Improving Protein Therapeutics Through Quantitative Molecular Engineering Approaches and A Cell-Based Oral Delivery Platform

Proteins, with their ability to perform a variety of highly specific biological functions, have emerged as an important class of therapeutics. However, to fully harness their therapeutic potential, proteins often need to be optimized by molecular engineering; therapeutic efficacy can be improved by modulating protein properties such as binding affinity/specificity, half-life, bioavailability, and immunogenicity. In this work, we first present an introductory example in which a mechanistic mathematical model was used to improve target selection for directed evolution of an aglycosylated Fc domain of an antibody to enhance phagocytosis of tumor cells. Several aspects of directed evolution experimental methods were then optimized. A model-guided ligation strategy was developed to maximize ligation yield in DNA library construction, and this design tool is freely available through a web server. Streamlined protocols for mRNA display and ribosome display, which are powerful in vitro selection methods, were also created to allow robust selection of a variety of therapeutic proteins, including monomeric Fc domains, designed ankyrin repeat proteins, a single-chain insulin analog (SCI-57), and leptin. Anti-ICAM-1 scFv antibody fragments were also optimized for ribosome display by grafting complementarity determining regions onto a stable human framework. In addition to engineering the proteins themselves, effective delivery systems are essential for maximizing the therapeutic benefit of these proteins in a clinical setting. We therefore also developed a novel oral delivery platform based on the food-grade bacterium Lactococcus lactis. SCI-57, leptin, and SCI-57-leptin fusion proteins have been successfully secreted from this host in vitro and preliminary studies in a diabetic mouse model show reduced glucose levels after oral administration of L. lactis secreting SCI-57. We then further improved the secretion potential of this host through directed evolution of a L. lactis signal peptide. In summary, our studies have provided important advances to the field of protein engineering through the development of mechanistic mathematical models, streamlined experimental methodologies, and polypeptides with improved properties. This work has also opened up the possibility of systemic delivery of protein therapeutics using live microorganisms