| The selection of functional macromolecules from diversified pools of polymeric sequences is a central feature of natural and laboratory-based molecular evolution. However, the monomers that are available for evolution are largely limited to those that are tolerated by polymerise enzymes or by the ribosome. We have investigated the DNA-templated polymerization of peptide nucleic acid building blocks, with the purpose of applying this reaction towards the directed evolution of functional synthetic polymers. Exploiting the high fidelity, efficiency, and rapid kinetics of DNA-templated reductive amination of PNA peptide aldehydes, we created libraries of DNA-encoded synthetic PNA polymers with diversities exceeding 109 members. Selection of such a library for binding to the protein papain revealed a sequence that was shown to have measurable affinity and specificity toward the target. Additionally, the mutation and reselection of this initial sequence led to second-generation synthetic polymers with improved protein-binding affinity. These findings represent the first in vitro evolution of a synthetic polymer.; The second part of this thesis describes the application of directed evolution principles to the study of natural proteins. Combinatorial and selection-based approaches are increasingly enabling the rapid and efficient characterization of protein complexes, which play a central role in almost all biological functions. Beyond characterization, the ability to re-engineer the specificity of protein-protein interactions can help us better understand complex networks of interacting proteins in living organisms. We have optimized an in vivo protein-protein selection system in E. coli based on fragment complementation of the enzyme dihydrofolate reductase. As opposed to existing in vitro systems such as phage display and in vivo systems such as the yeast two-hybrid, this system allowed for the randomization and selection of a complex, marginally stable mammalian protein, the nuclear receptor PPARgamma. We have used this system to characterize the interface between the PPARgamma ligand binding domain and several of its coactivator binding partners, with the ultimate goal of reengineering specificity in these biologically important complexes. |
