Bacteria-based methods for engineering and characterization of proteases and affinity proteins

Sammanfattning: This thesis is focused on the development of methods for characterization and engineering of both proteases and affinity proteins. In addition, a prodrug concept for small affinity proteins is developed.Two of the developed methods are for engineering and/or characterization of proteases. First, a method for substrate profiling and engineering of proteases was investigated (paper I). In this method, a protease and a reporter are co-expressed in E. coli. The reporter is comprised of an enzyme, which confers resistance to an antibiotic, fused to a substrate, and a degradation tag. In absence of site-specific proteolysis within the reporter, the degradation tag renders the entire reporter a substrate for the intracellular degradation machinery. Thus, by applying competitive growth in presence of the antibiotic, a substrate that is preferred by a model protease could be enriched relative to less efficiently hydrolyzed substrates. Then, an alternative method for substrate profiling was developed (paper III). Here, the substrate is instead displayed on the surface of bacteria, and located between two anti-idiotypic domains, where one blocks the other from interacting with a reporter. Site-specific proteolysis releases the blocking domain and is therefore reflected in reporter binding. After incubation with fluorescently labeled reporter, the proteolysis can be analyzed by flow cytometry. When large libraries of potential substrates for matrix metalloprotease 1 (MMP-1) were screened, a panel of substrates with the previously reported motif PXXXHy was enriched, thereby demonstrating the potential of the method. This method offers the possibility for high-throughput substrate profiling of proteases as well as engineering of substrates for use in for example protease-activated prodrugs.In another study, a new prodrug concept for small affinity proteins was developed to improve the tissue selectivity in future in vivo studies (paper II). This concept takes advantage of the local upregulation of proteases in the diseased tissue in various disorders. By fusing a targeting domain to an anti-idiotypic binding partner via a protease-sensitive linker, the targeting domain is masked from interacting with its target until activation by site-specific proteolysis within the linker. The concept was demonstrated for a small affinity protein (Affibody molecule). Bacterial display was employed to engineer the so-called pro-Affibody. When displayed on the bacterial surface, the pro-Affibody showed over 1.000-fold increase in apparent binding affinity upon activation by a disease-associated protease. Additionally, the activated pro-Affibody could bind to its target expressed on cancer cells, as opposed to the non-activated pro-Affibody. This concept is likely to be extendable to other small affinity proteins and opens up for the possibility to develop new such prodrugs to previously non-druggable targets.In the last study, a screening method for protein-based aggregation inhibitors was developed (paper IV). In this method, a reporter and an inhibitor are co-expressed in E. coli. The reporter is comprised of green fluorescent protein (GFP) fused to an aggregation prone peptide. Upon aggregation, the fluorescence is decreased, but it is then restored when the reporter is co-expressed with an inhibitor. In a model screening experiment, an Affibody molecule that targets the A? peptide (involved in Alzheimer’s disease) could be enriched from a background of non-inhibiting Affibody molecules. Also this method is likely to be extendable to other types of affinity proteins, and also to different aggregation prone peptides/proteins involved in other diseases.In conclusion, the methods and concepts presented in this thesis could in the future yield new means for the engineering and characterization of proteins with desired properties to be used in both biotechnological and medical applications.