The ability to redesign and even build completely new biological entities offers revolutionary opportunities for using biology to solve human problems. The dream, however, far outstrips the reality: engineering new biological machines, programming cells behaviors, and building new forms of life pose huge technical and conceptual challenges. Our research is designed to make some important first steps, by developing generally-applicable engineering methods for building synthetic gene networks. Implementation of even the most simple circuits in a biological system requires tedious optimization of a large number of poorly-understood parameters, many of which can neither be measured nor easily manipulated. Simulations can sometimes guide optimization, but we believe that biological systems are best optimized using nature's editing strategy, evolution. We believe that a combined approach of rational design based on computational predictions coupled with directed evolution-making mutations in the laboratory and selecting those organisms exhibiting the desired behaviors--will be fundamental to the progress of synthetic biology. We will in effect learn how to 'breed' useful synthetic gene networks, just as we have learned how to breed useful plants and animals. This approach mimics natural evolution in exploring the vast and complex landscape of functions available to a set of molecules making up an engineered regulatory pathway. Importantly, it circumvents our near-complete ignorance of how a DNA sequence encodes a specific set of biological functions, a detailed understanding that is required for any 'rational' design approach. By introducing random mutations into the DNA and screening for different functions that might be expressed by the mutant circuits, we can identify which functions are possible as well as the ranges of function available to the specific search process (e.g. random point mutation targeted to a specific gene). With further analysis, e.g. sequencing to identify the mutations and biochemical analysis of circuit components, we gain insights into the molecular mechanisms by which the overall function is achieved or modified. In this project we have three specific aims. The first is to validate a 'selection module' by which we can efficiently evolve components and circuits in the laboratory. This module connects proper circuit function to the ability of cells that express it to survive and grow. Cells with functioning circuits survive and grow; those that have not solved the problem do not. To program complex behaviors, we will also need components that respond to predefined ranges of input parameters with predictable output parameters. Thus our second aim is to use directed evolution to create a range of transcriptional activators based on the well-characterized framework protein, LuxR. Laboratory-evolved LuxR variants will activate gene transcription at different, nonnatural promoter sites on the DNA. Finally, we propose to investigate the range of circuit functions available to a predefined set of components via evolutionary exploration. Specifically, we will evolve a series of 'band detect' circuits that respond to a prespecified range of acyl-HSL concentrations. These circuits will be used to construct synthetic systems that form patterns of gene expression in the solid phase. Our ultimate goal is to develop a fundamental enabling technology for synthetic biology as well as for developing bio-inspired modes and architectures for computing. We envision that evolved circuits and the synthetic multicellular systems that can be constructed from them will be useful to researchers developing quantitative models of gene regulation, quorum sensing, and other aspects of cellular computing.
