Microscale dysfunction and mesoscale compensation in degenerating neuronal networks

Progressive neurodegenerative diseases involve neuronal dysfunction across cellular, circuit, and whole-brain levels. Despite differences in anatomical origins, vulnerable neuronal subtypes, and specific misfolded proteins, these diseases share key features. In presymptomatic phases, neural networks engage compensatory processes to maintain function, including increased centralization and reliance on a rich-club of hub nodes. While such mechanisms have supporting evidence in some disorders, they remain less established in amyotrophic lateral sclerosis (ALS), limiting understanding of potential shared presymptomatic responses. To address this, we investigated structural and functional properties of ALS patient-derived motor neuron networks compared with healthy controls using longitudinal multielectrode array recordings and graph theory-based analysis. We observed microscale dysfunction marked by TAR DNA-binding protein 43 proteinopathy, hyperactivity, and reduced spike amplitude. Structurally, ALS networks exhibited neurite hypertrophy, suggesting attempts to form new connections. Mesoscale analyses revealed functional reconfigurations, including increased rich-club connectivity and network assortativity, indicating compensatory centralization. Our findings provide novel evidence that ALS network features can be recapitulated in in vitro models, and that these networks progressively become more centralized to preserve computational capacity, imposing growing demands on hub nodes and predisposing them to further damage. These results support models proposing common network reconfiguration mechanisms across neurodegenerative diseases.