Altering the substrate specificity of proteases is a powerful process with possible applications in many areas of therapeutics as well as proteomics. Although the field is still developing, several proteases have been successfully engineered to recognize novel substrates. Previously in our laboratory, eight highly active OmpT variants were engineered with novel catalytic sites. The present study examined the roles of several residues surrounding the active site of OmpT while attempting to use rational design to modulate fine specificity enough to create a novel protease that prefers phosphotyrosine containing substrates relative to sulfotyrosine or unmodified tyrosine residues. In particular, a previously engineered sulfotyrosine-specific OmpT variant (Varadarajan et al., 2008) was the starting point for rationally designing fifteen new OmpT variants in an attempt to create a highly active protease that would selectively cleave phosphotyrosine substrates. Our design approach was to mimic the most selective phosphoryl-specific enzymes and binding proteins by increasing positive charge around the active site. Sulfonyl esters have a net overall charge of -1 near neutral pH, while phosphate monoesters have a net overall charge of -2. Selected active site residues were mutated by site-directed mutagenesis to lysine, arginine, and histidine. The catalytic activities and substrate specificities of each variant were characterized. Although several variants displayed altered substrate specificity, none preferred phosphotyrosine over sulfotyrosine containing peptides. Taken together, our results have underscored the subtle nature of protease substrate specificity and how elusive it can be to engineer fine specificity. Apparently, phosphotyrosine specific variants were not possible within the context of our starting sulfotyrosine specific OmpT derivative mutated to have single amino acid changes chosen on the basis of differential charge interactions. / text

Proteases are crucial to many biological processes and have become an important field of biomedical and biotechnological research. Engineering of proteases towards therapeutic applications has been limited due to the lack of high-throughput methods for characterization and selection. We have developed a novel high-throughput method for quantitative assessment of proteolytic activity in the cytoplasm of Escherichia coli bacterial cells. The method is based on coexpression of a protease of interest and a reporter complex consisting of an aggregation-prone protein fused to a fluorescent reporter. Cleavage of a substrate sequence situated between the two reporter complex proteins results in increased whole-cell fluorescence proportional to proteolytic activity, which can be monitored using flow cytometry. We have demonstrated that the method can distinguish efficiencies with which Tobacco Etch Virus (TEV) protease processes different substrates. We believe that this is the first method in the field of protease engineering that enables simultaneous measurement of proteolytic activity and protease expression levels and can therefore be applied for substrate profiling, as well as screening and selection of libraries of engineered proteases.

Proteases have a fundamental role in regulating diverse biological processes and are important in the biotechnological and medical fields. Therapeutic proteases have great potential but have been limited due to the lack of high throughput protein engineering methods. In this thesis, a method was developed for high throughput screening of protease libraries based on competitive growth in selective media. A proof-of-principle method using the Tobacco Etch Virus protease (TEVp) was developed. TEVp and the reporter consisting of an aggregation-prone peptide amyloidbeta42 (Aβ42) genetically fused to the antibiotic resistance enzyme chloramphenicol acetyltransferase (CAT), were co-expressed in Escherichia coli. The CAT enzyme makes the cells resistant to Chloramphenicol (Cml). Two different reporters containing different cleavage sites situated between the Aβ42 and CAT were used for which TEVp has distinguishable proteolysis efficiencies. Cleavage of the fusion protein would give the cell a growth advantage in media with Cml since the CAT enzyme would avoid misfolding due to Aβ42. The method demonstrates that cells with different substrates can be differentiated based on their survival. A 100-fold enrichment of clones expressing the efficient substrate was also demonstrated from a background of 1:1000 of clones expressing the inefficiently cleaved substrate. Moreover, a semi-rational TEVp library was successfully cloned and co-electroporated with the reporter into E. coli for future selection.