Pleiotropic Effects of Cis- And Trans-Regulatory Mutations

Abstract

Changes in gene expression are an important source of phenotypic diversity both within and between species. Mutations generating variation in gene expression can be cis-regulatory to a particular gene, which typically occur in promoters or enhancers and cause allele-specific changes in expression, or trans-regulatory to the gene, which are mediated by diffusible factors and thus do not result in allele specific changes in expression. Cis-regulatory mutations are hypothesized to be less pleiotropic, or impact fewer traits, than trans-regulatory mutations to the same gene. Also, mutations that are more pleiotropic are hypothesized to more frequently be deleterious than those that are less pleiotropic. Thus, lower pleiotropy could contribute to a preferential fixation of cis-regulatory mutations relative to trans-regulatory mutations over time. Here I test these hypotheses by examining the fitness effects and genome-wide effects on gene expression of cis- and trans-regulatory mutations in the baker’s yeast Saccharomyces cerevisiae. I first use RNA-sequencing data and fitness data for strains of bearing cis- and trans-regulatory mutations to the gene TDH3 to define a distribution of the pleiotropic fitness effects of trans-regulatory mutations relative to cis-regulatory mutations and show that most pleiotropic fitness effects are indeed detrimental. I then compare the extents of the mutations’ impacts on genome-wide gene expression and show that trans-regulatory mutations have a more widespread impact on gene expression than cis-regulatory mutations of similar effect size on the focal gene TDH3. In addition, I show that trans-regulatory mutations have pleiotropic effects on expression of genes affected by changes in the expression of the focal gene itself. I next use gene expression data for a set of of ~1400 gene deletion strains of Saccharomyces cerevisiae to compare the genome-wide impacts on gene expression of cis- and trans-acting deletions for all genes in the dataset, and find that for the vast majority of genes, trans-acting deletions have more widespread effects on gene expression, or are more pleiotropic, than cis-acting deletions. Furthermore, this pattern can be explained by the degree distribution of the regulatory network resulting in highly pleiotropic trans-regulatory factors serving as trans-regulatory deletions to many genes. Finally, I return to the RNA-sequencing data to explore a mechanism of active compensation for reduction in TDH3 expression by its paralog TDH2 that is dependent on the trans-regulators Gcr1p and Rap1p. This compensation occurs when TDH3 expression is lowered via cis-regulatory mutations, but not when it is lowered via mutations in these trans-regulators, resulting in the different downstream effects of changing TDH3 expression in cis and in trans. Together these analyses provide the first empirical description of the pleiotropic effects of cis- and trans-regulatory mutations on both fitness and gene expression.