Tuning and Engineering Protein Specificity towards Inhibition of Cancer Related Proteases

Tuning specificity for a multispecific proteins is a significant challenge, nevertheless, very useful for controlling biological processes regulated by such proteins. The challenges inherent in such a feat arise for the most part due to the high structural and sequence similarity among targets. Especially, proteolytic enzymes, which usually belongs to large families of closely similar enzymes and may also hydrolyze their designed target. To this end, we employed a combinatorial design method and yeast surface display to assess the effects of multiple mutations on binding specificity. As a proof of principle for this approach, we chose to design a specific inhibitor for one type of cancer related protease, namely human mesotrypsin.

Herein, we used a yeast surface display platform and novel competitive screening strategy to switch the selectivity of the human amyloid β-protein precursor Kunitz protease inhibitor domain (APPI) to mesotrypsin inhibition vs anionic trypsin, cationic trypsin and kallikrein-6 competitors. Using our strategy we were able to identify several selective mesotrypsin binders from APPI library. Measuring the on and off-target affinities of the purified APPI variants revealed a remarkable improvement, in one particular variant namely APPI_{P13W/M17G/I18F/F34V}, which exhibited specificity shifts ranging from 6500-fold up to 230000-fold versus the competitors tested relative to the wild-type APPI protein. Additionally, our experiments identified APPI positions tolerant to mutations as well as mutations that act as specificity switches for these targets. Using this strategy will thus help to better understand the binding landscape of multispecific proteins and to pave the way for design of new specific drugs and diagnostic tools targeting proteases in particular and other proteins in general.