The dominant mutations of STXBP1 responsible for EIEE4

Dominant mutations of STXPB1 (also called MUNC18-1) gene, which is located on chromosome 9q34.11, can cause sporadic EE. In autosomal dominant diseases, a weak penetrance (proportion of affected individuals among mutation carriers) can explain why parents are not both affected, one of them being an asymptomatic carrier. However, for deleterious phenotypes like EEs, most sporadic cases are due to de novo mutations. In this case, mutations that are eliminated from the population by their negative selective effect on patients’ reproduction are counter-balanced by the occurrence of neomutations. The mutation generally occurs in a gamete that will be involved later on in fecundation, the corresponding zygote being affected since the mutation is dominant. This is often the case for EE associated with mutations in STXBP1. STXBP1 gene contains 20 exons and encodes the syntaxin-binding protein 1. Forming a complex with the SNARE protein syntaxin1, the STXBP1 protein might mediate synaptic vesicle fusion. ⁴ In accordance with this hypothesis, deletion of Munc18-1 (paralog of STXBP1) in mice renders the brain synaptically silent, identifying Munc18-1 (STXBP1) as the currently most upstream essential protein in neurotransmitter release. ⁵ Saitsu and colleagues (2008) first described mutations in this gene in 5 unrelated cases among 13 patients with EIEE with suppression-burst (Ohtahara syndrome) for whom the main causes of EEIE were excluded. Missense, premature termination codon (non-sense mutations) and deletions were reported, leading to the conclusion that the disease is likely due to a haploinsufficiency (loss of function of the mutated copy of the gene) of STXBP1. ^{6, 7} EEIE4 is most often associated with neurological signs such as mental retardation, which can be profound, and spastic paraplegia. The rarity of reports on mutations in this gene has to be underlined and raises questions as to whether its systematic screening is indicated for patients with mental retardation and EIEE with suppression-burst.

SPTAN1, a newly identified gene responsible for sporadic EIEE5

Four patients presenting with early onset epileptic encephalopathy with poor attention, severe hypomyelination and reduction in cerebral white matter (EIEE5; 613477), were reported in 2008. On EEG, one patient presented with suppression-burst and the other three had hypsarrhythmia. The authors stated that this phenotype could be considered as a new clinical condition associated with early onset West syndrome (WS). ^{8, 9} A de novo 9q33.3-q34.11 microdeletion encompassing the STXBP1 gene has been detected in one of these patients. Since no mutations of STXBP1 were found in two of the remaining three subjects, the authors suspected that another gene within the deletion might contribute to this severe phenotype. SPTAN1 encoding alpha-II spectrin, which is involved in cell proliferation via arrest at cell cycle phase G1 and in myelination in zebrafish, ¹⁰ was localized in the deletion. In two subjects, an in-frame 3 bp deletion and a 6 bp duplication in SPTAN1, likely resulting in the synthesis of an abnormal protein, were found at the initial nucleation site of the alpha/beta spectrin heterodimer. ⁹ SPTAN1 was further screened in six unrelated individuals with WS and hypomyelination, but no mutations were found. In vitro functional expression studies suggested a dominant-negative effect of the two identified mutations on spectrin heterodimer stability, as well as perturbation of the axon initial segment.

Mutations in CDKL5/STK9 cause EE, preferentially in females

The CDKL5/STK9 gene was first incriminated in two unrelated girls with infantile spasm syndrome (ISSX2; 300672) associated with de novo balanced X-autosome translocations, t(X;7)(p22.3;p15) and t(X;6)(p22.3;q14), respectively. In both cases, the genomic rearrangements disrupted the CDKL5/STK9 gene. ¹⁴ Although CDKL5/STK9 was submitted to X inactivation in normal female somatic cells, the expressed protein was not functional in the two patients because of a preferential inactivation of the normal X. ¹⁴ Indeed, the inactivation of the translocated X might have diffused to the translocated autosome fragment, leading to a monosomy of the corresponding genomic region, which is lethal.

Mutations in CDKL5 were later on identified in female patients with atypical Rett syndrome. ¹⁵ These patients clearly had some Rett features such as deceleration of head growth, stereotypies and hand apraxia. In contrast, some signs were absent, such as a period of nearly normal development followed by regression with loss of acquired fine finger skill in early childhood and intensive eye communication, and the characteristic evolution of the Rett electroencephalogram. In the mouse brain, Weaving et al. (2004) showed that Cdkl5 expression overlapped with that of Mecp2 (OMIM: 300005), which is mutated in the classic Rett syndrome, ¹⁶ leading Tao et al. (2004) to postulate that these two genes may be involved in a common pathogenic pathway. ¹⁵

The characteristics of the epilepsy associated with CDKL5 mutations were defined by Bahi-Buisson and colleagues (2008), who showed in a retrospective study of 12 female patients with CDKL5 mutations that the epilepsy course had three successive stages: Stage I: early epilepsy (onset 1–10 weeks) with normal interictal EEG despite frequent convulsive seizures; Stage II: epileptic encephalopathy with infantile spasms and hypsarrhythmia; stage III: at the age of examination (3–19 years), seven patients were seizure free and six had developed refractory epilepsy with tonic seizures and myoclonia. ¹⁷ Moreover, early epilepsy with normal interictal EEG (stage I) and severe hypotonia seem to be key clinical features in identifying patients with CDKL5 mutations. ¹⁸

Epileptic encephalopathies as a part of the phenotypic spectrum associated with ARX mutations

Mutations in ARX were first identified in patients with three different phenotypes, including X-linked lissencephaly with abnormal genitalia (XLAG; MIM# 300215) ¹⁹, non syndromic X-linked mental retardation (MRX), ²⁰ and X-linked infantile spasms (ISSX; MIM# 308350). ²¹ The phenotypic spectrum associated with ARX mutations includes severe EEs with or without brain malformations. Shoudbrige and colleagues (2010) recently reviewed families with ARX mutations reported in the literature. ²² Among the 97 families identified, Ohtahara syndrome, X-linked myoclonic epilepsy and X-linked infantile spasms (West syndrome) were reported in 1, 1 and 12 families, respectively. Other syndromes including seizures were also identified, including Partington syndrome (intellectual disability with dystonic movements, ataxia, and seizures), mental retardation with tonic seizures with dystonia, and infantile epileptic-dyskinetic encephalopathy, in 7, 1 and 4 families, respectively.

The ARX gene, which contains five coding exons (GenBank: NM_139058.2) (MIM] 300382), maps to Xp22 and encodes a 562 amino acid mature protein, which is expressed predominately in the fetal and adult brain, testis, skeletal muscle, and pancreas. ^23–25 ARX is a paired-type homeobox protein that contains four polyalanine (PolyA) tracts, a homeodomain, and a conserved C-terminal aristaless domain. Male ARX-deficient mice have abnormal differentiation and deficient tangential migration of GABAergic (γ aminobutyric acid) interneurons in the ganglionic eminence, neocortex, and hippocampus, and abnormal testicular differentiation. ¹⁹

All types of mutations, missense, non-sense, insertions and deletions, have been reported in ARX. It is interesting to note that a phenotype-genotype correlation has recently emerged: ^{22, 26} non malformative syndromes including EE ²⁷ are mostly due to expansions of polyalanine tracts located in the N terminal fragment of the protein whereas syndromes with brain malformations are mostly associated with mutations leading to a premature termination codon (PTC) (non-sense or splice mutations) or deletions encompassing several exons. These mutations lead to a loss of function by haploinsufficiency since the mRNAs with PTC are mainly destroyed in the cells by the non-sense-mediated mRNA decay (NMD) system. ²⁸ Very recently, Kato et al. (2010) reported two families, comprising six males with Ohtahara syndrome in two generations, in which frameshift mutations in the terminal exon of the ARX gene, Ala524fsX534 and E536fsX672, were identified. It is noticeable that the PTC mutations localized in the last exon of genes escape the NMD system and most often cause the synthesis of a truncated protein. Interestingly, two patients developed West syndrome, and one of these later developed Lennox-Gastaut syndrome. The authors concluded that the analysis of ARX should be considered in sporadic or familial male patients with Ohtahara syndrome. ²⁹

SRPX2 is responsible for rolandic seizures with speech apraxia and mental retardation

Roll et al. (2006) reported a 3-generation French family with oral and speech dyspraxia (OSP), rolandic seizures (RS), and mental retardation. Eight individuals had OSP, rolandic seizures and mental retardation and three had only OSP and mental retardation without seizures. Epileptic patients had the electroencephalographic hallmark of RS (centro-temporal spikes that tend to occur in clusters and are strongly activated during sleep) during the active phase of the epileptic seizures. ³⁰ A whole-genome scan was performed and the gene was mapped to chromosome Xq. In the 20 cM large candidate interval, a missense mutation, N327S, was identified in the SRPX2 gene. SRPX2 is a secreted sushi-repeat-containing protein expressed in neurons of the human adult brain, including the rolandic area. The disease-causing mutation resulted in gain of glycosylation of the secreted mutant protein. ³⁰ Roll at al. (2006) also described a 12-year-old boy who presented with focal seizures beginning with numbness of the fingers of the right hand and leg and a sudden fall. Neurologic examination revealed clonus at both knees and generalized hyperreflexia with an equivocal right plantar response. Neuropsychologic examination at age 15 years showed low average to average intellect with weakness in mathematical ability. His EEG showed left centrotemporal epileptiform activity with a rolandic horizontal dipole. MRI showed bilateral posterior perisylvian polymicrogyria (MIM # 300388), more severe on the left and extending back to involve the parieto-occipital regions. Two of his maternal aunts had mild mental retardation, and his mother and another aunt were of normal intelligence. In this family, a second mutation was identified within the first sushi domain of SRPX2. These data suggest that SRPX2 may play a role in the development and/or function of the perisylvian region critical for language and cognitive development. Interestingly, the orthologous Srpx2 gene is not expressed during murine embryogenesis suggesting that this gene is involved in the development and functioning of language areas in human cortex.³⁰

Different types of mutation in SCN1A cause Dravet syndrome

All types of mutations have been identified in the coding sequence of the SCN1A gene in patients with Dravet syndrome: missense, nonsense, and splice-site mutations, small deletions and insertions. The mutations are located all along the SCN1A gene. Missense mutations are the most common mutation type identified (about 40%). ⁴² The main remaining mutation types (nonsense, splice-site mutations, frameshift) introduce PTCs into the mRNA, which are probably recognized and degraded via the NMD system. This hypothesis is compatible with the effects of the other mutation types detected, including alteration of the initiation codon and the microdeletions that delete essential regions of the proteins, introduce PTC, or delete the whole gene at the heterozygous state. Indeed, during the past few years, many techniques have been developed in order to detect genomic rearrangements such as deletions or duplications: quantitative PCR (QPCR), ⁴³ multiplex ligation-dependent probe amplification (MLPA) ⁴⁴ and microarrays. ⁴⁵ The analysis of series of patients without point mutations with these methods has identified several heterozygous rearrangements in SCN1A, variable in length and breakpoints and encompassing from a single exon to the whole gene. ^{42, 46, 47} Deletion of the whole gene was the most common rearrangement found, supporting haploinsufficiency as the main molecular mechanism responsible for Dravet syndrome. Analysis of the parents showed that most deletions occurred de novo. The nearby genes comprised in the deletion, including other genes encoding voltage-gated sodium channels (SCN7A and SCN9A), were different between patients, without evident variability of the phenotype. ⁴² However, patients with very large deletions could have consistent dysmorphic features including ear abnormalities, microcephaly, micrognathia and brachysyndactyly, likely related to the size of deletion and deletions of other genes than SCN1A. ⁴⁸

In spite of many functional studies, it remains unclear how missense mutations can cause a clinical phenotype indistinguishable from that of PTC mutations or whole-gene deletion and why some missense mutations are associated with a severe phenotype, such as Dravet syndrome, and others with curable epilepsies as in GEFS+ families. Kanai et al. (2004) suggested a preferential location of the missense mutations leading to Dravet syndrome in the S5 and S6 segments and the S5–S6 intracellular boucle, which together form the “pore” of the Nav1.1 channel.21. ⁴⁹ Depienne and colleagues (2009) reported that most missense .mutations causing Dravet syndrome affected highly conserved amino acids located in ion-transport sequences and resulted in chemically dissimilar changes in amino acid classes. However, these mutations were not preferentially located in S5–S6 segments, in contrast to the previous report (Figure 2). ⁴²

Figure 2

Schematic representation of the missense mutations and in-frame deletions in the Nav1.1 protein. Each star represents a missense mutation. Red stars: mutations identified in a single patient; green stars: recurrent mutations; blue triangles: inframe deletions (more...)

The proportion of patients with SCN1A mutations is highly variable in reported studies, ranging from 33% ^{39, 50} to 80–100%.^{40, 51, 52} These discrepancies may be due to the sizes of the series, the methods of screening, and use of different clinical criteria to define Dravet syndrome. Supporting this hypothesis, the proportion of positive patients is higher in SMEI using strict criteria than in epileptic syndromes closely related to SMEI including ICEGTC (intractable childhood epilepsy with generalised tonic–clonic seizures) or SMEB (borderline SMEI). ^{51, 53} Considering only the patients with typical Dravet syndrome, approximately 20% of them do not have mutations in SCN1A, even when microrearrangements have been excluded. ^{41, 42} In addition, recent studies have established that rare pathogenic mutations in SCN1A, some of which are de novo, can also be identified in other infantile EEs such as cryptogenic generalized epilepsies, cryptogenic focal epilepsies or infantile spasms.^{50, 53 54} This extension of the clinical spectrum related to SCN1A has called into question the initial concept of SMEI, because neither clinical nor genetic criteria are sufficient to delimit accurately the various syndromes.

Dravet syndrome in a familial context of epilepsy

Few patients with DS have a parent or relative with a milder epileptic phenotype. Depienne and colleagues (2010) showed that mutations inherited from an asymptomatic or mildly affected parent were identified in 10% of the DS patients. Mosaicism (the mutation is present only in a fraction of germinal [germinal mosaicism] or non-germinal [somatic mosaicism] cells) was the main event associated with inherited SCN1A mutations in DS patients (about 70%). Parental mosaicism in DS was not a rare situation since it was found in at least 7% of families with an SCN1A mutation. When the level of somatic mosaicism was high, the parent could present with seizures, although he/she was less severely affected than his/her child who carried the mutation in all cells. The clinical status of the mosaic parent appeared to be somehow correlated with the amount of the mutation in his/her blood cells although this correlation was not strict. ⁵⁵

Nevertheless, mosaicism is not the only situation accounting for inherited mutations as DS can also be encountered in the context of GEFS+ families. In that case, SCN1A missense mutations segregating in the family are associated with a wide phenotypic variability, with DS at the severe end of the spectrum. ³³ Distinguishing mosaicism from de novo constitutional mutations or other situations is of particular concern in disorders that are frequently sporadic, such as DS, in order to give appropriate genetic counseling. ⁵⁶ The risk of recurrence is, therefore, not null when the mutation is apparently de novo, which should be taken into account for genetic counseling.

A question that remains is whether SCN1A neomutations causing DS act through the same pathophysiological mechanisms as mutations found in GEFS+. Basically, mutations associated with clear loss-of-function are always associated with DS, except in one family recently described by Suls et al. (2010), in which a microdeletion of SCN1A segregated with a very variable phenotype in a 4-generation family. ⁵⁷ In mosaic patients, the mutations with a loss-of-function effect can be associated with milder phenotypes reminiscent of those seen in GEFS+ if present in only some neurons. Conversely, missense mutations found in GEFS+ are generally associated with mild epileptic phenotypes but can occasionally cause DS. From a mechanistic point of view, it is likely that the pathophysiological pathways are different: de novo mutations would be sufficient to cause DS whereas missense mutations associated with GEFS+ would not, and additional genetic or non-genetic factors would be necessary to cause DS in the latter case.

Dravet-like syndrome in females associated with mutations in PCDH19 gene

At least 20% of patients with Dravet syndrome are negative for mutations or rearrangements in SCN1A. In addition, rare mutations in GABRG2 have been described in cases of SMEI belonging to GEFS+ families ⁵⁸ but were not demonstrated in 29 sporadic cases. ³⁸ These findings suggest that Dravet syndrome could be genetically heterogeneous (i.e. other genes are involved in DS).

Mutations in PCDH19 mainly affect females with a Dravet-like syndrome

In order to identify new genes responsible for the disorder in SCN1A mutation-negative patients, Depienne and colleagues screened 41 patients with DS for micro-rearrangements with high-density SNP microarrays. Interestingly, a hemizygous deletion on chromosome Xq22.1, encompassing the PCDH19 gene, was identified in one male patient. To confirm that PCDH19 was responsible for a Dravet-like syndrome, its coding region was sequenced in 73 additional SCN1A mutation-negative patients. Different point mutations (four missense and five truncating mutations) were identified in 11 unrelated female patients (15%). The spectrum of mutations includes non-sense mutations, small deletions/insertions introducing a frameshift as well as missense mutations affecting highly conserved amino acids in the protein, predominantly in the extracellular domain, which is presumably involved in cell-cell interaction. These mutations are therefore predicted to result in a loss-of-function of the mutated allele. ⁵⁹

Protocadherin 19 is a 1148 amino acid transmembrane protein belonging to the protocadherin delta2 subclass of the cadherin superfamily, which is highly expressed in neural tissues and at different developmental stages. ^60–63 The precise functions of the protein remain so far unknown. However, Delta protocadherins were reported to mediate cell-cell adhesion in vitro and cell sorting in vivo, and could regulate the establishment of neuronal connections during brain development. ^{64, 65}

Patients with PCDH19 mutations could present with clinical features similar to patients with SCN1A mutations, including the association of early febrile and afebrile seizures, seizures occurring in clusters, developmental and language delays, behavioral disturbances, and cognitive regression. There were, however, slight but constant differences in the evolution of epilepsy, including fewer polymorphic seizures (in particular rare myoclonic jerks and atypical absences) in patients with PCDH19 mutations. These results show that PCDH19 plays a major role in epileptic encephalopathies, with a clinical spectrum overlapping that of DS. ⁵⁹

PCDH19-linked epilepsies: an unusual mode of inheritance

Mutations in PCDH19 were first reported to cause EFMR (epilepsy and mental retardation limited to females). The clinical features of EFMR, unlike those of DS, are highly variable, even in members of the same family: onset of seizures is between 6 and 36 months and affected females present with a combination of febrile and afebrile seizures of various types and a variable degree of psychomotor delay and cognitive impairment, ranging from mild to severe mental retardation. ⁶⁶ Dibbens et al. reported PCDH19 mutations in six large families and one small family with two affected sib pairs. ⁶⁰ All the patients were familial cases that were, for the most part, already adults at the time of examination, and appeared socially integrated in that most of them were married and had children.

PCDH19-related epileptic encephalopathy therefore mainly affects females. In a large series of PCDH19 mutation-positive index cases in whom inheritance could be assessed, half of the mutations occurred de novo and half were inherited from fathers who were healthy, had no cognitive impairment and had never had febrile seizures or epilepsy. ⁵⁹ (Depienne at al., personal communication) This heredity is very different from the X-linked mode of inheritance encountered for ARX or CDKL5 mutations since in this case only the heterozygous females are affected whereas the hemizygous males are asymptomatic and spared (Figure 3).

Figure 3

Segregation analysis of the PCDH19 deletion and point mutations in 12 families. del/+, m/+ and v/+ denote individuals heterozygous for the deletion, mutation or variant, respectively; +/+ denotes individuals carrying homozygous wild-type alleles. Squares (more...)

Several mechanisms have been suggested to account for the unusual mode of inheritance observed in PCDH19-linked epilepsy, one of which is cellular interference, a mechanism reminiscent of metabolic interference. ^{60 67, 68} This concept postulates that random inactivation of one X chromosome in mutated females generates tissue mosaicism (i.e., coexistence of PCDH19-positive an PCDH19-negative cells), which would be pathogenic by altering cell-cell interactions; normal individuals and mutated males, who are homogeneous for PCDH19-positive or PCDH19-negative cells, respectively, would not develop the disease (Figure 4). The identification of an affected male who was mosaic for the PCDH19 deletion in his fibroblasts, and therefore had PCDH19-positive and PCDH19-negative cells in this tissue, strongly supports the hypothesis of cellular interference as the main pathogenic mechanism associated with PCDH19 mutations. ⁵⁹

Figure 4

Schematic illustration of the cellular interference mechanism associated with PCDH19 mutations. A) In normal individuals characterized by a homogeneous population of PCDH19-positive cells, neurons are able to form normal neuronal networks. B) In male (more...)

Identifying new genes responsible for monogenic forms of epileptic encephalopathies

Identifying genes in sporadic epileptic encephalopathies is difficult since the classical genetic strategies based on the study of large families are not possible. However, alternative genetic approaches are available; they are various and related to the context of the disease:

1. As for ARX, CDKL5/STK9 and SRPX2, EE can be part of a complex phenotype with or without brain abnormalities. The strategy consisting in the screening of the responsible gene in series of patients with isolated EE of the same type can be fruitful. Some patients with isolated defined EEs, even if they were rare, had mutations in these genes.
2. The study of GEFS+ and Dravet syndrome highlights the possibility of a gene identification strategy based on the detailed description of syndromes: the occurrence of Dravet syndrome in families with GEFS+ and the high sensitivity to fever in both conditions led to Dravet syndrome being considered as a part of the GEFS+ phenotypic spectrum. Claes et al. (2001) to be homogeneous therefore tested SCN1A, which had firstly been incriminated in GEFS+, as a candidate gene in sporadic patients with DS and identified the first de novo mutations of SCN1A. ³⁶
3. Genomic rearrangements have been identified in most of the genes known to be responsible for infantile EE. Comparative genomic hybridization (CGH) and SNP microarrays have enabled the entire genome to be screened at various resolutions. The detection of a de novo rearrangement in a patient points to a candidate region in which the causative gene is localized. The candidate region often contains several genes and the involvement of one particular gene needs to be confirmed by the identification of point mutations located in the coding region of this gene in patients with the same phenotype. As we saw in previous sections, this strategy was applied to identify the SPTAN1 and CDKL5/STK9 genes. When the rearrangement is inherited from an asymptomatic parent, its causality in the patient’s phenotype is more difficult to prove, as it may correspond to a benign copy number variant (CNV) or copy number polymorphism (CNP). However, it cannot be excluded that a recessive mutation in a gene of the region has been transmitted on the undeleted remaining copy by the other parent. Association of an inherited CNV with a phenotype in more than one patient would argue in a favor of such a recessive gene and lead to the sequencing of candidate genes in the regions concerned.

Identifying susceptibility genes in polygenic-determined epileptic encephalopathies

Since the vast majority of cases of EE are sporadic, the most frequent inheritance of EE is usually thought of as being complex or multifactorial, resulting from the interaction between several genes (susceptibility genes) and environmental factors. The identification of these genes requires non-parametric analyses (genetic analyses that do not depend on parameters of monogenic diseases as the mode of inheritance, penetrance, rate of phenocopies, etc.). For these analyses, several conditions are needed: i) a large number of affected individuals (at least in the hundreds) and a similar number of controls well-matched for age, sex and ethnic background, ii) a previous exclusion of monogenic forms, and iii) a low genetic heterogeneity of the population (a great genetic heterogeneity could otherwise mask or dilute the effect of individual susceptibility genes). These conditions are difficult to obtain in EEs, which are rare diseases affecting children (problem of constituting a control group) with a well-known etiological heterogeneity (Table 1). However, such a strategy has already been performed, in which the entities Landau-Kleffner syndrome (LKS), continuous spike-and-waves during sleep syndrome (CSWS), and benign childhood epilepsy with centrotemporal spikes (BCECTS) and benign rolandic epilepsy, were considered to be part of a single continuous spectrum of disorders. ⁶⁹ The centrotemporal spikes (CTS) were considered to be an endophenotype determined by a common set of susceptibility genes, the different phenotypes being determined by other susceptibility genes and/or environmental factors. A genome-wide study showed a linkage of the CTS endophenotype to 11p13, especially with polymorphic markers in the ELP4 (Elongator Protein Complex 4) gene. ⁷⁰ The yeast Elp4 protein is one of the 6 subunits of Elongator. The Elongator complex is involved in transcription of several genes that regulate the actin cytoskeleton, cell motility and migration. These functions are crucial in the nervous system for nerve cell growth cone motility, axon outgrowth and guidance, neuritogenesis and neuronal migration during development. ^{71, 72} This gene is a good candidate-by-function. It remains to identify functional variants in the vicinity of the ELP4 gene to confirm its genetic involvement. This strategy could be applied to other clinical or EEG traits.

In conclusion, in the past ten years, a great amount of genetic data has accumulated, updating our knowledge on the etiologies and pathophysiological mechanisms of infantile epileptic encephalopathies. We can expect this chapter to become incomplete in the very next future and hope that all this basic science will open soon new therapeutic avenues to cure these devastating disorders.