Come see the latest PhD defense from the Buller Lab, Jonathan Ellis’ Defense! A quick overview of topics discussed:
Through the course of evolution, Nature has developed a swath of useful enzymes for catalyzing chemical reactions crucial to life. These enzymes are used in complex metabolic cascades to produce a wide variety of useful natural products in a chemo-, regio-, and stereoselective fashion. Through protein engineering, it is possible to further expand the applicability of these enzymes, and by using computational modelling tools, it is possible to accelerate the engineering process. Here, I discuss my efforts in engineering two useful pyridoxal 5’-phosphate-dependent enzymes, UstD and RgnTDC, to expand their substrate scopes and applicability. UstD is capable of directly installing an amino acid moiety into electrophilic aldehyde substrates through a decarboxylative aldol-like reaction from L-aspartic acid. Through multiple rounds of directed evolution, including linear regression guided evolution, we were able to improve the activity and expressibility of UstD. Complementary to UstD, RgnTDC is a L-tryptophan decarboxylase which is able to decarboxylate amino acids and generate precursors to medicinally relevant compounds. By using Substrate Multiplexed Screening (SUMS) in conjunction with logistic regression guided directed evolution and primer-doped mutagenesis, we have created a variety of generalist biocatalysts with diverse active site composition. To help facilitate these future engineering efforts and explore the utility of SUMS, I have also developed computational tools to process and analyze library screening data. These tools have been integrated into a single web-based program, titled the SUMS Interactive Processing Software (SIPS), and have been developed with broad applicability and community use in mind.