Genetic Dynamical Systems
Recombinant DNA technologies starting from PCR and enzymatic assembly, through to CRISPR, and de novo synthesis have generated an explosion of interest in engineering and using biology as technology. Some of the earliest successes in synthetic biology were in the creation of novel dynamical systems based on transcription regulation. By arranging genetic elements such as promoters and transcription factor coding sequences, researchers were able to create simple two-node (toggle switch) and three-node (oscillator) genetic dynamical systems. However these successes have proven difficult to capitalise on or extend to larger systems for lack of reliable methodologies for predicting circuit dynamics from measurements of genetic parts.
A major issue is the dependence of genetic parts on the host cell, its physiology and environment. In fact, DNA parts do very little outside the cell and are completely dependent on its machinery (RNA-polymerase, ribosomes, etc.) for their operation. The cell in turn is dependent on resources from its environment (sugars, amino acids, etc.). In this view it is clear that predictive measurement and analysis of genetic parts must include models of their interaction with the host cell, and indirectly its growth environment. Based on previous work in the Haseloff lab [LINK] at Cambridge University, and in collaboration with the Federici lab (LINK) we are developing methods using two-channel fluorescence measurement and ratiometric analysis based on biophysical and other mathematical models. The aim of this work is to extract intrinsic parameters of genetic parts that can be used to predict their operation in the context of different genetic dynamical systems, and different growth environments. We view the entire system, cells, media, and genetic constructs as a dynamical system and use methods such as parameter inference to characterise its behaviour.
One motivation for engineering genetic dynamical systems is to use them to engineer processes for multicellular coordination, as described by DPMs (see Multicellularity). We study the dynamics of genetic constructs in multicellular bacterial populations, colonies, or biofilms. We use time-lapse microscopy to measure growth and gene expression in Escherichia coli colonies (see figure 1), and computational modelling to simulate these processes. These approaches are combined with synthetic biology to create simple gene networks and analyse their spatio-temporal dynamics in relation to the physics of cell population growth. We continue to work on CellModeller (cellmodeller.org, developed in the Haseloff lab, Cambridge) and are developing simpler computational models. These models will enable large-scale exploration of the space of possible regulatory networks – the ‘design space’ of synthetic multicellularity. We use methods from machine learning, evolutionary algorithms and parameter inference with the aim of elucidating principle and engineering methodologies for multicellular coordination.
In close collaboration with the Federici lab (LINK), we are also applying these genetic networks to regulation of multicellular coordination processes such as cell growth rate, cell-cell signalling, cell shape and size. Using these techniques in combination with the computational modelling mentioned above we aim to generate synthetic morphogenetic systems that reliably create specified pattern and shape.