Brief Research Communication

Possible Association of a Silent Polymorphism inthe Neuronal Nicotinic Acetylcholine ReceptorSubunit a4 With Common IdiopathicGeneralized Epilepsies

Ortrud Steinlein,1* Thomas Sander,2 Jens Stoodt,1 Rebekka Kretz,2 Dieter Janz,2 andPeter Propping1

1Institute of Human Genetics, University of Bonn, Bonn, Germany2Epilepsy Genetic Group, Department of Neurology, Humboldt University of Berlin, Berlin, Germany

The a4 subunit gene of the neuronal nico-tinic acetylcholine receptor (CHRNA4) hasrecently been identified as the first gene un-derlying an idiopathic partial epilepsy syn-drome in human, autosomal-dominant noc-turnal frontal lobe epilepsy (ADNFLE).CHRNA4 is located in the candidate regionfor benign familial neonatal convulsionsand low-voltage EEG on chromosome 20q. Inthe present study, we examined the possiblerole of CHRNA4 in common subtypes of id-iopathic generalized epilepsy (IGE), com-prising childhood and juvenile absence epi-lepsy and juvenile myoclonic epilepsy(JME), by systematically screening the cod-ing region of the gene for sequence variants.We present here a population-based associa-tion study testing the hypothesis that vari-ants of the CHRNA4 gene confer genetic sus-ceptibility to common subtypes of IGE. Themissense mutation (Ser248Phe), associatedwith ADNFLE, and four silent polymor-phisms in the CHRNA4 gene were geno-typed in 103 IGE patients and 92 controls bypolymerase chain reaction and subsequentrestriction analysis. Without correction formultiple testing, the frequency of the T-allele of the silent CfoI bp595 polymorphismwas increased in the entire group of IGE pa-tients (f(T) = 0.085) compared to that in thecontrols (f(T) = 0.027). The allelic associationwas not restricted to any subgroup of IGEwith either JME or idiopathic absence epi-

lepsies. This result suggests that variationof the CHRNA4 gene, or so-far-undetectedsequence variants near the CHRNA4 locus,confer susceptibility to the common IGEsyndromes. Am. J. Med. Genet. 74:445–449,1997. © 1997 Wiley-Liss, Inc.

KEY WORDS: neuronal nicotinic acetylcho-line receptor; idiopathic gen-eralized epilepsy; association

Idiopathic generalized epilepsy (IGE) has a cumula-tive incidence of approximately 0.7% in the generalpopulation and accounts for about 40% of all epilepsiesup to age 40 [Beck-Mannagetta et al., 1989]. Regardingthe electroencephalographic and clinical characteris-tics, IGE can be divided into several common and raresubtypes. Childhood absence epilepsy (CAE), juvenileabsence epilepsy (JAE), and juvenile myoclonic epi-lepsy (JME) are the common IGE subtypes, while be-nign familial neonatal convulsions (BFNC) or benignfamilial infantile convulsions (BFIC) account for only aminority of IGE [Commission on Classification andTerminology of the International League Against Epi-lepsy, 1989].

A concordance rate of up to 95% in monozygotic twinspoints to a nearly exclusive genetic etiology of the com-mon IGE syndromes [Gedda and Tatarelli, 1971; Berk-ovic et al., 1990]. The problem of the underlying geneticmodel is still unsolved. Since first-degree relativeshave an empirical risk of about 5% to be affected [Beck-Mannagetta et al., 1989] and second-degree relatives ofabout 1.5%, a monogenic mode of inheritance can beexcluded. The familial clustering of the common IGEsubtypes, such as childhood absence and juvenile ab-sence epilepsy, juvenile myoclonic epilepsy, and epi-lepsy with generalized tonic-clonic seizures on awak-ening, as well as the combination of various IGE sub-types in patients, suggests the involvement of several

Contract grant sponsor: Deutsche Forschungsgemeinschaft;Contract grant numbers: SFB 400/B5, Be 1517/1-3.

*Correspondence to: Dr. Ortrud Steinlein, Institute for HumanGenetics, University of Bonn, Wilhelmstr. 31, 53111 Bonn, Ger-many.

Received 11 July 1996; Revised 7 February 1997

American Journal of Medical Genetics (Neuropsychiatric Genetics) 74:445–449 (1997)

© 1997 Wiley-Liss, Inc.

overlapping trait loci [Berkovic et al., 1987; Janz et al.,1992; Reutens and Berkovic, 1995]. Some of these locimight only unspecifically influence the seizure thresh-old, whereas others might determine the specific sei-zure subtype. From a neurobiological point of view, theIGE types with age-related onset appear to form a bio-logical continuum with age-dependent expression ofparticular seizure types. The seizure phenotypes seemto reflect the developmental stage of brain maturationat that point in time. Progression of brain maturationobviously changes the features of the original seizuretype, which can remit spontaneously or can evolve intoother seizure types with a later age of onset.

Several attempts have been made to map the genesunderlying the different subtypes of common IGE. Twosusceptibility genes (EJM1 and EJM2) have been as-signed to the short arm of chromosome 6 in familiesascertained through JME patients or in families whoexhibit no pyknoleptic absence seizures [Greenberg etal., 1988; Liu et al., 1995, 1996; Serratosa et al., 1996].Subsequent studies both confirmed [Weissbecker et al.,1991; Durner et al., 1991] or rejected [Whitehouse etal., 1993; Delgado-Escueta et al., 1994] the existence ofan IGE locus on chromosome 6. These contradictoryresults may be due to several reasons. In oligogenic orpolygenic traits, the chance of replicating an initiallinkage result depends on several factors, including ge-netic heterogeneity, differences in diagnostic criteria,the recombination frequency between the marker andthe disease locus, the share of contribution of the dis-ease locus to the trait, and the reliability of the initiallinkage observation [Suarez et al., 1994]. Furthermore,families ascertained through CAE or JAE patients maylack a major gene effect on chromosome 6 [Sander etal., 1995]. Other loci for a broad range of common IGEsyndromes on chromosome 8q and 3p have been sug-gested, but not yet confirmed [Zara et al., 1995].

In contrast to the common IGE syndromes, BFNCand BFIC are inherited as autosomal dominant traits.With an estimated penetrance between 70% and 90%,these diseases segregate through subsequent genera-tions. While no locus assignment for BFIC has yet beenconfirmed, two loci have been found for BFNC. Themajority of families show linkage to chromosome20q13.2-q13.3 (locus symbol ‘‘EBN1’’) [Leppert et al.,1989; Malafosse et al., 1992], some other families arelikely to be linked to chromosome 8q (locus symbol‘‘EBN2’’) [Lewis et al., 1993; Steinlein et al., 1995a].Recently, evidence suggestive of linkage to the EBN2region was found for a broad range of common IGEfamilies [Zara et al., 1995], raising the possibility thatat least some IGE syndromes and BFNC are under theinfluence of allelic variants of the same gene.

A variant of the normal human EEG, the low-voltageEEG (LVEEG) [Steinlein et al., 1992], and an idio-pathic partial epilepsy syndrome, autosomal dominantnocturnal frontal lobe epilepsy (ADNFLE) [Phillips etal., 1995], have been mapped to the same region onchromosome 20q as EBN1. The gene for the a4 subunitof the neuronal nicotinic acetylcholine receptor(CHRNA4), localized on chromosome 20q13.2-q13.3[Steinlein et al., 1994], is a reasonable candidate genefor epilepsies linked to chromosome 20q. Recently, a

missense mutation of a highly conserved serine (serineto phenylalanine; aa252, referring to the human a4subunit numbering [Steinlein et al., 1996]); aa248, re-ferring to the a subunit of Torpedo californica [Noda etal., 1982] in the second transmembrane domain ofCHRNA4 was found to be responsible for ADNFLE in alarge Australian pedigree [Steinlein et al., 1995b; Wei-land et al., 1997].

The discovery of CHRNA4 as the first gene respon-sible for a form of idiopathic partial epilepsy and itslocalization in the candidate region for a rare mono-genic epilepsy syndrome (BFNC) opens the possibilitythat this gene is involved in other idiopathic epilepsiesas well. Therefore, the present work was designed toidentify sequence variants in the CHRNA4 gene and totest the possibility of an association between CHRNA4and common IGE syndromes, comprising CAE, JAE,and JME.

The study protocol was approved by the Ethic‘s Com-mittee of the University Hospital Rudolf Virchow atHumboldt University of Berlin. Informed consent wasobtained from all participants. All IGE patients andcontrols were unrelated individuals of German origin.One hundred and three patients with either CAE, JAE,or JME were included in the present study andscreened for sequence variations in the CHRNA4 geneby single strand confirmation analysis (SSCA) andgenotyped for four silent polymorphisms. To increaseclinical homogeneity among the IGE patients, two sub-groups of patients were formed by the presence of ei-ther JME (60 individuals), or idiopathic absence epi-lepsy (CAE/JAE; 48 individuals). Both subgroups dis-play a partial overlap due to syndrome combinations ofJME with either CAE or JAE. The diagnoses were per-formed according to the revised ‘‘Classification of Epi-lepsies and Epileptic Syndromes of the InternationalLeague Against Epilepsy’’ [Commission on Classifica-tion and Terminology of the International LeagueAgainst Epilepsy, 1989]. Only patients were includedwho did not show any neurological or cognitive impair-ment and had no history of CNS trauma or disease. Thecontrols (94 individuals) were randomly chosen from apanel of anonymous blood donors of German descent.

The Ser248Phe mutation in the second transmem-brane domain of the CHRNA4 gene was detected asdescribed previously [Steinlein et al., 1995b]. None ofthe IGE patients and controls carried the Ser248Phemutation. For SSCA fragments covering the entire cod-ing region, including the exon-intron boundaries, of theCHRNA4 gene were amplified from genomic DNA (forprimer sequences, see Table I). Several band shifts in-dicating sequence variations were observed in SSCA.Sequencing of the respective fragments revealed tendifferent nucleotide exchanges (Table II). They wereeither localized in the introns of the CHRNA4 gene orin the third base of a coding triplet. None of themcaused an amino acid exchange. Most of the silent mu-tations were found in the less-conserved 38 part of thegene, which codes for the large intracellular loop. Thesilent mutations of the CHRNA4 gene were regardedas useful polymorphisms for association studies if bothalleles occurred in sufficient frequencies and created orabolished restriction sites, leading to easily detectable

446 Steinlein et al.

changes in the restriction pattern of fragments ampli-fied. Four polymorphisms meeting these criteria,HaeIII bp145−55, CfoI bp594, CfoI bp1545, and StyIbp1674+14, were used for association analysis (Fig. 1).

For each polymorphism, the heterozygotes and ho-mozygotes were counted and compared with the fre-quencies expected under Hardy-Weinberg equilibrium.All genotype frequencies were within Hardy-Weinbergexpectations. No linkage disequilibrium was observedbetween the four polymorphisms. The relative inci-dence of the different alleles and total x2 was esti-mated according to the method of Woolf [1955]. Thefrequencies of each allele were estimated for the entiregroup of IGE patients, as well as for two IGE subgroupscomprising either CAE and JAE or JME.

The allele frequencies in patients and controls areshown in Table III. Without correction for multipletesting, the x2 test showed that the frequency of theT-allele of the bp594 CfoI polymorphism was signifi-cantly higher in the entire group of IGE patients thanin unrelated control subjects (P 4 0.0216). This ismainly due to CAE/JAE patients (P 4 0.0397). The x2for the frequency of the bp594 T-allele of JME patientscompared with the control subjects did not reach sig-

nificance levels (P 4 0.0769). Individuals carrying theCfoI bp594/T-allele have odds ratios of 3.57 (1.31–9.72;whole sample), 2.96 (0.95–9.21; patients with JME)and 3.57 (1.16–11.02; patients with CAE/JAE), respec-tively (95% confidence bounds are given in parenthe-ses). With respect to the four polymorphisms tested,the observed differences between the allele frequenciesare not large enough to be significant after performinga Bonferroni correction (corrected P-value for bp594CfoI/T-allele 4 0.0864). The alleles of the HaeIIIbp145−55, CfoI bp1545, and StyI bp1674+14 polymor-phisms did not show significant differences to the allelefrequencies in the control subjects.

The present data suggest a possible allelic associa-tion of the bp594 CfoI T-allele with common IGE sub-types. Nevertheless, care should be taken in interpret-ing this association result. For example, populationstratification might lead to false-positive results, evenif both samples (patients and controls) were of the

TABLE II. Variants of the CHRNA4 Gene Detected by SSCA

Positiona

(bp) Aab Allelesc Enzyme Localization

145−8 — T/C +CfoI Intron 2145−55 — G/A −HaeIII Intron 2555 D C/T −HgaI Exon 5594 C C/T −CfoI Exon 51125 P G/T −AciI Exon 51143 C C/T −HgaI Exon 51545 S T/C +CfoI Exon 51575 A A/G +CfoI Exon 51674+11 — C/T −AciI Intron 51674

+14— G/A +StyI Intron 5

aThe nucleic acid numbering refers to the published CHRNA4 sequence[Steinlein et al., 1996].bAmino acid residue.cNucleic acid.

Fig. 1. Genotyping of polymorphisms: (from left to right) bp145−55

HaeIII, G-allele (homozygous), A/G-alleles (heterozygous); bp594 CfoI, C-allele (homozygous), T/C-alleles (heterozygous); bp1545 CfoI, C-allele (ho-mozygous), T-allele (homozygous); bp1674+14 StyI, A/G-alleles (heterozy-gous), A-allele (homozygous). Fragments smaller than 70bp are not shown.Asterisk 4 area of limited resolution.

TABLE I. Primer Sequences for Amplification of the CHRNA4 Gene

Primer A Primer B Exon

58 GGTGCGTGCGCCATGGAGC 38 58 GCAGTCAGGAGCCTGCCTC 38 158 ACCTGAGCCACTGGCCTGCC 38 58 CGACCTCAGTCACAGCGCAC 38 258 CCCGTCCACCATATCTTGC 38 58 GGCAGTGCCCTCCCACTC 38 3a

58 CATCCAGGAGTGGCACGACTA 38 58 ACCAAGGCCCTGTAGAGGAC 38 458 ACCCTTCGCTCTCTTCCTGC 38 58 CCACGATGACCCACTCGCC 38 558 CGGCTCCTGGACCTACGAC 38 58 CAGGCAGGGGATGATGAGG 38 558 GGCGAGTGGGTCATCGTGG 38 58 GATGACCAGTGAGGTGGACG 38 5b

58 CCTGCCCTCCGAGTGTGGC 38 58 GGGCATGGTGTGCGTGCGTG 38 558 ACGCACGCACACCATGCCC 38 58CGCGCAGGGTCCAGGCGAG 38 558 CCTTCCTGCAAGTCACCCTCC 38 58 TGCTCCGAGACCGGCACCG 38 558 GTGCTTTGGTGCTGCGGGTC 38 58 AGGGCTGGCCAAAGCCAGG 38 5c

58 CCCAAAGCGAAGCAGCCTGA 38 58 CTTCGGTGTCCCCGAGTGC 38 5d

58 GGCCGTGCTGGAGTGACG 38 58 GCCCCACAGAGTCCAGGG 38 6

58 GGCGAGTGGGTCATCGTGG 38 58 GGTGAGCGACAGCAGCCCG 38 5e

abp145−55/HaeIII.bbp 594/CfoI.cbp1545/CfoI.dbp1674+14/StyI.eSer248Phe.

Acetylcholine Receptor and Idiopathic Epilepsy 447

same ethnic origin. Furthermore, multiple tests havebeen performed. A possible correction could be obtainedby multiplying the P value by the number of allelestested (Bonferroni inequality). This approach is, how-ever, overstrict and increases the risk of a type-2 error.Therefore, it is generally preferable to check a possibleassociation against the results of another independentstudy in a new sample of IGE patients.

It is unlikely that the silent polymorphism itself isinvolved in the etiology of IGE, but that it is in linkagedisequilibrium with another variation which influencesseizure liability. Although intensive SSCA screening ofthe CHRNA4 gene has been performed in 103 patientswith common IGE subtypes in the present study, andin families with BFNC and LVEEG [Steinlein et al.,unpublished data], no mutations proving an etiologicalrole of this gene have been detected so far in the codingregion. The failure to detect mutations could be due toseveral factors. First, the detection rate of SSCA, evenif it is carried out under several temperature condi-tions, does not reach 100% [Hayashi and Yandell,1993]. Second, the promotor region of the CHRNA4gene has not yet been identified and sequenced. Thepossibility remains that mutations (or variations) inthe promotor exist which lead to changes in the expres-sion of the gene, thus influencing the seizure suscepti-bility. Furthermore, without RNA studies the possibil-ity of splice variants and/or intronic mutations cannotbe excluded. Third, despite the role of CHRNA4 in theetiology of ADNFLE [Steinlein et al., 1995b], the exis-tence of a second epilepsy gene in the vicinity ofCHRNA4 cannot be excluded at this time.

The frequency of the bp594 CfoI T-allele was in-creased in both IGE subgroups (CAE/JAE and JME,respectively), although the difference was not signifi-cant for the JME subgroup. This argues against aninfluence of the CHRNA4 locus on the seizure subtype,but rather points to a general effect on neuronal excit-ability determining the seizure threshold. This is con-sistent with the observation that approximately 12–

16% of the patients with BFNC exhibit seizures beyondthe neonatal period [Ryan et al., 1991; Ronen et al.,1993].

The extent of the possible impact of the CHRNA4locus on genetic susceptibility for generalized idio-pathic epilepsies remains to be determined. Results oflinkage studies argue against a frequent major geneinfluence. Negative lod scores have been observed withD20S19 (CMM6), a polymorphic marker located closeto CHRNA4, under four different trait models [Sanderet al., 1996]. Linkage studies, however, fail to uncovergenes of small effects [Lander and Schork, 1994; No-then et al., 1993].

The coincidence of positive linkage and/or associa-tion findings for IGE and BFNC on chromosome 20qand 8q [Leppert et al., 1989; Lewis et al., 1993; Zara etal., 1995] points to the possibility of an interactive age-dependent genetic influence on seizure liability frominfancy to early childhood [Sander et al., 1996]. It fur-ther supports the heuristic value of findings obtainedin rare monogenic idiopathic epilepsy syndromes forthe investigation of epilepsies with a more complex eti-ology.

ACKNOWLEDGMENTS

We thank M.M. Nothen for making the controlsavailable. The project was supported by the DeutscheForschungsgemeinschaft (SFB 400/B5; Be 1517/1-3).

REFERENCES

Beck-Mannagetta G, Anderson VE, Doose H, Janz D (eds) (1989): ‘‘Geneticsof epilepsies.’’ Stuttgart: Springer.

Berkovic SF, Andermann F, Andermann E, Gloor P (1987): Concepts ofabsence epilepsy: Discrete syndromes or biological continuum? Neurol-ogy 37:993–1000.

Berkovic SF, Howell RA, Hopper JL, Hay DA, Andermann E (1990): A twinstudy of epilepsy. Epilepsia 31:813.

Commission on Classification and Terminology of the International LeagueAgainst Epilepsy (1989). Epilepsia 30:389–399.

TABLE III. Allele Frequencies of CHRNA4 Polymorphisms in Unrelated Patients With IGEand IGE Subtypes and Unrelated Control Subjects

Whole sample SubgroupsCHRNA4Polymorphism

patients withIGE

patientsa

with JMEPatientsb

with CAE/JAE Controls

bp145−55/HaeIIIA-allele 15 (0.077) 11 (0.095) 5 (0.052) 12 (0.068)G-allele 179 (0.923) 105 (0.905) 91 (0.948) 164 (0.932)

bp 594/CfoIT-allele 16 (0.085)c 8 (0.073)d 8 (0.085)e 5 (0.027)C-allele 172 (0.915) 102 (0.927) 86 (0.915) 179 (0.973)

bp1545/CfoIT-allele 95 (0.490) 61 (0.508) 38 (0.422) 77 (0.475)C-allele 99 (0.510) 59 (0.492) 52 (0.578) 85 (0.525)

bp1674 + 14/StyIA-allele 155 (0.791) 93 (0.788) 74 (0.771) 155 (0.824)G-allele 41 (0.209) 25 (0.212) 22 (0.229) 33 (0.176)

Allele frequencies are given in parentheses.aIncluding two patients with JME and CAE, and seven patients with JME and JAE.bIncluding nine patients with CAE/JAE and JME.cx2 4 5.28, df 4 1, P 4 0.0216.dx2 4 3.13, df 4 1, P 4 0.0769 (n.s.).ex2 4 4.23, df 4 1, P 4 0.0397.Not all alleles could be typed for each individual.

448 Steinlein et al.

Delgado-Escueta AV, Serratosa JM, Liu A, Weissbecker K, Medina MT,Gee M, Treiman LJ, Sparkes R (1994): Progress in mapping humanepilepsy genes. Epilepsia 35:29–40.

Durner M, Sander T, Greenberg DA, Johnson K, Beck-Mannagetta G, JanzD (1991): Localization of idiopathic generalized epilepsies on chromo-some 6p in families ascertained through juvenile myoclonic epilepsypatients. Neurology 41:1651–1655.

Gedda L, Tatarelli R (1971): Essential isochronic epilepsy in MZ twin pairs.Acta Genet Med Gemellol 20:380–383.

Greenberg DA, Delgado-Escueta AV, Widelitz H, Sparkes RS, Treiman L,Maldonado HM, Parks MS, Terasaki PI (1988): Juvenile myoclonicepilepsy (JME) may be linked to the Bf and HLA loci on chromosome 6.Am J Med Genet 31:185–192.

Hayashi K, Yandell DW (1993): How sensitive is PCR-SSCP? Hum Mut2:338–346.

Janz D, Beck-Mannagetta G, Sander T (1992): Do idiopathic generalizedepilepsies share a common susceptibility gene? Neurology 42:48–55.

Lander ES, Schork NJ (1994): Genetic dissection of complex traits. Science265:2037–2048.

Leppert M, Anderson VE, Quattlebaum T, Stauffer D, O‘Conell P, Naka-mura Y, Lalouel JM, White R (1989): Benign familial neonatal convul-sions linked to genetic markers on chromosome 20. Nature 337:647–648.

Lewis TB, Leach RJ, Ward K, O‘Connell P, Ryan SG (1993 ): Geneticheterogeneity in benign familial neonatal convulsions: Identification ofa new locus on chromosome 8q. Am J Hum Genet 53:670–675.

Liu AW, Delgado-Escueta AV, Serratosa JM, Alonso ME, Medina MT, GeeMN, Cordova S, Zhao HZ, Spellman JM, Peek JRR, Donnadieu SR,Sparkes RS (1995): Juvenile myoclonic epilepsy locus in chromosome6p21.2-p11: Linkage to convulsions and electroencephalography trait.Am J Hum Genet 57:368–381.

Liu AW, Delgado-Escueta AV, Gee MN, Serratosa JM, Zhang QW, AlonsoME, Medina MT, Cordova S, Zhao HZ, Spellman JM, Donnadieu SR,Peek JR, Treiman LJ, Sparkes RS (1996): Juvenile myoclonus epilepsyin chromosome 6p12-p11: Locus heterogeneity and recombinations. AmJ Med Genet 63:438–446.

Malafosse A, Leboyer M, Dulac O, Navelet Y, Plouin P, Beck C, Laklou H,Mouchnino G, Grandscene P, Vallee L, Guilloud-Bataille M, SamolykD, Baldy-Moulinier M, Feingold J, Mallet J (1992): Confirmation oflinkage of benign familial neonatal convulsions to D20S19 and D20S20.Hum Genet 89:54–58.

Noda M, Takahashi H, Tanabe T, Toyosato M, Furutani Y, Hirose T, AsaiM, Inayama S, Miyata T, Numa S (1982): Primary structure of a-sub-unit precursor of Torpedo californica acetylcholine receptor deducedfrom cDNA sequence. Nature 299:793–823.

Nothen MM, Propping P, Fimmers R (1993): Association versus linkagestudies in psychosis genetics. J Med Genet 30:634–637.

Phillips HA, Scheffer IE, Berkovic SF, Hollway GE, Sutherland GR, Mul-ley, JC (1995): Localisation of a gene for autosomal dominant nocturnalfrontal lobe epilepsy to chromosome 20q13. Nature Genet 10:117–118.

Reutens DC, Berkovic SF (1995): Idiopathic generalized epilepsy of ado-lescence: Are the syndromes clinically distinct? Neurology 45:1469–1476.

Ronen GM, Rosales TO, Conolly M, Anderson VE, Leppert M (1993): Be-nign familial neonatal convulsions: Clinical and DNA analysis of alarge Newfoundland kindred. Epilepsia 32:72.

Ryan SG, Wiznitzer M, Hollman C, Torres MC, Szekeresova M, Schneider

S (1991): Benign familial neonatal convulsions: Evidence for clinicaland genetic heterogeneity. Ann Neurol 29:469–473.

Sander T, Hildmann T, Janz D, Wienker TF, Neitzel H, Bianchi A, BauerG, Sailer U, Berek K, Schmitz B, Beck-Mannagetta G (1995): The phe-notypic spectrum related to the human epilepsy susceptibility gene‘‘EJM1.’’ Ann Neurol 38:210–217.

Sander T, Hildmann T, Wienker TF, Ramel C, Beck-Mannagetta G, Bian-chi A, Sailer U, Berek K, Bauer G, Neitzel H, Schmitz B, Durner M,Johnson KJ, Janz D (1996): Common subtypes of idiopathic generalizedepilepsies: Lack of linkage to D20S19 close to candidate loci (EBN1,EEGV1) on chromosome 20. Am J Med Genet (Neuropsychiat Genet)67:31–39.

Serratosa JM, Delgado-Escueta AV, Medina MT, Zhang Q, Iranmanesh R,Sparkes R (1996): Clinical and genetic analysis of a large pedigree withjuvenile myoclonic epilepsy. Ann Neurol 39:187–195.

Steinlein O (1995): Detection of a CfoI polymorphism within exon 5 of thehuman neuronal nicotinic acetylcholine receptor a4 subunit gene(CHRNA4). Hum Genet 96:130.

Steinlein O, Anokhin A, Yping M, Vogel F (1992): Localization of a gene forthe human low voltage EEG on 20q and genetic heterogeneity. Genom-ics 12:69–73.

Steinlein O, Smigrodzki R, Lindstrom J, Anand R, Kohler M, Tocharoen-tanaphol C, Vogel F (1994): Refinement of the localization of the genefor neuronal nicotinic acetylcholine receptor a4 subunit (CHRNA4) tohuman chromosome 20q13.2-q13.3. Genomics 22:493–495.

Steinlein O, Schuster V, Fischer C, Haussler M (1995a): Benign familialneonatal convulsions: Confirmation of genetic heterogeneity and fur-ther evidence for a second locus on chromosome 8q. Hum Genet 95:411–415.

Steinlein O, Mulley JC, Propping P, Wallace RH, Phillips HA, SutherlandGR, Scheffer IE, Berkovic SF (1995b): A missense mutation in theneuronal nicotinic acetylcholine receptor a4 subunit is associated withautosomal dominant nocturnal frontal lobe epilepsy. Nature Genet 11:201–203.

Steinlein O, Weiland S, Stoodt J, Propping P (1996): Exon-intron structureof the human neuronal nicotinic acetylcholine receptor a4 subunit(CHRNA4). Genomics 32:289–294.

Suarez BK, Hampe CL, Eerdewegh PV (1994): Problems of replicatinglinkage claims in psychiatry. In Gershon ES, Cloninger CR (eds): ‘‘Ge-netic Approaches to Mental Disorder.’’ Washington, DC, London:American Psychiatric Press, pp 24–46.

Weiland S, Witzemann V, Villaroel A, Propping P, Steinlein O (1996) Anamino acid exchange in the second transmembrane segment of a neu-ronal nicotinic receptor causes partial epilepsy by altering its desensi-tization kinetics. FEBS Letters 398:91–96.

Weissbecker KA, Durner M, Janz D, Scaramelli A, Sparkes MA, SpenceMA (1991): Confirmation of linkage between juvenile myoclonic epi-lepsy and the HLA region on chromosome 6. Am J Med Genet 38:32–36.

Whitehouse WP, Rees M, Curtis D, Sunquist A, Parker K, Chung E, Bar-alle D, Gardiner RM (1993): Linkage analysis of idiopathic generalizedepilepsy (IGE) and marker loci on chromosome 6p in families of pa-tients with juvenile myoclonic epilepsy: No evidence for an epilepsylocus in the HLA region. Am J Hum Genet 53:652–662.

Woolf D (1955): On estimating the relation between blood group and dis-ease. Ann Hum Genet 19:251–253.

Zara F, Bianchi A, Avanzini G, Di Donato S, Castellotti B, Patel PI, Pan-dolfo M (1995): Mapping of genes predisposing to idiopathic generalizedepilepsy. Hum Mol Genet 4:1201–1207

Acetylcholine Receptor and Idiopathic Epilepsy 449