Behavior of host-cell-protein-rich aggregates in antibody capture and polishing chromatography.

Recent work has shown that aggregates in monoclonal antibody (mAb) solutions may be made up not just of mAb oligomers but can also harbor hundreds of host-cell proteins (HCPs), suggesting that aggregate persistence through downstream purification operations may be related to HCP clearance. We have examined this in a primary analysis of aggregate persistence through processing steps that are typically implemented for HCP reduction, demonstrating that the phenomenon is relevant to depth filtration, protein A chromatography and flow-through anion-exchange (AEX) polishing. Confocal laser scanning microscopy observations show that aggregates compete with the mAb to adsorb specifically in protein A chromatography and that this competitive interaction is integral to the efficacy of protein A washes. Column chromatography reveals that the protein A elution tail can have a relatively high concentration of aggregates, which corroborates analogous observations from recent HCP studies. Similar measurements in flow-through AEX chromatography show that relatively large aggregates that harbor HCPs and that persist into the protein A eluate can be retained to an extent that appears to depend primarily on the resin surface chemistry. The total aggregate mass fraction of both protein A eluate pools (∼ 2.4 - 3.6%) and AEX flow-through fractions (∼ 1.5 - 3.2%) correlates generally with HCP concentrations measured using enzyme-linked immunosorbent assay (ELISA) as well as the number of HCPs that may be identified in proteomic analysis. This suggests that quantification of the aggregate mass fraction may serve as a convenient albeit imperfect surrogate for informing early process development decisions regarding HCP clearance strategies.

Analytical characterization of host-cell-protein-rich aggregates in monoclonal antibody solutions.

Host-cell proteins (HCPs) and high molecular weight (HMW) species have historically been treated as independent classes of impurities in the downstream processing of monoclonal antibodies (mAbs), but recent indications suggest that they may be partially linked. We have explored this connection with a shotgun proteomic analysis of HMW impurities that were isolated from harvest cell culture fluid (HCCF) and protein A eluate using size-exclusion chromatography (SEC). As part of the proteomic analysis, a cross-digest study was performed in which samples were analyzed using both the standard and native digest techniques to enable a fair comparison between bioprocess pools. This comparison reveals that the HCP profiles of HCCF and protein A eluate overlap substantially more than previous work has suggested, because hundreds of HCPs are conserved in aggregates that may be up to ~50 nm in hydrodynamic radius and that persist through the protein A capture step. Quantitative SWATH proteomics suggests that the majority of the protein A eluate's HCP mass is found in such aggregates, and this is corroborated by ELISA measurements on SEC fractions. The SWATH data also show that intra-aggregate concentrations of individual HCPs are positively correlated between aggregates that were isolated from HCCF and protein A eluate, and species that have generally been considered difficult to remove tend to be more concentrated than their counterparts. These observations support prior hypotheses regarding aggregate-mediated HCP persistence through protein A chromatography and highlight the importance of this persistence mechanism.

Aggregates of α-chymotrypsinogen anneal to access more stable states.

Non-native protein aggregates present a variety of problems in fundamental and applied biochemistry and biotechnology, from quality and safety issues in pharmaceutical development to their association with a number of chronic diseases. The aggregated, often amyloid, protein state is often considered to be more thermodynamically and kinetically stable than (partially) unfolded or folded monomers except under highly denaturing conditions. However, evolution of the structure and stability of aggregated states has received much less attention. Here it is shown that under mildly-denaturing conditions (elevated temperature or [urea]), where the native monomer (N) is slightly favored compared to the unfolded state (U), α-chymotrypsinogen A (aCgn) non-native aggregates undergo a structural relaxation or annealing process to reach even more stable states. The annealed aggregates are more resistant to dissociation than aggregates that do not undergo this relaxation process. Aggregates without annealing dissociate via linear chain depolymerization, and annealing is accelerated under conditions that promote slow dissociation (partially denaturing conditions). This is consistent with a free energy landscape with multiple barriers and local minima that allows for a kinetic competition between aggregate dissociation and structural relaxation to more stable aggregate states. This highlights added complexities for protein refolding or aggregate dissociation processes, and may explain why it is often difficult to completely recover monomeric protein from aggregates.

Salting-in characteristics of globular proteins.

Protein solubility, and the formation of various solid phases, is of interest in both bioprocessing and the study of protein condensation diseases. Here we examine the the phase behavior of three proteins (chymosin B, ß-lactoglobulin B, and pumpkin seed globulin) previously known to display salting-in behavior, and measure their solubility as a function of pH, ionic strength, and salt type. Although the phase behavior of the three proteins is quantitatively different, general trends emerge. Stable crystal nucleation does not occur within the salting-in region for the proteins examined, despite the crystal being observed as the most stable solid phase. Instead, two types of amorphous phases were found within the salting-in region; additionally, an analog to the instantaneous clouding curve was observed within the salting-in region for chymosin B. Also, protein solutions containing sulfate salts resulted in different crystal morphologies depending on whether Li(2)SO(4) or (NH(4))(2)SO(4) was used.