OUP user menu

The genetic spectrum of a population-based sample of familial hemiplegic migraine

L. L. Thomsen, M. Kirchmann, A. Bjornsson, H. Stefansson, R. M. Jensen, A. C. Fasquel, H. Petursson, M. Stefansson, M. L. Frigge, A. Kong, J. Gulcher, K. Stefansson, J. Olesen
DOI: http://dx.doi.org/10.1093/brain/awl334 346-356 First published online: 2 December 2006

Summary

Familial hemiplegic migraine (FHM) is a rare subtype of migraine with aura and transient hemiplegia. FHM mutations are known in three genes, the CACNA1A (FHM1) gene, the ATP1A2 (FHM2) and the SCN1A (FHM3) gene and seem to have an autosomal-dominant mode of inheritance. The aim of this study was to search for FHM mutations in FHM families identified through a screen of the Danish population of 5.2 million people. FHM patients were diagnosed according to the International Classification of Headache Disorders and all FHM patients had a physical and neurological examination by a physician. A total of 147 FHM patients from 44 different families were identified; 43 FHM families participated in this study. Linkage analysis of these families shows clear linkage to the FHM locus (FHM1) on chromosome 19, supportive linkage to the FHM2 locus whereas no linkage was found to the FHM3 locus. Furthermore, we sequenced all exons and promoter regions of the CACNA1A and ATP1A2 genes and screened for the Q1489K mutation in the SCN1A gene. CACNA1A gene mutations were identified in three of the FHM families, two known FHM mutations, R583Q and T666M and one novel C1369Y mutation. Three FHM families were identified with novel mutations in the ATP1A2 gene; a family with a V138A mutation, a family with a R202Q mutation and a family with a R763C mutation. None of the Danish FHM families have the Q1489K mutation in the SCN1A gene. Our study shows that only 14% (6/42) of FHM families in the general Danish population have exonic FHM mutations in the CACNA1A or ATP1A2 gene. The families we identified with FHM mutations in the CACNA1A and ATP1A2 genes were extended, multiple affected families whereas the remaining FHM families were smaller. The existence of many small families in the Danish FHM cohort may reflect less bias in FHM family ascertainment and/or more locus heterogeneity than described previously.

  • familial hemiplegic migraine
  • mutations
  • genome scan
  • genetics
  • loci

Introduction

Familial hemiplegic migraine (FHM) is a rare subtype of migraine with aura, where the aura includes some degree of hemiparesis and where at least one first- or second-degree relative has attacks of migraine with some degree of hemiparesis. In most large FHM families studied FHM is inherited in an autosomal dominant manner. Mutation screening of such families has revealed missense mutations in the CACNA1A gene on chromosome 19p13 encoding a Cav2.1 calcium channel subunit (FHM1) (Ophoff et al., 1996); missense mutations in the ATP1A2 gene on chromosome 1q23 encoding a Na+/K+-ATPase pump subunit (FHM2) (De Fusco et al., 2003); or a heterozygote missense mutation in the SCN1A gene on chromosome 2q24 encoding a Nav2.1 sodium channel subunit (FHM3) (Dichgans et al., 2005). Seventeen different missense mutations have been identified in the CACNA1A gene worldwide causing FHM1 (Ophoff et al., 1996; Battistini et al., 1999; Ducros et al., 1999, 2001; Friend et al., 1999; Lykke Thomsen et al., 2002; Takahashi et al., 2002; Terwindt et al., 2002; Wada et al., 2002; Alonso et al., 2003; Kors et al., 2003, 2004; Vanmolkot et al., 2003; Beauvais et al., 2004; Jen et al., 2004; Jurkat-Rott et al., 2004; Spadaro et al., 2004). The T666M mutation is the most frequent CACNA1A mutation reported, having been found in 20 families worldwide (Ophoff et al., 1996; Ducros et al., 1999, 2001; Friend et al., 1999; Takahashi et al., 2002; Terwindt et al., 2002; Wada et al., 2002; Kors et al., 2003; Jen et al., 2004), and the R583Q mutation is the second most recurrent mutation reported in six families (Battistini et al., 1999; Ducros et al., 2001; Terwindt et al., 2002; Alonso et al., 2003). Most other CACNA1A mutations have been found in only one or two FHM families worldwide. In FHM1, patients with cerebellar ataxia tend to have hemiplegic attacks with interictal ataxia. There is though some phenotype variation in one family with the R583Q mutation in the CACNA1A gene, where some of the patients have only cerebellar ataxia (Alonso et al., 2003).

In the ATP1A2 gene 27 different missense mutations have been identified causing FHM2 (Headache Classification Committee of the International Headache Society, 1988; Russell et al., 1995; Rozen and Skaletsky, 2000; Kong et al., 2002; Thomsen et al., 2002; Bjornsson et al., 2003; De Fusco et al., 2003; Fossdal et al., 2004; Gudbjartsson et al., 2005; Eriksen et al., 2006). In FHM2 families cerebellar signs are rare; however, transient and permanent cerebellar signs have been reported in an Italian FHM family with a G301R mutation (Spadaro et al., 2004).

In the newly identified SCN1A gene on chromosome 2q24 a heterozygous missense mutation (Q1489K) was identified in 3 FHM families (18 FHM-affected individuals) of European origin. None of these patients had cerebellar symptoms; however, three patients had epileptic seizures during infancy. Thus co-occurrence of infantile seizures and FHM has been reported in FHM1, FHM2 and FHM3 (De Fusco et al., 2003; Vanmolkot et al., 2003; Beauvais et al., 2004; Jurkat-Rott et al., 2004; Kors et al., 2004; Spadaro et al., 2004). Mutations in the CACNA1A, ATP1A2 and SCN1A genes explain 50–70% of published families with FHM. However, these families are selected from hospitals or specialist practice and very likely represent families with higher penetrance and more severe symptomatology compared with cases from the general population. It is therefore possible that the frequency of mutations in the CACNA1A, ATP1A2 and SCN1A genes described previously may be different in families with FHM from the general population. In order to describe the full genetic spectrum of FHM in the reported genes and to identify new genes involved in FHM, we investigated a population-based sample of families with FHM.

Material and methods

Ascertainment of families with FHM

A systematic search was performed employing three different strategies in the entire Danish population of 5.2 million people. The search included a computer search of the Danish National Patient Register of all hospitalized patients with a discharge diagnosis of migraine with aura (MA) or migraine with complication (ICD-10 diagnosis DG431 or DG433), screening of >27 000 case records from headache clinics and practising neurologists in Denmark and additionally, advertisements for patients with hemiplegic migraine (HM) were placed in the major Danish Medical Journal and in the journal of the major Danish organization of headache patients (Lykke Thomsen et al., 2002).

The recruited patients received information in print about the project, before they were contacted by telephone. Out of 1828 recruited patients, 1446 took part in a screening telephone interview of whom 195 probands were diagnosed with HM (Lykke Thomsen et al., 2002). To stratify these patients into FHM or sporadic hemiplegic migraine (SHM) cases, we subsequently screened their relatives. All first-degree relatives were screened and second-degree, third-degree or more distant relatives were screened by two different strategies depending on whether an affected first-degree relative had been identified or not (Lykke Thomsen et al., 2002). All first-degree relatives above 15 years of age were contacted for a telephone interview. Relatives aged 15 years or below were only contacted for a telephone interview if suspected of having or previously having had headache or aura symptoms. If the screening interview diagnosed the proband with hemiplegic migraine, the proband and all available relatives were diagnosed according to the International Classification of Headache Disorders 1st Edition (ICHD-1) (Headache Classification Committee of the International Headache Society, 1988; Lykke Thomsen et al., 2002) in an extensive validated semi-structured telephone interview (Russell et al., 1995) performed by a trained physician. Out of 1486 (722 M; 764 F) recruited relatives, contact was sought in 903 living relatives of whom 859 took part in an interview (Lykke Thomsen et al., 2002).

In accordance with the ICHD-1 (Headache Classification Committee of the International Headache Society, 1988), HM patients with at least one affected first-degree relative were diagnosed as having FHM, whereas HM patients without any affected relatives were diagnosed as having SHM. All HM patients had a physical and neurological examination (Lykke Thomsen et al., 2002).

In total 291 patients were diagnosed with HM of whom 147 patients had FHM, 105 had SHM and 39 had unclassifiable hemiplegic migraine due to an unknown family history of hemiplegic migraine (Lykke Thomsen et al., 2002). The 147 patients with FHM were from 44 families and were available for this study. Overall, the number of FHM patients in each of the families was 11 (1 family), 7 (2 families), 6 (1 family), 5 (4 families), 4 (6 families), 3 (12 families), 2 (18 families). The project was approved by the Danish Ethics Committee.

Among the FHM families, one family (F6035) and two individuals from separate FHM families did not want to give blood for unspecified reasons. Twelve families consisted of an affected parent and child only (6008, 6011, 6013, 6014, 6022, 6024, 6028, 6031, 6036, 6043, 6047) or only with genotypes to that extent (family 6021), making them uninformative in linkage analysis. Thus, the genome-wide linkage scan was conducted on the 31 FHM families that were informative and available. In contrast, all 43 families that gave blood samples were screened by sequence analysis for exonic mutations in CACNA1A and ATP1A2, and by SNP assay for the known Q1489K mutation in the SCN1A gene.

After obtaining informed consent according to the Declaration of Helsinki, venous blood samples were collected from these families including all participating, unaffected relatives. Further details about material, participation/non-participation and the clinical characteristics of the families have been reported elsewhere (Thomsen et al., 2002).

Population controls were recruited by the Danish Headache Center, and are migraine-free as determined by standard headache diagnostic interview (Russell et al., 1995).

Genotyping and linkage analysis

The genotyping of microsatellite markers and linkage analysis were performed as described previously (Bjornsson et al., 2003). For the linkage analysis 141 FHM patients, of which 119 were in the 31 families informative for linkage and 567 unaffected relatives were genotyped using a genome-wide framework marker set of 1000 markers. The genetic position of markers used is according to deCODE's High-Resolution Genetic Map (Kong et al., 2002). The method of genotyping and genotype quality evaluation was as described previously (Fossdal et al., 2004). Linkage analysis was performed for both a non-parametric model (LOD-N) and autosomal dominant parametric model allowing for heterogeneity (HLOD-D) (allele frequency 0.0001 with autosomal penetrance: 0.9/0.9/0.01 and X chromosomal penetrance 0.9/0.9/0.01 for females and 0.9/0.01 for males), using an affected only analysis, with affected status for FHM patients and unknown status for relatives. LOD scores were generated with the Allegro 2 program (Gudbjartsson et al., 2005) and are based on multipoint calculations using all the genotyped markers. The Q1489K mutation in the SCN1A gene was typed with primers (SG02S408Fw: CCTTGAACCTGTTTATTGGTGTCATCATAG; SG02S408Rv: CTCA*TTTGGCA*GAGAAAACACT; probe 1: AT TTCAACCAGC; probe 2: TA ATT* TCAACCAGA; enhancer: GAAAAAGAAGATAAGTATTTCTAATAT; with * to indicate superbase), using the Centaurus SNP genotyping platform (Nanogen).

Sequence analysis

All exons and promoter regions of the CACNA1A and ATP1A2 genes were sequenced in at least two patients with FHM from each of the 42 FHM families. These patients were also genotyped with a SNP assay for the Q1489K mutation in the SCN1A gene. Because reported FHM mutations are rare and with high penetrance, mutations that were identified in two FHM patients of at least one family and not in healthy controls were considered as possible causative mutations (Ophoff et al., 1996). Co-segregation of the FHM phenotype in families with a possible mutation was established by sequencing the mutation-bearing exon in all the remaining patients and unaffected relatives in the relevant families. A population control group consisting of 92 unrelated persons without migraine, diagnosed in an extensive interview by a physician, was included in the initial sequence screen. For all novel mutations identified a total of 460 population controls were sequenced.

The 48 exons of the CACNA1A (19p13) and the 23 exons of the ATP1A2 (1q23) genes were sequenced, including promoter and flanking intron sequences (Eriksen et al., 2006). The primers were 18–30 bp long depending on the size of each exon and GC content of the region. The amplimers were selected from 200 to 750 bp, amplifying a minimum of 60–200 bp of the intronic sequence flanking the exons. PCR and cycle sequencing primers were designed by WinSeq1.6, a software based on the Primer3 software (Rozen and Skaletsky, 2000). PCR amplifications were set up on Sciclone ALH 300 and run on MJR Tetrad™. PCR products were verified for correct length on an agarose gel before being purified using AMPure™ from Agencourt. Cycle sequencing reactions were set up on Sciclone ALH 300, run on MJR Tetrad™ and excess dye terminators were removed using CleanSEQ™ from Agencourt. Amplimers were sequenced directly on an Applied Biosystems 3730 Capillary DNA Sequencer using an ABI PRISM® Fluorescent Dye Terminator System (Perkin–Elmer, Foster City, CA, USA). The sequence analysis was conducted with Clinical GenomeMiner™ 1.5, a deCODE software comparable to Consed (Gordon et al., 1998).

Results

The genomewide linkage analysis

The results of linkage analysis of the 31 Danish FHM families that are informative for linkage are shown in Fig. 1, for both dominant parametric models allowing for heterogeneity (HLOD-D) and a non-parametric model (LOD-N). A major linkage is observed only at the FHM1 locus, with a HLOD-D of 4.58 and LOD-N of 2.82 at marker D19S226 on chromosome 19. Interestingly, a HLOD-D reaching 0.60 is observed at the FHM2 locus at marker D1S2768 on chromosome 1. No linkage is seen at the FHM3 loci on chromosome 2. LOD scores between 1 and 2 are observed on chromosomes 5, 8 and 10. The best single-family LOD scores at the FHM1 locus are observed for family 6002 (LOD score of 1.4) and family 6034 (LOD score of 4.2), indicating that these families are the major contributors to the genomewide linkage to this locus. At the FHM2 locus, only family 6005 has a notable LOD score (2.1 at marker D1S2768). No single family has a LOD score exceeding 1 at the FHM3 locus. Regarding other genome locations, only family 6000 has a notable LOD score of 2.4 at marker D18S1152 on chromosome 18, which by itself is not significant.

Fig. 1

Genome-wide linkage of 31 Danish FHM families. Results are shown for both non-parametric linkage (LOD-N) (black line) and a dominant parametric model allowing for heterogeneity (HLOD-D) (red line). LOD scores are shown on the vertical axis and the cM position of genetic markers relative to the p end of the chromosome on the horizontal axis. The span of the vertical LOD score axis is 2 U for all chromosomes except chromosome 19 where it is 5.

Mutation screen

All exons in the CACNA1A and ATP1A2 genes were sequenced, and the Q1489K mutation in the SCN1A gene was screened by an SNP assay. A total of six exonic mutations were detected that are either probable or known FHM mutations, three mutations in the CACNA1A gene (R583Q, T666M and C1369Y) and three mutations in the ATP1A2 gene (V138A, R202Q and R763C). Also, two additional polymorphisms were discovered that did not correlate to the FHM phenotype: a R510S polymorphism in ATP1A2 and a R2233Q polymorphism in CACNA1A. None of the FHM families had the Q1489K mutation in the SCN1A gene. Pedigrees of the 6 FHM families with mutations are shown in Fig. 2. The location of the three mutations in the CACNA1A gene is indicated in Fig. 3A. The location of the three mutations in the ATP1A2 gene is shown in Fig. 3B. The clinical characteristics of all patients with FHM and mutations in either the CACNA1A or ATP1A2 genes are listed in Table 1.

Fig. 2

Pedigrees of FHM families with mutations in the CACNA1A or ATP1A2 genes. In the text persons are referred to by running numbers for each family (example; person 1 in family 6002 is referred to as P1F2). The clinical phenotype status is indicated by the symbols, with FHM denoting familial hemiplegic migraine, MA migraine with aura, MO migraine without aura, ETTH episodic tension type headache, MD migrainous disorder and CTTH chronic tension type headache. FHM mutation allele is indicated with m in each family and + for wild type allele. This is shown for all persons that were screened by sequence analysis, except for P16F2 and P25F2 (+/m*) who have the mutation according to haplotype analysis (see also Fig. 4).

Fig. 3

The location of FHM mutations in the (A) CACNA1A and (B) ATP1A2 gene.

Fig. 4

Haplotype analysis with available genotype data in Family 6002 with a T666M mutation in the CACNA1A gene. The affection status is as indicated by symbols and affected haplotype as a ribbon with the letter K. The allelic state and the name of corresponding markers are also shown. Persons with sequencing data support for the T666M mutation are indicated by arrows. Unaffected members of the family, P24F2, P12F2 and P25F2, also carry this haplotype. In the text persons are referred to by running numbers for each family (example; person 1 in family 6002 is referred to as P1F2)

View this table:
Table 1

Clinical characteristics of FHM-affected carriers of mutations in the CACNA1A and ATP1A2 genes

PersonFamilyAge/age at onset (years)Lifetime number of FHM attacksHemiplegia in arm/leg +/−Duration of hemiplegia (min)Duration of FHM attackHeadache accompanying hemiplegic attacks +/−Accompanying symptoms nausea/vomiting/photo/phonophobia +/−FHM triggered by minor head traumaCerebellar ataxiaCo-occurring attacks of MA/MO +/−
P52F34603424/710–49+/+604h-1d++/+/+/++p−/−
P47F34603452/4410–49+/+304h-1d++/+/+/++p+/+
P44F34603448/13>100+/+12004h-1d++/+/+/++p+/+
P56F34603425/5>100+/+1201-3d++/+/+/+++/+
P30F34603450/12>100+/+404h-1d++/+/+/++p−/+
P47F34603426/710–49+/−6001-3d++/−/−/++p−/−
P34F34603459/1550–100+/+204h-1d++/+/+/++p−/−
P49F34603434/610–49+/+1804h-1d++/−/+/++p−/−
P41F34603451/17>100+/−604h-1d++/+/+/++p+/−
P39F34603447/152–4+/+7201-3d++/+/−/++p−/−
P37F34603447/3510–49+/+2403-7d++/+/+/++p−/−
P4F2600277/16>100+/+3601-3d+−/+/+/++−/−
P13F2600247/710–49+/+301-3d++/+/+/−++−/−
P21F2600223/3>100+/+304h-1d++/+/+/+++−/−
P18F2600242/450–100+/+12004h-1d++/+/+/+++−/−
P16F2600238/1210–49+/+304h-1d+−/−/+/++−/+
P12F44604455/5>100+/+601-3 d++/+/+/+(+)+/+
P9F44604444/7>100+/+144030m-4h++/+/+/+−/−
P14F44604422/202–4+/−14401-3 d++/+/+/−+/+
P13F44604414/92–4+/+104h-1d+−/−/+/++/−
P4F7600738/7>100+/+451-3 d++/+/+/+++−/−
P8F7600710/82–4+/+12004h-1d++/+/−/−−/−
P7F7600712/710–49+/+601-3d++/+/+/+−/+
P10F5600562/4550–100+/+45?−/*−/−/−/−+−/−
P30F5600526/222–4+/−204h-1d++/+/−/++−/−
P32F5600527/2510–49+/+12004h-1d++/−/+/++/+
P29F5600552/355–9+/−120?−/*−/−/−/−++/−
P21F5600548/3050–100+/+16003-7da++/+/+/+++/+
P19F5600550/1050–100+/+12004h-1d++/+/+/++−/+
P24F5600553/1650–100+/−604h-1d++/−/+/+++/−
P20F16601613/102–4+/−6601-3d++/+/+/+−/+
P7F16601637/810–49+/−1804h-1d++/+/+/++/−
P12F16601647/?1+−/−
  • + = symptom present; − = symptom not present; ? = unable to recall; d = day; h = hours; m = minutes; +p = persistent cerebellar ataxia; + ataxia = paroxystic cerebellar ataxia.

  • *Hemiplegic attacks never associated with headache.

FHM mutations in the CACNA1A gene were found in 27 persons in 3 families. Only 18 of these persons have been diagnosed with FHM, which corresponds to a penetrance of 67%. For the ATP1A2 gene FHM mutations were found in 19 persons in 3 families. Only 12 of these persons have an FHM diagnosis which corresponds to a penetrance of 63%.

Mutation data and clinical characteristics

Family 6034 had a R583Q mutation [arginine (R, CGA) to glutamine (Q, CAA)] in the CACNA1A gene. The R583Q mutation is an FHM missense mutation (Battistini et al., 1999; Ducros et al., 2001; Terwindt et al., 2002; Alonso et al., 2003) reported previously. Eleven family members with FHM from family 6034 were included in the study (Fig. 2). Three patients with FHM had died and two patients did not want to participate for unspecified reasons. The sequencing data show that all patients with FHM have the mutation and three of the unaffected relatives (P22F34, P16F34, P53F34) (Fig. 2). The R583Q mutation was not identified in any of the 92 controls analysed.

Clinical characteristics

Among the 11 patients with FHM, 10 patients had slowly progressive cerebellar ataxia. Two relatives (P22F34 and P16F34) of 78 and 85 years of age, with the mutation, both had severe progressive cerebellar ataxia but denied ever having had attacks of migraine. The ataxia in this family was slowly progressive from 8 to 10 years of age and associated with cerebellar atrophy, present on MR scans in three individuals and CT scans in one patient (only 5 of these patients had previously been scanned). P53F34 was 24 years old and had exclusively had attacks of episodic tension type headache, never attacks of migraine.

Family 6002 had a T666M mutation [threonine (T, ACG) to methionine (M, ATG)] in the CACNA1A gene. The T666M missense mutation has been described before (Ophoff et al., 1996; Ducros et al., 1999, 2001; Friend et al., 1999; Takahashi et al., 2002; Terwindt et al., 2002; Wada et al., 2002; Kors et al., 2003; Jen et al., 2004). Five patients with FHM from family 6002 were included in the study. Sequence analysis shows that four family members with FHM and three family members without FHM (P18F2, P16F2, P20F2, P13F2, P12F2, P24F2, and P23F2) have the mutation. Sequencing was unsuccessful for one family member with FHM and one family member without FHM (P25F2 and P16F2), but haplotype analysis with available genotype data indicates that both should have the mutation since they share the haplotype that correlates to the mutation (Fig. 4). The T666M mutation was not identified in any of the 92 controls analysed.

Clinical characteristics

Among the five patients with FHM, one patient also had permanent ataxia. Four family members had the T666M mutation but did not express the FHM phenotype (P12F2, P23F2, P24F2 and P25F2). P12F2 was 56 years of age and did not fulfil the diagnostic criteria for FHM (ICHD-1), however, in childhood she had had a single attack where she became weak in both arms and legs, which was not followed by headache (described by her parents), furthermore she had attacks of migraine without aura (MO) but not MA. P23F2 was 25 years of age and did not suffer from headache disorders. P24F2 was 30 years of age and had had attacks of MO. P16F2 was 19 years of age but did not suffer from headache disorders.

Family 6044 had a C1369Y mutation [cysteine (C, TGT) to tyrosine (Y, TAT)] in the CACNA1A gene. The C1369Y mutation is a missense mutation that has not been described previously. Four patients with FHM were included in the study. Sequencing shows that 3 of 4 affected family members have the mutation. The C1369Y mutation is in a conserved amino acid residue. The C1369Y mutation was not identified in any of the 460 controls analysed.

Clinical characteristics

One person, 14 years of age (P13F44), expressed the FHM phenotype, had attacks of MA and tension type headache (TTH), but did not have the mutation. This is supported by the absence of a shared FHM1 locus haplotype for P13F44 with other family members with the C1369Y mutation, and is therefore a likely FHM phenocopy. One person 74 years of age (P5F44) had the mutation, had attacks of MA but did not express the FHM phenotype.

Family 6007 had a V138A mutation [valine (V, GTC) to alanine (A, GCC)] in the ATP1A2 gene. The V138A mutation is a missense mutation which has not been described previously. Three patients with FHM were included in the study. The sequencing data reveal that five family members are heterozygous for this mutation, including all three FHM patients. Valine to alanine is a minor change in amino acid properties; however, this residue is conserved in the protein. The V138A mutation was not identified in any of the 460 controls analysed.

Clinical characteristics

Two family members (P5F7 and P2F7) had the mutation but did not express the FHM phenotype and had never had attacks of any type of migraine.

Family 6005 had a R202Q mutation [arginine (R, CGG) to glutamine (Q, CAG)] in the ATP1A2 gene. The R202Q mutation is a novel missense mutation not described previously, and is seen in all seven FHM patients in this study. Four of the unaffected family members also have the mutation (P14F5, P34F5, P33F5 and P35F5). The R202Q mutation changed a conserved amino acid residue. The R202Q mutation was not identified in any of the 460 controls analysed.

Clinical characteristics

Patient P14F5 was 71 years of age and had exclusively had attacks of MA. Patients P34F5 and P33F5 are 22 and 33 years of age, respectively, and have only experienced attacks of MA at the time of our interview. Two patients with FHM had never had headache, but exclusively aura symptoms, in relation to their FHM attacks.

Family 6016 had a R763C mutation [arginine (R, CGC) to cysteine (C, TGC)] in the ATP1A2 gene. The R763C mutation is a novel missense mutation not described previously. Three patients with FHM were included in the study. The sequencing data indicate that four family members have the R763C mutation, including all but one FHM-affected. Patient P12F16 has FHM but does not have the mutation. The R763C mutation was not identified in any of the 460 controls analysed.

Clinical characteristics

Two relatives (P9F16 and P11F16) have the mutation. P9F16 suffered from attacks of MO and MA but did not fulfil the ICHD-1 criteria for FHM, because she had only experienced one attack of hemiplegic migraine. P11F16 had exclusive attacks of TTH and never attacks of migraine.

Linkage analysis excluding families with identified FHM mutations

Additionally we did a genomewide linkage analysis excluding the 6 FHM families with known or new mutations. The results are shown in Fig. 5. The linkage to the FHM1 locus on chromosome 19 decreases to linkage background level. No major linkage signal is observed. Linkage with LOD scores of 1 or greater is observed on chromosome 3p, 10p, 10q, 14q and Xp.

Fig. 5

Genomewide linkage of the FHM cohort with the 6 identified FHM mutation families excluded. Results are shown for both non-parametric linkage (LOD-N) (black line) and dominant parametric model allowing for heterogeneity (HLOD-D) (red line). LOD scores are shown on the vertical axis and the cM position of genetic markers relative to the p end of the chromosome on the horizontal axis.

Discussion

FHM is a rare disorder affecting 0.005% of the Danish population (Lykke Thomsen et al., 2002). A systematic nationwide search is the best way to collect a sample of families with FHM, which represents the whole spectrum of the disease in the population. This study is the first of a population-based sample of families with FHM. All patients were diagnosed according to generally accepted international classification criteria (ICHD-1). All patients fulfilled the diagnostic criteria and all patients had a physical and neurological examination which ruled out other possible causes of hemiplegic attacks. Overall, the clinical characteristics of our FHM families did not differ from FHM families described previously (Thomsen et al., 2002).

Genomewide linkage on the Danish FHM families revealed a linkage to the known FHM1 locus on chromosome 19, a minor linkage to the FHM2 locus on chromosome 1 and no linkage to the FHM3 locus on chromosome 2 (Fig. 1). The linkage data for individual families indicated that family 6002 and 6034 contribute to the FHM1 locus and family 6005 to the FHM2 locus. We have also screened the Danish FHM families by sequence analysis for FHM mutations in the CACNA1A (FHM1) and the ATP1A2 (FHM2) genes and by SNP assay for the recently identified Q1489K mutation in the SCN1 gene (FHM3). Three mutations were identified in each of CACNA1A and ATP1A2 genes but none in the SCN1 gene. Of the FHM families with identified mutations, three correspond to the families with notable single-family linkage at either the FHM1 or the FHM2 loci (families 6002, 6034 and 6005, Fig. 2). When families with FHM mutations in either the CACNA1A or the ATP1A2 genes are excluded the linkage on chromosome 19 disappears (Fig. 5), and no other suggestive locus is observed that would correspond to the remaining FHM family material. In general the FHM families described here are small and the FHM families that score with mutations at either the FHM1 or FHM2 locus tend to be larger than the remaining families (Table 2). The large number of small families in the remaining FHM family cohort makes it unlikely that novel FHM loci can be detected by linkage analysis.

View this table:
Table 2

Number of FHM cases and FHM families with or without mutation in the CACNA1A, ATP1A2 and SCN1A genes

FHM cases in each familyFHM familiesFHM families with mutation in CACNA1AFHM families with new mutation in ATP1A2FHM families with mutation in SCN1AMutationFHM families without mutations in the CACNA1A, ATP1A2 and SCN1A genes
217 (18)17
3122V138A, R763C10
461C1369Y5
541T666M3
611
721R202Q1
1111R583Q0

Methodological considerations regarding the new mutations

Families with two previously described CACNA1A mutations were identified, the R583Q mutation as found in FHM family 6034 and the T666M mutation in FHM family 6002.

The R583Q mutation co-segregated with the disease phenotype; however, the mutation was also identified in three relatives not expressing the FHM phenotype. Two of these relatives were older women with severe cerebellar ataxia but without migraine. One relative was 24 years of age and did not express either the FHM phenotype or cerebellar ataxia. This FHM family supports the clinical spectrum described previously in families with the R583Q mutation in the CACNA1A gene, where some family members express HM, others cerebellar ataxia and others both HM and cerebellar ataxia (Alonso et al., 2003).

The T666M mutation was identified by sequencing or haplotype analysis in all five FHM-affected family members and additionally in four unaffected family members. The family member with only one FHM attack (P12F2) does not fulfil the criteria for FHM; however, she could have FHM with reduced penetrance. Supporting this, both her children (age 25 and 30) had the mutation without expressing the FHM phenotype. The fourth family member with the mutation but without FHM was only 19 years of age, which makes it impossible to say if this is due to incomplete penetrance or young age.

We identified a novel mutation, C1369Y, in the CACNA1A gene encoding the Cav2.1 calcium channel pore-forming subunit. The subunit of the calcium channel contains four repeated domains, each encompassing six transmembrane segments. The mutation co-segregates with the disease in three of four patients with FHM. Person P13F44 most likely represents a phenocopy. Person P5F44 may be an instance of reduced penetrance. The mutation was absent in the 460 controls. Even though the C1369Y mutation is in a conserved residue, the absence of this mutation in one affected family member makes us only claim that this is a possible FHM mutation.

Furthermore we identified three novel missense mutations, V138A, R202Q, R763C, in the ATP1A2 gene encoding the Na+, K+-ATPase 2 subunit in three Danish families with FHM. The Na+, K+-ATPase 2 subunit consists of a N-terminal part containing four membrane spanning domains (M1-4) followed by a large intracellular loop and a C-terminal region with six membrane spanning domains (M5-10). All three mutations are located in a well-conserved region.

The V138A mutation in Family 6007 was identified in five individuals. The mutation followed the FHM phenotype in all three FHM-affected persons; however, it was also identified in two unaffected family members. Two other unaffected relatives did not have the mutation. The mutation was absent from 460 controls and changes a conserved residue. The V138A mutation therefore very likely is an FHM mutation with variable penetrance. However, no FHM mutations have previously been described in this region of the ATPase protein.

The R202Q mutation in Family 6005 followed the FHM phenotype in all seven FHM-affected family members. A total of 32 unaffected relatives were screened of whom four had the mutation. Because of a variable age of onset of FHM, it is still too early to tell whether family members P34F5, P33F5 and P14F5 represent truly non-penetrant individuals or may show the FHM phenotype later in life. The R202Q mutation was absent from the 460 control individuals. The R202Q mutation is therefore very clearly an FHM mutation.

The R763C mutation in Family 6016 followed the FHM phenotype in 4 FHM-affected individuals except in one individual (P12F16) who did not have the mutation but had FHM from other causes that are not linked to this mutation. A total of 17 unaffected family members were screened for the mutation and two of these individuals (P11F16 and P9F16) had the mutation. A child of P20F16 had the mutation, supporting that P20F16 represents reduced penetrance. The fact that P9F16 had had one attack of FHM previously supports that P9F16 represents reduced penetrance. The R763C mutation was not identified in the 460 controls. However, even though the R763C mutation changes a highly conserved amino acid residue it is possible that this is not truly the causative FHM mutation in this family, since one FHM family member did not have the mutation. Alternatively this family member could represent a phenocopy.

Our results compared with previous studies

In this study we identified two previously described CACNA1A mutations (T666M, R583Q), one new CACNA1A mutation and three new ATP1A2 mutations causing FHM. Previous functional studies show a dysfunction in the ion transport in the involved ion channels, a dysfunction that decreases the threshold for cortical spreading depression—the likely underlying mechanism for migraine aura (Leao, 1986; Goadsby, 2004). Functional studies of the new mutations in the CACNA1A and ATP1A2 genes need to be done in addition to further confirmation of these mutations in other FHM families.

The CACNA1A and ATP1A2 mutations in FHM have previously been reported in large highly selected families showing an autosomal dominant inheritance. Interestingly, in this study, CACNA1A and ATP1A2 mutations were identified in the large multi-generational families (6–15 FHM patients per family) in our population-based sample of FHM families whereas mutations in the known FHM genes were not found in the somewhat smaller FHM families (2–8 FHM patients per family) (Thomsen et al., 2002). Whether the symptoms are weaker in smaller families, the penetrance lower or inheritance more complex is not clear.

The FHM penetrance in the Danish families reported here is 67% for CACNA1A gene mutations and 63% for the ATP1A2 gene mutations. This is somewhat lower than the 80–90% penetrance published for the CACNA1A gene (Ducros et al., 1995) or the 81% penetrance for the ATP1A2 gene (Riant et al., 2005).

Surprisingly, this study showed that only 14% of FHM families had mutations in the CACNA1A or ATP1A2 gene. One explanation could be that a population-based sample of families with FHM from Denmark differs genetically from a population-based sample of families with FHM from other populations. Another explanation could be that previous studies are biased towards large families with many affected persons and clear Mendelian segregation of FHM. Thus, it is likely that there exist other FHM genes with causative variants with lower penetrance than the variants in the already known FHM genes.

In our survey of the Danish population we have also identified 105 SHM cases (Lykke Thomsen et al., 2002). 103 SHM patients have been screened for mutations in the known FHM genes. None of the identified mutations corresponds to the FHM mutations described here.

After removing families with known mutations our genome-wide scan shows linkage exceeding 1 at several locations and the linkage peaks on chromosome 10p and 14q overlap linkages published previously (Soragna et al., 2003; Lea et al., 2005) and may therefore harbour unidentified FHM genes.

In conclusion, FHM in the general population is most often caused by mutations in other genes than the CACNA1A, ATP1A2 and SCN1A genes. This study also raises questions regarding whether all FHM is actually dominantly inherited. The existence of many small families in a population-based sample of families with FHM indicates either low penetrance of a dominant gene or complex inheritance while recessive inheritance seems unlikely from the pedigrees.

Acknowledgments

We thank the patients with FHM and their unaffected relatives who agreed to participate. We also thank our colleagues for their excellent collaboration. The work was supported by deCODE genetics and grants from the Cool sorption Foundation of 1988, the Foundation for Research in Neurology, the Danish Headache Society, The A.P. Møller Foundation for advancement of medical science, the Novo Nordisk Foundation, the IMK-Almene Foundation, The Glaxo Wellcome Research Prize and Ms Else Torp Foundation.

References

View Abstract