MRC DTP: Competitive Dynamics in Bacterial Biofilms

Biofilms are social communities of microbial cells that underpin diverse processes including sewage bioremediation, plant growth promotion and plant protection, chronic infections and

industrial biofouling. They are hallmarked by the production of an extracellular polymeric matrix. One of the phenotypic consequences of biofilm formation is that resident microbes are highly resistant to physical stresses and antimicrobial agents. Of particular relevance to this project is that biofilms rarely develop as isolated, single-species colonies. Rather, biofilms typically comprise multiple species and distinct colonies compete with each other for space and resources.

Our aim is to comprehensively define and quantify the physical and mechanical properties of developing and mature biofilms of the ubiquitous soil bacterium Bacillus subtilis. This knowledge will contribute to an understanding of how these biofilms develop and compete in the natural environment and may inform future developments of biocontrol agents and strategies in agriculture.

In this project, we will focus on how single and multi-isolate biofilms form and compete for space and resources using a mathematical modelling approach. Competitive dynamics can be studied at a range of scales and in this project, we will utilize large scale ODE and PDE approaches to determine colony-scale interactions [1,2]. It may also be possible to investigate the theoretical underpinnings of competitive outcomes by adopting a game theoretic approach. At a finer scale, an individual-based approach will be taken. To make rapid progress we will utilise the freely available iDynoMics package. This individual-based-model (IMB) will detail (multiple) individual cells and will be capable of testing hypotheses regarding local heterogeneity in cell-type and matrix production and composition. As details emerge, we can increase the flexibility of the modelling approach by utilising either grid-based site and bond models or indeed grid-free models. The signaling fields can then be linked to the matrix model using a discrete-continuum hybrid approach.