Genomic analysis of regulatory networks for bacterial differentiation and multicellular behaviour

Bacteria are single-cell organisms typically viewed as living and acting independently of each other. In fact, most bacteria exist in large communities; a major advantage of this lifestyle is the greatly improved protection of individual cells from environmental stresses such as loss of water and nutrients. Swarming is one of the few main types of community behaviour that we observe across a large number of bacterial species. During swarming, bacteria align themselves in a large group and migrate together over a surface. This is a medically important phenomenon: swarming enables bacteria to travel rapidly to locations in the host organism that are otherwise inaccessible and initiate infections. For example, Proteus mirabilis is a bacterial species that frequently causes life-threatening infections acquired in hospitals. These bacteria swarm over and colonise artificial medical implants to enter the patients' urinary tract and ascend to the kidneys. In addition to its importance in medicine, swarming is an excellent model for studying two fundamental processes in biology: (a) how cells change, or 'differentiate' into distinct cell types that perform specific functions; and (b) how single cells transform into multicellular populations that coordinate their behaviour. Increasing our knowledge of these processes will help us understand how complex, multicellular organisms exist. Cells initiate swarming by sensing contact with a surface and with each other. This triggers a metamorphosis in which cells lengthen 20-fold and build long molecular propellers called flagella that extend outward from the cell surface. These elongated cells (that number in the billions) then align to form large bacterial rafts and migrate away propelled by synchronised flagella rotation. To drive the transition, a complex cascade of molecular signals convert the initial stimuli to activate or repress a specific set of genes required for swarming. We know the identity of many of these genes, such as those responsible for flagella construction. However, it is also clear that several hundred more are involved which are currently unidentified. Additionally, we only have a basic understanding of how incoming molecular signals are transmitted to control the activity of these genes. We will address these problems by building on our and others' previous work in bacterial genetics, bioinformatics and genomics. By using existing approaches and developing new techniques, we will identify the full complement of genes involved in bacterial swarming and uncover the mechanisms controlling them. In doing so, we will reveal new principles underlying cellular differentiation and multicellularity, as well as discover ways to prevent bacterial movement to infection sites.

Bacterial swarming, a paradigm for the developmental processes of differentiation and multicellularity, is an ideal model for systems-level study of regulatory networks. It is complex (involving the interplay of many cellular components), yet tractable (occurring in well-defined, simple prokaryotes amenable to established laboratory protocols). Swarming bacteria migrate over surfaces as coordinated populations, cycling periodically between distinct cell types. E. coli and Proteus exhibit variants of this behaviour that differ in their robustness with Proteus capable of swarming over less favourable surfaces. Molecular studies have identified discrete components required for swarming, but our understanding is incomplete. We propose a general systems approach integrating computational and experimental genomics to study how molecular networks effect a dynamic cellular program. (i) We will use tiling arrays to interrogate the E. coli and P. mirabilis transcriptomes at regular intervals during differentiation and swarming. (ii) We will measure DNA-binding for four orthologous transcription factors (TFs) in the two organisms using ChIP-chip techniques. (iii) We will computationally analyse the array data to compile a full compendium of genes involved in swarming, and characterise potential regulatory targets of the TFs. (iv) We will computationally compare the transcriptome and TF-binding for E. coli and Proteus to identify molecular changes underlying their divergent behaviour. (v) We will computationally integrate our results with public genomic data (eg, regulatory and protein interactions) to define active regulatory networks that coordinate swarming progression. Using established methods in graph theory we will describe cascades of regulatory events driving cellular transition. Our findings will be validated through ongoing molecular studies into swarming, and we will generate vast datasets for in silico modelling and simulation.