An Integrated Drug Discovery Platform for Neurodevelopmental Disorders Using Standardized Self-Organizing Human Cerebroids

Project Summary Autism Spectrum Disorders (ASDs) are extremely prevalent and have no cure. These diseases affect the way neurons are specified and communicate with one another. In this project, we will create an in vitro tool to reveal monogenic ASD phenotypes at the level of neuronal assemblies mimicking physiological and pathological neuronal network development in the human cortex, by creating self-organized networks of forebrain neurons on micropattern substrates: the cerebroids. Compared to existing techniques, our platform is extremely standardized and uses tools from the world of artificial intelligence to detect subtle morphological differences associated with the disease state. This uniquely allows us to test large numbers of molecules to discover new therapeutics. This project brings together an interdisciplinary team to develop an innovative solution for a problem that has no solution yet: the discovery of therapeutics for ASDs. Based on our preliminary data showing that neuronal progenitors can self-organize on micropattern substrates and give rise to consistent mutant phenotypes across banks of patient derived cells, we will first optimize our protocols and characterize the normal developmental time course of human cerebroids. This will create a baseline for normal development that will then be used to measure the pathogenic effects of highly penetrants mutations associated to five monogenic forms of ASDs associated with: ADNP, DDX3X, FOXP1, FOXG1 and SHANK3, creating a rich ASD-focused pipeline together with key patient led research foundations. Next, we will develop methods for probing neuronal communication within cerebroids based on live imaging of calcium reporters. This will allow the study of neuronal cell development and interactions at the level of a standardized micro-tissue and bridge an important gap in our ability to model and precisely quantify human neural networks. Finally, we will engage in a drug discovery project for small molecule therapeutics for one ASD gene using, as a primary screen, deep learning mediated morphological readouts of phenotypic reversal, and subsequent evaluation of the hits at the level of neuronal communication with a live imaging strategy. The successful molecules will be strong therapeutic candidates that will eventually be moved towards evaluation in mice models. At the technical level, this research is impactful as it realizes the dream of placing synthetic human micro tissues at the center of the drug discovery process. This will have long- lasting consequences in the way drugs will be found for neurological disorders in a filed currently characterized by large investments yet many failures. New tools lead to new discoveries - and our platform provides a unique opportunity to find new therapeutics in synthetic neural organoids for ASDs.