Modelling bacterial biofilms in spatially heterogeneous environments

Abstract

Biofi lms are communities of one or more species of microorganism which have adhered both together and to a surface. Biofi lms are ubiquitous in nature, with up to 80% of bacterial life on earth estimated to be found in a biofi lm. Bacterial biofi lms are far more resilient to both chemical and physical methods of removal than their planktonic counterparts, which presents numerous challenges in both clinical and industrial scenarios. Therefore, further research into the underlying mechanisms of how these biofi lms develop and survive is essential. This thesis aims to do so via the implementation of various computational modelling techniques.

Currently, most computational modelling of biofi lms is done under somewhat idealised conditions, such as uniform antibiotic concentrations and mono-species bio lms, which do not always reflect the complex conditions found in vivo. This thesis therefore also aims to address this problem by using computational models to understand how biofi lms proliferate and resist methods of removal in spatially heterogeneous environments, such as chemical gradients of nutrients and antibiotics, or non-uniform flow fi elds. The thesis takes the form of three distinct projects, which are linked together by this common theme of spatial non-uniformity.

Presented fi rst is an investigation into the coupling between nutrient availability and growth-dependent antibiotic susceptibility. This project uses a simple 1D Monte-Carlo model to simulate the advancement of a bacterial population along a spatial antibiotic concentration gradient. Bacterial replication consumes nutrients which in turn lowers the local growth rate, altering the antibiotic susceptibility. The results highlight the differing outcomes for antibiotics which target either slow-growing or fast-growing cells.

Following this, the next project investigates the initial stages of biofi lm formation on a surface. This chapter involves a pair of complementary models, a deterministic one, involving a system of differential equations; and a stochastic one, where the individual bacteria are simulated using a modifi ed Γ-leaping algorithm, both again in 1D. By modifying the rates for certain actions which the bacteria undertake, the models predict that under certain conditions biofi lm formation is highly predictable, but for other parameter regimes, bio lm formation becomes more stochastic.

In the third project, the stochastic biofi lm formation model described above is extended to develop a model for the formation of biofi lms on a surface which leaches an antimicrobial compound into the surrounding environment, similar to current antifouling coatings used to prevent marine biofouling in the shipping industry. A key difference in this model is the inclusion of multiple bacterial species, each with differing resistances to the applied biocide, intended to represent the biodiversity found in a typical marine environment.

Finally, a computational fluid dynamics model is presented, which is used to model the interaction between a micro-structured surface featuring shark skin-like riblets and an enveloping biofi lm, when exposed to an external flow fi eld of various incident flow angles. These riblets are a contemporary solution to reducing hydrodynamic drag, e.g., on ship hulls, but are only effective when their physical shape is unobstructed. Investigating how misaligned riblets can impede, or even prevent, the sloughing of bio lm matter is therefore crucial to optimising their performance.