Synthetic noise control in eukaryotic gene expression and signal transduction

A certain level of randomness is inherent to every biological process, causing individual cells in a clonally identical population to vary in the number of protein molecules. This variation that was termed gene expression noise arises from stochastic fluctuations and the variability of numbers and states of the involved expression machinery. Noise causes suboptimal protein concentrations, which can negatively affect biological processes. Nature has selected for noise-reducing mechanisms when they benefit cellular fitness. Most synthetic genetic circuits and signaling pathways, however, lack systems that control gene expression noise, which can reduce their functionality. Here, we report the construction of a synthetic noise tuning system in Saccharomyces cerevisiae. We present data acquired by flow cytometry, using a measurement setup that we optimized for minimal nonspecific biological and technical variations. The system we developed allows the tuning of expression noise of a target gene using externally added small molecules to control the transcription rate via inducible promoters and the mRNA degradation rate via inducible ribozyme sequences. We demonstrate the functionality of the noise tuner by achieving up to 3-fold noise differences in the expression of a fluorescence reporter gene. We benchmarked the performance of the noise tuner by comparing it to semi-synthetic systems with fixed mRNA degradation rates, mediated by native yeast terminators. Stochastic simulations of an analytical model that links gene expression to population-level distributions of protein numbers were used to reproduce the experimental findings and revealed the mechanisms underlying the observations: In the given parameter space, noise was mainly affected by the transcription rate, whereas the mean expression was governed by both, the transcription rate and the translational burst size defined by the mRNA degradation rate. The objective of the development of the noise tuner was twofold: the first goal was to reduce gene expression noise in contexts where it proves to be detrimental. The second goal was to establish the noise tuner as a tool to investigate the influence of noise in complex networks. We applied the noise tuner to different genes in the yeast mating pathway, a model signal transduction pathway and the basis for numerous studies in the field of synthetic biology. We determined that the noise tuner, when applied to different genes of the pathway resulted in detectable changes in pathway noise. Detailed analysis of the negative pathway regulator gene SST2 set to either high or low noise resulted in up to 50 % difference in pathway noise between the two settings. We demonstrated that the low noise setting of SST2 expression lead to improved information transmission through the pathway. Categorization of cell morphologies during stimulation with mating pheromone suggested a more precise, switch-like response in the low-noise SST2 cells. To investigate whether noise tuning principle we describe here was also applicable to native genes, we selected five yeast genes with reportedly extreme mRNA production and degradation rates. We used the corresponding promoters and terminators to drive the expression of a reporter gene to observe a general trend towards low noise for genes with high transcription and mRNA degradation rates and vice versa. The results of a gene ontology analysis of the two most extreme cases supported a hypothesis that noise levels are linked to protein function. In this thesis we report the design, construction and successful application of a synthetic noise tuner. Our results illustrate the impact of gene expression noise of individual components on pathway performance – but we also show that this can be controlled. We suggest that design principles for low-noise gene expression, such as those presented in this thesis, should be taken into account for the synthetic modification and de novo design of signal transduction pathways and other networks