Biophysical interactions of impurities, drug substance, and purification media in downstream bioprocessing
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
In the manufacturing of biologics, the downstream process includes several steps that cumulatively isolate the product from impurities, which range in size from cells to colloids such as host-cell proteins and nucleic acids. Unit operations that can remove these impurities commonly take advantage of the innate biophysical differences between product and impurity, such as charge and size, and are often themselves composed of purification media that preferentially interact with either product or impurity to enable isolation. There is an increasing incentive to understand the interactions that dominate these unit operations to mitigate potential risks in product development prior to final commercialization and develop novel technologies to keep pace with changes such as intensified upstream productivity or novel therapeutic modalities. The goal of this work is to better characterize these interactions across an array of size scales and unit operations, such as capture and harvest steps. ☐ Primary depth filtration is commonly used as a harvest step in downstream bioprocessing to remove cellular impurities. Despite its ubiquity, the biophysical interactions resulting in particle deposition in a depth filter are largely unstudied. Several popular depth filters from an array of manufacturers are used in this step in industry, and this work details the microstructural characterization of a library of many such layers, with an eye to assessing likely filtration mechanisms and efficiencies when the media layers are challenged with a typical industrial cell culture fluid. ☐ Typical cell culture fluids themselves are also analyzed, and the characterized filters and feeds were used in several depth filtration experiments where biophysical particle deposition mechanisms were inferred from increasing resistance to flow and measured filtrate turbidity as a function of fluid loading and observed by several imaging methods on the spent filter media. The mechanisms observed include adsorption, direct filtration (sieving), and filter caking, with the feed and filter in question playing a major role in the prevalence or even dominance of each mechanism. ☐ The knowledge gained from the characterization of cell culture fluid, filter, and their mechanistic interplay in-process is then used to develop a mechanistic model. Previous modeling and mechanistic work relevant to depth filtration is considered in the development of such a model, which also includes additional considerations specific to the filtration of mammalian cells. The resistance and filtrate turbidity data collected are used as a basis to validate the performance of the model across a wide range of relevant process parameters and material attributes, and additional industrial data are also used as a challenge to the model. ☐ Following harvest is typically capture, where challenges to the capture of virus-like particle (VLP) biologics are commonly seen due to their large size. This work describes a novel capture approach for VLPs wherein the VLP is not purified within a capture chromatography resin, as is common and feasible with small modalities, but is instead captured in bulk solution by a ligand with tunable solubility properties. These properties allow a continuous affinity precipitation and filtration approach to capture, where the binding and insolubility of a microparticle ligand-target complex is followed by the washing of soluble impurities into the filtrate of a hollow-fiber membrane module and subsequent resolubilization and separation of target from the ligand. ☐ In some bioprocessing situations, individual impurities may persist through capture step and even beyond, propagating until the final drug substance, where they can often have deleterious side effects on the product. This work adds additional context to an area under active investigation, the biophysical interactions enabling this impurity persistence through several downstream unit operations. One particular hypothesis addressed is the notion that these impurities strongly associate to the product, such that the impurities move through the process along with the product. The binding strengths of a set of these impurities is measured to a panel of archetypal products, and the binding locations associated with these binding strengths are assessed both experimentally and computationally.
Description
Keywords
Affinity precipitation, Depth filtration, Downstream bioprocessing