High Precision Single-Molecule and Single-Cell Measurements: DNA Polymerase Fidelity and Correlated Transcriptional Fluctuations

Abstract

Genome sequence and expression are highly dynamic. Understanding the forces that drive these changes is an important step in rationalizing the connection between an organism’s genome and its phenotype This thesis focuses on the development of next-generation sequencing assays to characterize DNA polymerase fidelity, which influences genomic variability, and correlated transcriptional fluctuations, which reveals the underlying gene co-regulation network controlling gene expression changes. DNA polymerases are essential for maintaining genetic stability. Although the general enzymatic mechanisms influencing fidelity are well understood, its contribution to DNA sequence variation is poorly characterized. Chapter 2 describes the development of a high sensitivity barcoding assay to measure the in vitro mutation spectrum of DNA polymerases. We used this to reproducibly measure the variations in error rate that occur over different template sequences and identify mutation hotspots. This assay has the potential to further extend our understanding of DNA polymerase fidelity into the genomic context. Transcriptional regulators coordinate the timely expression of thousands of genes within the genome. Since co-regulated genes are often functionally related, this co-regulatory network holds crucial information to guide our functional understanding of gene expression dynamics. Chapter 3 motivates a conceptual framework to reveal these co-regulatory networks by measuring correlations in steady-state gene expression fluctuations. To measure these correlations, we developed a novel single-cell RNA-seq method called Multiple Annealing and Looping Based Amplification Cycles for Digital Transcriptomics (MALBAC-DT) with substantially improved transcription detection efficiency and accuracy. We found numerous, highly functionally enriched and cell-type specific correlated transcriptional module (CTMs) from sequencing homogenous population of human cells. These CTMs are predictive of transcription factor binding targets and that they can be used to refine our understanding of differential gene expression to the pathway level. Chapter 4 shows the application of this correlational analysis to the non-steady-state system of mouse early embryo development during the preimplantation stage. We demonstrated that such correlations can be predictive of important developmental regulators and verified our predictions through perturbation experiments.