Abstract
Objective: To verify linkage to chromosome 19p13, to detect mutations in the CACNA1A gene, and to correlate genetic results to their clinical phenotypes in Italian families with familial hemiplegic migraine (FHM).
Background: FHM is an autosomal dominant disease, classified as a subtype of migraine with aura. Only a proportion of FHM patients have been associated with chromosome 19p13. Among these, four missense mutations within the CACNA1A gene in five unrelated families have been described.
Methods: A linkage study was performed in 19 patients affected by FHM from five families by studying microsatellite markers associated with the 19p13 region. All familial and seven additional sporadic patients with FHM were analyzed to search for mutations within the CACNA1A gene by applying the double gradient–denaturant gradient electrophoresis technique.
Results: Lod score values did not establish significantly linkage to chromosome 19. However, seven new genetic variants were detected: six were new polymorphisms. The seventh was a missense mutation present in family 1, and it was associated with a hemiplegic migraine phenotype without unconsciousness and cerebellar ataxia. Because this missense mutation is absent in the general population and cosegregates with the disease, it may be a pathologic mutation.
Conclusions: Genetic heterogeneity of FHM has been shown in familial and sporadic FHM patients of Italian origin. The new missense mutation—G4644T—is associated with milder clinical features compared with typical FHM.
The CACNA1A gene, located on chromosome 19p13, has been identified recently.1,2 It encodes a putative protein that is highly homologous to rabbit and rat brain-specific P/Q-type Ca2+ channel α1-subunit.3,4 Mutations within this class A voltage-dependent calcium channel α1-subunit were found to cause different dominant neurologic disorders, including familial hemiplegic migraine (FHM), episodic ataxia type 2 (EA-2), and spinocerebellar ataxia type 6 (SCA-6).1,2 In another recent report,5,6 the “tottering” and “leaner” mice were found to carry mutations in the homologue of the human CACNA1A gene. Of particular interest, the “tottering” phenotype shows typical seizures with a type and evolution that are remarkably similar to those seen in human absence epilepsy. If confirmed, this result might reveal at least one additional genetic disease due to mutations within the human CACNA1A gene. The cerebellar degeneration found in the “leaner” mice is also reminiscent of the cerebellar degeneration found in some of the most severe cases of the human FHM, EA-2, or SCA-6.
Calcium channels are multisubunit complexes in which ionophore activity is mediated mainly by the α1 pore-forming subunit. The β-, α2/δ-, and γ-subunits are accessory components that regulate channel activity. The high-threshold calcium channels have been defined in the peripheral and central neurons according to their electrophysiologic and pharmacologic properties by the letters L, N, P, Q, and T.7,8
The heterologously expressed class A subunits show pharmacologic properties similar to the P/Q-type voltage-operated calcium channels.9 Interestingly, the class A transcripts and the class A protein are particularly expressed in the cerebellum, either in Purkinje or granular cells, both pre- and postsynaptically.4 This high expression of class A proteins in the cerebellum is likely to be correlated with the cerebellar symptoms found in some of the patients affected by FHM.
FHM, classified as a subtype of migraine with aura, is usually characterized by transient hemiparesis lasting from minutes to several days, and recovering spontaneously with no clinical sequelae. In 20% of patients, cerebellar ataxia is also present.1,6 As part of the migraine spectrum, FHM is a good model for approaching the genetics of common types of migraine, a very common multifactorial disease, especially after the recent description of an involvement of the 19p13 region in common forms of migraine.10
The goal of the current study was to verify a possible linkage to chromosome 19p13, to characterize mutations, and to evaluate possible correlations with specific phenotypes in Italian individuals affected by FHM. Identification of new mutations, in fact, will be critical for elucidating structure–function relationships. We used double gradient–denaturant gradient electrophoresis (DG-DGGE)—a highly sensitive modification of DGGE analysis11—to assess all exons and exon–intron boundaries of the CACNA1A gene.
Patients and methods.
Clinical data.
The diagnosis of FHM was based on the diagnostic criteria of the International Headache Society.12 Twenty-six FHM patients (7 sporadic and 19 from 5 unrelated families) for a total of 51 individuals were analyzed. Pedigrees are shown in figure 1. Genetic analysis was performed on individuals who gave their consent to the study.
Figure 1. Pedigrees of families (Fam.) with familial hemiplegic migraine (FHM).
The first pedigree (family 1; see figure 1) is a five-generation Caucasian family originating from the northeastern part of Italy (Ferrara) with an average age at onset of 33.8 years (range, 10 to 56 years). FHM patients in the second generation had an onset of symptoms at 54 and 45 years of age. Among patients of the third generation, three patients had a mean age at onset of 23.3 years; the other two patients had unspecified juvenile onset. All patients had clinical symptoms preceded by aura (usually visual), followed by hemiparesis and various degrees of aphasia congruent with the hemispheric dominance of each individual. Of the numerous attacks over several years, patients never reported cerebellar ataxia or coma. The second pedigree (family 2; see figure 1) is a two-generation Caucasian family from Ferrara (epidemiologically and genetically unrelated to family 1), in which symptoms of the neurologic aura include impaired consciousness (in one patient proceeding into a coma, with ictal EEG showing persistent delta waves contralateral to the hemiplegic side). Three of four patients of this kindred reported persistence of attention and memory complaints as long as 2 months after resolution of hemiplegia and headache. The third pedigree (family 3; see figure 1) is a three-generation Caucasian family with three patients affected by FHM. All patients had clinical symptoms preceded by aura followed by hemiparesis. One of the patients in the second generation, with an onset of symptoms at 29 years, had recurrent attacks with coma lasting 1 to 2 days. Her sister had onset at 4 years and a milder phenotype with no unconsciousness reported. The patient in the third generation had an onset at 28 years of age, and he also shows a mild phenotype. Both the two affected patients of the fourth pedigree (family 4, figure 1) had clinical onset at age 14 years. The father had an average of three attacks per year during his late childhood. Attacks scaled down to one per year beginning in adulthood to the present. His son (age, 16 years), so far, has had only a single attack. Stereotyped ataxia in this pedigree consists of mild hemiplegia of approximately 1 hour duration, followed by lateralized headache and tiredness. The fifth pedigree (family 5; see figure 1) consists of a three-generation Caucasian family. The two affected subjects had juvenile onset (ages 14 and 15 years respectively) characterized by recurrent episodes of a 2-hour aura with mild hemilateral sensorimotor deficits (mostly confined to the oral rim and arm), followed by lateralized headache.
Sporadic patients had at least another relative with a history of signs and symptoms suggestive of FHM.13 Patient 1 (E.V.) is a 41-year-old woman who had onset at 29 years with hemiplegic aphasia followed rapidly by lateralized headache that persisted for 2 to 3 days. During the past 2 years the patient has experienced an increase in the frequency of attacks from one to four per year. Patient 2 (G.C.) is a 40-year-old Caucasian son of a migrainous woman. He experienced his first attack of hemiplegia followed by lateralized headache and coma in March 1993, followed by a single, additional, similar episode 5 months later. Patient 3 (D.B.) is a 27-year-old man with a family history consistent with FHM in relatives (his father and a cousin) who had several episodes of visual aura followed by hemiplegia and lateralized headache. Patient 4 (C.L.) is a 46-year-old Caucasian woman with a family history of FHM. She recalls onset at age 20 years associated with frequent episodes of alternating hemiplegia with aura characterized by vomiting, and photo- and phonophobia. A stronger attack occurred at age 35 years with intense left hemiparesis and coma. The left hemiparesis has not recovered completely since, and she now has a shuffling gait. Patient 5 (G.B.) is a 25-year-old Caucasian man with a history of recurrent episodes of migrainous headache with visual aura and tiredness that began at the age of 13 years and turned into hemiplegic migraine at age 19 with rapidly succeeding attacks of alternating hemiplegia, aphasia if hemiplegia occurred on the dominant side, and impaired consciousness. Brain MRI was normal on several occasions. Patient 6 (G.D.) is a 22-year-old Caucasian man with onset at age 18 years of hemiplegia, diplopia, and vomiting followed by lateralized headache occurring one to two times a year. Brain MRI was unrevealing. Patient 7 (E.C.) is a 44-year-old Caucasian man who presented at the age of 34 years with episodes of short-lasting alternating hemiplegia followed by lateralized, pulsating headache. Similar episodes were reported by the father, one of the brothers, and a cousin. Brain MRI, angiography, and SPECT were normal.
Control subjects.
A total of 75 unrelated individuals from the general population were recruited as control subjects. They had no family history for FHM or EA-2.
DNA extraction.
Genomic DNA was extracted from peripheral blood lymphocytes using standard techniques.14
Microsatellite analysis.
Genotyping was performed on all the components of the five families (see figure 1). The following polymorphic microsatellite markers surrounding the region where the CACNA1A gene is located were analyzed: D19S216, D19S391, D19S413, D19S394, D19S221, and D19S226. Microsatellite genomic sequences were amplified by PCR in the presence of a primer labeled terminally with either 32P or 6-carboxyfluorescein and analyzed with 6% polyacrylamide gel electrophoresis (PAGE) by autoradiography or on a 373-ABI Genescan system (Perkin-Elmer, Foster City, CA). Oligonucleotide sequences are available through the Human Genome Data Base.
Linkage analysis.
Linkage analysis was performed using the Linkage package, version 5.1.13 FHM was regarded as an autosomal dominant disease with an incomplete penetrance of 0.8 and a disease gene frequency of 0.0001. Four liability classes were used, as described previously by Ophoff et al.1 Lod scores were calculated for each marker for various recombination fractions. Marker allelic frequencies and number of alleles of the Genome Data Base were used. Asymptomatic individuals less than the age of 37 years were regarded as being of unknown status. Haplotypes were inferred by minimizing the number of crossovers.
Screening of previously described mutations.
All affected individuals with both familial and sporadic cases of FHM were tested initially for the absence of previously described mutations associated with FHM1 and for the CAG repeat in exon 47 associated with SCA-6. The CAG repeat was amplified as described previously1 and analyzed on a 373-ABI Genescan system.
PCR and DG-DGGE analysis.
The region analyzed with DG-DGGE comprises all but one of the exons of the CACNA1A gene and exon–intron boundaries, amplified in 51 fragments. PCR primers were based substantially on the sequences published by Ophoff et al.,1 with modifications required from the simulation of DNA melting profiles. Calculation of dissociation constants was performed by using the Melt 87 program.15 The complete listing of primers for DG-DGGE and their melting temperatures are shown in table 1. All fragments were GC-clamped,16 except the primer sequence of exon 1, which presents a high melting domain at the 5′ region, and the primer of exon 47, which was analyzed by sequencing. PCRs were performed on 300 ng genomic DNA extracted from peripheral blood. Annealing temperatures for each fragment are indicated in table 1. Further details are available on request to P.C. by e-mail or fax.
Primers for double gradient–denaturant gradient electrophoresis
PCRs were performed in 9600 and 2400 Perkin-Elmer thermal cyclers, with each step lasting 30 seconds. To obtain heteroduplexes, amplified fragments were denatured at 94 °C for 10 minutes and left to reanneal slowly at 56 °C for 60 minutes.
The general conditions used for DG-DGGE analysis have been described previously.11 DG-DGGE was performed in the D-CODE System (Bio-Rad; Hercules, CA). A total of 15 μL of each PCR product was loaded in a polyacrylamide gel containing a porosity gradient that ranged from 6.5 to 12% of monomers for fragments more than 250 base pairs (bp) and from 8 to 15% of monomers for fragments less than 250 bp. Table 2 shows the gradients of gel porosity, range of denaturants, temperature, running time, and voltage used for different fragments. Due to the uniformity of running times, different fragments presenting the same melting profile were loaded onto the same gradient.
Double gradient–denaturant gradient electrophoresis conditions
PCR and direct sequencing.
Double-stranded PCR products were analyzed directly with cycle sequencing using either the FS Dye Terminators or the FS dRhodamine Dye Terminators Ready Reaction Kit (Perkin-Elmer) and loaded on the ABI-373 and -377 DNA automatic sequencers. Sequencing data were analyzed with the DNA-Strider (version 1.0) and Navigator (version 1.0.1) software (Perkin-Elmer).
PCR-mediated site-directed mutagenesis (PSDM) system.
Generation of allele-specific restriction enzyme recognition sites was accomplished by PSDM.17 To detect the A-to-G variant rapidly at position 1474-31 in the general population, a mutagenic primer was designed: 5′ GAGAACTCATCCTCCAAATCC. The forward primer contains a C (italics) instead of a T at position 1474-33. This substitution inserts a novel MspI site CCGG in the amplified product of allele 1474-31G, whereas in the amplified product of allele 1474-31A the MspI site is absent.
The same strategy was utilized to detect the G-to-T substitution at position 4644 in exon 27. The mutagenic forward primer: 5′ CTCTGCTGACCCTCTTCAGC contains a G (italics) instead of a C at position 4642. This substitution inserts a novel AluI site AGCT in the amplified fragment of allele 4644T, whereas in the amplified product of allele 4644G the AluI site is not present.
Lastly, a PSDM was set up to detect the C-to-T substitution at position 3824 in exon 20. The mutagenic reverse primer 5′ AGAGGGTCTCACCTTGTTC contains a T (italics) instead of an A at position 3822. This substitution inserts a novel TaqI site TCGA in the amplified fragment of allele 3824C, whereas in the amplified product of allele 3824T the TaqI site is absent.
Restriction analysis.
Restriction analysis was performed on all the new variants except for Glu993Val. PCR products (approximately 500 ng) were digested with the appropriate restriction enzyme according to the manufacturer’s instructions. The resulting bands were analyzed on agarose or PAGE minigels.
Results.
Linkage study.
In the various families screened, a linkage study was performed to verify a possible association of the disease with the 19p13 region. Family 2 was excluded because of paternity uncertainty. Nevertheless, it was included in the screening for mutations because the mother was affected (see figure 1). Two-point lod score values between FHM and the selected markers were positive but not significant; the maximum combined lod scores of 2.94 and 2.81 were reached with markers D19S391 and D19S394 at θ = 0 respectively. The contribution of family 1 was particularly relevant, showing a maximum value of 2.72 at θ = 0 with the D19S394 marker. In the remaining families (families 3, 4, and 5), the maximum combined lod score of 0.58 was reached with marker D19S413 at θ = 0. The presence of meiotic crossovers contributed to lod score values < −2 in family 1 (−2.91 with D19S216, −2.84 with D19S226) and in family 3 (−2.61 with D19S221, −2.61 with D19S226). In the latter family, recombination at D19S221 was in contrast to the location of the gene proximal to this marker.1
Screening of previously described mutations.
Previously described mutations1 were not found in our patients. Also, the CAG repeat in exon 47 was always in the normal range (4-16 CAG).
DG-DGGE analysis.
A total of 11 fragments presenting a mobility variation were detected, showing typically two to four-band patterns (figure 2).
Figure 2. Double gradient–denaturant gradient electrophoresis separations of new variants within the CACNA1A gene. All the samples were run overnight at 60 V/cm. Increasing ranges of denaturant gradients are shown for fragments containing new variants. In each panel, fragments heterozygous (two- to four-band patterns) for a given variant and a normal control subject (single band) are shown. denat. = denaturant; int. = intron; ex. = extron.
Sequence variants detected within the CACNA1A gene.
All the variations revealed by DG-DGGE were confirmed and identified by direct sequencing. Among these, four variations were previously described polymorphisms: 1457 G-A in exon 8, 1635 G-A in exon 11, 2369 G-A in exon 16, and 4122 T-C in exon 23. The remaining seven are new variants (table 3). To investigate whether these new variants could represent a polymorphism of the CACNA1A gene, they were tested in 150 alleles from the general population. Especially for nonconservative changes, their frequency, together with cosegregation with the disease phenotype in the same family, was evaluated. Tables 4 and 5⇓ present the polymorphisms found in each patient. Two intronic polymorphisms were identified in positions 1253+114 G-A in IVS6 and 1474-31 A-G in IVS8. The latter showed a 0.4/0.6 heterozygosity and can be useful in linkage studies. Within the coding region, three point mutations not changing the codon were found in residues Arg1020, Val1183, and Ala1915, and two missense substitutions were found in residues Glu993 and Val1457. Interestingly, at codon 1020 in exon 19, the 3335-A polymorphic sequence is always associated with the 3253-T substitution at codon 993, and vice versa. This substitution causes the Glu993Val missense mutation, and it was detected in a sporadic case and in FHM family 5, where it cosegregated with the disease. However, it can be considered a polymorphic variant because both alleles are frequent in the general population, with the Glu allele in the 0.38 and the more common Val allele in the 0.62 of examined chromosomes. An additional missense Val1457Leu substitution in exon 27 was found in family 1. This substitution was completely absent in the general population and cosegregated with the disease in all eight affected subjects of this family, thus suggesting that the mutation might be causing FHM in this family. In figure 3, the sequencing electropherogram and the segregation analysis of the mutation are shown. Family 1 also showed polymorphisms in exon 8, in intron 8, and in exon 23. In the preliminary screening performed with single-strand conformation polymorphism18 and/or by denaturing high-performance liquid chromatography19,20 as reported by Ophoff et al.,1 the polymorphism in intron 8 and the mutation in exon 27 escaped detection, thus supporting the claim for higher sensitivity and detectability of the DG-DGGE technique compared with other methods. Lastly, table 6 lists nucleotide differences in the sequence reported by Ophoff et al.1 found in normal alleles.
New variants detected by double gradient–denaturant gradient electrophoresis in the CACNA1A gene
Polymorphisms of the CACNA1A gene in familial cases
Polymorphisms of the CACNA1A gene in sporadic cases
Figure 3. Missense mutation Val1457Leu in exon 27 of the CACNA1A gene. (A) Direct sequencing electropherogram of mutated and wild-type samples confirmed the heterozygosity of the patient with familial hemiplegic migraine. (B) Segregation analysis of the mutation in family 2 by PCR-mediated site-directed mutagenesis and AluI digestion as described in Methods. The 58-base pair (bp) band results from the mutated allele after digestion with AluI. (C) Normal control subject. Lane 1 = molecular weight marker FX174 DNA/HinfI.
Differences observed in CACNA1A published sequence
Discussion.
In the current report we describe a new Val1457Leu missense mutation associated with FHM. This mutation is putative until functional studies can prove that there is an alteration in channel function. However, its location, in the putative pore-forming (P) region between the S5-S6 transmembrane domains in motif III of the CACNA1A gene suggests a potential for interference in transmembrane conductance. The P region, in fact, is critical in determining ion selectivity and permeability of the Ca channels.21 In particular, ion selectivity involves conserved glutamate residues in the SS1-SS2 segments of domains I through IV of the α1-subunit. Interestingly, the contribution of glutamate in motif III has the strongest effect on ion selectivity, as shown for the L-type channel.22 The Val-to-Leu substitution is located four residues upstream to the glutamate. This Val residue is conserved throughout several Ca channels α1-subunits, and also in the rat skeletal muscle Na channel, as shown in figure 4. This substitution could interfere with the correct folding of this region, and in particular with the formation of a hairpin–loop structure.
Figure 4. A part of the putative pore-forming (P) region of domain III of various α1-Ca2+ channels and an α-Na+ channel subunit are shown (residues from 1452 to 1466 of the P/Q-type channel). Positionally conserved residues are capitalized. Glutamate negatively charged residues in Ca2+ channels, participating in ion binding, and the corresponding residue in the Na+ channel are in bold type. The Val1457Leu mutation in the human P/Q Ca2+ channel is in bold type and is underlined (V). Rat N = rat α1-Ca2+ channel, N type.
Six missense mutations have been described originally both in human and mutant “tottering” mouse class A Ca channel genes,1,5,23 all of them lying in transmembrane domains of the α1-subunit. Among these mutations, three are located within different P regions. The Thr666Met, located in the putative P region of domain II, was found in a family with hemiplegic migraine and mild interictal ataxia.1 In the tottering mouse, the Pro601Leu mutation, situated in the P region of domain II of the channel,5 is associated with ataxia and motor seizures. Finally, the Gly293Arg mutation has been reported in the P region of domain I in a family showing mixed episodic and severe progressive ataxia.23 In our family, carrying the Val1457Leu mutation, classic symptoms of hemiplegic migraine were found, with a variable age at onset ranging from 10 to 56 years, and with a different frequency of episodes in affected individuals. No ataxia was ever reported.
All these findings suggest that mutations in different P regions could occur quite frequently within the CACNA1A gene and that these mutations, located within functionally related domains of the channel, can result in different phenotypes. As mentioned earlier, glutamate residues in P regions exert a distinct contribution to ion selectivity, thus providing a partial explanation for the association of such mutants with different phenotypes. This high degree of phenotypic variability may represent a major problem in development of pharmacologic therapies.
Among the other new variants, five were silent polymorphisms (either intronic or conservative) and one was the clinically irrelevant missense mutation in exon 19. The significance of this Glu-to-Val substitution should be investigated for elucidating the role of this variant in the normal physiology of the channel. In fact, this variant resides in the linker region between domains II and III (LII-III) of α1A, within the region involved in the interaction with presynaptic proteins synaptosome-associated protein of 25 kDa and Syntaxin,24 and therefore could be important in modulating the interaction with these proteins.
In this study we were able to detect several new polymorphisms and a potential new mutation associated with FHM in family 1. In the other three pedigrees with slightly positive lod scores, and in sporadic patients, no mutations were found, thus suggesting genetic heterogeneity also in the Italian population, as reported previously in other ethnic groups.25,26 In a recent report,27 a single, large family with FHM was mapped on chromosome 1q31 in a region where the CACNA1E gene is located, adding more evidence to its heterogeneity. In our families with borderline lod scores we can postulate the existence of mutations in regulatory regions of the CACNA1A gene, the contribution of mutations located in alternatively transcribed regions of the same gene,2 or linkage to other candidate genes.
Acknowledgments
Acknowledgment
The authors are extremely grateful to the patients and their families for their collaboration. They thank Dr. M. Sessa for providing DNA samples of two sporadic patients.
Footnotes
-
See also pages 3, 34, and 38
-
P.G.R. is supported by A.I.R.C. and by Telethon (grant no. E.555). L.M.E.G. is supported by the Armenise–Harvard Foundation.
- Received July 14, 1998.
- Accepted March 23, 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.
You May Also be Interested in
Alert Me
Recommended articles
Variable clinical expression of mutations in the P/Q-type calcium channel gene in familial hemiplegic migraine
Familial typical migraine
High prevalence of CACNA1A truncations and broader clinical spectrum in episodic ataxia type 2
A new CACNA1A gene mutation in acetazolamide-responsive familial hemiplegic migraine and ataxia