Reverse Engineering and Quantifying Context Effects in Synthetic Gene Networks

Abstract

In the first part of this thesis, we undertake a quantitative investigation of how compositional context, the spatial arrangement and relative orientation of genes, affects individual gene expression in a genetic network. Taking a synthetic biology approach, we construct a series of simple two-reporter biocircuits, each expressing either an mRNA aptamer or a fluorescent protein, and show that by varying the relative orientation of the two genes we obtain a wide range of gene expression profiles, including context-dependent bimodality. We develop a mathematical model to describe the experimental trends observed based on concepts from DNA supercoiling theory. We validate the model through a series of in vitro supercoiling experiments and show that by relaxing positive supercoiling in the plasmids, we can significantly reduce the context effects in gene expression. Most importantly, these insights provide a framework for understanding how compositional context and supercoiling can impose feedback on the intended architecture of a synthetic gene network. As a proof of concept, we engineer a genetic toggle switch exploiting compositional context effects to improve its threshold detection and memory capabilities. In the second part of this thesis, we examine a series of theoretical and computational tools from dynamical systems theory that assist in engineering novel biochemical reaction networks. We briefly review the concept of dynamical structure functions and network reconstruction as tools for understanding biochemical reaction networks. In particular, we review the concept of resource-loading, show that resource-loading can lead to coupling interactions among biochemical species, and that by estimating a dynamical structure function from experimental data, it is possible to quantify resource loading effects in practice. We illustrate the importance of knowing these loading effects through several example systems, showing that crosstalk imbalance in feed-forward loops can lead to performance limitations. However, since biochemical reaction networks are generally large, in practice, only portions of the global network can be reconstructed at a time. We show, with a combination of theory, simulation, modeling and experiments, it is possible to reconstruct the dynamical structure function of a large-scale biochemical network using a series of network reconstruction experiments. We then demonstrate how the dynamical structure function can be used to analyze context interference and how these perturbations interfere with performance. We illustrate these ideas with several classes of standard biological networks, e.g. autocatalytic systems, cascade systems, and input-coupled systems. Finally, in the third part of this thesis, we consider models for context interference in stochastic chemical reaction networks. We address the problem of representing a biological system and its environment using a stochastic modeling framework. We first introduce a decomposition of the global chemical reaction system into two systems: a system of interest and its environment. We then present and derive a decomposition of the chemical master equation to achieve a representation describing the dynamics of the system of interest, perturbed by an environmental disturbance. We use this decomposition to model examples of two types of environmental disturbances: the disturbance a system experiences through loading effects from limited resources and the disturbance a system experiences when perturbed by an antibiotic that modifies transcription or translation rates.