Expanding the scope of DNA-templated organic synthesis and in vitro selection

Natural evolution is a powerful means of creating nucleic acids, proteins, and small molecules with new functional properties. To adapt evolution for the discovery of functional molecules in the laboratory, scientists have developed methods for the selection, amplification, and diversification of nucleic acids and proteins. The recent development of DNA-templated organic synthesis (DTS) has demonstrated that these principles can also be applied to the discovery of functional small molecules. To fully realize the potential of DTS to translate libraries of nucleic acid sequences into DNA-linked small molecules for in vitro selection and discovery requires the development of improved synthesis and encoding strategies. In this thesis we describe several efforts to expand the scope of DNA-templated organic synthesis and in vitro selection.; First, we describe a strategy for performing ordered multistep synthesis in a single solution. This system makes use of temperature-sensitive variations in DNA secondary structure to orchestrate an ordered multistep synthesis of a triolefin and a tripeptide in a single solution. This methodology provides a faster, higher-yielding route to multistep DNA-templated products and represents a mode of controlling chemical reactivity that is not accessible in traditional flask-based reaction formats.; Second, we describe a systematic study of the effect of secondary structure on DNA-templated reactivity. Our studies reveal that very structured and, unexpectedly, very unstructured DNA templates both exhibit compromised reactivity, suggesting that a modest degree of internal secondary structure is beneficial to maximize effective molarities between reactants after hybridization. We have incorporated these principles into the design of a genetic code suitable for encoding the synthesis of large libraries of DNA-linked small molecules.; Finally, we have compared the evolutionary potential of a nucleic acid library designed to contain an increased amount of predicted secondary structure with a fully random library for the discovery of functional nucleic acid aptamers. The results of a competitive selection experiment for binding to streptavidin reveal that the pre-structured library possesses more active species than the fully random library, suggesting that increasing the amount of structure in a nucleic acid library may be a general strategy for improving the likelihood of discovering functional molecules.