(340d) Predicting Biomolecule Partitioning in Aqueous Two-Phase Systems through Thermodynamic Modeling

Aqueous Two-Phase Extraction (ATPE) using an Aqueous Two-Phase System (ATPS) has long been shown to be a promising alternative in the purification of high-value biomolecules (e.g., biopharmaceuticals) from fermentation broth. Although offering a high biocompatibility, a gentle separation profile due to low interfacial tension and high efficiencies, ATPE has not yet made it into an industrial application. Reasons are several challenges, typically arising from state-of-the-art ATPS selection: (1) The high experimental effort required to identify a suitable ATPS by trial-and-error screenings. (2) The lack of knowledge about the influence of the ATPS solutes (e.g. phase formers, displacement agent and excipient) on the stability and activity of the high-value biomolecule. (3) The lack of suitable models that can predict the biomolecule partitioning in the ATPS, thus allowing for rapid preselection of potential ATPS and operating conditions.

To address the first two challenges, we recently presented a physically sound approach that enables the selection of optimized ATPS while maintaining conformational and colloidal stability of the biomolecule, enabling high yields, and running the process within an optimal process window.

Within this work, we now extend our approach towards predicting biomolecule partitioning within ATPS by thermodynamic modeling using the electrolyte perturbed-chain statistical associating fluid theory (ePC-SAFT) equation of state to further decrease time- and cost-intense experiments. A new, heterosegmental modeling approach is introduced, allowing for consideration of the biomolecule in ePC-SAFT and prediction of its phase behavior. For the fitting of biomolecule pure-component and binary interaction parameters, static light scattering data of aqueous biomolecule solutions is used. The predictive capability of the new approach was evaluated by predicting the partitioning behavior of Immunoglobulin G (IgG) in two different ATPS (1. monosodium glutamate-polyethylene glycol 2000 ATPS and 2. trisodium citrate-polyethylene glycol 2000 ATPS) containing different concentrations of sodium chloride as displacement agent. The predictions for the partitioning of IgG in both ATPS show a very good agreement with the experimental data, thus emphasizing the high potential of the new approach. Calculations can be performed in a few hours instead of several days for the respective experimental investigations. Prospectively, this will drastically reduce the experimental effort and pave the way for an accelerated ATPE development for downstream processing of high-value biomolecules.