Ya-Hui Chou｜Institute of Cellular and Organismic Biology, Academia Sinica

Although sculpted by similar developmental programs, individuals in a given species exhibit great variation in their genetic composition and behavioral patterns. How genetic, environmental, and experience-dependent factors collectively shape our brain circuits (thus individuality) is one of the central questions of biology. Our research attempts to understand how neuronal diversity and variability arise, how they contribute to neuronal circuit activity, dynamics, and behavioral outputs, and how experience shapes the variability of the neuronal circuit. Specifically, we try to explore the underlying mechanisms involved in neuronal variability in the fruit fly olfactory system, with a particular focus on the development, organization, variation and function of the local interneuron network. 

Development of local interneuron diversity. Olfactory local interneurons (LNs) form extensive synapses with different neurons in the antennal lobe (AL) to modulate odor information and exhibit extensive diversity and variability. We have built the developmental map of LNs and reagents to label several different types of LNs (Fig. 1). Three unbiased forward genetic screens to search for molecules involved in the LN development and wiring were conducted. Currently we are exploring the development of distinct types of LNs and the underlying molecular mechanisms. 

Local interneuron variability. The fact that Drosophila LNs exhibit variability across individuals offers an unprecedented opportunity to study the mechanisms that give rise to neuronal variability and the consequences of such variability on behavior. We recently demonstrated that a single identified LN can exhibit as many as 849 distinct innervation patterns, which are the results of developmental noise, sexual dimorphism, age and experience-induced plasticity. We are currently exploring the variability of distinct types of LNs with the reagents we built.

Modeling the function and dynamics of local interneuron network. With the powerful genetic tools we built, we can temporally label and manipulate single local interneurons. We are using calcium imaging system, optogenetic systems, different self-built behavioral paradigms (Fig 2), machine learning and mathematical modeling to address whether LN innervation variability leads to any variation in circuit activity, dynamics, and behaviors.