Aberrant Local Synchrony in Distinct Mouse Models of Epileptic Encephalopathy

Abstract

AbstractIdentifying and quantifying synchronous activity of primary neuronal networks using multielectrode arrays (MEAs) can potentially provide a medium-throughput platform to screen potential therapeutics for genetic epileptic encephalopathies (EEs). However, successfully identifying screenable synchrony phenotypesin vitroposes significant experimental and analytical challenges. Primary neuronal cultures quickly become highly synchronous and certain measures of synchrony tend to peak and plateau, while other network activity features remain dynamic. High levels of synchrony may confound the ability to identify reproducible phenotypesin vitrofor a subset of EEs. Reducing, or delaying the onset of, high levels of synchronyin vitromay increase the dynamic range of global synchrony measures to identify disease-relevant phenotypesin vitro,but such measures have not been established. We hypothesized that an emphasis on local (nearby) connectivity could elucidate reproducible disease-relevant synchrony phenotypes in cortical cultures not identified by current approaches. We show clear evidence of enriched local synchrony in 48-well MEAs that varies in amplitude during development of neuronal networks. Then, we show new topological-based measures are capable of identifying novel phenotypes of aberrant synchrony in distinct mouse models of EEs. Such topological synchrony measures may provide screenable phenotypes for certain brain diseases and may be further enhanced by experimental innovation reducing global levels of synchrony in primary neuronal networks.SignificanceIn vitrosynchrony phenotypes may provide disease-relevant features that can be used for screening potential therapeutic candidates for epileptic encephalopathies. Here, we incorporate inter-electrode distance to generate tools capable of identifying novel synchrony phenotypes in distinct neurodevelopmental disorders. We additionally report robust topological and globalin vitrosynchrony phenotypes, alongsidein vivosynchrony phenotypes inStxbp1+/-mice. While singular features of disease in anin vitromodel are unlikely to effectively test therapeutic candidates, compounds that reverse a larger subset of distinct features may translate to human patients, suggesting such a model may be ideally suited for therapeutic development using MEAs. Across multiple disease models, the topological tools developed here are complimentary to and expand upon those within meaRtools (Gelfman 2018), which is a suite of computational tools to identify network phenotypes using MEAs.