A method for extracting an approximated connectome from libraries of single cell reconstructions

Understanding how the neuronal circuit organization supports the spatio-temporal patterns that characterize the brain’s neuronal activity is still an open challenge. Despite a large number of approaches available to record and modulate neuronal activity at cell resolution and in living animals, limited data are available to map the functional information into a circuit wiring diagram supporting possible circuit working mechanisms. When available, the analysis of electron microscopy based high-resolution connectomes, leveraging synapse annotation, allows the unraveling of portions of the nervous system wiring diagram or specific circuit motives. However, even assuming a complete annotation of the synapses, extracting the general organization principles of the neuronal networks across the brain remains a challenging effort. In order to extend the available methods, we present an approach to reconstruct an approximated brain connectome starting from libraries of single cell reconstructions belonging to or co-registered in the same anatomical space. By leveraging both the Strahler numbering of the nodes characterizing the cell morphological reconstructions and a proximity range criterion, we inferred the general connectivity structures between the different cells, bypassing the need for synapse annotation. We applied this approach to extract an approximated connectome of the zebrafish larvae brain from a light microscopy-based dataset of about 3-thousand co-registered neuronal skeletonizations. Modularity analysis of the retrieved connectome provided a representation of the resulting graph organized in hierarchical structures, with neuronal modules capturing precise and topographically organized connection patterns mirroring identified functional circuit motives. In conclusion, we present a scalable, from-circuit-to-brain range approach amenable to revealing the neuronal architectures supporting brain mechanisms.