Objective: We hypothesized that combining multiple Parkinson’s disease (PD) relevant cellular phenotypes will increase the accuracy, stability, and usability of human midbrain dopaminergic (mDA) neuronal models for drug discovery purposes.

Background: In PD α‑synuclein (SNCA) protein aggregation and mitochondrial dysfunction can act together and cause mDA neuron loss. Current cellular models of PD for in vitro drug discovery often do not take complex pathological interactions into account and focus on a single readout or phenotype.

Method: We modeled the interaction of elevated α‑synuclein levels and mitochondrial dysfunction in patient-derived SNCA gene triplication-carrying induced pluripotent stem cell (iPSC) mDA neurons. We used automated fluorescence microscopy and plate-reader assays to measure multiple cellular phenotypes. Machine learning (ML) classification algorithms were used to differentiate between genotypes and drug-treated neurons. To minimize technical and biological variability we made use of robotic automation, isogenic cell line pairs and cryopreserved batches.

Results: Using microscopic imaging, we identified elevated levels of α-synuclein and its Serine 129 phosphorylated form (pS129), reduced dendritic complexity, and mitochondrial dysfunction in SNCA triplication mDA neurons. Complementing the imaging-based assays, functional plate-reader assays showed increased proteasome activity and decreased mitochondrial membrane potential in SNCA triplication mDA neurons. We applied ML classification and utilized the detected neuronal phenotypic features to accurately classify isogenic mDA neurons according to their SNCA genotype. Furthermore, we show that ML classification is sensitive enough to detect chemical compound treatments that lower α-synuclein levels and positively impact mitochondrial biology. Lastly, we automated procedures in 384-well plate format and assessed our approach in a small-scale drug screen.

Conclusion: To improve PD in vitro modelling we developed a screening-compatible strategy in patient-derived SNCA gene triplication mDA neurons. The computational combination of multiple PD-relevant phenotypes allowed the accurate identification of bioactive chemical molecules with effects on α-synuclein levels and mitochondrial function.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/multidimensional-phenotyping-of-human-stem-cell-derived-midbrain-dopaminergic-neurons-from-a-snca-triplication-carrier-for-drug-screening-applications/