Uncovering Biophysical Determinants Of Oxidoreductase Function Through De Novo Protein Design
Oxidoreductases are a diverse class of enzymes defined by their chemical role of transferring electrons from one substrate to another. Comprising the functional cores of photosynthetic, chemosynthetic, and cellular respiratory machinery, “redox proteins” and the catalysis they enable form the foundation biological energy transduction. Mastering the mechanisms by which these enzymes work is the key to understanding how life works and drives the improvement of an as of yet poorly developed theoretical framework for describing the multiscale chemistry and physics crucial to future developments in synthetic biology, mesoscale engineering, and macromolecular catalysis. Redox proteins are characterized by their near universal incorporation of one or more cofactors, non-protein molecules with notable electrochemical and/or photophysical properties that play a fundamental role in the enzyme’s behavior. The functions of these enzymes can then be thought of as having two aspects: the cofactor(s), whose physical chemistry dictates the general mechanism of energy transduction, and the protein scaffold, which via microenvironment, relative positioning, geometric strain, and other mechanical properties modulate the energetic landscape and gate the influx and efflux of reactants and products. Separation of these aspects in practice, however, is often hindered by the sheer structural complexity of the natural machinery in question, even the simpler ones. Billions of years of evolution can incorporate many inessential elements, creating what is known as irreducible Mullerian complexity. Observational studies are useful but have their fundamental limitations. True validation of understanding comes through construction and functional verification. In accordance with this philosophy, we set out to recreate two representative redox chemistries in minimally designed model protein scaffolds to establish the basic requirements for their respective catalytic processes. Specifically, we focus on the suppression of heme-based superoxide formation and the promotion of photo-induced flavin activation in two classes of four-helix bundle proteins with complementary biophysical properties. This thesis discusses the implementation of both active sites in each of the two classes of scaffolds, and draws conclusions about the essential components of each. With the long-term goals of engineerability and biological integration in mind, we also demonstrate the ability of these model scaffolds to interface with two types of natural membrane transport complexes. In addition, we show that the 4-helical bundle domain retains functional independence even as a polymer of up to 16 alpha helices. The thesis ends with a discussion on the present limits and future potential of minimal models and protein engineering as a whole.