Functional Recording of Neuronal Network Activity in Cortical Neurons Differentiated from Human Pluripotent Stem Revealed Mutation Dependent Patterns of Abnormalities in SHANK3 Haploinsufficiency Associated ASD Patients

Background: Human induced pluripotent stem cells (hiPSC), obtained by reprogramming of somatic cells of individuals with ASD, provide a promising route for in vitro modeling of the pathology by combining relevance, robustness and high availability of the biological resource, and offering a unique opportunity for powerful drug discovery approaches. One of these approaches relies upon using live cell imaging to study developing cortical networks differentiated from control or ASD iPSC in order to characterize functional abnormalities associated with the emergence of the symptoms. This will help understanding the impact of different genetic mutations on neuronal function and discover new, personalized pharmacological treatments.

Objectives: As a proof of principle we choose to model ASD associated with SHANK3 gene haploinsufficiency, also known as Phelan-McDermid syndrome, and compared the neuronal network activity of SHANK3-/+ neurons with controls in the context of different SHANK3 mutations.

Methods: We derived neuronal stem cells (NSC) from hiPSC of 4 controls and 5 ASD individuals bearing different heterozygotous mutations in SHANK3 gene. Each NSC line was differentiated into cortical neurons for up to 60 days. Neuronal phenotypes were analyzed using immunocytochemistry (beta-III Tubulin as a neuronal marker coupled with GABA and Glutamatergic markers) on days 14 and 28. Neuronal network activity was quantified using a fluorescent calcium probe Fluo4-AM from day 32 to 60. Calcium imagines were recorded during 3 min using the automated imaging system ImageXpress micro (Molecular Devices) and analyzed using FluoroSnap Software (Tapan Patel, University of Pennsylvania).

Results: This revealed that synchronized firing started in hiPSC derived neurons from day 50 where almost all the sampled cells participated in the synchronized firing. We observed synchronized firing defects in all SHANK3 +/- neuronal networks but we demonstrated that the type of defect differed with the type of mutation, suggesting that these defects are dependent of the type of isoforms spared by the mutation. Being totally amenable to automation, this read-out will be used to test the therapeutic potential of several pharmacologic compounds, more particularly, their ability to restore normal firing in SHANK3-/+ neuronal networks.

Conclusions: This approach may provide a pathway to an efficient stratification of ASD patients and a better personalization of their treatment.