Abstract
Asthmatics, unlike healthy subjects, experience bronchoconstriction in response to inhaled adenosine, and extracellular adenosine concentrations are elevated in the bronchoalveolar lavage fluid and exhaled breath condensate of asthmatic subjects. However, little is known about the location and expression of adenosine receptors in asthmatic airways. The aim of the present study was to investigate the distribution of adenosine A1 receptors in bronchial biopsy specimens from mildly asthmatic steroid-naïve subjects and then compare the degree of expression with that of healthy subjects.
Biopsy sections were immunostained using an adenosine A1 receptor antibody, the selectivity of which was validated in specific experiments. Image analysis was then performed in order to determine differences in immunostaining intensity.
Immunostaining of biopsy sections from the asthmatic subjects revealed strong expression of the A1 receptor, located predominantly in the bronchial epithelium and bronchial smooth muscle. In comparison, very weak immunostaining was observed in biopsy specimens obtained from healthy subjects. Image analysis revealed that the intensity of positive staining of the asthmatic bronchial epithelium and smooth muscle regions was significantly greater than that observed for the healthy epithelium and smooth muscle.
In conclusion, the sensitivity of asthmatics to inhaled adenosine coupled with increased adenosine A1 receptor expression implies that these receptors play a role in the pathophysiology of this disease.
Adenosine is a purine nucleoside that is expressed in all cells of the body and involved in a wide range of physiological processes. The effects of adenosine are mediated predominantly through specific cell surface receptors, of which four subtypes (A1, A2A, A2B and A3) have been described. It is now well recognised that extracellular levels of adenosine markedly increase under metabolically stressful conditions, such as hypoxia and inflammation, and, although an acutely elevated level of extracellular adenosine is considered to mediate anti-inflammatory and protective effects, chronic accumulation has been associated with pathological consequences 1.
In asthmatic subjects, it has been demonstrated that adenosine levels in bronchoalveolar lavage fluid and exhaled breath condensate are significantly higher than those occurring in healthy subjects 2, 3, and current evidence strongly suggests that they may contribute to the pathophysiology of asthma. For example, it has been recognised since the mid-1980s that inhalation of adenosine 5′-monophosphate (AMP; 5’-nucleotidase rapidly hydrolyses AMP to adenosine in the lung) in asthmatic but not healthy subjects results in dose-related bronchoconstriction 4. This is considered to be mediated predominantly, but not exclusively, by mast cell degranulation via A2B receptor activation (reviewed in 5). Furthermore, inhalation of AMP has been shown to increase airway eosinophilia 6.
A role for adenosine in asthma is further supported by the observation that plasma adenosine levels increase rapidly following allergen challenge in asthmatic subjects 7, suggesting a possible involvement of endogenous adenosine in the early-phase airway response to allergen. This possibility was supported by the blockade of allergen-induced bronchoconstriction through the use of an antisense oligonucleotide directed against the A1 receptor 8 and, in 2005, of a selective adenosine A1 receptor antagonist 9 in allergic rabbits. Furthermore, it has been demonstrated that the concentrations of adenosine are increased in both plasma and exhaled breath condensate during exercise-induced bronchoconstriction in subjects with asthma 10, 11. Finally, it has also been shown that isolated human asthmatic bronchial tissue contracts in response to adenosine via an A1-receptor-dependent mechanism 12.
In order to further investigate the role of endogenous adenosine in asthma, it is crucial that current knowledge concerning the distribution and specific functions of adenosine receptors in asthmatic airways is extended. To date, binding studies in healthy peripheral lung tissue have suggested that A2 receptor subtypes are more abundant than the A1 and A3 receptor subtypes 13, and expression of the A3 receptor was observed to be upregulated in asthmatic lung tissue, where it was located predominantly to eosinophils 14. Therefore, the aim of the present study was to extend these observations. Although the ultimate intention is to profile the expression of all four adenosine receptor subtypes and identify the subtype-specific cellular functions, it was hypothesised that characterisation of the distribution and expression of the A1 receptor would particularly further understanding of the role of adenosine in asthma given the observations detailed above. Thus the distribution of the adenosine A1 receptor in bronchial biopsy specimens from asthmatic subjects was investigated and then compared to the distribution and level of expression of biopsy specimens obtained from healthy subjects in order to provide further insight into the role of adenosine in asthma.
METHODS
Further details of all methods and procedures can be found in the supplementary material.
Subjects
All of the asthmatic volunteers provided a clinical history consistent with intermittent reversible airflow obstruction and atopy to common aeroallergens. Thus all of the asthmatic subjects (nonsmokers) were characterised on the basis of a previous history of wheeze, a forced expiratory volume in one second (FEV1) of >70% of the predicted value, atopy as defined by a positive cutaneous response to intradermal allergen, and airways hyperresponsiveness to both methacholine and AMP. In addition, all asthmatic subjects demonstrated an increase in baseline FEV1 of ≥15–20% in response to inhaled salbutamol (data not shown). Of the 16 mildly asthmatic subjects screened for the present study, 12 went on to complete bronchial provocation challenge with AMP and bronchoscopy (four subjects withdrew their consent to undergo bronchoscopy). Each subject exhibited infrequent symptoms, which were controlled using occasional inhaled short-acting β2-agonists alone. No subject was taking any regular anti-asthma therapy, and none had taken inhaled corticosteroids for ≥3 months prior to entry. None of the subjects had had an exacerbation of asthma or respiratory infection during the preceding 6 weeks, and all subjects were nonsmokers. Patient demographics are summarised in table 1⇓. For comparison, biopsy specimens from eight healthy age-matched nonsmoking subjects obtained under the same conditions in a separate but recent previous study were also analysed. The demographics of these subjects are also described in table 1⇓. Written informed consent was obtained from each subject, and the study was approved by the Ethics Committee of King’s College Hospital (London, UK).
Patient demographics
Study design
Visit 1 procedures for all subjects included medical history taking and physical examination, an asthma characterisation questionnaire, intradermal skin-prick tests, spirometry and methacholine challenge to determine the provocative concentration of agent causing a 20% fall in FEV1 (PC20). After ∼1 week, asthmatic subjects returned for visit 2, when they underwent AMP challenge in order to determine the PC20 to AMP. All subjects refrained from using rescue medication and caffeinated beverages for ≥8 h before each visit. After a period of 3–4 weeks, all subjects returned for visit 3, during which bronchoscopy was performed.
Immunohistochemistry
Using the methods described below to validate the adenosine receptor subtype specificity of an antibody, it was found that the specificity of commercially available antibodies against all adenosine receptor subtypes was unsatisfactory; hence an antibody directed against the adenosine A1 receptor was created specifically for the present study (the cost involved in obtaining a custom-made antibody was also a contributory factor in the decision not to concurrently profile the expression of the other adenosine receptor subtypes). Thus the custom-made affinity-purified A1 receptor polyclonal antibody was raised in the rabbit against a synthetic peptide corresponding to amino acids 309–326 (CQPAPPIDEDLPEERPDD) of the carboxy terminus of the human A1 receptor (Cambridge Research Biochemicals, Billingham, UK).
Brain cortex and cardiac tissue sections from three subjects were co-analysed as a positive control, since it is well established that these tissues highly express A1 receptors. Cadaveric brain cortex and cardiac tissue samples were obtained from the Institute of Psychiatry Brain Bank (London, UK) and Peterborough Tissue Bank (Peterborough, UK), respectively, following GlaxoSmithKline (Stevenage, UK) and UK (Human Tissue Authority, London, UK) guidelines for the acquisition and use of human tissues, including ethical approval and the use of appropriate consent forms.
Frozen biopsy specimens of brain cortex, cardiac tissue and bronchial tissue were sectioned at a thickness of 6 μm. Sections were placed on VectabondTM-coated slides (Vector Laboratories, Peterborough, UK), and the brain cortex and cardiac tissue sections fixed in ice-cold 4% weight (w)/volume (v) paraformaldehyde for 5 min (bronchial biopsy sections were fixed with 4% w/v paraformaldehyde prior to cryopreservation). Sections were then immunostained for the A1 receptor using an Optimax automatic immunostaining machine (Biogenex, San Ramon, CA, USA). Positive staining was detected with diaminobenzidine in addition to counterstaining with Mayer’s haematoxylin.
Image analysis
Fully automated densitometry of A1 receptor expression was developed using the Zeisss Vision KS400 system (Carl Zeiss, Göttingen, Germany). Stain intensity was described using arbitrary units (AU). All image analysis and measurements were performed blind by one observer.
Antibody validation with adenosine-receptor-transfected cells
Further antibody validation experiments were performed through flow cytometric analysis of A1 receptor antibody binding to each adenosine receptor subtype expressed on CHO-K1 cells (which lack any known subtype of adenosine receptor) transfected with recombinant human A1, A2A, A2B or A3 receptor. Comparative expression levels of the adenosine receptor subtypes in each cell line were first investigated using real-time RT-PCR analysis. Data are expressed as gene copy number per 50 ng of complementary DNA. For flow cytometric analysis, cell lines were analysed, following antibody labelling, by appropriate gating for immunofluorescence using a FACSCalibur flow cytometer (Beckton Dickinson, Oxford, UK) after excitation at 488 nm. At least 3,000 events were collected and mean fluorescence intensities recorded.
Statistics
Data were analysed using an unpaired t-test, and, unless otherwise stated, are presented as mean±sem.
RESULTS
Subjects
Patient demographics are described in table 1⇑. All asthmatic subjects who completed the present study were hyperresponsive to AMP and methacholine (table 1⇑). The PC20 to methacholine in the eight nonatopic healthy subjects were all >16 mg·mL−1. All asthmatic subjects exhibited a positive skin response to at least one allergen, whereas none of the healthy subjects responded to any of the allergens tested.
Immunohistochemistry
Positive immunostaining appeared brown against a blue background as a result of the Mayer’s haematoxylin counterstain. Immunostaining of bronchial biopsy specimens obtained from asthmatic subjects with the A1 receptor antibody consistently revealed strong positive staining on the bronchial epithelium and moderate-to-strong positive staining on the bronchial smooth muscle, these distinctive regions were identified morphologically (fig. 1⇓). The immunostaining appeared to be distributed uniformly throughout the whole of these regions, rather than localised to specific areas or structures within these regions. By comparison, immunostaining of bronchial biopsy specimens obtained from healthy subjects with the A1 receptor antibody was strikingly less intense, with weak positive staining of the epithelium and very weak or sometimes undetectable positive staining of the smooth muscle (fig. 2⇓). In all asthmatic and healthy bronchial biopsy sections, a negligible level of nonspecific staining was consistently observed with the isotype control antibody (data not shown).
Representative photomicrographs showing positive immunostaining of asthmatic bronchial biopsy sections with an A1 receptor antibody (a), with high expression of the A1 receptor on the epithelium (b) and smooth muscle (c), shown at higher magnification. Positive immunostaining appears brown against a blue background as a result of the Mayer’s haematoxylin counterstain. Scale bars = 25 μm.
Representative photomicrographs showing positive immunostaining of healthy bronchial biopsy sections with an A1 receptor antibody (a), with weak expression of the A1 receptor on the epithelium (b) and virtually no positive immunostaining of smooth muscle (c), shown at higher magnification. Positive immunostaining appears brown against a blue background as a result of the Mayer’s haematoxylin counterstain; scale bars = 25 μm.
Sections obtained from only 11 asthmatic subjects and seven healthy subjects were deemed of suitable quality for image analysis, since sections from one subject in each study group lacked sufficient epithelial and smooth muscle area for the analysis. Quantification of the staining intensity of the epithelium revealed that the level of positive staining of the asthmatic epithelium was significantly greater than that observed with the healthy epithelium (175.2±5.3 versus 78.9±30.9 AU; p<0.01). Similarly, the intensity of immunostaining of the asthmatic bronchial smooth muscle was significantly greater than that observed with the healthy bronchial smooth muscle (132.3±9.9 versus 58.1±24.4 AU; p<0.01; fig. 3⇓).
Image analysis of asthmatic (□) and healthy (▓) bronchial biopsy sections immunostained with an A1 receptor antibody. The intensity of positive staining specifically on the epithelium and smooth muscle areas of bronchial biopsy sections from 11 asthmatic subjects and seven healthy subjects (one section per subject) was quantified using Zeiss Vision KS400 software, from four different views of both epithelium and smooth muscle at a magnification of 1,000×. Data are presented as mean±sem. **: p<0.01 versus healthy subjects.
Submucosal glands were identified on sections from two asthmatic bronchial biopsy specimens and one healthy biopsy specimen. Again, A1 receptor expression appeared to be greater in the asthmatic biopsy specimens (fig. 4a⇓) compared with the healthy biopsy specimen (fig. 4b⇓).
Representative photomicrographs showing immunostaining of a) asthmatic, and b) healthy bronchial biopsy sections with an A1 receptor antibody. Positive immunostaining appears brown against a blue background as a result of the Mayer’s haematoxylin counterstain. There was high expression of the A1 receptor on asthmatic submucosal glands, but no apparent positive immunostaining of healthy submucosal glands. Scale bars = 25 μm.
Both the brain cortex and cardiac tissue samples from all three subjects immunostained very strongly for the A1 receptor (fig. 5a⇓ and c, respectively), with no detectable staining observed with the isotype control antibody (fig. 5b⇓ and d). The small degree of brown colouration observed in the cardiac tissue (fig. 5d⇓) was considered to be due to the presence of lipofuscin pigments.
Photomicrographs showing immunostaining of a, b) brain cortex, and c, d) cardiac tissue with the A1 antibody (a, c) and an isotype control antibody (b, d). Positive immunostaining appears brown against a blue background as a result of the Mayer’s haematoxylin counterstain. There was positive staining of brain cortex and cardiac tissue with the A1 antibody, but negligible staining of sections from the same samples using the isotype control. Scale bars = 25 μm.
A1 receptor antibody validation
In order to provide additional support for the observations resulting from the immunohistochemical analysis described above, antibody validation experiments were also performed in order to analyse the affinity and selectivity of the A1 receptor antibody for the A1 receptor.
Quantitative RT-PCR analysis of the CHO-K1 cell lines transfected with each of the four adenosine receptor subtypes revealed that each of the cell lines expressed very high transcript levels of the transfected receptor, and that the gene expression levels of each receptor were generally of a similar magnitude (fig. 6⇓), although slightly lower in the A3 cell line. Subsequent flow cytometric analysis of the CHO-K1 cell lines using the A1 receptor antibody showed a significant increase in mean fluorescence intensity compared with wild-type CHO-K1 cells in the A1 cell line alone (478±84 versus 35±8 AU; p<0.01). There were no significant differences between any of the other cell lines and the wild-type cells (fig. 6⇓).
a) Quantitative real-time RT-PCR, and b) flow cytometric analysis of CHO-K1 cells transfected with recombinant human adenosine A1, A2A, A2B or A3 receptor. Expression of each transfected receptor was analysed using the appropriate primer pair and probe, described in the table in the supplementary material. Phycoerythrin-labelled cells were analysed by appropriate gating for immunofluorescence after excitation at 488 nm. At least 3,000 events were collected and mean fluorescence intensities recorded. Data are presented as mean±sem and are from three separate experiments. An increase in fluorescence was observed only in the A1-transfected cell line. AU: arbitrary unit. **: p<0.01 versus wild-type cells
DISCUSSION
Although the expression and functions of adenosine receptors have been studied in individual cell types, usually obtained from healthy subjects, very little information is available concerning the distribution and relative expression of the four adenosine receptor subtypes in healthy or asthmatic airways. The aim of the present study, therefore, was to investigate the distribution of A1 receptors in bronchial biopsy specimens obtained from mildly asthmatic subjects and then to compare the level of expression with that in biopsy specimens taken from healthy subjects. The present data reveal that there is a low level of A1 receptor expression in healthy bronchial tissue, which appears to be predominantly located to the bronchial epithelium, whereas A1 receptor expression is markedly elevated in asthmatic bronchial tissue, particularly in the bronchial epithelium and bronchial smooth muscle. Although the precise location of this receptor was not determined, it is possible that it is expressed on ciliated epithelial cells 15, goblet cells 16, sensory nerves 17 and smooth muscle cells 18. Future studies using double labelling and confocal microscopy may discern the cell types involved.
In an attempt to quantify the increased expression in asthmatic tissue, image analysis was performed on the epithelial and smooth muscle areas of one tissue section from each biopsy specimen. Image analysis thus revealed that the mean intensity of the staining in these regions of asthmatic tissue was significantly greater than that observed with healthy tissue, which, therefore, confirmed the higher A1 receptor expression.
The A1 receptor antibody used in the present study was raised against a C-terminal portion of the A1 receptor, the aim being to limit recognition of the other adenosine receptor subtypes. In order to validate the antibody, experiments were performed to provide evidence that this A1 receptor antibody was able to selectively bind A1 receptors. Flow cytometric analysis of CHO-K1 cells transfected with recombinant human adenosine receptors showed a significant increase in fluorescence only for the cell line expressing A1 receptors compared with wild-type cells following incubation with the A1 receptor antibody. The caveat to this is the assumption that there is similar adenosine receptor protein expression in each cell line. Although a formal comparison of protein expression was not attempted in the present study, analysis of mRNA transcript levels revealed similar levels of gene expression in each cell line used in the present experiments, and, although the A3 receptor transcript level was slightly lower than that of the other subtypes, it was still of a magnitude expected to yield a very high degree of receptor protein expression. Further confirmation that the present antibody recognises and binds to the A1 receptor was demonstrated by a high level of positive immunostaining on the brain cortex and cardiac tissue samples, tissues recognised to exhibit high expression of the A1 receptor. Therefore, having validated the selectivity of the A1 receptor antibody, it was conclusively demonstrated, for the first time, that the A1 receptor is markedly upregulated in bronchial tissue from subjects with asthma who are responsive to AMP challenge. That this difference was a consistent observation was confirmed with image analysis. Given the limitations of immunohistochemistry and image analysis, it would be expected that the image analysis data are certainly not proportionally representative of total A1 receptor number but simply an indication that there is a difference in the pattern of A1 receptor expression between the mildly asthmatic and healthy phenotypes.
The present observations support a number of studies investigating adenosine receptor expression in animal models of allergic lung inflammation, which also identified an increase in airway A1 receptor expression compared with healthy naïve animals 8, 19–22. In a rabbit model of allergic lung inflammation, it was demonstrated that the A1 receptor was upregulated on airway smooth muscle, and a functional consequence of this increase was that it mediated bronchoconstriction following inhalation of adenosine 8, 23. Given the present observations, it could, therefore, be speculated that an increase in A1 receptor numbers on bronchial smooth muscle in atopic mildly asthmatic subjects may contribute to the bronchoconstrictor response to AMP observed in these subjects. In support of this possibility, an investigation into the spasmogenic effect of adenosine on isolated human bronchial strips showed a variable but small overall contractile response in tissue from healthy subjects, the magnitude of which was significantly increased in tissue from asthmatic subjects 12. The adenosine-induced contraction in isolated tissue from asthmatic subjects was significantly inhibited by the A1 receptor antagonist 2-thio-[(1,3-dipropyl)-8-cyclopentyl]-xanthine (DPCPX). The same study also showed that the response could be blocked by pre-incubation with a histamine H1 receptor antagonist and leukotriene synthesis inhibitor, suggesting that the response was mast cell-mediated. However, a subsequent investigation demonstrated that isolated human bronchial tissue has a high degree of intrinsic tone, and that histamine and leukotrienes constitute a major part of this basal tone 24. Thus it is difficult to ascertain to what degree the inhibition of adenosine-induced contractile responses by the H1 receptor antagonist and leukotriene synthesis inhibitor in the former study can be attributed to the inhibition of mast cell products released by adenosine, or simply to a reduction in basal tone. Hence further studies investigating the effects of selective adenosine receptor agonists and antagonists on isolated asthmatic bronchial tissue are required in order to provide further insight into the A1-mediated contraction of bronchial smooth muscle obtained from subjects with asthma.
On a cellular level, however, an investigation into adenosine receptor signalling pathways in isolated airway smooth muscle cells from healthy donors identified A2B receptors as the predominant adenosine receptor subtype 25. The A1 receptor is the subtype showing the greatest affinity for adenosine (reviewed in 26), and so it could be speculated that adenosine preferentially activates the upregulated A1 receptor on asthmatic bronchial smooth muscle cells. Since the A1 receptor is an inhibitory-G-protein-coupled receptor, its stimulation would result in a decrease in cyclic AMP level, leading to smooth muscle contraction. In support of this, it has recently been demonstrated that activation of the A1 receptor on healthy human airway smooth muscle cells with a selective A1 receptor agonist induces calcium mobilisation in vitro 18. Thus, given that H1 receptor antagonists do not completely block AMP-induced bronchoconstriction in asthmatic subjects and that co-administration of inhibitors of other mast cell products are not additive (reviewed in 4), it is likely that other mechanisms beside mast cell degranulation are involved in the response to AMP. Further studies are thus required in order to investigate these other mechanisms, including the potential direct spasmogenic effects of adenosine in asthmatic subjects.
The functional significance of the increased expression of the A1 receptor on the asthmatic bronchial epithelium remains to be established. However, a recent study demonstrated that activation of the A1 receptor on tracheal epithelial cells resulted in an increase in mucin gene expression 16. The present observation of an increase in A1 receptor expression on the bronchial epithelium in asthmatic subjects thus further implicates adenosine in the promotion of mucus hyperscretion via activation of the A1 receptor. Furthermore, submucosal glands were identified in tissue sections from three subjects. Two were from asthmatic subjects and showed high A1 receptor expression; however, this was not apparent in sections from the healthy subject, but these are only preliminary observations and further investigations are required before definitive conclusions can be reached. However, adenosine has been shown to induce canine tracheal mucus secretion in vivo through an A1-receptor-dependent mechanism 27, supporting the notion of adenosine as a secretagogue.
Interestingly, in 2006, it was demonstrated that subjects with chronic obstructive pulmonary disease, who also exhibit an increased extracellular level of adenosine in their airways, show increased A1, A2A and A3 receptor densities in the lung parenchyma compared with age-matched smoking controls 28. However, in order to investigate the role of adenosine in asthma further, the aim of future studies will be to characterise the expression of the other adenosine receptor subtypes in asthmatic subjects, in addition to investigating the effects of glucocorticosteroid treatment on adenosine receptor expression, since, clinically, glucocorticosteroids decrease hyperresponsiveness to AMP in subjects with asthma 29.
The present evidence of increased adenosine A1 receptor expression in asthmatic subjects reflects studies characterising adenosine receptor expression in mice partially lacking adenosine deaminase 30. Such mice accumulate high levels of adenosine extracellularly and exhibit severe pulmonary inflammation. Some of the pathological features reported in this model, such as mucus hypersecretion, were consistent with those observed in asthma, but the pulmonary inflammatory cell profile differed considerably and consisted mainly of macrophages, with no increase in eosinophil numbers. However, transcript levels of the A1, A2B and A3 receptors were found to be significantly increased, suggesting that an increase in extracellular adenosine promotes adenosine receptor signalling. The pulmonary inflammation in mice totally deficient in adenosine deaminase was even more severe, and the mice died aged 3 weeks due to respiratory distress 31. Interestingly, concurrent knockout of the adenosine A1 receptor was reported to exacerbate this inflammation further, suggesting a protective role of the A1 receptor in this model 32. However, eosinophil numbers in this model (<2% of total cells, compared with ∼40% of total in a murine model of allergic inflammation 33) were of a magnitude unlikely to result in alterations in airways responsiveness, and thus the relevance of these findings to asthma should be interpreted with caution.
In conclusion, therefore, it has been demonstrated, for the first time, that adenosine A1 receptor expression is markedly upregulated in bronchial tissue obtained from subjects with asthma who are responsive to adenosine 5’-monophophate challenge, particularly on bronchial smooth muscle and the airway epithelium.
Statement of interest
Statements of interest for M.J. Hurle, J.C. Denyer, D.E. Simcock, J.E. Coote, T.J. Savage, R.D. Murdoch, C.P. Page and D. Spina and the study itself can be found at www.erj.ersjournals.com/misc/statements.shtml
Acknowledgments
The authors would like to thank D. Hall and K. Saunders (GlaxoSmithKline, Stevenage, UK) for technical assistance with flow cytometry.
Footnotes
-
This article has supplementary material accessible from www.erj.ersjournals.com
- Received January 11, 2007.
- Accepted September 27, 2007.
- © ERS Journals Ltd
References
