β2‐adrenoceptor polymorphisms and asthma

There are two main approaches to the study of the genetics of asthma and its intermediate phenotypes. The first is to use positional cloning approaches starting with a genome screen to attempt to identify genes important in determining the development of disease. The second, known as the candidate gene approach, involves examining variants (polymorphisms) in genes whose products are known to be relevant to the pathophysiology of the disease [ 1]. One gene which has attracted extensive attention is the β2-adrenoceptor gene. This gene is situated on 5q32 [ 2], a region which has shown linkage in genome and chromosome screens for the intermediate phenotypes total IgE and bronchial hyperresponsiveness (BHR) [ 3]. For a candidate gene to potentially be important in the disease a number of criteria must be met. Firstly, the gene product must be relevant to the pathophysiology of the disease. Secondly, the gene must contain mutations either within the coding region or the regulatory regions controlling gene expression; these mutations need to be functionally relevant. The demonstration of functional relevance for a mutation is particularly important given the high rate of polymorphic variation within the human genome. This has been estimated to be around one in 1000 in coding DNA and around one in 500 base pairs in non-coding DNA. Thirdly, mutations which are functionally relevant should demonstrate association and/or linkage with an appropriate phenotype. Finally, for a mutation to contribute to the disease risk in the population at large the mutation must be relatively common: rare mutations may greatly increase the risk of developing asthma in individual families but are unlikely to be important in determining the population risk as a whole. It follows, however, that the size of the effects of common polymorphisms may be relatively small, because if major deleterious consequences occurred in individuals with a given polymorphism it would soon be lost from the population. The β2-adrenoceptor gene is a particularly interesting candidate because of the large number of polymorphic variants of the gene which exist. The coding region for the gene is known to contain at least nine single base substitutions, although because of degeneracy in the amino acid code only four of these result in amino acid changes in the resultant receptor [ 4]. Of these three appear to be functionally important. The Thr–Ile164 variant (threonine–isoleucine) shows altered agonist binding and altered receptor downregulation following agonist stimulation [ 5], however, this polymorphism is rare (the allelic frequency is around 3% in Caucasian populations: because every individual carries two copies of the β2-adrenoceptor gene only one in 900 individuals would be homozygous for this variant in theory). The amino acid substitutions at codon 16 (Arg–Gly16) (arginine–glycine) and 27. (Gln–Glu 27) (glutamine–glutamic acid) have been more extensively studied and are much commoner in the general population. Work in both cell lines [ 6] and in primary cultures of human airway smooth muscle cells [ 7] from individuals of known genotype has demonstrated an association of the Gly16 and Gln27 forms of the receptor with increased downregulation following agonist exposure, whereas the Glu27 form of the receptor appears to be protected from downregulation. Agonist binding to the receptor does not vary between the different forms of the receptor per se in cell culture model systems [ 6] and it is only following prolonged agonist exposure that differences in cell signalling can be observed [ 7]. These studies have been extended by in vivo work which demonstrates that these effects can be observed in bronchial epithelial cells following agonist stimulation [ 8] and, for the Gly16 form of the receptor, effects may also be seen in circulating lymphocytes in patients with nocturnal and non-nocturnal asthma [ 9]. Because of the clear functional associations of the amino acid 16 and 27 variants of the β2-adrenoceptor a number of groups have performed clinical studies which have addressed two main questions. Some studies have looked for association and/or linkage of β2-adrenoceptor polymorphisms with asthma or its intermediate phenotypes per se. In addition, given the important role for β-adrenoceptor agonists as bronchodilator agents, other studies have looked to see whether treatment responses or asthma severity are in part determined by β2-adrenoceptor polymorphism. The first large study to address the possibility that β2-adrenoceptor polymorphisms may be associated with asthma or intermediate phenotypes was performed in families multiplex for asthma recruited in Wessex [ 10]. The ethnic background of populations which have been studied is likely to be important because both Japanese and black South African populations are known to have different distributions of these two common polymorphisms with the Glu27 variant being rarer in those racial groups. In this study we found both association and linkage of the codon 27 polymorphism with IgE levels, with the Gln27 form of the receptor being associated with elevated IgE, although the overall effect was relatively small as a contributor to the level of total IgE. No association was observed between a semiquantitative score for asthma, a semiquantitative score for atopy or BHR and either of these polymorphisms. We subsequently extended this study into a large random population of 620 individuals from Nottingham: in this population, as perhaps might be predicted given its more heterogeneous nature, we failed to confirm the association of the codon 27 polymorphism with IgE [ 11]. Again no association was seen between β2-adrenoceptor polymorphism and doctor-diagnosed asthma, atopy or BHR. In contrast, we observed weak association of the codon 27 polymorphism with childhood asthma in a random population recruited from Aberdeen [ 12]. Whilst this may be a false positive result one could also argue that if IgE levels are associated with β2-adrenoceptor polymorphisms then the greater contribution of IgE to childhood asthma might account for this association. The study by Ramsay et al. in this issue of the journal [ 13] also failed to observe association of either the amino acid 16 or 27 β2-adrenoceptor polymorphisms and asthma, although they did observe an association between Glu27 and reduced responsiveness to histamine. This latter association is interesting: in a study of 60 carefully matched mild to moderate asthmatics we also found association of Glu27 with decreased airway responsiveness [ 14]. Given the results from these studies, what general conclusions can be drawn? First it is clear that neither the Gly16 nor Gln27 allele is a major determinant of asthma in adult populations. There are, however, consistent albeit weak effects suggesting that these polymorphisms are associated with total IgE levels and BHR. Although these effects were not observed in all studies this is not perhaps surprising given the strength of the association in those studies which have now reported positive findings. There are two potential explanations for these associations if one accepts that they are real. The first is that there is a causal relationship between β2-adrenoceptor polymorphism and control of IgE and/or BHR. The second possibility is that these polymorphisms are in linkage disequilibrium with another locus on chromosome 5, important in the control of IgE and/or BHR: intuitively this seems the more likely explanation [ 10]. The other group of studies which have been performed looking at β2-adrenoceptor polymorphism have attempted to determine whether these polymorphisms contribute in determining treatment response. Whilst one could hypothesize that circulating catecholamines or exogenously applied β-agonists may alter the level of expression of β2-adrenoceptors in the airways of individuals depending upon the β2-adrenoceptor genotype of an individual and so determine the degree of an individual's airway responsiveness, one might expect that the largest effect of these polymorphisms (given that their main functional role is in determining downregulation profile) would be following chronic dosing with β2-agonists. The most convincing data suggesting a functional role for these polymorphisms in vivo come from studies in this area. In a collaborative study with Dundee we assessed the degree of bronchodilator tachyphylaxis in a group of mild asthmatics following 2 weeks' treatment with high dose formoterol and found that individuals with the Gly16 genotype showed significantly greater loss of response to a subsequent challenge with inhaled formoterol [ 15]. Although the numbers in this study were small these data are consistent with results from in vitro studies. Martinez has also demonstrated that bronchodilator responses to β-agonists in children appear to be in part determined by the amino acid 16 β2-adrenoceptor polymorphism, with individuals being less likely to respond to inhaled salbutamol if they carry the Gly16 allele [ 16]. This effect was observed in both asthmatic and nonasthmatic children. The other effect of inhaled β2-agonist is to protect against subsequent bronchoconstriction when individuals are challenged with, e.g. methacholine or histamine − the effect of β2-adrenoceptor polymorphism upon this response has been less extensively studied, although in one small study of subjects on inhaled steroids we did not find an association of β2-receptor polymorphism with loss of bronchoprotection, which was observed in both Arg16 and Gly16 homozygous individuals [ 17]. In summary reasonably consistent associations of β2-adrenoceptor polymorphisms have been demonstrated both with IgE and BHR and with treatment response. However, a number of questions remain. Firstly the effects seen in most studies are relatively small and hence the clinical relevance of these observations remains uncertain. The most important question which requires answering is whether treatment response to long-acting β-agonists is determined by genotype to such an extent that it would be worth genotyping individuals before deciding on treatment for moderately severe asthmatics uncontrolled on moderate doses of inhaled steroid. Advances in DNA chip technology make this option technically feasible (and potentially cost effective) but the magnitude of effect attributable to β2-adrenoceptor polymorphism may well not be sufficiently large to make this approach clinically worthwhile. It might also be thought that if treatment response is in part determined by genotype, disease severity might also be genotype dependent. However, two recent studies have suggested that neither the risk of asthma death or near death, nor deteriorating asthma control on regular short-acting β-agonists is genotype dependent [ 18, 19]. Finally, it will be important to revisit the question of functional effects. Whilst the association of the amino acid 16 and 27 polymorphisms with altered receptor downregulation profiles is clear cut the mechanism underlying these effects is uncertain. Recently a number of novel polymorphisms within the β2-adrenoceptor promoter region have been described, one of which is in a short upstream leader peptide termed β upstream peptide (BUP), which is known to be important for controlling the level of expression of the β2-adrenoceptor in cells. This polymorphism (− 47 C-T) induces a non-conservative amino acid change (Cys–Arg in BUP) and in two recent studies has been shown to have functional effects on gene expression [ 20, 21]. Because this polymorphism is in genetic terms very close to the polymorphisms within the coding region of the gene, linkage disequilibrium exists between them (and the other β2-adrenoceptor promoter polymorphisms). It may be, therefore, that the functional effects previously ascribed to the amino acid 16 and 27 polymorphisms are in fact due to functionally important polymorphisms within the gene promoter. It may prove necessary to study extended haplotypes across this region to fully determine the potential importance of β2-adrenoceptor polymorphisms to receptor expression and function. It is intriguing to speculate on the reason for the high polymorphic rate within this genomic region. To date 15 polymorphisms have been described within a 3-kb region containing the β2-adrenoceptor gene and its upstream regulatory regions, a much higher rate of polymorphism than exists in many other regions. For example we have recently screened the coding region of both the muscarinic M2 receptor (Fenech and Hall, unpublished data) and histamine H1 receptor [ 22], both of which are of similar size to the β2-adrenoceptor gene and in both cases have only found one degenerate polymorphism. The explanation for the highly polymorphic nature of the β2-adrenoceptor gene remains to be determined.