Abstracts

Rationale: The genetic basis of developmental and epileptic encephalopathies (DEEs) is now well established with over 50 genes associated with this disorder; the majority of pathogenic variants arise de novo. The unbiased approach of whole exome sequencing to gene discovery has led to many new biological mechanisms being implicated in epilepsy, including the mTOR pathway, synaptic vesicle transport and chromatin remodeling. Moreover, genes that are markers of cell-type identity during development are increasingly implicated in epilepsy, including FOXG1 (progenitors) and RORB (layer IV). In this study we identify pathogenic variants in upper cortical layer II/III marker, CUX2, in patients with DEE. Methods: Though a multicenter collaboration we identified patients with CUX2 pathogenic variants and performed detailed clinical phenotyping in these individuals. Functional studies using patient derived induced pluripotent stem cell (iPSC) lines and heterozygous knockout CUX2 cell lines are ongoing. Results: We identified nine patients with a de novo CUX2 p.Glu590Lys variant by either whole exome or gene panel sequencing in a clinical or research setting. The p.Glu590Lys variant locates to the first DNA-binding domain of the transcription factor, CUX2 and disrupts a highly conserved residue that is predicted to be damaging (CADD score -33, Polyphen score - 1). All nine patients presented with DEE. Median age at seizure onset was six months and ranged from 2-12 months, one patient had later onset at 9 years. Multiple seizure types were observed at onset including myoclonic (n=4), absence (n=3) and focal (n=2) seizures.  Seizures were ongoing in the majority of patients (6/9) and more prominently included myoclonic (n=5), absence (n=3) and generalized tonic clonic (n=3) seizures. MRI was normal in 6/9 individuals; findings in the other three patients included bilateral lesions in the globus pallidus, hippocampal asymmetry and thin corpus callosum. 8/9 patients had profound or severe intellectual disability and were non-verbal; the remaining patient is only 6 months of age. Conclusions: We report the recurrent p.Glu590Lys CUX2 pathogenic variant in nine patients with DEE and poor prognosis. To date, de novo CUX2 variants have not been reported at any other amino acid position in CUX2 in patients with epilepsy or other neurodevelopmental disorders. Moreover, the gene is intolerant to variation in the general population. Our results suggest that the p.Glu590Lys variant causes DEE by a distinct pathogenic mechanism that may include partial loss of function. Neuronal differentiation and functional experiments in stem cells from patients and genome edited heterozygous knockout lines (loss-of-function) to dissect the pathogenic effect on CUX2 function are ongoing. Our results will shed light not only on CUX2 function, but also broaden our appreciation for early neuronal development in the pathogenesis of developmental and epileptic encephalopathy. Funding: CURE Taking Flight Award, K99/R00 to GLC