Extreme Small-World, Modular, and Rich-Club Topology of Single-Neuron Networks in Mouse Primary Visual Cortex

Abstract

Understanding whether the canonical topologies of macroscale connectomes, such as small-world architecture, hub dominance, rich-club cores, and modularity, extend to local cortical microcircuitry has remained challenging due to limitations in simultaneously recording large neuronal populations in vivo. Here, using ultra-large-scale, high-resolution calcium imaging, we tracked spontaneous activity from approximately 2,000 neurons across the mouse primary visual cortex (V1). Across multiple mice and correlation thresholds, V1 neuronal networks exhibited hallmark characteristics of efficient brain organization, but in a markedly intensified form compared to macroscopic brain networks. Local clustering coefficients remained an order of magnitude above random levels, while characteristic path lengths approached those observed in random networks, yielding an exceptionally high small-world index that substantially exceeded typical values previously reported at macroscopic scales. Degree distributions followed a power-law, identifying highly connected hub neurons whose interconnections formed a robust rich-club integrative core. Community detection analyses showed robust modularity upon pruning weak connections, indicating functionally specialized neuronal clusters interconnected predominantly through hubs. These findings provide one of the first direct in vivo evidence that single cortical microcircuits not only recapitulate but intensify network topologies observed at macroscopic scale, implying evolutionarily conserved design principles underlying brain organization from neurons to systems.