BioSystems 90 (2007) 115–120 Dual selection of a genetic switch by a single selection marker Yoko Nomura, Yohei Yokobayashi ∗ Department of Biomedical Engineering, University of California, Davis, 451 E. Health Sciences Dr., Davis, CA 95616, USA Received 23 February 2006; received in revised form 10 July 2006; accepted 12 July 2006 Abstract Forward engineering of synthetic genetic circuits in living cells is expected to deliver various applications in biotechnology and medicine and to provide valuable insights into the design principles of natural gene networks. However, lack of biochemical data and complexity of biological environment complicate rational design of such circuits based on quantitative simulation. Previously, we have shown that directed evolution can complement our weakness in designing genetic circuits by screening or selecting functional circuits from a large pool of nonfunctional ones. Here we describe a dual selection strategy that allows selection of both ON and OFF states of genetic circuits using tetA as a single selection marker. We also describe a successful demonstration of a genetic switch selection from a 2000-fold excess background of nonfunctional switches in three rounds of iterative selection. The dual selection system is more robust than the previously reported selection system employing three genes, with no observed false positive mutants during the simulated selections. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Genetic circuit; Genetic switch; Directed evolution; Genetic selection 1. Introduction While significant progress has been made recently in computational modeling of biological systems at the molecular and genetic level, the daunting complexity of real biological systems and environment in which they operate often preclude accurate quantitative sim- ulation of even the smallest synthetic genetic circuits. Simulations provide crucial information and insights into genetic circuit design but significant experimen- tal efforts are still required to implement such circuits. We have demonstrated that well planned and executed screening or selection of functional genetic circuits from a large pool of nonfunctional variants can be an effi- cient strategy for implementing or optimizing genetic circuits (Yokobayashi and Arnold, 2005; Yokobayashi et ∗ Corresponding author. Tel.: +1 5307549676; fax: +1 5307545739. E-mail address: yoko@ucdavis.edu (Y. Yokobayashi). al., 2003, 2002). This approach is based on the directed evolution technologies that have been successfully used to improve various protein functions (Otten and Quax, 2005). Screening may be well suited for directed evolution of genetic circuits because many circuits provide simple ON/OFF output that can be coupled to autofluorescent protein (e.g., GFP) expression which can be measured in high-throughput. Yokobayashi et al. (2002) demon- strated the screening of genetic inverters in Escherichia coli. Guet et al. (2002) generated a small library of randomly connected gene networks and screened for various Boolean logic functions by fluorescence screening. However, screening often requires significant cost and labor as the library size becomes larger because each mutant in a library must be physically isolated and individually evaluated. Selection of genetic circuits by coupling the circuit output with the survival of the host cells would allow sorting of larger libraries which may be necessary 0303-2647/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.biosystems.2006.07.006