GE Healthcare at ESACT 2019
Accelerate your bioprocess journey

Cell culture perfusion processes are considered optimal for a truly integrated continuous biomanufacturing pipeline. The nutrient-rich but balanced media should be designed to support very low cell-specific perfusion rates (CSPR) that minimize media consumption but maximize viable cell days and productivities. Optimized processes at low CSPR drastically reduce equipment costs, lab space, and product dilution. Finally, operating at very low CSPR enables running mammalian cell bioprocesses as true chemostat cultures in the future. We demonstrate a general workflow to develop high-performing perfusion media using small-scale models and transferred the process to 50 L scale at CSPR of 20 pL/c/d.

Recombinant CHO cells were evaluated at small scale in shaking tubes. Cells were grown in HyClone ActiPro or CDM4NS0 basal media, and optimal spike concentrations of HyClone Cell Boost supplements were determined using a DoE-supported workflow. The identified high-performing perfusion medium was applied to ReadyToProcess WAVE 25 and XDR-50 bioreactor runs. Different strategies were tested to find the critical minimum CSPR and maximum supported viable cell density (VCD). The obtained product profile was compared between scales, as determined by glycan-, charge-, and size-variant distribution.

Scale-down models were leveraged to define high-performing media and applied to bioreactor runs at constant volumetric perfusion rate, VCD, or CSPR. CSPR values as low as 10 pL/c/d at 2 × 10⁸ c/mL were achieved. These results make high-density perfusion processes suitable for inoculum preparation (N-1) or high cell density cryopreservation. The developed perfusion processes supported steady-state production at constant 5 × 10⁷ c/mL by applying a continuous cell bleed and were scaled to 50 L.

Worldwide, Rotavirus infection is a major cause of acute gastroenteritis with dehydrating diarrhea in infants and young children throughout the world. A number of live-attenuated, oral rotavirus vaccines are prequalified and available globally. Improved production technologies are needed to reduce costs and increase quality. Rotavirus is mainly produced in T-flasks, cell factories, or roller bottles using adherent Vero cells as cell substrate and porcine trypsin for rotavirus activation. As regulatory requirements become stricter, live vaccine production is moving away from animal-derived materials. Furthermore, closed bioreactor systems will help mitigate cross-contamination risks associated with open handling in current methods. Improved methods and solutions arising from academic and industrial research will result in updated production processes. These will benefit rotavirus production in the future.

In this work, we developed methods for rotavirus production in bioreactors using microcarriers and animal component-free materials. We describe different approaches for serum-free Vero cell cultivation using pre-sterilized microcarriers. Vero cells were grown in spinner flasks and a single-use WAVE bioreactor system, then infected with rotavirus during exponential growth. Infectious virus concentration was determined by fluorescence focus assay.

The results show significant differences between tested cell culture media, as well as suitability of recombinant trypsin for rotavirus activation. Pre-sterilized microcarrier cultures give similar cell-specific productivity compared with cells grown in T-flasks. Finally, the described culture conditions successfully propagate rotavirus in a WAVE bioreactor system.

Industry 4.0 is changing our manufacturing concepts (1). Biologicalisation uses 4.0 principles in concert with biological and bio-inspired materials, chemistries, and functions to support efficient and sustainable manufacturing. From product design to development and manufacturing, biomimetic product designs and bio-integrated manufacturing systems describe this biological transformation of manufacturing (2).

Many processes, chemistries, systems, and supply chains are enhanced by harmonizing digital manufacturing principles with biological structures and chemistries. This allows efficient and robust production to support a global circular economy. Progress in understanding biological elements, phenomena, materials, and chemistries enables this revolution. In fact, a 2018 Nobel Prize winner in Chemistry, Frances H. Arnold, illustrates this point. She invented systems directing the evolution of enzymes now routinely used to develop tools such as manufacturing catalysts. This technology also supports other 4.0 goals of more sustainable manufacturing of pharmaceuticals and renewable fuels (3).

The history of acetic acid manufacturing illustrates our move from biological to synthetic to bio-integrated manufacturing. The German Method of acetic acid production percolated an alcoholic solution through wood shavings with Acetobacter. We next generated it synthetically from inorganic and chlorinated intermediates (4). Today the biologicalisation of the process through genetic and metabolomic engineering, as well as digital manufacturing-based advances in fermentation, are promising more sustainable acetic manufacturing. Benefits include lower energy requirements and the possibility of low-cost carbohydrate sources such as organic wastes and agricultural residues.

In process development it is desirable to test process parameters in small scale and later scale up the process into larger production scale. This work demonstrates scalability from micro reactors up to large production-sized reactors. A theoretical scale-up study was performed on a mAb process based on physical characterizations done on the Xcellerex XDR platform. The theoretical study was the basis for the scalability study in XDR systems from 10 to 1000 L. Agitation and gassing settings suitable for the cell line and cultivation process were calculated and tested. Gas settings included sparger selection, gas management, and appropriate oxygen-to-air ratio for efficient CO² removal.

An in-house CHO cell line in ActiPro medium was cultivated in a fed-batch process in scales from 15 mL to 1000 L in the following bioreactors: ambr 15, XDR-10, XDR-50, XDR-200, and XDR-1000. From day 3 the reactor was fed with Cell Boost 7a and Cell Boost 7b once a day. Glucose addition started at day 5 when glucose levels decreased below 2 g/L. Each bioreactor was sampled daily; viable cell density (VCD), viability, product titer, pH, gases, nutrients, and metabolites were analyzed. Product quality (charge variants, glycosylation pattern, and aggregate to main fragment ratio) was analyzed from samples taken at day 7 and at harvest (day 13).

All investigated bioreactor systems displayed similar trends over time and acceptable variability between cultures for VCD and viability profile, product titer, lactate, and product quality. The work demonstrates process scalability from ambr 15 to XDR systems up to 1000 L scale. It also establishes general agitation and gassing settings suitable for the CHO-based mAb production process.

Hydrocyclones (HC) have been explored in a few past publications as cell retention devices for perfusion cultures. These low-cost, high-capacity devices are compact and free of clogging issues, and can be produced by 3D printing. These features make them an ideal cell retention device for single-use perfusion processes. However, previous reports on the use of HCs for mammalian cell perfusion were limited to non-disposable lab-scale bioreactors and to relatively low cell densities (up to ~10 million cells/mL). Thus, the aim of the present work was to evaluate the HC in an ADCF CHO process as a cell retention device coupled to a 50-L single-use bioreactor and to test its steady-state operation at cell densities in the range of 50 million cells/mL.

An HC prototype was connected to a single-use bioreactor bag (XDR-50, GE Healthcare). Three perfusion runs with CHO cells were performed at 40-L working volume, evaluating different ways of connecting the HC underflow port (concentrated stream) to the bag. Due to its high processing capacity, the HC was operated intermittently by easily connecting a simple timer to the conventional peristaltic pump used for cell recirculation.

We obtained for the first time in literature perfusion runs reaching steady states of up to 50 million cells/mL using an HC as cell retention device. The perfusion runs were operated at high cell viabilities, using cell-specific perfusion rates (CSPR) of 50 down to 15 pL/cell/d for 20-25 days. Pressure drops in the HC of 2 bar provided total separation efficiencies up to 96% and allowed a natural cell bleed to occur through the diluted overflow orifice. Comparisons of perfusion runs with this cell line and medium using other cell retention devices (ATF and inclined settler) are ongoing.