Salt Modulation and Optimized CHO Media Highlighted in Antibody Bioprocessing Studies
With over 170 antibody products approved globally, bioprocess studies aim to improve manufacturing. Optimized OPM-CHO media increased fed-batch productivity by >40%, while salt addition was evaluated to improve antibody pool filterability.
Antibody-based products are the largest class of recombinant-protein therapeutics, and antibody therapeutics are applied across oncology, immunology, hematology, and infectious disease. By the end of 2023, more than 170 antibody products had received commercial approval globally. Two studies examine strategies to improve antibody manufacturing: one focuses on optimizing cell-culture media for Chinese hamster ovary (CHO) cells to enhance fed-batch productivity, and the other investigates the use of salts to improve the filterability of antibody pools.
For upstream processing, most recombinant-protein therapeutics are produced in mammalian cell culture, with CHO cells being the dominant production platform. Monospecific immunoglobulin G (IgG) molecules account for about 74% of all approved antibody therapeutics. Typical mAb expression titers in the late 1990s were often 1–2 g/L or less, whereas titers of 5–8 g/L are routine today. Optimization of cell-culture media and feed remains one of the most significant and scalable approaches to improving upstream process efficiency.
In a demonstration of OPM-CHO media performance, CHOZN GS–/– host cells engineered for stable expression of a known IgG4 biosimilar mAb were cultured in fed-batch shake flasks. Cells were inoculated at 1.0 × 10^6 cells/mL in basal medium, with glucose maintained at concentrations >1 g/L. Combinations of OPM media and feeds supported higher growth profiles and peak viable-cell densities (VCDs) than two global competitor products did, while maintaining similar or better cell viabilities. With process extension to day 16, cell viabilities of OPM 1-2, 2-1, and 2-2 remained near or >90%. Final titer measurements on day 16 demonstrated superior performance of OPM products, with >40% increase in volumetric productivity. Analytical characterization of purified product from group OPM 1-1 revealed limited product fragmentation and aggregation and similar charge-variant distribution to that of the reference molecule.
For downstream processing, filtration frequently is used during antibody harvest processes and intermediate polishing steps, and early fouling of filters can lead to increased overall processing time and reduced yield. Fouling mechanisms include pore constriction, pore blockage, and cake formation, driven by protein aggregates, protein–protein interactions, and/or protein–impurity interactions. One strategy to mitigate membrane fouling involves diluting filter load material, but that approach might be limited by facility-fit constraints. Previous work has demonstrated that increasing the ionic strength of process fluids can reduce protein–protein interactions that arise from electrostatic interactions at both high and low protein concentrations. Salts provide charge shielding and interrupt protein–protein interactions. In ultrafiltration/diafiltration processes, increasing ionic strength in a diafiltration buffer reduces solution viscosity and therefore improves process pressure and the capability to reach high concentrations. Both arginine and sodium chloride have been shown to reduce the viscosity of protein solutions at pH 7.5 and pH 5.0.
To evaluate whether increasing ionic strength through addition of different salts could improve the filterability of antibody pools through an intermediate membrane filter, a study used two immunoglobulin G1 (IgG1) molecules: a conventional antibody (Ab1) with an isoelectric point (pI) of 7.4 and a multispecific antibody (Ab2) with a pI of 8.4. Each salt was added to a glycine-based protein A elution buffer, and pooled eluate was adjusted to pH 3.4 and held for two hours, mirroring a low-pH viral inactivation step. At the highest salt-additive concentration, the pre-titrated eluate pool exhibited a slightly lower pH (4.0) than did other eluate pools (4.3), yielding ≈1% v/v variation in the amount of low-pH titrant used.