Amyotrophic lateral sclerosis (ALS) is traditionally viewed as a late-onset motor neuron disease, yet how cortical dysfunction originates and contributes to pathogenesis remains unresolved. In this study, we reconstruct the developmental trajectory of cultured cortical networks derived from SOD1G93A mouse embryos using a multimodal approach, by combining morphometric, electrophysiological, pharmacological, molecular, computational, and machine-learning techniques. We prove that ALS neurons fail to acquire mature polarization and connectivity, displaying a transient phase of hyperexcitability that precedes a progressive collapse of network organization. Astrocytic dysfunction emerges early and impairs synchronization, establishing a causal link between glial dysfunction and neuronal instability. The analysis of synaptic transmission reveals an excitatory bias followed by maladaptive inhibitory recruitment and GABA/glutamate co-release, causing fragmented and inefficient network topologies. Finally, in silico modelling identified deficient intrinsic adaptation as a key driver of hyperexcitability. Together, our findings position ALS as a developmentally rooted disorder of cultured cortical network homeostasis, driven by glial, synaptic, and intrinsic adaptation failures. By demonstrating that cortical dysfunction is embedded before degeneration, this work provides a unifying framework connecting early network instability to disease progression and establishes electrophysiological network signatures, detected by machine learning classifiers, as candidate biomarkers for early diagnosis and therapeutic screening.
Lunga et al. (Thu,) studied this question.