Dutch Migraine Genetic Research Group (DMGRG)

Introduction of our research:

Migraine research has intensified and gained direction with the discovery of the first familial hemiplegic migraine (FHM) gene, CACNA1A, by our research group a few years ago (Ophoff et al., 1996). The finding that the pore-forming subunit of a P/Q-type calcium channel was mutated in FHM families, gave new insights in the pathophysiology of this disease. It is now a challenge to test the functional consequences of the mutations and to unravel the exact pathways that are affected. Genetic evidence already indicates that the same gene might play a role in common forms of migraine (see below). It is clear that a better understanding of the dysfunction of the FHM gene, will be important for the development of new anti-migraine drugs. In this section, you will find information on the disease migraine, the search for FHM genes and on the characterisation of FHM mutations.


Clinical diagnosis: Migraine with and without aura

Migraine is a common neurological paroxysmal disorder affecting up to 16% of the general population and is more frequent in women than in men. The disease is characterised by recurrent attacks of disabling, mostly unilateral headache, associated with other symptoms such as nausea, vomiting, photo- and phonophobia, and malaise (migraine without aura; MO). In about one-third of patients, the attacks are preceeded or accompanied by transient focal neurological aura symptoms (migraine with aura; MA) that usually do not last longer than 60 minutes. The headache phase, that can vary from hours to days in MA and MO, shows similar features, but may be less severe or of shorter duration in MA patients (according to Headache Classification Committee (1988)). According to IHS criteria, a migraine patient, has had at least 2 attacks of MA or 5 attacks of MO. Family and twin studies have provided evidence that genetic as well as environmental factors are involved in the common forms of migraine. Importantly, not the migraine attack itself, but the repeated recurrence of attacks is abnormal. Apparently, in migraine patients “the threshold” to get attacks is lowered or triggers (such as stress, exertion, lack of sleep) are particularly strong or frequent in patients.

Familial Hemiplegic Migraine

FHM is a rare, autosomal dominant inherited, subtype of migraine with aura characterised by transient hemiparesis or hemiplegia (one-sided weakness or paralysis of the body) during the attacks (according to Headache Classification Committee (1988)). This ictal hemiparesis may last from minutes to hours or even weeks. In addition, other aura symptoms may be present that are also observed in patients with MA. Occasionally, FHM attacks are accompanied by confusion or psychosis, alterations of consciousness, fever or an aseptic meningeal reaction. In about 20% of the  families,  FHM symptoms include permanent cerebellar ataxia. Importantly, the symptoms of the headache and aura phase of FHM and “normal” migraine attacks are very similar, and both types may alternate within individuals and co-occur within families. These observations strongly suggest that FHM is part of the migraine spectrum. Thus, FHM can be used as a model to study the complex genetics and pathophysiology of the common types of migraine.


Pathophysiology of migraine:

The pathophysiology of migraine is still not well understood (Ferrari, 1998). Abnormal activation of the trigeminovascular system seems to be important. This  gives rise to abnormal transmission of nociceptive information to higher CNS centers and results in excessive release of vasoactive peptides at nerve endings that surround pial vessels. Consequently, these vessels are dilated which causes the throbbing pulsating headache.  Aura symptoms are believed to be caused by a depolarising wave, known as cortical spreading depression. When this wave propagates across the brain cortex it causes neuronal silencing, reduced ion homeostasis, and massive efflux of excitatory amino acids.


The search for FHM genes

Our genetic research on migraine started with the search for a FHM gene using a genome-wide linkage analysis in two of our largest FHM families. The publication by Joutel et al. (1993) that showed linkage of FHM to chromosome 19p13, was confirmed in some of our Dutch families. Through positional cloning, the gene encoding the pore-forming a1A (Cav2.1) subunit of the brain-specific P/Q-type calcium channel (CACNA1A) was identified and mutations detected in five FHM families (Ophoff et al., 1996). Mutations were identified in the same channel subunit in patients with episodic ataxia type 2 (EA-2) (truncating mutations) and spinocerebellar type 6 (SCA-6) (moderate expansions of a carboxyl terminal, polyglutamine (CAG)-repeat). However, recent reports, have shown that also missense mutations or moderate CAG-expansions can occur in EA-2 patients. Only in fifty percent of the FHM families reported, the CACNA1A gene located on chromosome 19p13 caused the disease. A second FHM locus has been mapped to chromosome 1q. Additional loci are expected, since a number of FHM families are linked neither to chromosome 19p nor 1q. Clearly, these observations display the genetic heterogeneity of FHM. It will be important to learn how functional defects in different genes converge into a common pathophysiological mechanism responsible for the neuronal instability observed in migraine patients.


FHM mutations in the CACNA1A gene

At the moment at least 13 FHM mutations in the CACNA1A gene have been reported.

figure 1

All of these mutations are single base-pair missense mutations, that lead to a substitution of a single amino acid residue in the  a1A subunit. The mutations involve highly conserved amino acid residues in various functional domains of the protein. For instance, mutations R192Q, R583Q and R1667W all involve Arginine residues in S4 segments of the voltage sensors. So, it is expected that these mutations might affect gating properties of the channel. Other mutations, like T666M and V1457L, are located in P-loops close to key glutamate residues that form the binding site for divalent cations, and have a role in ion-selectivity and permeability of the channels. Mutations V714A, I1811L and D715E are located in or near S6 segments that contribute to the lining of the part of the pore internal to the selectivity filter and control the channel’s inactivation properties (Hockerman et al., 1997; Hering et al., 1997). Therefore, these mutations may well interfere with the inactivation function. Similarly, mutation K1335E is located in the S3-S4 linker of repeat 3, a region which controls the time course and voltage dependence of channel activation subunit of N-type channels.  It is at present unclear, how mutations in S5 segments (e.g.. Y1384C and V1695I) or at the cytoplasmic face nearby (L1682P and W1683R) might affect calcium channel functioning. The observation that mutations in the S4-S5 linker of potassium channel Shaker B alter the stability of the inactivated state and channel conductance, might hint at a similar function of this region in P/Q-type calcium channels.

The fact that only missense mutations are associated with FHM suggests a molecular mechanism common to other channelopathies. Although no definite proof has been obtained yet, both alleles are likely to be expressed with the mutant allele resulting in gain-of-function variants of the a1A calcium channel pore-forming subunit. Similar findings have been described for mutations in the a subunit of the skeletal muscle sodium channel associated with hyperkalemic periodic paralysis.


Clinical variation with FHM mutations

A clinical comparison of FHM families linked and unlinked to chromosome 19p did not show significant differences for e.g. age of onset or frequency and duration of attacks (Terwindt et al., 1996). However, unconsciousness during attacks and provocation of attacks by mild head trauma was reported more in the 19p-linked families. However, in about half of the FHM families linked to chromosome 19, chronic cerebellar ataxia is part of the clinical phenotype. This is interesting, because cerebellar ataxia is also present in patients with episodic ataxia type 2 (EA2). This observation together with recent studies suggest a considerable overlap between the clinical phenotype of  FHM and EA-2). With the identification of an increasing number of mutations in CACNA1A, genotype-phenotype correlation studies have become feasible. Terwindt et al. (1998) compared the phenotypical consequences of the I1811L and the V714A mutation. Interestingly, cerebellar ataxia was observed only with the I1811L mutation. No other significant differences could be identified, except that loss of consciousness during attacks was reported more often in patients with the V714A mutation. Also patients with the T666M mutation have been reported with both increased loss of consciousness and cerebellar ataxia. In the pedigree published by Elliott et al. (1996) all patients had abnormal eye movements which is consistent with vestibulocerebellar dysfunction and probably an early manifestation of the cerebellar atrophy.

In addition, permanent cerebellar ataxia was also reported in FHM patients with mutations D715E, R583Q, R1668W and W1683R. In contrast, ataxia was never observed with mutations R192Q, V714A, K133E, V1457L and V1695I. Although the number of patients with the mutations mentioned above is very small it seems that the position of the mutation and/or the specific amino acid change controls the development of ataxia in a yet unknown manner. The mutations causing ataxia are neither clustered nor located in homologous domains of  a1A (see figure 1). Comparison of the functional consequences of mutations with the clinical phenotype might explain for instance the ataxic phenotype observed with some mutations.


Involvement of chromosome 19p13 locus in common forms of migraine

Knowing the CACNA1A gene is implicated in familial hemiplegic migraine, the next step is to study whether the same gene is also involved in the common types of migraine. A direct way to investigate such an involvement is by direct sequencing of the whole gene, including regulatory regions. However, the gene is rather large with almost 50 exons distributed over 350 kb of genomic sequence (Ophoff et al., 1996) which makes direct sequencing very labor intensive. In addition, locus heterogeneity of the common types of migraine with and without aura is likely to exist, meaning that only a fraction of the patients may carry a variant in this gene associated with migraine susceptibility. A study by Hovatta et al. (1994), showed no linkage to 19p in 4 families with migraine with aura and positive family history for migraine, probably because of the stringent criteria used to find linkage (one gene assumption, large contribution of the locus). In contrast, Nyholt et al. (1998) did find positive linkage in an Australian family with migraine with aura. This genetic heterogeneity also impedes a case-control study for migraine, using the intragenic polymorphisms of the CACNA1A gene: extremely large sample sizes will be required. A useful but indirect approach for gene mapping is the affected sib-pair analysis. In the affected sib-pair  analysis method, the observed sharing of alleles among affected siblings is compared with the expected sharing for a random locus. Since a pair of affected siblings has the same disease, these patients are expected to share a chromosomal region around the gene involved in the etiology of the trait. Tentative evidence was found in a first affected sib-pair analysis migraine study showing that the familial hemiplegic migraine containing region on chromosome 19p13 is involved in the common types of migraine (May et al., 1995). However, the results were inconclusive as to the magnitude of the involvement and the relative importance of migraine with aura and migraine without aura. A second affected sib-pair study was performed in an independent sample of 36 extended Dutch families with the common types of migraine (Terwindt et al., 2001). Significant increased sharing was found for migraine with aura. No such sharing was observed for migraine without aura. These two studies provide independent evidence of the involvement of the chromosome 19p13 region containing the P/Q-type calcium channel gene in the etiology of migraine, especially migraine with aura. The exact nature of this involvement, however, remains to be elucidated. Ultimately, the involvement of the CACNA1A gene has to be demonstrated by mutational or functional studies.