Approaches to determining the three-dimensional structure and dynamics of bacterial chromosomes

Abstract

The information in genomes is only partially contained in the linear sequence of their nucleotides. Their folding into dynamic three-dimensional structures creates spatial relationships between loci that likely play important functional roles. Yet so far only the broad outlines of this spatial organization have been discerned. In chapter 2 of this thesis I describe a general constraint-based framework for defining the configuration space of chromosomes. Analogous to protein structure determination through NMR, such a framework allows the quantitative reduction of the conformation space down to the level of a single structure or an ensemble of structures. It is compatible with both experimentally determined and theoretical constraints, particularly those motivated by evolutionary optimality. In chapter 3., I describe the first method to search for signals of large-scale three- dimensional structure in genome sequences. The results suggest that there is strong selection for three-dimensional relationships within the chromosome, particularly those related to transcription. The signals generated recapitulate both known structural data from microscopy and functional data on genome-wide transcription levels.

(cont.) Moreover, a detailed analysis of these signals in E. coli suggests previously unknown structural features including chromosome-long periodic looping and an axis of high transcriptional activity. There are immediate applications to other bacteria and potentially to eukaryotes.