Abstract
It is very likely that genetic factors play a role in the pathophysiology of cluster headache (CH). As CH shares its paroxysmal character with migraine, and migraine has been described in coexistence with CH in some families, we hypothesized that both diseases might share a genetic aetiology. In this study, we tested whether the migraine CACNA1A gene on chromosome 19 is involved in CH in an extended pedigree. Haplotype analysis did not reveal an obvious disease haplotype, and SSCP analysis of all 47 exons of the CACNA1A gene did not reveal a causative mutation. CH in this family is not caused by mutations in the CACNA1A gene.
Introduction
The exact pathogenesis of cluster headache (CH) is unknown. Previous hypotheses suggested that CH attacks are caused by abnormal vasodilatation of the internal carotid artery in the region of the cavernous sinus, but recent studies have shown that also central brain structures are involved (1). A genetic predisposition underlying CH is likely for several reasons: family studies have shown an increased risk for CH in first- and second-degree family members of probands with CH (2), and CH has been described in multiple members of some families as well as in monozygotic twin-pairs. The only molecular genetic analysis of CH published so far, has revealed a point mutation in platelet mitochondrial tRNALeu(UUR) only in one CH patient (3). However, subsequent studies could not confirm this finding in other patients with CH (4, 5). When looking at the reported pedigrees with multiple CH patients, it is clear that an autosomal dominant mode of inheritance (probably with a low penetrance) is more likely than maternal mitochondrial inheritance, as in several families transmission from father to offspring has been reported. This was confirmed in a complex segregation analysis study that showed autosomal dominant inheritance in CH (6).
Investigation of the genetic cause of CH is difficult for several reasons. First, virtually all CH pedigrees reported are too small for conventional parametric linkage studies. Secondly, for allele sharing studies such as non-parametric sib-pair analysis, a large number of families with multiple affected sibs is needed, which are not available in CH. Furthermore, it is difficult to indicate a priori candidate genes in CH as this headache syndrome is seldom described in association with other (hereditary) disorders, and virtually no clues exist as to an underlying pathophysiology.
The only disease that has been described in familial association with CH thus far is migraine (7–9). Importantly, CH and migraine share clinical features: both diseases have a paroxysmal character and vascular as well as central mechanisms are believed to be involved in the pathophysiology of both.
Two genetic loci have been implicated in migraine so far, both in familial hemiplegic migraine (FHM): the CACNA1A gene on chromosome 19 (10) and an as yet unknown gene on chromosome 1 (11). In this study we regarded the CACNA1A gene as a candidate gene for CH and performed a haplotype and mutation analysis of the CACNA1A gene in a family with multiple members suffering from CH.
Patients and methods
The male proband (III-1 in the pedigree) had been suffering from episodic attacks of CH since 1988, starting at the age of 48. The duration of the right-sided attacks is between 25 min and 2 h, always with tearing and redness of the ipsilateral eye. The patient still has yearly attack-episodes lasting 3–5 weeks. The attacks occur mostly at night. Interictal neurological examination revealed no abnormalities, except for a right-sided Horner syndrome. Neuroimaging was unremarkable. During CH episodes, headache attacks are successfully treated with verapamil (maximal dose 3 × 120 mg) and oxygen inhalation.
One of the daughters of the patient (IV-7) suffered from primary chronic right-sided CH, starting at the age of 16, responding to prednisone and verapamil. One maternal cousin of the patient (III-6) suffers from episodic CH. None of the other family members suffered from cluster headache or migraine, although the youngest (IV-5) son of the patient had episodic tension type headache.
Genomic DNA samples
Blood samples were collected and genomic DNA was isolated from leucocytes by a salting out procedure, as described by Miller et al. (12).
Haplotype analysis
Microsatellite markers D19S221, D19S1150, SCA-6, D19S226, all from the chromosome 19p13 region, were tested by PCR. Markers D19S1150 and SCA-6 are intragenic markers of the CACNA1A gene. Oligonucleotide sequences are available through the Human Genome Data Base (GDB) except for SCA-6 (forward primer 5′-GCCCCCTCAACATCTGGTA-3′ and reverse primer 5′-GACCCGCCTCTCCATCCT-3′). PCRs for all markers were performed using the same protocol. The reaction was performed in a 10-μl reaction volume, containing 10 pmol of each primer, 1 × PCR Buffer II (Applied Biosystems, Fostercity, California, USA), 2.5 m
SSCP analysis
In the proband (III-1) PCR products of all 47 exons of the CACNA1A gene were screened for sequence aberrations by SSCP analysis (13). The exons were amplified using exon-specific primer sets and PCR conditions previously described by Ophoff et al. (10).
Essentially, for SSCP analysis, amplification was performed in two consecutive rounds, in such a way that PCR products were radioactively labelled by the incorporation of [α-32P]dCTP in the second round of amplification. Subsequently, samples were denatured in formamide buffer and subjected to SSCP analysis on an 8% polyacrylamide (19:1) gel containing 10% of glycerol. Electrophoresis was carried out at room temperature at a constant power of 28 W. Finally, SSCP gels were dried and exposed for autoradiography. Gels were visually inspected for the presence of aberrant banding patterns.
Results
Haplotype analysis showed that the three affected family members did not share CACNA1A gene markers (Fig. 1). The portion of the allele that is containing the CACNA1A gene, and that is transmitted from the proband (III-1) to his affected daughter (IV-7), is also shared by other children (IV-4 and IV-5) and the brother of the proband (III-3). Clinical diagnoses gave no indications that these individuals are suffering from CH. The other allele of the proband was not transmitted to his offspring. The cousin of the proband with CH (III-6) shared only one CAG allele with the proband, but from missing information on the oldest generations from the pedigree it is hard to see the relevance. In the proband (III-1), mutation analysis of all 47 exons by SSCP analysis did not reveal any mutations in the CACNA1A gene.
Family pedigree. □ ○ unaffected; ▪ • cluster headache;
Discussion
Based on a previously reported familial association of CH with migraine, we considered the FHM CACNA1A gene a good candidate gene for CH, and performed genetic analysis of this gene in a family with three affected members with CH. This gave us the unique opportunity to study haplotype segregation of this candidate gene. Four markers in the direct area of the CACNA1A gene were tested, but no obvious disease haplotype could be identified. Although CH has a low frequency, which argues against the chance of both male affected carriers having inherited different disease genes, we could not exclude this possibility. Therefore, we decided to do a mutation analysis of all 47 exons of the CACNA1A gene and look for possible causative mutations in the proband. SSCP analysis did not reveal any CACNA1A mutations.
These negative SSCP results in the proband can be explained in several ways. First, as SSCP is not 100% sensitive, there is always the possibility that we missed a disease-causing mutation. Still, SSCP is a reasonably sensitive (75–98%) method for mutation screening as long as the fragments are not longer than 200 bases (14). Secondly, it is possible that the genetic predisposition of CH is not associated with the CACNA1A gene. In that case, other ion-channel genes still remain candidate genes, as mutations in ion channel genes are associated with paroxysmal phenomena, such as, for example, migraine, epilepsy and episodic ataxia (15). Thirdly, it is still possible that the CACNA1A gene plays a role in the aetiology of CH, but that – as in migraine – genetic heterogeneity exists, and that the family studied here is one of the families without a ‘CACNA1A CH mutation’, whereas in other families such mutations might be present.
Footnotes
Acknowledgements
JA van Vliet is supported by the Asclepiade Foundation, EE Kors is supported by the Netherlands Science Organization (NWO).