Mixed-mode membrane chromatography: mechanistic-modeling-based optimization and scale-up of antibody charge variant separation using a novel membrane adsorber

Abstract

BACKGROUND: Efficient manufacturing of complex biopharmaceuticals increasingly relies on innovative downstream processing strategies. Membrane chromatography offers a promising alternative to resin-based systems by enabling high-throughput separations at elevated flow rates. Typically used in flow-through mode, its industrial applications range from contaminant removal (DNA, host cell proteins) to purification of large biomolecules like viruses, virus-like particles and protein complexes. The Natrix® CH membrane, a novel mixed-mode membrane integrating ion-exchange and hydrophobic interaction functionalities, has shown potential in complex or elevated salt conditions (ca 0.1 mol L−1). This study evaluates its performance in bind-and-elute mode for separating monoclonal antibody charge variants and establishes a scalable mechanistic model. RESULTS: Initial comparisons with a traditional ion-exchange membrane showed that the Natrix® CH had enhanced selectivity, as demonstrated by better separation of charge variants and an overall increase of retention up to one pH unit. Both might be attributed to its additional hydrophobic ligand. Linear gradient elution experiments enabled calibration of a stoichiometric displacement model under low-loading conditions. High-loading experiments up to 35 mg mLMV−1 refined a steric mass action model using a reverse fitting procedure to capture the nonlinear range of the adsorption isotherm. These models were applied in silico to design optimized separation conditions, including an antiparallel dual pH and salt gradient. This dual-gradient strategy achieved high-resolution variant separation and was validated experimentally. A scale-up from laboratory scale (1.06 mL) to a bench-scale (8.8 mL) membrane device (eightfold increase) demonstrated consistent elution behavior and separation performance, even at elevated flow rates (up to 10 membrane volumes per minute), with only minor reductions in dynamic binding capacity (e.g. DBC10% decreased by 7%). CONCLUSION: Mechanistic in silico modeling enabled accurate prediction, optimization and successful scale-up of charge variant separation on Natrix® CH membranes. The approach demonstrated good applicability and scalability within the calibrated range, advancing membrane chromatography as a high-throughput alternative for efficient protein purification. © 2025 The Author(s). Journal of Chemical Technology and Biotechnology published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry (SCI).