Evolution of an artificial allylic alkylase based on the biotin-streptavidin technology

The PhD thesis presented here summarizes the work and the scientific effort done in the research group of Prof. Dr. Ward at the University of Basel during the years 2013 – 2017. The Ward group has a long-term knowledge in the design and evolution of artificial metalloenzymes capable of catalyzing reactions including transfer hydrogenation, ring-closing metathesis, C-H activation, Suzuki-coupling and many more. Artificial metalloenzymes are formed by the incorporation of a catalytically active transition-metal complex into a host protein. This allows combining the advantageous features of both homogeneous catalysis and enzyme catalysis. The protein forms a defined reaction environment (i.e. a second coordination sphere) around the metal cofactor. Thus, artificial metalloenzymes can be evolved by chemical modification of the metal cofactor or by genetic engineering of the host protein. In the Ward group often the biotin-streptavidin technology is applied to generate artificial metalloenzymes. This system relies on the ultra-high affinity of the protein streptavidin for the small molecule biotin. Attachment of a biotin-anchor to a transition-metal complex ensures its incorporation into the streptavidin scaffold.
In this thesis the design, expression and evolution of an artificial allylic deallocase based on the biotin-streptavidin technology is described. A biotinylated ruthenium complex was synthesized, incorporated into streptavidin and a crystal structure of the resulting artificial metalloenzyme was determined. The activity of the hybrid catalyst in a deallocation reaction was investigated. An O-allyl carbamate caged pro-fluorescent coumarin derivative was deprotected in the presence of the artificial metalloenzyme. The in vitro performance of the artificial allylic deallocase was evolved by genetic modification of the host protein. In a next step, the artificial metalloenzyme was displayed on the surface of E. coli cells. The activity of the hybrid catalyst was further evolved by in vivo screening of several single-site saturation mutagenesis libraries. It was aimed to further increase the throughput of the screening assay by application of a microfluidic system in combination with fluorescence-activated droplet sorting. In a third step, a biogenetic switch based on O-allyl carbamate caged inducer molecules was designed. By the action of the artificial allylic deallocase, the caged inducer was deprotected and subsequently induced the expression of a green fluorescent protein (GFP)-reporter. By substitution of the GFP with another natural protein, a cascade reaction can be envisioned. In parallel, a series of streptavidin mutants with lid-like amino acid structures on top of the biotin-binding vestibule was designed. This approach aimed gaining a better control of the second coordination sphere of the metal cofactor in order to increase the activity and selectivity of the artificial metalloenzyme.
In summary, these efforts should allow a straightforward design, expression and evolution of new artificial metalloenzymes for in vivo applications. During the time in the Ward group a deeper knowledge on protein design and expression, molecular biology, synthesis of organometallic cofactors, in vivo catalysis and high-throughput screening based on microfluidics was garnered.