A high-density microelectrode array integrated microfluidic platform for quantitative investigation of functional connectivity in interconnected human dopaminergic neurospheroids

The quantitative investigation of functional connectivity across spatially organized neuronal networks remains a major challenge in human three-dimensional (3D) in vitro models. Although neurospheroids, brain organoids, and microfluidic systems enable controlled structural organization and guided neurite growth, their integration with high-resolution electrophysiological interfaces remains limited, often requiring post hoc adaptations that compromise recording stability and long-term reproducibility. Here, we present a polydimethylsiloxane (PDMS) microfluidic platform natively engineered for seamless integration with high-density microelectrode arrays (HD-MEAs), enabling the controlled interconnection and long-term electrophysiological interrogation of human neurospheroids within a defined geometry. The platform supports reproducible axonal growth between spatially separated spheroids while preserving full compatibility with HD-MEA recording requirements and electrical stability. Structural characterization revealed the formation of a stable inter-spheroid axonal fascicle enriched in neuronal and synapse-associated markers, supported by ultrastructural analyses across multiple spatial scales. Long-term electrophysiological recordings revealed progressive electrophysiological maturation-associated changes, characterized by increased mean firing rate, sustained and spatially confined neuronal activity within each spheroid, and changes in burst dynamics. Quantitative cross-correlation analysis identified lag-consistent interaction patterns compatible with both local (intra-spheroid) and long-range (inter-spheroid) putative functional connectivity, enabling the distinction between local and distributed network dynamics within the modular system. By combining controlled 3D architecture, stable HD-MEA integration, and quantitative electrophysiological readouts, this platform establishes a robust, reproducible and human-relevant framework for investigating activity patterns compatible with functional connectivity dynamics in interconnected neuronal assemblies, with potential applications in disease modeling and connectopathies-related studies.