Computational design and in vivo implementation synthetic gene Boolean gates

View all posters

Mario Andrea Marchisio, Fabian Rudof, Joerg Stelling

ETH Zurich, Switzerland

We developed a computational tool for the automatic design of digital synthetic gene circuits (Marchisio and Stelling, PLoS Comp.Biol. 7, e1001083, 2011). We chose digital circuits because they can be used as biosensors that, for instance, produce a fluorescent protein as a response to the presence of chemicals. In the latest version of our tool, the user has to specify the input (chemicals) number and type, fill in the truth table, and indicate if the chemicals bind transcription factors or riboswitches. The truth table is converted into Boolean formulas via the Karnaugh Map method and formulas are translated into circuits organized in two or three layers of gates. Boolean gates exploit transcriptional and translational control at promoter and RBS (ribosome binding site) level, respectively. Digital circuits computed by our tool require three kinds of basic Boolean gates: YES, NOT, and AND. All the schemes corresponding to a given input are ranked according to a complexity score that quantifies the effort required for a practical implementation. A solution chosen by the user is encoded into MDL (Model Description Language) files and is visualized with ProMoT. Circuits can be finally exported to a format suitable for simulations such as SBML or Matlab. As a prof of concept, we implemented in yeast several configurations of basic Boolean gates such as YES, NOT, and AND. They were realized by integrating bacterial repression systems into the yeast genome. Each gate was characterized by the signal separation i.e. the distance between 1 and 0 outputs at steady state. Small digital circuits that sense two or three inputs and arise from the combination of our library of basic gates are currently under construction.