(499e) Docking Simulations to Predict Binding Performance of Affinity Ligands for Purification of Butyrylcholinesterase

An ever-increasing demand for affordable biotherapeutics manufactured under stringent regulatory guidelines is driving the development of innovative, more efficient, lower-cost downstream purification technologies. The use of an affinity separation step in downstream purification can significantly reduce the number of process steps, buffer use, production time and overall costs. Affinity resins using antibody and other protein ligands can be prohibitively expensive. Synthetic peptide ligands offer a robust, lower-cost alternative to biological affinity ligands. This work illustrates the use of computational tools in identification, development and characterization of peptide based ligands for the purification of butyrylcholinesterase - an enzyme important for its ability to degrade cocaine, hydrolyze ester-containing drugs, and as an antidote to various nerve gas and organo-phosphorous compounds. This paper describes the use of High Ambiguity Driven Molecular Docking (HADDOCK) software to identify peptides that will bind to the protein through its active site. Docking calculations were performed for in-silico peptide libraries of three amino acids. Using characteristics of reported butyrylcholinesterase inhibitor molecule(s) as a guideline, peptides containing aromatic residues were prioritized for creating the peptide library. Residue cysteine forms cystine through a disulfide linkage, while methionine forms methionine sulfoxide (MetO) by addition of oxygen to its sulfur atom and were not considered while designing the library. Out of a total possible 8000 peptides, a biased library of 2256 peptides was designed with peptides containing one or more aromatic residues. Based on the available description of the active site of human butyrylcholinesterase, a targeted docking of these chosen 2256 peptides to the active site of butyrylcholinesterase was performed. A three-stage approach was adopted for this docking simulation – rigid, semi-flexible and finally water refined fully flexible docking. The docked structures were grouped in clusters based on root-mean-square-distance (RMSD) and scored using Firedock software to calculate free energies of binding. Procainamide, a known ligand to human butyrylcholinesterase, was also docked using the same protocol and its binding energy was used as a reference. Binding characteristics of a few chosen peptide ligands were investigated by synthesizing them on chromatographic resins via Fmoc based solid phase peptide synthesis. Experimental binding results for synthesized peptides were compared to predictions made by HADDOCK.