Liquid-Phase Peptide Synthesis Optimisation: Selection of Quenching Reagents and Peptide Solubility Estimation

Abstract

With the increasing demand on peptides market, the peptide synthesis process must be kept upgraded. Therefore, the focus of this project is the optimisation of liquid-phase peptide synthesis. During peptide synthesis, the by-products and the amino acid excess must be kept controlled to not interfere with the main reactions and, consequentially, with the final product yield and purity. In order to achieve this, a quenching reagent is added to inactivate the amino acid in excess. An optimisation study was performed for the quenching reaction, where it was found that 2 or 3 amino acids equivalents reacting with 2.1 quenching reagent equivalents allowed the best conditions for a controlled reaction. From there, with the conditions established, three reagents – piperidine, aniline, thiomalic acid – were used in the quenching reaction, where the quenching efficiency was evaluated by the quenching rate constant (kq). The kq values for piperidine, thiomalic acid and aniline we deduced to 2.962, 1.849, and 0.020 min-1 respectively. Piperidine was then determined as the best quenching reagent, followed by thiomalic acid and aniline, respectively. The second problem of peptide synthesis that this project approached, was the determination of the peptide solubility. Currently, the solubility of peptides is measured experimentally, which is a time-consuming and wasteful process once it needs to be measured in different solvents, thus increasing the process cost. The second part of the project investigated two theoretical approaches – Solubility Parameters and COSMO-RS – to predict the best and worst solvents for free peptides and peptides attached to a hub (wang and rink amide nanostar). DMSO was found the best solvent for most of the peptides, followed by NFM, which also presented itself as the best green solvent. The validation of the COSMO-RS method was tested and exhibited a RMSE of between 1.10 – 0.74 logarithmic units. Therefore, this model is presented as a tool to be implemented in determining the solubility of the amino acids.