LAUSR

Abstract Psychiatric disorders are etiologically heterogeneous and complex. For example, it has been demonstrated that hundreds of genetic loci contribute to risk for autism or schizophrenia. These loci include rare variation of major effect as well as common variation: Rare variants of major effect can be a primary contributor to risk in a given individual, while common variation functions in a multifactorial context, including polygenic risk and other factors. Modeling etiologically complex disorders can be quite challenging. For rare variants, targeted disruption or mutation of a gene in cell and animal models will have strong construct validity. However, in the case of oligogenic and polygenic risk, such approaches are less relevant. Study of human samples from patients with oligogenic or polygenic risk represents a means of understanding pathobiology in such cases. However, given the etiological heterogeneity and the small effect size of common variation, large and even ultra-large cohorts will be required before studies are sufficiently powered. Over the past 5 years, we have been developing large cohorts of human samples from psychiatric disorders, while also piloting methods to robotically reprogram and differentiate these samples. We are taking three approaches to sample collection. First, for rare genetic disorders such as Phelan-McDermid Syndrome (PMS) – a severe neurodevelopmental disorder that presents with intellectual disability, autism, and additional neurological and psychiatric manifestations, we are collecting case-sibling pairs (n~18) for reprogramming. Second, idiopathic disorders, such as autism, schizophrenia, and PTSD, we are collecting large sample cohorts, targeting hundreds of samples for each disorder. Finally, in disorders where there are known or suspected dysregulated pathways, we can increase power by challenging cells with perturbing agents, including IGF-1 in PMS and cortisol in PTSD. In parallel with collection, we have piloted robotic reprogramming methods with these sample sets. NYSCF has developed automated protocols for fibroblast expansion, iPSC reprogramming and validation. In pilot studies with our schizophrenia cohort (n>120), ~10% of the total sample is in the final stages of reprogramming on this automated platform, hence ensuring feasibility. For each individual, two clonal iPSC lines are validated and banked. Furthermore, we have developed robotic methods to induce neurons from iPSC. In pilot studies, we have already evaluated two automated protocols for rapid induction of excitatory neurons induces either iPSCs or NPCs into functional neurons with nearly 100% yield in less than 2 weeks by expression of doxycycline-inducible Ngn2 lentiviral vectors, combined with puromycin selection to increase the purity of the cultures. Over the next 3 years, we anticipate reprogramming, differentiating, and analyzing hundreds of iPSC lines derived from patients with autism, schizophrenia, PTSD, or rare genetic disorders associated with autism or schizophrenia.