Variants in the a2A AR Adrenergic Receptor Genein Psychiatric Patients

Jinong Feng,1 Janet L. Sobell,2 Leonard L. Heston,3 David Goldman,4 Edwin Cook Jr.,5Henry R. Kranzler,6 Joel Gelernter,7 and Steve S. Sommer1*

1Department of Molecular Genetics, City of Hope National Medical Center, Duarte, California2Division of Molecular Medicine, City of Hope National Medical Center, Duarte, California3Department of Psychiatry, University of Washington, Seattle, Washington4Laboratory of Neurogenetics, National Institute on Alcohol Abuse and Alcoholism, Bethesda, Maryland5Department of Child Psychiatry, University of Chicago, Chicago, Illinois6Department of Psychiatry, University of Connecticut Health Center, Farmington, Connecticut7Department of Psychiatry, Yale University, New Haven, Connecticut

In various studies of psychiatric patients,alterations in adrenergic receptor (AR) ex-pression or function have been suggested.Herein, the a2A AR gene was screened in 206patients with schizophrenia, attention defi-cit hyperactivity disorder (ADHD), autism,alcohol dependence, or cocaine depen-dence. The entire coding region was exam-ined for single base pair changes, using re-striction endonuclease fingerprinting(REF), a screening method that can detectvirtually 100% of mutations in 2-kb DNA seg-ments. In the approximately 600 kb ofscreened sequence, six novel nucleotidechanges were identified. The changes re-sulted in four missense changes (A25G,N251K, R368L, and K370N), and a sequencein the 3* untranslated region. In addition, asilent change (G363G) was found at high fre-quency in Asians and Native Americans. Ofthe four missense changes, two found in pa-tients with alcohol/drug dependence occurin highly conserved amino acids, suggestingthat these are of likely functional signifi-cance. As the a2A ARs are widely distributedboth pre- and postsynaptically, and as manypharmacological agents with multiple ef-fects target these receptors, the novel mis-sense changes described herein may be can-didates for involvement in alcohol/drug de-pendence, in other clinical disorders ortraits, or in differential response to pharma-

cotherapy. Am. J. Med. Genet. (Neuropsy-chiatr. Genet.) 81:405–410, 1998.© 1998 Wiley-Liss, Inc.

KEY WORDS: schizophrenia; attentiondeficit hyperactivity disor-der; autism; alcohol depen-dence; cocaine dependence

INTRODUCTION

VAPSE-based analysis is one type of candidate geneapproach for delineating genes that predispose to com-plex diseases (Weinshenker and Sommer, unpublishedobservations). VAPSE-based analysis is a molecularepidemiological approach in which the regions of likelyfunctional significance in candidate genes are scannedfor variants affecting protein structure or expression(VAPSEs). Promising VAPSEs are analyzed by case-control association studies for the disease(s) of interest.Positive studies may be confirmed by additional tiers ofcase-control studies, by linkage analysis in the familiesin which the VAPSE segregates, and by searching foradditional VAPSEs in the candidate gene that may beassociated with the disease. Often, case-control studiesare negative. To determine if a VAPSE predisposes toother diseases or traits, a genotype-to-phenotypesearch can be undertaken by defining individuals withthe VAPSE in a large population-based sample inwhich detailed medical histories are available. In thehypothesis-generating phase, those diseases or traitsfound at higher frequency among patients with theVAPSE relative to ethnicity, age, and sex-matched con-trols generate the hypotheses. Subsequent tiers ofcase-control studies serve to test these hypotheses.

The adrenergic receptors (AR), members of the su-perfamily of G protein-coupled receptors, mediate thediverse metabolic and neuroendocrine actions of epi-nephrine and norepinephrine. There are three major

Contract grant sponsor: PHS (Public Health Services); Con-tract grant numbers: MH44276, AA00239, DAAO4060,MH01387, AA11330.

*Correspondence to: Steve S. Sommer, Department of Molecu-lar Genetics, City of Hope National Medical Center, Duarte, CA91010. E-mail: ssommer@smtplink.coh.org

Received 30 September 1997; Revised 16 March 1998

American Journal of Medical Genetics (Neuropsychiatric Genetics) 81:405–410 (1998)

© 1998 Wiley-Liss, Inc.

types of AR: a1, a2, and b; each of these types has atleast three subtypes. Within subtype, amino-acid iden-tity is about 75% in the transmembrane domains; how-ever, this drops to 30–40% between types [Lefkowitz etal., 1995]. The types also couple to different effectormolecules, e.g., the b-type receptors couple to GS withresulting activation of adenylyl cyclase, while a2-typereceptors can couple to various Gi or Go proteins [re-viewed in Lefkowitz et al., 1995].

The a2 ARs are present at both pre- and postsynapticsites and are dispersed in a wide variety of tissues.Based on quantitative ribonuclease protection assaysof messenger RNA expression in the cardiovascularand peripheral tissues, the overall concentration of a2ARs has been shown to be greatest in the human kid-ney, followed by aorta, spleen, heart, and lung[Berkowitz et al., 1994]. By subtype, the a2A predomi-nates in the pancreas and small intestine, while a2Bpredominates in liver and a2C predominates in heart,lung, aorta, spleen, kidney, and adrenal gland.Berkowitz et al. [1994] also examined two regions ofthe central nervous system (CNS) for receptor sub-types; in the cerebral cortex, the a2C subtype showedthe greatest concentration, followed by the a2A sub-type. In the cerebellum, a2A and a2C showed higherlevels than a2B. In the spinal cord, in situ hybridizationstudies have shown a2A to be highest in the cervicalregion, with the a2B subtype predominating in the tho-racic, lumbar, and sacral regions [Smith et al., 1995].

Pharmacological data reflect the wide distribution ofa2A ARs. Clonidine, an a2A AR agonist with majortherapeutic use as an antihypertensive agent, also hasefficacy in the treatment of a variety of other disorders[reviewed in Hoffman and Lefkowitz, 1995]. Diabeticpatients suffering from diarrhea as a result of auto-nomic neuropathy have been successfully treated withclonidine, presumably through stimulation of a2A ARsin the gastrointestinal tract, leading to increased ab-sorption of sodium chloride and fluid and inhibition ofbicarbonate secretion.

Additionally, clonidine appears to have some efficacyas an adjunct to anesthesia; in some patients, it ap-pears to decrease the anesthetic dose requirement andto increase hemodynamic stability [Maze and Tran-quilli, 1991]. Another a2 agonist, dexmedetomidine,has been demonstrated to be efficacious in preoperativesedation, anxiolysis, drying of secretions, and analge-sia [Lefkowitz et al., 1995]. Apraclonidine, a relativelyselective a2A AR agonist, has been clinically useful as alocal agent to reduce intraocular pressure.

As reviewed by Lefkowitz [1995], other pharmaco-logical agonists related to clonidine include guanfacine,a more select a2A AR agonist than clonidine, and gua-nabenz. These agents have efficacy as antihyperten-sives and, like clonidine, appear to activate brain-stemreceptors with a resultant suppression of sympatheticnervous system activity. Methyldopa is also thought toreduce blood pressure in a manner similar to that ofclonidine.

Antagonists of a2A ARs, such as yohimbine, in-creased sympathetic output and potentiate the releaseof norepinephrine from nerve endings, leading to acti-vation of a1 and b1A ARs in the heart and peripheral

vasculature, with a consequent rise in blood pressure.Another antagonist, phenoxybenzamine, is efficaciousin the treatment of pheochromocytomas, tumors of theadrenal medulla, and sympathetic neurons that se-crete large quantities of catecholamines and cause hy-pertension. Phentolamine, a competitive a adrenergicantagonist, has similar affinities for both a1 and a2AARs. Clinically, phentolamine and the related com-pound tolazoline have been useful for short-term con-trol of hypertension.

Involvement of the adrenergic systems also has beensuggested in several psychiatric disorders. Use of thea2 agonist clonidine has been found to be efficacious insubgroups of children with attention-deficit hyperac-tivity disorder (ADHD). For those children with higharousal states, hyperactivity, impulsivity, defiance, ir-ritability, lability, and explosivity, clonidine appears tomodulate mood and activity level as well as to improvecooperation and raise frustration thresholds [Dulcan etal., 1995]. In autism, studies suggest that clonidinemay be efficacious in reducing impulsivity, self-stimulation, and hyperarousal behaviors [Freimer etal., 1996; Fankhauser et al., 1992].

Clonidine has also shown efficacy in the treatment ofwithdrawal from narcotics, alcohol, and nicotine[Glassman et al., 1984; Manhem et al., 1985; Charneyet al., 1981]. The elevated sympathetic nervous activityand the substance craving associated with withdrawalappear to be ameliorated with treatment [Lefkowitz etal., 1995]. Noradrenergic as well as dopaminergic andopioid systems may be involved in the rewarding ef-fects of ethanol [Nevo and Hamon, 1995], and centralnoradrenergic activity may be involved in craving[Liljeberg et al., 1987]. Reduced levels of noradrenalinehave been reported in alcoholics, and some studieshave reported the loss of central noradrenergic neuronsin the locus coeruleus of alcoholics [Baker et al., 1994].Recent studies also have demonstrated that ethanolincreases a2B and a2C adrenergic receptor mRNA ex-pression in cell culture [Hu et al., 1993].

An initial double-blind, placebo-controlled trial com-bining clonidine with haloperidol demonstrated effi-cacy in improving thought disorder ratings in 12schizophrenic patients who were medication-free for2–4 weeks before treatment [Maas et al., 1995]. In an-other trial with 17 treatment-resistant patients, theaddition of an a2A AR antagonist (idazoxan) augmentedresponse to typical neuroleptics under double-blind,placebo-controlled conditions [Litman et al., 1996].Clozapine, an atypical neuroleptic with efficacy in asubset of treatment-resistant schizophrenics, has anti-alpha-adrenergic, antihistaminic, and anticholinergicproperties, in addition to its potent binding affinity forserotonergic receptors [Meltzer, 1994].

In studies of depression and suicide, depressed pa-tients appear to have increased a2 receptor bindingsites in platelets, and increased a2 receptor densitieshave been noted in the brains of suicide victims[Nathan et al., 1995].

These pharmacological and other data suggest thatalterations in alpha-adrenergic receptor expression orfunction may contribute to certain psychiatric pheno-types. ARs may be involved in the pathogenesis of psy-

406 Feng et al.

chiatric disorders through missense changes that re-sult in protein structures which eliminate function orspecifically alter ligand binding and signal transduc-tion characteristics. To begin to test this hypothesis, wescreened the a2A adrenergic receptor gene in patientswith schizophrenia, ADHD, autism, alcohol depen-dence, or cocaine dependence (N total, 206). With re-striction endonuclease fingerprinting (REF), a screen-ing method that can detect virtually 100% of mutationsin DNA segments as large as 2 kb [Liu and Sommer,1995], the entire coding region of the gene was exam-ined in one lane of a gel.

In this report, we confirm the utility of REF as arapid, highly sensitive method by scanning 600 kb ofgenomic sequence from 93 patients with schizophreniaand pilot samples of patients with four other psychiat-ric disorders. Four VAPSEs were found, including twoat highly conserved amino acids in patients with sub-stance abuse. These VAPSEs may be tested for asso-ciation with a variety of different disorders.

MATERIALS AND METHODSPatient Samples

All schizophrenic patients met criteria for the dis-ease as defined by the Diagnostic and StatisticalManual, Third Edition, Revised (DSM-III-R), as de-scribed previously [Sobell et al., 1993]. The majority ofpatients were ascertained through state mental insti-tutions in Minnesota, Washington, and Oregon. Alco-holic patients of Finnish ethnicity were ascertainedthrough collaborative efforts involving the National In-stitute on Alcohol Abuse and Alcoholism (NIAAA) andthe University of Helsinki, Finland (by D.G.). South-western Native American patients with alcohol depen-dence were ascertained through the NIAAA (by D.G.).Two additional alcohol-dependent patients, one Cauca-sian and one African-American, were ascertainedthrough the West Haven Veteran’s AdministrationMedical Center (by J.G.). ADHD patients were ascer-tained at the University of Chicago using the followingcriteria: 1) a child or adolescent with a DSM-III-R di-agnosis of ADHD made in a consensus diagnostic con-ference in which a child psychologist, child psychia-trist, and a developmental pediatrician presented find-ings from each of their evaluations; and 2) consent toparticipate by both parents and child. Patients withautistic disorder also were ascertained at the Univer-sity of Chicago if the following criteria were met: 1)child or adolescent with a mental age of at least 18months and intelligence quotient greater than 35; 2)DMS-IV diagnosis of autistic disorder by a child psy-

chologist and child psychiatrist; 3) diagnosis of autisticdisorder by Autism Diagnostic Interview-Revised [Lordet al., 1994]; 4) absence of specific identified etiologybased on history and physical examination; and 5) con-sent to participate from both parents and child. Pa-tients with a history of DSM-III-R cocaine dependencewere ascertained through the University of Connecti-cut Health Center (by H.R.K.) and the West HavenVeteran’s Administration Medical Center (by J.G.), asdescribed previously [Freimer et al., 1996]. Selected de-mographic characteristics of all patients are displayedin Table I.

Laboratory Methods for REF

DNA amplification. DNA was extracted as previ-ously described [Gustafson et al., 1987]. An initial PCRamplification was performed in a total volume of 25 mlwith 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5 mMMgCl2, 200 mM of each deoxyribonucleoside triphos-phate, 0.1 mM of primers a2A AR (Hs)-I1 (−172)-24D:GGCTCCTGCCCCTCCCTATGTGA; a2A AR (Hs)-I2(1472)-25U: CAGACGAGGAAACGCAGAGCAGGCA),1 U of AmpliTaq (Perkin-Elmer, Norwalk, CT), and ap-proximately 2 ng of genomic DNA (initial denaturationat 95°C for 1 min followed by 15 cycles of denaturationat 95°C for 20 sec, annealing at 65°C for 1 min, andelongation at 72°C for 3 min with the Perkin-ElmerGeneAmp PCR System 9600). The informative primernames and numbering system for the gene were previ-ously described [Fraser et al., 1989; Stoflet et al., 1988].A second nested round of PCR was performed usinginternal primers a2A AR (Hs)-I1 (−125)-47D: CGGAG-GAAGACGAGGACCCACGG + tail; a2AAR (Hs)-I1(1413)-46U: CCTCGATGCCCCCCGCTGCCTG + tail),using 1 ml of the first-round product. An oligonucleotide‘‘tail’’ containing two recognition sequences for restric-tion endonucleases was added to each of the nestingprimers. The additional restriction sites prevented theamplified and digested products from having uniformfragment sizes at the ends. The tail sequences, with therecognition sequences underlined, were:

Downstream: GCCTGGTCGGCCGGGATCGAACGCC

EagI AlwI

Upstream: TTACCAGCTGGCAATCGGCCGTAA

AluI EagI

The primers were added to a volume of 75 ml with thesame reagents as above. PCR cycles were conducted atthe same temperatures and times for a total of 30cycles and a final elongation for 10 min at 72°C. The

TABLE I. Distribution of Patients by Disease and Race/Ethnicity

Disease Total Caucasian AsianAfrican-

American FinnishSouthwest

Indian Hispanic

Schizophrenia 93 74 1 18Alcohol dependence 44 1 1 14 28ADHD 19 17 1 1Autism 15 14 1Cocaine dependence 35 20 13 2Total 206 126 2 33 14 28 3

Adrenergic Receptors, Psychiatric Illness 407

nested amplification products (1.64 kb) were purifiedusing Microcon-100 columns (Amicon, Beverly, MA)and diluted to 20 ng/ml.

Restriction endonuclease digestions. The restric-tion enzyme digestions were performed as describedpreviously [Liu and Sommer, 1995], with the exceptionthat both DNA and enzyme amounts were halved. Inaddition, 0.5 U of calf intestinal alkaline phosphatase(CIAP) (Life Technologies, Gaithersburg, MD) wereadded to the reaction to remove the phosphate group atthe 58 end of the digested DNA fragments. Selection ofthe enzyme battery was aided by REF Select Software(Sommer et al., unpublished observations) softwareavailable on request. Fifty nanograms of purified DNAwere digested in separate tubes by five groups of endo-nucleases (AluI/BglII; EagI; AlwI/SapI; NlaIII/XmnI;and SfaNI/StyI). All enzymes and buffers were purchasedfrom New England Biolabs (Beverly, MA). Digestionreactions were incubated at 37°C overnight, followedby enzyme inactivation by heating at 80°C for 30 min.

End-labeling and electrophoresis. The digestionreaction products were combined. Four nanograms of di-gested DNA products were 58 end-labeled with [g-33P]ATP (Amersham, Arlington Heights, IL) and 1 U of T4polynucleotide kinase (Promega, Madison, WI) in abuffer of 50 mM Tris-HCl (pH 7.4), 10 mM MgCl2, and5 mM DTT in a total reaction volume of 2 ml. Incuba-tion was at 37°C for 30 min. One microliter of the end-labeled product was electrophoresed on a 0.5 × MDE™gel with 9% urea (1.5 M), using the Model SE 1500Poker Face™ sequencing apparatus (Hoefer Scientific,San Francisco, CA) with a 50 mM Tris-borate (pH 8.3)buffer at 15 W constant power. Autoradiographs of therestriction endonuclease fingerprints were interpretedas previously described [Liu and Sommer, 1995].

Sequencing. Samples with variant banding pat-terns were analyzed by cycle-sequencing. After 1 mindenaturation at 95°C, 30 cycles were performed withdenaturation at 95°C for 20 sec, annealing at 60°C for30 sec, and elongation (after a ramping time of 1 min)at 72°C for 1 min.

RESULTS

Of the 206 patients screened (∼600 kb of total se-quence), several variant electrophoretic fingerprintswere found. By direct sequencing, six changes were

identified, including four missense mutations (A25G ina Finnish alcoholic patient; N251K in an African-American schizophrenic patient; R368L in an African-American with cocaine dependence; and K370N in aFinnish alcoholic patient), one silent change (G363G),and a polymorphism in the 38 untranslated region(Table II).

The silent change at residue 363, resulting from aC → A transversion, was found in different ra-cial/ethnic backgrounds, with allele frequencies being afunction of population of origin rather than disease di-agnosis (Table III). Among Caucasian patients of pri-marily Western European descent, the frequency of theminor allele was 1.2%, whereas the minor allele waspresent in 41.1% of Southwest Native American alco-holic patients and 28% of Asians. There was significantheterogeneity between the different populations stud-ied (x2

5df 4 110; P < 10−5).The polymorphism in the 38 untranslated region

(C1371G) was found only in one African-American pa-tient with a diagnosis of cocaine dependence (patientidentification code: Coc078).

DISCUSSION

In the present analysis, six sequence changes in thea2A adrenergic receptor were identified in patientswith a variety of psychiatric disorders. These sequencealterations included four missense changes, two ofwhich occurred in highly conserved amino acids(R368L and K370N) (Table II). The R368L change cor-responds to one of the sequencing discrepancies in thetwo Genbank reports (Genbank accession numbersM23533 and M18415).

The R368 and K370 residues are conserved in allmammalian a2 genes, and also in a goldfish a2 gene(Table II; Fig. 1). Expression studies can demonstratethe deleterious nature of these missense changes, butthese studies are laborious and a negative result isdifficult to interpret, since some functionally signifi-cant changes may not be detected. However, it appearslikely that these changes are of physiological signifi-cance, because the likelihood of a missense change al-tering the function of the allele may be estimated by itslevel of evolutionary conservation [Bottema et al.,1991]. A multitude of data suggests that the level ofconservation is a good indicator of this likelihood; for

TABLE II. Sequence Variants Identified in the AR a2A Gene

Nucleotidea Structural change Amino-acid conservationb Disease and patient Race/ethnicity

C73G A25G Mammalian ARa2A Alcohol dependence, FS1195 FinnishC752G N251K Mammalian ARa2A Schizophrenia, S589 African-AmericanG1102T R368L Mammalian ARa2A,B,C Cocaine dependence, Coc032 African-American

Goldfish ARa2c

G1109C K370N Mammalian ARa2A,B,C Alcohol dependence, FT2870 FinnishGoldfish ARa2

No structural change

C1091A G363G Silent polymorphism See Table III See Table IIIC1371G 38 UTR — Cocaine dependence, Coc078 African-American

aNumbering according to Fraser et al. [1989].bConservation based on alignments performed using pileup software from the Genetics Computing Group, Inc. (gap weight, 12; gap length weight, 4).cGoldfish a2 subtype has highest similarity to the a2C subtype based on our alignments, but is not an a2C subtype conclusively.

408 Feng et al.

example, analyses of the relationship between mis-sense mutations causing hemophilia B and the extentto which residues in factor IX are conserved duringevolution have indicated that almost all missense mu-tations that occur at residues conserved in the factor IXgene family (more than 2 billion years of evolutionarydivergence) are functionally deleterious [Bottema etal., 1991]. This same relationship has been found forother genes in which many mutations have been de-scribed (e.g. p53, phenylalanine hydroxylase, and thecystic fibrosis transmembrane conductance regulator).

The data from the factor IX gene also suggest that asignificant proportion of residues conserved in multiplemammalian species are not essential. As a result oftheir comparison of missense mutations causing hemo-philia B with evolutionary conservation of residuesfrom amino-acid sequence alignments of factor IX andrelated coagulation proteases, Bottema et al. [1991]proposed that most residues in factor IX were of twoextreme types. At least 25%, and possibly up to 40%,are ‘‘critical’’ residues in which any missense changesdisrupt function. Most of the remaining residues func-tion as ‘‘spacer’’ residues in which virtually all possiblemissense changes are tolerated, but the main chain is

necessary to keep the critical residues in proper regis-ter. This model, termed the critical/spacer model of pro-tein structure, implies that important vs. inconsequen-tial amino acid substitutions in a human protein can bepredicted by analysis of appropriately diverged se-quences (for the factor IX and related genes, this rep-resents 2.5+ billion years). Four levels of evolutionaryconservation for the factor IX and related genes weredefined: generic, partially generic, specific, and non-conserved residues. ‘‘Specific’’ residues are defined asthose conserved in the available factor IX sequencesbut different in the three members of the gene family(factor VII, factor X, and protein C); ‘‘partially-generic’’residues are defined as those conserved in the factor IXgene and in one or two other members of the gene fam-ily; and ‘‘generic’’ residues are defined as those con-served in all members of the gene family. Relative toresidues that are not conserved in mammalian factorIX sequences, deleterious missense mutations are 5-fold more likely to occur in factor IX-specific residues,13-fold more likely to occur in partially-generic resi-dues, and 33-fold more likely to occur in generic resi-dues.

Given the hypothesis that the missense changes inhighly conserved amino acids are likely to be of func-tional significance, case-control analyses with ethni-cally similar controls, genotype relative-risk studieswith parental controls [Schaid and Sommer, 1993], andfamily studies may be conducted to determine if theR368L and K370N substitutions are associated withsubstance use disorders.

If no association is found between these variants andsubstance abuse, additional studies may be conductedto determine if the aberrant alleles are associated withother disorders or clinical traits plausibly related toadrenergic dysfunction. The adrenergic receptors havediverse effects on cardiovascular physiology, smoothmuscle contraction, neurotransmitter release, andmodulation of sympathetic outflow from the centralnervous system. In cardiac function, bA ARs and a1AARs are largely involved in myocardial iontrophy,while the a2A AR appears to be involved in the func-tioning of coronary arteries and prejunctional cardiacnerves [Berkowitz et al., 1994]. However, recent stud-ies have also demonstrated the presence of low levels ofa2A ARs in the myocardium. In other peripheral tis-sues, a2A ARs are involved in functions of the liver(glycogenolysis and gluconeogenesis), pancreas (de-creased secretion of insulin from islet B cells), spleen(contraction of capsule), and kidney (role in water re-absorption at proximal and distal tubules), and inplatelet aggregation [Lefkowitz et al., 1995; Berkowitzet al., 1994]. Additionally, the a2A ARs modulate spinalpain pathways by mediating sedative and hypnotic ef-fects of anesthetic agents [Maze and Tranquilli, 1991].In the brain, a2A ARs are involved in modulation ofblood pressure, control of affect, sedation, and nocicep-tion [Berkowitz et al., 1994].

To determine if the R368L or K370N amino-acid sub-stitutions are associated with other disorders or clini-cal traits, a large number (perhaps thousands) of DNAsamples may be screened to identify a subset of indi-viduals with the variant alleles. Association and/or

Fig. 1. Amino-acid alignment of a2-adrenergic receptors. Alignmentswere conducted using the ‘‘pileup’’ software from the Genetics ComputerGroup, Inc. Gap weight was set at 12 and gap length penalty was set at 4.Goldfish showed highest similarity to the a2C subtype, but could not beconclusively designated as such. Both the R368L and the K370N missensechange occurred in highly conserved amino acids.

TABLE III. Distribution of the Silent Change at G363 byRace/Ethnicity

Race/ethnicity Frequency Percentage

Caucasian (Western European) 3/256 1.2Finnish 0/28 0.0African-American 4/68 5.9Native American (Southwestern) 23/56 41.1Asiana 9/32 28.1Hispanic 0/6 0.0Total 39/446 8.7

aIncludes 14 Asian individuals without psychiatric illness.

Adrenergic Receptors, Psychiatric Illness 409

family studies comparing the medical histories of indi-viduals with R368L or K370N and those homozygousfor the wild-type allele may be conducted in the hope ofelucidating clinical phenotypes [Sobell et al., 1996].

ACKNOWLEDGMENTS

Sally Satel, M.D., helped with the clinical assess-ment of patients with substance dependence.

REFERENCESBaker KG, Halliday GM, Harper CG (1994): Effect of chronic alcohol con-

sumption on the human locus coeruleus. Alcohol Clin Exp Res 18:1491–1496.

Berkowitz DE, Price TE, Bello EA, Page SO, Schwinn DA (1994): Local-ization of messenger RNA for three distinct a2-adrenergic receptorsubtypes in human tissues. Anesthesiology 81:1235–1244.

Bottema CDK, Ketterling RP, Ii S, Yoon H, Phillips JA III, Sommer SS(1991): Missense mutations and evolutionary conservation of aminoacids: Evidence that many of the amino acids in factor IX function as‘‘spacer’’ elements. Am J Hum Genet 49:820–838.

Charney DS, Sternberg DE, Kleber HD, Heninger GR, Redmond DE Jr(1981): The clinical use of clonidine in abrupt withdrawal from metha-done. Effects on blood pressure and specific signs and symptoms. ArchGen Psychiatry 38:1273–1277.

Dulcan MK, Bregman JD, Weller EB, Weller RA (1995): ‘‘Treatment ofChildhood and Adolescent Disorders.’’ Washington, DC: American Psy-chiatric Press, 673.

Fankhauser MP, Karumanchi VC, German ML, Yates A, Karumanchi SD(1992): A double-blind, placebo-controlled study of the efficacy of trans-dermal clonidine in autism. J Clin Psychiatry 53:77–82.

Fraser CM, Arakawa S, McCombie WR, Venter JC (1989): Cloning, se-quence analysis, and permanent expression of a human alpha 2-adrenergic receptor in Chinese hamster ovary cells. Evidence for inde-pendent pathways of receptor coupling to adenylate cyclase attenua-tion and activation. J Biol Chem 264:11754–11761.

Freimer M, Kranzler HR, Satel S, Lacobelle J, Skipsey K, Gelernter J(1996): No association between D3 dopamine receptor (DRD3) allelesand cocaine dependence. Addiction Biol 1:281–287.

Glassman AH, Jackson WK, Walsh BT, Roose SP, Rosengeld B (1984):Cigarette craving, smoking withdrawal, and clonidine. Science 226:864–866.

Gustafson S, Proper JA, Bowie EJW, Sommer SS (1987): Parameters af-fecting the yield of DNA from human blood. Anal Biochem 165:294–299.

Hu G, Querimit LA, Downing LA, Charness ME (1993): Ethanol differen-tially increases alpha 2-adrenergic and muscarinic acetylcholine recep-tor gene expression in NG108-15 cells. J Biol Chem 268:23441–23447.

Lefkowitz RJ, Hoffman B, Taylor P (1995); ‘‘The Pharmacological Basis ofTherapeutics,’’ 9th ed. New York: McGraw-Hill, 128.

Liljeberg P, Mossberg D, Borg S (1987): Clinical conditions and centralnoradrenergic activity in long term abstinent alcoholic patients. Alco-hol Alcohol 1:615–618.

Litman RE, Su TP, Potter WZ, Hong WW, Pickar D (1996): Idazoxan andresponse to typical neuroleptics in treatment resistant schizophrenia.Comparison with the atypical neuroleptic, clozapine. Br J Psychiatry168:571–579.

Liu Q, Sommer SS (1995): Restriction endonuclease fingerprinting (REF):A sensitive method for screening mutations in long, contiguous seg-ments of DNA. BioTechniques 18:470–477.

Lord C, Rutter M, LeCouteur A (1994): Autism Diagnostic Interview-Revised: A revised version of a diagnostic interview for caregivers ofindividuals with possible pervasive developmental disorders. J AutismDev Disord 24:659–685.

Maas JW, Miller AL, Tekell JI, Fimderburg L, Silva JA, True J, Velligan D,Berman N, Bowden CL (1995): Clonidine plus haloperidol in the treat-ment of schizophrenia/psychosis. J Clin Psychopharmacol 15:361–364.

Manhem P, Nilsson LH, Moberg AL (1985): Alcohol withdrawal: Effects ofclonidine treatment on sympathetic activity, the renin-aldosterone sys-tem, and clinical symptoms. Alcohol Clin Exp Res 9:238–243.

Maze M, Tranquilli W (1991): Alpha-2 adrenoceptor agonists: Defining therole in clinical anesthesia. Anesthesiology 74:581–605.

Meltzer HY (1994): An overview of the mechanism of action of clozapine. JClin Psychiatry 55:47–52.

Nathan KI, Musselman DL, Schatzberg AF, Nemeroff CB (1995): ‘‘Text-book of Psychopharmacology.’’ Washington, DC: American PsychiatricPress, 452.

Nevo I, Hamon M (1995): Neurotransmitter and neuromodulatory mecha-nisms involved in alcohol abuse and alcoholism. Neurochem Int 26:305–336.

Schaid DJ, Sommer SS (1993): Genotype relative risks: Methods for designand analysis of candidate-gene association studies. Am J Hum Genet53:1114–1126.

Smith MS, Schambra UB, Wilson KH, Page SO, Hulette C, Light AR,Schwinn DA (1995): a2-adrenergic receptors in human spinal cord:Specific localized expression of mRNA encoding a2-adrenergic receptorsubtypes at four distinct levels. Brain Res Mol Brain Res 34:109–117.

Sobell J, Heston L, Sommer S (1993): Novel association approach for de-termining the genetic predisposition to schizophrenia: Case-control re-source and testing of a candidate gene. Am J Med Genet 48:28–35.

Sobell JL, Sigurdson DC, Heston LL, Byerley WF, Sommer SS (1996):Genotype-to-phenotype analysis: Search for clinical characteristics of amissense change in the GABAA-b1 receptor gene. Am J Med Genet67:81–84.

Stoflet ES, Koeberl DD, Sarkar G, Sommer SS (1988): Genomic amplifica-tion with transcript sequencing. Science 239:491–494.

410 Feng et al.