Association of β2-adrenergic receptor gene polymorphisms (rs1042713, rs1042714, rs1042711) with asthma risk: a systematic review and updated meta-analysis

Asthma is a chronic respiratory inflammation disease characterized by airway hyperresponsiveness, reversible airway obstruction and airway wall remodeling [1]. It is believed to be a multifactorial disorder with a strong genetic component in its pathogenesis [2, 3]. So far, many studies have evaluated the association between genetic variants and asthma susceptibility. Numerous genes have been identified as asthma-susceptible genes, from which the β-2 adrenergic receptor (ADRB2) is the most widely studied [4‐6].

ADRB2 is encoded by an intronless gene on chromosome 5q31, which is abundantly expressed on many airway cells including smooth muscle cells [7, 8]. ADRB2 transcript has a 5′ leader cistron (5′ LC) harboring a short open reading frame (ORF) that encodes a 19-amino acid peptide, which regulates mRNA translation and controls the cellular expression of ADRB2. A variation at position 19 that causes a change from cysteine (Cys) to arginine (Arg) was reported in the 5′ LC, and this variation plays a role in regulating ADRB2 gene expression [9‐11]. However, little is known regarding the possible role of this polymorphism in asthma. In addition, two missense variations (rs1042713, c.G46A, p.Gly16Arg and rs1042714, c.G79C, p.Gln27Glu) that occur in high allelic frequency in the general population have been identified, corresponding to a change from glycine (Gly) to arginine (Arg) at amino acid position 16 and glutamate (Gln) to glutamine (Glu) at amino acid position 27 [12]. Studies in vitro [13] and primary cultures of cells expressing these endogenous variants [14] illustrated the different phenotypes between the polymorphic receptors. The Gly16 receptor could enhance agonist-promoted downregulation of receptor expression compared with the Arg16 receptor. In contrast, the Glu27 receptor is relatively resistant to agonist-promoted downregulation compared with the Gln27 receptor [13, 14]. Genetic studies have indicated that these variations not only affect the risk of asthma, but also affect the therapeutic outcomes of inhaled β2-adrenergic receptor agonists [15‐20].

Considering the impact of the asthma risk potentially resulting from ADRB2 gene variations, a number of case-control studies have explored the association between the ADRB2 gene polymorphisms and asthma risk in different ethnicities [21‐23]. However, these results are conflicting and inconclusive, which are possibly due to the limitations associated with an individual studies and small sample size. To shed light on these contradictory results and to more precisely evaluate the relationship between the ADRB2 gene polymorphisms and asthma risk, several meta-analyses concerning the association between ADRB2 gene polymorphisms and asthma have been reported [15, 24‐29]. However, these meta-analyses have also shown inconsistent results. After publication of these meta-analyses, many additional case-control studies about the ADRB2 polymorphisms on asthma risk were carried out [30‐34]. Therefore, we present the results of a comprehensively updated meta-analysis of all relevant published data to investigate the associations between ADRB2 gene polymorphisms and asthma risk with a focus on rs1042713, rs1042714 and rs1042711 polymorphisms.

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement guidelines.

Characteristics of the studies included in the meta-analysis.

The flow chart in Fig. 1 outlines the study selection process. After a comprehensive search of the PubMed, Web of Science, EMBASE, CNKI and Wanfang databases, a total of 1557 articles were identified. First, we excluded 882 duplicated studies. After reading the abstracts and titles, 530 studies were excluded. The remaining 145 studies were then assessed for inclusion. Of these, 72 studies were excluded because 48 studies were not irrelevance, 16 studies lacked detailed genotypes, and 8 studies lacked a case-control design. Finally, a total of 73 studies met the inclusion criteria and were included in the meta-analysis [8, 12, 17, 21‐23, 30‐34, 36‐97]. The characteristics of each eligible study are shown in Additional file 1. Of these 73 studies, each contains two independent studies, and the data is extracted accordingly. During the process, if the sum of genotype distribution of one study is over than 1, the study is excluded from the meta-analysis. The genotype, allele distribution and P-value of HWE for the rs1042713, rs1042714 and rs1042711 polymorphisms are respectively shown in Additional files 2, 3 and 4.

In this study, the associations between three ADRB2 gene SNPs (rs1042713, rs1042714 and rs1042711) and the risk of asthma were conducted base on the data from 73 studies involving 13,493 asthmatic patients and 22,931 controls. This meta-analysis showed that the rs1042713 polymorphism was not a risk factor for overall asthma susceptibility, which was consistent with most of the previous meta-analyses [15, 16, 25, 27‐29]; however, the data contradict one of the latest meta-analyses [24]. The difference between these results is seemingly due to the different inclusion criteria. The inclusion criteria in the Xie et al. [24] meta-analysis used P-value of the HWE; however, when stratified by the P-value of the HWE, there was still no association between the Gly16Arg polymorphism and asthma in the HWE (P > 0.05) subgroup in our study and in that of Liang et al. [25]. When reviewing Xie et al.’s literature list, we found that they missed some studies that satisfied their inclusion criteria [31, 54, 77‐79, 87, 88, 91, 92, 95], which may be a reason for the discrepancies.

Some studies reported that the genotype frequency and allele frequency of the rs1042713 polymorphism vary among different ethnic groups [98, 99]. A search of 1000 Genomes Project or Hapmap data showed an approximately 15% difference in allele frequency for the rs1042713 polymorphism. The Gly16 homozygous genotype frequency is more common in the non-Hispanic White than Chinese and more frequent in non-Hispanic White compared to non-Hispanic Black. Therefore, there is high heterogeneity for the rs1042713 polymorphism in the overall analysis. Even though we excluded some studies that did not meet the HWE, heterogeneity was not decreased. Selective bias in the literature has an important effect on the results of the overall meta-analysis. Large sample sizes can better reflect the truth of the effects of the rs1042713 polymorphism on the asthma risk.

Because the genotype frequency and allele frequency of the rs1042713 polymorphism vary among the different ethnic groups, we divided the ethnic groups according to the studies reported by Ortega et al. [100, 101], such that (1) non-Hispanic Whites of European ancestry were designated as Caucasian, (2) Mexicans and South Americans were designated as Hispanic-Latinos, (3) African Americans and non-Hispanic Blacks from Europe and Africa were designated as non-Hispanic Blacks; and (4) Chinese, Japanese and Korean individuals were considered separate Asian ethnic groups and designated as the East-Asian subgroup. In addition, Indian and Arab descendants were designated as separate ethnic groups because these ethnic subgroups have different genotypes and allele frequencies [98]. The analysis stratified by ethnicity showed a significant risk association in Hispanic-Latinos in the dominant model comparison, consistent with a previous study where a significant association in the South American population was found [25]. In addition, the previous study claimed to divide the cohorts by ethnicity, but for the most part, the cohorts were divided by continent when they combined Indian, Arab, Japanese and Han Chinese individuals as the Asian population [24, 25]. These populations have different genotypic and allelic frequencies for the rs1042713 polymorphism [98]. After the stratified analysis, the heterogeneities in the Arab and Indian population were decreased, and a protective association in the Indian population in the dominant model comparison and a risk association in the Arab population in the dominant model comparison and the homozygote genotype comparison were found. However, there is a need for further studies with larger sample sets.

For the rs1042714 polymorphism in the current meta-analysis, benefitting from the inclusion of more case-control studies, a protective effect was found not only in the pediatric subgroup in the recessive model comparison and the homozygote genotype comparison but also in the overall population in the recessive model comparison, the homozygote genotype comparison and the allelic genetic model. This was consistent with previous reports by Liang et al. [25] and Ammarin et al. [27] that showed a protective effect in the pediatric subgroup in the recessive model comparison and the homozygote genotype comparison, confirming that the Glu27 polymorphism was a protective effect for asthma. A genetic study showed that Glu27 homozygotes had less reactive airways than Gln27 homozygotes, and these results could further suggest a protective role for the Glu27 polymorphism in asthma [102]. In addition, in vitro and ex vivo functional studies indicated that Glu27 allele enhanced resistance to agonist-induced down regulation of the receptor, suggesting a protective role of Glu27 polymorphism in regard to receptor desensitisation [13, 14].

In the analysis stratified by HWE according to the P-value for the rs1042714 polymorphism, significant associations were found in the subgroup with P > 0.05 in the dominant model comparison and the allelic genetic model but not in the P < 0.05 subgroup. These results need to be interpreted with caution. The reason the control group population was not in HWE may be selection for a particular phenotype or that the population was not sufficiently large or random.

For the rs1042711 polymorphism, no significant associations with the risk of asthma were found in any comparison model. More research is needed because only seven case-controls were included in this study. There might not be sufficient statistical evidence to clarify the association between the rs1042711 polymorphism and the risk of asthma.

There could be several potential limitations to this meta-analysis. The first problem relates to the limitations of the literature. All available literature should be included in the meta-analysis, but we only included literature published in English and Chinese, thus neglecting studies published in other languages. Second, even though the existing literature had acceptable quality, detailed information was not provided such as asthma definition varied among different articles and this may be a confounding factor. Using a self- or physician-diagnosis of asthma can be confounded by individuals who do not have asthma such as older subjects with a smoking history who could have COPD. These physician-diagnosed cohorts many times do not have an objective diagnostic basis of asthma based on methacholine BHR or beta agonist responsiveness which could result in confounded and undetected associations. In addition, the participation rates for cases and controls were not reported in the majorities of included studies; thus, our meta-analysis was unable to explore the selection bias. Moreover, with limited information about maternal constitutional and environmental risk factors for asthma (such as smoking history), we could not evaluate the gene-gene and gene-environmental interactions. The different definitions of asthma and the environmental factors in individual studies were obvious sources of clinical heterogeneity and may produce bias. Therefore, moderate-to-high heterogeneities were found in some genetic models for the Gly16Arg polymorphism. Stratification by ethnicity may help to improve homogeneity among studies, but it may also influence statistical power. In addition, some meta-analysis studies claimed to divide the cohorts by ethnicity but it seems like the cohorts were actually being divided by continent for the most part, it will induce the contradictory findings. Third, the Gly16Arg and Gln27Glu polymorphisms tagging for rare variants modulated therapeutic responses and contributed to asthma risk [103]; however, these variants were not specifically genotyped.

In conclusion, the present meta-analysis indicates that the rs1042714 polymorphism is an important genetic protective factor to decrease the risk of developing asthma, especially in children. The rs1042713 polymorphism may be involved in the risk of asthma in the Arab and Hispanic-Latino populations and a protective factor in the Indian population. However, more well-designed and high-quality studies with larger sample sizes should be conducted to support this finding in various ethnic groups.