Navigating Spaces of the Possible: In Vivo Genotype-Phenotype Maps of Gene Regulation

Abstract

Life’s diversity arises from the intricate interplay of multiple organizational layers, from molecular networks to entire organisms, all shaped by continuous evolutionary processes. At the heart of this diversity lies transcriptional regulation, a process driven by transcription factors (TFs) and their interactions with specific DNA regions known as transcription factor binding sites (TFBSs). Changes in these elements can reshape their interaction dynamics and, in turn, gene regulation, facilitating biological innovations that fuel organismal diversity. This thesis investigates the evolutionary origins and potential of TFBSs, focusing on how these binding sites emerge through Darwinian mutation-selection processes and subsequently influence gene expression levels. By combining massively parallel reporter assays with the adaptive landscape framework, I mapped the in vivo genotype-to-phenotype relationships for thousands of TFBS variants, providing insights into the evolutionary dynamics of gene regulation. The first comprehensive in vivo adaptive landscape of a local bacterial TF, TetR, reveals a rugged landscape with 2,092 distinct peaks, many of which are evolutionarily accessible. Extending this analysis to global Escherichia coli transcriptional regulators CRP, Fis, and IHF, I mapped their adaptive landscapes, demonstrating the feasibility of de novo adaptive evolution of gene regulation, influenced by chance, historical contingency, and adaptive biases. Additionally, I explored the exaptive evolution of TFBSs for CRP, Fis, and IHF, uncovering smooth, navigable landscapes where Darwinian evolution can create strong binding sites for different TFs through a few adaptive mutations. This study provides the first in vivo landscapes of exaptive evolution in bacterial TFBSs, highlighting the role of regulatory crosstalk in the diversification of gene regulation. This thesis emphasizes the critical role of integrating modern high-throughput technologies with adaptive landscape theory to advance our understanding of the evolutionary potential of TFBSs in shaping gene regulation. It also provides empirical evidence demonstrating the relative ease with which TFBSs can evolve, whether through de novo processes or exaptation, suggesting that these evolutionary pathways may be more prevalent than previously believed, even in prokaryotes with longer and more complex TFBS architectures compared to their eukaryotic counterparts.