Comparison between Protein−Polyethylene Glycol (PEG) Interactions and the Effect of PEG on Protein−Protein Interactions Using the Liquid−Liquid Phase Transition

TLDR

Phase transitions of protein aqueous solutions are crucial for protein crystallization and biomaterials science, but additives such as polyethylene glycol (PEG) introduce significant thermodynamic complexity. The study investigates how small amounts of PEG1450 influence the liquid‑liquid phase separation of bovine serum albumin (BSA) and demonstrates that protein‑PEG‑buffer mixtures can exhibit two distinct LLPS types, providing a tool to validate and refine interaction models. Experimental measurements of LLPS temperature shifts with PEG concentration and BSA/PEG partitioning were combined with thermodynamic perturbation theory to link these observables to protein‑protein and protein‑PEG interactions, and the resulting model was evaluated against both properties. A depletion‑interaction model successfully accounts for both the PEG‑induced changes in protein‑protein interactions and the protein‑PEG interactions observed in the system.

Abstract

Phase transitions of protein aqueous solutions are important for protein crystallization and biomaterials science in general. One source of thermodynamic complexity in protein solutions and their phase transitions is the required presence of additives such as polyethylene glycol (PEG). To investigate the effects of PEG on the thermodynamic behavior of protein solutions, we report measurements on the liquid-liquid phase separation (LLPS) of aqueous bovine serum albumin (BSA) in the presence of relatively small amounts of PEG with an average molecular weight of 1450 g/mol (PEG1450) as a model system. We experimentally characterize two thermodynamically independent properties of the phase boundary: (1) the effect of PEG1450 concentration on the LLPS temperature, (2) BSA/PEG1450 partitioning in the two liquid coexisting phases. We then use a thermodynamic perturbation theory to relate the first property to the effect of PEG concentration on protein-protein interactions and the second property to protein-PEG interactions. As criteria to determine the accuracy of a microscopic model, we examine the model's ability to describe both experimental thermodynamic properties. We believe that the parallel examination of these two properties is a valuable tool for verifying the validity of existing models and for developing more accurate ones. For our system, we have found that a depletion-interaction model satisfactorily explains both protein-PEG interactions and the effect of PEG concentration on protein-protein interactions. Finally, due to the general importance of LLPS, we will experimentally show that protein-PEG-buffer mixtures can exhibit two distinct types of liquid-liquid phase transitions.