Abstract
Background: Inherited and de novo mutations in sodium channel genes underlie a variety of channelopathies. Mutations in SCN2A, encoding the brain sodium channel NaV1.2, have previously been reported to be associated with benign familial neonatal infantile seizures, febrile seizures plus, and intractable epilepsy of infancy.
Methods: We evaluated the clinical characteristics in a patient with a neonatal-onset complex episodic neurologic phenotype. We screened SCN2A for mutations and carried out in vitro electrophysiologic analyses to study the consequences of the identified mutation. We studied the developmental expression of NaV1.2 in cerebellum by immunohistochemical analysis.
Results: The patient presented with neonatal-onset seizures and variable episodes of ataxia, myoclonia, headache, and back pain after 18 months of age. The patient carries a de novo missense mutation (p.Ala263Val) in SCN2A, which leads to a pronounced gain-of-function, in particular an increased persistent Na+ current. Immunohistochemical studies suggest a developmentally increasing expression of NaV1.2 in granule cell axons projecting to Purkinje neurons.
Conclusions: These results can explain a neuronal hyperexcitability resulting in seizures and other episodic symptoms extending the spectrum of SCN2A-associated phenotypes. The developmentally increasing expression of NaV1.2 in cerebellum may be responsible for the later onset of episodic ataxia.
Sodium channel defects underlie different epileptic phenotypes ranging from simple febrile seizures to severe epileptic encephalopathies. SCN2A encodes NaV1.2, one of 4 sodium channels expressed in mammalian brain. The most common phenotype caused by SCN2A mutations is benign familial neonatal-infantile seizures (BFNIS),1 but febrile seizures plus and intractable epilepsy of infancy also have been associated with this gene.1,–,4 BFNIS-associated SCN2A missense mutations have revealed mainly gain-of-function defects predicting an increase in neuronal excitability.3,5 A truncating SCN2A mutation and missense mutations showing strong, mainly loss-of-function biophysical defects have been described in severe early-onset epilepsies.3,4 We describe a de novo missense mutation revealing a pronounced NaV1.2 gain-of-function that expands SCN2A-associated phenotypes to more complex neurologic symptoms.
METHODS
The patient was examined in Helsinki and recruited for a genetic study of SCN2A in SCN1A mutation-negative patients with neonatal-infantile-onset intractable epilepsy. All SCN2A exons were sequenced from PCR-amplified genomic DNA of the patient (table e-1 on the Neurology® Web site at www.neurology.org). Exon 7 was sequenced in the parents and 93 Finnish controls. Parental allele transmission was confirmed by microsatellite markers.
The c.788C>T mutation was introduced into cDNAs of neonatal and adult NaV1.2 human splice variants, coexpressed in tsA201 cells with β1- and β2-subunits and studied using the whole-cell patch-clamp technique. Data are shown as means ± SEM; Student t-test was applied for statistical analysis. Immunohistochemical staining of unfixed cerebellar mouse brain slices was performed using a monoclonal antibody against NaV1.2 channels. For experimental procedures, see e-Methods.
Standard protocol approvals, registrations, and patient consents.
An Institutional Review Board of the Helsinki University Central Hospital approved the study. Written informed consent and signed patient consent-to-disclose form for the video were obtained from parents. Local authorities (Regierungspraesidium Tübingen, Germany) approved the animal experiments.
RESULTS
The patient, born in 1999, presented with hypomotor followed by tonic-clonic seizures on alternating sides with contralateral ictal EEG discharges from the first day of life. Family history is positive for migraine with visual aura (figure 1A). Extensive investigations including metabolic workup, interictal EEG, and MRI were normal. Seizures were controlled by high levels of phenobarbital and phenytoin, but recurred upon viral infections. Hypomotor, focal and bilateral motor seizures continued weekly to monthly until 1.3 years. A tonic-clonic seizure after a minor head trauma occurred at 3.5 years during acetazolamide and at 6.5 years during topiramate monotherapy. He is currently seizure-free on acetazolamide. Recent neuropsychological testing showed normal intelligence but specific problems in visual processing, fine motor function, and tactile sensation.
(A) Pedigree of the family. The proband is indicated with an arrow. Individuals with migraine with visual aura are marked with a gray upper left quarter. Unspecific headache is marked with a gray lower right quarter. (B) Chromatograms of the patient and the unaffected parents showing a heterozygous c.788C>T alteration in the patient. (C) Structure of the human NaV1.2 channel with localization of the p.Ala263Val mutation (star) in the S5 segment of domain I. (D) Ala263 (boxed) and the surrounding amino acids show evolutionary conservation.
Episodes with poor balance, ataxia, slurred speech, intermittent myoclonic jerks, and severe distress with headache, back pain, hypermotor activity, hyperventilation, and retching or vomiting started at 1.5 years. During the symptoms, he is fully conscious, has normal muscle strength and reflexes, and has no tonic-clonic activity. The episodes occur 1 to 3 times per month, lasting minutes to several hours. Mild motor dyspraxia but no ataxia is present between the episodes. A video-EEG including several episodes at 4.5 years showed occasional (poly)spikes frontocentrally, but no ictal epileptic discharges. MRI showed mild cerebellar atrophy at 5 years. Phenytoin was started at 1 month and withdrawn at 2.8 years. Phenytoin levels varied between 40 and 100 μmol/L with occasional higher levels (maximum 114 μmol/L at 6 months). Valproate, acetazolamide, oxcarbazepine, gabapentin, clobazam, levetiracetam, and topiramate as well as rescue medications (diazepam, anti-inflammatory drugs, and sumatriptan) have been ineffective. Characteristic symptoms are shown in the supplementary video.
A novel heterozygous SCN2A mutation, c.788C>T, predicting the amino acid exchange p.Ala263Val in the highly conserved transmembrane segment D1/S5 (figure 1, B and C) was identified in the patient. It was absent in the parents and in 93 controls.
Electrophysiologic analysis revealed profound changes for mutant channels using both neonatal and adult NaV1.2 splice variants. Most prominent was a 3-fold increase in persistent Na+ current (figure 2, A–D). Shifts and slope changes of steady-state activation and fast inactivation curves led to an increase in window current. Fast inactivation was significantly slowed, whereas recovery was not changed (figure 2, E–G). Recovery from slow inactivation was accelerated only for the neonatal variant (figure 2H), while slow inactivation was not changed otherwise (figure e-1). Current density was not significantly changed (see table e-2 for all electrophysiologic results). We thus found prominent gain-of-function changes predicting an increase of membrane excitability in neurons expressing mutant NaV1.2 channels.
(A, B) Families of whole-cell Na+ currents recorded from tsA201 cells transfected with either WT (A) or A263V mutant (B) channels, both in the background of the neonatal splice variant. Na+ currents were elicited by step depolarizations ranging from −105 to +67.5 mV from a holding potential of −140 mV. (C) Representative WT and A263V persistent Na+ currents showing a largely increased persistent current for the mutation in the background of both the neonatal and the adult splice variants. Current amplitudes were recorded at the end of a 70-msec depolarization to 0 mV and are normalized to the peak amplitude (ISS/IPEAK). Upon application of the specific Na+ channel blocker TTX (20 nM), the persistent current was completely abolished, as shown in the inset. (D) Voltage dependence of the persistent currents. (E) Voltage dependence of steady-state Na+ channel activation and fast inactivation revealing an increase in window current (area under the overlap of both curves) for A263V channels compared to the WT. Lines represent fits of Boltzmann functions. (F) Voltage dependence of the fast inactivation time constant, τh, for WT and A263V mutant channels revealing a slowing of fast inactivation for the mutant channels. (G) Voltage dependence of the time constant of recovery from fast inactivation, τrec, which was not significantly different between WT and mutant channels. (H) Time course of recovery from slow inactivation determined at −120 mV revealing an acceleration for A263V mutant channels only in the background of the neonatal splice variant. Lines represent fits of exponential functions. The values of the electrophysiologic results, numbers of experiments, and p values are given in table e-2. All values are shown as means ± SEM.
Immunohistochemical stainings of mouse cerebellum with a NaV1.2-specific antibody revealed a prominent increase in staining of the molecular layer during postnatal development. Since unmyelinated parallel fibers of the granule cells are densely packed in the molecular layer, an NaV1.2 expression increasing with maturation could be the correlate of the observed developmental change in staining pattern (figure 3).
Immunohistochemical stainings of cerebellar mouse brain slices at different stages of development (postnatal days P5, P8, P14, P20). The signal stemming from a monoclonal antibody directed against the C-terminus of NaV1.2 channels (shown in red) is increasing in the molecular cell layer (m) with maturation. This suggests that the expression of these channels is upregulated during development in the unmyelinated parallel fibers of granule cells, which are densely packed in the molecular layer and project to Purkinje neurons. Nuclei are stained with DAPI (blue fluorescence). The granule cell layer (g) is divided into an internal (gi) and external (ge) part early in development. p = Purkinje cell layer.
DISCUSSION
The neonatal-onset seizures in our patient carrying the de novo SCN2A mutation persisted unusually long for BFNIS and were difficult to treat. Moreover, the described late-onset episodic symptoms (ataxia/myoclonia/pain) further broaden the spectrum of SCN2A-associated phenotypes. Neuromyotonia, rolandic epilepsy, and epileptic encephalopathy have previously been described in KCNQ2-associated benign familial neonatal seizures6 and patients with benign familial infantile seizures may present with paroxysmal dyskinesia or migraine.2 “Benign” neonatal-infantile epilepsies are thus not always benign and can be accompanied by other neurologic symptoms.
The p.Ala263Val mutation affects a highly conserved amino acid in NaV1.2, was not found in controls, and has prominent gain-of-function consequences predicting neuronal hyperexcitability. We recently described similar, but less pronounced changes for 2 other mutations in the same channel region in patients with BFNIS.5 We showed that in axon initial segments of hippocampal and cortical principal neurons, NaV1.2 channels are expressed early in development and that NaV1.6 channels replace them partially with increasing maturation, providing an explanation for seizure remission in BFNIS.5 Thus there is ample evidence for p.Ala263Val being responsible for the neonatal-onset seizures in our patient. The more severe epilepsy can be explained by more pronounced biophysical defects compared to BFNIS-associated SCN2A mutations.3,5 It remains unclear why severe loss-of-function mutations of NaV1.2, which should decrease neuronal excitability, can cause epileptic encephalopathy.3,4 Possible reasons comprise different compensatory mechanisms or altered network function by knock-out of one allele, whereas relatively small gain-of-function changes may “only” provoke increased excitability of pyramidal neurons.
Mild cerebellar atrophy on MRI may be due to phenytoin treatment, but does not explain episodic ataxia in our patient. NaV1.2 mRNA is upregulated in rodent cerebellar granule cells from P15 on,7,8 the channels are highly expressed in unmyelinated parallel fibers (the axons of granule cells in the molecular layer),7,9 and NaV1.2 expression in unmyelinated fibers generally increases with maturation.5,7 Combined with these data, our immunohistochemical investigations suggest a strong developmental increase of NaV1.2 expression in granule cell axons of the molecular layer. Thus, a gain-of-function NaV1.2 mutation could explain a hyperexcitability of these fibers projecting to Purkinje neurons thereby changing cerebellar output signals with subsequent ataxia. The differential developmental expression of NaV1.2 channels—early expression developmentally decreasing in axon initial segments of principal neurons and the opposite in unmyelinated axons5,7—may explain why seizures started before ataxia. However, the cerebellar network seems to be less vulnerable to NaV1.2 gain-of-function changes, as ataxia has been associated exclusively with the mutation displaying the most severe defect. Similar to ataxia, episodic pain might be due to the expression of NaV1.2 channels in unmyelinated spinal cord fibers.9 As symptoms observed in our patient were more diffuse than in episodic pain disorders caused by mutations in SCN9A,10 NaV1.2 may only play a minor role in pain fibers. Other genetic factors, such as those increasing susceptibility to migraine in both branches of the family, may also contribute to the episodic symptoms, which are reminiscent of migraine and episodic ataxia type 2.
DISCLOSURE
Dr. Liao and Dr. Anttonen report no disclosures. Dr. Liukkonen has served on a scientific advisory board for Eisai Inc.; has received funding for travel from Nutricia Advanced Medical Nutrition; and has received research support from Eisai Inc. and the Arvo and Lea Ylppö Foundation. Dr. Gaily has received funding for travel from UCB and receives research support from the Arvo and Lea Ylppö Foundation. Dr. Maljevic, S. Schubert, and A. Bellan-Koch report no disclosures. Dr. Petrou serves on the editorial board of the Neurobiology of Disease; is listed as an inventor on patents re: A method for the diagnosis of SMEI, A method of drug design, and Modulation of an ion channel or receptor; has received speaker honoraria from Sepracor Inc.; and receives research support from the NHMRC and the Australian Research Council. Dr. Ahonen reports no disclosures. Dr. Lerche serves on scientific advisory boards for Eisai Inc., GlaxoSmithKline, Pfizer Inc., UCB, and Valeant Pharmaceuticals International; has received funding for travel from GlaxoSmithKline and UCB; receives speaker honoraria from Desitin Pharmaceuticals, GmbH, Eisai Inc., GlaxoSmithKline, Pfizer Inc., and UCB; and receives research support from Sanofi-Aventis, UCB, DFG, BMBF, and the European Union. Dr. Lehesjoki received a speaker honorarium from Pfizer Inc.; receives research support from the Academy of Finland, the European Union, and the Sigrid Jusélius Foundation grants; and has received royalties from Licentia Ltd. for a patent re: Cystatin B Mutants.
ACKNOWLEDGMENT
The authors thank the patient and his parents for participating in this study and Katariina Mattila for technical assistance.
Footnotes
Study funding: Supported by the Folkhälsan Research Foundation, the University of Helsinki, the Academy of Finland (Center of Excellence Programme 2006–2011), the German Research Foundation (DFG Le1030/10-1, /8-2), the National Genome Network of the Federal Ministry for Education and Research (BMBF: NGFNplus/01GS08123), the European Union (EPICURE: LSH 037315), and the University of Ulm.
-
- BFNIS
- benign familial neonatal-infantile seizure
Supplemental data at www.neurology.org
- Received December 23, 2009.
- Accepted June 30, 2010.
- Copyright © 2010 by AAN Enterprises, Inc.
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.