SCN2A – this is what you need to know in 2015

SCN2A. The next gene on our list of epilepsy genes to review is SCN2A. Within less than three years, SCN2A has risen from a gene for a very rare benign familial epilepsy syndrome to one of the most prominent genes associated with neurodevelopmental disorders to date. Epilepsies due to SCN2A mutations can have a broad range of phenotypes that are still not fully understood. Here is our 2015 post on the gene that we refer to as the genetic shapeshifter.

Genotype vs. Phenotype. Phenotypes associated with SCN2A may broadly be grouped into three categories; (1) mutations causing epileptic encephalopathies including Ohtahara Syndrome, unclassified epileptic encephalopathies with or without dystonia, and Infantile Spasms or Lennox-Gastaut Syndrome, (2) mutations resulting in Benign Familial Neonatal-Infantile Seizures (BFNIS) and (3) mutations resulting in autism and intellectual disability (ID). Notably, truncation mutations and gene deletions (not shown) are only found in patients with autism or ID. The information in this graph has been compiled from various sources in the literature [1,2,3,4,5,6]

Genotype vs. Phenotype. Phenotypes associated with SCN2A may broadly be grouped into three categories; (1) mutations causing epileptic encephalopathies including Ohtahara Syndrome, unclassified epileptic encephalopathies with or without dystonia, and Infantile Spasms or Lennox-Gastaut Syndrome, (2) mutations resulting in Benign Familial Neonatal-Infantile Seizures (BFNIS) and (3) mutations resulting in autism and intellectual disability (ID). Notably, truncation mutations and gene deletions (not shown) are only found in patients with autism or ID. The information in this graph has been compiled from various sources in the literature, which is referenced in our previous post about SCN2A.

Phenotype. There is no comprehensive overview of SCN2A phenotypes available to date, and the full spectrum of neurodevelopmental disorders due to mutations in this gene is not fully understood. From my current understanding of the literature, I would breakdown the SCN2A phenotypes into four different groups including (1) benign neonatal or infantile epilepsies, (2) autism and intellectual disability, (3) Infantile Spasms, and (4) early onset epileptic encephalopathies including Ohtahara Syndrome and severe neonatal epilepsy. All phenotypes within the SCN2A spectrum can be conceptualized along three dimensions including (a) cognitive outcome, (b) seizures, and (c) movement disorders. Here are the main phenotypes of this spectrum, given in historical order.

Benign familial neonatal/infantile seizures (BFNIS). The familial syndrome of benign familial neonatal/infantile seizures was the first SCN2A phenotype to be described, and it is probably the least common presentation of an SCN2A mutation. Mutations were identified in large families where affected individuals had neonatal and/or infantile seizures and a normal neurodevelopmental outcome. This familial form of SCN2A has been reported in independent studies after the initial discovery, but it became obvious quickly that this familial syndrome is much less common than familial neonatal seizures due to mutation in KCNQ2 or infantile seizures due to PRRT2.

Autism, intellectual disability, schizophrenia. After the initial discovery of SCN2A as a gene for BFNIS, it became silent around SCN2A. Interestingly, this gene re-emerged as one of the most obvious candidate genes in many neurodevelopmental disorders in large exome studies. Several studies found that de novo mutations in SCN2A and leading to neurodevelopmental phenotypes including intellectual disability, autism, and schizophrenia are usually truncating mutations, suggesting that haploinsufficiency plays a major role.

Infantile Spasms. We have broken down the epileptic encephalopathy phenotypes caused by mutations in SCN2A in two broad categories – epileptic encephalopathies presenting with Infantile Spasms (usually at the age of 3-6 months) or severe epilepsies starting earlier. Even though there is probably a gradient between both phenotypes, we felt that breaking down the SCN2A epileptic encephalopathies into two major categories may make sense from a clinical perspective, as we usually think about different causes for seizures in children presenting with epilepsy as neonates compared to the epilepsy in children developing Infantile Spasms after an initial normal development. With regards to Infantile Spasms, it is interesting to note that the patients in the Epi4K publication (Infantile Spasms) had the identical R853Q mutation, suggesting some genotype/phenotype correlation. Some patients with SCN2A-related Infantile Spasms are also reported to have movement disorders.

Severe early-onset epileptic encephalopathies. There is an entire spectrum of severe epilepsies due to mutations in SCN2A that start neonatally or within the first two months of life. For some epilepsies, it is possible to classify them into known epilepsy syndromes such as Ohtahara Syndrome. For other epilepsies, a classification is more difficult, and these conditions are referred to as unclassified epileptic encephalopathies. Again, movement disorders seem to be common in this group of severe epilepsies. As of 2015, there is no established genotype/phenotype correlation, complicating the interpretation of genetic findings in this group of epilepsies. There are, however, several recurrent mutations that are emerging that may give a hint at the emerging genetic landscape of SCN2A encephalopathy.

Genotype. The spectrum of SCN2A-related epilepsies are due to mutations in the gene for the voltage-gated sodium channel, neuronal type II, alpha subunit. There are various types of genetic mechanisms implicated in the disease etiology. For some mutations, the phenotypic spectrum is obvious. For other mutations, the phenotypic range remains to be explored. For neurodevelopmental disorders other than epilepsy, mutations seem to be protein-disrupting. For example, truncation mutations are found in these phenotypes but are not observed in patients with epilepsy phenotypes. Some of the mutations seen in epileptic encephalopathies are recurrent, suggesting that other mechanisms such as dominant-negative or activating mechanisms are involved.

Mutational spectrum. Here are some of the mutations found in SCN2A. This list is currently only a small representation of the mutational spectrum found in this gene. There is currently no publicly available SCN2A-related gene database that includes all published mutations. Organizing this data in a phenotype-specific way is part of our Epilepsiome project. Some of the information about de novo mutations was taken from the most recent download of the NPdenovo database (see attached table). All annotations were re-annotated with wANNOVAR and SCN2A:NM_001040143 was used as a reference. With the exception of the BFNIS mutations, these variants were reported as de novo mutations. This list of mutations is incomplete and will be expanded in the following months to provide you with a better idea about the spectrum of SCN2A-related disorders.

Phenotype Mutation Amino acid ClinVar
BFNIS c.3988C>T p.Leu1330Phe ClinVar
BNFIS c.4687C>G p.Leu1563Val ClinVar
Autism c.476+1G>A Splicing Link
Autism c.2877C>A p.Cys959X Link
ID c.4193G>A p.Trp1398X Link
ID c.4561C>T p.Gln1521X Link
Schizophrenia c.2150-2A>G Splicing Link
Infantile Spasms c.2558G>A p.Arg853Gln PubMed
EOEE c.5645G>A p.Arg1882Gln ClinVar
Ohtahara Syndrome c.788C>T p.Ala263Val PubMed

My patient has a mutation in SCN2A – what does it mean? Given that SCN2A is found in a variety of genetic epilepsies and neurodevelopmental disorders, this answer is not very easy. For many neurodevelopmental disorders, it has been shown that SCN2A mutations arise de novo; therefore confirming that a particular variant is de novo is one important step. This is particularly relevant as there is no single phenotype that is tightly linked with SCN2A. If a variant in SCN2A is detected in a patient, it may therefore be that this variant is inherited and an innocent bystander and the patient may carry a different causative mutation in another gene. Some SCN2A variants are recurrent and may therefore be considered pathogenic without parental confirmation. Other SCN2A variants such as splice or truncating mutations are more likely to be identified in autism and intellectual disability than in patients with epileptic encephalopathies.

Mechanism. The SCN2A protein is located at the axon initial segment, the key structure in neurons where the decision is made whether the addition of all excitatory and inhibitory inputs will result in the firing of an action potential. Therefore, the function of SCN2A is pivotal to neurons. While the loss of function obviously results in haploinsufficiency for SCN2A as seen in some patients with autism and intellectual disability, the functional impact of other mutations is less clear. There is some suggestion that the missense mutations seen in patients with epileptic encephalopathies are activating whereas some of the BFNIS mutations have reduced cell surface expression, resulting in a net loss of SCN2A function. Systematic data on a battery of SCN2A variants implicated in human epilepsy is currently not readily available, leaving some of the functional aspects difficult to address. This is particularly pertinent as there is currently no good mouse model available.

Community. The SCN2A community is very active and maintains scn2a.org, which provides a link to family organizations and Facebook groups.

 

Ingo Helbig is a child neurologist and epilepsy genetics researcher working at the Children’s Hospital of Philadelphia (CHOP), USA.

Twitter