|Home | About | Journals | Submit | Contact Us | Français|
To perform a clinical and genetic study of a family with benign familial infantile seizures (BFIS) and, upon finding a PRRT2 gene mutation, to study a cohort of probands with a similar phenotype. We extended the study to all available family members to find out whether PRRT2 mutations cosegregated with additional symptoms.
We carried out a clinical and genealogic study of a 3-generation family and of 32 additional probands with BFIS (11 families), infantile convulsions and paroxysmal choreoathetosis (ICCA) (9 families), BFIS/generalized epilepsy with febrile seizures plus (5 families), and sporadic benign neonatal or infantile seizures (7 probands/families). We performed a genetic study consisting of linkage analysis and PRRT2 screening of the 33 probands/families.
We obtained a positive linkage in the 16p11.3-q23.1 chromosomal region in the large BFIS family. Mutation analysis of PRRT2 gene revealed a c.649dupC (p.Arg217Profs*8) in all affected individuals. PRRT2 analysis of the 32 additional probands showed mutations in 10, 8 familial and 2 sporadic, probands. Overall we found PRRT2 mutations in 11 probands with a mutation rate of 11 out of 33 (33%). BFIS co-occurred with migraine and febrile seizures in 2 families, with childhood absence epilepsy in one family and with hemiplegic migraine in one family.
Our results confirm the predominant role of PRRT2 mutations in BFIS and expand the spectrum of PRRT2-associated phenotypes to include febrile seizures, childhood absence seizures, migraine, and hemiplegic migraine.
Benign familial infantile seizures (BFIS) is an autosomal dominant disorder characterized by the occurrence of afebrile focal seizures, often in clusters, beginning between the 3rd and 12th months of life in otherwise normal infants.1,2 Seizures have a very benign outcome and disappear by age 2 years in most children who have normal psychomotor development. During childhood or adolescence, some patients may develop paroxysmal kinesigenic choreoathetosis.3 The term infantile convulsions and paroxysmal choreoathetosis (ICCA) has been used to describe patients in whom seizures and paroxysmal choreoathetosis co-occur.4
Linkage studies of informative families with autosomal dominant paroxysmal kinesigenic choreoathetosis, ICCA, and BFIS identified a locus on the pericentromeric region of chromosome 16p11.2-q12.1.4–8 Recently, using whole exome sequencing, mutations in the proline-rich transmembrane protein 2 (PRRT2) gene were identified in 8 families with autosomal dominant paroxysmal kinesigenic choreoathetosis.9 PRRT2 gene maps to 16p.11.2 and encodes for the proline-rich transmembrane protein 2, which seems to be involved in the synaptic vesicle membrane docking and fusion pathway, interacting with SNAP25.9 Additional mutations in PRRT2 have been recently identified in families with ICCA and BFIS.10–14
We studied a large family with BFIS and obtained a positive lod score to the 16p12.3-q23.1 region, which prompted us to test PRRT2, identifying it as the disease-causing gene. We subsequently extended mutation analysis of PRRT2 to an additional cohort of either sporadic or familial cases with infantile seizures and ICCA. We identified PRRT2 mutations in 11 probands, including the original BFIS large pedigree, 5 additional families with BFIS, 3 families with ICCA, and 2 sporadic patients with benign infantile seizures. Clinical analysis of affected family members revealed that febrile seizures, migraine, hemiplegic migraine, and childhood absence seizures were additional phenotypic manifestations cosegregating with the mutated allele and identifies an additional genetic cause of migraine and hemiplegic migraine. This observation widens the spectrum of phenotypes associated to PRRT2 mutations.
We studied a 3-generation Italian family (family 1) ascertained through the proband (III-25, figure 1). We made a field trip to obtain information of available family members and constructed a large pedigree with 9 affected individuals over 3 generations (figure 1).
Upon finding a PRRT2 mutation in family 1 (see genetic analysis), we extended the study to a cohort of 32 familial or sporadic cases with BFIS (11 probands), ICCA (9 probands), BFIS and generalized epilepsy with febrile seizures plus (5 probands), and sporadic benign neonatal or infantile seizures (7 probands), in order to evaluate how common PRRT2 mutations are. Probands with infantile-onset seizure were recruited from pediatric epilepsy and neurology clinics. We obtained clinical information, EEG recordings, and neuroimaging of all probands. We collected genealogic information to identify familial cases and extended the clinical evaluation to available family members, to evaluate whether PRRT2 mutations were associated with not yet described additional symptoms.
All participants and their parents/legal guardians, in the case of minors, gave informed consent. The study was approved by the Human Research Ethics Committees of the Meyer Children's Hospital.
Blood samples were obtained after informed consent. Genomic DNA was extracted from blood leukocytes using an automated DNA isolation robot, according to the manufacturer's protocol (QIASymphony, Qiagen, Hilden, Germany).
Genome-wide linkage analysis was performed in family 1 using a set of 400 highly polymorphic fluorescent-labeled markers, at an average reciprocal distance of about 10cM (ABI PRISM Linkage mapping sets v 2.5; Life Tech, Foster City, CA). PCR products were analyzed on an ABI Prism 3130XL sequencer (Life Tech) and statistical calculations for 2-point and multipoint linkage analysis were performed using the easyLINKAGE package,15 modeling the disease as autosomal dominant and assuming a penetrance of 65%.
Considering that the PRRT2 gene maps to the critical region identified in family 1, and that mutations in this gene have recently been associated with phenotypes close to that observed in our patients, we considered this gene as the best candidate to be screened. The 4 exons of the PRRT2 gene (Entrez Gene, GeneID: 112476, accession number: NM_145239.2) and their respective intron-exon boundaries were amplified by PCR and cycle sequenced using the BigDye Terminator v.1.1 chemistry (Life Tech). Exon 2 was amplified as a fragment of about 2 kb and cycle sequenced using internal primers. The sequence reactions were analyzed on a 3130XL sequencer (Life Tech). Primer sequences and PCR/sequencing conditions are available as supplemental material (table e-1 on the Neurology® Web site at www.neurology.org). The identified PRRT2 variants were not found in a control population of 190 ethnically matched subjects. In silico analysis performed using Kaviar (http://db.systemsbiology.net/kaviar/cgi-pub/Kaviar.pl) excluded that the identified variants had been reported as polymorphisms. The possible pathogenic effect of the identified variants has been evaluated using Polyphen2 (http://genetics.bwh.harvard.edu/pph2/index.shtml), Mutation Taster (http://www.mutationtaster.org/), and SIFT (http://sift.jcvi.org/www/SIFT_enst_submit.html) for missense mutations and the BDGP Splice Site Prediction (http://www.fruitfly.org/seq_tools/splice.html), the Human Splicing Finder (http://www.umd.be/HSF/), and the NetGene2 Server (http://www.cbs.dtu.dk/services/NetGene2/) for the splice site mutation.
A brief summary of relevant clinical information of affected individuals, including 9 familial and 2 sporadic patients, is summarized in the table.
Family 1 included 9 clinically affected individuals who presented with clusters of focal, afebrile seizures, at a mean of age 7 months (range 5–8) (figure 1). All affected individuals, with an age at the time of the study ranging from 2.5 to 44 years, mean 19 years, are at present seizure-free.
We also studied 32 probands, either sporadic or familial, whose phenotypes were 1) BFIS in 11 probands; 2) ICCA in 9 probands; 3) BFIS/generalized epilepsy with febrile seizure plus spectrum in 5 probands; 4) sporadic benign infantile seizures in 4 probands; and 5) neonatal seizures in 3 probands. We identified PRRT2 gene mutations in 10 probands (figure 2), bringing the total number of probands with mutations to 11/33 (33%).
Overall, the whole cohort of families/patients carrying PRRT2 mutations (including family 1) had a total of 27 individuals who had had focal seizures (figure 3) at a mean age of 7 months (median 7, range 3.5–8 months), 11 individuals with paroxysmal dyskinesia migraine, or both, and hemiplegic migraine and 4 in whom infantile seizures were variably associated with migraine, paroxysmal dyskinesia, and hemiplegic migraine (figure 2). A single individual (figure 2, family 8) had had only febrile seizures. Seizures manifested in clusters in all probands. In 5 of them, they were characterized by unresponsiveness followed by perioral cyanosis, staring, and hypotonia, lasting 30 to 60 seconds. During seizures, 3 of these 5 probands had lateral eye deviation and 2 also exhibited oral automatisms associated to a hiccup-like sound. In the remaining 6 probands, following initial unresponsiveness and staring, a hypertonic phase ensued, with a diffuse vibratory tonic component. In addition to focal seizures, 4 of the 11 probands also exhibited tonic-clonic seizures that were reported to be generalized from onset. All affected individuals had normal developmental milestones and cognition.
Nine of the 11 probands had a family history of seizures or other paroxysmal disorders, or both, consistent with a diagnosis of 1) BFIS in 6 families, one of which had a family member who had absence seizures in childhood and one with hemiplegic migraine in 5 individuals; 2) ICCA in 3 families with migraine and febrile seizures in 2. There were also 9 additional healthy individuals from family 1 (figure 1) and one unaffected obligate carrier from family 11 (figure 2) carrying PRRT2 mutations.
Linkage analysis, performed in family 1, revealed a maximum 2-point lod score of 3.31 at marker D16S415, at θ = 0.0 (table e-2), while multipoint lod score analysis reached a maximum value of 2.71 at marker D16S415 (figure e-1). The highest probability haplotype inferred using HaploPainter16 was consistent with key recombinant events between markers D16S3046 and D16S3068 in patient III-14 and between markers D16S5415 and D16S503 in patient III-24, outlining the common haplotype shared by all affected members (figure 1).
Sequencing analysis of the PRRT2 gene in the proband of family 1 demonstrated the duplication of a single base (c.649dupC) in exon 2, resulting in a frameshift with the insertion of a premature stop codon (p.Arg217Profs*8). The same mutation was identified in all affected individuals, in 1 obligate carrier, and in 8 asymptomatic individuals (figure 1). The p.Arg217Profs*8 is a recurrent frameshift, loss of function mutation13 that has also been identified in families and in sporadic patients with paroxysmal kinesigenic dyskinesias with or without infantile convulsions and benign familial infantile epilepsy.9–14 The analysis of additional BFIS families led us to identify the c.649dupC in 3 additional familial cases and in the 2 sporadic cases (table, figure 2). In one of the 2 sporadic patients, the mutation occurred de novo. No DNA could be obtained for the parents of the other sporadic patient. In the remaining families, we found 3 novel missense mutations (c.916G > A; p.Ala306Thr, c.970G > A; p.Gly324Arg, and c.971G > A; p.Gly324Glu) and one novel splicing mutation (c.880-2A > T).
Linkage analysis in family 1, with 9 affected individuals, showed a significant lod score in the 16p12.3-q23.1 region, partially overlapping previously published BFIS, ICCA, and autosomal dominant paroxysmal kinesigenic choreoathetosis loci.2 Screening of the PRRT2 gene in available members of family 1 identified a frameshift mutation, p.Arg217Profs*8, in all affected individuals, in one obligate carrier, and in 8 asymptomatic individuals. The identification of asymptomatic individuals carrying the mutation is not surprising, since a reduced penetrance for BFIS has been already described.2 The p.Arg217Profs*8 mutation, which is a loss of function mutation,13 is emerging as a very recurrent mutation in patients with paroxysmal kinesigenic dyskinesia, with or without infantile convulsions and BFIS.10–14
After identifying the c.649dupC mutation in family 1, we assembled and studied a cohort of 32 probands with BFIS, ICCA, and other idiopathic infantile seizures, either familial or sporadic, and identified PRRT2 mutations in 10 additional probands (11/33; 33%). Family history revealed that 8 probands had a familial segregation: 5 had pure BFIS, 1 had ICCA, 1 had a co-occurrence of BFIS and childhood absence epilepsy, 1 had a cosegregation of BFIS and hemiplegic migraine, and 2 had a combination of BFIS, febrile seizures, paroxysmal dyskinesia, and migraine. The 2 sporadic patients had benign infantile seizures with de novo PRRT2 in one, whereas DNA could not be obtained for the parents of the other sporadic patient. In our study, there were 6/11 BFIS families carrying PRRT2 mutations, corresponding to 54.5%, and 3/8 with ICCA, corresponding to 37.5%. These figures are smaller than previously reported, in which 80% of BFIS and ICCA families carried PRRT2 mutations.14
Of the 34 affected individuals carrying a PRRT2 mutation, 3 (from 2 different families) manifested migraine and 5 (from a single family) had hemiplegic migraine consisting of yearly attacks of a tingling sensation in one arm, leg, and face followed by the inability to move the affected limbs and aphasia. Attacks, sometimes accompanied by vomiting, were associated with severe headache and lasted several hours. These episodes are therefore clearly distinguishable from epileptic seizures. The association with hemiplegic migraine is remarkable and suggests that PRRT2 gene might be added to the list of genes associated with this disorder including SCN1A, ATP1A2, and CACNA1A.17 These 3 best known hemiplegic migraine genes all encode ion transporters, suggesting that disturbances in ion and neurotransmitter balances in the brain are responsible for this migraine type. PRRT2, which is involved in the synaptic vesicle membrane docking and fusion pathway, could cause migraine through a different, yet unknown mechanism. The c.649dupC mutation identified in the family with BFIS and hemiplegic migraine was also identified in 5 other families in our series (figure 2) and in several already published families.10–14 However, hemiplegic migraine was present only in this family (figure 2, family 11), possibly suggesting an interfamilial phenotypic variability and the interaction of PRRT2 with genetic modifiers.
We found the c.649dupC mutation in 6/10 mutated families and in a sporadic patient. This finding is in agreement with previous reports and confirms that the nucleotide c.649 may be considered as a mutation hotspot in PRRT2. In addition, we found 4 novel mutations (3 missense and a splicing mutation). Two missense mutations (c.970G > A; p.Gly324Arg and c.971G > A; p.Gly324Glu) affect the Gly324 residue and the remaining missense mutation (c.916G > A) the residue Ala306. The transmembrane Gly324 residue is highly conserved and the 2 mutations affecting it cause nonconservative amino acid changes that are predicted to be highly pathogenic. The Ala306Thr substitution is predicted to be more tolerated, since alanine and threonine share similar chemical and physical properties. However, the Ala306 residue has a high evolutionary conservation which, taken together with its segregation in both affected family members on family 9 (figure 2), supports its pathogenic role. The fourth novel mutation we identified (c.880-2A > T) affects the canonical acceptor splice site of the exon 3 and is likely to alter the splicing process, leading to the skipping of exon 3 resulting in a truncated protein.
Our findings confirm that screening for PRRT2 gene mutations is indicated both in familial and sporadic cases of infants presenting with clusters of afebrile seizures with normal psychomotor development, normal background EEG activity, and normal imaging. Although in our series seizure onset was within the first year of life, with peak at around the 7th month, later onset is possible, usually within the 18th month.14 Familial cosegregation of other paroxysmal neurologic disorders such as paroxysmal dyskinesia, migraine, or hemiplegic migraine is strongly suggestive of a PRRT2 gene–related disorder. Infants without clear evidence of additional affected family members should also be tested since seizures occurring very early in infancy may not be on record. To reach a genetic diagnosis is important in order to avoid expensive and invasive investigations and for genetic counseling.
The authors thank the patients and their families for participating in the research.
Study concept and design: Dr. Marini, Dr. Conti, D. Mei, and Dr. Guerrini. Acquisition of data: Dr. Marini, Dr. Conti, D. Mei, Dr. Battaglia, Dr. Lettori, Dr. Losito, Dr. Bruccini, Dr. Tortorella, and Dr. Guerrini. Analysis and interpretation of data: Dr. Marini, Dr. Conti, Dr. Battaglia, Dr. Lettori, Dr. Losito, and Dr. Guerrini. Drafting of the manuscript: Dr. Marini, Dr. Conti, and Dr. Guerrini. Critical revision of the manuscript for important intellectual content: Dr. Guerrini. Obtained funding: Dr. Marini. Administrative, technical, and material support: Dr. Conti, D. Mei, Dr. Battaglia, Dr. Bruccini, Dr. Tortorella, and Dr. Guerrini. Study supervision: Dr. Guerrini.
C. Marini receives research support from the Italian Minister of Health, Research Program Section (RF-2009-1525669). V. Conti, D. Mei, D. Battaglia, D. Lettori, E. Losito, G. Bruccini, and G. Tortorella report no disclosures. R. Guerrini has received honoraria from Biocodex, UCB, Eisai Inc, ValueBox, and EMA (European Medicine Agency). He receives research support from the Italian Ministry of Health, the European Community Sixth Framework Thematic Priority Life Sciences, Genomics and Biotechnology for Health, the Italian Ministry of Education, University and Research, the Tuscany Region, the Telethon Foundation, and the Mariani Foundation. Go to Neurology.org for full disclosures.