswith asthma (3, 18).

It has been suggested that adenosine may cause airway smoothmuscle contraction via the release of neuropeptides and subse-quent activation of mast cells (19, 20), although inhibition of neu-tral endopeptidase by phosphoramidon did not affect bronchialresponses to 5-AMP in subjects with asthma (21). Moreover, theneurokinin A receptor antagonist, SR48968C, did not affect bron-chospasm induced by AMP in subjects with asthma (22), althoughthis response is partly mediated by histamine and leukotrienes(23, 24) and reflex activation of parasympathetic nerves (23, 25,26). Similarly, preclinical studies have also confirmed an involve-ment of neural pathways, as bronchoconstriction to an adenosinereceptor agonist was inhibited by vagal section and pretreatmentwith capsaicin in allergic guinea pigs (27).

In the present study, we have used a range of potent and selec-tive adenosine receptor agonists and antagonists (see the online

(Received in original form February 16, 2011 and in final form June 17, 2011)

This work was supported by Pneumolabs Inc., UK grants to C.P.P. to support

Ph.D. studentship for L.C.

Author Contributions: conception and design: D.S., M.C., C.P.P., M.G.M.; aqui-

sition of data or analysis and interpretation of data: L.C., D.S., M.C., C.P.P., F.F.,

E.A.R., M.G.M.; drafting the manuscript for important intellectual content: L.C.,

D.S., M.C., C.P.P., M.G.M.

Correspondence and requests for reprints should be addressed to Domenico

Spina, Ph.D., The Sackler Institute of Pulmonary Pharmacology, Institute of Pharma-

ceutical Science, King’s College London, London SE1 9NH, UK. E-mail: domenico.

spina@kcl.ac.uk

This article has an online supplement, which is accessible from this issue’s table of

contents at www.atsjournals.org

Am J Respir Cell Mol Biol Vol 45. pp 1222–1231, 2011

Originally Published in Press as DOI: 10.1165/rcmb.2011-0056OC on june 14, 2011

Internet address: www.atsjournals.org

supplement) to identify the role of these receptors in contractionof human airway smooth muscle.

MATERIALS AND METHODS

Tissue Preparation

Ethical approval for these studies was obtained from the InstitutoNazionale tumori Regina Elena and Istituto dermatologico San Galli-cano (Rome, Italy), and all patients gave informed consent (NationalCommittee of Bioethics, National Committee of Bio-Safety, Biotech-nology, and Sciences: collection of biological samples for research pur-poses, February 16, 2009, Italy). Details of the patients and recovery oflung tissue can be found in the online supplement.

Drugs

All the drugs used in this study were obtained from Sigma-Aldrich(St. Louis, MO), and their source and method of preparation can befound in the online supplement.

Tension Measurement

Human segmental bronchi (third generation bronchi), typically 3–5 mmin diameter, were placed in organ baths containing Krebs-Henseleitbuffer solution (378C) bubbled with 95%O2/5%CO2, and suspendedunder passive tension (0.5–1.0 3 g). Airways were allowed to equili-brate for 90 minutes, with repeated changes of KH buffer solutionevery 10 minutes. Changes in isometric tension were measured witha transducer (Fort 10 transducer, World Precision Instruments, Berlin,Germany).

Passive Sensitization

Human isolated bronchial rings were rotated overnight at room temper-ature in tubes containing KH buffer solution in the absence (nonsensi-tized control rings) or presence of 10% vol21 sensitizing serum(sensitized rings). Patients suffering from atopic asthma (total IgE .250 U ml21 specific against common aeroallergens) during exacerba-tion (28, 29) provided signed consent for serum donation. Sera wereprepared by centrifugation of whole blood and sera samples were fro-zen at2808C in 200-ml aliquots until required. The next morning, afterremoval of adhering alveolar and connective tissues, bronchial ringswere transferred into an organ bath containing KH buffer (378C) andcontinuously gassed with 95% O2/5% CO2. In some experiments, thebronchial epithelium was also mechanically removed.

Study Design of Isolated Organ Bath Experiments

Logarithmic concentration-response curves (CRCs) to carbachol(0.1 mM–1 mM) were performed in all tissues as an internal refer-ence, followed by a washout period of 90 minutes until the baselinewas achieved. The effect of adenosine receptor–selective agonistsand antagonists on baseline tone was addressed using experimentaldesigns to ascertain the receptor subtype(s) relevant for contraction,and the mechanism of this response, and is detailed in the onlinesupplement.

Statistical Analysis

Concentration-response data were fitted to a four-parameter logisticequation to obtain estimates of the maximal response (Emax) andthe concentration required to cause a 50% maximal effect (EC50)and slope (30, 31). Contractile responses to the various adenosineligands were expressed as a percentage of the carbachol Emax or ten-sion (mg), and EC50 values were converted to negative logarithmicvalues (pD2) for statistical analysis (30). All values are presented asarithmetic mean (6SEM), or geometric mean (6confidence interval[CI]) for EC50, of n subjects for each treatment group. Statistical sig-nificance was assessed by Student’s t test or one-way ANOVA, withDunnett’s or Bonferroni’s post tests. All data analyses were performedusing computer software (GraphPad Prism version 5.00 for Windows,GraphPad Software, San Diego, CA). The level of statistical signifi-cance was defined as P , 0.05.

RESULTS

Baseline Characteristics of the Human Isolated

Bronchial Rings

Tissues from 35 different patients were used in the experiments,and there were no significant differences (P. 0.05) between thebaseline characteristics of the human isolated bronchial ringsregarding the wet weight (198 6 13 mg). The contractile po-tency to carbachol was 1.79 mM (95% CI, 0.50–3.07) and theEmax was 1.51 6 0.04 g. The contraction induced by acetylcho-line (100 mM) was 0.63 6 0.12 g.

The Effect of Adenosine Agonists and Antagonists on Human

Isolated Bronchial Tone

Selective and nonselective adenosine receptor agonists induceda concentration-dependent contractile response over the rangeof concentrations (1 nM–3 mM). It was possible to distinguishthree different trends, where adenosine A1 receptor activationproduced the greatest degree of contraction (15.30 6 1.13% car-bachol Emax; n ¼ 4), adenosine A2B receptor activation pro-duced a modest contractile response (8.07 6 3.48%; n ¼ 4),whereas the contractile response to adenosine A2A, and A3 re-ceptor–selective agonists and nonspecific agonists (5-AMP or 5’-[N-Ethylcarboxamido]adenosine) produced variable responsesless than 5% of the maximum response induced by carbachol(Figures 1A and 1B, Table 1).

Contraction in response to repeated adenosine A1 receptorstimulation with N6-cyclopentyladenosine was characterized bytachyphylaxis. The maximum contraction induced by CPA ob-tained 60 and 120 minutes after the first response was signifi-cantly reduced (Figures 1C and 1D, Table 1). Despite thereduction of subsequent contractions induced by CPA, thepD2 values did not significantly change (Figures 1C and 1D,Table 1). For this reason, when it was necessary to comparethe response of adenosine receptor activation under variousexperimental conditions, the CRCs were performed on differentsamples from the same patient to avoid an underestimation ofthe response to CPA due to tachyphylaxis.

We next investigated the effect of various selective adenosinereceptor antagonists on adenosine receptor agonist responses inbronchial tissues using antagonist concentrations two logarithmshigher than the pA2, pKB, or IC50 values for DPCPX (0.1 mM),for ZM241,385 (0.1 mM), for MRS1754 (1.0 mM), and forMRS1220 (0.5 mM) (32) (see the online supplement for furtherdetails). The treatment of human bronchial rings with selectiveadenosine receptor antagonists reduced the potency and themaximal effect mediated by adenosine receptor activation.DPCPX significantly reduced the contractile potency of5-AMP (Figure 2, Table 1; P , 0.05). In contrast, the adenosineA2B receptor antagonist, MRS1754, did not alter the contrac-tion induced by 5-AMP. Furthermore, selective adenosine re-ceptor antagonism significantly reduced the potency ofadenosine receptor agonists (P , 0.05), except for antagonismof the adenosine A2A receptor, where the Emax induced by itsagonist was not detectable. Therefore, in this case, the pD2 andEC50 values were not calculated, because, in the presence ofadenosine receptor antagonists, the sigmoid curves induced byselective adenosine receptor agonists were noticeably shifteddownward (Figure 2, Table 1).

NECA, a nonselective adenosine agonist, induced a weakcontraction of human bronchial rings. This may have beendue to the potential of NECA inducing opposing effects on dif-ferent adenosine receptors, as it is known to activate all fourreceptor subtypes. To investigate this further, bronchial rings

Calzetta, Spina, Cazzola, et al.: Adenosine and Human Airways 1223

were pretreated with combinations of the adenosine A1 receptorantagonist (DPCPX, 0.1 mM), the adenosine A2A receptor an-tagonist (ZM241385, 0.1 mM), the adenosine A2B receptor an-tagonist (MRS1734, 1.0 mM), and the adenosine A3 receptorantagonist (MRS1220, 0.5 mM) to selectively inhibit three outof four of the adenosine receptor subtypes, and to allow theeffect of NECA on one adenosine receptor subtype to be inves-tigated. Under these conditions, NECA produced significantchanges in bronchial smooth muscle tension of a similar profileto that observed with the corresponding adenosine receptor–selective agonist. There were no significant differences in Emaxand pD2 values (P . 0.05) among CRCs induced by selectiveadenosine receptor agonists and NECA under the different ex-perimental conditions (Table 2).

The adenosine A1 receptor appeared to be the predominantreceptor leading to the contraction of human isolated bronchialtissue, and thus the selective adenosine A1 receptor antagonist,DPCPX, was pharmacologically characterized to estimate thepA2 value. DPCPX (1–30 nM) caused parallel rightward dis-placements of the CRC induced by CPA, with a suppressionof Emax values (41.2 6 0.6%). In contrast, DPCPX (100 nM)

almost totally abolished responses to CPA (Figure 3A). TheSchild analysis of DPCPX (1–30 nM) versus CPA gave a linearplot (r2 ¼ 0.99) with a slope of 1.07 6 0.10, corresponding toa pA2 of 8.95 and to a KB of 1.50 6 0.18 nM (Figure 3B).

The Role of Neutral Endopeptidase and Capsaicin-Sensitive

Sensory Nerves on Human Isolated Bronchial Contraction

Mediated by Adenosine A1 Receptor Activation

The influence of the neutral endopeptidases and capsaicin-sensitive sensory nerves were studied as possible modulatorsof isolated bronchial contraction induced by the selective activa-tion of adenosine A1 receptors. Capsaicin (10 mM) induceda contraction of human isolated bronchi (22.56 9.0% carbacholEmax; n ¼ 4) and five consecutive capsaicin administrationsresulted in a significant tachyphylaxis (Figure 4A; P , 0.05).Moreover, the desensitization induced by capsaicin significantlyreduced the contractile response induced by CPA (6.73 61.01% carbachol Emax; n ¼ 4). On the other hand, the treat-ment of tissues with thiorphan (1 mM) significantly increasedthe contractile potency to CPA (pD2; 7.156 0.25 versus control;

Figure 1. Concentration-response curves

to adenosine receptor agonists (A and B).

Concentration-response curves to CPA athourly intervals in the same human iso-

lated bronchi (C). Bar graph representa-

tion of contraction (tension, mg) of

human isolated preparations in responseto consecutive administration of CPA

(3 mM) (D). Each symbol or column

represents the mean (6SEM) from exper-iments performed with bronchial prepa-

rations from different subjects (n ¼ 4–5).

***P , 0.001 of the maximal response

(Emax) versus CPA; #P , 0.05 versusCV1808 1 ZM241385 (A and B); **P ,0.01, ***P , 0.001 versus CPA at time

0 (C and D). AR, adenosine receptor;

CPA, N6-cyclopentyladenosine; IB-MECA, 1-Deoxy-1-[6-[[(3-iodophenyl)

methyl]amino]-9H-purin-9-yl]-N-methyl-

b-D-ribofuranuronamide; NECA, 5’-(N-Ethylcarboxamido)adenosine.

1224 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 45 2011

P , 0.05), but did not modify the maximal response. Further-more, thiorphan did not alter the response to CPA in capsaicin-desensitized bronchi (Figure 4B).

The Effect of Passive Sensitization on Human Isolated

Bronchial Contraction Mediated by Adenosine

Receptor Activation

The control responses of bronchial tissues incubated in the presenceof non-asthmatic serum was similar compared with healthy bronchinot exposed to atopic serum (P . 0.05). Furthermore, passive sen-sitization did not alter the contractile response to carbachol com-pared with healthy bronchi not exposed to atopic serum (P. 0.05).

In contrast, passive sensitization significantly increased thecontraction in response to non-selective and adenosine A1

receptor–selective agonists (Table 2; P , 0.01). The contractileresponse to the adenosine A1 receptor agonist, CPA, was 25.863.5% carbachol Emax (n ¼ 4), and the contractile potency wassignificantly increased (pD2 7.01 6 0.13; P , 0.05) in sensitizedtissues when compared with nonsensitized bronchi. In contrast,passive sensitization did not significantly alter the contractileresponse to adenosine A2A, A2B, and A3 receptor activationcompared with the response of nonsensitized bronchi. The con-tractile potency of 5-AMP, and after adenosine A1 and A2B

receptor stimulation, was significantly reduced in passively sen-sitized bronchi pretreated with selective adenosine receptor

TABLE 1. PHARMACOLOGICAL CHARACTERISTICS OF ADENOSINE RECEPTOR AGONISTS AND EFFECTS OF ADENOSINE RECEPTORANTAGONIST TREATMENT

Emax EC50 Emax EC50

Agonist Selectivity (mg) (mM) Antagonist (mg) (mM)

5-AMP Nonselective 70.8 6 10.4 0.17 (0.07–0.27) 1DPCPX 66.1 6 15.1 1.22 (0.01–2.44)*

1MRS1745 72.9 6 9.4

NECA Nonselective 50.9 6 23.7 0.13 (0.01–0.26)

CPA 1st A1 231.0 6 17.1 0.28 (0.15–0.41) 1DPCPX 41.5 6 20.5† ND

CPA 2nd 124.7 6 11.6‡ 0.10 (0.03–0.17)

CPA 3rd 78.6 6 5. 9x 0.10 (0.01–0.19)

CGS21680 A2A 38.2 6 12.5 0.18 (0.08–0.28) 1ZM241385 18.6 6 9.7 ND

CV1808 1 ZM241385 A2B 121.9 6 52.6 0.07 (0.02–0.13) 1MRS1754 18.4 6 10.3* ND

IB-MECA A3 61.8 6 23.6 0.09 (0.07–0.10) 1MRS1220 11. 8 6 5.9* ND

Definition of abbreviations: 5-AMP, adenosine monophosphate; CPA, N6-cyclopentyladenosine; EC50, concentration required to cause a 50% maximal effect; Emax,

estimate of maximal response; IB-MECA, 1-Deoxy-1-[6-[[(3-iodophenyl)methyl]amino]-9H-purin-9-yl]-N-methyl-b-D-ribofuranuronamide; ND, not detectable; NECA,

5’-(N-Ethylcarboxamido)adenosine.

Values shown are the arithmetic mean 6 SEM (Emax) and geometric mean6 95% confidence interval (EC50, mM) from experiments performed with samples of n ¼ 4

different subjects (5-AMP: n ¼ 5).

* P , 0.05 versus the respective adenosine receptor activation.y P , 0.001 and versus the respective adenosine receptor activation.z P , 0.01 versus CPA at time 0.x P , 0.001 versus CPA at time 0.

Figure 2. Concentration-response curves to 5-AMP (A), CPA (B), CGS21680 (C), CV1808 (D), and IB-MECA (E) in the absence and presence of

adenosine receptor antagonists. Each symbol represents mean (6SEM) from experiments performed with bronchial preparations from differentsubjects (n ¼ 4). *P , 0.05, **P , 0.01, ***P , 0.001 versus selective or nonselective adenosine receptor activation.

Calzetta, Spina, Cazzola, et al.: Adenosine and Human Airways 1225

antagonists (P , 0.05). The mean reduction of pD2 induced bythe selective adenosine A1 and A2B receptor antagonists in pas-sively sensitized bronchi was greater than half a logarithm(0.66 6 0.11) (Table 3).

The Effect of Muscarinic Receptor Antagonism on Human

Isolated Bronchial Contraction Mediated by Adenosine A1

Receptor Activation

The maximal contractile response to CPA was significantly (P,0.001) reduced after incubation of both epithelium-intact andepithelium-denudedhumanbronchial preparationswith atropine(0.1 mM). The magnitude of the receptor antagonism caused byatropine against the contractile potency to CPA was similar inepithelium-intact (pD2; atropine: 6.50 6 0.15 versus control:6.55 6 0.11; n ¼ 4; Figure 5A) and epithelium-denuded bronchi(pD2; atropine: 6.61 6 0.13 versus control: 6.69 6 0.13; n ¼ 3;Figure 5B), although, in both cases, atropine significantly reducedthe maximum contraction to CPA (Figure 5).

In contrast, the M2-selective receptor antagonist, methoctr-amine (0.1 mM), did not significantly (P . 0.05) affect the con-tractile response to CPA in both epithelium-intact (pD2;methoctramine: 6.55 6 0.07 versus control: 6.55 6 0.11; n ¼ 4;Figure 5A) and epithelium-denuded (pD2;methoctramine 6.6360.05 versus control: 6.696 0.13; n ¼ 3; Figure 5B) bronchi.

Atropine (0.1 mM) significantly (P , 0.05) inhibited themaximal contraction of isolated, passively sensitized human

bronchi with an intact epithelium without a significant changein the contractile potency to CPA (pD2: atropine 6.97 60.25 versus control: 7.01 6 0.13; n ¼ 3; Figure 5C). Methoctr-amine did not alter the contractile potency to CPA (pD2:methoctramine 6.91 6 0.12 versus control: 7.01 6 0.13; n ¼ 3;P , 0.05) nor maximum contractile response (Figure 5C).

The Role of H1 and Cys-LT1 Receptors in the Contraction of

Human Isolated Bronchial Preparations to Adenosine

Receptor Agonists

The selective H1-receptor antagonist, epinastine, and the selec-tive Cys-LT1 receptor antagonist, MK571, both at 1 mM, signif-icantly reduced baseline tone (% carbachol Emax) in humanisolated bronchi (20.8 6 3.6%, n ¼ 4, and 56.0 6 6.5%, n ¼ 4,respectively; Figure 6A). The simultaneous incubation with epi-nastine and MK571 induced a relaxation similar to MK571alone (47.4 6 7.4%; Figure 6A).

Passive sensitization did not alter the relaxation mediated byepinastine, whereas it did reduce the relaxation mediated byMK571 (11.8 6 6.9%; P , 0.01) and by epinastine andMK571 together (12.9 6 6.5%; P , 0.05) compared with thevalues observed in nonsensitized tissues or in tissues incubatedwith nonasthmatic serum (Figure 6A).

The contractile response to the adenosine A1 receptor agonist,CPA, in nonsensitized tissues was not significantly altered byepinastine or MK571 alone, although the simultaneous treatment

TABLE 2. SELECTIVE ACTIVATION OF ADENOSINE RECEPTORS BY THE NONSELECTIVE ADENOSINERECEPTOR AGONIST, 59-(N-ETHYLCARBOXAMIDO)ADENOSINE

Emax EC50

Agonist 1Antagonist Selectivity (mg) (mM)

NECA (nonselective) 1ZM241385 A1 189.7 6 66.2 0.11 (0.05–0.18)

1MRS1754

1MRS1220

NECA (nonselective) 1DPCPX A2A 56. 8 6 9.7 0.05 (0.01–0.09)

1MRS1754

1MRS1220

NECA (nonselective) 1DPCPX A2B 137.6 6 52.1 0.14 (0.01–0.28)

1ZM241385

1MRS1220

NECA (nonselective) 1DPCPX A3 53.0 6 40.9 0.19 (0.11–0.27)

1ZM241385

1MRS1754

Definition of abbreviations: EC50, concentration required to cause a 50% maximal effect; Emax, maximal effect; NECA, 5’-(N-

Ethylcarboxamido)adenosine.

Values shown are the arithmetic mean 6 SEM (Emax) and geometric mean 6 95% confidence interval (EC50, mM) from

experiments performed with samples of n ¼ 4 different subjects (A1 selectivity: n ¼ 5).

Figure 3. Concentration-response curveto CPA in the absence and presence of

increasing concentrations of the adeno-

sine A1 receptor antagonist, DPCPX

(1 nM–100 nM) (A). Calculation of pA2

value for DPCPX using modification of the

Schild analysis (see MATERIALS AND METHODS

for description) (B). Each symbol repre-

sents mean (6SEM) from experimentsperformed with bronchial preparations

from different subjects (n ¼ 4). DR ¼dose ratio.

1226 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 45 2011

with the H1-receptor and Cys-LT1 receptor antagonists reducedthe contractile response to CPA by 37.9 6 2.5% (Figure 6B).

After passive sensitization, human bronchi became more sen-sitive to the inhibitory effect of epinastine and MK571 (Figure6B). Thus, the contractile response to CPA was reduced by33.3 6 3.8% in the presence of epinastine alone (P , 0.05),MK571 alone (P , 0.05), and in combination; this treatmentabolished the contractile response to CPA (Figure 6B).

The contractile response induced by the nonselective activa-tion of adenosine receptors using 5-AMP was significantlyinhibited by epinastine and MK571 in nonsensitized humanbronchial preparations (Figure 6B; P , 0.05). After passivesensitization, epinastine significantly (P , 0.05) reduced thecontractile potency () of 5-AMP (geometric mean [95% CI]:passive, 0.30 [0.20–0.41] mM versus control, 0.13 [0.07–0.19]mM), but not Emax compared with control bronchi not treatedwith the epinastine. In contrast, pretreatment of either nonsen-sitized or sensitized bronchial preparations with MK571, aloneor in combination with epinastine, completely abolished thecontractile response to 5-AMP (Figure 6B).

DISCUSSION

The present study demonstrates that the activation of specificadenosine receptor subtypes can induce changes in the tone of hu-man isolated bronchi. We have demonstrated that the stimulationof adenosine receptors, by selective and nonselective adenosine re-ceptor agonists, induces a typical sigmoidal response to cumulative

increases in concentration of selective agonists. The pharmacolog-ical characterization of the response to adenosine agonists, per-formed by employing potent and selective full adenosinesubtype receptor agonists (online supplement), showed a rank or-der of activity of A1 . A2B .. A2A ¼ A3 (33). In addition, wealso investigated the function of each receptor subtype in re-sponse to a nonspecific agonist, such as 5-AMP or NECA, in thepresence of selective adenosine receptor antagonists, and havedemonstrated that all the antagonists employed were selectivefor their adenosine receptor subtypes. This experimental ap-proach of selectively blocking three out of four of the adenosinereceptor subtypes allowed us to reveal that adenosine A1 recep-tors and, to a lesser degree, adenosine A2B receptors were in-volved in the contraction of human bronchial smooth muscle.The intrinsic activity of NECA is comparable to CPA whenassessed against the human cloned adenosine A1 receptor(34), and exhibited the greatest intrinsic activity at the clonedadenosine A2B receptors compared with other adenosine ago-nists (35). Consequently, the differences in contractile potencyand magnitude of the contractile response to the different aden-osine agonists used in this study cannot be accounted for bydifferences in intrinsic activity per se, but are likely a conse-quence of the degree of adenosine receptor subtype expressionin human bronchial tissue (34).

It was apparent that adenosine A1 receptor signaling pro-duced the greatest contractile responses in human bronchialpreparations, and this response was characterized by tachyphy-laxis with repeated administration. This is consistent with

Figure 4. Linear regression of contraction

(tension, mg) induced by consecutive ad-

ministration of capsaicin (10 mM) at

hourly intervals in human isolated bron-chi (A). Concentration-response curve to

CPA in untreated tissues (open circles) and

tissues either treated (closed symbols)

with capsaicin, thiorphan (1 mM) or theircombination (B). Each symbol represents

mean (6SEM) from experiments per-

formed with bronchial preparations fromdifferent subjects (n¼ 4). #P, 0.05, ##P,0.01 versus the first capsaicin-induced

contraction (A); **P , 0.01 versus control

bronchi (nondesensitized and nonthior-phan treated) (B).

TABLE 3. EFFECT OF PASSIVE SENSITIZATION ON THE RESPONSE OF HUMAN ISOLATED RESPONSE TO ADENOSINE RECEPTORSTIMULATION AND TREATMENT WITH SELECTIVE ADENOSINE RECEPTOR ANTAGONISTS

Emax EC50 Emax EC50

Agonist Selectivity (mg) (mM) Antagonist (mg) (mM)

5-AMP Nonselective 137.7 6 9.5* 0.13 (0.07–0.19) 1DPCPX 134.8 6 8.3 0.36 (0.29–0.43)†

NECA Nonselective 143.0 6 18.2* 0.10 (0.03–0.17) 1DPCPX 134.4 6 9.7 0.59 (0.27–0.90)†

CPA A1 389.7 6 52.8* 0.10 (0.04–0.16) 1DPCPX 355.4 6 43.8 0.62 (0.27–0.97)†

CGS21680 A2A 54.7 6 9.1 0.08 (0.07–0.10) 1ZM241385 ND ND

CV1808 1 ZM241385 A2B 156.6 6 55.0 0.05 (0.03–0.07) 1MRS1754 123.7 6 10.4 0.21 (0.13–0.28)†

IB-MECA A3 61.5 6 15.6 0.08 (0.06–0.09) 1MRS1220 ND ND

Definition of abbreviations: 5-AMP, adenosine monophosphate; CPA, N6-cyclopentyladenosine; EC50, concentration required to cause a 50% maximal effect; Emax,

maximal effect; IB-MECA, 1-Deoxy-1-[6-[[(3-iodophenyl)methyl]amino]-9H-purin-9-yl]-N-methyl-b-D-ribofuranuronamide; ND, not detectable; NECA, 5’-(N-Ethylcar-

boxamido)adenosine.

Values shown are the arithmetic mean 6 SEM (Emax) and geometric mean6 95% confidence interval (EC50, mM) from experiments performed with samples of n ¼ 4

different subjects (5-AMP: n ¼ 5).

* P , 0.01 versus the nonsensitized bronchi.yP , 0.05 versus passively sensitized bronchi not treated with antagonists.

Calzetta, Spina, Cazzola, et al.: Adenosine and Human Airways 1227

clinical data reporting tachyphylaxis to inhaled adenosine insubjects without and those with asthma (36, 37). The weakerbronchial contraction induced by the nonselective adenosinereceptor stimulation, in comparison to that mediated by theselective A1 adenosine receptor activation with CPA, may beexplained when considering the antithetical transductional

effects after activation of different adenosine receptor subtypes,which can either decrease or increase adenylyl cyclase activity, forinstance, thus leading to a functional antagonism offsetting the con-tractile response induced by adenosine A1 receptor activation (38).

The competitive adenosine A1 receptor antagonist, DPCPX(33, 39), caused a noticeable decrease in the maximum response

Figure 5. The effect of muscarinic receptor antagonism on concentration-response curve to CPA (control, open circles) in epithelium intact (A),

epithelium-denuded (B), and passively sensitized (C) bronchi (treated tissues, half closed symbols). Each symbol represents mean (6SEM) from

experiments performed with bronchial preparations from different subjects ([A] n ¼ 4; [B and C] n ¼ 3). *P, 0.05, ***P, 0.001 versus adenosine A1

receptor agonist.

Figure 6. Bar graph representing the reduction in baseline tension (relaxation) of human isolated bronchi in the presence of the H1 receptor

antagonist, epinastine, and Cys-LT1 receptor antagonist, MK571 (both 1 mM) (A). The effect of epinastine and MK571 (both 1 mM) on the maximal

contraction induced by CPA and 5-AMP in human isolated bronchial preparations (B). Each symbol represents mean (6SEM) from experimentsperformed with bronchial preparations from different subjects (n ¼ 4). Nonsensitized tissue is represented by lightly shaded columns and passively

sensitized tissue is represented by darkly shaded columns. #P , 0.05, ##P , 0.01 versus nonsensitized bronchi (A); *P , 0.05, **P , 0.01, ***P ,0.001 versus control bronchi (B).

1228 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 45 2011

to CPA, which was unlikely to be a consequence of a nonspecificaction on airway smooth muscle contractility, as the contractileresponse to acetylcholine was not affected by DPCPX. Further-more, in all experiments, we allowed CPA to achieve equilibriumin the presence of antagonists, and we confirmed the absence ofany overt toxicity by DPCPX by comparing the effect of 100 mMacetylcholine at the beginning and at the end of experiments(data not shown) (40). This profile of loss of Emax to CPA inthe presence of a competitive antagonist resembles the responseobserved for indirectly acting substances that release endogenousintermediaries, the concentrations of which are the limiting factorand therefore cannot compete with antagonist-occupied recep-tors (40, 41). Furthermore, DPCPX, employed at concentrationstwo logarithmic units higher than its pA2, was effective in induc-ing a rightward shift of the 5-AMP curve of one logarithmic unit.This finding may be explained by considering a dominant in-volvement of A1 receptors, but the possibility remains that, inconditions of adenosine A1 receptor blockade, other subtypesmay participate in this response.

We next explored the possibility that the contractile responseafter activation of adenosine A1 receptors was mediated bydirect activation of airway smooth muscle, or indirectly by therelease of endogenous substances in both nonsensitized andsensitized human bronchial tissue. The demonstration of tachy-phylaxis to the contractile response to CPA could infer thedepletion of endogenous chemical substances after repeatedapplication or receptor desensitization. However, we confirmedthe involvement of neural mechanisms after activation of aden-osine A1 receptors in human bronchial tissue. Our data areconsistent with the hypothesis that sensory nerves may regulatethe human airway smooth muscle contraction after adenosineA1 receptor activation (26, 42, 43), because capsaicin-induceddesensitization inhibited the adenosine A1 receptor mediationof human bronchial tissues. Although we have demonstratedthat thiorphan increased the contractile potency of CPA, thiswas absent in capsaicin-desensitized tissues, further suggestinga role of capsaicin-sensitive sensory nerves in the contractileresponse of human isolated airways. Clinical studies have dem-onstrated a neural contribution to 5-AMP–induced bronchocon-striction in patients with asthma after inhalation of the loopdiuretics, furosemide and bumetanide (25, 44). Furthermore,the bronchoconstriction induced by adenosine in guinea pigswas attenuated by capsaicin in vivo (27).

We also demonstrate that a component of the contractile re-sponse induced by CPAwas dependent on the activation of para-sympathetic neurones, an observation of interest, as adenosineA1 and A2A receptors have been found on cholinergic neurones(45). The stimulus-coupling mechanism appears to be weak,because atropine depressed the Emax response to CPA, whichis characteristic of the effect of competitive antagonists on func-tional responses to endogenously released neurotransmitters(40, 41). The possibility that CPA stimulated the release ofacetylcholine from nonneuronal sources (e.g., epithelium) can-not be excluded (46). Indeed, we have previously shown theexpression of adenosine A1 receptors on airway epithelium(3), and activation of these receptors on the epithelium causesa functional response, albeit an increase in mucin gene expres-sion and mucus hypersecretion (47). However, first, we haveconclusively demonstrated that the contractile response toCPA was not dependent on the presence of the epitheliumand, second, the degree of antagonism induced by atropinewas similar in epithelium-intact and -denuded tissues. Thesedata exclude extraneuronal sources of acetylcholine contribut-ing to the contractile response to CPA in nonsensitized andpassively sensitized human bronchial preparations. Our dataare therefore consistent with clinical studies that report that

airway obstruction induced by 5-AMP and CPA was signifi-cantly inhibited by atropine as a consequence of activation ofparasympathetic reflex in the airways (48).

Clinical studies and preclinical models have consistently shownthat adenosinemediates airways obstruction in subjects with asthmaor in models of allergic inflammation, but not in healthy subjects (2,5, 18, 48–51). Accordingly, with the demonstration that there is anelevated expression of adenosine A1 receptors in bronchial biopsyspecimens from subjects with asthma (3), our results demonstrate,for the first time, a role for these receptors in the contraction ofpassively sensitized airways. Passive sensitization using humanasthmatic serum is frequently used to model asthmatic airways(28, 29). The contraction induced by CPA was almost twofoldgreater than that observed in normal bronchi, which was alsoinhibited by a selective adenosine A1 receptor antagonist. Weruled out a role for mast cell degranulation via activation of aden-osine A2B receptors in mediating the adenosine-induced contrac-tion of human airways (5), because an adenosine A2B receptorantagonist was without significant effect on this response, and,indeed, adenosine A2B receptor activation appears to play a negli-gible role in the airway contraction induced by 5-AMP. Nonethe-less, this modest role of adenosine A2B receptors may be importantin situations in which adenosine A1 receptors are antagonized.Thus, activation of adenosine A2B receptors could stimulate therelease of endogenous substances in the human isolated bronchi,which, in turn, stimulate bronchial smooth muscle. We considerour results clinically relevant, as they have been obtained usinghuman tissue and are consistent with other literature (28, 29).

We also used selective H1-receptor and Cys-LT1 receptorantagonists to re-evaluate the original findings of Bjorck andDahlen (18) concerning the role of adenosine in mediating con-traction of human bronchial tissue from subjects with asthma,which was suggested to be secondary to the release of leuko-trienes and histamine. In our study, human isolated airwaysdemonstrated a high degree of intrinsic tone, mediated by thespontaneous generation of histamine and leukotrienes. Para-doxically, this intrinsic tone was stronger in nonsensitized thanin passively sensitized samples. This could have been due to lossof endogenous mediators during the sensitization protocol.However, this was unlikely in view of the inhibitory effect ofthese receptor antagonists on CPA-induced contractions in sensi-tized bronchi. Of particular interest was the finding that theleukotriene-dependent component of the contractile responsewas more evident in passively sensitized tissues. The mechanismby which passive sensitization increased the contractile responseto CPA is unclear, but it is known that IgE binding to mast cellscan cause sensitization (28, 52). However, we have demon-strated that the adenosine A2B receptor is not involved in thecontraction induced by 5-AMP, and therefore rule out a role forthese receptors. Whether IgE binding to non–mast cell sites (53,54) leads to sensitization of the response to CPA remains to beestablished. It is also very interesting to note that, in passivelysensitized bronchi, the contraction induced by CPA or 5-AMPwas more sensitive to antagonism of Cys-LT1 than by H1 recep-tors. Our findings support the results of Bjorck and Dahlen, butcall in to question whether the effects of adenosine are primar-ily mediated via mast cell degranulation,

This study provides clear evidence for the role of adenosineA1

receptors in eliciting human airway smooth muscle contraction,both directly and indirectly, after the activation of capsaicin-sensitive sensory nerves and the release of leukotrienes.

Author Disclosure: L.C. received institutional grant support from Pneumolabs IncUK. D.S. was a consultant for Veronapharma plc. M.C. does not have a financialrelationship with a commercial entity that has an interest in the subject of thismanuscript. C.P.P. is a Board member and consultant for Veronapharma plc. Hewas an expert witness for Almirall and Teva, and received lecture fees from

Calzetta, Spina, Cazzola, et al.: Adenosine and Human Airways 1229

Merck. He owns stocks, or options of Veronapharma plc and received travel/accommodations from Almirall. F.F., E.A.R., and M.G.M. do not have a financialrelationship with a commercial entity that has an interest in the subject of thismanuscript.

References

1. Polosa R, Holgate ST. Adenosine bronchoprovocation: a promising

marker of allergic inflammation in asthma? Thorax 1997;52:919–923.

2. Brown RA, Spina D, Page CP. Adenosine receptors and asthma. Br J

Pharmacol 2008;153:S446–S456.

3. Brown RA, Clarke GW, Ledbetter CL, Hurle MJ, Denyer JC, Simcock

DE, Coote JE, Savage TJ, Murdoch RD, Page CP, et al. Elevated

expression of adenosine A1 receptor in bronchial biopsy specimens

from asthmatic subjects. Eur Respir J 2008;31:311–319.

4. Spicuzza L, Bonfiglio C, Polosa R. Research applications and implica-

tions of adenosine in diseased airways. Trends Pharmacol Sci 2003;24:

409–413.

5. Spicuzza L, Di Maria G, Polosa R. Adenosine in the airways: implica-

tions and applications. Eur J Pharmacol 2006;533:77–88.

6. Olah ME, Stiles GL. The role of receptor structure in determining

adenosine receptor activity. Pharmacol Ther 2000;85:55–75.

7. Cronstein BN, Van de Stouwe M, Druska L, Levin RI, Weissmann G.

Nonsteroidal antiinflammatory agents inhibit stimulated neutrophil

adhesion to endothelium: adenosine dependent and independent

mechanisms. Inflammation 1994;18:323–335.

8. Blackburn MR. Too much of a good thing: adenosine overload in

adenosine-deaminase–deficient mice. Trends Pharmacol Sci 2003;24:

66–70.

9. 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–adenosine

amplification pathway. J Clin Invest 2003;112:332–344.

10. Lasley RD, Rhee JW, Van Wylen DG, Mentzer RM Jr. Adenosine A1

receptor mediated protection of the globally ischemic isolated rat

heart. J Mol Cell Cardiol 1990;22:39–47.

11. Linden J. Molecular approach to adenosine receptors: receptor-mediated

mechanisms of tissue protection. Annu Rev Pharmacol Toxicol 2001;41:

775–787.

12. Rudolphi KA, Schubert P, Parkinson FE, Fredholm BB. Neuro-

protective role of adenosine in cerebral ischaemia. Trends Pharmacol

Sci 1992;13:439–445.

13. Pauwels RA, Van der Straeten ME. An animal model for adenosine-

induced bronchoconstriction. Am Rev Respir Dis 1987;136:374–378.

14. Hannon JP, Tigani B, Schuurman HJ, Fozard JR. Suppression of

adenosine A(3) receptor–mediated hypotension and mast cell de-

granulation in the rat by dexamethasone. J Pharmacol Exp Ther 2002;

302:725–730.

15. Hannon JP, Tigani B, Wolber C, Williams I, Mazzoni L, Howes C,

Fozard JR. Evidence for an atypical receptor mediating the aug-

mented bronchoconstrictor response to adenosine induced by allergen

challenge in actively sensitized brown norway rats. Br J Pharmacol

2002;135:685–696.

16. Thorne JR, Danahay H, Broadley KJ. Analysis of the bronchocon-

strictor responses to adenosine receptor agonists in sensitized guinea-

pig lungs and trachea. Eur J Pharmacol 1996;316:263–271.

17. Feoktistov I, Polosa R, Holgate ST, Biaggioni I. Adenosine A2b recep-

tors: a novel therapeutic target in asthma? Trends Pharmacol Sci

1998;19:148–153.

18. Bjorck T, Gustafsson LE, Dahlen SE. Isolated bronchi from asthmatics

are hyperresponsive to adenosine, which apparently acts indirectly by

liberation of leukotrienes and histamine. Am Rev Respir Dis 1992;145:

1087–1091.

19. Rajakulasingam K, Polosa R, Church MK, Howarth PH, Holgate ST.

Effect of inhaled frusemide on responses of airways to bradykinin and

adenosine 59-monophosphate in asthma. Thorax 1994;49:485–491.

20. Holgate ST. The Quintiles Prize Lecture 2004: the identification of the

adenosine A2b receptor as a novel therapeutic target in asthma. Br J

Pharmacol 2005;145:1009–1015.

21. Polosa R, Santonocito G, Magri S, Paolino G, Armato F, Pagano C, Crimi

N. Neutral endopeptidase inhibition with inhaled phosphoramidon: no

effect on bronchial responsiveness to adenosine 59-monophosphate

(AMP) in asthma. Eur Respir J 1997;10:2460–2464.

22. Kraan J, Vink-Klooster H, Postma DS. The NK-2 receptor antagonist SR

48968c does not improve adenosine hyperresponsiveness and airway

obstruction in allergic asthma. Clin Exp Allergy 2001;31:274–278.

23. Polosa R, Phillips GD, Rajakulasingam K, Holgate ST. The effect of in-

haled ipratropium bromide alone and in combination with oral terfena-

dine on bronchoconstriction provoked by adenosine 59-monophosphate

and histamine in asthma. J Allergy Clin Immunol 1991;87:939–947.

24. Rorke S, Jennison S, Jeffs JA, Sampson AP, Arshad H, Holgate ST.

Role of cysteinyl leukotrienes in adenosine 59-monophosphate in-

duced bronchoconstriction in asthma. Thorax 2002;57:323–327.

25. O’Connor BJ, Chung KF, Chen-Worsdell YM, Fuller RW, Barnes PJ.

Effect of inhaled furosemide and bumetanide on adenosine 59-monophosphate– and sodium metabisulfite–induced bronchocon-

striction in asthmatic subjects. Am Rev Respir Dis 1991;143:1329–

1333.

26. Polosa R, Rorke S, Holgate ST. Evolving concepts on the value of

adenosine hyperresponsiveness in asthma and chronic obstructive

pulmonary disease. Thorax 2002;57:649–654.

27. Manzini S, Ballati L. 2-Chloroadenosine induction of vagally-mediated

and atropine-resistant bronchomotor responses in anaesthetized

guinea-pigs. Br J Pharmacol 1990;100:251–256.

28. Rabe KF. Mechanisms of immune sensitization of human bronchus. Am

J Respir Crit Care Med 1998;158:S161–S170.

29. Watson N, Bodtke K, Coleman RA, Dent G, Morton BE, Ruhlmann E,

Magnussen H, Rabe KF. Role of IgE in hyperresponsiveness induced

by passive sensitization of human airways. Am J Respir Crit Care Med

1997;155:839–844.

30. Goodman LS, Gilman A, Brunton LL, editors. Goodman & Gilman’s

manual of pharmacology and therapeutics. New York: McGraw Hill

Medical; 2008.

31. Motulsky H, Christopoulos A. Fitting models to biological data using

linear and nonlinear regression: a practical guide to curve fitting.

Oxford: Oxford University Press; 2004.

32. Kenakin TP. The Schild regression in the process of receptor classifi-

cation. Can J Physiol Pharmacol 1982;60:249–265.

33. Klotz KN. Adenosine receptors and their ligands. Naunyn Schmiede-

bergs Arch Pharmacol 2000;362:382–391.

34. Cordeaux Y, Briddon SJ, Megson AE, McDonnell J, Dickenson JM, Hill

SJ. Influence of receptor number on functional responses elicited by

agonists acting at the human adenosine A(1) receptor: evidence for

signaling pathway–dependent changes in agonist potency and relative

intrinsic activity. Mol Pharmacol 2000;58:1075–1084.

35. Cooper J, Hill SJ, Alexander SP. An endogenous A2b adenosine re-

ceptor coupled to cyclic AMP generation in human embryonic kidney

(HEK 293) cells. Br J Pharmacol 1997;122:546–550.

36. Crimi N, Palermo F, Oliveri R, Polosa R, Vancheri C, Palermo B,

Mistretta A. Tachyphylaxis to repeated challenges with inhaled

adenosine in asthmatic subjects. Ann Allergy 1990;65:134–138.

37. Daxun Z, Rafferty P, Richards R, Summerell S, Holgate ST. Airway

refractoriness to adenosine 59-monophosphate after repeated inhala-

tion. J Allergy Clin Immunol 1989;83:152–158.

38. Torphy TJ, Rinard GA, Rietow MG, Mayer SE. Functional antagonism

in canine tracheal smooth muscle: inhibition by methacholine of the

mechanical and biochemical responses to isoproterenol. J Pharmacol

Exp Ther 1983;227:694–699.

39. von der Leyen H, Schmitz W, Scholz H, Scholz J, Lohse MJ, Schwabe U.

Effects of 1,3-dipropyl-8-cyclopentylxanthine (DPCPX), a highly se-

lective adenosine receptor antagonist, on force of contraction in guinea-

pig atrial and ventricular cardiac preparations. Naunyn Schmiedebergs

Arch Pharmacol 1989;340:204–209.

40. Kenakin T, Jenkinson S, Watson C. Determining the potency and mo-

lecular mechanism of action of insurmountable antagonists. J Phar-

macol Exp Ther 2006;319:710–723.

41. Black JW, Jenkinson DH, Kenakin TP. Antagonism of an indirectly

acting agonist: block by propranolol and sotalol of the action of

tyramine on rat heart. Eur J Pharmacol 1980;65:1–10.

42. Reynolds SM, Docherty R, Robbins J, Spina D, Page CP. Adenosine

induces a cholinergic tracheal reflex contraction in guinea pigs in vivo

via an adenosine A1 receptor–dependent mechanism. J Appl Physiol

2008;105:187–196.

43. Hua X, Erikson CJ, Chason KD, Rosebrock CN, Deshpande DA, Penn

RB, Tilley SL. Involvement of A1 adenosine receptors and neural

1230 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 45 2011

pathways in adenosine-induced bronchoconstriction in mice. Am J

Physiol Lung Cell Mol Physiol 2007;293:L25–L32.

44. Polosa R, Rajakulasingam K, Prosperini G, Church MK, Holgate ST.

Relative potencies and time course of changes in adenosine

59-monophosphate airway responsiveness with inhaled furosemide

and bumetanide in asthma. J Allergy Clin Immunol 1993;92:

288–297.

45. Dixon AK, Gubitz AK, Sirinathsinghji DJ, Richardson PJ, Freeman TC.

Tissue distribution of adenosine receptor mRNAs in the rat. Br J

Pharmacol 1996;118:1461–1468.

46. Kummer W, Lips KS, Pfeil U. The epithelial cholinergic system of the

airways. Histochem Cell Biol 2008;130:219–234.

47. McNamara N, Gallup M, Khong A, Sucher A, Maltseva I, Fahy J,

Basbaum C. Adenosine up-regulation of the mucin gene, MUC2, in

asthma. FASEB J 2004;18:1770–1772.

48. Keir S, Boswell-Smith V, Spina D, Page C. Mechanism of adenosine-

induced airways obstruction in allergic guinea pigs. Br J Pharmacol

2006;147:720–728.

49. Smith N, Broadley KJ. Adenosine receptor subtypes in airways responses

of sensitized guinea-pigs to inhaled ovalbumin. Pulm Pharmacol Ther

2010;23:355–364.

50. Wilson CN. Adenosine receptors and asthma in humans. Br J Pharmacol

2008;155:475–486.

51. Polosa R, Blackburn MR. Adenosine receptors as targets for therapeutic

intervention in asthma and chronic obstructive pulmonary disease.

Trends Pharmacol Sci 2009;30:528–535.

52. Schmidt D, Rabe KF. Immune mechanisms of smooth muscle hyperre-

activity in asthma. J Allergy Clin Immunol 2000;105:673–682.

53. Williams CM, Galli SJ. The diverse potential effector and immunoreg-

ulatory roles of mast cells in allergic disease. J Allergy Clin Immunol

2000;105:847–859.

54. Ikeda RK, Miller M, Nayar J, Walker L, Cho JY, McElwain K, McElwain

S, Raz E, Broide DH. Accumulation of peribronchial mast cells in

a mouse model of ovalbumin allergen induced chronic airway inflam-

mation: modulation by immunostimulatory DNA sequences. J Immunol

2003;171:4860–4867.

Calzetta, Spina, Cazzola, et al.: Adenosine and Human Airways 1231