Genetics of the epilepsies: genetic twists in the channels and other tales.
Ingrid E Scheffer,Yue-Hua Zhang,Jozef Gecz,Leanne Dibbens
DOI: https://doi.org/10.1111/j.1528-1167.2009.02440.x
IF: 6.74
2010-01-01
Epilepsia
Abstract:Recent studies of the genetics of the epilepsies have identified surprising mechanisms and novel patterns of inheritance. The greatest challenge is to solve the genetics of the common idiopathic generalized epilepsies (IGEs). The common IGEs follow complex inheritance, where multiple genes are likely to contribute and environmental factors may also play a role. Little is known about the susceptibility genes contributing to complex inheritance, although there is evidence that the calcium channel subunit gene, CACNA1H, and another gene whose function is poorly understood, EFHC1, play a role (Helbig et al., 2008). In contrast, the 19 or so genes identified for idiopathic epilepsies have largely been found in monogenic epilepsies, often in large multiplex families. Most of these encode ion channels encompassing both voltage-gated and ligand-gated ion channel subunits. More recently, unexpected inheritance patterns have emerged. The most important is that a severe epileptic encephalopathy of infancy, Dravet syndrome, has turned out to be a monogenic disease. Second, an unrecognized, major group of IGEs called genetic epilepsy with febrile seizures plus (GEFS+) has become a model of “complex monogenic” epilepsy, exemplified in large dominant families in which genes have been found. The dominant genes are not the whole story in these families, as the heterogenous phenotypes are likely to be due to additional genetic factors. Third, the molecular basis for a fascinating novel mode of inheritance in epilepsy has been discovered for a familial epilepsy in which only females are affected. This X-linked inheritance pattern requires clinicians to think differently about the family histories they obtain. The only current genetic test that has clinical utility is mutational analysis of the sodium channel alpha-1 subunit gene, SCN1A, which is abnormal in 70–80% of individuals with Dravet syndrome. Dravet syndrome, or severe myoclonic epilepsy of infancy (SMEI), begins in a previously normal infant at around 6 months of age, and always before 14 months. The infant presents with febrile status epilepticus, which may be hemiclonic or generalized. Over the next 6 months, the infant presents every month or two with convulsive seizures. If they have hemiclonic attacks, the subsequent seizure may involve the other side. Between one and 4 years, the child develops other types of seizures including myoclonic, partial, atonic, and absence seizures. Hyperthermia, such as fever or a warm bath, often precipitates seizures. Development in the first year is normal but subsequently slows and may regress. Children may develop ataxia and pyramidal signs with time. Despite frequent seizures, their electroencephalography (EEG) may be normal until age 2 years when generalized spike wave activity is seen; approximately 10% are photosensitive. The magnetic resonance imaging (MRI) scan is either normal or shows nonspecific features such as atrophy (Dravet et al., 2005; Mullen & Scheffer, 2009). Many adult patients with Dravet syndrome remain undiagnosed unless the clinician takes a careful history regarding the first few years of life. This usually necessitates talking to the patient’s parents and sourcing the hospital records pertaining to early admissions and investigations. Adults with Dravet syndrome usually continue to have seizures. They have brief nocturnal convulsions, which occur at least monthly, if not weekly. In our review of 14 patients older than 18 years of age, half also had complex partial seizures and a few had ongoing absence, myoclonic, and atonic seizures. Thirteen patients had intellectual disability, which was severe in six, moderate in four, and mild in three. Neurologic examination revealed cerebellar signs in four and pyramidal signs in five patients. Most patients remained fully dependent (Jansen et al., 2006). SCN1A mutations are found in >70% of patients with Dravet syndrome; 95% arise de novo. Around 40% are truncation and 40% missense mutations. It is important to note that 10% of patients who are negative on conventional (sequence-based) mutational analysis, have copy number variation including exonic deletion or duplication and that can involve several exons or the whole gene (Madia et al., 2006; Mulley et al., 2006; Suls et al., 2006; Depienne et al., 2009b; Marini et al., 2009). Japanese authors identified a subgroup of patients with borderline Dravet syndrome. This is where a patient misses a key feature such as myoclonic seizures, generalized spike wave on EEG, or has normal development (Ohmori et al., 2003; Fukuma et al., 2004). Within the borderline subgroup, there is a subset coined intractable childhood epilepsy with generalized tonic–clonic seizures or severe idiopathic generalized epilepsy of infancy (Fujiwara et al., 1992; Doose et al., 1998). We asked whether the distinction between the borderline and the classical form of Dravet syndrome was meaningful. Our findings suggested it would be better to merge the borderline and classical forms under the term Dravet syndrome, as the mutation rates were ∼70% in the borderline group and 80% in the classical group. Truncation and missense mutations were observed in both groups (Harkin et al., 2007). We looked for other early childhood epileptic encephalopathies due to SCN1A mutations. We identified a new phenotype termed symptomatic infantile multifocal epilepsy for which three of five cases had SCN1A mutations. This phenotype was characterized by multiple different types of focal seizures with varying seizure semiology with onset in the first year. EEG showed abundant multifocal epileptiform discharges without generalized spike wave activity. Imaging was normal. On some occasions, patients had myoclonic, tonic, or tonic–clonic seizures. These individuals had normal development initially with later cognitive decline, with one patient having documented deterioration after 6 years (Harkin et al., 2007). Our large study also found that 6 of 25 cases with cryptogenic generalized epilepsy had SCN1A mutations. Two of these cases had a more severe Dravet-like phenotype, with onset of seizures before 2 months and abnormal early development. Another important finding in Dravet syndrome is that germline mosaicism may result in siblings with Dravet syndrome in cases where the parents were initially thought not to carry an SCN1A mutation. More sophisticated genetic analysis shows that one parent has low level mosaicism for the mutation. The parent is either clinically unaffected or has a mild phenotype such as febrile seizures (Depienne et al., 2006; Gennaro et al., 2006; Marini et al., 2006; Morimoto et al., 2006). Since we first described genetic (formerly generalized) epilepsy with febrile seizures plus (GEFS+), many large families have been reported and four genes identified. GEFS+ is a familial epilepsy syndrome, diagnosed on the basis of at least two individuals with GEFS+ phenotypes in a family. The GEFS+ spectrum denotes the phenotypic heterogeneity observed in families including febrile seizures (FS) and febrile seizures plus (FS+) at its mildest end, and myoclonic-astatic epilepsy and Dravet syndrome at its severe end. Our analysis of 20 new families with Dravet syndrome, selected on the basis of two individuals in a family with a GEFS+ phenotype, showed that partial seizures are more frequent than previously thought, accounting for >10% of the cases. Overlap with classical idiopathic generalized epilepsy (IGE) is also seen. A review of all of our 45 families shows that FS remains the most common phenotype, followed by FS+. In recent years, it has become apparent that partial seizures are not uncommon in GEFS+, and hence the name change to genetic epilepsy with febrile seizures plus (Scheffer et al., 2009). GEFS+ was originally recognized because of remarkable large autosomal dominant pedigrees with low penetrance of 60–70%. It is likely that most cases, however, arise in small families or even occur sporadically. In these instances GEFS+ probably follows complex inheritance. Quite often there is a family history of seizures on both sides of the family supporting a polygenic basis. The understanding that GEFS+ is usually complex has important genetic counseling ramifications as it implies a low risk to siblings overall in smaller GEFS+ families. Both sodium channel and γ-aminobutyric acid (GABA)A receptor subunit genes have been implicated in GEFS+. The most important GEFS+ gene is SCN1A, in which missense mutations occur (Helbig et al., 2008). The beta-1 sodium channel subunit gene, SCN1B, has a major role in regulating the kinetics of the sodium channel and was the first gene identified in GEFS+ (Wallace et al., 1998). SCN1B mutations have been associated with temporal lobe epilepsy in GEFS+ families (Scheffer et al., 2007). The γ-2 subunit of the GABAA receptor has been associated with GEFS+ in several large families (Baulac et al., 2001; Wallace et al., 2001; Harkin et al., 2002; Helbig et al., 2008), whereas the subunit of the GABAA receptor has been implicated in small GEFS+ families, a finding that requires confirmation (Dibbens et al., 2004). Therefore, even where a gene of major dominant effect is identified in a GEFS+ family, further research is required to understand the genes that interact to produce the phenotypic heterogeneity characteristic of GEFS+. X-linked inheritance is typically associated with females carrying or expressing a disorder and males being more severely affected. In 1971, a very unusual form of X-linked inheritance was identified in a large North American family with epilepsy limited to females with mental retardation (EFMR) reported three times over 26 years (Juberg & Hellman, 1971; Fabisiak & Erickson, 1990; Ryan et al., 1997). Our attention recently turned to EFMR when we studied two Israeli and two Australian families. In this disorder, females are affected and unaffected males transmit the disorder. Therefore, EFMR should be considered when only girls are affected in a family. A female has a 50% chance of their daughters having the disorder and their sons have a 50% chance of being a carrier. Just as importantly, every daughter of a transmitting male is affected (Scheffer et al., 2008). Affected females have seizure onset between 4 and 36 months of age. They typically present with tonic–clonic seizures and partial seizures, often triggered initially by fever. Other seizure types include absence, tonic, myoclonic, and atonic seizures. The developmental trajectory of affected females varies considerably. Approximately one-third are normal, half regress often with seizure onset or in the second to third year, and 20% are delayed throughout. Intellectual outcome varies from normal to profound intellectual disability. Psychiatric features are prominent and include autistic features, aggression, obsessionality, and depression (Scheffer et al., 2008). We identified PCDH19, encoding protocadherin 19, as the causative gene in 2008 (Dibbens et al., 2008). Mutations were found in seven families, including the original large American family. The mutations included five protein truncation and two missense mutations; they all occurred in the extracellular domain of the protein. Mutations cosegregated with EFMR in the respective pedigrees and were not identified by testing 750 control chromosomes. Protocadherin 19 is part of the protocadherin delta-2 subclass of the cadherin superfamily. The cadherin superfamily includes cell–cell adhesion molecules. Protocadherin members are expressed predominantly in the nervous system, where they have a role in establishing neuronal connections and in signal transduction at synaptic membranes. The biologic role of protocadherin 19 is not known (Dibbens et al., 2008). The mechanism of an X-linked gene causing female-only affected disease has been the source of considerable conjecture (Ryan et al., 1997). Recently light has been shed on this question by Depienne and colleagues (Depienne et al., 2009a), who reported an affected male who is mosaic for a chromosomal deletion containing the PCDH19 gene, showing that the theory of cellular interference is likely to hold true. This theory suggests that if an individual has two populations of protocadherin cells then a phenotype occurs. In the normal female or a transmitting male where only one protocadherin population of cells is seen, protocadherin-positive or protocadherin-negative cells, respectively, then no phenotype is demonstrated. The mechanism requires further clarification. There is still a huge amount of research to be performed in the genetics of the epilepsies. New discoveries are turning up every day and answering questions, but also raising many more in their stead. Professors Eva and Fred Andermann have been the pioneers of epilepsy genetics research, showing us how to perform family studies and also how to think more broadly about the genetics of the epilepsies. Their work will continue to instill academic thinking, partly through their personal training of many leaders in this area. We cannot express highly enough our gratitude and admiration for this incredible partnership of intellectual rigor and visionary thinking, married with humanity and care for those who have had the good fortune to be inspired by them both. The authors have nothing to disclose.