A Systems Biology Approach to Drug Discovery in Autism

Background:  Autism spectrum disorder (ASD) has high heritability and a prevalence of ca. 1% worldwide, but heterogeneity has made identifying the underlying pathobiology difficult. By focusing on monogenic disorders with high penetrance for causing ASD, common pathobiological pathways might be identified. Phelan-McDermid syndrome (PMS) is one such monogenic ASD-associated syndrome that is caused by haploinsufficiency of the gene SHANK3, which encodes for a scaffolding protein of the post-synaptic density of glutamatergic synapses. While animal models provide great insight into the pathways involved in PMS, some features of the disease may not be captured because of neurobiological variation across species. One approach to deal with this shortcoming is to generate induced pluripotent stem cells (iPSCs) from patients that can then be differentiated into neural progenitor cells and neurons.

Objectives:  Results from a recent study indicate that iPSC-derived neurons from PMS patients show excitatory synaptic deficits similar to those seen in animal models. This provides further support for our hypothesis that expression analysis from such cells can provide valuable insight into the underlying pathology and can be mapped to the expression profiles of FDA-approved drugs to identify candidates for repositioning as novel PMS therapeutics. Therefore, we aim to 1) generate high-quality iPSC clones from PMS patients and siblings; 2) differentiate them into neurons that capture the neurobiological phenotype of PMS in patients; 3) identify PMS-associated differential gene expression in iPSC-derived neurons by RNA sequencing; and 4) identify candidate drugs by comparing gene expression patterns for FDA-approved drugs with PMS-associated expression.

Methods:  Blood samples from patients with PMS and unaffected siblings have been collected for 12 patient/sibling pairs and reprogrammed using a modified non-integrating Sendai virus to express reprogramming factors. At least three clones are selected for each patient after quality control (QC). Clones are then transfected with lentiviruses carrying vectors to induce expression of NGN2 under the control of doxycycline, and the transduced cells are selected with puromycin. Induced neurons are grown for 3 weeks, with mouse astrocytes being added on day 2 to aid in synapse formation, and then cells are harvested and processed for RNA isolation. All samples are subjected to RNA sequencing. The PMS-associated changes in gene expression are then compared to known gene expression profiles of FDA-approved drugs and used to identify candidate PMS therapeutics based on anti-correlation between disease and drug gene expression.

Results:  Three patient/sibling pairs have been reprogrammed and high quality clones have been obtained after QC. The remaining 9 patient/sibling pairs are currently being reprogrammed with many clones also having already passed QC. Neuronal induction has been performed on clones from the initial 3 pairs and is being started with clones from additional pairs as they become available.

Conclusions:  iPSCs from PMS patients offer a powerful tool for disease characterization, drug identification, and screening. Generating an expression profile for these patient-derived neurons will provide a unique perspective on the transcriptional signature of PMS that can be used in conjunction with other models of the disease and known drug expression profiles to identify new therapeutics.