Quantitative Experimental and Mathematical Approaches to Extract Information about Transcription Factor Pulsing

The picture of intracellular information processing is changing from a passive, predictable mechanism to an active, spontaneous dynamic system. Together with an increasing appreciation of the stochastic nature of signaling pathways and regulatory networks, a key discovery in this transition was that many transcription factors show stochastic pulsing: these transcription factors repetitively shuttle in and out of nuclei when cells are in a constant environment. Although functional explanations of stochastic pulsing have been proposed, many outstanding questions remained when I started my thesis research: What are the key regulators that lead to pulsing? Is the regulation encoded in the amino acid sequence of these transcription factors? Is stochastic pulsing beneficial to the organism or a byproduct of other biological functions? How does stochastic pulsing evolve? Two systems biology approaches have been widely applied to study stochastic single-cell dynamics: mathematical modeling, which helps us explain complex cellular phenomena, and quantitative experiments, which help us falsify model predictions. Hence, my thesis research aims to understand the mechanism underlying stochastic pulsing through quantitative, mathematical, and experimental approaches. I first characterized the mechanism at the level of the signaling pathway by using a stochastic pulsing transcription factor, Crz1, which is downstream of the calcium signaling pathway. I found that Crz1 shows damped oscillations after rapid fluctuations in cytosolic calcium concentration. Using a stochastic model, I found that the stochastic pulsing of Crz1 can be explained as a synchronous molecular movement with delay to upstream signals. I next characterized the mechanism at the level of peptide sequence by taking the advantage of wide sequence variation among Crz1 orthologues. By developing a reporter system for the calcium/Crz1 pathway, I found that the intrinsically disordered regions (IDRs) of the orthologous Crz1 show differences in their signaling dynamics. I also found that a few evolutionary changes in the IDRs are sufficient to introduce pulsing and improve fitness, supporting the hypothesis that pulsing is beneficial. Finally, I contributed to developing inference methods for quantitative studies on stochastic signaling pathways, which are known to include IDRs. Starting with the high osmolarity glycerol (HOG) pathway, I quantified the effects of conserved elements in IDRs of the HOG pathway. I developed a method to fit data with non-symmetrical errors and demonstrated the current limit of model interpretability. Overall, my thesis clarifies key challenges in systems biology and provides a promising direction to investigate the mechanism of stochastic signaling pathways.Ph.D