UC Berkeley

Synthesis of S-adenosylmethionine (SAM) Analogs and Development of High-Throughput Screens to Target SAM-I Riboswitches

By

Scott Forsythe Hickey

Doctor of Philosophy in Chemistry

University of California, Berkeley

Professor Ming C. Hammond, Chair

The understanding of fundamental regulatory mechanisms in bacteria has been aided by the discovery of a class of non-coding RNAs known as riboswitches. Riboswitches are ligand-binding RNAs that control the expression of a wide variety of genes in response to changes in intracellular metabolite concentrations. Bacterial genomes contain riboswitches that can exert control over entire collections of metabolic genes that are important for cell survival, homeostasis, and adapting to environmental changes. Coupled with the presence of riboswitches within the genomes of bacterial human pathogens, these features suggest that riboswitches are promising new targets for therapeutic development. In addition, the unique gene control mechanisms of riboswitches make them well-suited as tools for synthetic biology. However, investigations into riboswitch-ligand interactions has been hindered by the lack of available high-throughput screening tools. The ability to screen riboswitch binding properties in a rapid and scalable fashion will augment drug discovery and target validation efforts for bacterial riboswitches and other RNA targets in general.

Here, we describe the development of a series of riboswitch-ligand binding assays with increased throughput and utility as compared to traditional assays. The techniques described are demonstrated for riboswitch model systems, but are also applicable to RNA-ligand interactions in general. Out of the 20 known riboswitch classes that have been described to date, seven of them bind the small molecule S-adenosylmethionine (SAM). The most abundant of these is the SAM-I riboswitch family, which we chose as a model system for our studies. First, a synthetic procedure is described for the preparation of SAM analogs that are of interest for both riboswitch- and protein methyltransferase- drug discovery. SAM analogs prepared using our procedure were profiled against a panel of human protein methyltransferases, and a fluorescently-tagged SAM analog was shown to target the SAM-I riboswitch family. The fluorescent analog was applied to the development of a high-throughput screening approach to identify riboswitch agonists or antagonists by a competitive ligand displacement assay. A similar strategy could be used for identifying protein methyltransferase inhibitors.

In addition, the development of a microfluidic mobility shift screening platform is described for the rapid characterization of riboswitch-ligand interactions with extremely low resource consumption. The method is highly scalable, offers significant advantages over traditional methods, and will enable screening efforts for a variety of riboswitch-ligand interactions. Finally, a mechanism-based magnetic bead assay is described with applications both for riboswitch in vitro selection and investigation of kinetically-controlled riboswitch regulatory responses. It is anticipated that the assays described here will forward the development of new riboswitch-ligand pairs for antibiotic discovery as well as increase fundamental understanding of riboswitch binding and selectivity principles.