Daniel G. Bracewell

Soft sensors in bioprocessing: A status report and recommendations

The following report with recommendations is the result of an expert panel meeting on soft sensor applications in bioprocess engineering that was organized by the Measurement, Monitoring, Modelling and Control (M3C) Working Group of the European Federation of Biotechnology - Section of Biochemical Engineering Science (ESBES). The aim of the panel was to provide an update on the present status of the subject and to identify critical needs and issues for the furthering of the successful development of soft sensor methods in bioprocess engineering research and for industrial applications, in particular with focus on biopharmaceutical applications. It concludes with a set of recommendations, which highlight current prospects for the extended use of soft sensors and those areas requiring development.

Host cell protein dynamics in the supernatant of a mAb producing CHO cell line

The characterization of host cell protein (HCP) content during the production of therapeutic recombinant proteins is an important aspect in the drug development process. Despite this, key components of the HCP profile and how this changes with processing has not been fully investigated. Here we have investigated the supernatant HCP profile at different times throughout culture of a null and model GS‐CHO monoclonal antibody producing mammalian cell line grown in fed‐batch mode. Using 2D‐PAGE and LC‐MS/MS we identify a number of intracellular proteins (e.g., protein disulfide isomerise; elongation factor 2; calreticulin) that show a significant change in abundance relative to the general increase in HCP concentration observed with progression of culture. Those HCPs that showed a significant change in abundance across the culture above the general increase were dependent on the cell line examined. Further, our data suggests that the majority of HCPs in the supernatant of the cell lines investigated here arise through lysis or breakage of cells, associated with loss in viability, and are not present due to the secretion of protein material from within the cell. SELDI‐TOF and principal components analysis were also investigated to enable rapid monitoring of changes in the HCP profile. SELDI‐TOF analysis showed the same trends in the HCP profile as observed by 2D‐PAGE analysis and highlighted biomarkers that could be used for process monitoring. These data further our understanding of the relationship between the HCP profile and cell viability and may ultimately enable a more directed development of purification strategies and the development of cell lines based upon their HCP profile. Biotechnol. Bioeng. 2012; 109:971–982.

Advances in product release strategies and impact on bioprocess design

Intracellular products such as recombinant insulin, which are typically produced in microbial host cells, demand a product release step to remove them from the cell. How this is performed determines the quantity of released contaminants, the particle size distribution of cell debris and the physical properties of the resultant process stream, which all impact on the performance of the downstream operations. Thus, achieving selective release of the desired product is crucial for improving the process economics. Advances in upstream processing (the bioreactor phase) have been successful in achieving high product titres, and downstream costs now typically dominate the overall manufacturing costs. Here, we review and discuss the selective release of products as a possible means of improving the efficiency of downstream processing.

Bioprocess Engineering Issues That Would Be Faced in Producing a DNA Vaccine at up to 100 m3 Fermentation Scale for an Influenza Pandemic

The risk of a pandemic with a virulent form of influenza is acknowledged by the World Health Organization (WHO) and other agencies. Current vaccine production facilities would be unable to meet the global requirement for vaccine. As a possible supplement a DNA vaccine may be appropriate, and bioprocess engineering factors bearing on the use of existing biopharmaceutical and antibiotics plants to produce it are described. This approach addresses the uncertainty of timing of a pandemic that precludes purpose‐built facilities. The strengths and weaknesses of alternative downstream processing routes are analyzed, and several gaps in public domain information are addressed. The conclusion is that such processing would be challenging but feasible.

Optimising the design and operation of semi-continuous affinity chromatography for clinical and commercial manufacture

This paper presents an integrated experimental and modelling approach to evaluate the potential of semi-continuous chromatography for the capture of monoclonal antibodies (mAb) in clinical and commercial manufacture. Small-scale single-column experimental breakthrough studies were used to derive design equations for the semi-continuous affinity chromatography system. Verification runs with the semi-continuous 3-column and 4-column periodic counter current (PCC) chromatography system indicated the robustness of the design approach. The product quality profiles and step yields (after wash step optimisation) achieved were comparable to the standard batch process. The experimentally-derived design equations were incorporated into a decisional tool comprising dynamic simulation, process economics and sizing optimisation. The decisional tool was used to evaluate the economic and operational feasibility of whole mAb bioprocesses employing PCC affinity capture chromatography versus standard batch chromatography across a products lifecycle from clinical to commercial manufacture. The tool predicted that PCC capture chromatography would offer more significant savings in direct costs for early-stage clinical manufacture (proof-of-concept) (∼30%) than for late-stage clinical (∼10-15%) or commercial (∼5%) manufacture. The evaluation also highlighted the potential facility fit issues that could arise with a capture resin (MabSelect) that experiences losses in binding capacity when operated in continuous mode over lengthy commercial campaigns. Consequently, the analysis explored the scenario of adopting the PCC system for clinical manufacture and switching to the standard batch process following product launch. The tool determined the PCC system design required to operate at commercial scale without facility fit issues and with similar costs to the standard batch process whilst pursuing a process change application. A retrofitting analysis established that the direct cost savings obtained by 8 proof-of-concept batches would be sufficient to pay back the investment cost of the pilot-scale semi-continuous chromatography system.

The dynamics of the CHO host cell protein profile during clarification and protein A capture in a platform antibody purification process

Recombinant protein products such as monoclonal antibodies (mAbs) for use in the clinic must be clear of host cell impurities such as host cell protein (HCP), DNA/RNA, and high molecular weight immunogenic aggregates. Despite the need to remove and monitor HCPs, the nature, and fate of these during downstream processing (DSP) remains poorly characterized. We have applied a proteomic approach to investigate the dynamics and fate of HCPs in the supernatant of a mAb producing cell line during early DSP including centrifugation, depth filtration, and protein A capture chromatography. The primary clarification technique selected was shown to influence the HCP profile that entered subsequent downstream steps. MabSelect protein A chromatography removed the majority of contaminating proteins, however using 2D‐PAGE we could visualize not only the antibody species in the eluate (heavy and light chain) but also contaminant HCPs. These data showed that the choice of secondary clarification impacts upon the HCP profile post‐protein A chromatography as differences arose in both the presence and abundance of specific HCPs when depth filters were compared. A number of intracellularly located HCPs were identified in protein A elution fractions from a Null cell line culture supernatant including the chaperone Bip/GRP78, heat shock proteins, and the enzyme enolase. We demonstrate that the selection of early DSP steps influences the resulting HCP profile and that 2D‐PAGE can be used for monitoring and identification of HCPs post‐protein A chromatography. This approach could be used to screen cell lines or hosts to select those with reduced HCP profiles, or to identify HCPs that are problematic and difficult to remove so that cell‐engineering approaches can be applied to reduced, or eliminate, such HCPs. Biotechnol. Bioeng. 2013; 110: 240–251.

Host cell protein adsorption characteristics during protein A chromatography.

Protein A chromatography is a critical and ‘gold‐standard’ step in the purification of monoclonal antibody (mAb) products. Its ability to remove >98% of impurities in a single step alleviates the burden on subsequent process steps and facilitates the implementation of platform processes, with a minimal number of chromatographic steps. Here, we have evaluated four commercially available protein A chromatography matrices in terms of their ability to remove host cell proteins (HCPs), a complex group of process related impurities that must be removed to minimal levels. SELDI‐TOF MS was used as a screening tool to generate an impurity profile fingerprint for each resin and indicated a number of residual impurities present following protein A chromatography, agreeing with HCP ELISA. Although many of these were observed for all matrices there was a significantly elevated level of impurity binding associated with the resin based on controlled pore glass under standard conditions. Use of null cell line supernatant with and without spiked purified mAb demonstrated the interaction of HCPs to be not only with the resin back‐bone but also with the bound mAb. A null cell line column overload and sample enrichment method before 2D‐PAGE was then used to determine individual components associated with resin back‐bone adsorption. The methods shown allow for a critical analysis of HCP removal during protein A chromatography. Taken together they provide the necessary process understanding to allow process engineers to identify rational approaches for the removal of prominent HCPs.

Determining antibody stability: creation of solid-liquid interfacial effects within a high shear environment.

The purpose of this study was to assess the stability of protein formulations using a device designed to generate defined, quantifiable levels of shear in the presence of a solid‐liquid interface. The device, based on a rotating disk, produced shear strain rates of up to 3.4 × 104 s−1 (at 250 rps) and was designed to exclude air‐liquid interfaces and enable temperature to be controlled. Computational fluid dynamics (CFD) was used to study the fluid flow patterns within the device and to determine the shear strain rate (s−1) at a range of disk speeds. The device was then used to study the effect on a monoclonal IgG4 of high levels of shear at the solid‐liquid interface. Monomeric antibody concentration and aggregation of the protein in solution were monitored by gel permeation HPLC and turbidity at 350 nm. High shear strain rates were found to cause significant levels of protein aggregation and precipitation with reduction of protein monomer following first‐order kinetics. Monomer reduction rate was determined for a range of disk speeds and found to have a nonlinear relationship with shear strain rate, indicating the importance of identifying and minimizing such environments during processing.

An automated microscale chromatographic purification of virus‐like particles as a strategy for process development

The development of fermentation processes for recombinant vaccines requires optimizing expression while maintaining high product quality. Changes to cell fermentation conditions are typically evaluated following cell disruption, with expression levels quantified by immunoassay, liquid chromatography or enzyme activity. However, assay titres do not always predict the effects that intracellular aggregation, proteolysis, post‐translational modifications and differences in relative impurity levels can have on purification yield and product purity. Furthermore, heterogeneity in the size and surface properties inherent in viral particles makes unit operations such as chromatography less predictable. In these cases, the purification procedure (or a mimic thereof) must be carried out to give accurate information on the impact of changes in fermentation conditions on purification process performance. This was demonstrated for the development of a recombinant vaccine against human papillomavirus produced in Saccharomyces cerevisiae, where the most informative feedback on fermentation variables was obtained by completing a multistep chromatographic purification to evaluate process yield and product purity. To increase the purification throughput and reduce labour, the chromatography was miniaturized 1000‐fold from the laboratory scale using microlitre volumes of adsorbent in a pipette tip and automated on a robotic workstation. The microscale purification is shown to be predictive of the laboratory‐scale purification in terms of yield and purity, while providing over a 10‐fold increase in throughput and allowing for increased monitoring of fermentation processes. In addition, by reducing the volume of cells needed for this assessment, the fermentation can be correspondingly reduced in scale and carried out in parallel for additional throughput gains.

A microscale yeast cell disruption technique for integrated process development strategies

Miniaturizing protein purification processes at the microliter scale (microscale) holds the promise of accelerating process development by enabling multi‐parallel experimentation and automation. For intracellular proteins expressed in yeast, small‐scale cell breakage methods capable of disrupting the rigid cell wall are needed that can match the protein release and contaminant profile of full‐scale methods like homogenization, thereby enabling representative studies of subsequent downstream operations to be performed. In this study, a noncontact method known as adaptive focused acoustics (AFA) was optimized for the disruption of milligram quantities of yeast cells for the subsequent purification of recombinant human papillomavirus (HPV) virus‐like particles (VLPs). AFA operates by delivering highly focused, computer‐controlled acoustic radiation at frequencies significantly higher than those used in conventional sonication. With this method, the total soluble protein release was equivalent to that of laboratory‐scale homogenization, and cell disruption was evident by light microscopy. The recovery of VLPs through a microscale chromatographic purification following AFA treatment was within 10% of that obtained using homogenization, with equivalent product purity. The addition of a yeast lytic enzyme prior to cell disruption reduced processing time by nearly 3‐fold and further improved the comparability of the lysate to that of the laboratory‐scale homogenate. In addition, unlike conventional sonication methods, sample heating was minimized (≤8 °C increase), even using the maximum power settings required for yeast cell disruption. This disruption technique in combination with microscale chromatographic methods for protein purification enables a strategy for the rapid process development of intracellularly expressed proteins.