SCN1A duplications and deletions detected in Dravet syndrome: Implications for molecular diagnosis

Authors

  • Carla Marini,

    1. Child Neurology Unit, Children’s Hospital A. Meyer, University of Florence, Florence, Italy
    Search for more papers by this author
  • Ingrid E. Scheffer,

    Corresponding author
    1. Department of Medicine (Neurology), University of Melbourne and Austin Health, Heidelberg, Australia
    2. Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Melbourne, Australia
      Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au
    Search for more papers by this author
  • Rima Nabbout,

    Corresponding author
    1. Service de Neurologie Pédiatrique, AP-HP, Hopital Necker-Enfants Malades, Centre de Référence Épilepsies Rares, Paris, France
      Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au
    Search for more papers by this author
  • Davide Mei,

    Corresponding author
    1. Child Neurology Unit, Children’s Hospital A. Meyer, University of Florence, Florence, Italy
      Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au
    Search for more papers by this author
  • Kathy Cox,

    Corresponding author
    1. Department of Genetic Medicine, Women’s and Children’s Hospital, Adelaide, Australia
      Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au
    Search for more papers by this author
  • Leanne M. Dibbens,

    Corresponding author
    1. Department of Genetic Medicine, Women’s and Children’s Hospital, Adelaide, Australia
    2. School of Pediatrics and Reproductive Health, University of Adelaide, Adelaide, Australia
      Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au
    Search for more papers by this author
  • Jacinta M. McMahon,

    1. Department of Medicine (Neurology), University of Melbourne and Austin Health, Heidelberg, Australia
    Search for more papers by this author
  • Xenia Iona,

    1. Department of Genetic Medicine, Women’s and Children’s Hospital, Adelaide, Australia
    Search for more papers by this author
  • Rochio Sanchez Carpintero,

    1. Paediatric Neurology Unit, Department of Paediatrics, Clinica Universitaria de Navarra, University of Navarra, Pamplona, Spain
    Search for more papers by this author
  • Maurizio Elia,

    1. Department of Neurology, Oasi Institute for Research on Mental Retardation and Brain Aging (IRCCS), Troina, Enna, Italy
    Search for more papers by this author
  • Maria Roberta Cilio,

    1. Division of Neurology, Bambino Gesu Children’s Hospital, Roma, Italy
    Search for more papers by this author
  • Nicola Specchio,

    1. Division of Neurology, Bambino Gesu Children’s Hospital, Roma, Italy
    Search for more papers by this author
  • Lucio Giordano,

    1. Department of Child and Adolescent Neurology and Psychiatry, Spedali Civili, Brescia, Italy
    Search for more papers by this author
  • Pasquale Striano,

    1. Muscular and Neurodegenerative Disease Unit, Institute G. Gaslini, University of Genova, Genova, Italy
    Search for more papers by this author
  • Elena Gennaro,

    1. Laboratory of Genetics, E.O. Ospedale Galliera, Genova, Italy
    Search for more papers by this author
  • J. Helen Cross,

    1. UCL Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust, London, United Kingdom
    Search for more papers by this author
  • Sara Kivity,

    1. Schneider Children’s Medical Center, Petaq Tikva, Israel
    Search for more papers by this author
  • Miriam Y. Neufeld,

    1. Department of Neurology, Tel Aviv Sourasky Medical Centre, Tel Aviv, Israel
    Search for more papers by this author
  • Zaid Afawi,

    1. Department of Neurology, Tel Aviv Sourasky Medical Centre, Tel Aviv, Israel
    Search for more papers by this author
  • Eva Andermann,

    1. Department of Neurology and Neurosurgery, Montreal Neurological Institute and Hospital, McGill University, Montreal, Canada
    Search for more papers by this author
  • Daniel Keene,

    1. Department of Pediatrics, Children’s Hospital of Eastern Ontario, Ottawa, Ontario, Canada
    Search for more papers by this author
  • Olivier Dulac,

    1. Service de Neurologie Pédiatrique, AP-HP, Hopital Necker-Enfants Malades, Centre de Référence Épilepsies Rares, Paris, France
    Search for more papers by this author
  • Federico Zara,

    1. Muscular and Neurodegenerative Disease Unit, Institute G. Gaslini, University of Genova, Genova, Italy
    Search for more papers by this author
  • Samuel F. Berkovic,

    1. Department of Medicine (Neurology), University of Melbourne and Austin Health, Heidelberg, Australia
    Search for more papers by this author
  • Renzo Guerrini,

    1. Child Neurology Unit, Children’s Hospital A. Meyer, University of Florence, Florence, Italy
    Search for more papers by this author
  • John C. Mulley

    1. Department of Genetic Medicine, Women’s and Children’s Hospital, Adelaide, Australia
    2. School of Pediatrics and Reproductive Health, University of Adelaide, Adelaide, Australia
    3. School of Molecular and Biomedical Sciences, University of Adelaide, Adelaide, Australia
    Search for more papers by this author

  • 1

    These authors are considered to have contributed equally to this work.

Address correspondence to Renzo Guerrini, Child Neurology Unit, Pediatric Hospital A. Meyer-University of Firenze, Viale Pieraccini 24, 50139, Florence, Italy or John Mulley, Department of Genetic Medicine, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia. E-mail: r.guerrini@meyer.it or john.mulley@cywhs.sa.gov.au

Summary

Objective: We aimed to determine the type, frequency, and size of microchromosomal copy number variations (CNVs) affecting the neuronal sodium channel α 1 subunit gene (SCN1A) in Dravet syndrome (DS), other epileptic encephalopathies, and generalized epilepsy with febrile seizures plus (GEFS+).

Methods: Multiplex ligation-dependent probe amplification (MLPA) was applied to detect SCN1A CNVs among 289 cases (126 DS, 97 GEFS+, and 66 with other phenotypes). CNVs extending beyond SCN1A were further characterized by comparative genome hybridization (array CGH).

Results: Novel SCN1A CNVs were found in 12.5% of DS patients where sequence-based mutations had been excluded. We identified the first partial SCN1A duplications in two siblings with typical DS and in a patient with early-onset symptomatic generalized epilepsy. In addition, a patient with DS had a partial SCN1A amplification of 5–6 copies. The remaining CNVs abnormalities were four partial and nine whole SCN1A deletions involving contiguous genes. Two CNVs (a partial SCN1A deletion and a duplication) were inherited from a parent, in whom there was mosaicism. Array CGH showed intragenic deletions of 90 kb and larger, with the largest of 9.3 Mb deleting 49 contiguous genes and extending beyond SCN1A.

Discussion: Duplication and amplification involving SCN1A are now added to molecular mechanisms of DS patients. Our findings showed that 12.5% of DS patients who are mutation negative have MLPA-detected SCN1A CNVs with an overall frequency of about 2–3%. MLPA is the established second-line testing strategy to reliably detect all CNVs of SCN1A from the megabase range down to one exon. Large CNVs extending outside SCN1A and involving contiguous genes can be precisely characterized by array CGH.

Introduction

Dravet syndrome (DS), otherwise known as severe myoclonic epilepsy of infancy (SMEI) [MIM 607208], is an epileptic encephalopathy presenting in the first year of life (Dravet et al., 2005). The discovery of mutations in the neuronal sodium channel α1 subunit gene (SCN1A) on chromosome 2q in DS (Claes et al., 2001) prompted deeper molecular analysis of this severe epilepsy syndrome. The frequency of SCN1A mutations in DS is approximately 70–80% (Mulley et al., 2005; Harkin et al., 2007; Marini et al., 2007). Most mutations are de novo, but familial SCN1A mutations occur, and relatives have generalized epilepsy with febrile seizures plus (GEFS+) phenotypes (Scheffer & Berkovic, 1997; Singh et al., 1999). SCN1A germline mosaic mutations are on record (Depienne et al., 2006; Gennaro et al., 2006; Marini et al., 2006), and they need to be incorporated into recurrence risks in the context of genetic counseling.

Based on the number of patients in whom mutations have been identified, SCN1A is currently the most clinically relevant epilepsy gene. It is predominantly associated with DS, although about 10% of GEFS+ (Escayg et al., 2001; Wallace et al., 2001; Marini et al., 2007) and some other rare epileptic encephalopathies (Harkin et al. 2007) are also associated with SCN1A mutations. Despite intensive investigation, the etiology of about 20–30% of DS remains unknown.

Deletions or duplications of one or more exons of a causative gene are not detectable by standard polymerase chain reaction (PCR)–based mutation screening. Multiplex ligation-dependent probe amplification (MLPA) is a diagnostic approach for the quantitative detection of deletions, duplications, and other copy number variants (CNVs) affecting one or more exons of a target gene (Schouten et al., 2002). Previously MLPA was applied to DS patients who did not have SCN1A coding mutations (Mulley et al., 2006; Marini et al., 2007; Wang et al., 2008). These independent studies found that 10% (7 of 69) (Wang et al., 2008), 15% (2 of 13) (Mulley et al., 2006), and 11% (2 of 18) (Marini et al., 2007) patients had microchromosomal alterations ranging in size from a single exon to deletions extending beyond SCN1A. These findings were complemented by other studies using different technologies and showing similar small chromosomal rearrangements (Madia et al., 2006; Suls et al., 2006). These studies followed an earlier report of a larger chromosomal anomaly incorporating SCN1A (Pereira et al., 2004).

Through a collaborative international study we assembled a large DS cohort that had tested negative for SCN1A coding mutations. These were analyzed by MLPA to establish an estimate for frequency of microchromosomal alterations involving SCN1A. Analysis was extended to other severe epilepsies with infantile onset and to GEFS+. Microchromosomal deletions extending beyond SCN1A as inferred by MLPA were further characterized by comparative genome hybridization (CGH) using a commercially available chromosome 2 specific array as the probe to determine the size of the abnormality and the identities of any additional genes involved.

Subjects and Methods

Patients

Patients with DS who tested negative for SCN1A sequence-based mutations were collected from Italy, France, and internationally through an Australian study. We collectively defined SMEI and its borderline phenotypes (SMEB: severe myoclonic epilepsy borderline) as DS. DS is an epileptic encephalopathy with seizures beginning before one year of age in a previously normal baby. Polymorphic seizure types evolve and include febrile and afebrile hemiclonic and generalized seizures, which are often prolonged, atypical absences, focal, and myoclonic seizures (Engel, 2001; Dravet et al., 2005). This conceptualizes SMEI and various subgroups of SMEB as different shades of the same disorder, DS, and is consistent with similar mutational distributions observed in these subgroups (Harkin et al., 2007). Written informed consent was obtained from parents or legal guardians in all cases.

Italy

This was an internationally derived cohort comprising 74 DS patients from child neurologists and epileptologists throughout Italy, Spain, and the United Kingdom. Part of this cohort has already been included in previous studies (Nabbout et al., 2003), where preliminary MLPA findings were also reported (Marini et al., 2007). Three GEFS+ patients were also included.

France

This cohort comprised 29 DS patients previously described and published as part of a joint Italian and French study (Nabbout et al., 2003).

Australia

This internationally derived cohort from the Infantile Epileptic Encephalopathy Referral Consortium was previously analyzed for sequence-based mutations (Harkin et al., 2007). Preliminary outcomes from the initial application of MLPA within this cohort were reported previously (Mulley et al., 2006). The cohort included 23 DS patients including two affected sisters (counted as a single case) from a nonconsanguineous Israeli family of Ashkenazi Jewish (Hungarian/Polish) descent. There were also 20 symptomatic generalized epilepsy (SGE), 12 Lennox-Gastaut syndrome (LGS), 5 West syndrome (WS), 12 myoclonic-astatic epilepsy (MAE), 7 symptomatic partial epilepsy (SPE), and 10 cryptogenic partial epilepsy (CPE) cases. We also included a cohort of 94 new Australian GEFS+ patients; none had been sequenced for SCN1A mutations.

Multiplex ligation-dependent probe amplification assay

The MLPA SCN1A kit (SALSA P137; MRC-Holland, Amsterdam, The Netherlands) contains 25 paired probes from the SCN1A region (covering 25 of 26 exons—with exon 9 untested due to technical issues with probe design) and 14 control probes to detect sequences located in other chromosomal regions. Deletion/duplication screening was performed according to the manufacturer’s protocol. Patients from Italy were tested in Italy as described previously (Marini et al., 2007). The French and Australian-derived cases were tested in Australia as previously described (Mulley et al., 2006; Harkin et al., 2007). A benign deletion variant at low frequency was detected for the 6p25 control probe, and was ignored.

For MLPA analysis any putative single exon “deletion” was routinely sequenced in order to exclude allele dropout caused by mismatch interference to probe binding sites, which mimics deletion. Confirmation of deletion, duplication, or amplification was shown by repeatability in a second MLPA reaction, and where more than one SCN1A exon was involved, by observation that all deleted exons were contiguous. Minimal scatter from plots generated by GeneMarker software (SoftGenetics, State College, PA, U.S.A.) of MLPA probe reactions confirmed adequate DNA purity.

Bioinformatics

The SCN1A full-length isoform was taken from Genbank AB093548 for the purpose of assessing whether SCN1A intragenic deletions were in or out of frame.

Array CGH

Hybridizations were carried out by Nimblegen Systems Inc. (http://www.nimblegen.com) using their chromosome 2 specific oligonucleotide array. Median spacing along the chromosome for probes on the array was 575 bp and probe sizes on the chip varied from 50-mers to 75-mers (Roche Nimblegen, Madison, WI, U.S.A.).

Results

A total of 289 patients with a clinical diagnosis of DS (N = 126), GEFS+ (N = 97). and other severe epileptic encephalopathies (N = 66) were analyzed by MLPA. Table 1 summarizes the total number of patients included in the study and their epilepsy syndrome diagnosis. Table 2 summarizes clinical features of the 16 patients, 15 DS and one SGE, with positive MLPA findings and the results of further characterization of the extragenic SCN1A abnormalities by array CGH (Fig. S1). Fig. 1A shows the breakpoints within SCN1A as detected by MLPA and Fig. 1B represents all of the deletions extending beyond SCN1A as initially detected by MLPA and further characterized by array CGH. Most breakpoints were within or close to SCN1A, but at the sequence level there does not appear to be any breakage hotspot.

Table 1.   Summary of the epilepsy phenotypes of all patients included in the study and of the MLPA detected SCN1A deletions, duplications, and amplification
DisorderPositive on MLPANo. of patients examined
  1. MLPA, Multiplex ligation-dependent probe amplification.

  2. aPatients with partial and whole gene deletions, including two siblings (counted as one case), patient with amplification and patient with duplication.

  3. bPatient with duplication.

Dravet syndrome15a126
Lennox-Gastaut syndrome0 12
Symptomatic generalized epilepsy1b 20
West syndrome0 5
Myoclonic astatic epilepsy012
Symptomatic partial epilepsy0 7
Cryptogenic partial epilepsy010
Generalized epilepsy with febrile seizures plus097
Total16289
Table 2.   Clinical features of patients with molecular characterization of deletions/duplications/amplification by MLPA and array CGH
Patient ID/GenderAgea1st sz, type, durationSz type during follow-upDevelopment and outcomeMLPAArray CGH
  1. Ab, absences; Amp, amplification; AT, atonic; Bz, benzodiazepine; CLB, clobazam; CL, clonic; CP, complex partial; Del, deletion; dup, duplication; ex, exon; F, female, Feb, febrile; FS, febrile seizures, GTCS, generalized tonic–clonic seizures; M, male; MR, mental retardation; mo, months; My, myoclonic; ND, not detected; NT, not tested; PB, phenobarbital; PHT, phenytoin; SE, status epilepticus; sz, seizures; STP, stiripentol; T, tonic; TC, tonic–clonic; TPM, topiramate; VPA, valproic acid.

  2. Prime symbol denotes minutes.

  3. aAt the time of the study, years.

  4. bSiblings.

  5. cPatient with symptomatic generalized epilepsy.

1/F4.26 mo, FS, 20′FS, TC, My, CPModerate MR, drug resistant szDel ex 3NT
2/M10.69 mo, Feb SE, 60′FS, Feb SE, TC, My, CPModerate–severe MR, good response to VPA + TPMDel ex 2 – 6NT
3/F (Marini et al., 2007)11.34.5 mo, hemiCL, 20′HemiCL, Feb SE, TC, Ab, My, CPModerate–severe MR, drug resistant sz, short statureDel ex 12-14NT
4/M9.210 mo, FS, <5′CL, FS, TC, SE, CPModerate MR, good response to PHTDel ex 1-13∼186kb up to 3′ end of SCN9A
5/F (Marini et al., 2007)10.34 mo, FS, 30′FS, TC, Ab, My, CPModerate–severe MR, drug resistant szDel SCN1A∼4.87Mb SCN1A-Myo3B
6/F14.46 mo, Feb hemiCL, 30′HemiCL; TC; My; Ab; SLI+Moderate–severe MR; ataxia; drug resistant szDel SCN1A588kb GALNT3
(5′end)-SCN9A
7/M46 mo, FS, 5′FS, Feb SE, TC, Ab, My, CPMild MR; partial response to TPM + PB + Bz, mild facial dysmorphic featuresDel SCN1A6Mb TTC21B to SLC25A12 (3′end)
8/M2.93 mo, Feb SE, 50′Feb SE, TC, My, CPEarly severe MR with autistic features, mild facial dysmorphic features, drug resistant sz, extreme photosensentivityDel SCN1A∼9.3Mb GRB14 to ZAK
9/F114 mo, TC, 15′FS, Feb SE, CP, TC, Ab, MyMild MR; drug resistant szDel SCN1A∼1.49Mb FAM130A2 to XIRP2 (5′ end)
10/F104.5 mo, FS, 10′FS, Feb SE, My, CP, AbsSevere MR, partial response to STP + VPA + CLBDel ex 7-2684kb, only SCN1A
11/M (Mulley et al., 2006)77.5 mo, Feb HemiCL, 7–8′FS, TC, SE, My, AT, HemiCL, T, focalModerate MR; ataxia, drug resistant szDel ex21∼6.5kb
12/M (Mulley et al., 2006)84 mo, HemiCL, 2′FS, TC, CP, My, HemiCL, Ab, T, SEMild MR; autistic featuresDel ex 21-26∼90kb up to TTC21B ex 1-7
13b
 A/F1.73 mo, Feb TC, 10′My, TC, AT, Ab, focalModerate MR; died in sleepDel ex 1-22∼1.64Mb SCN1A to XIRP2
 B/F35.5 mo, hypotonia, 10′TC, SE, FS, AT, AbModerate MR  
14/F306 mo, TCMy, AT, TC, SE, AbMild MRAmp ex 26ND
15/Fc263 mo, brief FSTonic, hemiCL, GTCS, My, focal, SEProfound MR, microcephaly, generalized spasticityDup ex 26ND
16b
 A/F143 moGTCSMild MR, kyphoscoliosis, mild ataxia, pyramidal signsDup ex 8-16ND
 B/M44 moHemiCLSevere MR, clumsy gait, mild pyramidal signs, hyperactivity, aggressive behavior ND
Figure 1.


Microchromosomal abnormalities characterized by multiplex ligation-dependent probe amplification (MLPA) and array comparative gene hybridization (CGH). (A) Showing SCN1A intragenic single or multiple exon deletions in patients 1, 2, 3, 10, and 11; exon amplification in patient 14, and exon duplications in patients 15 and 16 (two siblings) identified by MLPA. (B) Showing gene content of deletions extending beyond SCN1A in patients 5–9 and 13 (two sisters) as further characterized by array CGH.

Dravet syndrome

MLPA analysis showed SCN1A abnormalities in 16 probands comprising one duplication (identified in two siblings and their mother), one amplification, and 13 different deletions of one or more exons (Table 2).

DS with amplification/duplication

MLPA of patient 14 (Table 2, Fig. 1A) showed a signal displacement consistent with an estimated total copy number of 5–6. This amplification involved only exon 26 and extended for an unknown distance beyond the 3′ end of SCN1A. The amplification was not detectable in the array profile, even with examination of the region of exon 26 at high resolution.

An intragenic duplication of the genomic segment spanning exons 8–16 of SCN1A was identified in two siblings (Table 1, patients 16 A and B) with typical DS. MLPA analysis in the parents indicated that the duplication was present in the mother, who had had simple febrile seizures during the first year of life. The relative ratio of averaged peak areas for SCN1A exons 8–16 in the mother comprised between 1.0 (normal) and 1.4 (duplication), with an averaged relative ratio of 1.28. These data indicate that the mother is, at least in the blood, a mosaic carrier of the duplication. By assuming that the averaged relative ratio values of 1.0 and 1.4 represent 0% and 100% of mutant cells, respectively, and a linear relationship between averaged relative ratio values and proportion of mutant cells, the mother shows that 65% of cells carry the mutation.

DS with deletions

Four of the 13 deletion cases: exon 12–14 deletion (patient 3, Table 2), whole gene deletion (patient 5, Table 2), exon 21 (patient 11, Table 2), and exon 21–26 deletion (patient 12, Table 2) were previously reported (Mulley et al., 2006; Marini et al., 2007). Deletions in patients 5 and 12 were confirmed and further characterized by array CGH (Table 2, Fig. 1B); however, in the case of patient 12 this relatively small deletion was not visually evident on the array profile until directed to the specific site for a closer look based on the MLPA result. The exon 21 deletion in patient 11 was similarly not visually detectable from inspection of the chromosome 2 array profile, but could be seen on the array profile by specifically looking at the exon 21 signal. Size estimated from the array profile was ∼6.5 kb, close to the exact size of 6,499 bp previously determined by sequencing the junction fragment (Mulley et al., 2006).

For the additional cases, deletion of exon 3 was identified in one patient (patient 1, Table 2, Fig. 1A). Three deletions (patients 2, 4, and 10, Table 2, Fig. 1B) removed multiple contiguous SCN1A exons, two of which extended an unknown distance beyond SCN1A in the 5′ direction (patient 4, 1–13 exons deleted) and in the 3′ direction (patient 10, 7–26 exons deleted). Patient 4 was confirmed from the array profile, and the size of the deletion was characterized (Table 2, Fig. 1B). Patient 10 (Table 2) was found to have a small deletion excising SCN1A (Fig. 1A), but no adjacent genes. It was not detected on the array profile until directed to the SCN1A site on the basis of the MLPA result.

Five patients (patients 5–9, Table 2) had the entire SCN1A gene deleted, with the deletion extending unknown distances beyond both the 3′ and 5′ ends of SCN1A (Fig. 1B). Parental testing was carried out in four patients where DNA was available from both parents, and confirmed de novo origin of these deletions (patients 3, 6, 7, and 8, Table 2). Deletions in all four patients were confirmed from the array profile and their extent characterized.

An exon 1–22 deletion was detected in two affected sisters (patients 13A and 13B, Table 2) but not in either parent. The presence of two affected sisters from unaffected and nonconsanguineous parents infers the presence of gonadal mosaicism in one of the parents. Additional microsatellite markers genotyped from the family confirmed the accuracy of the pedigree. The deletion was also detected on the CGH profile and its extent determined (Table 2, Fig. 1B).

Early onset epileptic encephalopathy with duplications

Of the 66 patients with early onset epileptic encephalopathies, one patient had a duplication of exon 26 (patient 15, Table 1, Fig. 1A) extending for an unknown distance in the 3′ direction; this abnormality was also not detected on the array profile. The patient was a disabled 26-year-old woman born to consanguineous Lebanese parents. She had her first febrile seizure at 3 months and a left hemiparesis was noted. At 5 months, she presented with a cluster of afebrile generalized seizures and at 6 months, she developed frequent myoclonic jerks. She had frequent febrile and afebrile generalized, hemiclonic, and tonic seizures. At 17 years, she had daily tonic, focal, and generalized myoclonic seizures, and monthly left hemiclonic secondarily generalized seizures. At 19 years, she experienced clusters of tonic–clonic seizures every few minutes and had several admissions for status epilepticus over the next few years. Electroencephalography (EEG) studies were normal in the first years of life but by 18–26 years showed 2-Hz bifrontal or diffuse sharp–slow wave activity in addition to polyspike wave, multifocal discharges, and diffuse background slowing. Computerized tomography (CT) brain scan at 14 months showed generalized cerebral atrophy. Her development was always delayed and regression did not occur. At 4 years, she was not interactive and did not make eye contact. By adult life, she smiled and vocalized but had no words, and could sit but never walked. She had a marked kyphoscoliosis with generalized spasticity and contractures and her head circumference was 52 cm (<2nd percentile).

Generalized epilepsy with febrile seizures plus (GEFS+)

No copy number abnormalities were detected in the 97 GEFS+ cases.

Combining the Italian-, French-, and Australian-derived cases with DS, diagnostic closure was achieved for 14 of the previous 125 unsolved cases and for one of the 20 SGE cases (Table 1). Because the Australian- (N = 23) and Italian- (N = 73) derived cases of DS were previously tested for sequence-based mutations with near absolute sensitivity, and accounted for 12 of the 14 DS microchromosomal abnormalities, the best estimate for the frequency of microchromosomal abnormalities in those DS patients testing negative by sequencing is 12 of 96 (12.5%). The two siblings with typical DS and their mother were not included in the estimate of the frequency as they were not part of the consecutive DS cases previously tested. Relatively small deletions in three patients ranging in size from 6.5–90 kb were not detectable from the CGH profile until directed to the exact site of the abnormality using prior knowledge of the MLPA result. Deletions ranging in size from 186 kb to 9.3 Mb were clearly visible on the high-resolution chromosome 2 CGH array profile (Fig. S1). Hence, the limit of definitive detection by the CGH array used in these experiments, without MLPA, lies in the region of CNVs between 90 kb and 186 kb.

Discussion

Several molecular approaches have demonstrated small chromosomal abnormalities affecting SCN1A in DS. Deletions of the SCN1A gene, sometimes encompassing as many as 20 additional genes, have been found using haplotype analysis with microsatellite markers and single-nucleotide polymorphisms (SNPs) and multiplex amplicon quantification (MAQ) (Madia et al., 2006; Suls et al., 2006). These non-MLPA approaches are less sensitive for detection of small deletions/duplications, and might underestimate the overall SCN1A deletion/duplication frequency.

This study obtained a frequency of 12.5% MLPA-detected chromosomal rearrangements involving deletions and an amplification among otherwise mutation-negative DS patients. The percentage from our cohort is higher than that from Suls et al. (2006) (2.7%) and Madia et al. (2006) (7.7%), perhaps due to the fact that MLPA detects all deletions and duplications down to single exons (except for the SCN1A exon 9 for which there is no reliable MLPA probe). The frequency of SCN1A MLPA-detected anomalies obtained in this study (12.5%) is slightly higher than that of Wang et al. (2008) (10%), possibly due to mild differences in phenotyping or to the larger number of patients analyzed.

We report the first amplification and two duplications involving SCN1A, which are considerably rarer and smaller than deletions. The amplification and one duplication were observed in three patients with a typical history of DS. The duplication was present in two siblings and it was inherited from a mildly affected mother with only febrile seizures. Her milder phenotype might be partly explained by the fact that the duplication was mosaic and present in only about 65% of ectodermal derived cells.

An additional duplication was also identified in a woman with SGE with seizure onset in the first year and a predilection to seizures with fever. She had prominent tonic seizures, abnormal early development without regression, microcephaly, profound intellectual disability, and generalized spasticity.

Patients with chromosomal abnormalities involving SCN1A are also subject to gonadal mosaicism, and may recur within a sibship. Therefore, the recurrence risk for DS associated with a SCN1A mutation, needs to take into account the possibility of recurrence caused by gonadal mosaicism.

Microchromosomal rearrangements extending beyond SCN1A and including variable numbers of contiguous genes are potentially associated with additional dysmorphic features, depending on the genes involved. Suls et al. (2006) showed that two such patients had additional clinical features but a third patient had no additional features. In our cohort, seven patients (4–9 and 13; Table 2) had deletions removing SCN1A and contiguous genes including SCN3A, SCN2A, SCN9A, and SCN7A. Of these, SCN2A and SCN3A genes have been associated with human epilepsy (Sugawara et al., 2001; Heron et al., 2002; Kamiya et al., 2004; Holland et al., 2008). SCN9A—associated with three inherited disorders: primary erythermalgia (Yang et al., 2004), familial pain (Fertleman et al., 2006), and congenital insensitivity to pain (Cox et al., 2006)—was removed in all seven of our patients, none had clinical symptoms suggestive of erythermalgia, which is consistent with the gain of function induced by missense mutations in that disorder (Yang et al., 2004; Fertleman et al., 2006). Patient 8 (Table 2) had the largest deletion, removing all sodium channel genes and 49 contiguous genes (Fig. 1B); his epilepsy phenotype was not distinct from DS and he did not have additional dysmorphic features. Overall, none of our DS patients had obvious dysmorphic features and there did not appear to be significant clinical differences between patients in whom the deletions involved only SCN1A and those in whom contiguous genes were also removed. Our findings suggest that the other deleted genes do not exhibit haploinsufficiency and caused no additional symptoms in this cohort.

SCN1A mutations appear to be randomly distributed throughout the gene (Harkin et al. 2007; Marini et al., 2007). The MLPA-detectable breakpoints that occur within SCN1A (Fig. 1A) fail to show any pattern suggestive of sequence related chromosomal breakpoint hotspots within SCN1A. Moreover, there does not appear to be breakage propensity limited to any specific region outside of SCN1A when contiguous genes are excised, in addition to SCN1A.

We looked at the consequences of the deletions within SCN1A affecting the reading frame. Those cases with internal deletions were examined for disruption of the reading frame in the coding DNA sequence downstream of the deletion, to assess their potential effects on severity. Four of the 13 deletion cases (31%) have deletions internal to SCN1A (Table 2, Fig. 1A). Only one of these (exons 2–6) alters the downstream reading frame. The remaining internal deletions do not disrupt the reading frame (exons 3, 12–14, and 21), but these patients are affected in a manner similar to that of DS patients with truncating mutations and the patient with the exon 2–6 deletion where the reading frame was disrupted.

What causes DS when there is no detectable SCN1A involvement? This question remains unanswered. We reported a single case with a mutation in the γ2 subunit gene of the γ-aminobutyric acid (GABA)A receptor, GABRG2 (Harkin et al., 2002). However, in French (Nabbout et al., 2003) and Australian (Mulley et al., 2006) cohorts of DS, GABRG2 mutations have not been detected (unpublished results). The DS so far unexplained might have polygenic etiologies, which will pose the same challenges as for any of the epilepsies with complex genetic inheritance.

We have provided a robust estimate of SCN1A deletion/duplication frequency in DS cases without detectable sequence-based mutations. Moreover, this has reinforced MLPA as a rapid highly sensitive and relatively economical technology. This extension to the diagnostic repertoire increases the overall detection rate of mutations in DS to approximately 12.5% of those patients that do not have sequence-based mutations. Array CGH has the sensitivity to precisely characterize any deletions extending beyond SCN1A.

Accepting the frequency of sequencing detectable mutations in DS, around 70–80% (Mulley et al., 2005; Harkin et al. 2007; Marini et al., 2007), and using our estimate of 12.5% frequency of microchromosomal abnormalities in the residual, the overall frequency of microchromosomal abnormalities is 2–3%. Therefore, SCN1A sequence-based mutations are a priori far more likely and should be searched for first. The MLPA approach is far less likely to show positive findings and should be used as the second-line test. Nevertheless, for those individuals diagnosed with SCN1A alterations through MLPA, this diagnostic test is invaluable as it provides a definitive diagnosis, removes the need for ongoing testing, and brings closure for the family. Array CGH is useful in refining MLPA results to determine the extent of a detected deletion beyond SCN1A but, at present, MLPA is the method of choice for detection of small microchromosomal anomalies.

Acknowledgment

We thank the patients and their families for their participation in this study. The Italian series was supported by the Genetic Commission of the Italian League Against Epilepsy; the French series was supported by Genethon and ARGE sponsored by Sanofi; and the Australian series was supported by the National Health and Medical Research Council of Australia, Thyne-Reid Charitable Trusts, and the MS McLeod Research Fund. We thank the Infantile Epileptic Encephalopathy Referral Consortium as acknowledged by Harkin et al., 2007 for referral of the infantile encephalopathy patients.

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Disclosure: None of the authors has a financial interest in this publication.

Ancillary