Abstract
Neuronal mitochondria display distinct morphologies across compartments, with dendritic mitochondria being elongated and axonal ones shorter, and their morphologies are dynamically changed via fusion and fission machineries. Mitochondrial structural abnormalities are common in neurodegenerative diseases such as Alzheimer's disease and Parkinson's disease, yet systematic evaluation of therapeutic targets remains limited. Here, we tested key mitochondrial shape regulators, mitofusin 1/2 for fusion and Mff/Fis1 for fission, in an α-synucleinopathy model. Using MitoVis, a deep learning-based neuronal mitochondrial image analysis tool, we achieved rapid, compartment-specific analysis of mitochondrial morphologies. Among all interventions, Fis1 knockdown most effectively protected mitochondrial structure to control levels without inducing over-elongation of axonal mitochondria, which was linked to abnormal Ca(2+) dynamics. While all manipulations preserved dendritic spine loss, Fis1 optimally maintained axonal mitochondrial function. These findings demonstrate a high-throughput screening approach for mitochondrial regulators and highlight Fis1 as a promising preventive/therapeutic target. Our results support targeting mitochondrial morphology as a viable strategy for treating α-synucleinopathy and potentially other mitochondria-related neurodegenerative diseases.