Quantitative transcriptional input/output relationships in the Ciona notochord

Abstract

Embryonic development proceeds under the control of an intricate network of signaling molecules and transcription factors that control the spatial and temporal dynamics of gene expression. These gene regulatory networks depend on enhancers and related cis-regulatory DNA sequences that integrate information from multiple upstream regulators to control the transcription of individual downstream genes. This integration involves precise relationships between input signals and output responses, and while the direct activators and repressors of many enhancers are known, the quantitative details of these input/output relationships are poorly understood. The invertebrate tunicate chordate, Ciona robusta, is an excellent model system for examining transcriptional input/output relationships thanks to a small sequenced genome, rapid embryogenesis, simple and stereotyped cell lineages, and straightforward transgenesis by electroporation. In this dissertation, I leverage these advantages to address multiple questions about tissue-specific transcriptional regulation using the Ciona notochord as a model. RNA-seq on notochord-enriched cell populations and in situ hybridization validation of those data is used to establish a comprehensive Ciona notochord transcriptome and identify new expression domains in subregions of the notochord. I dissected the enhancer regions of a gene specific to the posterior notochord and identified upstream regulators of this novel expression pattern. This cis-regulatory analysis identified multiple regions capable of driving the same expression pattern, consistent with growing evidence that seemingly redundant distributed enhancers are common. I developed a reporter-based assay based on the graded pharmacological inhibition of an upstream signal to quantify input/output relationships for two distributed enhancers of the key notochord regulator, Brachyury. Despite driving similar notochord-specific expression patterns, I found that these two enhancers have fundamentally different dose-response relationships to MAPK signaling activity. This indicates that they are not truly redundant genetic elements and supports a model in which distributed enhancers play an important role in shaping transcriptional input/output curves.