Biological development is a complex and elegant process that generates the vast array of form and shape found in nature. This process is an example of self-organization, whereby a global pattern of cells is generated by the propagation of local intercellular interactions to larger scales. Understanding this process is one of the fundamental challenges in biology and presents an opportunity to engineer the organization of biological matter. The work described in this thesis has developed a simplified bacterial system to study biological self-organization, focussing on quantifying the influence of physical cellular properties and interactions on global cellular arrangements within colonies. The approach of this project was to use Escherichia coli bacteria, building a range of synthetic genetic tools to generate and regulate cellular properties and interactions, aiming to regulate colony organization. This approach of regulating control over cellular properties within a simplified framework has significant advantages in probing biological self-organization, avoiding the complexity present in natural multicellular systems. A range of microscopy techniques and image processing tools were established to measure cellular properties and cellular arrangements, as well as the dynamics of colony growth. The experimental work was closely integrated with biophysical computational cellular models using the CellModeller multicellular modelling software.

In order to regulate cellular organization within colonies, the first objective was to understand the native organization in E. coli colonies. For this, segmentation was performed on whole colonies, as well as time-lapse microscopy of growing colonies to find the spatial pattern of growth. The computational model was informed using this experimental data, setting the properties of single cells, then used in combination with the data to understand the processes driving global cellular arrangements. The results showed that bacterial colonies contain multiple domains of locally aligned cells, with domain size dependent on cell length. As aligned domains grow, they buckle and break alignment, leading to the development of fractal-like boundaries between lineages of cells. As a colony grows larger, the central region ceases growth, creating a velocity gradient that aligns cells with the radial direction of the colony. The experimentally observed cellular arrangements were also generated in the computational model through the designation of only single cell parameters, indicating the order within colonies was generated by physical interactions.

A toolbox of synthetic genetic mechanisms was constructed to alter cellular and colony properties, alongside computational models of the mechanisms. Genes were used to alter cell shape, resulting in spherical cells of various sizes, as well rod shaped and filamentous cells. Intercellular adhesive interactions were generated through the expression of an adhesin, which generated altered microcolony morphology. Furthermore, synthetic symmetry breaking mechanisms were created to generate distinct domains of gene expression in colonies arising from single cells. Spatial patterns were generated through a system of segregating incompatible plasmids, and temporal patterns of gene expression with a quorum sensing circuit.

Two of the engineered mechanisms were combined to investigate the structural effects of intercellular adhesion. This was done by combining characterized mechanisms for intercellular adhesion and spatial patterning, generating colonies with distinct spatial domains with varying adhesive strengths. Quantification of the morphology of the spatial domains found that adhesion elongated the fractal-like boundary between domains, but only when both domains were adhesive. Modelling further indicated the physical origin of this result, finding the same result in simulated colonies. Time-lapse microscopy of bacterial colonies found that adhesive interactions increased rotational motion during growth, which elongated the boundaries between domains, expanding the area of interaction.

The work described in this thesis has demonstrated that physical interactions can have significant impact on bacterial self-organization, and provides a platform to study and regulate such processes. The work has shown that even relatively simple organisms collectively display a remarkable amount of order, generated by physical processes. Furthermore, the results display a close correspondence of computational models to experimental data, providing a computational platform to study and design self-organizing processes in bacterial systems, which can subsequently be built in vivo. Given that the properties of individual cells are relatively simple to engineer, understanding their propagation to larger scales can radically simplify the engineering of macroscopic biological structure.