Genome-wide analyses of large cohorts of people with autism spectrum disorder (ASD) have enabled the identification of numerous de novo and inherited rare DNA variants linked to ASD. While there has been a lot of excitement about the progress in gene discovery, translating these findings to an understanding of the underlying pathobiology of ASD has remained largely unrealized. Many studies have been devoted to the categorization of ASD genes according to their biological and cellular function, protein-protein network and human cell type expression, with the aim of identifying points of convergence. Whereas these studies provide valuable groupings of genes/proteins, they do not provide an explicit hypothesis about their mechanistic relationship to the condition as a whole, particularly with respect to specific consequences of these mutations for signal processing in human neuronal networks.

A widely recognized general concept in the field is that mutations in ASD risk genes converge on a disturbed balance between excitatory and inhibitory inputs within neuronal networks (E/I balance). Recently, Nael Nadif Kasri and colleagues have shown that the interrogation of neuronal network activity by growing human neurons derived from induced pluripotent stem cells (hiPSCs) on micro-electrode arrays (MEAs) offers a robust, efficient and physiologically relevant readout for disease-specific network-level phenotypes¹.

In this project, Kasri and his team propose to use their recently established hiPSC-derived E/I culture protocol² for growing defined compositions of excitatory (glutamatergic) and inhibitory (GABAergic) human neurons in the presence of human astrocytes on MEAs, to establish an all-human iPSC-based neuronal network platform for a rigorous and multiparametric assessment of the consequences of specific mutations in ASD risk genes on neuronal network function. Using hiPSC lines obtained from individuals with ASD and/or isogenic hiPSC lines generated by gene editing, they aim to reveal the gene-to-function relationship for 18 distinct ASD risk genes from the SFARI priority list and, employing an all-human iPSC-based MEA platform, to build a phenotypic map based on their network activity. Functional phenotypes will be correlated with transcriptional data by using single-cell transcriptomics coupled to the individual MEA recordings.

This approach will allow them to define subsets of interrelated genes and assess how neuronal networks are influenced by these gene sets. They will extend the phenotypic map of the distinct SFARI ASD risk genes by assessing selected ASD gene variants of unknown significance and for which variants in context with ASD have so far not been confirmed. Finally, they will exploit the virtues of the hiPSC-MEA platform as a first-tier in vitro selection platform for testing  antisense oligonucleotides for the amelioration of neuropathological phenotypes associated with genetically defined ASDs.