Continuous Evolution of Proteases With Altered Specificity

Abstract

The following thesis work aims to establish a method for the generation of proteases with tailor-made substrate specificities. This programmability will enable the design of proteases that modulate the activity of a target protein of biotechnological or therapeutic relevance. In the first phase of this project, we developed and validated a system for the phage-assisted continuous evolution (PACE) of protease enzymes. In pilot feasibility studies, we used this system to evolve hepatitis C virus proteases that are resistant to small-molecule protease inhibitors. We then established that this PACE selection could also evolve specificity changes in two different model proteases: human rhinovirus (HRV) protease and tobacco etch virus (TEV) protease. After these proof-of-concept experiments, we began the phage-assisted continuous evolution (PACE) of TEV protease, which canonically cleaves ENLYFQS, to cleave a very different target peptide sequence, HPLVGHM, that is present in IL-23, a cytokine implicated in inflammatory disease. A protease emerging from approximately 2,500 generations of PACE contains 20 non-silent mutations, cleaves human IL-23 at the target peptide bond, and inhibits IL-23 mediated signaling in cultured primary murine splenocytes. We characterized the substrate specificity of this evolved enzyme using a protease specificity profiling method, revealing a mixture of shifted and broadened specificity at the six positions in which the target amino acid sequence differed. On-going studies seek to expand the scope of protease PACE through: (1) the development of a negative selection scheme to ablate off-target proteolysis, (2) the adaptation of disulfide compatible E. coli strains for the PACE of human circulating proteases, and (3) the evolution Botulinum toxin proteases to enable the destruction of intracellular target proteins.