INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS Bronchial Hyperresponsiveness DISCUSSION REFERENCES

Bronchial hyperresponsiveness (BHR) is an abnormal airways constriction following the exposure to a provoking stimulus. This condition is found in most asthmatics and represents a risk factor for asthma (1, 2), so that an increasing reliance is being placed on its presence to diagnose the disease. Nevertheless, BHR can be observed also in normal individuals (3) as well as in patients with other respiratory diseases, although varying to a great extent according to environmental exposure, airways viral infections, therapeutic intervention, gender, and atopic status (6).

A heritable component of BHR has been shown both in humans and in animals (9), but the basis of such genetic susceptibility is still poorly understood and the responsible gene (or genes) remains unknown. One candidate is the ADRB2 gene, coding for the ₂-adrenergic receptor (₂AR) which is expressed on the surface of airway smooth muscle cells and plays a key role in tuning airways reactivity (13). It has long been hypothesized that defects in this receptor might represent a pathogenic factor in asthma (14). Four polymorphisms have been described for this gene, giving rise to amino acid changes at position 16 (Arg or Gly), 27 (Glu or Gln), 34 (Val or Met), and 164 (Thr or Ile) of the encoded protein (15). Notably, polymorphisms 16 and 27 have been shown to correlate with some clinical manifestations of asthma in different studies: Gly at position 16 associated with a more severe, steroid-dependent form of the disease (15), with the nocturnal phenotype of asthma (16), and with a greater degree of bronchodilator desensitization (17); Arg 16 has been recently shown to be associated with a higher response to albuterol (18); at position 27, Glu correlates with lower airway reactivity among asthmatics (19) whereas Gln is associated with high levels of total IgE in asthmatic families (20). Despite these observations, different distributions of these polymorphisms between patients and control subjects were never observed and, accordingly, a disease-modifying role only has been assigned to this gene. However, most of the association studies performed so far have focused on single site genotypes in small samples of patients and control subjects. Therefore, no information is available on the distribution of the haplotypic combinations of these polymorphisms and their possible relevance at the population level.

To this purpose, we analyzed the ADRB2/BHR relationship in a large, highly homogeneous sample of 248 young subjects sharing race, gender, age, and remarkably, current living environment. Three serial methacholine challenge tests have been performed on each subject during a 10-mo observation period and a strict definition of BHR was used, by narrowing the analysis to those subjects showing a constant positive or negative response. An improved definition of genetic variability at ADRB2 locus was used, by assessing all four haplotypic combinations that can occur at relevant polymorphic positions 16 and 27. The results so far obtained were finally adjusted for serum levels of total IgE and for IgE sensitization toward house dust mite (HDM).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS Bronchial Hyperresponsiveness DISCUSSION REFERENCES

Subjects

The present study is part of an extended program of preventive medicine devoted also to investigating the individual and environmental risk factors for allergy and asthma in the military. It was performed in 1993-1994 in the Italian Air Force (IAF) School Base of Caserta (Campania, Italy). The study population consisted of a community of 285 white males, age 18 to 25, all attending a residential course for ground personnel. The students of this particular course had passed a physical examination which did not include allergy or pulmonary function tests; therefore subjects with a history of allergy or asthma had not been excluded from it. The observation period lasted from September 1993 to July 1994. Two hundred forty-eight out of 285 students (87%) completed the study after written informed consent. No differences in known characteristics were observed between those who did and those who did not complete the survey. The study design was approved by the IAF review board authorities.

ADRB2 Genotyping

To identify haplotypic combinations of the two relevant polymorphisms at positions 16 and 27, we set up a procedure based on the amplification of the region of interest by polymerase chain reaction (PCR), sequence specific oligonucleotide (SSO) hybridization, and enzyme restriction analysis. After standard genomic DNA extraction from whole blood, all samples were amplified for 30 cycles with previously described ADRB2 gene-specific primers (19), spanning both polymorphisms coding for aa 16 and 27 (upstream: 5'-CCCAGCCAGTGCGCTTACCT-3'; downstream: 5'-CCGTCTGCAGACGCTCGAAC-3'). Each cycle was as follows: denaturation at 94° C for 30 s, annealing at 62° C for 30 s and extension at 72° C for 30 s. An additional step of 10 min extension was added at the end of the reaction to prevent heteroduplexes formation which could affect the following characterization by enzyme restriction analysis. One-microliter aliquots of PCR products were transferred onto nylon membranes and hybridized in 6× SSPE (NaCl 0.9 M; NaH₂PO₄ 70 mM; EDTA 6 mM), 0.5% sodium dodecyl sulfate (SDS), 5× Denhardt's, and 100 µg/ml salmon sperm DNA with SSO probes specific for Arg16 (5'-GCACCCAATAGAAGCCATG-3'), Gly 16 (5'-GCACCCAATGGAAGCCATG-3'), Gln27 (5'-GTCACGCAGCAAAGGGAC-3'), and Glu27 (5'-GTCACGCAGGAAAGGGAC-3'). Hybridizations were carried out in stringent conditions at the melting temperature of each oligonucleotide. Filters were washed twice at room temperature in 2× SSPE, 0.1% SDS for 10 min, once 2° C above the melting temperature in 6× SSPE, 1% SDS for 10 min and then exposed to autoradiography. This allowed the assignment of haplotypic combinations to all individuals homozygous at either position 16 or 27, or both. Samples appearing heterozygous at both positions were further characterized by enzyme restriction analysis and hybridization of digestion products. Briefly, 7 µl of PCR products were treated with Bbv I restriction enzyme (New England Biolabs, Boston, MA), for 3 h at 37° C, run on a 12% polyacrylamide minigel, stained with ethidium bromide and photographed under ultraviolet light. After electrophoresis, digestion products were transferred on nylon membranes by Southern quick blot and filters were hybridized with either Arg16 or Gly16 SSO probes. Haplotype assignment was then made by the localization of the radioactive signal. An example of the entire procedure is shown in Figure 1. Following this approach, haplotypes for polymorphic positions 16 and 27 could be assessed for all 248 analyzed subjects. SSO typing results were confirmed on 10 randomly selected samples by direct sequencing of PCR products with AmpliCycle Sequencing kit (Perkin Elmer, Norwalk, CT), according to the manufacturer's instructions.

View larger version (18K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   View larger version (50K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 1.   Bbv I restriction analysis and SSO hybridization procedure for the identification of ADRB2 haplotypes in samples heterozygous at both polymorphic positions 16 and 27. (A) Expected Bbv I digestion patterns of the four possible haplotypic combinations for polymorphisms at positions 16 and 27. Bbv I recognizes and cuts only Gln27- but not Glu27-coding sequence, 112 bp from the 3'-end of the amplified product, in addition to a nonpolymorphic site located 67 bp from the 5'-end (arrowheads). Thus, Bbv I digestion yields DNA fragments of 55, 67, and 112 bp for either Arg16-Gln27 or Gly16-Gln27 haplotypes, versus fragments of 67 and 167 bp for either Arg16-Glu27 or Gly16-Glu27 haplotypes. (B) Schematic representation of the expected electrophoretic patterns, characterized by all four restriction fragments, resulting from Bbv I digestion of samples heterozygous at both positions 16 and 27. The two possible combinations are shown: Arg16-Gln27/Gly16-Glu27 (left) and Gly16-Gln27/Arg16-Glu27 (right). (C ) Polyacrylamide gel electrophoresis of Bbv I digestion products. Lanes 1 and 2: Gly16-Glu27 and Gly16-Gln27 homozygous reference samples, respectively; lanes 3-7: five samples from our panel, randomly selected among those resulting heterozygous at both positions 16 and 27; lane 8: DNA molecular weight marker V (Boehringer Mannheim GmbH, Mannheim, Germany). (D) Hybridization pattern obtained with a Gly16-specific SSO probe, after Southern blot transfer of the gel shown in C. Haplotypes are assigned based on which band gives a (positive) radioactive signal (asterisk in B): samples in lanes 3-7 are all Arg16-Gln27/Gly16-Glu27. Consistent results can be obtained by alternatively hybridizing the filters with an Arg16-specific SSO (not shown).

Bronchial Challenge with Methacholine

Bronchial reactivity to methacholine was tested on all subjects in three different occasions (October 1993, February and July 1994) by a dosimeter method. Diluent (phosphate-buffered saline) and cumulative doses of 50, 100, 200, 400, 800, and 1,600 µg of methacholine in diluent (Lofarma, Milano, Italy), were delivered by a metered nebulizer dosimeter. At each point FEV₁ was measured (Cosmed, Milano, Italy). The test was continued until either a fall of 20% or more in FEV₁ was obtained or the last dose was reached. Results were expressed as the provocative dose, in micrograms, of methacholine causing a 20% fall in FEV₁ from the postdiluent value (PD₂₀%FEV₁), which was calculated on the cumulative log dose-response curve by linear interpolation between the last two points. A PD₂₀%FEV₁ < 1,600 was regarded as positive. No subject received ₂-agonist administration for at least 1 wk prior to methacholine challenge testing.

Total and Specific IgE

Total and specific IgE measurements were performed on blood samples obtained in February 1994. Total IgE and IgE specific for a number of common inhalant allergens and Dermatophagoides pteronyssinus (house dust mite, HDM) were determined by immunoenzymatic fluorescence assays (IgE-FEIA and RAST-FEIA; Pharmacia CAP system, Uppsala, Sweden). Results were expressed as kU/L, according to the manufacturer's instructions.

Statistical Analysis

Association of ADRB2 haplotypes with BHR was assessed by Fisher's exact test. A multiple logistic regression analysis was used to adjust the effect of total and specific IgE levels on that of ADRB2 haplotypes in conferring susceptibility to BHR, chosen as the dependent variable (21).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS Bronchial Hyperresponsiveness DISCUSSION REFERENCES

ADRB2 Haplotypes

A total of 248 individuals underwent ADRB2 genotyping by an ad hoc procedure that allowed the identification of all four possible haplotypic combinations of relevant polymorphisms at positions 16 and 27. The relative haplotype frequencies, obtained for this sample, are shown in Table 1. Notably, three of the four expected combinations, namely Arg16-Gln27, Gly16-Gln27, and Gly16-Glu27, showed similar frequencies, whereas we found no single instance of the Arg16-Glu27 haplotype. There was indeed a strong linkage disequilibrium between the two polymorphisms (p < 0.00001 for the deviation from a random association). The relative frequencies of single polymorphisms did not differ substantially from those already published for other Caucasian populations (15, 16, 18), in that Gly16 and Gln27 were found to be commoner than Arg16 and Glu27 at the corresponding positions (65% and 69% versus 35% and 31%, respectively).

	Bronchial Hyperresponsiveness

TOP ABSTRACT INTRODUCTION METHODS RESULTS Bronchial Hyperresponsiveness DISCUSSION REFERENCES

BHR was measured on all subjects, by methacholine challenge test, in October 1993, February and June 1994. The diverse patterns of bronchial response to methacholine challenge during the observation period and their relative frequencies are reported in Table 2. Sixty-seven (27%) individuals showed a positive response in October, 45 (18.1%) in February, and 52 (20.9%) in June. Fifty-two subjects (21%) showed a variable response, being positive only once or twice (patterns B through G). Season-related intervening factors could account for the observed fluctuations in the percent of BHR positives. For example, among the 30 subjects showing BHR only once (patterns B, C, and D), the majority (16/30, 53.3%) was positive in October, possibly as a residual consequence of respiratory viral infections exchanged at the beginning of the course. Conversely, the lowest prevalence of BHR was found in February when, compared with June and October, some environmental triggering agents, like allergenic pollens, are almost absent in this area.

Serum Total and Specific IgE

Total IgE had a geometric mean of 69.0 ± 1.1 (SEM) kU/L in the studied population. As expected, a relation between total IgE and BHR was observed: subjects with at least one positive methacholine response were only 12.9% in the lowest quartile of total IgE, 21.0% in the second quartile, 43.5% in the third quartile, up to 53.8% in the highest quartile (p for linear trend 0.0001).

IgE sensitization against common inhalant allergens including HDM were measured, the latter showing the highest prevalence in our sample: subjects with a detectable concentration (> 0.35 kU/L) of IgE against HDM were 40.7% (101/ 248). HDM sensitization was also by far the most strongly associated with the occurrence of at least one positive methacholine response in the overall period (48.5% versus 22.45% of BHR positives in HDM-sensitized and nonsensitized individuals, respectively; p < 0.0001).

Association of ADRB2 Haplotypes with BHR

ADRB2 haplotype frequencies at each measurement of methacholine challenge response taken in October, February, and June were similar for the Arg16-Gln27 variant. Conversely, the other two haplotypes were generally overrepresented (Gly16-Gln27) and decreased (Gly16-Glu27) among BHR-positive subjects, with the highest differences observed in February (47.8 versus 31.3 and 23.3 versus 32.7 in BHR-positive and BHR-negative individuals for Gly16-Gln27 and Gly16- Glu27, respectively). To avoid potential biases given by BHR variability over time, statistical analysis of the association between BHR and ADRB2 haplotypes was limited only to those subjects showing a consistent negative or positive methacholine response (patterns A and H in Table 2). Accordingly, 52 individuals (patterns B through G in Table 2) were excluded from any further analysis, and haplotypes of subjects with persistent BHR were compared with those of subjects consistently negative to methacholine challenge. As shown in Table 3, the Gly16-Gln27 haplotype was found to be significantly associated with BHR (p = 0.016), whereas no association was found with other haplotypes and with single polymorphic positions 16 or 27 (not shown).

Multivariate Analysis of ADRB2 Gly16-Gln27 Haplotype, IgE Levels, and BHR

No relevant association was observed for either ADRB2 haplotypes or single polymorphisms with total or specific IgE levels. However, serum IgE are a well-known BHR-associated factor (22, 23) and Gln27 has been recently shown to associate with high total IgE levels in asthmatic families (20). We therefore evaluated if the association with the Gly16-Gln27 haplotype persisted in a logistic regression analysis where total and HDM-specific IgE (both strongly associated with BHR in our sample) were introduced as covariates. As shown in Table 4, total IgE, HDM-specific IgE, and the Gly16-Gln27 ADRB2 haplotype were all significantly and independently associated with BHR. Interestingly, there was a linear association between BHR and the number of Gly16-Gln27 variants carried by each subject (p for trend 0.004), suggesting haplotype additivity. A descriptive analysis of the independent contribution of HDM sensitization and the number of ADRB2 Gly16- Gln27 haplotypes to a persistent BHR phenotype is reported in Figure 2. In the subgroup of subjects with high levels of HDM-specific IgE (> 3.5 kU/L), a persistent BHR was observed in 80% (4/5) of Gly16-Gln27 homozygous subjects but only in 33.3% (5/15) of those negative for this haplotype. No evidence of association with BHR was observed when identical analyses were performed for the haplotypes Gly16-Glu27 and Arg16-Gln27 (not shown).

View larger version (24K): [in this window] [in a new window]   Figure 2.   Prevalence of persistent BHR according to HDM-specific IgE levels and ADRB2 Gly16-Gln27 haplotype. Gly16-Gln27 negatives (white box); Gly16-Gln27 heterozygotes (grey box); Gly16- Gln27 homozygotes (black box). The number of subjects for each category is reported below the abscissa.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS Bronchial Hyperresponsiveness DISCUSSION REFERENCES

This study was undertaken to evaluate at a population level the association of BHR with ₂-adrenergic receptor (ADRB2) polymorphisms. Compared with previous surveys on this topic, the novelties of our approach reside in that: (1) the studied population has not been selected for any disease-related phenotype, (2) BHR was strictly defined as a constant response to serial methacholine challenge tests, and (3) genetic variability at ADRB2 locus has been more precisely defined by the assessment of all four haplotypic combinations that can occur for the relevant polymorphic positions 16 and 27.

A major finding of ADRB2 genotyping in our sample is the absence of the Arg16-Glu27 variant of the receptor. Similar results have been recently obtained by Martinez and coworkers (18) in a sample of Hispanic and Caucasian children. By studying single site polymorphisms, they found only three Arg16-Glu27-carrying chromosomes, out of all those for which the haplotype could be deduced (3/136, 2.2%). Moreover, from existing literature on the frequencies of single genotypes at positions 16 and 27, it can be derived that this haplotype is rare also in other Caucasoid populations (15, 16). This observation will be relevant for future studies and suggests the opportunity of an eventual reappraisal of previous findings based on the analysis of single ADRB2 polymorphisms.

The functional properties of the diverse ₂ARs have been studied through their transfection into Chinese hamster fibroblasts (24) and, in that system, the Arg16-Glu27 variant was strongly resistant to agonist-induced downregulation (24). Accordingly, it is tempting to speculate that a virtual absence of this haplotype could be the result of a negative selection acting on such a poorly "flexible" signaling protein.

We observed an increased frequency of the Gly16-Gln27 haplotype among BHR-positive subjects. This finding gains relevance because we used a very strict and specific definition of BHR, by excluding from the analysis all subjects showing variable methacholine responses. This approach should limit the bias caused by generic variability of individual BHR over time and, consistently, the possibility of finding poorly reproducible positive/negative results. Many individuals would have been misclassified on the basis of single methacholine challenge test responses. On the other hand, no associations could be observed by examining only single polymorphisms at positions 16 and 27. Thus, repeated BHR measurements and haplotypes identification may represent a key strategy to reveal otherwise hidden ADRB2-BHR associations.

A further characterization was made in this study by ruling out the potential confounding variable due to the atopic status, expressed here through the levels of total and HDM-specific IgE. The logistic regression analysis showed that the association of the Gly16-Gln27 haplotype with BHR was independent of these factors, proposing this variant as an unrelated risk factor for the development of BHR.

The effect of ADRB2 polymorphisms on the functional properties of the corresponding ₂AR variants has been studied in ₂AR-transfected Chinese hamster fibroblasts and in human bronchial smooth muscle cell lines (24, 25). It was shown that amino-terminal polymorphisms 16 and 27 affect the agonist-promoted downregulation in that Gly16 confers increased downregulation when compared with Arg16, as does Gln27 when compared with Glu27 (24, 25). Thus, the association of the Gly16-Gln27 ₂AR haplotype with BHR in our sample could be explained by the fact that this combination has the greatest potential for receptor downregulation. However, functional data on ADRB2 haplotypic combinations have been derived solely from transfected Chinese hamster fibroblasts (where Gly16-Gln27 but also Gly16-Glu27 showed enhanced downregulation) (24). Moreover, in vitro work on human pulmonary smooth muscle cells has provided information only on single site polymorphisms, as none of the studied subjects was homozygous for ADRB2 haplotypes (25). Thus, studies on bronchial cell lines obtained from individuals homozygous at both polymorphic sites (16 and 27) are needed to ascertain the real properties of the Gly16-Gln27 variant as well as those of the other ₂ARs.

On the other hand, it cannot be excluded that other uncharacterized polymorphisms in the promoter region of ADRB2 gene might contribute to the genetic inheritance of BHR while being in linkage with those studied here. We are currently investigating this hypothesis.

In conclusion, this is the first report of a link between ₂-adrenoceptor gene and BHR in a population sample not referred to an asthma clinic. This observation implies that this gene, along with its disease-modifying effect in asthma, does play also an inducing role contributing to the genetic background that predisposes to BHR.