Engineering dynamic protein binding pairs for protein purification and delivery
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
A substantial number of proteins that bind to target molecules with great
specificity and high affinity has been identified. Through genetic engineering,
synthetic biologists have adapted these domains for the self-assembly of artificial
protein complexes for applications such as medical diagnostics and therapeutics.
Typically, many binding proteins are designed to spontaneously assemble with
corresponding ligands, forming static complexes; however, these same protein binding
pairs can be adapted to have switch-like activity through engineering. Throughout this
work, several strategies were employed to create signal-responsive alteration of
binding affinity between a protein and its corresponding ligand for applications in both
protein purification and drug delivery. ☐ SH3 domains and corresponding binding ligands were combined to create a
new affinity capture pair to use in elastin-like polypeptide (ELP) mediated affinity
capture (EMAC). SH3 domains were chosen as the capture domain, because the short
binding ligands were likely to have minimal impact on folding or expression as fusion
partners to proteins of interest (POI). Two binding ligands that have a reported 100-
fold difference in KD with SH3 domains were used to bind and elute POI from ELPSH3.
The effect on purification of the number of tandem repeats of both SH3 and
binding ligands with respect to purification of monomeric and multimeric POIs was
studied. Furthermore, dissociation of ligands from SH3 domains using a low pH buffer
highlighted potential for a recycling scheme. ☐ Binding affinity of a protein binding pair was also modulated using disulfide-bonds
for use in drug delivery. The PDZ domain and its binding ligand were mutated
to form intermolecular disulfide bonds that would lock-up the heterodimerization pair
in an oxidizing environment, but release in a reducing environment. Redox responsive
binding pairs were a key part of our drug delivery strategy to take advantage of highly
reducing nature of cytosolic environments that can be used to release protein drugs
from delivery vehicle complexes. The use of a relatively weak binding pair (KD = 10
μM) was important to the strategy as the desired heterodimer would be driven to form
because of the affinity between the binding pair and lock up in oxidizing conditions,
but would still dissociate in reducing environments. Using ELP-PDZ(N24C) and GFPdsPDZlig
as test proteins, the responsiveness of affinity-driven disulfide bond
formation and dissolution to redox balances was confirmed. ☐ Lastly, Protein M (protM), a novel antibody binding protein, was adapted to
modulate antibody function. protM was touted for its ability to bind to many different
types of antibodies, and to block antibody/antigen interactions through a steric clash
with its C-terminus. We theorized that if the antibody binding and antigen blocking
regions of protM could be clearly identified, protM could be used to universally
modulate antibody function through an external signal. Through incremental
truncations to the C-terminus, protM was reduced down to a stable fragment, pM420,
which was able to bind to antibodies but with greatly reduced antigen-blocking ability. In addition, antibody functionality was modulated by controlling the presence of
protM C-terminus through ligation and subsequent thrombin proteolysis. Splitting
protM in this manner unlocked the potential for creating a universal and customizable
switch for antibodies.