NRXN1 deletions and the double hit hypothesis of idiopathic epilepsy

Old friends. Structural genomic variants or Copy Number Variations (CNVs) play an important role in many neurodevelopmental disorders including epilepsy, autism, schizophrenia and intellectual disability. Many of the CNVs representing genetic risk factors overlap between these diseases. Now, a recent study in Epilepsia reports on the exon-disrupting deletions in NRXN1 as genetic risk factors for Idiopathic Generalised Epilepsy. NRNX1 deletions were previously reported in several other neurodevelopmental disorders. However, there is an interesting and unanticipated twist to the story.

Neurexins and neuroligins. During synapse formation, cell adhesion proteins are important for synapse differentation and maturation. In the presynapse, these cell-adhesion proteins are the neuroligins. In the postsynapse, this is accomplished through neurexins. Historically, these proteins were first identified as the binding partner of alpha-latrotoxin, the venom of the black widow spider. This demonstrates that both neurexins and neuroligins also have additional functions in the mature CNS. In the last few years, CNVs and mutations in the neurexins and neuroligins were found to be risk factors of autism, schizophrenia and intellectual disability, emphasizing the role of impaired synapse maturation in these disorders. But where does epilepsy fit into this?

NRXN1 deletions in IGE. Møller and collaborators screened the EPICURE IGE cohort for microdeletions in the NRXN1 gene and compared these findings to controls. They identified 5 out of ~1500 patients with microdeletions in NRXN1, which are exon-disruptive. In contrast, similar deletions were only found in a very small subset (2/6000) in controls. This suggests that deletions in this gene are a significant risk factor for IGE, increasing the risk of developing IGE roughly by the factor of 10. Interestingly, 4/5 probands were from small IGE families. This is a surprising finding and in contrast to other microdeletions with more frequent sporadic cases. Therefore, the frequency of NRXN1 deletion in familial Idiopathic Generalized Epilepsy might even be higher. Segregation of the NRXN1 deletion in the families was variable, which is in line with the behavior of moderate genetic risk factors in families. Even though NRXN1 deletions are a genomic chameleon with a broad phenotypic range, there were few other phenotypes seen except for some patients with intellectual disability. Also, the NRXN1 deletions were not the only variant seen in the families.

NRXN1 deletions and IGE. 5/1500 IGE patients with NRXN1 deletions were identified in the study by Møller and collaborators. 3/5 probands carried additional, possibly pathogenic structural genomic variants

NRXN1 deletions and IGE. 5/1500 IGE patients with NRXN1 deletions were identified in the study by Møller and collaborators. 3/5 probands carried additional, possibly pathogenic structural genomic variants.

The second hit. In three out of the five families, another large and possible pathogenic structural genomic variant was identified. In one family, this variant included a 1q21.1 microdeletion, which segregated with the phenotype. In both other families, the variant was de novo or segregation was not tested. This finding, despite the small sample size, is intriguing. Double hits, i.e. the presence of two or more possible pathogenic structural genomic variants, has been reported in intellectual disability. The data by Møller and collaborators now begs the question whether a similar double hit model might be applicable to Idiopathic Generalized Epilepsy.

Double hit or no double hit. The double hit hypothesis was initially conceptualized in intellectual disability based on findings from large study samples. This hypothesis suggests that the genetic morbidity in an individual patient can be traced back to two identifiable variants in this patient. This model implies that a few rare and strong genetic risk factors act together to produce disease (the oligogenic model). Despite the fact that similar suggestions have been around for decades, statistical evidence for this on the basis of CNVs was only provided recently. Can IGE be explained through a similar model? Prior to the paper by Møller and collaborators, I would have been skeptical. IGE a is relatively mild disorder compared to severe intellectual disability and –at least to me- the double hit hypothesis was limited to severe diseases. However, at the same time, it is obvious that the epilepsy phenotype of an individual patient cannot be sufficiently explained by a single risk variant. Future investigations will tell whether this seemingly high frequency of second hits is part of a recurring pattern.

What does NRNX1 add? With 0.3% of IGE patients carrying NRXN1 deletions, these genetic alterations are one of the more common of the rare variants. This finding further emphasizes the close relationship between IGE and other neurodevelopmental disorders, but also stresses that the basic mechanisms that lead to an IGE phenotype and protect the patient from autism are not at all understood.

Double hit and exomes. Returing to the complex genetics of epilepsies might seem like an anachronism in 2013. However, it is not. Family exome sequencing studies are sometimes interpreted as finally getting through to the monogenic causation of human disease. However, in contrast to the assumption, there is a similar complexity in exome studies. The genes identified in neurodevelopmental disorders again overlap, leaving sufficient room for the genetic background. For example, the reason why one SCN1A mutation leads to autism, while a different mutation results in Dravet Syndrome is not understood. Testing the validity of models for oligogenic or complex genetic inheritance will be an important step towards tackling this question.

Ingo Helbig

Child Neurology Fellow and epilepsy genetics researcher at the Children’s Hospital of Philadelphia (CHOP), USA and Department of Neuropediatrics, Kiel, Germany

Facebook Twitter