Title: Quantification and Analysis of Intrinsic Variation in Gene Expression during Embryonic Development
The molecular scale interactions responsible for turning genes on and off have an inescapable degree of randomness — a consequence of their thermally driven motion and their small numbers. As a result, genetically identical cells exposed to identical signals may differ substantially in the level and timing of their transcriptional response to that signal. Such variation poses a challenge during animal embryonic development, where the growth and behavior of many different cells must be tightly coordinated to build a properly proportioned animal. Regulatory mechanisms much exist to minimize it. At other points in development, cellular variation needs to be amplified rather than suppressed, allowing for the formation of heterogeneous tissues without the need for complex spatial-temporal signalling (for example the distribution of different photo-receptor types in your eye). Consistent with these observations, recent research suggests that substantial regulatory mechanisms have evolved in multicellular systems to modulate the degree of randomness in gene expression.
To study such regulation, we need tools to quantify transcriptional differences with single cell resolution in the context of developing animals. Here, I will discuss our progress in developing such tools in the fruit fly and in applying them to understand the developmental importance of some surprising aspects of cis-regulation. I will focus on three experimental methods; single-molecule counting approaches to measuring absolute levels of expression and expression variation among cells, analysis of active transcription frequencies, and measurements of variability in transcription initiation kinetics. The frequency distributions generated from these quantitative data place in many cases provide new insights into the types of regulatory architectures which lie behind them. To facilitate this inference, I will present a mathematical framework to study how the shape of these distributions depends on the type of biochemical mechanisms used to modulate gene expression. Together, these experimental tools and modeling framework can generate bold predictions to further our understanding of the genetic control of development.