The importance of robustness and the regulation of noise as mechanisms that maintain high evolutionary fitness is obvious. Yet, we still have a relatively poor understanding of how robustness is achieved and how is noise being regulated at the molecular level. In model organisms, robustness and evolvability can be studied using experimental evolution approaches. Relative robustness is typically quantified with respect to the change in variation of a trait when the experimental perturbation is applied. In such experiments, the phenotypic outcomes rather than the underlying mechanisms of robustness are measured. With few exceptions, experimental evolution studies have always considered population-average measurements of phenotypes using entire organisms, tissues, or cell cultures. However, as was often suggested in the literature, to truly understand how robustness is established and encoded in the genome, one needs to consider variation in phenotypes across individual cells. To take first steps towards understanding how robustness is regulated in humans, we propose to focus a molecular phenotype, namely gene expression levels. We will characterize the loci in which genetic variation is associated with inter-individual differences in gene expression robustness, validate a subset of such loci, and develop an understanding of the underlying mechanics. To do so, we propose to apply a QTL approach to map inter-individual variation in regulatory noise using single-cell gene expression level phenotypes measured in human induced pluripotency stem cells (iPSCs). Our preliminary results indicate that the level of gene-specific regulatory noise is a trait whose largest component of variation is associated with the individual origin of the sample. Other studies, mostly in yeast, have shown that heterogeneity across single cells in the expression of certain genes is highly heritable and placed under complex genetic control, suggesting that the level of noise in gene regulation may also differ between individuals of multicellular organisms depending on their genetic background. Follow-up studies further demonstrated that gene expression noise mediated by promoter variants could provide a fitness benefit at times of environmental stress in yeast, highlighting the direct role of genetically controlled stochastic cell-cell variation in evolutionary robustness. We will therefore test the hypothesis that the regulation of gene expression noise and robustness is genetically encoded. Specifically, we will collect single cell RNA-seq from iPSCs of 70 Yoruba individuals (Aim 1), map genetic loci that are associated with inter-individual variation in gene expression noise (robustness QTLs;
Aim 2), and study the mechanisms underlying robustness QTLs (Aim 3), by analyzing the single cell data in combination with functional genomic data from the same samples.
The goals of this study are to characterize regulatory noise in single-cell gene expression data, identify genomic loci in which genetic variation is associated with inter-individual differences in regulatory noise (robustness QTLs), and develop an understanding of the mechanisms that underlie gene regulatory robustness.