hide
Free keywords:
gene expression; expression noise; gene regulatory networks; evolutionary simulations; in silico evolution
Abstract:
One of the keystones of evolutionary biology is the study of how organismal traits change in time. Technological advancements in the past twenty years have enabled us to study the variation of an important trait, gene expression level, at single cell resolution. One of the sources of gene expression level variation is gene expression noise, a result of the innate stochasticity of the gene expression process. Gene expression noise is gene-specific and can be tuned by selection, but what drives the evolution of gene-specific expression noise remains an open question. In this thesis, I explore the selective pressure and evolvability of gene-specific expression noise in gene regulatory networks. I use evolutionary simulations by applying rounds of mutation, recombination and reproduction to populations of model gene regulatory networks in different selection scenarios. In the first chapter, I investigate the response of gene-specific expression noise in gene regulatory networks in constant environments, which imposes stabilizing selection on gene expression level. In these simulations, the expression noise was allowed to evolve over thousands of generations. The probability of responding to selection and the strength of the selective response was affected by local network centrality metrics. Namely, genes with higher centrality metrics had higher probability of responding to selection and a higher reduction in gene-specific expression noise in response to stabilizing selection. Furthermore, global network features, such as network diameter, centralization and average degree affected the average expression variance and average selective pressure acting on constituent genes. In the second chapter, I investigate the response of mean gene expression level and gene-specific expression noise in isolated genes and genes in gene regulatory networks in changing environments. In these simulations, both gene expression level and gene-specific expression noise were allowed to evolve over thousands of generations under directional or fluctuating selection. Gene-specific expression noise of genes increased under fluctuating selection, indicating the evolution of a bet-hedging strategy. Under directional selection gene-specific expression noise transiently increased, showing that expression noise plays a role in the adaptation process towards a new mean expression optimum. In both selective scenarios, target genes, genes regulated by other genes, were more likely to respond than regulator genes. These results show that selection at the network level leads to differential selective pressure at the gene level, and local and global network characteristics of gene regulatory networks are an essential component of gene-specific expression noise evolution. They further demonstrate that increased expression noise can be utilized as an adaptive strategy and that the gene network background imposes evolutionary constraints on the evolution of mean expression level and gene-specific expression noise. These findings represent a step forward in understanding the evolution of gene expression noise in gene networks.
