Title : Early network-level signatures of Parkinson’s disease in human midbrain organoids
Abstract:
Human midbrain organoids are 3-dimension neural tissues derived from iPSC-stem cells. As they self-assemble into layered neural networks, these models capture key aspects of developing human dopaminergic circuitry in a controlled in-vitro setting, enabling mechanistic studies that are difficult in real patients. One important application is Parkinson’s disease, a neurodegenerative disorder marked by dysfunction and eventual loss of dopaminergic neurons. Specifically, by modeling these processes in a human-derived system, midbrain organoids offer a powerful platform to detect early electrophysiological abnormalities and to study how disease-like perturbations reshape neuronal networks. Here, we analyzed midbrain organoid multielectrode-array recordings (MaxTwo; Sharf Lab) from healthy controls and 6-hydroxydopamine (6-OHDA)–treated organoids. Spiking activity was isolated using KiloSort2 and quality filtered to ensure robust neuronal signals. Spike trains were then converted into functional connectivity networks, enabling comparison across complementary axes: network complexity, topology, and directed information flow. Across analyses, PD-induced organoids exhibited reduced coordinated network computation. Higher-order interaction metrics revealed a shift toward greater redundancy and reduced synergistic signaling relative to controls (p < 0.0001). Network topology was weakened, consistent with disrupted hub-like structure and reduced balance between integration and segregation, while directed communication measures suggested less efficient signal propagation. To assess early detectability, an autoencoder trained on control recordings learned baseline dynamics and flagged treated samples as deviations within short windows (~120s, 94% AUC). Together, these results indicate that PD-like perturbations produce quantifiable network-level dysfunction detectable from organoid electrophysiology, supporting brain organoids as a scientific resource for further biomarker discovery and research.
