Mammalian cell retention devices for stirred perfusion bioreactors

Within the spectrum of current applications for cell culture technologies, efficient large-scale mammalian cell production processes are typically carried out in stirred fed-batch or perfusion bioreactors. The specific aspects of each individual process that can be considered when determining the method of choice are presented. A major challenge for perfusion reactor design and operation is the reliability of the cell retention device. Current retention systems include cross-flow membrane filters, spin-filters, inclined settlers, continuous centrifuges and ultrasonic separators. The relative merits and limitations of these technologies for cell retention and their suitability for large-scale perfusion are discussed. This revised version was published online in August 2006 with corrections to the Cover Date.     

Acoustic cell filter: a proven cell retention technology for perfusion of animal cell cultures

This article is a review highlighting the application of the acoustic filter as a reliable cell retention device during the long-term perfusion of animal cell cultures. Critical operating parameters such as duty cycle, perfusion and re-circulation flow rates, acoustic power and backflush frequency are discussed with regard to influence on the separation efficiency and optimal operating ranges have been identified. Perfusion data gathered from the literature have been complemented with original data from a series of perfusion experiments carried out in the context of industrial projects for industrially relevant cell lines including NS0, HEK-293, SP2-derived hybridoma and insect cells in different serum-supplemented and serum-free media at different perfusion rates and acoustic chamber volumes. Finally, scale-up potential of the acoustic filter for large-scale industrial applications is discussed.     

Process intensification strategies toward cell culture-based high-yield production of a fusogenic oncolytic virus

We present a proof-of-concept study for production of a recombinant vesicular stomatitis virus (rVSV)-based fusogenic oncolytic virus (OV), rVSV-Newcastle disease virus (NDV), at high cell densities (HCD). Based on comprehensive experiments in 1 L stirred tank reactors (STRs) in batch mode, first optimization studies at HCD were carried out in semi-perfusion in small-scale cultivations using shake flasks. Further, a perfusion process was established using an acoustic settler for cell retention. Growth, production yields, and process-related impurities were evaluated for three candidate cell lines (AGE1.CR, BHK-21, HEK293SF)infected at densities ranging from 15 to 30 × 10⁶ cells/mL. The acoustic settler allowed continuous harvesting of rVSV-NDV with high cell retention efficiencies (above 97%) and infectious virus titers (up to 2.4 × 10⁹ TCID₅₀/mL), more than 4–100 times higher than for optimized batch processes. No decrease in cell-specific virus yield (CSVY) was observed at HCD, regardless of the cell substrate. Taking into account the accumulated number of virions both from the harvest and bioreactor, a 15–30 fold increased volumetric virus productivity for AGE1.CR and HEK293SF was obtained compared to batch processes performed at the same scale. In contrast to all previous findings, formation of syncytia was observed at HCD for the suspension cells BHK 21 and HEK293SF. Oncolytic potency was not affected compared to production in batch mode. Overall, our study describes promising options for the establishment of perfusion processes for efficient large-scale manufacturing of fusogenic rVSV-NDV at HCD for all three candidate cell lines.     

An on-line method for the reduction of fouling of spin-filters for animal cell perfusion cultures

The main limitation in the use of spin-filters during perfusion cultures of animal cells was revealed to be filter fouling. This phenomenon involves cell-sieve interactions as well as cell attachment to, and growth on, the filter surface. The cell attachment effect has been analysed in the present study during long-term perfusion simulations with CHO animal cells. It was demonstrated that at low filter acceleration, below 6.2 m/s2, a high perfusion rate of 25 cm/h induced rapid filter pore clogging within 3 days, whereas increasing the filter acceleration to 25 m/s2 increased filter longevity from 3 to 25 days, for filters with a pore size of 8.5 microm. Increasing the filter pore size to 14.5 microm improved filter longevity by 84% with less viable and dead cell deposits on the filter surface. However, it was demonstrated that filter longevity was not necessarily dependent on the amount of cell deposit on the filter surface. In the second part of this study, ultrasonic technology was used to reduce filter fouling. Filter vibration, induced by a piezo actuator, improved filter longevity by 113% during CHO cells perfusion cultures.     

Construction of multicomponent catalytic films based on avidin-biotin technology for the electroenzymatic oxidation of molecular hydrogen

The ability to predict the performance of large-scale processes is central to the rapid development of successful operations at the pilot and industrial scale. In this article, we examine the operation, at laboratory scale, of precipitation reactors and centrifuges for protein precipitate recovery and dewatering and how they might best mimic large-scale reactors and centrifuges, in this case, a pilot-scale batch stirred-tank reactor and a multichamber-bowl centrifuge. Novel approaches to bench-top centrifuge operation are provided, in particular with a view to delivery of material for subsequent high-resolution purification, which would be obtained at full pilot scale. Results are presented in terms of properties of the protein precipitates, the fraction of solids recovered, and the extent of dewatering achieved. Good agreement was obtained at bench scale (a 1000-fold scale down factor) for all of these parameters for pilot-scale, batch-feed operation. In addition, the methodology developed allows identification of the extent of break-up that occurs in continuous-feed centrifuges when processing shear-sensitive materials such as the protein precipitates studied here.     

Optimization of an acoustic cell filter with a novel air-backflush system

Increasing worldwide demand for mammalian cell production capacity will likely be partially satisfied by a greater use of higher volumetric productivity perfusion processes. An important additional component of any perfusion system is the cell retention device that can be based on filtration, sedimentation, and/or acoustic technologies. A common concern with these systems is that pumping and transient exposure to suboptimal medium conditions may damage the cells or influence the product quality. A novel air-backflush mode of operating an acoustic cell separator was developed in which an injection of bioreactor air downstream of the separator periodically returned the captured cells to the reactor, allowing separation to resume within 20 s. This mode of operation eliminated the need to pump the cells and allows the selection of a residence time in the separator depending on the sensitivity of the cell line. The air-backflush mode of operating a 10L acoustic separator was systematically tested at 10(7) cells/mL to define reliable ranges of operation. Consistent separation performance was obtained for wide ranges of cooling airflow rates from 0 to 15 L/min and for backflush frequencies between 10 and 40 h(-1). The separator performance was optimized at a perfusion rate of 10 L/day to obtain a maximum separation efficiency of 92 +/- 0.3%. This was achieved by increasing the power setting to 8 W and using duty cycle stop and run times of 4.5 and 45 s, respectively. Acoustic cell separation with air backflush was successfully applied over a 110 day CHO cell perfusion culture at 10(7) cells/mL and 95% viability.     

A novel scale-down mimic of perfusion cell culture using sedimentation in an automated microbioreactor (SAM)

Continuous upstream processing in mammalian cell culture for recombinant protein production holds promise to increase product yield and quality. To facilitate the design and optimization of large-scale perfusion cultures, suitable scale-down mimics are needed which allow high-throughput experiments to be performed with minimal raw material requirements. Automated microbioreactors are available that mimic batch and fed-batch processes effectively but these have not yet been adapted for perfusion cell culture. This article describes how an automated microbioreactor system (ambr15) can be used to scale-down perfusion cell cultures using cell sedimentation as the method for cell retention. The approach accurately predicts the viable cell concentration, in the range of about 1 × 10⁷ cells/mL for a human cell line, and cell viability of larger scale cultures using a hollow fiber based cell retention system. While it was found to underpredict cell line productivity, the method accurately predicts product quality attributes, including glycosylation profiles, from cultures performed in bioreactors with working volumes between 1 L and 1,000 L. The spent media exchange method using the ambr15 was found to predict the influence of different media formulations on large-scale perfusion cultures in contrast to batch and chemostat experiments performed in the microbioreactor system. The described experimental setup in the microbioreactor allowed an 80-fold reduction in cell culture media requirements, half the daily operator time, which can translate into a cost reduction of approximately 2.5-fold compared to a similar experimental setup at bench scale.     

Shear stress analysis of mammalian cell suspensions for prediction of industrial centrifugation and its verification

This article describes the use of ultra scale-down studies requiring milliliter quantities of process material to study the clarification of mammalian cell culture broths using industrial-scale continuous centrifuges during the manufacture of a monoclonal antibody for therapeutic use. Samples were pretreated in a small high-speed rotating-disc device in order to mimic the effect on the cells of shear stresses in the feed zone of the industrial scale centrifuges. The use of this feed mimic was shown to predict a reduction of the clarification efficiency by significantly reducing the particle size distribution of the mammalian cells. The combined use of the rotating-disc device and a laboratory-scale test tube centrifuge successfully predicted the separation characteristics of industrial-scale, disc stack centrifuges operating with different feed zones. A 70% reduction in flow rate in the industrial-scale centrifuge was shown to arise from shear effects. A predicted 2.5-fold increase in throughput for the same clarification performance, achieved by the change to a centrifuge using a feed zone designed to give gentler acceleration of the bioprocess fluid, was also verified at large-scale.     

Modeling industrial centrifugation of mammalian cell culture using a capillary based scale-down system

Continuous-flow centrifugation is widely utilized as the primary clarification step in the recovery of biopharmaceuticals from cell culture. However, it is a challenging operation to develop and characterize due to the lack of easy to use, small-scale, systems that can be used to model industrial processes. As a result, pilot-scale continuous centrifugation is typically employed to model large-scale systems requiring a significant amount of resources. In an effort to reduce resource requirements and create a system which is easy to construct and utilize, a capillary shear device, capable of producing energy dissipation rates equivalent to those present in the feed zones of industrial disk stack centrifuges, was developed and evaluated. When coupled to a bench-top, batch centrifuge, the capillary device reduced centrate turbidity prediction error from 37% to 4% compared to using a bench-top centrifuge alone. Laboratory-scale parameters that are analogous to those routinely varied during industrial-scale continuous centrifugation were identified and evaluated for their utility in emulating disk stack centrifuge performance. The resulting relationships enable bench-scale process modeling of continuous disk stack centrifuges using an easily constructed, scalable, capillary shear device coupled to a typical bench-top centrifuge.     

A computer-aided approach to compare the production economics of fed-batch and perfusion culture under uncertainty

Fed-batch and perfusion culture dominate mammalian cell culture production processes. In this paper, a decision-support tool was employed to evaluate the economic feasibility of both culture modes via a case study based upon the large-scale production of monoclonal antibodies. The trade-offs between the relative simplicity but higher start-up costs of fed-batch processes and the high productivity but higher chances of equipment failure of perfusion processes were analysed. Deterministic analysis showed that whilst there was an insignificant difference (3%) between the cost of goods per gram (COG/g) values, the perfusion option benefited from a 42% reduction in capital investment and a 12% higher projected net present value (NPV). When Monte Carlo simulations were used to account for uncertainties in titre and yield, as well as the risks of contamination and filter fouling, the frequency distributions for the output metrics revealed that neither process route offered the best of both NPV or product output. A product output criterion was formulated and the options that met the criterion were compared based on their reward/risk ratio. The perfusion option was no longer feasible as it failed to meet the product output criterion and the fed-batch option had a 100% higher reward/risk ratio. The tool indicated that in this particular case, the probabilities of contamination and fouling in the perfusion option need to be reduced from 10% to 3% for this option to have the higher reward/risk ratio. The case study highlighted the limitations of relying on deterministic analysis alone.     

Stable hybridoma cultivation in a pilot-scale acoustic perfusion system: long-term process performance and effect of recirculation rate

Perfusion systems have the possibility to be operated continuously for several months. It is important that the performance of the cell retention device does not limit the operation time of a perfusion process used in the production of active pharmaceutical ingredients. Therefore, the aim of this study was to investigate the reliability and long-term stability of an acoustic perfusion process using the 200 L/d BioSep. As the BioSep is an external device, it is possible that dependent on the recirculation rate nutrient gradients occur in the external loop, which could affect the cell metabolism. Therefore, the effect of possible nutrient gradients on cell metabolism, viability and productivity was studied by varying the recirculation rate. In this study, it is shown that a perfusion process using a pilot-scale acoustic cell-retention device (200 L/d) is reliable and simple to operate, resulting in a stable 75-day cultivation of a hybridoma cell line producing a monoclonal antibody. The recirculation rate had a significant effect on the oxygen concentration in the external loop, with oxygen being depleted within the cell-retention device at recirculation rates below 6 m3/m(reactor)3.d (=600 L/d). The oxygen depletion at low circulation rates correlated with a slightly increased lactate production rate. For all other parameters no effect of the recirculation rate was observed, including cell death measured through the release of lactate dehydrogenase and specific productivity. A maximum specific productivity of 12 pg/cell.d was reached.     

Scale-up and optimization of an acoustic filter for 200 L/day perfusion of a CHO cell culture

Acoustic cell retention devices have provided a practical alternative for up to 50 L/day perfusion cultures but further scale-up has been limited. A novel temperature-controlled and larger-scale acoustic separator was evaluated at up to 400 L/day for a 10(7) CHO cell/mL perfusion culture using a 100-L bioreactor that produced up to 34 g/day recombinant protein. The increased active volume of this scaled-up separator was divided into four parallel compartments for improved fluid dynamics. Operational settings of the acoustic separator were optimized and the limits of robust operations explored. The performance was not influenced over wide ranges of duty cycle stop and run times. The maximum performance of 96% separation efficiency at 200 L/day was obtained by setting the separator temperature to 35.1 degrees C, the recirculation rate to three times the harvest rate, and the power to 90 W. While there was no detectable effect on culture viability, viable cells were selectively retained, especially at 50 L/day, where there was a 5-fold higher nonviable washout efficiency. Overall, the new temperature-controlled and scaled-up separator design performed reliably in a way similar to smaller-scale acoustic separators. These results provide strong support for the feasibility of much greater scale-up of acoustic separations.     

Compact Cell Settlers for Perfusion Cultures of Microbial (and Mammalian) Cells

As microbial secretory expression systems have become well developed for microbial yeast cells, such as Saccharomyces cerevisiae and Pichia pastoris, it is advantageous to develop high cell density continuous perfusion cultures of microbial yeast cells to retain the live and productive yeast cells inside the perfusion bioreactor while removing the dead cells and cell debris along with the secreted product protein in the harvest stream. While the previously demonstrated inclined or lamellar settlers can be used for such perfusion bioreactors for microbial cells, the size and footprint requirements of such inefficiently scaled up devices can be quite large in comparison to the bioreactor size. Faced with this constraint, we have now developed novel, patent‐pending compact cell settlers that can be used more efficiently with microbial perfusion bioreactors to achieve high cell densities and bioreactor productivities. Reproducible results from numerous month‐long perfusion culture experiments using these devices attached to the 5 L perfusion bioreactor demonstrate very high cell densities due to substantial sedimentation of the larger live yeast cells which are returned to the bioreactor, while the harvest stream from the top of these cell settlers is a significantly clarified liquid, containing less than 30% and more typically less than 10% of the bioreactor cell concentration. Size of cells in the harvest is smaller than that of the cells in the bioreactor. Accumulated protein collected from the harvest and rate of protein accumulation is significantly (> 6x) higher than the protein produced in repeated fed‐batch cultures over the same culture duration. © 2017 American Institute of Chemical Engineers Biotechnol. Prog., 33:913–922, 2017     

Neutrophil retention in model capillaries: deformability, geometry, and hydrodynamic forces

The initial retention of neutrophils within the pulmonary microvascular bed occurs in both physiological and pathological states, yet the factors responsible for this retention are poorly understood. Because the diameter of the neutrophil is approximately 7.03 micron and the mean pulmonary capillary diameter is 5.5 micron, we postulated that geometric constraints imposed by the microvascular bed, the deformability of the neutrophil, and the hydrodynamic characteristics of blood were important determinants of neutrophil retention. We used a filtration system wherein 111In-labeled human neutrophils (111In-N) suspended in a serum-containing buffer were passed through Nuclepore filters of known pore size. Compared with 99mTc-labeled erythrocytes (99mTc-RBC), the passage of 111In-N was delayed and a higher percentage was retained within the filter. Because the neutrophil and RBC are approximately equal in diameter, the deformability of the neutrophil must be less than that of RBC. As the flow rate increased, retention in the filters decreased logarithmically from 72 +/- 5% (flow rate 0.5 ml/min) to 15 +/- 4% (10.0 ml/min). As the number of RBC in the buffer increased, neutrophil retention in 5-micron filters decreased in a linear fashion from 65 +/- 6% at hematocrit of 0 to 33 +/- 2% at hematocrit of 10. The perfusion pressure and shear stress were of critical importance, and there was a logarithmic relationship between retention and perfusion pressure or shear stress (tau), whether the increase in pressure or tau was generated by increasing flow or by increasing the hematocrit of the perfusate. As the pore size of the filter increased, the retention of neutrophils decreased in a logarithmic fashion: from 75 +/- 5% in the 3-micron filter to 4 +/- 1.3% in the 12-micron filter.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization and optimization of acoustic filter performance by experimental design methodology

Acoustic cell filters operate at high separation efficiencies with minimal fouling and have provided a practical alternative for up to 200 L/d perfusion cultures. However, the operation of cell retention systems depends on several settings that should be adjusted depending on the cell concentration and perfusion rate. The impact of operating variables on the separation efficiency performance of a 10-L acoustic separator was characterized using a factorial design of experiments. For the recirculation mode of separator operation, bioreactor cell concentration, perfusion rate, power input, stop time and recirculation ratio were studied using a fractional factorial 2(5-1) design, augmented with axial and center point runs. One complete replicate of the experiment was carried out, consisting of 32 more runs, at 8 runs per day. Separation efficiency was the primary response and it was fitted by a second-order model using restricted maximum likelihood estimation. By backward elimination, the model equation for both experiments was reduced to 14 significant terms. The response surface model for the separation efficiency was tested using additional independent data to check the accuracy of its predictions, to explore robust operation ranges and to optimize separator performance. A recirculation ratio of 1.5 and a stop time of 2 s improved the separator performance over a wide range of separator operation. At power input of 5 W the broad range of robust high SE performance (95% or higher) was raised to over 8 L/d. The reproducible model testing results over a total period of 3 months illustrate both the stable separator performance and the applicability of the model developed to long-term perfusion cultures.     

Production of Modified Vaccinia Ankara Virus by Intensified Cell Cultures: A Comparison of Platform Technologies for Viral Vector Production

Modified Vaccinia Ankara (MVA) virus is a promising vector for vaccination against various challenging pathogens or the treatment of some types of cancers, requiring a high amount of virions per dose for vaccination and gene therapy. Upstream process intensification combining perfusion technologies, the avian suspension cell line AGE1.CR.pIX and the virus strain MVA-CR19 is an option to obtain very high MVA yields. Here the authors compare different options for cell retention in perfusion mode using conventional stirred-tank bioreactors. Furthermore, the authors study hollow-fiber bioreactors and an orbital-shaken bioreactor in perfusion mode, both available for single-use. Productivity for the virus strain MVA-CR19 is compared to results from batch and continuous production reported in literature. The results demonstrate that cell retention devices are only required to maximize cell concentration but not for continuous harvesting. Using a stirred-tank bioreactor, a perfusion strategy with working volume expansion after virus infection results in the highest yields. Overall, infectious MVA virus titers of 2.1–16.5 × 10⁹ virions/mL are achieved in these intensified processes. Taken together, the study shows a novel perspective on high-yield MVA virus production in conventional bioreactor systems linked to various cell retention devices and addresses options for process intensification including fully single-use perfusion platforms.     

CFD simulation of an internal spin-filter: evidence of lateral migration and exchange flow through the mesh

In the present work Computational Fluid Dynamics (CFD) was used to study the flow field and particle dynamics in an internal spin-filter (SF) bioreactor system. Evidence of a radial exchange flow through the filter mesh was detected, with a magnitude up to 130-fold higher than the perfusion flow, thus significantly contributing to radial drag. The exchange flow magnitude was significantly influenced by the filter rotation rate, but not by the perfusion flow, within the ranges evaluated. Previous reports had only given indirect evidences of this exchange flow phenomenon in spin-filters, but the current simulations were able to quantify and explain it. Flow pattern inside the spin-filter bioreactor resembled a typical Taylor-Couette flow, with vortices being formed in the annular gap and eventually penetrating the internal volume of the filter, thus being the probable reason for the significant exchange flow observed. The simulations also showed that cells become depleted in the vicinity of the mesh due to lateral particle migration. Cell concentration near the filter was approximately 50% of the bulk concentration, explaining why cell separation achieved in SFs is not solely due to size exclusion. The results presented indicate the power of CFD techniques to study and better understand spin-filter systems, aiming at the establishment of effective design, operation and scale-up criteria.     

Cell separator operation within temperature ranges to minimize effects on Chinese hamster ovary cell perfusion culture

A cell retention device that provides reliable high-separation efficiency with minimal negative effects on the cell culture is essential for robust perfusion culture processes. External separation devices generally expose cells to periodic variations in temperature, most commonly temperatures below 37 degrees C, while the cells are outside the bioreactor. To examine this phenomenon, aliquots of approximately 5% of a CHO cell culture were exposed to 60 s cyclic variations of temperature simulating an acoustic separator environment. It was found that, for average exposure temperatures between 31.5 and 38.5 degrees C, there were no significant impacts on the rates of growth, glucose consumption, or t-PA production, defining an acceptable range of operating temperatures. These results were subsequently confirmed in perfusion culture experiments for average exposure temperatures between 31.6 and 38.1 degrees C. A 2(5-1) central composite factorial design experiment was then performed to systematically evaluate the effects of different operating variables on the inlet and outlet temperatures of a 10L acoustic separator. The power input, ambient temperature, as well as the perfusion and recycle flow rates significantly influenced the temperature, while the cell concentration did not. An empirical model was developed that predicted the temperature changes between the inlet and the outlet of the acoustic separator within +/-0.5 degrees C. A series of perfusion experiments determined the ranges of the significant operational settings that maintained the acoustic separator inlet and outlet temperatures within the acceptable range. For example, these objectives were always met by using the manufacturer-recommended operational settings as long as the recirculation flow rate was maintained above 15 L day(-1) and the ambient temperature was near 22 degrees C.     

Mechanistic failure mode investigation and resolution of parvovirus retentive filters

Virus retentive filters are a key product safety measure for biopharmaceuticals. A simplistic perception is that they function solely based on a size‐based particle removal mechanism of mechanical sieving and retention of particles based on their hydrodynamic size. Recent observations have revealed a more nuanced picture, indicating that changes in viral particle retention can result from process pressure and/or flow interruptions. In this study, a mechanistic investigation was performed to help identify a potential mechanism leading to the reported reduced particle retention in small virus filters. Permeate flow rate or permeate driving force were varied and analyzed for their impact on particle retention in three commercially available small virus retentive filters. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:959–970, 2016