Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Modulation of airway smooth muscle (ASM) contractile state is the principal means of regulating airway diameter. In asthma, airway inflammation begets high levels of contractile agents that cause ASM contraction, airway constriction, and symptomatic wheezing. G protein-coupled receptors (GPCRs) that activate the heterotrimeric G protein G_s-adenylyl cyclase pathway play an important role in counteracting procontractile stimuli. For example, activation of beta-2-adrenergic receptors (₂ARs) on ASM promotes functional antagonism of bronchoconstricting agents through multiple mechanisms associated with cyclic adenosine monophosphate (cAMP)-mediated activation of protein kinase (PK) A. Consequently, inhaled beta-agonists are the most widely used agents in asthma therapy and are universally recognized as the treatment of choice for acute asthma attacks. The critical role of beta-agonists and ₂ARs in asthma therapy is further underscored by the numerous studies examining the impact of the asthma state or asthma treatment on ₂AR responsiveness.

Conversely, the endogenous nucleoside adenosine is believed to promote bronchoconstriction through indirect, and possibly direct, effects on ASM. Endogenous adenosine, whose concentration is elevated in the asthmatic lung (1) where inflammatory cells and platelets represent potential sources, activates adenosine receptors (ARs) on mast cells to induce histamine release and ASM contraction (2). In addition, A1 adenosine receptors (A1AR) on ASM may also be a target of adenosine. In an allergic rabbit model, administration of aerosolized antisense oligodeoxynucleotides specific for the A1AR reduced ASM A1AR density and attenuated adenosine or allergen-induced bronchoconstriction (3). However, studies to date have failed to identify A1AR-mediated responses in human ASM from either nonasthmatic or asthmatic tissues, suggesting that A1AR expression in ASM may be species-dependent.

In this study, we characterize the effects of AR activation on second messenger generation and GPCR regulation in human ASM cultures. With a pharmacology consistent with activation of the G_s-coupled A_2b subtype (A_2bAR) of ARs, acute exposure of human ASM cultures to exogenous adenosine ligands results in cAMP accumulation, homologous desensitization of the A_2bAR, and heterologous desensitization of the ₂AR. Interestingly, human ASM is shown to release endogenous adenosine that modulates basal intracellular cAMP and promotes desensitization of G_s-coupled receptors, effects also consistent with A_2bAR activation. However, chronic administration of AR ligands or manipulation of adenosine transport/conversion to enhance accumulation of extracellular adenosine elicited a dual effect of G_s-coupled receptor desensitization and pertussis toxin-sensitive sensitization of adenylyl cyclase. These data demonstrate that human ASM is a direct target of exogenous and autocrine adenosine, with effects determined by differential contributions of A_2bARs and A1ARs that are time-dependent. Accordingly, the relative distribution and activation of AR subtypes in ASM may influence airway function in diseases such as asthma and warrant consideration in therapeutic strategies that target ARs or alter nucleotide/nucleoside levels in the airway.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Materials

Xanthine amine congener (XAC), 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), MRS 1220, erythro-9-(2-hydroxy-3-nonyl)adenine hydrochloride (EHNA), dipyridamole (DIP), nitrobenzylthioinosine (NBTI), N⁶-(4-aminobenzyl)-9-[5-(methylcarbonyl)--D-ribofuranosyl]adenine (AB-MECA), and 5-iodotubercidin (5-IODO) were purchased from Research Biochemicals International (Natick, MA). 2,8-[³H]adenine, 8-[¹⁴C]cAMP, [¹²⁵I]adenosine 3', 5'-cyclicphosphoric acid (2,200 Ci/mmol), and [³H]DPCPX (120 Ci/mmol) were purchased from NEN Dupont (Boston, MA). Na[¹²⁵I] was purchased from Amersham-Pharmacia (Piscataway, NJ). cAMP antibody was a gift from Mario Ascoli (University of Iowa, Iowa City, IA). pEGFPN1 was purchased from Clontech (Palo Alto, CA). All other reagents were purchased from Sigma (St. Louis, MO) or from sources described previously (4).

Human ASM Cell Culture

Human ASM cultures were established as described by Panettieri and coworkers (5) from human tracheae obtained from lung transplant donors, in accordance with procedures approved by the University of Pennsylvania Committee on Studies Involving Human Beings. Characterization of these cell lines with regard to immunofluorescence of smooth muscle actin and agonist-induced changes in cytosolic calcium has been previously reported (4, 6, 7). Third to fifth passage cells were plated at a density of 10⁴ cells/ cm² in either 24-well (for cAMP accumulation assays in intact cells) or 15-cm plates (adenylyl cyclase assays) and maintained in fetal bovine serum (FBS)-supplemented Ham's F12 medium as described previously (8). Confluent cells were growth-arrested by refeeding cells with Ham's F12 supplemented with 5 µg/ml each insulin and transferrin (IT medium) for 24 h before pretreatment.

Accumulation of cAMP in Intact Cells

Except where noted, growth-arrested human ASM cultures were pretreated for 30 min or 18 h with the appropriate vehicle or various agents: 0.5 U/ml adenosine deaminase (AD); 0.1 to 1 µM XAC; 50 to 100 nM DPCPX; 100 nM MRS 1220; 5 to 100 µM 5'-(N-ethylcarboxamido)-adenosine (NECA); 0.01 to 1.0 mM N⁶-cyclopentyladenosine (CPA); 1 to 10 µM EHNA; 1 to 10 µM DIP, 0.1 to 1 µM NBTI; or 0.1 to 1 µM 5-IODO. After pretreatment(s), cells were washed twice with cold phosphate-buffered saline (PBS) and subsequently stimulated with 500 µl PBS containing 300 µM ascorbic acid, 1 mM RO-20-1724 (phosphodiesterase inhibitor), and either vehicle (basal), isoproterenol (ISO), prostaglandin (PG)E₂, CPA, NECA, 2-p-(2-carboxyethyl) phenethylamino-5'-N-ethylcarboxamido adenosine (CGS 21680), or forskolin (FSK) at the indicated concentrations for 10 min at 37°C. cAMP was isolated and quantified by radioimmunoassay as described previously (4).

Radioligand Binding

[¹²⁵I]AB-MECA was synthesized as described previously (9). For radioligand binding assays, human ASM cells in 100-mm dishes were washed twice with ice-cold 10 mM Tris, pH 7.4, at 5°C and 5 mM ethylenediaminetetraacetic acid (EDTA), and then scraped into the same buffer. Cells were disrupted by dounce homogenization, and the homogenates were centrifuged at 43,000 × g at 5°C. The resulting membrane pellet was resuspended in a binding buffer of 50 mM Tris, pH 8.26, at 5°C, 10 mM MgCl₂, 1 mM EDTA, and AD was added to give a final concentration of 4 U/ml. In [¹²⁵I]AB-MECA saturation binding assays, nonspecific binding was defined with 100 µM CPA. Endogenous A1AR density was estimated from one-point binding analyses using ~ 2 nM [¹²⁵I]AB-MECA in the presence or absence of 1 µM DPCPX to determine nonspecific binding. Expression of A1AR-green fluorescent protein (GFP) was estimated from one-point binding analyses using 10 nM [³H] DPCPX in the presence or absence of 100 µM CPA. Predicted potency and efficacy of all ligands were confirmed in parallel experiments examining binding to A1AR or A3AR heterologously expressed in Chinese hamster ovary (CHO) cells. All binding assays were conducted at 37°C for 1 h and terminated by filtration using a Brandel (Gaithersburg, MD) cell harvester and three rapid washes with binding buffer supplemented with 0.01% 3-(3-cholamidopropyl) diethy-ammonio-1 propanesulfonate.

Bioassay of Adenosine from Conditioned Media from Human ASM Cultures

To test for the presence of adenosine in human ASM culture medium, confluent human ASM cultures were fed serum-free Ham's F12 medium for 24 h. This medium was harvested and used to stimulate COS-1 cells (as described above for human ASM cells) that had been grown to confluence in 24-well plates and washed twice with cold PBS. On the same plate, wells were stimulated with IT medium containing concentrations of adenosine ranging from 0 to 1 µM. Duplicate wells for each condition contained 0.5 U/ml AD. After 10 min of stimulation at 37°C, cAMP was isolated and quantified by radioimmunoassay as described previously.

Adenylyl Cyclase Assay in Cell Homogenates

After pretreatment with various agents as described previously, cells from 15-cm plates were washed with cold PBS, harvested by scraping into 10 ml of ice-cold PBS, and pelleted by centrifugation at 200 × g for 10 min followed by snap-freezing. Adenylyl cyclase activity was subsequently measured in cell homogenates using a competitive protein binding assay and [8-³H]cAMP as described previously (10).

Plasmid Construction

Constructs encoding chimeras of an enhanced variant of green fluorescent protein (EGFP, from Aequorea victoria) (11) fused to the C-termini of the human A1ARs and the rat A_2bARs were generated by polymerase chain reaction (PCR) cloning. Sense and antisense primers were designed to amplify the open reading frame of the A1AR (from pCMV5A1AR, provided by Tim Palmer, University of Glasgow, Glasgow, UK) with a 5' HA tag (sense, 5' GCCAAGCTTACCATGGGTTACCCTTATGATGTGCCAGATTATGCCTCTCCCATGCCGCCCTCCATCTCA 3'; antisense, 5' ACGCGTCGACTGGTCATCAGGCCTCTCTCT 3') and the A_2bAR (from pcDNA3A_2bAR, provided by Eamonn Kelly, Bristol University, Bristol, UK) (sense, 5' GCCAAGCTTACCATGCAGCTAGAGACGCAGGAC 3', antisense, 5' GCGGATCCTGCAAGCTCAGACTGAAAGT 3'). The PCR products were digested and cloned into the HindIII/Sal1 (pA1AR-GFP) and HindIII/BamH1 (pA_2bAR-GFP) sites of pEGFPN1 such that receptor sequence remained in frame with the downstream GFP sequence. DNA was amplified in DH5 cells and purified by the cesium chloride method. Orientation, in-frame alignment, and sequence were confirmed by dideoxynucleotide sequencing. When transfected into CHO cells, A1AR-GFP displayed receptor radioligand binding properties ([³H]DPCPX binding) and mediated agonist-activated p42/p44 phosphorylation similar to that of the wild-type A1AR (data not shown).

Human ASM Transfection

Human ASM cells were passaged at a density of 1.5 × 10⁴ cells/ml onto 15-cm plates and grown in 27 ml Ham's F12/10% FBS. The medium was replaced with fresh Ham's F12/10% FBS 24 h later. Cells were then transfected by addition of a N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid-based CaPO₄ mixture (GIBCO-BRL, Grand Island, NY) containing 15 µg of carrier DNA, and 45 µg of either pEGFPN1, pA1ARGFP, or pA_2bARGFP. Cultures were then incubated for 24 h, washed three times with Ca²⁺- and Mg²⁺-free PBS, and refed Ham's F12/10% FBS. Cells were harvested 24 h later, and human ASM cells expressing GFP were sorted to > 98% purity using a Coulter Epics Elite ESP Flow Cytometer (Beckman Coulter, Fullerton, CA). Cell autofluorescence was established in mock-transfected (no DNA) cells. From each sort, a population of cells with minimal fluorescence (comparable to that of mock-transfected cells) was also sorted in parallel with GFP-containing cells and served as an additional control. In subsequent analyses, these GFP-negative cells displayed near-identical properties to cells transfected with pEGFPN1 and sorted as GFP-positive (data not shown). After sorting, cells were plated at a density of 3 × 10⁴ cells/cm² in 48-well plates and grown in Ham's F12/10% FBS. Cells were refed IT media 24 h later for 24 h, then pretreated and subsequently stimulated as described previously. Determination of protein density per well was assessed in parallel wells using the Bradford assay.

Data Presentation and Statistical Analysis

Data points from individual assays represent the mean values from duplicate or triplicate measurements. Radioligand binding data were analyzed using GraphPad Prism software (GraphPad, San Diego, CA). Data are presented as mean ± standard error (SE). Statistically significant differences among groups were assessed by either analysis of variance (ANOVA) with Fisher's protected least significant difference (PLSD) post-hoc analysis (Statview 4.5; Abacus Concepts, Berkeley, CA) or by t test for paired samples, with P values < 0.05 sufficient to reject the null hypothesis.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Human ASM Cultures Exhibit Functional Responses Consistent with A_2bAR Expression

Initial studies characterized the response of human ASM cells to various agonists known to stimulate members of the AR family. Stimulation with 1 mM adenosine or 100 µM NECA (a nonspecific AR agonist) elicited a rapid production of cAMP that plateaued at ~ 10 min (Figure 1A). Conversely, CGS 21680 (100 nM), a specific A_2aAR agonist, failed to stimulate cAMP, as did adenosine triphosphate (1 mM) (data not shown). Examination of the dose-dependent response to NECA (Figure 1B) and adenosine (Figure 1C) exhibited a low potency consistent with a response mediated by the A_2bAR. Specific inhibition of possible A1AR activation by 50 nM DPCPX did not alter the dose-dependent response to NECA or adenosine (data not shown), whereas 100 nM DCPCX caused a small inhibition of agonist-stimulated cAMP production at submaximal doses of NECA (Figure 1B) or adenosine (Figure 1C). Specific inhibition of potential A3AR activation with 100 nM MRS 1220 had no effect. CPA (an A1/A3 AR selective agonist) inhibited FSK-stimulated cAMP production but with low affinity and efficacy, requiring a concentration of 0.1 to 1 µM to inhibit ~ 15% of the response (Figure 1D). AB-MECA, also an A1/A3 AR selective agonist, displayed a similarly weak dose-dependent inhibition of FSK-stimulated cAMP production.

View larger version (26K): [in this window] [in a new window]   Figure 1.   Characterization of AR responsiveness in human ASM cultures. (A) Time-dependent cAMP accumulation in human ASM. Human ASM cells grown to confluence on 24-well plates were stimulated with either 1 µM ISO (open circles), 100 µM NECA (solid circles), 1 mM adenosine (open squares), or 1 µM CGS 21680 (solid squares). Values for vehicle-stimulated (basal) cAMP accumulation did not differ from those for CGS 21680-stimulated cells (not shown). Data represent mean ± SE of four experiments. (B, C) Human ASM cultures were stimulated with 10 nM to 100 µM NECA (B) or 1 mM adenosine (C) in the presence or absence of 100 nM DPCPX (solid circles) or 100 nM MRS 1220 (solid squares). Data represent mean ± SE of four experiments. (D) Effect of CPA (open circles) or AB-MECA (solid circles) on ISO- and FSK-stimulated cAMP accumulation. Cultures were stimulated with 100 µM FSK ± 0.01 nM to 1 µM CPA or AB-MECA. Data represent mean ± SE of three to four experiments. Absolute FSK-stimulated cAMP accumulation (pmol/well, paired group values): CON, 93.8 ± 11.0 versus +CPA (1 µM), 78.3 ± 8.4; CON, 99.9 ± 10.8 versus +AB-MECA (1 µM), 88.2 ± 6.3. All reactions were terminated after 10 min of stimulation at 37°C. cAMP was isolated and measured by radioimmunoassay as described in MATERIALS AND METHODS. (E) Saturation isotherm of [125I]AB-MECA binding in human ASM membranes. Saturation binding analysis was performed as described in MATERIALS AND METHODS, using 100 µM CPA to establish nonspecific binding.

To further characterize potential A1/A3 ARs in human ASM cultures, radioligand binding analysis using [¹²⁵I]AB-MECA (9) was performed. In saturation binding analyses employing CPA to define nonspecific binding, [¹²⁵I]AB-MECA bound to human ASM membranes derived from three separate cultures with high affinity (dissociation constant [K_d] = 2 ± 1 nM), with calculated B_max values of 33 ± 4 fmol/mg protein (n = 3) (Figure 1E). In a fourth culture, specific [¹²⁵I]AB-MECA binding was not apparent, suggesting variability among cultures in A1/A3 AR expression. In competition binding assays from four cultures, DPCPX inhibited [¹²⁵I]AB-MECA binding (inhibition constant [K_i] = 93 ± 47 nM, n = 4) but typically displaced less than 25% of total binding and suggested a low level of A1AR expression (13 ± 5 fmol/mg protein, defined by specific [¹²⁵I]AB-MECA binding in the presence of 1 µM DPCPX, n = 4); however, in two other cultures, we failed to observe displacement of [¹²⁵I]AB-MECA binding by DPCPX. A total of 1 µM MRS 1220 did not inhibit [¹²⁵I]AB-MECA binding in any cultures (data not shown). Collectively, these data suggest that human ASM cultures express A_2bARs, low levels of A1ARs that test the limits of detection, and possibly additional undetermined AR subtypes.

Acute NECA Treatment Causes Desensitization of the G_s-Coupled Receptors in Human ASM

We next examined the capacity of NECA treatment to evoke an agonist-specific or homologous desensitization of the A_2bAR. A 30-min pretreatment with 100 µM NECA caused a decrease in cAMP accumulation in human ASM cells subsequently challenged with 10 nM to 100 µM NECA (Figure 2A). The cAMP response to 100 µM NECA was reduced 33 ± 3% (Figure 2A, inset; n = 4, P < 0.05). Basal and FSK-stimulated levels of cAMP were not significantly altered by 30 min of pretreatment with 100 µM NECA (data not shown).

View larger version (22K): [in this window] [in a new window]   Figure 2.   NECA induces desensitization of the A2bAR and 2AR in human ASM. Human ASM cultures were pretreated for 30 min with vehicle (0.3% dimethyl sulfoxide; CON [open circles]) or 100 µM NECA (solid circles), washed twice with cold PBS, then rechallenged with 10 nM to 100 µM NECA (A) or 1 nM to 100 µM ISO (B). Both graphs reflect mean ± SE values from three experiments. NECA pretreatment significantly reduced the cAMP production elicited by 100 µM NECA (A inset; n = 4, CON, 14.0 ± 2.1 pmol/well, 30' NECA treated, 9.4 ± 1.6, * P < 0.05) and 1 µM ISO (B inset; CON, 25.5 ± 3.0 pmol/well, 30' NECA treated, 19.6 ± 2.4, n = 4, *P < 0.05).

We have previously demonstrated that ₂AR responsiveness in human ASM is subject to PKA-mediated heterologous desensitization when treated with either a G_s-coupled receptor agonist (PGE₂) or a direct activator of adenylyl cyclase (FSK) (4). To determine whether A_2bAR activation could promote a similar desensitization, human ASM cells were pretreated for 30 min with 100 µM NECA, washed, then rechallenged with 0 to 100 µM ISO. NECA pretreatment shifted the ISO dose-response curve down and to the right, with the maximal response to ISO reduced 23 ± 7% (Figure 2B, inset; n = 5, P < 0.05). Similar effects were observed on the dose-dependent response to PGE₂ (data not shown). Thus, A_2bAR activation is capable of eliciting homologous desensitization of the A_2bAR, as well as heterologous desensitization of other G_s-coupled receptors in human ASM.

Autocrine Adenosine Modulates Receptor-Mediated cAMP Production in Human ASM

Previous studies have reported that adenosine released from vascular smooth muscle cells can function as an autocrine factor and regulate vascular smooth muscle growth via A_2bAR activation (12, 13). Because human ASM muscle expresses functional A_2bARs with a demonstrated capacity to regulate G_s-coupled receptors, we examined whether the potential release of endogenous adenosine from human ASM cells influences G_s-coupled receptor responsiveness. Cultures were pretreated for 18 h with 0.5 U/ml AD and subsequently challenged with vehicle, NECA, ISO, or FSK. Vehicle-stimulated (basal) cAMP accumulation was significantly reduced 25 ± 7% (control [CON] 0.95 ± 0.13 pmol/well versus AD-treated 0.72 ± 0.11 pmol/well, n = 10, P < 0.05) by AD pretreatment, whereas FSK-stimulated cAMP production was not significantly altered (data not shown). AD pretreatment caused a significant increase in A_2bAR responsiveness (Figure 3A); cAMP accumulation in response to 100 µM NECA increased 20 ± 6% (Figure 3A, inset; n = 10, P < 0.05). AD treatment also had a small but significant effect on ₂AR responsiveness (Figure 3B). The dose-dependent response to ISO was shifted up and to the left, with maximal cAMP production elicited by 1 µM ISO increased by 15 ± 7% (Figure 3B, inset; n = 10, P < 0.05). Similar effects of AD treatment were observed on the dose-dependent effect of PGE₂ on cAMP production (data not shown).

View larger version (23K): [in this window] [in a new window]   Figure 3.   AD treatment increases NECA- and ISO-stimulated cAMP production. Human ASM cultures were pretreated for 18 h with vehicle (CON; open circles) or 0.5 U/ml AD (solid circles), washed twice with cold PBS, then rechallenged with 10 nM to 100 µM NECA (A) or 1 nM to 10 µM ISO (B). Both graphs reflect mean ± SE values from three experiments. AD pretreatment significantly increased the maximal cAMP production elicited by 100 µM NECA (A inset; CON, 15.0 ± 1.9 pmol/well, AD treated, 17.8 ± 2.2, n = 10, * P < 0.05) and 1 µM ISO (B inset; CON, 27.8 ± 1.4 pmol/well, AD treated, 33.0 ± 2.2, n = 10, * P < 0.05).

To examine whether the effects of AD treatment were related to receptor regulation per se and not to alterations in substrate/product disposition somehow effected by AD, the responsiveness of receptor, G protein, and adenylyl cyclase was examined in an in vitro assay using cell homogenates (Figure 4). In these experiments, the effects of AD treatment on ISO, NECA, and PGE₂ -stimulated adenylyl cyclase activity were actually more pronounced (~ 40% increase for each agent) than those observed on agonist-stimulated cAMP accumulation in intact cells (Figure 3). Minimal changes were observed in NaF- and FSK-stimulated adenylyl cyclase activity. An 18-h treatment with XAC, a nonspecific inhibitor of ARs, caused a similar effect to that of AD, although the response to NECA was curiously unaltered. Similar effects of XAC pretreatment on receptor and FSK responsiveness were observed in intact cell assays (data not shown). The failure of XAC pretreatment to alter the cAMP response to NECA suggests either retained antagonist (despite extensive washing) or a failure of 300 nM XAC to fully suppress mechanisms of homologous A_2bAR desensitization under chronic conditions (see subsequent text). An 18-h pretreatment with 50 nM DPCPX affected neither adenylyl cyclase activity in cell homogenates nor agonist-stimulated cAMP accumulation in intact cell assays (data not shown). These data of both intact cell cAMP production and cell-free adenylyl cyclase activity strongly suggest that AD pretreatment effects on receptor-mediated cAMP production are a result of changes in receptor-G protein coupling and are not related to alterations in downstream effectors or in the intracellular nucleoside/nucleotide pool.

View larger version (34K): [in this window] [in a new window]   Figure 4.   Effects of AD on adenylyl cyclase activity in ASM homogenates. Human ASM cultures grown on 15-cm plates were treated for 18 h with vehicle (CON; open bars), 0.5 U/ml AD (solid bars), or 300 nM XAC (bars with white dots). Cells were then washed with cold PBS and harvested by scraping into 10 ml of ice-cold PBS. Adenylyl cyclase activity was subsequently measured in cell homogenates using a competitive protein binding assay and [8-3H]cAMP as described previously (8). Data represent mean ± SE values from three to four observations. Statisically significant differences (* P < 0.05) conferred by treatment effects (AD or XAC) for each stimulus were analyzed by ANOVA and Fisher's PLSD multiple comparisons test.

Collectively, these data suggest that human ASM releases a low level of endogenous adenosine (or nucleotides that are converted to adenosine by ectonucleotidases) which activates human ASM A_2bARs and causes a small but discernable homologous A_2bAR desensitization as well as heterologous desensitization of other G_s-coupled receptors. To test for the presence of adenosine in the medium of human ASM cultures, we used COS-1 cells in a bioassay of AR activation. As demonstrated in Figure 5, exogenous adenosine produces a potent, dose-dependent effect on cAMP generation in COS-1 cells, suggestive of an A_2aAR-mediated response. Conditioned human ASM medium stimulated cAMP generation (4.8 ± 0.2 pmol/well, n = 6), which was inhibited by inclusion of AD (to basal levels of ~ 2.4 pmol/well) in the medium. Inclusion of 100 nM XAC (two experiments, data not shown) had an identical effect to that of AD. By comparison with the response of COS-1 cells to exogenous adenosine, these data suggest the presence of low nanomolar concentrations of adenosine in human ASM-conditioned medium.

View larger version (20K): [in this window] [in a new window]   Figure 5.   Adenosine is present in media conditioned by human ASM. COS-1 cells were used in a bioassay testing for the presence of adenosine in culture media conditioned by human ASM. Confluent cultures of human ASM grown in 15-cm plates were washed with PBS then fed 30 ml Ham's F12 medium supplemented with IT medium for 24 h. COS-1 cells grown to confluence in 24-well plates were washed with PBS then stimulated for 10 min at 37°C with either human ASM-conditioned medium (CM) or IT medium containing 0 to 1 mM adenosine, each containing 1 mM RO-20-1724 and the presence or absence of 0.5 U/ml AD (solid bars). cAMP was subsequently isolated and measured as described in MATERIALS AND METHODS. Data represent mean ± SE values from six observations (CON, open bars).

Functional Consequences of Manipulating Adenosine Transport/Conversion in Human ASM

Previous studies have reported that agents that inhibit adenosine transport, adenosine kinase, or AD in vascular smooth muscle cells increase the accumulation of adenosine in the culture media, such that mitogen-stimulated growth can be inhibited by associated A_2bAR activation (12, 14). We therefore tested whether chronic treatment of human ASM with inhibitors of adenosine transport (DIP and NBTI), adenosine kinase (5-IODO), adenosine deaminase (EHNA), or a combination of EHNA+5-IODO+ NBTI could effect an A_2bAR-mediated desensitization of GPCRs similar to that depicted in Figure 2. An 18-h treatment with each of the agents resulted in minimal changes in the absolute cAMP accumulation elicited by NECA (data not shown). However, pretreatment with a combination of NBTI+EHNA+5-IODO caused a significant loss of NECA-stimulated cAMP (26 ± 7%, n = 5, P < 0.05), suggesting that human ASM employs at least three mechanisms (adenosine transport, phosphorylation, and deamination) to regulate extracellular adenosine levels (Figure 6A). Chronic treatment with CPA had no effect on NECA- stimulated cAMP, whereas chronic treatment with either low (3 µM) (Figure 6A) or high (30 µM) (data not shown) concentrations of NECA eliminated A_2bAR responsiveness. Interestingly, XAC at a concentration as high as 1 µM had no effect on 30 µM NECA pretreatment and recovered only 49% of the CON NECA response after a 3-µM NECA pretreatment (Figure 6A), despite its capacity to inhibit acute (30 µM) NECA-stimulated cAMP production with an IC₅₀ of ~ 20 nM (data not shown). These data suggest that the effects induced by chronic GPCR activation can occur at low levels of receptor occupancy and have lower EC₅₀ values than those associated with effects elicited by acute activation of the receptor. Consequently, higher levels (than those suggested by reported K_i values) of antagonists appear required to inhibit such chronic effects.

View larger version (44K): [in this window] [in a new window]   Figure 6.   Effects of chronic AR activation on receptor, adenylyl cyclase responsiveness. (A, B) Cultures were pretreated with vehicle (CON; open bars), 100 µM DPCPX (bars with black dots), or 1 µM XAC (bars with white dots) for 30 min, or 100 ng/ ml pertussis toxin (PTX; solid bars) 8 h before an 18-h treatment with vehicle (CON), 0.1 µM CPA (CPA-7M), a combination of 0.1 µM NBTI, 1 µM EHNA, and 0.1 µM 5-IODO (N+E+5), or 3 µM NECA. Cells were washed twice with cold PBS and stimulated with 100 µM NECA (A) or 100 µM FSK (B), and cAMP isolated and quantified by radioimmunoassay as described in MATERIALS AND METHODS. Data represent mean ± SE from four to six paired observations. * P < 0.05, vehicle pretreated CON versus (vehicle-pretreated) N+E+5 and NECA groups (NECA-stimulated cAMP production); vehicle pretreated CON versus (vehicle-pretreated) CPA-7M, N+E+5, and NECA groups (FSK-stimulated cAMP production). Absolute values of cAMP production (pmol/well) among groups (paired group comparisons, no prior DPCPX, XAC, or PTX treatment): NECA stimulation: CON, 15.1 ± 1.4 versus CPA-7M treated, 15.0 ± 1.5; CON, 17.1 ± 1.8 versus N+E+5 treated, 12.5 ± 0.7; CON, 15.1 ± 1.8 versus NECA 3 µM treated, 1.5 ± 0.4; FSK stimulation: CON, 81.1 ± 11.2 versus CPA-7M treated, 105.8 ± 13.0; CON, 80.3 ± 8.4 versus N+E+5 treated, 100.6 ± 9.8; CON, 68.6 ± 5.7 versus NECA 3 µM treated, 91.4 ± 12.0.

Although absolute ISO-stimulated cAMP production was essentially unchanged after DIP, NBTI, EHNA, 5-IODO, or combined pretreatment (data not shown), FSK-stimulated cAMP production was unexpectedly increased by each of the pretreatments. DIP, NBTI, EHNA, or 5-IODO each caused an ~ 20 to 30% increase in FSK-stimulated cAMP (data not shown), whereas combined EHNA+ 5-IODO+ NBTI pretreatment resulted in a 35 ± 8% increase (n = 5, P < 0.05) (Figure 6B). This effect was surprising in that chronic activation of G_s-coupled receptors (e.g., with ISO or PGE₂) in human ASM cultures typically results in a slight loss of FSK-stimulated cAMP production ([15] and R. B. Penn, unpublished observations). Because we have previously determined that chronic activation of G_i-coupled receptors in human ASM results in increased responsiveness to FSK (i.e., adenylyl cyclase sensitization) (8), these results were compared with those obtained after chronic pretreatment with NECA or CPA. Interestingly, both chronic, nonspecific AR activation (by NECA) and specific A1/A3 AR activation (by CPA) resulted in adenylyl cyclase sensitization (Figure 6B). These effects, as well as those elicited by combined NBTI+EHNA+5-IODO treatment, were completely reversed by pertussis toxin and, to a lesser degree, by 100 nM DPCPX pretreatment, implicating the A1AR in contributing to this sensitization elicited by exogenous and autocrine AR agonists. Inhibition of PKC by pretreatment with either bisindolylmaleimide I or IX had no affect on the adenylyl cyclase sensitization elicited by any agent (data not shown). Importantly, because adenylyl cyclase responsiveness appears to limit cAMP production by G_s-coupled receptors in human ASM (8), this sensitization of adenylyl cyclase suggests a greater loss of A_2bAR and ₂AR responsiveness than that suggested by analysis of absolute cAMP production elicited by NECA and ISO, respectively.

Effects of Heterologously Expressed A1AR and A_2bAR

To assess the effect of increased A1AR or A_2bAR expression in human ASM cultures, constructs encoding GFP chimeras of the human A1AR and rat A_2bAR were generated and transfected into human ASM cultures as described in MATERIALS AND METHODS. This strategy was employed to enable enrichment of cells expressing the construct of interest by cell sorting of GFP-positive cells. Transfected cells were harvested and the fluorescence profile analyzed (Figure 7). Cells exhibiting fluorescence greater than that associated with mock-transfected cells were sorted and subsequently plated into 48-well plates. After growth arrest, cells were pretreated with or without AD for 18 h then stimulated as indicated in Figure 8. Overexpression of A1AR-GFP did not significantly affect basal or NECA-stimulated cAMP production in comparison with GFP-expressing cells. However, FSK-stimulated cAMP was significantly reduced compared to CON GFP values (34 ± 8% loss, n = 4, P < 0.05), and the inhibitory effect of CPA was augmented (56 ± 5% loss compared to CON GFP values, n = 4, P < 0.05). Overexpression of A_2bAR-GFP caused a large increase in basal (693 ± 217% of CON GFP values, n = 3, P < 0.05) and NECA-stimulated (370 ± 83%, n = 3, P < 0.05) cAMP levels. Interestingly, FSK-stimulated cAMP production was also significantly increased (193 ± 38% of CON GFP values, n = 3, P < 0.05), most likely a result of a high level of steady-state G_s-adenylyl cyclase interaction. Pretreatment with AD reduced basal levels in A_2bAR-GFP-expressing cells but did not alter responses to NECA or FSK. These data suggest that modulation of AR expression in ASM cells can impact basal as well as agonist-dependent regulation of adenylyl cyclase.

View larger version (54K): [in this window] [in a new window]   Figure 7.   Distribution of cellular fluorescence in human ASM cells transfected with GFP or A1AR- or A2bAR-GFP chimeras. Human ASM cultures plated on 15-cm plates were subject to mock transfection (no DNA; A) or transfection with pEGFPN1 (GFP; B), pA1AR-GFP (A1AR-GFP; C), or pA2bAR-GFP (A2bAR-GFP; D) as described in MATERIALS AND METHODS. Populations of cells expressing (circumscribed by line B in each panel ) or lacking (line D) fluorescence indicative of GFP were sorted to > 98% purity using a Coulter Epics Elite ESP Flow Cytometer and plated into 48-well plates for subsequent analysis.

View larger version (17K): [in this window] [in a new window]   Figure 8.   Overexpression of A1AR- and A2bAR-GFP chimeras in human ASM alters basal and agonist-stimulated cAMP. Homogeneous populations of human ASM cells expressing either GFP, A1AR-GFP, or A2bAR-GFP were plated in 48-well plates, growth-arrested, and stimulated with (A) vehicle (BASAL), (B) 100 µM NECA, or (C) 100 µM FSK ± 100 nM CPA (solid bars) for 10 min at 37°C. BASAL and NECA-stimulated cells were pretreated with either vehicle (open bars) or 0.5 U/ml AD (solid bars) 18 h before stimulation. cAMP was subsequently isolated and measured as described in MATERIALS AND METHODS. A1AR-GFP expression was confirmed by specific [3H]DPCPX binding in two separate experiments measuring 1.78 and 1.36 pmol/mg whole cell protein. Group cAMP values were normalized to CON GFP values that were: BASAL, 0.10 ± 0.02 pmol cAMP/µg whole cell protein; NECA, 1.16+ 0.21 pmol/µg protein; FSK, 8.20 ± 2.10 pmol/µg protein. *P < 0.05 denotes a statistically significant difference between experimental and CON GFP values.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study finds that the G_s-coupled A_2bAR is the predominate AR subtype mediating acute responses to adenosine in human ASM cultures. A_2bAR activation elicited a rapid increase in cAMP production and promoted both homologous desensitization of the A_2bAR and heterologous desensitization of ₂AR and PGE₂ receptors. In addition, quiescent human ASM cells were found to be a source of extracellular adenosine, and this autocrine adenosine was shown to regulate both basal cAMP levels and GPCR responsiveness, presumably via a low level of A_2bAR activation. By increasing the accumulation of extracellular adenosine through a combined inhibition of adenosine transport, phosphorylation, and deamination, a significant pertussis toxin-sensitive sensitization of adenylyl cyclase was observed. This effect was mimicked by chronic treatment with both nonspecific and specific A1/A3 AR agonists and was partially reversed by the A1AR-specific antagonist DPCPX. These data suggest that effects of chronic adenosine exposure on human ASM are mediated by combined activation of A_2b and G_i-coupled AR subtypes.

The family of known ARs includes A1, A_2a, and A_2b, and the A3 ARs. The G_s-coupled A_2aARs and A_2bARs are distinguished by their high and low affinity, respectively, for adenosine. Recently, A_2bARs have also been shown to mediate Ca²⁺ influx via activation of G_q (16, 17). The A1AR is coupled to G_i, inhibits adenylyl cyclase, and is capable of activating phospholipase C through subunits. The A3AR is coupled to both G_i and G_q, and has been shown to inhibit adenylyl cyclase, stimulate phosphoinositide accumulation, and increase intracellular calcium. To date, differentiation and analysis of AR subtypes in cells and tissues have been hindered by the lack of subtype-selective agonists and antagonists. Moreover, a recently appreciated difference in the pharmacologic properties of the AR subtypes among species has further complicated this field of study (for a current review of AR pharmacology, see Olah and Stiles [18]).

In the human ASM cultures used in the present study, we observed a clear activation of adenylyl cyclase by the nonspecific AR agonist NECA that was unaltered by inclusion of A1AR- or A3AR-specific antagonists. The A_2aAR-specific agonist CGS 21680 failed to stimulate cAMP production, and the dose-response response to NECA exhibits a low potency of NECA in stimulating cAMP production, suggesting human ASM cultures predominantly express A_2bARs. Thus, (1) acute production of cAMP upon stimulation with NECA; (2) desensitization of NECA, ISO, and PGE₂ activation of adenylyl cyclase after 30 min NECA pretreatment; and (3) effect of AD pretreatment in increasing receptor responsiveness appear to be mediated through the A_2bARs. We have previously demonstrated that heterologous desensitization of the ₂AR by either G_s-coupled receptor activation or direct adenylyl cyclase activation was PKA-dependent (4). Although the possibility exists that an A_2bAR-mediated Ca²⁺ signal may play some role in our results, preliminary studies from our lab have yet to demonstrate a Ca²⁺ flux in human ASM cells upon stimulation with NECA (data not shown).

Numerous studies have identified adenosine as an important regulator of airway cell and lung function, although the specific receptor subtypes mediating these responses among various cell types are not firmly established (17, 19). In both asthmatic subjects and animal models of asthma, activation of ARs (possibly A1AR, A_2bAR, or A3AR, dependent on species and phenotype [20]) on mast cells is known to cause histamine release, which in turn promotes ASM contraction. Inhibition of ARs on inflammatory cells is believed to be a contributory mechanism mediating the therapeutic benefits of theophylline and caffeine. However, a more recent study by Nyce and Metzger (3) also proposes A1ARs in ASM as important in adenosine-mediated bronchoconstriction. In a dust mite-conditioned allergic rabbit model of asthma, sensitized rabbits were administered aerolized antisense oligodeoxynucleotides targeting the A1AR, which significantly inhibited brochonconstriction induced by challenge with adenosine or dust-mite allergen. In addition, ASM dissected from antisense-treated, sensitized animals displayed a significant reduction in A1AR expression.

Although the majority of studies examining whole lung challenge or isolated ASM strips have attributed the bronchoconstrictive effect of adenosine to secondary release of histamine or leukotrienes from inflammatory cells, the identification of A1ARs in rabbit ASM is intriguing. Whether A1AR or A3AR expression levels in human ASM are regulated as a function of the inflammatory or asthmatic state is unknown, as is their potential functional impact. In an analysis of ARs in human airways and peripheral lung, Joad and Kott (21) found no evidence of A1ARs in human airways from apparently nonasthmatic subjects. The majority of studies to date have reported little or no capacity of adenosine to contract ASM strips harvested from nonasthmatic subjects or unsensitized animals. Examining human bronchi from asthmatic subjects, Bjorck and colleagues (22) demonstrated an A1AR-mediated contraction that was blocked by a combination of leukotriene and histamine antagonists, suggesting that inflammatory cells imbedded in ASM strips from asthmatic subjects are the primary targets of adenosine. However, Ali and associates (23) suggest that bronchial hyperresponsiveness to adenosine in an allergic rabbit model of asthma can be explained by an upregulation of A1ARs in ASM strips that were devoid of mast cells upon histologic observation. In addition, Abebe and Mustafa (24) recently characterized an A1AR-mediated inositol 1,4,5 trisphosphate generation in these same rabbits, providing further evidence of A1AR induction. A similar upregulation of A3ARs may occur in sensitized guinea pig ASM, where specific A3AR activation mediates contraction (25), and in rat ASM, where A3ARs can potentiate Ca²⁺ release stimulated by procontractile agonists (26). Whether asthma-associated A1/A3 AR expression is species-specific or exists in discrete subpopulations of asthmatics is unknown.

Our initial studies failed to provide clear evidence of A1/A3 AR activation upon acute exposure of human ASM cells to AR ligands. Specific A1AR or A3AR inhibition affected neither adenosine- nor NECA-stimulated cAMP production, and neither NECA nor CPA elicited a Ca²⁺ transient or significant phosphoinositide production (data not shown). However, a small inhibition of FSK-stimulated cAMP production by CPA and AB-MECA was observed, and low levels of specific [¹²⁵I]AB-MECA binding to human ASM membranes were detected using the A1AR-specific antagonist DPCPX, suggesting a low level of A1AR expression.

More substantial evidence of functional G_i-coupled ARs was provided in analyses of the chronic effects of AR ligands or increased extracellular accumulation of endogenously derived adenosine. Chronic treatment of cultures with CPA, NECA, or inhibitors of adenosine deaminase, kinase, and transport caused a sensitization of adenylyl cyclase similar to that previously shown to be induced by chronic G_i-coupled receptor activation (8). This sensitization was inhibited by pertussis toxin and partially reversed by DPCPX. The failure of DPCPX to fully inhibit sensitization suggests that other (unidentified) G_i-coupled AR subtypes, in addition to the A1AR, may also be involved. Adenylyl cyclase sensitization occurred against a backdrop of significant A_2bAR and, to a lesser extent, ₂AR desensitization. Thus, G_s-coupled (A_2b) as well as G_i-coupled (A1 and possibly others) AR subtypes appear to contribute to the chronic effects of adenosine in human ASM cultures.

Physiologic Significance of ARs in Human ASM

The identification of A_2bARs and regulatory potential of endogenously derived adenosine in human ASM cultures add an additional layer of complexity toward understanding the manner in which adenosine regulates ASM function in vivo. Moreover, our study suggests that altered AR subtype expression can impact basal and agonist-stimulated adenylyl cyclase activity in human ASM. Although studies to date have failed to identify A1AR or A3AR subtypes in human ASM in vivo, such identification may depend on the development of subtype-selective radioligands with high specific activity, particularly if expression levels are as low as those suggested in human ASM cultures. Direct effects of adenosine may depend on the distribution of AR subtypes in ASM in vivo, the local level of adenosine generated from paracrine (inflammatory cells) and autocrine sources, and the presence or absence of therapeutic agents such as theophylline, caffeine, or beta-agonists. A_2bARs on ASM could subserve relaxation through activation of PKA, and the net effect on the ASM contractile state could be determined by the balance between A_2bARs and (procontractile) A1ARs or A3ARs. However, A_2bAR activation may diminish the bronchorelaxant effect of beta-agonists by promoting ₂AR desensitization. This effect may be modulated by methylxanthine inhibition of ARs.

In addition, the cellular stress imposed on ASM in vivo during airway hypoxia or contraction would likely increase ASM production of adenosine and magnify any of the autocrine effects on receptor-mediated processes observed in quiescent cultures. Under such circumstances the A_2bARs may respond to the "retaliatory metabolite" (18) adenosine and promote, via PKA activation, a negative feedback mechanism. However, under conditions where G_i-coupled ARs might be upregulated and dominate the functional response to adenosine, this phenomenon may be supplanted by a positive feedback mechanism. Furthermore, adenosine released from ASM may exert a paracrine influence on local inflammatory cells, thus perpetuating a cycle of procontractile events.

Future studies analyzing the distribution of AR subtype expression in ASM among nonasthmatic and asthmatic populations, and the mechanisms that regulate nucleoside/ nucleotide release in ASM will help to establish the relative role of AR subtypes and the effects of autocrine adenosine in modulating GPCR responsiveness and contractile state in ASM.