Mutations in the Gene Encoding for the 1J2-adrenergic Receptor inNormal and Asthmatic SubjectsEllen Reihsaus, Michael Innis, Neil MacIntyre, and Stephen B. Liggett

Departments of Medicine (Pulmonary), Molecular Genetics, and Pharmacology, University of CincinnatiCollege of Medicine, Cincinnati, Ohio; Pulmonary Division, Duke University Medical Center, Durham,North Carolina; and Chiron Corporation, Department of Molecular Biology, Emeryville, California

It has long been hypothesized that a defective ti,-adrenergic receptor (ti,AR) may be a pathogenic factorin bronchial asthma. We examined the gene encoding the ti,AR to assess the frequency of polymor­phisms in 51 patients with moderate to severe asthma and 56 normal subjects. Nine different point muta­tions were found in both heterozygous and homozygous forms at nucleic acid residues 46, 79, 100, 252,491,523, 1053, 1098, and 1239. No mutations resulting in large deletions or frame shifts were detected.Of these nine polymorphisms, four were found to cause changes in the encoded amino acids at residues16, 27, 34, and 164. The most frequent polymorphisms were arginine 16 to glycine (Arglo-eOly) andglutamine 27 to glutamic acid (Gln27-Glu). The other two polymorphisms, valine 34 to methionine,and threonine 164 to isoleucine, occurred in only four subjects. The incidence of ti,AR homozygouspolymorphisms was no greater in asthmatic patients as compared with controls (Arglti-t-Gly: 53% versus59%, Gln27-Glu: 24% versus 29%, respectively; P = NS). Some subjects were found to have both ofthese polymorphisms simultaneously, but there was no difference in incidence between the two groups,with 23 % of asthmatics and 28 % of normal subjects being homozygous for both polymorphisms. Theapparently normal subjects with both polymorphisms did not have subclinical hyperreactive airways dis­ease as determined by methacholine challenge testing. In the asthma group, one mutation (ArgI6-Gly)identified a subset of patients with a distinct clinical profile. Patients with this polymorphism were morelikely to be steroid dependent and to require immunization therapy, as compared with those without thispolymorphism. Other parameters of asthma severity or medication use were not related to the presenceof a polymorphism at any locus. Thus, asthma does not appear to be primarily caused by a genetic defectin the ti,AR. However, the Argl6-Gly polymorphism may be associated with a different clinical status,suggesting that an alteration in the gene encoding for the ti,AR plays an accessory role in the pathogene­sis of asthma in certain patients.

In 1968, Szentivanyi hypothesized that one of the underlyingpathogenic mechanisms of asthma may be the imbalance be­tween bronchoconstricting and bronchodilating adrenergicreceptors (1). Although our understanding of asthma now in­cludes roles for a large group of inflammatory componentsand nonadrenergic mediators of bronchomotor tone (2), sev­eral lines of evidence continue to suggest that ti,-adrenergicreceptors (ti,ARs) may be abnormal in asthma. Barnes and

(Received in original form September 10, 1992 and in revised form No­vember 10, 1992)

Address correspondence to: Stephen B. Liggett, M.D., Division of Pulmo­nary/Critical Care Medicine, University of Cincinnati Medical Center, 231Bethesda Ave., Cincinnati, OH 45267-0564.

Abbreviations: arginine, Arg; f3,-adrenergic receptor, f3,AR; glutamine,GIn; glutamic acid, Glu; glycine, Gly; isoleucine, lie; methionine, Met;polymerase chain reaction, PCR; temperature gradient gel electrophoresis,TGGE; threonine, Thr; valine, Val.

Am. J. Respir. Cell Mol. BioI. Vol. 8. pp. 334-339, 1993

colleagues provided the first such evidence, showing a de­crease in ti,AR expression in lungs of guinea pigs with oval­bumin-induced asthma (3), a finding confirmed by severalothers (4, 5). The ti,ARs of lymphocytes, thought to berepresentative of receptors in the less accessible lung, havebeen reported to be either dysfunctional or decreased innumber in patients with asthma as compared with normalcontrols (6, 8). Such observations with the ti,ARs of lym­phocytes have not been universal, though, and in some casesthe alterations in the receptors were found to be due to con­comitant use of tiAR agonists in patients with asthma (9, 10).Inother studies examining in vivo responses to infused adren­ergic agonists, several reports have shown depressed ti-ad­renergic or enhanced a-adrenergic responses in patients withasthma (11, 12). In pulmonary macrophages, ti,AR func­tion has been found to be depressed in patients with asthma(13). In the Basenji-Greyhound dog model of asthma, bron­chial smooth muscle ti,-adrenergic relaxation in responseto agonist has been shown to be depressed in these animals

Reihsaus, Innis, MacIntyre et al.: {32-adrenergic Receptor Mutations in Asthma 335

as compared with mongrel dogs (14). In vitro relaxation ofbronchial smooth muscle in response to {3AR agonists hasbeen reported to be depressed in samples obtained from sub­jects who died of asthma (15). A recent study by Sharma andJeffery, however, with human lungs obtained at autopsy,showed decreases in bronchial epithelium and submucosalgland {32ARs but not bronchial smooth muscle (16). Thelimitations of the above studies include the use of lympho­cyte, macrophage, or peripheral lung {32ARs (as opposed tothe receptors on bronchial smooth muscle), the use of animalmodels of asthma that are not necessarily representative ofasthma in humans, or the lack of direct measurements of{32AR function. A relationship between defects in the {32ARsystem and asthma has not, therefore, been clearly eluci­dated.

The gene encoding the human {32AR has been clonedand sequenced (17). It is an intronless gene that has beenlocalized to q31-q32 of chromosome 5. The deduced aminoacid sequenceconsists of 413 amino acids, with seven clustersof hydrophobic residues thought to represent transmembranespanning domains. The N-terminus is extracellular, contain­ing two sites for asparagine-linked glycosylation. The trans­membrane spanning domains are connected by three ex­tracellular and three intracellular loops. The C-terminus isintracellular. We (18, 19) and others (20, 21) have shown,using site-directed mutagenesis, that mutations involvingsmall regions of the {32AR, including changes of a singleamino acid, can markedly alter the functional properties ofthe receptor. Given the evidence for dysfunctional {32ARs inasthma, and the apparent genetic predisposition for asthma(reviewed in reference 22), we wondered whether mutationsof the {32AR might be present, and a patho-genic factor, insome patients with asthma. We therefore established a highlyaccurate method for screening of the {32AR gene for muta­tions and then studied 56 normal subjects and 51 patientswith moderate to severe asthma.

Materials and MethodsSubjects

Fifty-one unrelated patients with asthma (mean age, 46 yr;range, 23 to 74 yr) were enrolled in this study from the Al­lergy and Pulmonary Clinics at Duke University MedicalCenter. All had asthma for more than 2 yr and were takingat least one medication on a continuous basis. None smokedcigarettes or had occupational exposures known to causeasthma. A chart review and a survey assessing symptomsand anti-asthma medication use were then obtained. Briefly,this survey determined the number of emergency departmentand hospital admissions for asthma treatment, the number oftimes intubated for severe respiratory distress, the presenceof exercise-induced asthma, the presence of allergies oratopic dermatitis, and the use of the following medications:oral steroids (continuous or intermittent), inhaled steroids,{3AR agonists (regular schedule or on an as-needed basis),methylxanthines, cromolyn, methotrexate, and allergy im­munizations.' Blood was then drawn for {32AR gene analy-

1 A detailed description of the patient characteristics is available fromthe corresponding author upon request.

sis as described below. Fifty-six unrelated, age-matched nor­mal subjects acted as controls, who had no known medicalillness or family history of asthma, hypertension, or cardiacor endocrine disorders and were taking no medications. Thepersonnel responsible for patient care, those administeringthe survey, those carrying out methacholine challenge test­ing, and those performing the DNA analysis were blinded asto results pertaining to each of the parties' respective ac­tivities.

Identification of Mutations in the {32AR Gene

Blood (5 ml) from each patient or control subject was drawnin sodium citrate and the lymphocytes were isolated byFicollgradients (23). Genomic DNA was prepared by incubatinglymphocyte pellets in a 0.5% Tween, 50 mM Tris, 1 mMEDTA buffer with 400 1tg/ml proteinase K at 55°C for 1 h.After centrifugation, the supernatant was subjected to 95°Cfor 10 min to denature the proteinase K, and then an aliquotwas utilized for the polymerase chain reaction (PCR). TheDNA encoding for the intronless {32AR gene (17) was thenamplified by PCR using five sets of oligonucleotide primers,each providing products of rv250 bp. The sense primer ofeach set included a 5' GC clamp (24). PCR conditions wereas described (25) using Taq polymerase, with temperaturecycling as follows: 95°C for 1.5 min, 55°C for 1 min, andnoc for 1 min, for 30 cycles. The PCR products were thensubjected to parallel temperature gradient gel electrophore­sis (TGGE) (26). Samples were prepared for TGGE by dena­turing the PCR product at 98°C for 5 min, then renaturationat 50°C for 15 min with the corresponding wild-type DNAderived from PCR of the cloned gene. Electrophoresis of thesamples was then carried out on a 5 % acrylamide, 8 M ureagel over a temperature gradient (26). The temperature gra­dient was optimized for each of the {32AR DNA segments,with a typical gradient being rv35° to rv60°C and requiring4 to 6 h of run time. A control wild-type was included in eachrun. Preliminary experiments using site-directed mutated{32AR cDNA showed that these methods could detect singlebase changes in the {32AR sequence, which were exhibitedby the presence of multiple bands after TGGE, as comparedwith a single band that occurred with nonmutated sequence(see RESULTS). Repeated PCRs followed by TGGE showedthat these techniques gave reproducible results. All samplesfound to be positive for mutations byTGGE were then se­quenced by the dideoxy method (27) to determine the muta­tions present. The sequence encoding for the N-terminalportion of the {32AR (amino acids 1-50) was not amenableto TGGE, and the PCR products representing this regionwere sequenced directly. PCR reactions intended to provideproducts for sequencing were performed with biotinylatedprimers, and single-stranded DNA was isolated using mag­netic streptavidin beads (28). For the purpose of thesestudies, the normal (wild-type) sequence of the human{32AR gene is considered as published in the originaldescription of the cloning of the receptor by Kobilka and col­leagues (17). However, as will be shown, the frequency ofsome polymorphisms in normal subjects is sufficiently highto suggest that at some loci there is no consensus wild-typesequence.

336 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 8 1993

Methacholine Challenges

Some of the normal subjects who were found to have 132ARmutations were studied with methacholine challenges to as­sess nonspecific airway hyperreactivity using the standardAmerican Thoracic Society protocol (29). The study wasconsidered positive if sRaw increased by 66% or sRaw de­creased by 45 %.

Miscellaneous

The relationshipsbetween mutations, the presence of asthma,and clinical characteristics were assessed by analysis offrequency distribution by the chi-square test of indepen­dence, with significance imparted at the P <0.05 level. Anal­ysis was performed using software from the SAS Institute(Cary, NC). Oligonucleotides were made on an Applied Bio­systems oligonucleotide synthesizer. Taq polymerase wasfrom Perkin-Elmer Cetus (Norwalk, CT). [25S]dATP wasfrom New England Nuclear (Boston, MA). Sequenasewas from U.S. Biochemical (Cleveland, OH). Streptavidinmagnetic beads were from Dynal (Great Neck, NY). Otherreagents were from standard commercial sources as listedelsewhere (19, 25).

Results and DiscussionShown in Figure I are the results of a typical TGGE reveal­ing the presence of mutations of the 132AR gene from sev­eral subjects, and a representative sequencing gel identifyinga mutation and delineating the homozygous and heterozy­gous forms. In all, nine different mutations were detected,at nucleic acid residues 46, 79, 100, 252, 491, 523, 1053,1098, and 1239 (Figure 2 and Table 1). These occurred withrelatively high frequency and are henceforth referred to aspolymorphisms. Five of these, contained within the tripletcodons for amino acids 84, 175, 351, 366, and 413 (Figure2, darkened circles) did not change the encoded amino acids.

These occurred with a frequency as high as 51%, were notdistributed differently between asthma patients and normalsubjects (Table 1, upper section), and are not further dis­cussed. In subjects with asthma, three polymorphisms of the132AR gene were found that did change the encoded aminoacid sequence. These occurred both in the heterozygous andhomozygous forms. As shown in Table 1, the most commonpolymorphisms in asthma were found to cause changes inamino acid 16, from arginine to glycine (Argle-eGly), andat amino acid 27, changing glutamine to glutamic acid(Gln27-+Glu). Of the 51 asthma patients, 53% were foundto be homozygous for ArgI6-+Gly, and 24% were homozy­gous for Gln27-+Glu. Twelve patients (24%) were homozy­gous for both polymorphisms occurring simultaneously.One patient was heterozygous for a mutation at amino acid34 changing valine to methionine.

As shown in Table 1 (lower panel), with the normal sub­jects we found no difference between the frequency of poly­morphisms at amino acids 16 and 27 (both heterozygous andhomozygous forms), as compared with those with asthma(P = NS). One additional mutation was observed in the nor­mal group, where amino acid 164 was changed from threo­nine to isoleucine. This occurred in the heterozygous formin three subjects. To assess whether the apparently normalsubjects who had changes in amino acids 16 or 27 may havehad subclinical abnormal pulmonary mechanics or hyper­reactive airway disease, 10 subjects who were homozygousfor both Arglo-eGly and Gln27-+Glu underwent pulmonaryfunction and methacholine challenge testing. All of thesesubjects were found to have normal baseline lung mechanics.Only one of these subjects had hyperreactive airways aftermethacholine challenge.

We also considered whether asthmatics who had muta­tions of the 132AR might have different clinical characteris­tics as compared with those who did not. We recognize thatthe varied clinical practices of the physicians caring for these

SAMPLE:

CTRL

, 1 234567

TGGE

I I I I I IGATCGATCGATC

SEQUENCING

Figure1. Representative TGGEand sequencing results frompeR-amplified portions of the{J2AR gene. Shown are the re­sults from seven samples sub­jected to TGGE screening formutations. Samples 2,3,5, and 6displayed multiple bands, whichidentifies the presence of mu­tations. Single bands (samples 1and 7) represent the wild-typesequence. The arrows in the se­quencing gel show the presenceof an adenine (A) at nucleic acidposition 46 of the wild-type{J2ARsequence, an adenine and a gua­nine (G) in this position repre­senting the heterologous form ofthe Arg16-+Gly polymorphism,and a guanine in the homologousform of the polymorphism.

Reihsaus, Innis, MacIntyre et al.: {32-adrenergic Receptor Mutations in Asthma

Figure 2. Localization of poly­morphisms in the human f32AR.

The deduced amino acid sequenceand proposed membrane topol­ogy of the human f32AR isshown, with the location of poly­morphisms that caused aminoacid changes denoted. A dark­ened circle indicates the positionwhere a nucleic acid within a trip­let codon was found to be mu­tated, but the encoded amino acidis not changed.

366

HVE

6SQEGTNGNSSYGN

337

Q EkE N K L LeE 0 LPG TED F VG

m ~COOH LSDNTSCNRSQSDINDSPVT G

patient make such an assessment difficult. Also, the asth­matic patients that we have chosen represent a heterogenousgroup. Given these caveats, analysis showed no associationbetween the presence of any polymorphism and the age orsex of the patient, the presence of exercise-induced asthma,the number of emergency department visits, or the numberof hospitalizations for the treatment of asthma. We alsohypothesized that an increase in the use of therapeutic agentsin patients who were homozygous for the polymorphismsmight occur. There were no differences in the use of {3-ad­renergic agonists, cromolyn, or methylxanthines in patientswith any of the homozygous polymorphisms. However, inthose with Arg16-+Gly, there was an association between thepresence of this polymorphism and the use of corticosteroids

and immunotherapy. Of those patients that required con­tinuous oral steroids, 75% were homozygous for Arglo-eGly(P < 0.05). The presence of any oral corticosteroid use (in­termittent or continuous) was also related to the presence ofthis mutation. Of those requiring either regimen, 64% werehomozygous for Arg16-+Gly (P < 0.03). Patients with thismutation also were clearly segregated regarding the need forimmunization therapy. Of those who required such, all(l00 %) had this mutation in the heterozygous or homozy­gous forms. For heterologous mutations, we considered thatif one allele was sufficient to cause a clinical effect, that effectwould be seen in the entire population of patients (homozy­gotes and heterozygotes) that had the polymorphism. Usingthis criterion, patients with polymorphisms of a single allele

TABLE 1

Distribution of {32AR gene polymorphisms in normal and asthmatic subjects*

Mutated Asthma (n = 51) Normal (n = 56)

Designation Nucleic Acid Heterozygous Homozygous Heterozygous Homozygous

No amino acid changeLeu84 252 18 2 18 4Arg175 523 15 2 16 4G1y351 1053 19 3 22 5Tyr366 1098 0 0 1 0Leu413 1239 21 6 26 7

Amino acid changesArg16-+Gly 46 19 27 16 33G1n27-+Glu 79 26 12 23 16Val34-+Met 100 1 0 0 0Thr164-+Ile 491 0 0 3 0Arg16-+Gly + 46 + 79GIn27-+Glu 12 16

* Shown are the number of asthma patients or normal subjects who had each of the indicated po1ymorphisms.

338 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 8 1993

were not associated with a different clinical status or medica­tion profile as compared with all asthmatics. Some patientshad not only one or two nonsilent mutations but also severalsilent mutations as well. The total number of mutations perpatient was not found to correlate with any of the above clini­cal parameters.

While a number of G-protein-coupled receptors havebeen cloned, very little is known about the frequency of poly­morphisms in these genes. The current study provides thefirst data on the frequency of polymorphisms of any hormone/neurotransmitter G-protein-coupled receptor gene to date.We found nine different nucleic acid mutations (Table 1).Five cause no change in the encoded amino acids. Thesepolymorphisms occurred primarily in the heterozygous form,and are distributed throughout the receptor coding block(Figure 2). On the other hand, the majority of polymor­phisms that do cause changes in the amino acid sequence oc­curred in the N-terminus region, at amino acids 16 and 27.Mutations altering amino acids 34 (1 asthmatic) and 164 (3normal subjects) were infrequent and were only found in theheterozygous form. We found no difference between the fre­quency of polymorphisms of the 112AR gene in the asthmapopulation as compared with the normal group. Indeed,the percentage of homozygous and heterozygous polymor­phisms at any locus was virtually identical between asth­matics and normal subjects (Table 1). We considered two ad­ditional factors: (1) that some apparently normal subjectswith polymorphisms had subclinical abnormal lung functionor hyperreactive airways, and (2) that those asthmatics whohad polymorphisms had different clinical profiles than asth­matics with the normal 112AR genotype. Regarding the firstpoint, 10 of 10 normal subjects who were homozygous forthe simultaneous presence of Arglo-eGly and Gln27-Gluhad normal baseline lung function, and only one of the 10had clearly hyperreactive airways. Therefore, we concludethat given the nearly identical frequency of these polymor­phisms in the normal and asthmatic populations, they do notconfer asthmatic pathophysiology. We also assessed the clin­ical features of asthma patients to determine if 112AR poly­morphisms might be associated with certain characteristics.We found an association between those patients who had theArglo-eGly polymorphisms and the use of corticosteroidsand immunization therapy. Further studies with well-definedgroups of asthmatic patients will be necessary to furtherevaluate whether this common polymorphism is associatedwith a specific asthmatic phenotype. One might speculatethat this mutation may impart some dysfunction in the 112ARwhich is not clinically apparent in those without asthma.But, in those with asthma, such a defect may result in a lackof benefit from I1AR agonist therapy, thus necessitatingoral corticosteroids and/or immunotherapy.

It is interesting to compare the frequency of 112AR genepolymorphisms that we have found here with those of thegene encoding for rhodopsin, the "receptor" for light, whichis the only other G-protein-coupled receptor gene in whichthe frequency of mutations has been examined extensively(30, 31). In patients with autosomal dominant retinitis pig­mentosa, which has a clearly defined phenotype and geneticinheritance pattern, 18 to 24% have polymorphisms of therhodopsin gene. The most common polymorphism in therhodopsin gene causing this disorder occurs at amino acid

23, which is in a similar position (the extracellular N-termi­nal domain) as the two most frequent polymorphisms that wefound in the current study with the 112AR. Unlike what wefound with the 112-adrenergic receptor in asthma, though,there is a zero percent incidence of polymorphisms in therhodopsin gene that cause amino acid changes in normal(i.e., those without retinitis pigmentosa) subjects. Thus, thefrequency of background polymorphisms in apparently nor­mal subjects of the 112AR is significantly higher than that ofthe related G-protein-coupled receptor rhodopsin. Less isknown about the role of the N-terminal extracellular domainof the 112ARas compared with other regions that have beenstudied extensively. Dixon and colleagues (32) have shownthat mutant receptors lacking amino acids 21-30 were poorlyprocessed by the cell to the full 67 kD protein. In anothermutation, where amino acid residues 6-15 were deleted, noimmunoreactive protein of the mature receptor wasdetected.Subsequent studies where the sites ofN-linked glycosylationin the extracellular domain were specifically mutated (resi­dues 6 and 15) have shown important roles for these sites incellular processing of the receptor (33). The most commonpolymorphisms that we have found are adjacent to one of theglycosylation sites or in the other region found by Dixon andcolleagues (32) to be important for cellular processing (resi­due 27).

Our current findings have resolved a long-standing con­troversy regarding dysfunctional 112ARs in asthma. Prior tothis study, it has not been known whether such dysfunctionis due to abnormal (i.e., mutated) receptors or whetherthe dysfunction is acquired, due to the multiple regulatorymechanisms that are known to occur with the 112AR. Wehave now definitively shown that the frequency and distribu­tion of polymorphisms of the 112AR gene are no higher inpatients with asthma than that of the normal population.Thus asthma does not appear to be primarily caused by agenetic defect in the 112AR. Interestingly, though, one mu­tation of this receptor may portend a different clinical profileand thus may play an accessory role in the pathogenesis ofasthma.

Acknowledgments:We wish to thank Drs. Peter Kussin, Herbert Saltzman, andEdward Buckley for allowing us to study their patients, and Brian Holt, AndreaWilkes, and Judy Carlsen for their excellent technical assistance. Statistical anal­ysis was provided by D. F. Harris Associates, Inc. (Chapel Hill, NC). This workwas supported in part by NIH Grant HL45967 to S.B.L.

References

1. Szentivanyi, A. 1968. The beta-adrenergic theory ofthe atopic abnormalityin bronchial asthma. J. Allergy 42:203-232.

2. Barnes, P. J. 1989. A new approach to the treatment of asthma. N. Engl.J. Med. 321:1517-1527.

3. Barnes, P. J., C. T. Dollery, and J. MacDermot. 1980. Increased pulmo­nary a-adrenergic and reduced l3-adrenergic receptors in experimentalasthma. Nature 285:569-571.

4. Gatto, C., T. P. Green, M. G. Johnson, R. P. Marchessault, V. Seybold,and D. E. Johnson. 1987. Localization of quantitative changes in pulmo­nary beta-receptors in ovalbumin-sensitized guinea pigs. Am. Rev.Respir. Dis. 136:150-154.

5. Cheng, J. B., and R. G. Townley. 1982. Effects of chronic histamine andovalbumin aerosols on pulmonary beta adrenergic receptors in sensitizedguinea pigs. Res. Commun. Chem. Pathol. Pharmacol. 36:507-510.

6. Parker, C. W., and J. W. Smith. 1973. Alterations in cAMP metabolismin bronchial asthma. J. Clin. Invest. 52:48-59.

7. Sano, Y., G. Watt, and R. G. Townley. 1983. Decreased mononuclear cellbeta-adrenergic receptors in bronchial asthma: parallel studies of lympho­cyte and granulocyte desensitization. J. Allergy Clin. Immunol. 72:495-503.

Reihsaus, Innis, MacIntyre et al.: ,62-adrenergic Receptor Mutations in Asthma 339

8. Kariman, K. 1980. ~-adrenergic receptor binding in lymphocytes from pa­tients with asthma. Lung 158:41-51.

9. Galant, S. P., L. Duriseti, S. Underwood, S. Allred, and P. A. Insel. 1980.Beta-adrenergic receptors of polymorphonuclear particulates in bronchialasthma. J. Clin. Invest. 65 :577-585.

10. Conolly, M. E., and J. K. Greenacre. 1976. The lymphocyte ~-adrenore­

ceptor in normal subjects and patients with bronchial asthma. J. Clin. In­vest. 58:1307-1316.

II. Kaliner, M., J. H. Shelhamer, P. B. Davis, L. J. Smith, and J. C. Venter.1982. Autonomic nervous system abnormalities and allergy. Ann. Int.Med. 96:349-357.

12. Middleton, E., and S. R. Fink. 1968. Metabolic responses to epinephrinein bronchial asthma. J. Allergy 42:288-299.

13. Bachelet, M., D. Vincent, N. Havet, R. Marrash-Chahla, A. Pradalier, J.Dry, and B. B. Vargaftig. 1991. Reduced responsiveness of adenylate cy­clase in alveolar macrophages from patients with asthma. J. Allergy Clin.Immunol. 88:322-328.

14. Tobias, J. D., R. A. Sauder, and C. A. Hirshman. 1990. Reduced sensitivityof beta-adrenergic agonists in Basenji-Greyhound dogs. J. Appl. Physiol.69:1212-1219.

15. Goldie, R. G., D. Spina, P. J. Henry, K. M. Lulich, and J. W. Paterson.1986. In vitro responsiveness of human asthmatic bronchus to carbachol,histamine, ~-adrenoceptor agonists and theophylline. Br. J. Clin. Phar­macol. 22:669-676.

16. Sharma, R. K., and P. K. Jeffery. 1990. Airway ~-adrenoceptor number incystic fibrosis and asthma. Clin. Sci. 78:409-417.

17. Kobilka, B. K., R. A. E Dixon, T. Frielle, H. G. Dohlman, M. A.Bolanowski, I. S. Sigal, T. L. Yang-Feng, U. Francke, M. G. Caron, andR. J. Lefkowitz. 1987. eDNA for the human ~2-adrenergic receptor: aprotein with multiple membrane-spanning domains and encoded by a genewhose chromosomal location is shared with that of the receptor forplatelet-derived growth factor. Proc. Nat!. Acad. Sci. USA 84:46-50.

18. Liggett, S. B., M. Bouvier, B. E O'Dowd, M. G. Caron, R. J. Lefkowitz,and A. DeBlasi. 1989. Substitution of an extracellular cysteine in the ~2­

adrenergic receptor enhances receptor phosphorylation and desensitiza­tion. Biochem. Biophys. Res. Commun. 165:257-263.

19. Liggett, S. B., M. G. Caron, R. J. Lefkowitz, and M. Hnatowich. 1991.Coupling of a mutated form of the human ~2-adrenergic receptor to G,and Gs: requirements for multiple cytoplasmic domains in the couplingprocess. J. Bio!. Chem. 266:4816-4821.

20. Strader, C. D., I. S. Sigal, R. B. Register, M. R. Candelore, E. Rands, andR. A. E Dixon. 1987. Identification of residues required for ligand bindingto the ~-adrenergic receptor. Proc. Nat!' Acad. Sci. USA 84:4384-4388.

21. Fraser, C. M., E Z. Chung, C. D. Wang, and J. C. Venter. 1988. Site-

directed mutagenesis of human ~-adrenergic receptors: substitution ofaspartic acid-130 by asparagine produces a receptor with high-affinityagonist binding that is uncoupled from adenylate cyclase. Proc. Natl.Acad. Sci. USA 85:5478-5482.

22. Levitt, R. c., W. Mitzner, and S. R. Kleeberger. 1990. A genetic approachto the study oflung physiology: understanding biological variability in air­way responsiveness. Am. J. Physiol. 258:Ll57-Ll64.

23. Kellogg, D. E., and S. Kwok. 1990. Detection of human immunodeficiencyvirus. In PCR Protocols: A Guide to Methods and Applications. M. A.Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White, editors. AcademicPress, Orlando, FL. 341-342.

24. Sheffeld, V. C., D. R. Cox, L. R. Lerman, and R. M. Myers. 1989. Attach­ment of a 40-base-pair G+C-rich sequence (GC-clamp) to genomic DNAfragments by the polymerase chain reaction results in improved detectionof single-base changes. Proc. Natl. Acad. Sci. USA 86:232-236.

25. Innis, M A., and D. H. Gelfand. 1990. Optimization of PCRs. In PCR Pro­tocols: A Guide to Methods and Applications. M. A. Innis, D. H. Gel­fand, J. J. Sninsky, and T. J. White, editors. Academic Press, Orlando, FL.1-12.

26. Wartell, R. M., S. H. Mosseini, and C. P. Moran Jr. 1990. Detecting basepair substitutions in DNA fragments by temperature-gradient gel elec­trophoresis. Nucleic Acids Res. 18:2699-2706.

27. Sanger, E, S. Nicklen, and A. R. Coulson. 1977. DNA sequencing withchain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467.

28. Hultman, T., S. Stahl, E. Homes, and M. Uhlen. 1989. Direct solid phasesequencing of genomic and plasmid DNA using magnetic beads as solidsupport. Nucleic Acids Res. 17:4937-4946.

29. Guideline for bronchial inhalation challenges with pharmacologic and anti­genic agents. Position paper, American Thoracic Society, ATSNews, 1980;11-18.

30. Dryja, T. P., T. L. McGee, B. A. Lauri, L. B. Hahn, G. S. Cowley, 1. E.Olsson, E. Reichel, M. A. Sandberg, and E. L. Berson. 1990. Mutationswithin the rhodopsin gene in patients with autosomal dominant retinitispigmentosa. N. Eng!. J. Med. 323:1302-1307.

31. Sung, C. H., C. M. Davenport, J. C. Hennessey, I. H. Maumenee, S. G.Jacobson, J. R. Heckenlively, R. Nowakowski, G. Fishman, P. Gouras,and J. Nathans. 1991. Rhodopsin mutations in autosomal dominant retini­tis pigmentosa. Proc. Nat!' Acad. Sci. USA 88:6481-6485.

32. Dixon, R. A. F., S. S. Irving. M. R. Candelor, R. B. Register, W. Scatter­good, E. Rands, andC. D. Strader. 1987. Structural features required forligand binding to the ~-adrenergic receptor. EMBO J. 6:3269-3275.

33. Rands, E., M. R. Candelore, A. H. Cheung, W. S. Hill, C. D. Strader, andR. A. E Dixon. 1990. Mutational analysis of ~-adrenergic receptorglycosylation. J. Bio!. Chem. 265:10759-10764.