The advent of numerous monoclonal antibody (mAb) therapeutics has expanded the biopharmaceutical market appreciably over the last two decades, partially enabled by the establishment of general platform processes to support mAb downstream purification [1], [2], [3], [4], [5], [6]. These platform processes typically begin with clarification of harvested cell culture fluid (HCCF) using centrifugation and/or depth filtration, followed by protein A chromatography for product capture [7,8]. One or more polishing operations using ion-exchange, hydrophobic interaction or multimodal chromatography are also typically employed to clear impurities that could jeopardize the therapeutic's safety and stability [9,10]. Impurity classes in mAb solutions are often categorized as process- or product-related; process-related impurities include cell debris, lipids, host-cell proteins (HCPs), nucleic acids and adventitious viruses, whereas product-related impurities include mAb aggregates and fragments [11,12]. Although downstream operations typically remove multiple impurity classes simultaneously, the clearance of each one is often analyzed independently.

The distinction between process- and product-related impurities has been blurred by studies that found significant HCP content in mAb solution aggregates [13], [14], [15], [16], [17], [18], [19], [20], [21]. This was variously attributed to the electrostatic association of HCPs to chromatin–DNA complexes [13], [14], [15], [16], [17], [18], [19], the transient formation of charged mAb clusters [20] and to the unfolded protein response in cell culture [21]. Regardless of the origins, these structures appear to contribute to HCP persistence and we recently showed in a cross-digest proteomic analysis of high molecular weight (HMW) impurities that more HCPs were conserved among HCCF and protein A eluate aggregates than previously reported [22]. Such HCP-rich aggregates could potentially function as HCP reservoirs, as a typical mAb drug substance contains < 100 ppm HCP but ∼0.1–1.0% HMW species [23], [24], [25], [26]. How these HMW species may mediate HCP persistence through downstream operations remains an open question.

Some downstream operations are heuristically known to clear either HCPs or HMW impurities but not necessarily both. For instance, protein A chromatography is known to reduce the HCP content of HCCF substantially and several washes have been explored to improve this clearance. High-pH and arginine washes have usually provided meaningful benefits but their effects on aggregate persistence are less well understood [20,27,28]. Protein A chromatography is not necessarily considered beneficial for clearing HMW species because the product concentration effect coupled with the low-pH elution can promote aggregation [29], [30], [31], [32]. In polishing, cation-exchange (CEX) and hydrophobic interaction chromatography (HIC) are primarily used to remove HMW impurities but the HCP concentration can be decreased as well [33,34]. Flow-through anion-exchange (AEX) chromatography is known to impact HCP concentrations because the majority of HCPs are more acidic than the typical mAb but this step is not usually considered to be especially effective for aggregate removal [35], [36], [37].

Despite some apparent dissimilarities, the relationship between HCP and HMW species may be useful in developing polishing operations for HCP clearance. Enzyme-linked immunosorbent assays (ELISAs) are typically used for estimating HCP concentrations but such measurements are inherently semi-quantitative and do not identify individual HCP species [38,39]. Shotgun proteomic techniques using liquid chromatography-tandem mass spectrometry (LC-MS/MS) could overcome these deficiencies but they are generally time-intensive and prone to false negative identifications, which may preclude confident assertions about the removal of individual HCPs [22,40]. If the HMW content of mAb solutions could serve as an approximate surrogate measure of HCP concentrations, it may be useful for informing early process development decisions such as which resin should be selected for a given polishing operation.

In this article we report a primary analysis of aggregate clearance in processing steps that are typically implemented for HCP reduction. Specifically, factors that contribute to aggregate persistence through protein A and flow-through AEX chromatography are studied on the length scales of both resin beads and the column. Confocal laser scanning microscopy (CLSM) is used to probe aggregate retention profiles and to make qualitative comparisons of adsorption affinity under a variety of conditions, and column studies are used to probe resin capacities for aggregate retention. The total HCP concentration of flow-through AEX polishing pools is compared with their HMW content and a quantitative proteomic analysis is presented. The relevance of aggregate-mediated HCP persistence to depth filtration is also demonstrated in a scale-down format. The observations from this study may be useful in developing process understanding of impurity persistence phenomena.