Low-volume solubility assessment during high-concentration protein formulation development.

Objective Solubility is often one of the limiting factors for high-concentration protein formulation (HCF) development. Determination of protein solubility is challenging and requires high amount of material. Therefore, low-volume and predictive approaches are desired. Methods This work presents a simple and material-saving approach using static light scattering to describe non-ideal solution behaviour of HCF. Non-ideality can be related to protein-protein interactions in solution. The type and strength of these interactions indicate maximum protein solubility at actual formulation compositions. Interactions of four therapeutic model proteins at multiple formulation compositions were investigated, and deduced solubility was compared to apparent solubility behaviour determined by ether turbidity or content measurements. Key findings Protein-protein interactions and deduced solubilities matched actual solubility data for all tested formulations. Protein solubility was found to be lowest at pH values near the isoelectric point of each model protein. Buffer salts and ionic strength were also found to strongly influence protein solubility. In addition, sucrose and a combination of arginine and glycine enhanced protein solubility, whereas surfactants such as polysorbate 20 did not influence protein solubility. Conclusions The introduced screening procedure is a powerful tool during (early) protein formulation development. It meets several requirements of HCF development and enables reliable prediction of protein solubility based on determination of protein interactions. In addition, rare data about the influence of several common excipients on apparent solubility of therapeutic proteins were shown.