Establishment of an optimized and automated workflow for whole brain probing of neuronal activity

Background Behaviors are encoded by widespread neural circuits within the brain that change with age and experience. Immunodetection of the immediate early gene c-Fos has been successfully used for decades to reveal neural circuits that are active during specific tasks or conditions. Our objective was to develop and benchmark a workflow that circumvents the classical temporal and spatial limitations associated with c-Fos quantification. Methods We combined c-Fos immunohistochemistry with c-Fos-driven Cre-dependent tdTomato expression in TRAP2 mice to visualize and directly compare neural circuits activated at different times or during different tasks. Using open-source software (QuPath and ABBA), we established a workflow that optimizes and automates cell detection, cell classification (e.g., c-Fos vs. c-Fos/tdTomato), and whole-brain registration. Results We demonstrated that this workflow, based on fully automatic scripts, allows accurate cell number quantification with minimal interindividual variability. Furthermore, interrogation of brain atlases at different scales (from simplified to detailed) allows for gradual zooming in brain regions to explore the spatial distribution of activated cells. We illustrate the potential of this approach by comparing patterns of neuronal activation in various contexts (two vigilance states, complex behavioral tasks, etc.), in separate groups of mice, or at two time points in the same animals. Finally, we explored software (BrainRender) for an intuitive representation of the results. Conclusions Altogether, this automated workflow, accessible to all labs with some expertise in histology, allows an unbiased, fast, and accurate analysis of the whole brain activity pattern at the cellular level in various contexts.