Bacteria may be single celled organisms but they very rarely exist as single cells on their own. Instead, bacteria form colonies made up of many cells, all growing and dividing together. These colonies are often ordered in shape and form and various physical systems have been created to model this self-organisation; for example small vibrating rods in close proximity which organise themselves into patterns. However these systems often miss out two of the most crucial factors of bacterial colony development: all the cells are both growing and dividing continuously.

A group of researchers at Cambridge university used synthetic biology techniques to observe the patterns seen in a growing colony of the bacteria E. coli. In order to observe the patterns of cell organisation the genes for either a red, green or blue florescent protein were introduced into the bacteria which were then allowed to grow and develop on a flat surface.

The patterns of growing bacteria: A, B, C and D shown at different magnifications, in the final one individual cells can be seen. Scale bars, 1 mm in A, 100 μm inB, 100 μm in C, 10 μm in D. Image by J. Haseloff and F. Federici.

There are many different reasons why bacteria should form patterns during growth and development. It might be due to signalling molecules moving between the bacteria, adhesive forces holding them together, or the physics of dividing shapes pushing against each other. The researchers used a computational modelling system called CellModeller to see if they could make a system of dividing rod-shapes that produced the patterns seen in the physical colonies.

The model was put together with a series of assumptions. Firstly, it contained only rigid elongating capsules, the same shape as the dividing E. coli. These capsules would grow to a certain length and then divide in half. Each capsule did not move under its own propulsion, but only when subjected to outside forces. Finally they constrained the growth with viscous drag forces caused by the interactions of cells pressing and growing against each other. The CellModeller model was then compared with the naturally growing E. coli, as shown below (the graph on the right shows the fractal dimension measurements of the bacteria and the model):

Comparison between the bacterial growth (labelled BW27783) and the model (CellModeller). Image by J. Haseloff and F. Federici.

As shown by the computational model, these patterns rely not on genetic or cellular features of the bacteria, but on the physical interactions between the growing rods and their surrounding physical environment. The forces of the rods pressing up against each other, and jostling for space as they grow, creates the patterns observed.

In order to explore this further, the researchers used a mutant of their E. coli strain that is round, rather than rod shaped (strain KJB24). These cells grow and divide in the same way as the rod shaped cells, but don't form the same pretty spiky patterns. Instead smoothed domain boundaries are seen, and large disappointing block-shapes instead of the colourful peaks and troughs.

Comparison of the rod-shaped bacteria (left) and the circular bacteria (middle). The graph on the right shows the fractal dimension measurements of the two strains. Image by J. Haseloff and F. Federici.

This model uses physical interactions to create the growth patterns seen in single layer of E. coli cells. Despite the apparent simplicity of the system, emergent fractal patterns are still seen forming. The development of more complex models to include genetic and cellular factors seen in more sophisticated bacterial colonies, such as those exhibiting swarming, swimming or biofilm behaviour could provide a useful way of researching bacterial colony development and growth.


Reference: Cell Polarity-Driven Instability Generates Self-Organized, Fractal Patterning of Cell Layers, Timothy J. Rudge, Fernn Federici, Paul J. Steiner, Anton Kan, and Jim Haseloff, ACS Synthetic Biology 2013 In Press DOI: 10.1021/sb400030p

EDIT: This research has also been covered over at the Oscillator!