Multimodal Analysis of Cell Types in a Hypothalamic Node Controlling Social Behavior in Mice

Citation

Kim, Dong-Wook (2020) Multimodal Analysis of Cell Types in a Hypothalamic Node Controlling Social Behavior in Mice. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/RGVK-9962. https://resolver.caltech.edu/CaltechTHESIS:12162019-183140887

Abstract

The advent and recent advances of single-cell RNA sequencing (scRNA-seq) have yielded transformative insights into our understanding of cellular diversity in the central nervous system (CNS) with unprecedented detail. However, due to current experimental and computational limitations on defining transcriptomic cell types (T-types) and the multiple phenotypic features of cell types in the CNS, an integrative and multimodal approach should be required for the comprehensive classification of cell types.

To this end, performing multimodal analysis of scRNA-seq in hypothalamus would be very beneficial in that hypothalamus, controlling homeostatic and innate survival behaviors which known to be highly conserved across a wide range of species and encoded in hard-wired brain circuits, is likely to display the more straightforward relationship between transcriptomic identity, axonal projections, and behavioral activation, respectively. In my dissertation, I have been focused on the cell type characterizations of a hypothalamic node controlling innate social behavior in mice, the ventrolateral subdivision of the ventromedial hypothalamus (VMHvl). VMHvl only contains ~4,000 neurons per hemisphere in mice but due to its behavioral, anatomical, and molecular heterogeneity, which T-types in VMHvl are related to connectivity and behavioral function is largely unknown.

In Chapter II, I described my main thesis work to perform scRNA-seq in VMHvl using two independent platforms: SMART-seq2 (~4,500 neurons sequenced) and 10x (~78,000 neurons sequenced). Specifically, 17 joint VMHvl T-types including several sexually dimorphic clusters were identified by canonical correlation analysis (CCA) in Seurat, and the majority of them were validated by multiplexed single-molecule FISH (seqFISH). Correspondence between transcriptomic identity, and axonal projections or behavioral activation, respectively, was also investigated. Immediate early gene analysis identified T-types exhibiting preferential responses to intruder males versus females but only rare examples of behavior-specific activation. Unexpectedly, many VMHvl T-types comprise a mixed population of neurons with different projection target preferences. Overall our analysis revealed that, surprisingly, few VMHvl T-types exhibit a clear correspondence with behavior-specific activation and connectivity.

In Chapter III, I will discuss about future directions for a deeper and better understanding of VMHvl cell types. Briefly, my previous data from whole-cell patch clamp recording in VMHvl slices suggested that there were at least 4 distinct electrophysiological cell types (E-types). Additionally, two distinct neuromodulatory effects on VMHvl were observed (persistently activated by vasopressin/oxytocin vs. silenced by nitric oxide) by monitoring populational activities using two-photon Ca2+ imaging in slices. Based on the results from the first part and combined with advanced molecular techniques (e.g. Patch-seq and CRISPR-Cas9), we can further dissect out the cellular diversity in VMHvl and their functional implications.

Thesis Files