UCLA

Animal brains have as few as a couple hundred to as many as tens of billions of neurons. Each neuron in the developing brain recognizes and makes synapses with appropriate partners to build functional circuits. The principles of wiring have been studied in many scales, from cell-fate specification, cell recognition, local signaling to neural activity. Cell surface molecules have been shown to contribute to wiring in different stages and ways. However, the complexity of the brain hindered a larger and detailed search for the molecular underpinnings of wiring, especially synaptic specificity. The three studies presented in this dissertation investigate molecular principles of wiring by profiling gene expression programs of the whole visual system and by referencing synaptic connectome as their final map of wiring choices. I’ll describe in Chapter 1 an overview of anatomical and genomic methods that are developed up to date and an introduction to the drosophila motion detection circuit to contextualize it as a system where the two lineages of methods can converge and be used for developmental studies. Chapter 2 and 3 show the transcriptomic approach to profile gene expression programs of individual cell types in the drosophila visual system and the transcriptomic architecture for synaptic wiring. In Chapter 2, I will present how single-cell RNA sequencing of the motion detector revealed the genetic architecture of the motion detector that reflects its wiring patterns. Eight subtypes of motion detectors derived from the same progenitors were profiled during the wiring period. They did not show cell-type specific genetic markers, instead revealed modular architecture of gene expression programs that correlated with dendritic and axonal connections. In Chapter 3, I’ll present an expansion of genetic effort to build a transcriptional atlas that describes developmental genetic programs of all the neurons in the fly visual system. We found that (i) the gene expression programs that are common to all developing neurons which reflect general developmental progress and (ii) the programs that are cell-type specific which endow cell-type specific characteristics and connectivity. This atlas is by far the largest and the most detailed developmental transcriptome that can be linked to the connectome. In Chapter 4, by converging transcriptome and connectome, I’ll show that the genetic underpinnings of the brain wiring can be identified systematically. We show that transcriptionally closely related neurons making alternative synaptic choices do so via pairs of immunoglobulin superfamily molecules. This study paves a way for further studies in larger brains by showing what can be achieved by combining developmental transcriptome and connectome. As the single-cell transcriptome became a household method and the synapse level connectome is being reconstructed for larger mammalian brains, it will elucidate common and unique molecular principles by applying this framework.