Ž .Clinica Chimica Acta 308 2001 17–24www.elsevier.comrlocaterclinchim

Review article

Automated detection of single nucleotide polymorphism in b-2adrenergic receptor gene using LCxw

Hong Yu a,), Barbara Merchant a, Christi Scheffel a, Chandri Yandava b,Jeffrey M. Drazen b, Jeffrey Huff a

a Department 9NA, AP 20, Abbott Diagnostic DiÕision, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064, USAb Brigham and Women’s Hospital, Boston, MA, USA

Received 8 May 2000; accepted 23 January 2001

Abstract

Ž .b- adrenergic receptor B2AR agonists are the most widely prescribed rescue agents used in the treatment of asthma.2

Recent studies have indicated a relationship between a polymorphism at codon 16 of the B2AR gene, and the response torecurrent b-agonist therapy. The B2AR polymorphism of interest involves a single nucleotide change from A to G, resulting

Ž . Ž .in an amino acid change from Arginine Arg to Glycine Gly . Clinical efforts to further investigate this relationship requirean accurate, reliable and inexpensive method for detecting the polymorphism.

In this study, we report an LCxw assay for the detection of a single nucleotide polymorphism at codon 16 of the b- 2

adrenergic receptor. This assay is capable of detecting patients harboring any of the three possible genotypes at this locus,namely, homozygous wild type, homozygous variant or heterozygous individuals with a single genomic DNA sample of25–500 ng. It requires minimum hands-on time with automated detection. The assay would be suitable for use in researchlabs for screening of a large number of samples. We believe that this type of assay will facilitate research and clinicalinvestigations in elucidating the association of SNPs with disease states, diagnosis, prognosis and treatment. q 2001 ElsevierScience B.V. All rights reserved.

Keywords: Genotyping; Polymerase chain reaction; Asthma; Adrenergic receptor; Polymorphism

1. Introduction

Adrenergic receptors mediate catecholaminetransmission and regulate a vast number of biochem-

Ž .ical processes. The b- adrenergic receptor B2AR2

is a major subtype expressed in the smooth muscleof vascular, gastrointestinal, genito-urinary, skeletal

w xand bronchial tissues 1 . b -agonists are a class of2

) Corresponding author. Tel.: q1-847-938-1109; fax: q1-847-938-8777.

Ž .E-mail address: judy.yu@add.ssw.abbott.com H. Yu .

pharmaceutical compounds that activate the B2AR,leading to the relaxation of the bronchial smoothmuscle and airway dilation.

b -agonists are the most widely prescribed rescue2

medication for the treatment of asthma. Clinicalguidelines currently recommend the use of b -2

agonists only on an as-needed basis, but as patientshave disease exacerbations, they are likely to use this

w xform of treatment many times a day 2 . Polymor-phisms, especially the polymorphism at codon 16, ofB2AR gene are believed to play a role in the variable

w xresponse to b -agonists 3–7 . The codon 16 poly-2

0009-8981r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0009-8981 01 00421-1

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–2418

morphism involves a single nucleotide change fromA to G, resulting in an amino acid change from

Ž . Ž .Arginine Arg to Glycine Gly at codon 16. Clini-cal efforts to further investigate this relationshiprequire an accurate and, most preferably, an auto-mated method for the detection of the polymorphism.

Commonly used genotyping methods include al-Ž .lele-specific PCR AS-PCR or amplification refrac-

Ž .tory mutation systems ARMS , restriction fragmentŽ .length polymorphism RFLP and sequencing. All of

these methods are manual and tedious to perform.Such manual methods are especially prohibitive whenscreening large numbers of samples. We have devel-oped a semi-automated, sensitive and specific methodto detect the single nucleotide change using theAbbott LCxw instrument. The LCxw assay requiresminimal hands-on time and has a low potential forcross-contamination of samples.

The Abbott LCxw instrument system was devel-oped for DNA probe-based assays. The samplepreparation and the amplification reaction setup re-quire minimum hands-on time and detection is com-pletely automated after loading the amplificationproduct into the LCxw analyzer. The B2AR assayutilizes the polymerase chain reaction to amplify aspecific region of the human b-2 adrenergic receptorgene. The assay is a duplex assay yielding tworeadings per each test sample. One reading corre-

Ž .sponds to the wild type A sequence, and the other,Ž .the variant type G sequence. Therefore, the assay

affords the assignment of homozygous wild type,heterozygous or homozygous variant from a singlealiquot of test sample.

This new LCxw assay, as described in this paper,enables the reduction of hands-on time, labor and thecomplexity involved in the detection of single nu-

Ž .cleotide polymorphisms SNPs . Thus, it is antici-pated that the development of this assay could greatlyfacilitate the effort in clinical utility studies on hu-man SNPs and their relationship with disease diag-nosis, prognosis and response to therapy.

2. Materials and methods

2.1. DNA isolation

ŽFresh blood samples received within 24 h at.room temperature were ordered from Interstate

Blood Bank. The blood tubes were bar code labeled,without any donor identity. DNA was extracted fromwhole blood using several different commercial kits.

w Ž .Puregene by Gentra Systems Minneapolis, MNw Žand QIAamp Blood Maxi kit by Qiagen Valencia,

.CA were used to extract genomic DNA from up to10 ml of whole blood. Gentra Generationw andQIAampw Blood Mini kits were also used for DNApreparation from 200 ml of whole blood. See themanufacturer’s instructions for detailed procedures.

2.2. Sequencing

PCR was performed and the resultant, a 368base-pair amplicon, was prepared from 100 ng DNA

Ž .isolated from fresh blood as described above usinga forward and a reverse primer at 200 nM concentra-tion each. The reaction was carried out in 0.2-mlPerkin Elmer microAMP tubes at 50 ml total vol-ume. The 1= PCR buffer consisted of the follow-ing: 1.5 mM MgCl , 100 uM dNTPs, 0.4 Urml2

AmpliTaq Gold. Cycling parameters were: meltingfor 30 s at 958C, annealing for 20 s at 558C, exten-sion for 30 s at 728C, for a total of 30 cycles. ThePCR product was checked by agarose gel elec-trophoresis. The amplicon was purified by passingthe reaction product through a MicroSpine S-400HR pre-packed resin column equilibrated in TE bufferŽ .pH 7.6 and the concentration was calculated basedon the absorbance at 260 nm.

Sequencing was performed on ABI Prisme 377DNA sequencer using Big Dyee Terminator chem-istry. Templates were sequenced with reagents fromABI Ready Reaction Kit, according to the kit man-ual. Briefly, 100 ng purified PCR template wasadded to a total of 20 ml, containing 3.2 pmolprimers, 8 ml Termination Ready Reaction Mix withAmpliTaq FS. The mixture was subjected to 25rounds of cycle sequencing on an MJ ResearchPTC-225 Peltier Thermal Cycler. Unincorporated BigDye nucleotides were removed by passing the se-quencing reaction mix over a AntoSeq G-50 spincolumn.

2.3. ARMS

The amplification–refractory mutation systemŽ .ARMS used for the detection of mutation at the

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–24 19

amino acid 16, human b-2-adrenergic receptor genewas carried out according to the methods described

w xearlier 8,9 . The primers used for detecting theŽmutation, A™G at nucleotide 1633 Genbank ac-

.cession no. M15169 corresponding to the aminoŽ .acid change at position 16, Arg™Gly R™G were:

Ž XArg-specific forward primer A1 5 -GCCTTCTTGCX .TGGCACCCAA AA-3 , corresponding to nu-

cleotides 1612–1633, except the penultimate base atX Ž .the 3 end underlined was changed from T to A,

Ž XGly-specific forward primer A2 5 -GCCTTCTTGCX .TGGCACCCAA AG-3 , differs from the Arg-

specific primer at the last nucleotide at the 3X endŽ . Ž Xshown in bold , and the reverse primer 5 -

X.CAGACGCTCGAACTTGGCCATG-3 . The under-lined base was modified from the published se-quence.

Amplification by PCR of the genomic DNA ofeach sample included two separate reactions for eachassay: one with primers A1 and reverse primer forwild type, and the other, with primers A2 and re-verse primer for mutation type. Each PCR reaction

Ž .contained 5-ml template genomic DNA 250 ng ,Ž .PCR buffer Boehringer-Mannheim , 1.5 mM MgCl2

Ž .Promega , 12.5 pmol of each primer, 200 mMŽ .dNTPs Pharmacia , 1.5 U of Taq polymerase

Ž .Boehringer-Mannheim , 0.05 U of Perfect Match,Ž .PCR enhancer Stratagene in a total volume of 25

ml. Conditions for PCR were: an initial hot startperiod of 5 min at 948C, and temperature was held at808C after the hot start, and during this period,dNTPs and Taq polymerase were added. This wasfollowed by 35 cycles for 1 min at 948C, 1 min at628C, 1 min at 728C, with a final extension time of 5min at 728C. Thin-walled, 96-micro-well platesŽ .Costar with mineral oil were used for the amplifi-

Žcation reactions in a PTC-100 thermal cycler MJ.Research, Watertown, MA . After amplification,

about 20 ml of the reaction mixture was resolved byelectrophoresis on a 2.0% agarose gel and stainedwith ethidium bromide for analysis.

2.4. RFLP

Genotyping at B2AR-16 was carried out accord-w xing to the method described by Martinez et al. 6 .

The b AR-16 containing region was amplified using2

the primers: 5X-GCCTTCTTGCTGGCACCCCAT-3X

and 5X-CAGACGCTCGAACTTGGCCATG-3X. Theunderlined bases were modified from the reportedsequence to create NcoI site. The reverse primercontains an NcoI site and digests products from bothalleles. The PCR product from the Gly-16 allelecontains the NcoI site in the 5X-primer. The PCRreaction conditions were similar to the conditionsdescribed for B2AR-27, except a total reaction vol-ume of 25 ml was used.

For the restriction enzyme digestion, 10 ml of thePCR product was digested with 2 U of NcoI and

Ž .Buffer 4 New England Biolabs, Boston for 2 h at378C. The PCR products were resolved by elec-trophoresis on a 4% NuSieve agarose gel and stainedwith ethidium bromide. The uncut PCR product was167 bp in length. The enzyme digests the PCRproduct to produce a 145-bp fragment from the16-Arg allele and a 127-bp fragment from the 16-Glyallele.

2.5. PositiÕe control cloning

A homozygous wild type and homozygous varianthuman genomic DNA samples were amplified withthe appropriate primer sets. The amplification prod-uct was purified using the QIAquick PCR Purifica-

Ž .tion Kit Qiagen, CA , following the manufacture’sinstructions. Purified PCR products were analyzedon 1% agarose gel before proceeding to ligation.Ligation was performed using 1 mg of purified PCR

Žproduct and Sureclone Ligation Kit Amersham.Pharmacia , following the manufacturer’s instruc-

tions. The ligated DNA was transformed into Super-Ž .competent XL1 Blue cells Stratagene, CA using the

heat shock procedure provided by the manufacturer.ŽCells were plated on LB ampicillin plates Microdi-

.agnostics and white colonies were selected. Colonieswere screened by boiling the colonies, followed byPCR, using vector primers. The solution containingboiled colonies were also tested and confirmed usingthe LCxw assay. One hundred milliliters of culturewere then grown for colonies containing the insert,and the plasmid purified using Qiagen Plasmid MaxiPurification Kit, following the manufacturer’s in-struction. The purified plasmids were confirmed bysequencing as well as with the LCxw assay.

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–2420

2.6. NegatiÕe control

The negative control contains 5 mM EPPS buffer,2.5 mM KOH, 2.5 ngrml Salmon Testes DNA, 11mM EDTA and 15.4 mM NaN .3

2.7. LCX w OH–PCR

The OH–PCR reaction mixture contains 1= OHŽ .buffer Abbott , 150 mM of each of the dNTPs

Ž .made from 10 mM stock from Pharmacia , 110 nMŽof each of the forward and reverse primers forward

primer: AAC GGC AGC GCC TTC TTG C, reverse.primer: ACA TGA CGA TGC CCA TGC C , 110

ŽnM wild type probe Adamantane-TTTTTTTTTT-.CAATAGAAGCCATGC and 50 nM mutant probe

Ž .dansyl-TTTTTTTTTT-CC CAA TGG AAG CC , 5U of rTth DNA polymerase and 3.25 mM MnCl .2

Ž .The OH–PCR reaction mixture 180 ml is aliquotedinto each LCxw reaction tube and 20 ml of the testsample is added, resulting in a total reaction volumeof 200 mlrtest sample. OH–PCR is then carried outin LCxw cyclers. The initial melt is at 978C for 2min, followed by 45 cycles at 948C for 40 s, 558Cfor 40 s and 728C for 40 s. After 45 cycles, thereactions are melted again at 978C for 5 min andthen allowed to drop down to 128C rapidly and thesamples are incubated for at least 10 min at 128C.Oligo probe hybridization occurs at the last melt-and-cool down cycle.

2.8. LCX w detection

After OH–PCR, the tubes are taken from thecycler and spun down in a microcentrifuge for 10 s.The tubes are then transferred without opening thelids to the LCxw analyzer for automated detectionŽreagents and instrumentation provided by Abbott

.Labs, IL . The results are printed out at the end ofthe detection.

3. Results and discussion

3.1. LCX w genotyping assay principle

The B2AR LCxw assay utilizes polymerase chainreaction to amplify a specific region of human b-2

adrenergic receptor gene. The assay mixture containsall the reagents necessary for PCR, including buffer,dNTP, rTth DNA polymerase, Mn2q, a pair ofprimers and a pair of probes, one being the wild typeprobe, and the other, the variant probe, each labeledwith a distinct hapten. The primers, and therefore theamplified products, carry a hapten that is recognizedby an antibody attached to the MEIA microparticles.Thus, the amplified product, as well as excessprimers, will be captured on the microparticles.

Ž . Ž .Probes with wild type A andror variant G se-quences hybridize to the amplified target dependingupon whether the target sequence is a homozygous

Ž . Ž .wild type AA , heterozygous AG or homozygousŽ .variant GG . The amplified targetrprobe complex

serves as the detection product in the LCxw ana-lyzer. The wild type and the variant probes arelabeled with different haptens that are recognized byunique antibodies bioconjugated to alkaline phos-phatase and b-galactocidase, respectively. Alkalinephosphatase catalyzes a hydrolytic reaction with the

Ž .substrate methylumbelliferyl phosphate MUP toyield a fluorescent product, methylumbelliferone, andb-galatocidase catalyzes a deglycosylation of the

Ž . Žsubstrate AUG N- 2-hydroxyethyl -2- 7-b-D-galac-.topyranosyloxy-2-oxo-2H-chromen-4-yl acetamide,

.US Patent 5,998,593 to yield a coumarin-basedproduct in sequential fashion. Therefore, a homozy-gous wild type target yields a positive MUP signaland homozygous variant a positive AUG signal,whereas heterozygous sample yields positive signalsin both MUP and AUG channels, as illustrated inFig. 1. The assay is considered a duplex assay in thatthe output per sample has dual readings, an MUPsubstrate reading and an AUG substrate reading. Theduplex nature of the assay enables the assignment ofhomozygous wild type, heterozygous or homozygousvariant, with only one aliquot of a test sample in asingle tube.

3.2. Sample preparation

Several commercial DNA extraction kits havebeen used with the assay, including Gentra Pure-genew , Generationw kit and the Qiagen Qiaampw

Maxi and Mini blood kit. The Puregenew andQiaampw Maxi kits can be used for archiving large

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–24 21

Fig. 1. One hundred nanograms of three known genotype sampleswere applied to LCxw assay following the procedure in Section 2.

Ž .Each sample has dual readings, MUP slashed bar and AUGŽ .dotted bar . Wild type sample shows only the positive MUPsignal with minimum background AUG readings, and vice versa

Ž .for the variant sample MUT . The heterozygous sample showsboth positive readings.

amounts of DNA extracted from up to 10 ml ofwhole blood. The Puregenew procedure is carriedout with bench top centrifuges and the Qiaampw

Maxi procedure requires an ultra-high-speed cen-Ž .trifuge 4500=g . Generally, the archived DNA

samples are stored at known concentrations. Theassay protocol designates 20 ml total target volumeto be added to the reaction mixture. One hundrednanograms of DNA is recommended for the assay.

ŽConsequently, target dilution of 100 ngr20 ml 5.ngrml is desirable. However, target amounts rang-

ing from 25 to 500 ng were all shown to be accept-able. Both DNA isolation procedures are easy to

Žperform and 20 ml of the DNA eluent both manu-facturer’s procedures elute off DNA in 200 ml of

.elution buffer can be used directly in the assaywithout further dilution or concentration. All theabove methods provided virtually equivalent LCxw

Ž .results in our assay data not shown .

3.3. Signal to noise

The signal to noise, or discrimination ratio, forthis assay is defined as the probe signal from match-ing target versus that from the mis-matched target.Specifically, MUP channel signal to noise is definedhere as LCxw MUP counts that are associated withthe wild type probe label from the wild type target,

divided by the background signal observed from themutant target in the same channel. AUG channelsignal to noise is the LCxw AUG counts that areassociated with the mutant probe label from themutant target, divided by the background signal inthat channel observed from the wild type target. Thisratio is an effective representation of the power ofthe assay to discriminate single base changes in agiven sequence. As shown in Table 1, MUP signal tonoise is 17 for homozygous sample and 11 forheterozygous samples. AUG signal to noise is 21 forhomozygous sample and 13 for a heterozygous sam-ple. The homozygous AUG signal to noise is ob-tained by dividing the AUG reading from a homozy-gous variant sample by the background AnoiseB inthe same AUG channel observed with a homozygouswild type sample. The heterozygous AUG signal tonoise is obtained by dividing the AUG reading fromthe heterozygous sample by the signal obtained inthe same channel using a homozygous wild typesample. The MUP signal to noise is calculated usingthe same principle.

For all of the 108 samples screened, the MUPŽbackground was at 54"13 average"standard de-

.viation . The MUP heterozygous signal was at 499"78 and homozygous signal at 773"98. The AUGbackground was at 48"4. The AUG heterozygoussignal at 566"87 and the homozygous at 1008"121. Both the MUP and AUG channels, representingthe heterozygous and the homozygous populations,are well separated from the background with morethan 3s distance. The heterozygous and the ho-mozygous populations are not as well separated inboth MUP and AUG channels. However, there is no

Table 1LCxw readings and signal-to-noise ratio

AUG reading MUP reading

Homozygous wild type 49.7 873.6Heterozygous 629.7 552.3Homozygous variant 1063.6 51.4Homozygous signal to noise 21 17Heterozygous signal to noise 13 11

One hundred nanograms of three known genotype samples wereapplied to LCxw assay, following the procedure in Section 2. TheLCxw raw counts for the three samples are shown along with thesignal to noise.

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–2422

ambiguity in assigning heterozygous or homozygousgenotypes because the former would have both AUGand MUP signals well above the background, whereasthe latter would have only one positive signal.

Theoretically, the homozygous signal might beexpected to be twice as high as that of heterozygotesin a quantitative sense. However, due to the nature ofthe PCR process, especially after 45 cycles of am-plification, the relative quantity of amplified prod-ucts may not reflect the relative quantity of initialtarget. Nonetheless, significantly lower counts havebeen observed for the heterozygous samples, whichare close to half of the counts for the homozygoussamples.

3.4. PositiÕe and negatiÕe controls

A wild type and a variant B2AR codon 16 regionsequence were inserted into the pUC18 vector togenerate a wild type and a variant plasmid control.Equal amounts of the wild type and the variantplasmid were then mixed to generate a heterozygouscontrol. A shorter and a longer insert of each of thewild type and the variant sequence were made usingtwo sets of primers. One set of primers is the sameas the one used in the LCxw assay that generates anamplicon of 106 bp long. Another set of primers wasalso used to generate a 368-bp insert. Both the longand the short inserts span the region of the interestedmutation. Table 2 shows the LCxw counts of plas-mid clones generated using both inserts which workequally well. It was determined after titration that 5pg of each of the wild type and the variant controlplasmid will be used in the assay. The heterozygous

Table 2LCxw readings of four control plasmids

w wClone ID Insert size and type LCx AUG LCx MUPreading reading

Clone 26 106 bp, wild type 52.3 808.4Clone 46 386 bp, wild type 49.6 806.9Clone 7 106 bp, variant 1195.9 45.8Clone 23 386 bp, variant 1170.5 43.2

Five picograms of four different plasmid clones were analyzed onLCxw following the assay procedure in Section 2. All providedequivalent LCxw readings and all are suitable for the purpose ofassay positive controls.

ŽFig. 2. Various amounts of three known genotype samples one.wild type, one heterzygote, and one variant were applied to the

Ž . Ž .assay. The AUG signal is shown in a and the MUP signal in b .

control will contain 2.5 pg each of the wild type andthe variant plasmid.

3.5. Target titration

The amount of DNA isolated from differentaliquots of blood from the same donor vary, depend-ing on the DNA purification procedure. Furthermore,the amount of DNA isolated from the same volumeof blood from different donors will also vary depend-ing on many factors, including white blood cellcounts. In order to access the variation in LCxw

counts due to target amount variation, a target titra-tion experiment was carried out. As illustrated inFig. 2a and b, B2AR LCxw genotyping assay isinsensitive to the amount of DNA within a largetarget range. Therefore, we do not expect to see alarge variation in LCxw counts from donor to donorand different aliquots of blood from the same donor.

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–24 23

Our data, collected from 108 samples, is evidence ofthis target amount insensitivity.

3.6. Reproducibility

The assay reproducibility was tested in six runson 6 different days. Factors that may introduce vari-ability are different instruments including thermalcycler and LCxw analyzer. Therefore, different cy-clers and LCxw analyzers were used during differentruns. Fig. 3a and b demonstrates that the assay hasexcellent reproducibility. The percentage of CV iswithin 10%.

3.7. Assay specificity

A total of 108 individual samples were analyzedon LCxw and a genotype assigned at base position1633 unequivocally. All 108 samples were geno-

Fig. 3. One hundred nanograms of three known genotype sampleswere analyzed on 6 different days using three different cyclers and

w Ž .two different LCx analyzers. The AUG signal is shown in aŽ .and the MUP signal is shown in b .

Fig. 4. Thirty DNA samples were analyzed on LCxw , 10 of whichwere purified using Qiagen Mini whole blood kit and 20 ml of theeluent per sample was used in the assay. Twenty samples werepurified using Qiagen Maxi whole blood kit and archived atknown concentrations. In this case, 100 ng of each sample wasapplied to the assay. Samples 1–12 are homozygous variants;samples 13–26 are heterozygous; samples 27–30 are homozygouswild type.

typed by at least another method such as ARMS,RFLP and sequencing. LCxw genotyping results andresults from other methods are in complete concor-

Ž .dance. Out of 108 samples analyzed, 14 13% wereŽ . Ž .found to be of the wild type A , 34 31% wereŽ . Ž .found to be of the variant type G and 60 56%

Ž .were found to be heterozygotes AG .Fig. 4 illustrates the raw data from LCxw instru-

ment for only 30 of these 108 samples due to spacelimitation.

4. Conclusion

As an increasing number of SNPs are related tovarious disease states, including potential utilities forprognosis and response to therapy, clinical verifica-tion of such correlation will require highly accurategenotyping methods on significantly large numbersof patient samples. Traditional genotyping methodsare time consuming, tedious, as well as technicallychallenging. It is anticipated that automated or semi-automated genotyping methods will be in high de-mand. The basic requirements for these tests willinclude minimal hands-on time, high specificity, highthroughput and cost effectiveness. The B2AR LCxw

genotyping assay described in this paper is very easyto perform and permits SNP genotyping in an auto-mated format. Aside from the DNA extraction time,

( )H. Yu et al.rClinica Chimica Acta 308 2001 17–2424

the setup requires only two pipetting steps and takesless than 30 min for genotyping 48 samples. Ampli-fication and detection of the 48 samples takes about5.5 h of hands-off time and affords excellent speci-ficity. All 108 samples we tested were assignedcorrect genotype without any ambiguity in one run.We believe that this assay will facilitate clinicalresearch in the field of asthma treatment response tob-adrenergic agonists.

Acknowledgements

The authors at Abbott Laboratories would like tothank David Katz, Maria Davey and Steve ConradŽ .Pharmaceutical Product Division, Abbott Labs forproviding part of the samples used in the study andfor helpful discussions with them during the devel-opment of the assay.

References

w x1 Johnson M. The beta-adrenoceptor. Am J Respir Crit CareMed 1998;158:S146–53.

w x2 Drazen JM, Israel E, Boushey HA, Chinchilli VM, Fahy JV,Fish HE, Lazarus SC, Lemanske RF, Martin RJ, Peters SP,Sorkness C, Szefler SJ. Comparison of regularly scheduledwith as needed use of albuterol in mild asthma. AsthmaClinical Research Network. N Engl J Med 1996;335:841–7.

w x3 Drazen JM, Weiss ST, Cooper D. Beta 2 adrenoceptor poly-morphisms. Thorax 1996;51:1168.

w x4 Scott MG, Swan C, Wheatley AP, Hall IP. Identificationof novel polymorphisms within the promoter region of thehuman beta2 adrenergic receptor gene. Br J Pharmacol 1999;126:841–4.

w x5 McGraw DW, Forbes SL, Kramer LA, Liggett SB. Polymor-phisms of the 5X leader cistron of the human b2 adrenergicreceptor regulate receptor expression. J Clin Invest 1998;102:1927–32.

w x6 Martinez FD, Graves PE, Baldini M, Solomon S, Erickson R.Association between genetic polymorphisms of the b2-adren-oceptor and response to albuterol in children with and withouta history of wheezing. J Clin Invest 1997;100:3184–8.

w x7 Liggett SB. Polymorphisms of the b2-adrenergic receptor andasthma. Am J Respir Crit Care Med 1997;156:S156–62.

w x8 Cooper DM, Silverman EK, Speizer FE, Liggett SB, Troisi R,Grobholz JK, Schneyder M, Drazen JM. Population geneticsof polymorphisms of the b -adrenergic receptor in asthma.2

Am J Respir Crit Care Med 1997;155:A256.w x9 Yandava CN, Zappulla DC, Korf BR, Neufeld EJ. ARMS test

for diagnosis of factor V Leiden mutation, a common causeof inherited thrombotic tendency. J Clin Lab Anal 1996;10:414–7.