An Integrative Multi-Level Systems Biology Approach to Understanding Enterobacterial Responses to Oxygen

In the 'molecular age', biologists have investigated the mechanisms underlying life by analysing individual components in great detail / tissues, cells, subcellular bits and pieces, and molecules. This approach has been highly productive but new methods and technology now allow scientists to look at the 'big picture' and attempt to understand how all these pieces work together. Bacterial cells provide ideal material for this kind of analysis since they are (relatively) small and simple and we understand a lot about their component molecules. In this project we will attempt to reassemble a picture of how a well-understood bacterium, Escherichia coli, responds to a stress that it will frequently encounter, particularly in its dual lifestyle, alternating between the animal (including human) gut and the outside world. In the former, the bacterium's environment is largely oxygen-free and the organism must adapt its physiology via gene regulation, protein synthesis and metabolic control to utilise other (non-oxygen) respirable substrates. When E. coli leaves the large intestine, it is suddenly faced with a colder, more oxygen-rich but nutritionally less satisfying environment and must make rapid and major changes to gene regulation, protein synthesis and metabolic processes. We will perform careful laboratory experiments in which the bacterium is cultivated under reproducible conditions, cataloguing the changes that occur in the cell when it encounters various levels of oxygen. Then a mathematical simulation (or model) of these processes will be attempted in which we try to integrate the data on molecules and pathways into the 'big picture' (Systems Biology). It is hoped that the model will reveal new aspects of the cell's response and, importantly, highlight those facts that we still need to gather to refine the simulation. The results of this work will be a much better understanding of how a simple cell responds to life-threatening stresses and may inform our approaches to controlling the survival of this and other bacteria in disease, as well as optimising the growth of this organism in commercial bioprocesses. Moreover, the project will stimulate cooperation between biological and mathematical/computing scientists in the UK, the Netherlands and Germany, promoting a concerted research activity in the new area of Systems Biology.

Escherichia coli is able to grow aerobically with oxygen as terminal electron acceptor or anaerobically with numerous alternative oxidants and adapts rapidly to changes in oxygen availability, allowing survival in diverse environments (the gut, water, industrial bioprocesses). Reductionist approaches have been highly successful in characterising the numerous components of the respiratory apparatus (low potential dehydrogenases, quinones, high potential oxidases and terminal reductases) and unravelling mechanisms of gene regulation that underpin respiratory adaptation. However, a holistic understanding of the responses to oxygen, incorporating post-genomic and modeling approaches, is lacking. We seek to describe via a transnational collaborative and interactive approach how functional network modules emerge from molecular interactions, how cellular behaviour emerges from the interplay of network modules, and how population behaviour emerges from the behaviour of a single cell. In Sheffield, we will obtain new internally consistent and quantitative time-resolved transcriptomic, proteomic, metabolomic and biochemical data sets under highly controlled growth conditions in a chemostat and investigate the consequences of mutations in key genes. In parallel, we will contribute to a multi-level mathematical modelling effort. We will develop a simple mathematical model of E. coli respiratory adaptation and enrich this model by incorporating new data. We will use agent-based models, in which each component of the system will be represented as an active agent that behaves according to a set of rules that are triggered by the conditions experienced at any moment in time. Our approach permits interoperability and data sharing with other model frameworks. We will integrate new data from our project partners in the Netherlands and Germany, as well as different hierarchical levels of agent-based, kinetic and reduced-order models.