D.R. Nyholt, BSc Rons; R.A. Lea, BSc Rons; P.J.Goadsby, MD, PhD; P.J.Brimage, FRACP; and L.R. Griffiths, PhD

D.R. Nyholt, BSc Rons; R.A. Lea, BSc Rons; P.J.Goadsby, MD, PhD; P.J.Brimage, FRACP; and L.R. Griffiths, PhD Article abstract-migraine is a frequent familial disorder that, in common with most multifactorial disorders, has an unknown etiology. The authors identified several families with multiple individuals affected by typical migraine using a single set of diagnostic criteria and studied these families for cosegregation between the disorder and markers on chromosome 19, the location of a mutation that causes a rare form of familial hemiplegic migraine (FHM). One large tested family showed both cosegregation and significant allele sharing for markers situated within or adjacent to the FHM locus. Multipoint GENEHUNTER results indicated significant excess allele sharing across a 12.6-cM region containing the FHM Ca2+ channel gene, CACNLl.A4 (~aximum nonparametric linkage Z score = 6.64, p = 26), with a maximum parametric lod score of 1.92 obtained for a (CAG)n triplet repeat polymorphism situated in exon 47 of this gene. The CAG expansion did not, however, appear to be the cause of migraine in this pedigree. Other tested families showed neither cosegregation nor excess allele sharing to chromosome 19 markers. HOMOG analysis indicated heterogeneity, generating a maximum HLOD score of 3.6. It was concluded that Chr19 mutations either in the CACNLlA4 gene or a closely linked gene are implicated in some pedigrees with familial typical migraine, and that the disorder is genetically heterogeneous. NEUROLOGY 1998;50:1428-1432 Migraine is a common complex disorder that shows strong familial aggregation. The disorder is generally characterized by chronic episodic headache that is usually associated with nausea and vomiting. Interest in the localization of a common migraine gene has been stimulated by the recent mapping and subsequent identification of a gene involved in the rare familial hemiplegic migraine (FHM) subtype. The prevalence of migraine has been shown to vary between 10%1 and 23%2 depending on the population studied, with a recent large study in the United States indicating that 18% of women, 6% of men, and 4% of children suffer from the disorder.3 Migraine shows strong familial aggregation, although the mode of transmission is controversial. A recent review of twin, spouse, and family studies strongly suggested that the two major types of migraine, migraine with aura and migraine without aura, are genetically determined, with the mode of inheritance most likely multifactorial. However, autosomal- From the Genomics Research Centre (Drs. Griffiths, Nyholt, and Lea), Griffith University-Gold Coast, Queensland, Australia; The National Hospital for Neurology and Neurosurgery (Dr. Goadsby), London, England; and the Institute of Neurological Sciences (Dr. Brimage), Prince ofwales Hospital, Randwick, Australia. Supported by funding from Griffith University-Gold Coast and the Australian Government Employees Medical Research Fund. The work of Dr. Peter Brimage was supported by a Headache Fellowship provided by Glaxo Wellcome, and Dr. Peter Goadsby is a Wellcome Senior Research Fellow. Received September 9, 1997. Accepted in final form December 16, 1997. Address correspondence and reprint requests to Dr. Lyn R. Griffiths, Genomics Research Centre, School of Health Science, Griffith University, Gold Coast, PMB 50 GCMC, Qld. Australia, 9726. 1428 Copyright @ 1998 by the American Academy of Neurology

dominant inheritance with reduced penetrance could not be excluded in either subtype of migraine.4 Segregation analyses by Mochi et al.5 suggested that there may be both a common genetic background for the two types of migraine and a major gene contributing to the disease. This idea has been supported by similarities in medication response in the two types of migraine6 and also by the fact that the two types can occur in the same family and even in the same individual. In 1993, a gene for a rare, distinct subtype of migraine, FHM, was mapped to chromosome 19p13.7 Subsequent studies indicated heterogeneity, with about 50% of tested families showing linkage to this genomic region.7-10 More recently, mutations in a newly characterized brain-specific P/Q-type Ca2+ channel a1-subunit gene, CACNLlA4 located on chromosome 19p13, have been shown to be involved in some FHM pedigrees. A similar etiology may be involved in more common types of migraine.11 However, two studies that have tested this region for linkage to typical migraine have produced conflicting results. The first, using two-point and multipoint parametric linkage analysis, excluded the FHM locus in typical migraine,12 whereas the second, using nonparametric affected sibling-pair analysis in 28 unrelated families, suggested that the FHM locus is involved in migraine. The second study, using quite a small number of samples, found the number of shared alleles in affected siblings was 18.4 versus 9.2 nonshared alleles, resulting in a significant p value of 0.04.13 The differing results from these two studies, combined with the small sample size used in the second study, provide inconclusive evidence for a typical migraine susceptibility locus on chromosome 19p13. Therefore, further investigation of chromosome 19 involvement in the common forms of migraine with and without aura is warranted. We investigated chromosome 19 involvement in typical migraine in four large pedigrees. Using seven microsatellite markers that span the FHMimplicated Ca2+ channel CACNLlA4 gene on 19p13,11 we simultaneously tested for genetic linkage and heterogeneity of common migraine. One of the tested fainilies gave evidence of chromosome 19 involvement. To examine the role of this chromosome further in this typical migraine pedigree, we performed a complete linkage scan of chromosome 19 using 16 microsatellite markers.~ Methods. Families and diagnoses. Before commencement, all research was approved by Griffith University's Ethics Committee for Experimentation on Humans. All individuals donating blood samples gave informed consent and all were Caucasian. Diagnosis of migraine was performed by a clinical neurologist following the criteria specified by the International 'Headache Society (IHS).14 Affected individuals were classified as migraine with aura or migraine without aura, as previously described.15 DNA analysis. A standard SDS-proteinase K method16 incorporating a salting-out procedure17 was used to extract DNA from blood samples. Primer sequences for microsatellite markers were obtained from both published reports18 and the Genome Database.19 Polymerase chain reactions (PCR) for microsatellite markers were performed using 25 ng genomic DNA, 200 nm of each primer, and 1.75 mm MgC12 in a final volume of 20 ~L. Samples were subjected to an initial denature of 5 minutes at 94 C, followed by 35 cycles of 1 minute at 94 C, 20 seconds at 55 C, 20 seconds at 57.5 C, and 20 seconds at 60 C with a final extension period of 2 minutes at 72 C. Fluorescent-labeled PCR products were fractionated by capillary electrophoresis through a 4% high-resolution denaturing polymer using an ABI Prism 310 genetic analyzer with alleles determined using GENESCAN software (Applied Biosystems, Perkin- Elmer, Foster City, CA). Linkage analyses. Allele frequencies for 10 of the 16 microsatellite markers were calculated directly from a genotyped control population. Allele frequencies for the remaining six markers were calculated from family data.2 Genotypes for pedigree members were assessed and analyzed for linkage using parametric and nonparametric techniques. Pairwise lod scores between each marker and migraine were calculated using the parametric FASTLINK program.21-23 Multipoint lod scores were calculated using the VITESSE program.24 For these parametric lod score calculations, the exact mode of transmission of the disease needs to be specified. A conservative model similar to that used by Hovatta et al. (1994), assuming an autosomaldominant mode of inheritance with 70% penetrance and a phenocopy rate of 0.7%, was used.12 The frequency of the disease was set at 12%.3 Two-point lod scores, at recombination fractions (8) of, 0.05, 0.1,, 0.3, and 0.4, calculated using F ASTLINK, were then used as input for the F ASTMAP program.25 Using FASTMAP, we evaluated multipoint lod scores at 20 equidistant points across the 12.6-cM region between the and D198410 microsatellite markers. The resulting multipoint lod scores contained information from all seven tested microsatellite markers across a large number of recombination fractions and were subsequently used in a test for heterogeneity using the HOMOG program.26 For GENEHUNTER version 1.1 analysis,27 "max bits" was set at 20, the "score all" option and Haldane map function were used, and the "skip large" option was switched off,.thereby allowing "trimming" of the pedigrees. Use of the skip large off option resulted in the elimination of unaffected individuals starting from the pedigree base. This was necessary in MFl, MF7, and MF14. In addition, a lateral branch of the MFl pedigree, involving an affected spouse lineage, was also excluded. This was made necessary due to constraints on the size offamilies able to be processed by GENEHUNTER. As a compromise between processing demands and maximization of the information content across the region under investigation, information from only five of the seven markers was used in the initial multipoint GENE- HUNTER analysis. For the chromosome scan, the following genetic map order was used, with map distances (recombination fractions) given in parentheses: D198209--{0.130)-IN8R- ( 0.110 )--( 1 )-D 198394-( 0.020 )-D 198221- (O.O34)-D 1981150-(1H CAG)n --{0.024)-D 198179- (0. 016)-D 198226--(0 )-D 198415--{0.020 )-D 198410- May 1998 NEUROLOGY 50 1429

( 0. 100)-D 198414-( 0.100 )-D 198220-( 0.060 )-D 198420- (60)-D 198418-(0.080)-D 198210. Results. Families. Four large multigenerational families identified in Australia were recruited for this study. These included 121 individuals for whom DNA was available and 118 who underwent neurologic examinations. For this report, 52 individuals (51 with DNA available) were classified as affected after having a clinical diagnosis of either migraine with aura or migraine without aura. Blood samples collected from 96 unaffected individuals were used as a control population for the calculation of microsatellite marker allele frequencies. Parametric lod score analyses. Pairwise parametric lod score calculations for the seven tested microsatellite markers (,,,,, and ) spanning the FHM locus on chromosome 19p13 were carried out with the FASTLINK program package,21-23 with the maximum pairwise lod scores summarized in table 1. Three of the four migraine families, MF7, MFI4, and MFI5, produced mostly negative lod scores, suggesting a lack of involvement of a chromosome 19p13 locus in these three families. One pedigree (MFl), however, produced positive lod scores for all of the initial seven markers tested, the highest being 1.56 for the marker. Parametric 8-point multipoint lod scores were calculated for MFl by the VITESSE program,24 producing a maximum lod of 1.5 at 10 cm from the marker. However, two-point lod and 8 calculations could be considered more robust than multipoint calculations, if there are questions about model parameters or diagnostic criteria, as is common in complex diseases.28.29 Consequently, we used two-point lod scores calculated by FASTLINK to estimate multipoint lod scores using the FASTMAP program.25 FASTMAP results for MFl gave a maximum multipoint lod score of 2.0776 at the marker. The FASTMAP multipoint lod scores were also used in our subsequent test for heterogeneity using the HOMOG program.26 HOMOG analysis, using information from all seven markers in the four tested pedigrees, gave evidence for heterogeneity to the 2.8% significance level with a likelihood ratio in favor of heterogeneity of 36.13 and a combined lod score (HLOD) supporting linkage and heterogeneity of 3.59. Also, HOMOG calculated a high posterior probability (0.99) of MFl being linked to chromosome 19. A complete chromosome 19 scan using 16 microsatellite markers including two markers (D19S1150 and a 3'(CAG) repeat) located within the CACNLIA4 gene was then undertaken in MFl. Analysis of these chromosome scan results further implicated the 12.6-cM region between and, with a maximum two-point lod score of 1.92 (8 = 0.06) obtained for the (CAG)n trinucleotide repeat polymorphism, which lies in exon 47 of the CACNLIA4 gene.11 Analysis of this polymorphism in MFl did not reveal a higher than normal level of repeats,11.30 and expansion was not indicated (figure 1). In addition, FASTMAP 10-point analysis using the,,, D19S1150, (CAG)n,,,, and markers produced a maximum lod score of 2.4326 near the locus (figure 2). Nonparametric analyses. To overcome problems arising from misspecification of transmission model parameters, we also used the GENEHUNTER program package.27 1430 NEUROLOGY 50 May 1998 Tab~ 1 Pairwise FASTLINK lad scores Familynocus & Zmax Family1 Family7 Family 14 Family 15 0.1 0.05 0.05 0.4 0.4 0.3 0.1 0.80 0.47 1.56 1.54 0.64 1.00 0.02 0.10 0.07 0.13 GENEHUNTER extracts complete multipoint data from affected relatives in general pedigrees of modest size, implementing a nonparametric linkage (NPL) analysis. Initial multipoint GENEHUNTER analysis for the four pedigrees used information from the,,,, and markers. Maximum NPL Z scores and their corresponding p values for all four families are summarized in table 2. For MF1 only, we obtained significant p values for the five markers, indicating a large excess of allele sharing, with a maximum NPL Z score of 5.54 (p = 3235) near. When an analysis of the 16 chromosome 19 markers was undertaken in MF1, GENEHUNTER results from the complete chromosome scan showed an even higher degree of allele sharing across this genomic region, with a maximum NPL Z score of 6.64 (p = 2686) near the locus (figure 3). 0.16 0.17 0.06 0.03 0.60

Figure 1. Pedigree of migraine family 1 (MFl). All individuals were diagnosed as having migraine with aura (MA), having migraine without aura (MO), or as being unaffected (blank), following International Headache Society guidelines, and are given directly under each symbol. The number of repeats for the CACNLIA4 exon 47 (GAG) trinucleotide repeat polymorphism are indicated below migraine status. Proband is indicated by an arrow and the letter P. Discussion. Sequencing of the CACNLlA4 gene has found a number of deleterious mutations in FHM and another neurologic disorder, hereditary paroxysmal cerebellar ataxia (also known as episodic ataxia type 2 or EA-2). Four missense mutations in five unrelated families have been implicated in FHM, and two mutations, disrupting the reading frame of CACNLlA4, have been found in two families with EA-2.1l In addition, a third type of mutation in CACNLlA4 has been linked to another form of human ataxia. Specifically, Zhuckenko et al.3 were able to link CAG-expanded versions of CACNLlA4 to a slowly progressing form of ataxia, designated spinocerebellar ataxia 6 (SCA6). Thus, these three neurologic disorders are allelic channelopathies involving the CACNLlA4 gene on chromosome 19. Ophoff et au1 raised the possibility that a similar defect may be involved in the common forms of migraine and suggested investigation of the polymorphic trinucleotide (CAG) repeat in the common forms of migraine with and without aura. The observed repeat length of (CAG)n in control populations varies from n = 4 to n = 16.11,30 Analysis of this repeat in our large Chrl9-linked migraine pedigree (MFl) showed no evidence of expansion within the family, with the largest number of repeats being 15 (figure 1). The CAG repeat therefore does not appear to be implicated causally in MFl. Other mutations in the CACNLlA4 gene, however, may be found responsible, although our linkage results gave a maximum lod at a distance approximately 6 cm from the CACNLlA4 gene. There may be similar but uncharacterized voltage-gated channel genes located within this genomic region. Two Ca2+ channel genes, CACNLlA3 and CACNLlA6, have already been shown to be colocated on chromosome lq31-q3231 and lq25-q31,32,33 respectively. Another possible candidate gene, with a known location near CACNLlA4, Figure 2. Results of F ASTMAP multipoint lod score analysis of nine chromosome 19p13 markers in migraine family 1 (MF1). Table 2 GENEHUNTER* analysis Family ~ 1 14 15 NPL statisticmax 5.539304 0.882726-0.124424 0.469345 ~fj~ p Value 3235 0.148438 0.387390 0.082703 * Using markers,,,,. NPL = nonparametric linkage.

is a potassium voltage-gated channel gene (KCNA7), which has been implicated in episodic ataxia/ myokymia type 1 (EA-1).34.35 The results of our chromosome 19 scan involving four familial typical migraine pedigrees and several marker loci provide support for genetic heterogeneity of typical migraine and suggestive evidence for the presence of a susceptibility locus on chromosome 19p13. Therefore, similar to results obtained for FHM, there is now evidence that common migraine is a heterogeneous disorder and that one gene for typical migraine is located on chromosome 19p13. There is obviously an urgent need to determine whether mutations in the CACNLlA4 gene are involved in the common forms of migraine, or whether typical migraine is in fact caused by mutations in a gene other than the FHM-implicated CACNLlA4. Acknowledgments The authors thank R. Williamson in particular, and also M. Eadie, J. MacMillan, S. Wilson, J. Ott, and L. Kruglyak for helpful discussions. The authors also thank Ms. Sharon Quinlan for assistance in collecting pedigree blood samples and Mr. Robert Curtain for assistance in the collection of genotypic data. References 1. Kurtzke JF.The current neurologic burden of illness and injury in the United States. Neurology 1982;32:1207-1214. 2. Dalsgaard-Nielsen T, Ulrich J. Prevalence and heredity of lnigraine and lnigrainoid headaches among 461 Danish doctors. Headache 1972;12:168-172. 3. Stewart WF, Lipton RB, Celentano DD, Reed ML. Prevalence of lnigraine headache in the United States. JAMA 1992;267: 64-69. 4. Russell MB, Olesen J. The genetics of lnigraine without aura and lnigraine with aura. Cephalalgia 1992;13:245-248. 5. Mochi M, Sangiorgi S, Cortelli P, et al. Testing models for genetic detennination in lnigraine. Cephalalgia 1993;13:389-394. 6. Blau IN.Migraine with and without aura are not different entities. Cephalalgia 1995;15:186-189. 7. Joutel A, Bousser M-G, Biousse V, et al. A gene for familial helniplegic lnigraine maps to chromosome 19. Nat Genet 1993; 5:40-45. 8. OphoffHA, Van Eijk R, Sandkuijl LA, et al. Geneticheterogeneity of familial helniplegic migraine. Genolnics 1994;22:21-26. 9. OphoffHA, Terwindt GM, Vergouwe MN, et al. A 3 Mb region for the familial helniplegic lnigraine locus on 19p13.1-pI3.2; exclusion of PRKCSH as a candidate gene. Eur J Hum Genet 1996;4:321-328. 10. Joutel A, Ducros A, Vahedi K, et al. Genetic heterogeneity of familial helniplegic. Am J Hum Genet 1994;55:1166-1172. 11. Ophoff:RA, Terwindt GM, Vergouwe MN, et al. Familial helniplegic lnigraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene CACNL1A4. Cell 1996;87: 543-552. 12. Hovatta L, Kallela M, Farkkila M, Peltonen L. Familial Inigraine: exclusion of the susceptibility gene from the reported locus of familial helniplegic lnigraine on 19p. Genolnics 1994; 23:707-709. 13. May A, Ophoff HA, Terwindt GM, et al. Familial helniplegic lnigraine locus on 19p13 is involved in the common forms of migraine with and without aura. Hum Genet 1995;96:604-608. 14. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias, and facial pain. Cephalalgia 1988;8(suppl 7):19-28. 15. Griffiths LR, Nyholt DR, Curtain RP, Goadsby PJ, Brimage PJ. Migraine association and linkage studies of an endothelial nitric oxide synthase (NOS3) gene polymorphism. Neurology 1997;49:614-617. 16. Blin N, Stafford DW. Isolation of high molecular-weight DNA. Nucleic Acids Res 1976;3:2303. 17. Miller SA, Dykes DD, Plensky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 1988;16:1215. 18. Gyapay G, Morissette J, Vignal A, et al. The 1993-94 Genethon human genetic linkage map. Nat Genet 1994;7:246-339. 19. Johns Hopkins University School of Medicine, host. The genome database. Baltimore (http:llgdbwww.gdb.org/gdb/ gdbtop.html/). 20. Boehnke M. Allele frequency estimation from data on relatives. Am J Hum Genet 1991;48:22-25. 21. Lathrop GM, Lalouel JM, Julier C, Ott J. Multilocus linkage analysis in humans: detection of linkage and estimation of recombination. Am J Hum Genet 1985;37:482-498. 22. Cottingham RW Jr, Idury RM, Schaffer AA. Faster sequential genetic linkage computations. Am J Hum Genet 1993;53:252-263. 23. Schaffer AA, Gupta SK, Shiram K, Cottingham RW Jr. Avoiding recomputation in linkage analysis. Hum Hered 1994;44: 225-237. 24. O'Connell JR, Weeks DE. The VITESSE algorithm for rapid exact multilocus linkage analysis via genotype set-recoding and fuzzy inheritance. Nat Genet 1995;11:402-408. 25. Curtis D, Gurling HMD. A procedure for combining two-point lod scores into a summary multipoint map. Hum Hered 1993; 43:173-185. 26. Ott J. Analysis of human genetic linkage. Baltimore: Johns Hopkins University Press, 1991. 27. Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. Parametric and non-parametric linkage analysis: a unified multipoint approach. Am J Hum Genet 1996;58:1347-1363. 28. Risch N. Linkage strategies of genetically complex traits. I. Multilocus models. Am J Hum Genet 1990;46:222-228. 29. Terwilliger JD, Ott J. Handbook of genetic linkage. Baltimore: Johns Hopkins University Press, 1994. 30. Zhuchenko 0, Bailey J, Bonnen P, et al. Autosomal dominant cerebellar ataxia (SCA6) associated with small polyglutamine expansions in the alpha(1a)-voltage-dependent calcium channel. Nat Genet 1997;15:62-69. 31. Diriong S, Lory P, Williams ME, Ellis SB, Harpold MM, Taviaux S. Chromosomal localisation of the human genes for alpha-1a, alpha-1b, and alpha-1e voltage-dependent Ca2+ channel subunits. Genomics 1995;30:605-609. 32. Gregg RG, Couch F, Hogan K, Powers P. Assignment of the human gene for the alpha-1 subunit of the skeletal muscle DHP-sensitive Ca2+ channel (CACNLlA3) to chromosome 1q31-32. Genomics 1993;15:107-112. 33. lies DE, Segers B, Weghuis DO, et al. Refined localisation of the alpha-1 subunit of the skeletal muscle L-type voltage dependent calcium channel (CACNL1A3) to human chromosome 1q32 by in situ hybridization. Genomics 1994;19:561-563. 34. Litt M, Kramer P, Browne DL, et al. A gene for episodic ataxia/myokymia maps to chromosome 12p13. Am J Hum Genet 1994;55:702-709. 35. Browne DL, Gancer ST, Nutt JG, et al. Episodic ataxia/ myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNAl. Nat Genet 1994;8: 136-140. 1432 NEUROLOGY 50 May 1998