Chromosomal deletions of 2q21-q31 have been known to be closely related to seizures and epilepsies, and the band 2q24 is the smallest commonly deleted segment in these patients (Davidsson et al., 2008). Several authors including us have reported that deletions of the voltage-gated sodium channel (SCN) cluster at 2q24 are associated with several epilepsy syndromes (Takatsuki et al., 2010). The deletions of the locus, including SCN1A, were found in patients with Dravet syndrome (Marini et al., 2009; Suls et al., 2010). The chromosomal deletions of 2q including SCN1A and SCN2A were reported in some patients with severe epilepsy of infantile onset, developmental delay, and dysmorphic features (Pereira et al., 2004; Langer et al., 2006;Pereira et al., 2006). A small deletion between the locus of SCN2A and SCN3A was seen in a patient with infantile seizures, mental retardation, and behavioral and psychiatric abnormalities (Bartnik et al., 2010). Compared to that, the duplications of this region have been identified in only two families (Heron et al., 2010; Raymond et al., 2011). Recently, we identified a chromosomal duplication of 2q24.2q24.3, in which SCN gene cluster is located, in an infant with refractory neonatal epilepsy and severe developmental delay. We report on this patient and discuss the genotype–phenotype correlation.
There are only two reports on epileptic patients associated with microduplication of 2q. We found a de novo duplication of chromosome 2q24.2q24.3 in another infant with neonatal epilepsy. The patient had refractory focal seizures since the third day of life. Her seizures were refractory against phenobarbital and levetiracetam, but were controlled by valproate. Array comparative genomic hybridization revealed a 5.3-Mb duplication of 2q24.2q24.3, where at least 22 genes including a cluster of voltage-gated sodium channel genes (SCN1A, SCN2A, SCN3A, SCN7A, and SCN9A) and one noncoding RNA are located.
A female infant was born spontaneously at 40 weeks of gestation following an uncomplicated pregnancy. She was the first product of unrelated healthy parents. Her birth weight was 2,796 g and head circumference was 33 cm. No dysmorphic features were recognized.
Her mother had noticed mild convulsive movement and staring lasting for 10–20 s since the third day of life. The frequency of paroxysmal events gradually increased and she was admitted to the neonatal intensive care unit of Iwaki Kyoritsu Hospital at 9 days of age. Physical examinations were unremarkable other than for mild hypotonia. Head magnetic resonance imaging (MRI) and blood examination including metabolic screening of amino acids and organic acid analyses revealed no abnormalities. Interictal electroencephalography (EEG) showed markedly abnormal background activities with spiky transients. She was diagnosed as having neonatal seizures and treated with phenobarbital and midazolam. Although her seizures were refractory against these antiepileptic drugs, seizures were transiently controlled by 20 mg/kg of oral phenobarbital after 23 days of age.
Her seizures recurred at 2 months of age and she was referred to Juntendo University Hospital. She had 5–20 seizures every day. Although the dose of phenobarbital was increased to reach the serum level of 61.0 μg/ml, her seizures could not be stopped. Next, levetiracetam was added up to the dose of 40 mg/kg. However, the frequency of seizures did not decrease.
Karyotypic chromosomal analysis revealed a normal female karyotype. Array comparative genomic hybridization (aCGH) analysis was performed at 5 months of age and revealed duplication of 2q24.2q24.3 as mentioned below. This prompted us to change antiepileptic drugs. Phenobarbital and levetiracetam were substituted for valproate, and her seizures were completely controlled by 50 mg/kg of valproate (serum level, 69.1 μg/ml). Seizures provoked by fever had never been recognized, although she had several febrile illnesses. Her developmental milestones were severely delayed. Although social smile and eye following were recognized at 3 and at 4 months of age, respectively, head control or eye-hand coordination was not achieved at the last follow-up at 10 months of age. Marked generalized hypotonia was observed with normal deep tendon reflexes.
To investigate the chromosomal aberration, we performed aCGH analysis using Human Genome CGH Microarray 44A (Agilent Technologies, Palo Alto, CA, U.S.A.) with genomic DNA extracted from the patient’s peripheral blood according to the method described elsewhere (Takatsuki et al., 2010). Metaphase or prometaphase chromosomes were prepared from phytohemagglutinin-stimulated peripheral blood lymphocytes for two-color fluorescence in situ hybridization (FISH) analysis using bacterial artificial chromosome (BAC) clones as probes as described previously (Takatsuki et al., 2010). BAC clones were selected from an in silico library (UCSC Human Genome Browser, March 2006 http://genome.ucsc.edu/).
aCGH analysis identified an aberration in chr2(161,704,227-167,042,361) with average log2 ratio of 0.55, which indicated a 5.3-Mb duplication of 2q24.2q24.3 (Fig. 1). Physical positions referred to NCBI36/hg18. Similar variations were not identified in the Database of Genomic Variants (http://projects.tcag.ca/variation/). Two-color FISH analysis confirmed the duplication in the patient (Fig. 2), and parental FISH analysis showed normal results indicating de novo occurrence of the duplication in the patient. According to the UCSC genome browser, at least 22 genes and one noncoding RNA are included in the duplicated region of this patient, in which there is a cluster of SCN genes including SCN1A, SCN2A, SCN3A, SCN7A, and SCN9A (Fig. 2).
In the literature, there are only two reports on epileptic patients with microduplication of 2q. The first is a family with neonatal seizures and intellectual disability caused by a 1.57 Mb microduplication of chromosome 2q24.3 containing SCN2A, SCN3A, and the 3′ end of SCN1A (Heron et al., 2010). This family comprised four individuals with neonatal onset seizures and learning difficulties. In three of them, seizures commenced on days 2–3 and settled by 5 months. In the remaining patient, seizures commenced on day 18 and ceased by 20 months. All of them had intellectual disability ranging from full-scale intelligence quotient (FSIQ) <40 to borderline intellect. The authors considered that the phenotype of the family is likely to be attributable to one or more of the duplicated (SCN2A, SCN3A) or partially duplicated (SCN1A) sodium channel subunit genes. The second is a girl with neonatal-infantile epilepsy and delayed development (Raymond et al., 2011). She had frequent seizures since 2–3 weeks of age and was unable to sit independently at 6 months of age. This patient had a 2.0 Mb microduplication at 2q24.3 including SCN1A, SCN2A, and SCN3A.
We identified another patient with severe seizures of neonatal onset and severe developmental delay. The size of the duplicated region in our patient was much larger than that in the previously reported patients, which may explain the more severe seizure prognosis and severe developmental delay. However, phenotype–genotype correlation is hard to establish because the patients reported by Raymond and by us are too young and the full spectrum of the phenotypic features have not been sufficiently understood. At least, what we can say is that SCNs in 2q24q25 are dose-sensitive in both loss and gain of genomic copy numbers and contribute to the severe seizure disorders.
Marini et al. (2009) reported microchromosomal copy number variations affecting SCN1A in Dravet syndrome, other epileptic encephalopathy, and generalized epilepsy with febrile seizures plus. They found a partial SCN1A duplication in two siblings with typical Dravet syndrome and a partial SCN1A amplification of five to six copies in another patient with Dravet syndrome. However, the phenotype of patients with chromosomal duplications including entire SCN gene clusters is different from that of the patients with Dravet syndrome, in terms of neonatal onset seizures, no seizures induced by fever, and focal seizures alone. This information would give us an important clue to help reveal the genetic mechanism of neonatal seizures.
At least 17 genes other than the SCN genes are present in the duplicated region of our patient. Some of them can contribute to seizures. SLC4A10 encodes an electroneutral sodium bicarbonate exchanger. Gurnett et al. (2008) reported a patient with a disruption of SLC4A10, who had focal seizures with an onset of 7 years of age and moderate mental retardation. Krepischi et al. (2010) reported that aCGH revealed a common disruption of SLC4A10 in two patients with epilepsy and mental retardation. One of them commenced refractory seizures since 2 months of age and had severe mental retardation. This phenotype is relatively similar to that of our patient. KCNH7 encodes a pore-forming subunit of the voltage-gated potassium channel. Although some genes encoding voltage-gated potassium channel such as KCNQ2 and KCNQ3, have been known to contribute to benign familial neonatal seizures (Schroeder et al., 1998), there have been no reports on contribution of KCNH7 to seizures or epilepsies. There is a possibility that the duplication of these genes may affect the phenotype of our patient by the modification of neuronal excitability through the altered ion channel function. As to the other duplicated genes, there have been no reports on the relation to epilepsy, brain anomaly, or developmental disorder of the brain, although future studies may unveil some possible role that these genes will play.
The identification of microchromosomal copy number variations was useful to plan antiepileptic treatment of our patient. Our patient has neonatal-onset focal seizures. Phenobarbital and phenytoin has been generally considered to be preferable for the treatment of neonatal seizures, whereas valproate has been considered to be less appropriate (Wheless et al., 2007). We considered phenobarbital as a drug of choice in our patient. On the other hand, valproate is usually considered to be preferable for the treatment of Dravet syndrome (Dravet & Bureau, 2008). After the microchromosomal aberration including SCN1A was identified, we presumed that valproate will be more appropriate than phenobarbital and levetiracetam. Eventually, the alteration of antiepileptic drugs resulted in seizure freedom. This indicates that aCGH may add useful information not only for the diagnosis of epilepsy but also for the determination of the treatment.
In conclusion, our findings show that the duplication of 2q24.2q24.3 including the whole SCN cluster will contribute to refractory neonatal epilepsy and severe mental retardation. This indicates that SCNs in this region are dose-sensitive in both loss and gain of genomic copy numbers.
This work was partly supported by the grants from the Ministry of Ministry of Health, Labour, and Welfare (H21-Shinkou-Ippan-010 and H22-Nanji-Ippan-028).
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. None of the authors has any conflict of interest to disclose.