Search for: All records

ABSTRACT To enable adeno‐associated viral vectors (AAV) to achieve their maximum potential, next‐generation manufacturing processes must be developed to make gene therapies more affordable and accessible. This study focused on the design of two different intensified AAV downstream manufacturing processes at bench and pilot scale. Novel clarification methods were studied at bench scale, including the use of BioOptimal™ MF‐SL tangential flow microfilters for continuous removal of cell debris. Membrane adsorbers were used for further clarification, including DNA removal. Single pass tangential flow filtration (SPTFF) was implemented at bench scale by feeding the clarified cell lysate (CCL) into two Pellicon XL50 cassettes with 100 kDa regenerated cellulose membranes. At pilot scale, a multi‐membrane staged SPTFF module was designed to concentrate 10 L of AAV CCL. Both SPTFF systems provided 12X inline volumetric concentration with AAV yield > 99% after an appropriate buffer chase. Host cell protein removal was 48% and 37% for the bench and pilot scale processes, respectively. As an initial proof‐of‐concept, an integrated process was developed at pilot‐scale which linked clarification, SPTFF, and affinity chromatography. The integrated process offered an 81% reduction in total operating time (due to the reduced volume of load material for the affinity column after preconcentration by SPTFF), 36% improvement in affinity resin utilization (due to the higher AAV concentration in the column load), and an estimated 10% reduction in raw material costs. These improvements translated to an 8.5‐fold increase in overall productivity compared to an equivalent batch process, underscoring the potential for SPTFF to intensify large‐scale AAV downstream processing. 

Abstract Surfactants like polysorbate (Tween®) are commonly used as excipients in the production of monoclonal antibodies and other recombinant proteins. The retention behavior of these excipients in the final ultrafiltration step can be difficult to predict due to the presence of both monomers and micelles. This study examined the retention of polysorbate during ultrafiltration through cellulose and polyethersulfone membranes with nominal molecular weight cutoffs of 10, 30, and 100 kDa. Novel flux stepping experiments were performed to examine the effects of concentration polarization on surfactant transmission. Polysorbate 20 transmission through the 30 kDamembrane was a strong function of the surfactant concentration, decreasing from nearly 100% for a 2.5 mg/L solution to <10% for a 50 mg/L solution due to high retention of the micelles. Polysorbate transmission was lower for the polyethersulfone membrane due to polysorbate adsorption. A simple mathematical model was developed to describe the polysorbate transmission accounting for the effects of concentration polarization as well as the presence of surfactant monomers and micelles. Model calculations were in good agreement with the experimental data, providing a framework for the analysis and design of ultrafiltration/diafiltration processes for biopharmaceutical formulations containing surfactants. 

Abstract The design of effective ultrafiltration/diafiltration processes for protein formulation requires the use of membranes with very high protein retention. The objective of this study was to examine the effects of specific buffers on the retention of a model protein (bovine serum albumin) during ultrafiltration. Albumin retention at pH 4.8 was significantly reduced in phosphate buffer compared with that in acetate, citrate, and histidine. This behavior was consistent with a small change in the effective albumin hydrodynamic diameter as determined by dynamic light scattering. The underlying conformational changes leading to this change in diameter were explored using circular dichroism spectroscopy and differential scanning calorimetry. These results provide important insights into the factors controlling protein retention during ultrafiltration and diafiltration. 

Abstract Virus filtration is a crucial step in ensuring the high levels of viral clearance required in the production of biotherapeutics produced in mammalian cells or derived from human plasma. Previous studies have reported that virus retention is often reduced in the presence of therapeutic proteins due to membrane fouling; however, the underlying mechanisms controlling this behavior are still not well understood. Experimental studies were performed with a single layer of the commercially available dual‐layer Pegasus^TMSV4 virus removal filter to more easily interpret the experimental results. Bacteriophage ФX174 was used as a model parvovirus, and human immunoglobulin (hIgG) and Bovine Serum Albumin (BSA) were used as model proteins. Data obtained with 5 g/L solutions of hIgG showed more than a 100‐fold reduction in virus retention compared to that in the protein‐free solution. Similar effects were seen with membranes that were pre‐fouled with hIgG and then challenged with ФX174. The experimental data were well‐described using an internal polarization model that accounts for virus capture and accumulation within the virus filter, with the hIgG nearly eliminating the irreversible virus capture while also facilitating the release of previously captured virus. These results provide important insights into the performance and validation of virus removal filters in bioprocessing. 

Protein fouling can significantly reduce the filtrate flux, capacity, and virus retention during processing of plasma- or mammalian cell-derived biopharmaceuticals through virus removal filters. We use focused ion beam (FIB) milling and scanning electron microscopy (SEM) to directly evaluate changes in 3D pore structure in a Viresolve® Pro membrane due to fouling by human serum immunoglobulin G. Protein fouling causes a significant reduction in the membrane porosity, which decreases by approximately 40% in the size-selective region near the exit of the highly asymmetric Viresolve® Pro membrane after the filter is fouled to 90% flux decline. There is a corresponding reduction in the number of small pores by more than a factor of two. Model simulations of flow and particle transport in the protein-fouled membrane are in good agreement with independent experimental measurements of the permeability and location of particle capture. Simulations show an upstream shift in the location of nanoparticle capture (away from the filter exit) by about 0.4 µm for the membrane fouled to 90% flux decline. This is due to pore constriction from protein deposition, highlighting how fouling redistributes flow paths within the membrane. These results demonstrate the capability of using FIB-SEM to directly evaluate the effects of protein fouling on the 3D pore structure in virus removal filters, providing important insights into how protein fouling alters the performance of these highly selective membranes. 

The rapid development of adeno-associated viral vectors (AAV) to treat genetic disease has placed increased emphasis on the design of efficient downstream manufacturing processes. This study investigated the potential of using single pass tangential flow filtration (SPTFF) as a novel means of concentrating and purifying AAV clarified cell lysate (CCL). AAV stability studies revealed the shear-sensitive nature of the AAV capsids, with evidence of aggregation and fragmentation following repeated passages through a peristaltic pump (as would occur during batch ultrafiltration). SPTFF experiments focused on first identifying the membrane(s) that permitted high yield of AAV (negligible sieving into the permeate) along with substantial host cell protein (HCP) removal. Experiments were then performed at various permeate fluxes, which revealed that stable SPTFF processes can be achieved by operating below a critical flux for fouling (Jfoul). 300 kDa regenerated cellulose (RC) membranes were identified as optimal for this application, given their ability to provide complete AAV retention with high removal of HCP (>90%) when operated below Jfoul. The critical flux during SPTFF was increased by preconditioning the CCL through a positively-charged adsorptive filter, which reduced the concentration of foulants prior to SPTFF. These studies provide the first demonstration of SPTFF for the concentration and purification of AAV clarified cell lysate while minimizing shear exposure. 

Tangential flow microfiltration is easily adapted for batch and continuous bioreactor clarification. The permeate can be introduced directly to the subsequent capture step. However, the commercial use of tangential flow filtration (TFF) is limited by membrane fouling, leading to a compromised performance. Here, we explored the possibility of reducing membrane fouling by integrating a hydrocyclone as the primary clarification operation. The overflow from the hydrocyclone was introduced directly as the feed to the microfiltration module. Chinese hamster ovary cells were used as the feed stream to investigate the feasibility of this integrated process. A range of cell viabilities from 0% (cell lysate) to 96% were investigated. The cell densities ranged from 0.9 to 10 million cells per mL. Two commercially available hollow fiber microfiltration membranes were used, an essentially symmetric membrane and a reverse asymmetric membrane where the more open support structure faced the feed stream. The reverse asymmetric membrane was more resistant to fouling in the absence of an integrated hydrocyclone. Integrating a hydrocyclone led to a reduction in the flux decline for the symmetric membrane, but did not affect the performance of the reverse asymmetric membrane. The careful choice of membrane morphology and pore size is important when designing an integrated process. 

Virus filtration is used to ensure the high level of virus clearance required in the manufacture of biopharmaceutical products such as monoclonal antibodies. Flux decline during virus filtration can occur due to the formation of reversible aggregates consisting of self-assembled monomeric monoclonal antibody molecules, particularly at high antibody concentrations. While size exclusion chromatography is generally unable to detect these reversible aggregates, dynamic light scattering may be used to determine their presence. Flux decline during virus filtration may be minimized by pretreating the feed using a membrane adsorber in order to disrupt the reversible aggregates that are present. The formation of reversible aggregates is highly dependent on the monoclonal antibody and the feed conditions. For the pH values investigated here, pretreatment of the feed using a hydrophobic interaction membrane adsorber was the most effective in minimizing flux decline during virus filtration. Ion exchange membranes may also be effective if the monoclonal antibody and membrane are oppositely charged. Consequently, the effectiveness of ion exchange membrane adsorbers is much more dependent on solution pH when compared to hydrophobic interaction membrane adsorbers. Size based prefiltration was found to be ineffective at disrupting these reversible aggregates. These results can help guide the development of more effective virus filtration processes for monoclonal antibody production. 

Recent advances in the use of viral vectors for gene therapy has created a need for efficient downstream processing of these novel therapeutics. Single-pass tangential flow filtration (SPTFF) can potentially improve final product quality via reductions in shear, and it can increase manufacturing productivity via simple implementation into continuous/intensified processes. This study investigated the impact of variations in pressure and flow rate along the length of the membrane on overall SPTFF performance. Constant-flux filtration experiments at feed fluxes from 14 to 420 L/m2/h (Reynolds numbers <20) were performed using Pellicon® 3 TFF cassettes with fluorescent nanoparticles as model viral vectors. The location of nanoparticle accumulation shifted towards the filter outlet at high conversion and was also a function of the permeate flow configuration. These phenomena were explained using a newly developed concentration polarization model that predicts the distribution in local wall concentration over the length of the membrane. The model accurately captured the observed nanoparticle accumulation trends, including the effects of the permeate flow profile (co-current, divergent, or convergent flow) on nanoparticle accumulation within the SPTFF module. Nanoparticle accumulation at moderate conversion was more uniform using convergent flow, but nanoparticle accumulation at 80 % conversion (5x concentration factor) can be minimized using a divergent flow configuration. The local wall concentration model was also used to evaluate the critical flux by assuming that fouling occurs when the nanoparticle concentration at any point along the membrane surface exceeds 15 % by volume. These results provide important insights for the design and operation of SPTFF technology for inline concentration of viral vectors. 

In the biopharmaceutical industry, virus filters are crucial for ensuring the removal of endogenous and adventitious viruses as part of the viral clearance strategy. Although traditionally described as a size-exclusion mechanism, virus retention has a pro-cess-dependent nature where challenging conditions, such as process disruptions, may compromise membrane retention and significantly increase virus filtrate concentrations. The detailed mechanisms underlying this loss of retention are challenging to determine using traditional breakthrough experiments. In this work, single particle tracking and kinetic simulations were employed to connect individual particle behavior to the observed macroscopic losses in virus retention. Our experiments, using fluorescently labeled ΦX174 bacteriophage as a model parvovirus, replicated conditions representative of process disruptions within the Pegasus SV4, a homogeneous polymeric virus filtration membrane. During flow, phage particles retained were trapped within relatively large cavity spaces that had downstream constrictions aligned with the flow direction; the trapped particles were dynamic and exhibited significant intra-cavity motion. Upon flow stoppage, particles escaped from these retention locations rapidly, with approximately 90% of previously trapped particles being remobilized for process dis-ruption time ranging from 2 to 10 minutes, suggesting that local cavity escape had reached saturation at these timescales. Diffusion experiments within the membrane revealed isotropic and Fickian motion, hindered by more than an order of mag-nitude compared to diffusion in unconfined liquid. Despite the reduced mobility within the membrane, the substantial diffusion coefficient of 4.19 ± 0.06 µm²/s indicated that virus particles could travel tortuous but non-retentive pathways through the membrane on length scales equal to or greater than the membrane thickness during a disruption event. A 1D kinetic Monte-Carlo simulation successfully connected single-particle behavior to macroscopically observed virus release, indicating that significant diffusive release into the filtrate can occur even without the resumption of flow. This work provides crucial insights into the retention behavior of homogeneous membranes during periods of disruption, enabling the design of more robust mitigation strategies and filter designs.