Theses Doctoral

Modeling Gene Therapy for Intractable Developmental and Epileptic Encephalopathy

Aimiuwu, Osasumwen Virginia

Childhood epileptic encephalopathies (EE) are severe neurodevelopmental diseases that manifest in early development. EE is characterized by abnormal electroencephalographic (EEG) activity, intractable seizures comprising of various seizure types, as well as cognitive, behavioral and neurological defects. Developmental and epileptic encephalopathies (DEEs) are a subclass of EEs where the progressive and permanent cognitive and neurophysiological deterioration is not caused by seizure activity alone, but is caused by the same underlying etiology. Recent advances in whole exome sequencing revealed an important role for synaptic dysregulation in DEE and identified multiple new causative variants in synaptic genes. Indeed, mutations in various genes associated with neuronal functions like synaptic transmission and recycling, including transporters, neurotransmitter receptors, and ion channels, have all been identified as causative of DEE. In total, pathogenic DEE-causing variants in over eighty-five genes have been identified and more are likely to follow as next-generation sequencing becomes widely available. DEEs comprise a large group of genetically and phenotypically heterogenous diseases that have been difficult to treat. While in many cases the etiology is unknown, de novo heterozygous missense mutations have often been identified as the underlying cause of DEE. Existing pharmacological interventions by way of antiepileptic drugs leave approximately seventy-percent of DEE patients with intractable seizures. Moreover, these pharmacological treatments do not address the cognitive impairments and associated comorbidities caused by the underlying pathophysiological mechanism. In fact, treatment with antiepileptic drugs may actually worsen cognitive comorbidities due to side effects. Additionally, there are no pharmacological treatments for these cognitive comorbidities other than mood stabilizers and antipsychotics. Therefore, alternative approaches to treatments that address the underlying genetic etiology are necessary. Indeed, the recent utilization of gene therapeutic approaches in other genetic disease models such as spinal muscular atrophy (SMA) has spurred the investigation of gene therapies to treat DEEs.

Here, we executed a molecular, behavioral and functional characterization of three preclinical mouse models of DEE involved in synaptic function (Dnm1) and ion channel function (Kcnq3). The human orthologs of the Dnm1 and Kcnq3 genes cause some of the most severe DEE syndromes. Understanding the pathophysiological mechanisms by which mutations in these genes cause disease, is important in identifying and assessing future gene therapeutic interventions.

Patients with heterozygous DNM1 pathogenic mutations present with early onset seizures, severe intellectual disability, developmental delay, lack of speech and ambulation, and hypotonia. For the DNM1 dominant-negative model of DEE, we first characterized the Dnm1Ftfl mouse which phenocopies the key disease-defining phenotypes and comorbidities observed in DNM1 patients. Further, we modelled a gene therapy approach in Dnm1Ftfl mice using an RNA interference-based, virally delivered treatment construct. Dnm1Ftfl homozygous mice showed early onset lethality, seizures, growth deficits, hypotonia, and severe ataxia. Molecular analysis of Dnm1Ftfl homozygous mice showed gliosis, cellular degeneration, increased neuronal activation and aberrant metabolic activity, all indicative of recurrent seizure activity. Importantly, our gene therapy treatment significantly rescued all the severe phenotypes associated with DEE, including seizures, early-onset lethality, growth deficits, and aberrant neuronal phenotypes. Thus, our gene therapy approach provided a proof-of-principle for the efficacy of gene silencing to treat DEEs caused by dominant-negative mutations.

Second, a DNM1 human variant modelled in mice was generated and characterized. The Dnm1G359A mutation, unlike the Dnm1Ftfl mouse-specific mutation has been identified in patients suffering from DNM1 DEE. Thus, this model allows for a more clinically relevant assessment of the impact of a human DNM1 mutation in mice. In the long run, this model will help validate gene therapeutic approaches that may be clinically relevant to DNM1 DEE patients. The Dnm1G359A mutation, like the Dnm1Ftfl mutation, led to early onset seizures, growth deficits, and lethality, establishing the Dnm1G359A mouse model as a viable model to study DNM1 DEE.

In the gain-of-function KCNQ3 model of DEE, Kcnq3R231H mice were characterized molecularly and behaviorally. Patients with KCNQ3 mutations show electrical status epilepticus during sleep (ESES), as well as cognitive and behavioral impairments. The Kcnq3R231H variant led to severe spike-wave discharge phenotype on EEG, decreased maximal seizure threshold, and anxiety-like behavior. Additionally, Kcnq3R231H led to increased localization of Kcnq3 protein at neuronal membranes, suggesting a role for membrane aggregation on disease phenotypes.

Altogether, these findings show the viability of preclinical models of both dominant-negative and gain-of-function mutations in replicating key disease-defining phenotypes associated with severe DEEs. Additionally, the results presented here establish a proof-of-principle demonstration that gene silencing can rescue severe phenotypes caused by dominant-negative mutations in DEE. Future studies on both dominant-negative and gain-of-function models should enable an in-depth understanding of mechanistic implications for each mutation, and lead to gene therapeutic strategies to mitigate the debilitating phenotypes of these DEEs.


  • thumnail for Aimiuwu_columbia_0054D_16922.pdf Aimiuwu_columbia_0054D_16922.pdf application/pdf 4.43 MB Download File

More About This Work

Academic Units
Genetics and Development
Thesis Advisors
Frankel, Wayne
Ph.D., Columbia University
Published Here
October 20, 2021