Designing Proteins and Short Peptides for Biochemical Applications

Date of Award

December 2018

Degree Type


Degree Name

Doctor of Philosophy (PhD)




Ivan V. Korendovych

Subject Categories

Physical Sciences and Mathematics


Since their discovery, enzymes have been the focus of attention in several research areas. Many of their properties make them appealing for catalyzing various chemical reactions. Natural enzymes catalyze complex chemical transformations under mild conditions, with high efficiency and selectivity. Although present day enzymes are readily available, their production, narrow substrate range and lack of stability can be a disadvantage. Because of these limitations, there is great interest in tailoring the properties of existing enzymes and engineering new ones. A range of strategies, combining computational methods and high-throughput combinatorial techniques were successfully developed before and applied to engineer proteins and peptides for various biochemical applications, like catalysis, therapeutics and drug delivery. In this work, we used a minimalistic approach and high-throughput screening to design proteins and peptides with novel functions.

Previously we demonstrated that short, self-assembling peptides are capable of enzyme-like catalysis and efficiently promote hydrolysis and oxidation of model substrates in the presence of metal ions. In this study, we show that amyloids can catalyze the transformation of more complex substrates and they can be used in flow devices to create a heterogeneous catalyst. Moreover, non-covalent mixture of peptides demonstrates synergistic interactions, enabling us to screen multiple coordination spheres and further improve catalytic efficiency. The ease with which we have discovered robust catalysts opens a new path for the development of biocatalysts and suggests, that short peptides might have served as intermediates in the evolution of enzymes.

Additionally, we show that a de novo designed protein can stabilize a highly reactive semiquinone radical in the presence of different metal ions and the yield of radical formation depends on the amount of metal ion bound to the protein. Nuclear magnetic resonance and computational modeling studies suggest that the radical is stabilized by hydrophobic interactions, resulting in a lower redox potential of the radical/catechol couple as compared to bulk aqueous solution. These results open new possibilities for development of mechanistic models for metalloenzymes.

In this work, we also propose a new method for early, in vivo detection of hepatocellular carcinoma (HCC) by using Positron Emission Tomography and a 18F-based radiotracer, fluorodeoxyglucose (FDG). HCC is one of the deadliest cancers and when discovered late, there is no cure. Fusing an FDG-protein complex to an antibody, specific for HCC cells, would enable quick, non-invasive detection of the diseases in its early stages.

Finally, we established the pKa of the active residue in a de novo designed Kemp eliminase using a combination of multidimensional NMR experiments at different pH values. Establishing the pKa value helped us to better understand the mechanism of the catalysis.


SURFACE provides description only. Full text may be available to ProQuest subscribers. Please ask your Librarian for assistance.

This document is currently not available here.