Chemotactic Sensing towards Ambient and Secreted Attractant Drives Collective Behaviour of E. coli

We simulate the dynamics of a suspension of bacterial swimmers, which chemotactically sense gradients in either ambient or self-secreted attractants (e.g. nutrient or aspartate respectively), or in both. Unlike previous mean field models based on a set of continuum partial differential equations, our model resolves single swimmers and therefore incorporates stochasticity and effects due to fluctuations in the bacterial density field. The algorithm we use is simple enough that we can follow the evolution of colonies of up to over a million bacteria for timescales relevant to pattern formation for E. coli growing in semisolid medium such as agar, or in confined geometries. Our results confirm previous mean field results that the patterns observed experimentally can be reproduced with a model incorporating chemoattractant secretion, chemotaxis (towards gradients in the chemoattractant field), and bacterial reproduction. They also suggest that further experiments with bacterial strains chemotactically moving up both nutrient and secreted attractant field may yield yet more dynamical patterns.