Abstract
Objective: To search for mutations in the calcium channel gene CACNA1A and to study the genotype–phenotype correlation in a family with a severe familial hemiplegic migraine (FHM) phenotype and a slowly progressive cerebellar ataxia.
Background: CACNA1A gene mutations on chromosome 19 are involved in approximately 50% of FHM families. The association of FHM and cerebellar ataxia has been reported in a small number of FHM families, all linked to chromosome 19.
Methods: The proband, in addition to typical hemiplegic migraine attacks, experienced severe episodes during which hemiplegia was associated with acutely altered consciousness and fever lasting several days. She, as well as her affected sister, developed a permanent, late-onset cerebellar ataxia and cerebellar atrophy evident on MRI. Linkage analysis was performed and the whole CACNA1A gene, 47 exon–intron boundaries, was analyzed by double gradient–denaturing gradient gel electrophoresis (DG-DGGE).
Results: Genetic studies suggested linkage to chromosome 19p13, and DG-DGGE analysis detected a heteroduplex fragment in exon 13 of the CACNA1A gene. By direct sequencing, a G-to-A substitution resulting in an arginine to glutamine change at codon 583 in the second putative voltage sensor domain of the channel α1A-subunit, was identified, possibly representing the disease-causing mutation. The proband and her affected sister were treated with acetazolamide, reporting freedom from new FHM attacks but no benefit in the progression of ataxia.
Conclusions: The combination of episodic dysfunction and permanent deficit could depend on the variety of functions of calcium channels and their distribution in the nervous system.
Familial hemiplegic migraine (FHM), according to the International Headache Society (IHS) classification, is a subtype of migraine with aura showing autosomal dominant inheritance.1 Onset usually occurs during childhood or adolescence, although later onset has been reported.2 Attacks are characterized by some degree of hemiparesis in addition to other migraine and aura symptoms.3 FHM attacks are occasionally associated with fever, drowsiness, confusion, or coma, which can be prolonged for days or weeks.4,5
Genetic heterogeneity of FHM has been clearly established. Approximately 50% of the reported FHM families showed genetic linkage to chromosome 19p13.6,7 Recently, a second locus for FHM has been assigned to chromosome 1q in four pedigrees.8,9 With respect to clinical features, patients from FHM families linked to chromosome 19 more often reported unconsciousness during attacks than patients from families not linked to this chromosome.10 In addition, a small number of FHM families linked to chromosome 19, but none of the unlinked families, had progressive cerebellar ataxia and cerebellar atrophy on MRI.6,7,10-14
The brain-specific P/Q-type calcium channel α1A-subunit gene CACNA1A on chromosome 19p13 has been implicated recently in a wide range of disease phenotypes, some episodic, with no or minimal interictal sequelae; others, chronic and progressive.15 Four different missense mutations in this gene have been identified in five unrelated FHM families, whereas two mutations disrupting the reading frame have been shown to cause familial episodic ataxia type 2 (EA-2).16 EA-2 is another autosomal dominant paroxysmal disorder characterized by attacks of cerebellar ataxia, migrainous symptoms, and interictal nystagmus, usually lasting a few hours, which respond to acetazolamide treatment.17 In addition, progressive cerebellar dysfunction17-19 and cerebellar atrophy on MRI20 have been described in some patients. A CAG repeat expansion in the CACNA1A gene has been found in patients with spinocerebellar ataxia 6 (SCA-6), a late-onset progressive cerebellar ataxia of the limbs and gait with dysarthria, nystagmus, and mild vibratory and proprioceptive sensory loss.21 Based on these different classes of mutations identified in FHM, EA-2, and SCA6, a genotype–phenotype correlation has been suggested.16,21 However such a correlation is not as clear as it initially seemed, following the identification of a CAG repeat expansion in an EA-2 family22 and, more recently, the identification of a missense mutation in a family with both severe progressive and episodic ataxia syndrome.23 These findings indicate that different mutations within the CACNA1A gene are associated with marked phenotypic heterogeneity.
In the current study we describe a previously unreported Italian family with FHM associated with altered consciousness, fever, and slowly progressive cerebellar ataxia due to a novel missense mutation in the CACNA1A gene. Moreover, in our patients acetazolamide was an effective prophylactic therapy for their attacks, suggesting acetazolamide responsiveness in FHM.
Methods.
Patients.
All available family members of an Italian family with FHM (figure 1) were evaluated clinically and interviewed. Information on unavailable individuals was obtained from other family members and by reviewing medical records. The diagnosis of FHM, migraine with and without aura, was made according to the criteria of the IHS.1
Figure 1. Pedigree of family under study and segregation analysis of the missense mutation Arg583Glu by BanII digestion as described in Methods. The 245-base pair band results from the mutant allele. It was present in both affected family members (Individuals I-4 and I-6) and also in an asymptomatic subject (Individual II-4). Black symbols indicate affected individuals. Gray symbols indicate individuals with migraine without aura. Empty symbols represent unaffected individuals. C indicates normal control subject. The generations are indicated by Roman numerals.
Patient I-6.
The proband, a 71-year-old woman, experienced recurrent attacks of hemiparesis and paresthesia beginning at the age of 17. These attacks usually started in her right hand, spread to the arm and tongue, and were associated with a migrainous headache that lasted a few hours. In addition to typical attacks ranging in frequency of two to four per year, at age 59 she experienced her first severe episode of migraine associated with confusion and fever lasting approximately 24 hours and necessitating admission to the hospital. Her interictal neurologic examination and EEG were normal. During the subsequent years her attacks became more frequent, approximately once or twice a month, and severe. She was admitted to the hospital three times for episodes of acutely altered consciousness, varying from confusion to stupor, and fever during attacks, lasting 2 to 3 days. Beginning in her 60s, apart from the migraine attacks, she noticed an increasing unsteadiness of gait. At age 66 her neurologic examination revealed a mild cerebellar ataxia.
She first presented to us in 1995 at age 68, when she experienced a sudden, mild right hemiparesis and paresthesia associated with aphasia, blurred vision, headache, severe confusion, and fever (38 °C). The attack resolved after approximately 3 days. An EEG performed on admission was characterized by persistent high-amplitude theta and delta waves with isolated sharp waves in the left frontotemporal region and moderate slowing of the background rhythms on the left hemisphere. After having been hospitalized at age 69 for another episode of right hemiparesis and paresthesia associated with headache, stupor, and fever, she returned for reevaluation in 1997 at age 70 because her gait had deteriorated. Her interictal neurologic examination showed dysmetria in the upper extremities with intention tremor, and a definite cerebellar ataxia. The remainder of her neurologic examination was normal. She was started on acetazolamide 250 mg twice a day. During a follow-up examination she reported freedom from new attacks over a 17-month period while taking acetazolamide, whereas in the 17 months preceding the treatment she had experienced a total of 26 typical and 2 severe attacks. She reported no benefits to her gait, which had worsened, and she had experienced several falls. Her neurologic examination showed dysmetria of the upper and lower extremities, mild dysarthria, and a marked cerebellar ataxia.
Patient I-4.
The proband’s 68-year-old sister recalled at age 40 a first episode of sudden right hemiparesthesia, including the hemiface, triggered by emotional stress, lasting minutes, associated with headache. She had suffered one to two attacks per year since that time. At age 53 she was admitted to the hospital for a short-lasting episode of hemiparesis and paresthesia in her right hand and arm, dysarthria, and blurred vision associated with migrainous headache. On suspicion of TIA, Doppler ultrasonography of the carotid and vertebral–basilar arteries and a cerebral angiography were performed, with normal results. Her interictal neurologic examination was also normal. During the subsequent years she reported that the number of attacks increased to approximately one a month, with hemiparesis occurring more frequently on the right side of the body. Beginning at age 57 she noticed mild imbalance. Brain CT performed at that time revealed the presence of vermian and hemispheric cerebellar atrophy. When she first presented to us in 1997 at age 67, her neurologic examination showed slight but definite cerebellar ataxia and an unsteady gait. She was started on acetazolamide 250 mg twice a day and reported freedom from new attacks over a 17-month period of treatment compared with the 16 attacks that she had experienced during the 17 months preceding the treatment. Her neurologic examination showed uncoordination of the upper extremities and a mild cerebellar ataxia.
Other family members.
The proband reported that her mother experienced similar attacks of hemiplegic migraine beginning in her 20s and developed progressive ataxia in her later years. She died at age 49 of unknown causes. The maternal grandmother apparently had gait difficulties and migraine. She died at age 89 of unknown causes. Further details could not be obtained. Individuals I-2 (61 years old) and II-5 (44 years old) had migraine without aura. Both reported that the attacks became less frequent and less severe as they aged. Individuals I-1, I-3, I-5, II-1, II-2, II-3, and II-4 were apparently healthy, and neurologic examinations were normal.
Magnetic resonance imaging.
Brain MRI was carried out on Individuals I-2, I-4, and I-6 using a Philips (Best, The Netherlands) Gyroscan imaging system at 0.5 T. Sagittal and axial spin-echo T1-weighted images, and axial and coronal turbo spin-echo T2-weighted images were obtained. All MR images were evaluated independently by two experienced neuroradiologists who did not know the individuals’ clinical histories.
Genotyping.
Three extragenic polymorphic microsatellite markers from the Genethon linkage map (D19S394, D19S221, and D19S226) flanking the CACNA1A gene were analyzed. In addition, four intragenic polymorphisms were analyzed: the D19S1150 microsatellite repeat located in intron 7,16 the 1474-31 A–G in intron 8, the 3253 A–T in exon 19 polymorphisms (authors’ unpublished data), and the CAG repeat 3′ to the gene in exon 47.16 The analyzed region spans a 8-cM interval between D19S394 and D19S226.
Genomic DNA was extracted from peripheral blood lymphocytes according to standard techniques.24 Microsatellites16,25 and the CAG repeat16 genomic sequences were amplified by PCR, labeled fluorescently, and analyzed using 6% polyacrylamide gel electrophoresis (PAGE) on a 373-GeneScan Automatic System (Perkin Elmer—ABI, Foster City, CA). Polymorphisms in intron 8 and exon 19 were amplified and analyzed by restriction digestion or allele-specific oligonucleotide hybridization.
Genotyping was performed on the following individuals, including two with FHM and two with migraine without aura, who gave their informed consent to the study: I-1, I-2, I-3, I-4, I-5, I-6, II-1, II-2, II-3, II-4, and II-5.
DG-DGGE analysis.
The whole CACNA1A gene, 47 exons and exon–intron boundaries, was analyzed by DG-DGGE, as described previously by our group,26 in the proband affected by FHM (Individual I-6).
DNA sequence analysis.
PCR products from exon 13 amplified for Patient I-6 and a normal control subject were excised from 1% low melting gels, after gel electrophoresis, and purified by using the QIA quick gel extraction kit (Qiagen; Chatsworth, CA). Purified fragments were analyzed by cycle sequencing using the FS dRhodamine Dye Terminators Ready Reaction Kit (Perkin-Elmer) according to the manufacturer’s instructions, and were loaded on a 377DNA automatic sequencer (Perkin Elmer—ABI). Sequencing data were analyzed with the DNA Strider and Navigator software (Perkin Elmer–ABI).
Restriction analysis.
Exon 13 was amplified as described16 for all family members and for 80 unaffected control subjects from the general population. The 312-base pair (bp) PCR product (approximately 500 ng) was digested with BanII according to the manufacturer’s instructions. The resulting bands were analyzed on agarose minigels. In normal individuals, 123-, 122-, and 67-bp fragments are detected, whereas in affected individuals an additional 245-bp band resulting from the mutated allele is present.
PCR and gene scan detection of exon 1 variant.
The GGC polymorphic repeat was detected by fluorescent in vitro amplification with primers described previously16 and separation on a 6% polyacrylamide gel on a 373-Gene-Scan Automatic System (Perkin Elmer–ABI).
Results.
Magnetic resonance imaging.
Brain MRI was performed on the two family members affected with hemiplegic migraine (Individuals I-4 and I-6) and on one family member with migraine without aura (Individual I-2). The two affected members had cerebellar atrophy with a moderate enlargement of the cerebellar sulci (figure 2). In Individual I-2, MRI results were normal.
Figure 2. Brain MR image of the proband (Individual I-6). This T1-weighted, midsagittal image shows cerebellar vermian atrophy.
Genotyping.
To evaluate the association of FHM with chromosome 19p in this family, several polymorphic markers were analyzed. Patients with FHM (Individuals I-4 and I-6) shared the same haplotype on both chromosomes. Haplotype analysis was uninformative because of the small size of this family. However, haplotype segregation did not exclude a possible linkage to the 19p13 region. All families with FHM and cerebellar ataxia tested so far by linkage analysis or direct CACNA1A mutation screening were linked to an alteration of this gene on chromosome 19. Based on this observation we decided to search for mutations directly within the CACNA1A gene.
Mutation detection.
Mutation analysis was performed using DG-DGGE on DNA obtained from the proband (Individual I-6). All fragments presenting a mobility variation were confirmed and identified by direct sequencing. Among these, although no previously described mutations were present, a new variant in exon 13 was found, consisting of a G-to-A substitution at nucleotide 2023 changing a Gln for an Arg at codon 583, located within the putative IIS4 transmembrane segment. Within the family, segregation analysis was performed by restriction digestion because the 2023G-A substitution abolishes a BanII restriction site (see figure 1). In addition to the proband, the mutation was found in Patient I-4 and her son (Individual II-4), who was apparently unaffected at age 32. No other family members or 150 normal control alleles were positive for the presence of this mutation.
An additional variant in exon 1 was detected in a control DNA sample, and it is an insertion of a GGC triplet in the 5′ untranslated region, 53 bp upstream to the ATG initiation codon.16,21 This insertion was localized within a stretch of seven GGC repeats. The frequency in 150 alleles from the general population was 96% for the (GCC)7 allele and 4% for the (GCC)8 allele, thus it can be considered a polymorphic variant.
Discussion.
The current study describes a family with FHM in which the two affected family members, in addition to the typical paroxysmal manifestations of FHM, exhibited permanent, late-onset cerebellar symptoms. Mutations analysis revealed the presence of a new Arg583Gln missense mutation located in the second putative voltage sensor domain of the S4 segment of the CACNA1A gene. In our family, as well as in other FHM families reported,7 individuals with migraine without aura were found next to individuals with typical hemiplegic migraine attacks. Nevertheless, migraine without aura was not shown to be associated with the mutation in exon 13 of the CACNA1A gene.
The mutation detected in this family results in a major change because a positively charged arginine is substituted with a neutral, nonpolar glutamine within one of the S4 domains that represent the voltage sensors of the channel and are highly conserved in Ca2+ and Na2+ channels (figure 3). In particular, Arg583 is the outermost of the three Arg located in each S4 transmembrane segment.
Figure 3. Alignment of deduced calcium and sodium channel sequences (single-letter code) from human and rabbit over a part of the IIS4 domain (residues from 578 to 597 of the P/Q-type channel). Positionally conserved residues are in bold type. Positively charged arginine (Arg) residues are underlined. The mutation in the human P/Q Ca2+ channel involves a highly conserved Arg 583 (boxed).
Two other mutations involving S4 segments have been described so far. The first was identified by Ophoff et al.16 in a pure FHM family (It-II), linked to chromosome 19, showing the substitution of the first Arg of the IS4 domain of the CACNA1A gene with a Gln. Another missense mutation involving the functional corresponding Arg (IIS4 domain) in the CACNA1S gene, expressing the α1-subunit of the skeletal muscle L-type voltage-gated calcium channel, has been reported.27 In that case, the mutation was an Arg528His substitution, one of the more common mutations found in hypokalemic periodic paralysis (HypoPP)—an autosomal dominant muscle disease.27 Changes in physiology and protein structure that result from the Arg528His substitution in HypoPP have been investigated.28,29 Based on these studies it may be suggested that also the Arg583Gln mutant in FHM can lead to a global distortion of positively charged amino acids within the IIS4 domain and ultimately may affect gating properties of the P/Q-type calcium channel.
Interestingly, incomplete penetrance was observed for the Arg528His mutation in women from families with HypoPP.30 In our family, the missense mutation Arg583Gln was found in affected family members but also in an asymptomatic individual, who denied any history of headache. This finding could be explained on the basis of incomplete penetrance of the disease in this man. In addition, either the onset of the disease later in life, like his mother had, or features of the normal allele could be hypothesized to modify the phenotype.
Although we did not demonstrate that the G-to-A substitution at codon 583 represents the disease-causing mutation in our family, this can be reasonably assumed considering the absence of the mutation from the general population; its location in a functionally important, highly conserved domain; and evidence of a similar disease-causing mutation associated with the IIS4 domain in HypoPP.
The two affected members of this family developed, in their 60s, the additional feature of progressive cerebellar ataxia with cerebellar atrophy evident on MRI. The association of FHM and cerebellar ataxia has been reported in a small number of FHM families, all linked to chromosome 19.6,7,10-14 In some of the patients described previously, the cerebellar manifestations developed during the FHM attacks and remained thereafter; in others, they developed independently from the attacks.13 In our patients the cerebellar symptoms were permanent and not influenced by migraine attacks, developing around the sixth decade when the attacks became severe and frequent. This clinical course resembles that of Patient 1 described by Ohta et al.31
So far, only two missense mutations (Thr666Met and Ile1811Leu) within the CACNA1A gene have been associated with cerebellar ataxia in three families with FHM linked to chromosome 19.12,14,16 Thus, our family represents the fourth chromosome 19-linked FHM family associated with progressive cerebellar ataxia in which the molecular defect has been identified. Mutations found in FHM families with cerebellar ataxia are all located in different domains of the α1A-subunit and this could partially explain differences in the clinical course of cerebellar manifestations. In addition, by looking at the FHM phenotypes in these families, a marked phenotypic variability was described.10,14 In this regard, in our family the clinical course in the two affected members differed from that of other reported patients with FHM9,13 because the frequency and severity of attacks increased after middle age. In addition, the older, affected member experienced severe episodes of acutely altered consciousness and fever during FHM attacks, whereas Individual II-4 is completely asymptomatic. In conclusion, a clear genotype– phenotype correlation is not easy, and is probably influenced by environmental factors, modifying genes, and other factors such as hormone levels, as demonstrated previously for voltage-gated sodium currents.32
Interestingly, in the two affected members, acetazolamide was an effective prophylactic therapy for the migraine attacks, although it did not show any effect on the progression of the ataxia. Oral acetazolamide has proved to be a very effective treatment in EA-2,33 HypoPP,34 and other periodic neurologic diseases associated with an ion channel defect.35,36 It has been proposed that it acts by stabilizing the abnormal ion channels through changes in extracellular pH.37 Regarding the basis of the overlap in the phenotypes of FHM to EA-2, the usefulness of acetazolamide has been speculated as a treatment for FHM.12,18 However, besides the current study, the only report of acetazolamide responsiveness in patients with FHM was from Athwal and Lennox.38 In addition, Baloh et al.39 reported a beneficial effect of acetazolamide in a family with dominantly inherited migraine. Taken together these observations strongly support the need of formal trials of acetazolamide in FHM and in nonhemiplegic migraine.
From the study of the families with FHM associated with ataxia, it remains difficult to determine whether cerebellar ataxia represents an independent phenotypic feature. In the current study, our patients developed a progressive cerebellar ataxia near the sixth decade, when the attacks became more frequent and severe. This observation could support the hypothesis that episodic dysfunction and permanent deficit may share a similar pathogenic mechanism. However, in our patients the prophylactic effect of acetazolamide was limited to the migraine attacks but it did not affect the progression of ataxia, thus suggesting different pathophysiologic mechanisms for the two clinical features. In other ion channel disorders, such as HypoPP, treatment with acetazolamide up to a maximum dosage of 1,500 mg/day has been shown to improve the permanent deficit in some patients, even though it had minimal or no effect in others.40 Accordingly, in our patients the dosage of acetazolamide was increased to a total of 750 mg/day. However, both patients experienced severe side effects (paresthesias and dizziness) and therefore they were placed again on the initial dosage. Further studies on the response to acetazolamide in additional patients with FHM and cerebellar ataxia would be helpful to understand better the relationship between the two clinical features.
As in HypoPP, in which clinical symptoms (either transient attacks or permanent muscle weakness) mirror the dual function of L-type Ca2+ channel,15 as well as in the P/Q-type channel-associated diseases, different phenotypes could develop depending on the variety of functions in which the channel is involved in the cerebellum, brainstem, and cortex.41,42 In this way, transcriptional isoforms of the gene would play an important role too.
Acknowledgments
Supported in part by a grant of the Italian government (Ministero della Università e della Ricerca Scientifica e Tecnologica: project 60%).
Footnotes
-
See also pages 3, 26, and 34
- Received January 5, 1999.
- Accepted March 16, 1999.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
