Genetic selection provides the most powerful method to assay large libraries of biomolecules for function, and harnessing the power of genetic selection for the detection of specific, non-endogenous small molecule targets would greatly facilitate the cloning of biosynthesis genes, the directed evolution of enzymes, and the engineering of metabolic pathways. While significant progress has been made in developing in vivo genetic selections for small molecule targets, there remain limitations to both the sizes of libraries and the types of molecules and chemical transformations that may be selected for. This proposal details a comprehensive effort to develop a method capable of performing genetic screens and selections for non-endogenous small molecule targets in bacteria. The proposed method addresses many of the limitations of current methods and rests on two well-established principles: The first is that it is possible to use in vitro selection to generate RNA sequences (aptamers) that tightly and specifically recognize small molecules, while discriminating against related structures. The second is that small molecule metabolites can interact with messenger RNAs to modulate gene expression in vivo through a mechanism known as riboswitch control. Results from our lab and from others show that it is possible to combine these ideas to generate synthetic riboswitches that recognize non-endogenous small molecules and control gene expression in E. coli. Herein we propose a method for producing synthetic riboswitches capable of mediating genetic selections for small molecules. This approach will be validated by using synthetic riboswitch-mediated'genetic selections to clone and direct the evolution of enzymes within two important alkaloid biosynthesis pathways. Concurrent mechanistic studies will provide evidence of how the new synthetic riboswitches function and will likely illuminate principles that govern the function of natural riboswitches in vivo.