Comparative Connectomics: Revealing change and conservation in whole brain circuits across evolution

Animals display remarkable diversity in their behaviors, yet very little is known about how nervous systems evolve to generate suchvariation. CircuitEvolution will answer two fundamental questions:1. How does brain wiring change across species as novel behavioral traits evolve?2. What core features of brain circuits are deeply conserved across species?We will use electron microscopy connectomics, the only method that enables unbiased, brain-wide structural analysis of the synapticconnectivity between all neurons. Building on recent rapid progress in Drosophila melanogaster, the frontier of whole brainconnectomics, we will target drosophilids at a strategically selected range of evolutionary divergences. Two PIs (Cardona, Jefferis) arepioneers in Drosophila connectomics, and our third PI (Ruta) and collaborators (Prieto-Godino, Benton) have pioneered moleculargenetic investigation of behavioral and circuit evolution across drosophilids. Crucially, with our fourth PI (statistical physicist, SalesPardo), we will identify conserved and divergent neurons and circuits across increasingly distant species, developing new inferencemethods for connectome alignment, simplification, and neuron annotation. Working together on this project we can develop a newfield of whole-brain comparative connectomics.With cutting edge technology, we will obtain and share multiple connectomes for 7 species. These unique datasets will allowcomparisons across species, sexes and developmental stages and support three areas of biological investigation within this grant:olfactory circuits, courtship behavior and the conserved core of the insect brain. This will generate many hypotheses that will betested experimentally by the PIs, collaborators and laboratories worldwide. The insights and methods will have a major impact onunderstanding conserved and evolving circuit function across phyla all the way to mammalian brains and impact the design ofartificial intelligence systems.