On-line gas analysis in animal cell cultivation: I. Control of dissolved oxygen and pH

To monitor gas reaction rates in animal cell culture at constant dissolved oxygen concentration (DO) and constant pH it was necessary to develop improved control methods. Decoupling of both controllrs was obtained by manipulation of molar fractions of oxygen and carbon dioxide in the gas phase. Two pairs of DO and pH controllers were designed and tested both in simulation and exprimental runs. The first controller pair was developed for headspace aeration only, whereas the second controller pair was designed for bubble aeration using a microsparger and flushing the headspace with helium. pH was controlled by a conventional discrete PID controller in its velocity form. For DO control two linear state space feedback controllers with parameter adaptation were established. In these controllers the oxygen uptake rate (OUR) was considered as a disturbance and was not included in the mathematical model. The feedback gain adaptation was based on the difference between the actual molar fraction of oxygen at time step n and the initial molar fraction. This difference is related to OUR and was used to increase or decrease the state feedback controller gain (k and k(1), respectively) in a slow manner. With these controllers it was possible to get an excellent online estimate of OUR. In the case of bubble aeration a simple gas phase mass balance was sufficient, whereas during the headspace aeration a liquid phase balance was required. It has been shown that determination of OUR using gas balance requires a significantly better controller performance compared to just keeping DO and pH within reasonable limits. (c) 1995 John Wiley & Sons, Inc.     

Overcoming challenges in WAVE Bioreactors without feedback controls for pH and dissolved oxygen

The biopharmaceutical industry is increasing its use of the WAVE Bioreactor for culturing cells. Although this disposable bioreactor can be equipped to provide real-time pH and dissolved oxygen (DO) monitoring and control, our goal was to develop a process for culturing CHO cells in this system without relying on pH and DO feedback controls. After identifying challenges in culturing cells without controlling for pH and DO in the WAVE Bioreactor, we characterized O(2) and CO(2) transfer in the system. From these cell-free studies, we identified rock rate and rock angle as key parameters affecting O(2) transfer. We also identified the concentration of CO(2) in the incoming gas and the rate of gas flow into the headspace as key parameters affecting CO(2) transfer--and therefore pH--in the disposable culture chamber. Using a full-factorial design to evaluate the rock rate, rock angle, and gas flow rate defined for this WAVE Bioreactor process, we found comparable cell growth and pH profiles in the ranges tested for these three parameters in two CHO cell lines. This process supported cell growth, and maintained pH and DO within our desired range--pH 6.8-7.2 and DO exceeding 20% of air saturation--for six CHO cell lines, and it also demonstrated comparable cell growth and viability with the stirred-tank bioreactor process with online pH and DO control. By eliminating the use of pH and DO probes, this process provides a simple and more cost-effective method for culturing cells in the WAVE Bioreactor.     

Twenty-four well plate miniature bioreactor system as a scale-down model for cell culture process development

Increasing the throughput and efficiency of cell culture process development has become increasingly important to rapidly screen and optimize cell culture media and process parameters. This study describes the application of a miniaturized bioreactor system as a scaled-down model for cell culture process development using a CHO cell line expressing a recombinant protein. The microbioreactor system (M24) provides non-invasive online monitoring and control capability for process parameters such as pH, dissolved oxygen (DO), and temperature at the individual well level. A systematic evaluation of the M24 for cell culture process applications was successfully completed. Several challenges were initially identified. These included uneven gas distribution in the wells due to system design and lot to lot variability, foaming issues caused by sparging required for active DO control, and pH control limitation under conditions of minimal dissolved CO2. A high degree of variability was found which was addressed by changes in the system design. The foaming issue was resolved by addition of anti-foam, reduction of sparge rate, and elimination of DO control. The pH control limitation was overcome by a single manual liquid base addition. Intra-well reproducibility, as indicated by measurements of process parameters, cell growth, metabolite profiles, protein titer, protein quality, and scale-equivalency between the M24 and 2 L bioreactor cultures were very good. This evaluation has shown feasibility of utilizing the M24 as a scale-down tool for cell culture application development under industrially relevant process conditions.     

Efficient and reproducible mammalian cell bioprocesses without probes and controllers?

Bioprocesses for recombinant protein production with mammalian cells are typically controlled for several physicochemical parameters including the pH and dissolved oxygen concentration (DO) of the culture medium. Here we studied whether these controls are necessary for efficient and reproducible bioprocesses in an orbitally shaken bioreactor (OSR). Mixing, gas transfer, and volumetric power consumption (P(V)) were determined in both a 5-L OSR and a 3-L stirred-tank bioreactor (STR). The two cultivation systems had a similar mixing intensity, but the STR had a lower volumetric mass transfer coefficient of oxygen (k(L)a) and a higher P(V) than the OSR. Recombinant CHO cell lines expressing either tumor necrosis factor receptor as an Fc fusion protein (TNFR:Fc) or an anti-RhesusD monoclonal antibody were cultivated in the two systems. The 5-L OSR was operated in an incubator shaker with 5% CO(2) in the gas environment but without pH and DO control whereas the STR was operated with or without pH and DO control. Higher cell densities and recombinant protein titers were obtained in the OSR as compared to both the controlled and the non-controlled STRs. To test the reproducibility of a bioprocess in a non-controlled OSR, the two CHO cell lines were each cultivated in parallel in six 5-L OSRs. Similar cell densities, cell viabilities, and recombinant protein titers along with similar pH and DO profiles were achieved in each group of replicates. Our study demonstrated that bioprocesses can be performed in OSRs without pH or DO control in a highly reproducible manner, at least at the scale of operation studied here.     

Adaptive control of dissolved oxygen concentration in a bioreactor

A new adaptive DO (dissolved oxygen) concentration control algorithm considering DO electrode dynamics with response time delay has been developed. A system model with two time-varying parameters was used to relate the DO concentration with two control variables: air flow rate and agitation speed. Parameters of this model were estimated on-line using a regularized constant trace recursive least-squares method. An extended Kalman filter was used to remove the effect of noises from the DO concentration measurements and thus to improve control performance. A discrete one-step ahead control scheme was adopted to determine control actions based on the parameter estimation results. Experimental results showed that the new adaptive DO concentration control algorithm performed better than other algorithms tested, a PID controller and adaptive algorithms without the DO electrode dynamics.     

High cell density induces spontaneous bifurcations of dissolved oxygen controllers during CHO cell fermentations

High cell density cultures of CHO cells growing in a bioreactor under dissolved oxygen control were found to undergo spontaneous bifurcations and a subsequent loss of stability some time into the fermentation. This loss of stability was manifested by sustained and amplified oscillations in the bioreactor dissolved oxygen concentration and in the oxygen gas flow rate to the reactor. To identify potential biological and operational causes for the phenomenon, linear stability analysis was applied in a neighborhood of the experimentally observed bifurcation point. The analysis revealed that two steady state process gains, K(P1) and K(P2), regulated k(l)a and gas phase oxygen concentration inputs, respectively, and the magnitude of K(P1) was found to determine system stability about the bifurcation point. The magnitude of K(P1), and hence the corresponding open-loop steady state gain K(OL1), scaled linearly with the bioreactor cell density, increasing with increasing cell density. These results allowed the generation of a fermentation stability diagram, which partitioned K(C)-N operating space into stable and unstable regions separated by the loci of predicted critically stable controller constants, K(C,critical), as a function of bioreactor cell density. This consistency of this operating diagram with experimentally observed changes in system stability was demonstrated. We conclude that time-dependent increases in cell density are the cause of the observed instabilities and that cell density is the critical bifurcation parameter. The results of this study should be readily applicable to the design of a more robust controller.     

Computer control of pH and DO in a laboratory fermenter using a neural network technique

In this contribution, the advantages of the artificial neural network approach to the identification and control of a laboratory-scale biochemical reactor are demonstrated. It is very important to be able to maintain the levels of two process variables, pH and dissolved oxygen (DO) concentration, over the course of fermentation in biosystems control. A PC-supported, fully automated, multi-task control system has been designed and built by the authors. Forward and inverse neural process models are used to identify and control both the pH and the DO concentration in a fermenter containing a Saccharomyces cerevisiae based-culture. The models are trained off-line, using a modified back-propagation algorithm based on conjugate gradients. The inverse neural controller is augmented by a new adaptive term that results in a system with robust performance. Experimental results have confirmed that the regulatory and tracking performances of the control system proposed are good.     

High‐throughput miniaturized bioreactors for cell culture process development: Reproducibility,scalability, and control

Decreasing the timeframe for cell culture process development has been a key goal toward accelerating biopharmaceutical development. Advanced Microscale Bioreactors (ambr?) is an automated micro‐bioreactor system with miniature single‐use bioreactors with a 10–15 mL working volume controlled by an automated workstation. This system was compared to conventional bioreactor systems in terms of its performance for the production of a monoclonal antibody in a recombinant Chinese Hamster Ovary cell line. The miniaturized bioreactor system was found to produce cell culture profiles that matched across scales to 3 L, 15 L, and 200 L stirred tank bioreactors. The processes used in this article involve complex feed formulations, perturbations, and strict process control within the design space, which are in‐line with processes used for commercial scale manufacturing of biopharmaceuticals. Changes to important process parameters in ambr? resulted in predictable cell growth, viability and titer changes, which were in good agreement to data from the conventional larger scale bioreactors. ambr? was found to successfully reproduce variations in temperature, dissolved oxygen (DO), and pH conditions similar to the larger bioreactor systems. Additionally, the miniature bioreactors were found to react well to perturbations in pH and DO through adjustments to the Proportional and Integral control loop. The data presented here demonstrates the utility of the ambr? system as a high throughput system for cell culture process development. © 2014 American Institute of Chemical Engineers Biotechnol. Prog., 30:718–727, 2014     

Bioreactor studies on the effect of dissolved oxygen concentrations on growth and differentiation of carrot (Daucus carota L.) cell cultures

A bioreactor control system was used to investigate the effects of two dissolved oxygen concentrations (10% and 100%) on the growth and differentiation of Daucus carota L. cell cultures. The strategy used allowed the dissolved oxygen concentration to be controlled without the need for changing either the agitator speed or the total gas flow rate. During the proliferation phase, reducing oxygen resulted in a lower growth rate and in a delay in sugar uptake kinetics. Nonetheless, varying levels of oxygen were observed to have no effect on the final dry biomass. The higher alcohol dehydrogenase activity obtained under reduced oxygen conditions suggests that proliferating cultures adapted to the hypoxic environment by inducing alcoholic fermentation. Cell differentiation was highly sensitive to reduced oxygen since under this condition, the somatic embryo production was inhibited by about 75%. Sugar uptake and embryo formation were also delayed.Abbreviations ADH alcohol dehydrogenase - 2,4-D 2,4-dichlorophenoxyacetic acid - DO2 dissolved oxygen - SE somatic embryos - Tris tris(hydroxymethyl)-aminoethane     

Correlating on-line monitoring parameters, pH, DO and ORP with nutrient removal in an intermittent cyclic process bioreactor system

The paper presents the study correlating the profile of on-line monitoring parameters and nutrient removal in an intermittent cyclic process bioreactor (ICPBR) system, thereby utilizing the parameters as operational tool. A laboratory scale ICPBR was employed to treat low C/N ratio domestic wastewater from a township. The study was conducted for correlating biological nutrient removal and on-line monitored parameters pH, dissolved oxygen (DO) and oxidation-reduction potential (ORP). The results revealed that pH, DO and ORP related with the dynamic behavior of nutrient concentration (NH4-N, NO3-N, and PO4-P) during treatment in an ICPBR system. The variation in pH and ORP of the reactor liquor correlate to conversion of ammonia (NH4-N) and nitrate (NO3-N) concentrations, respectively. As the bioconversion of ammonia nitrogen and phosphorus are related to the varying profile of the on-line monitored parameters, the profiles could possibly be used as onsite process control parameters.     

Novel micro‐bioreactor high throughput technology for cell culture process development: Reproducibility and scalability assessment of fed‐batch CHO cultures

With increasing timeline pressures to get therapeutic and vaccine candidates into the clinic, resource intensive approaches such as the use of shake flasks and bench‐top bioreactors may limit the design space for experimentation to yield highly productive processes. The need to conduct large numbers of experiments has resulted in the use of miniaturized high‐throughput (HT) technology for process development. One such high‐throughput system is the SimCell? platform, a robotically driven, cell culture bioreactor system developed by BioProcessors Corp. This study describes the use of the SimCell? micro‐bioreactor technology for fed‐batch cultivation of a GS‐CHO transfectant expressing a model IgG4 monoclonal antibody. Cultivations were conducted in gas‐permeable chambers based on a micro‐fluidic design, with six micro‐bioreactors (MBs) per micro‐bioreactor array (MBA). Online, non‐invasive measurement of total cell density, pH and dissolved oxygen (DO) was performed. One hundred fourteen parallel MBs (19 MBAs) were employed to examine process reproducibility and scalability at shake flask, 3‐ and 100‐L bioreactor scales. The results of the study demonstrate that the SimCell? platform operated under fed‐batch conditions could support viable cell concentrations up to least 12 × 10⁶ cells/mL. In addition, both intra‐MB (MB to MB) as well as intra‐MBA (MBA to MBA) culture performance was found to be highly reproducible. The intra‐MB and ‐MBA variability was calculated for each measurement as the coefficient of variation defined as CV (%) = (standard deviation/mean) × 100. The % CV values for most intra‐MB and intra‐MBA measurements were generally under 10% and the intra‐MBA values were slightly lower than those for intra‐MB. Cell growth, process parameters, metabolic and protein titer profiles were also compared to those from shake flask, bench‐top, and pilot scale bioreactor cultivations and found to be within ±20% of the historical averages. Biotechnol. Bioeng. 2010; 106: 57–67. © 2010 Wiley Periodicals, Inc.     

A novel in situ probe for oxygen uptake rate measurement in mammalian cell cultures

The newly developed in situ oxygen uptake rate (in situ OUR) probe presented in this article is based on the in situ microscope technology platform. It is designed to measure the oxygen uptake rate (OUR) of mammalian cells, an important parameter for metabolic flux analysis, inside a reactor (in situ) and in real-time. The system isolates a known volume of cell culture from the bulk inside the bioreactor, monitors the oxygen consumption over time, and releases the sample again. The sample is mixed during the measurement with a new agitation system to keep the cells in suspension and prevent oxygen concentration gradients. The OUR measurement system also doubles as a standard dissolved oxygen (DO) probe for process monitoring when it is not performing OUR measurements. It can be equipped with two different types of optical sensors (i.e., DO, pH) simultaneously or a conventional polarographic DO-probe (Clark type). This new probe was successfully tested in baby hamster kidney perfusion cell cultures.     

Control of carbon source supply and dissolved oxygen by use of carbon dioxide concentration of exhaust gas in fed-batch culture

Accurate and automatic control strategies for a feedback-control system of volatile carbon source feeding and dissolved oxygen (DO) level were investigated. To maintain the optimal ethanol concentration for microbial growth, carbon dioxide concentration in exhaust gas was used as a stepwise control parameter of ethanol feeding. A proportional-differential (PD) control program was used to correct the errors. The coefficient of stepwise control was calculated stoichiometrically, and parameters of PD were experimentally preset and were not changed during cultivation. DO was also controlled by the PD control and the stepwise program based on carbon dioxide concentration of the exhaust gas. Agitation speed and partial pressure of oxygen of the inlet gas were changed stepwise in accordance with the oxygen consumption rate. The stepwise coefficients were estimated from stoichiometry and material balance of molecular oxygen. The PD control program was only used for the agitation speed control to correct the fluctuations of DO level. The parameters did not need to be changed during cultivation. By use of these sophisticated control programs for fed-batch culture of Candida brassicae, ethanol concentration and DO level were accurately controlled at 3.4–3.7 g/l and 2.2–2.8 ppm, respectively, while cell mass concentration reached about 80 g/l. No manual operation was needed.     

Long-term continuous monitoring of dissolved oxygen in cell culture medium for perfused bioreactors using optical oxygen sensors

For long-term growth of mammalian cells in perfused bioreactors, it is essential to monitor the concentration of dissolved oxygen (DO) present in the culture medium to ascertain the health of the cells. An optical oxygen sensor based on dynamic fluorescent quenching was developed for long-term continuous measurement of DO for NASA-designed rotating perfused bioreactors. Tris(4,7-diphenyl-1,10-phenanthroline) ruthenium(II) chloride is employed as the fluorescent dye indicator. A pulsed, blue LED was chosen as the excitation light source. The sensor can be sterilized using an autoclave. The sensors were tested in a perfused rotating bioreactor supporting a BHK-21 (baby hamster kidney) cell culture over one 28-day, one 43-day, and one 180-day cell runs. The sensors were initially calibrated in sterile phosphate-buffered saline (PBS) against a blood-gas analyzer (BGA), and then used continuously during the entire cell culture without recalibration. In the 180-day cell run, two oxygen sensors were employed; one interfaced at the outlet of the bioreactor and the other at the inlet of the bioreactor. The DO concentrations determined by both sensors were compared with those sampled and measured regularly with the BGA reference. The sensor outputs were found to correlate well with the BGA data throughout the experiment using a single calibration, where the DO of the culture medium varied between 25 and 60 mm Hg at the bioreactor outlet and 80-116 mm Hg at the bioreactor inlet. During all 180 days of culture, the precision and the bias were +/-5.1 mm Hg and -3.8 mm Hg at the bioreactor outlet, and +/- 19 mm Hg and -18 mm Hg at inlet. The sensor dynamic range is between 0 and 200 mm Hg and the response time is less than 1 minute. The resolution of the sensor is 0.1 mm Hg at 50 mm Hg, and 0.25 mm Hg at 130 mm Hg.     

Influence of aeration–homogenization system in stirred tank bioreactors,dissolved oxygen concentration and pH control mode on BHK-21 cell growth and metabolism

This work focused on determining the effect of dissolved oxygen concentration (DO) on growth and metabolism of BHK-21 cell line (host cell for recombinant proteins manufacturing and viral vaccines) cultured in two stirred tank bioreactors with different aeration-homogenization systems, as well as pH control mode. BHK-21 cell line adapted to single-cell suspension was cultured in Celligen without aeration cage (rotating gas-sparger) and Bioflo 110, at 10, 30 and 50 % air saturation (impeller for gas dispersion from sparger-ring). The pH was controlled at 7.2 as far as it was possible with gas mixtures. In other runs, at 30 and 50 % (DO) in Bioflo 110, the cells grew at pH controlled with CO₂ and NaHCO₃ solution. Glucose, lactate, glutamine, and ammonium were quantified by enzymatic methods. Cell concentration, size and specific oxygen consumption were also determined. When NaHCO₃ solution was not used, the optimal DOs were 10 and 50 % air saturation for Celligen and Bioflo 110, respectively. In this condition maximum cell concentrations were higher than 4 × 10⁶ cell/mL. An increase in maximum cell concentration of 36 % was observed in batch carried out at 30 % air saturation in a classical stirred tank bioreactor (Bioflo 110) with base solution addition. The optimal parameters defined in this work allow for bioprocess developing of viral vaccines, transient protein expression and viral vector for gene therapy based on BHK-21 cell line in two stirred tank bioreactors with different agitation–aeration systems.     

Control of dissolved oxygen concentration in mammalian cell culture using a computer-coupled mass flow controller

Summary A simple proportional control system for dissolved oxygen (DO) concentration in cell culture medium was developed by using a computer-coupled mass flow controller. The DO levels were very stable during the cultivation of Vero-6, while flow rates of air and/or oxygen enriched air were gradually changed depending on the DO concentration and the preset DO level. Vero-6 cells could grow normally to the confluence in the range of 30% and 50% of DO. Growth of Vero-6 at 10% of DO was markedly retarded.     

Identification and control of dissolved oxygen in hybridoma cell culture in a shear sensitive environment

The productivity of mammalian cells can be enhanced by facilitating adequate oxygen transfer into the cultivation medium. However, current methods of controlling dissolved oxygen (DO) fail to account for alterations in medium composition during the course of the fermentation. These changes, which directly affect gas solubility and overall mass transfer coefficient, may be significant and deteriorate controller's performance in the long run. In this paper, the applications of Generalized Predictive Controllers (GPC) to DO control were investigated in a shear sensitive environment and compared to PID and Model Predictive Controllers (MPC). Input and output data for system identification were initially generated by varying the composition of oxygen fed into the bioreactor from 0 to 0.21 mol % while keeping the total inlet gas flow rate at 8.75 vvm. The process was identified using an AutoRegressive model with eXogeneous inputs (ARX) model and tested on different data sets. The model parameters were then correlated with the overall mass transfer coefficients. In simulation tests, the output of the PID controller switched from minimum to maximum values while more continuous control signals were obtained with the MPC and GPC controllers. When tested in a cell-free medium, all three controllers were able to track setpoint changes with some chattering observed in the control signals. The GPC outperformed the MPC and PID controllers when applied to the cultivation of hybridoma cells.     

Implementing Process Analytical Technology for the Production of Recombinant Proteins in Escherichia coli Using an Advanced Controller Scheme

The performance of a bioreactor in meeting process goals is affected by the microorganism used, medium composition, and operating conditions. A typical bioreactor uses a supervisory control and data acquisition (SCADA) system for control, and a combination of software and hardware tools for real‐time data analysis. However, when the process is disrupted by utility or instrumentation failure, typical process controllers may be unable to reinstate normal operating conditions before the cells in the reactor shift to unfavorable metabolic regimes. The objective of this study is to examine how the response of a controller affects process recovery when disruptive incidences occur under a process analytical technology (PAT) framework. The process used for this investigation is the production of lethal toxin‐neutralizing factor (LTNF) by Escherichia coli (E. coli), which is controlled by a decoupled input–output‐linearizing controller (DIOLC). The performance of the DIOLC is compared to a proportional integral derivative (PID) controller subjected to the same conditions. The disruptions are introduced manually and the effect of controller action on process recovery and LTNF synthesis is measured in terms of peak purity and concentration. It is observed that DIOLC performs better after reinstating operating conditions and results in a meaningful improvement in performance.     

Production of L(+)-lactic acid from glucose and starch by immobilized cells of Rhizopus oryzae in a rotating fibrous bed bioreactor

A rotating fibrous-bed bioreactor (RFB) was developed for fermentation to produce L(+)-lactic acid from glucose and cornstarch by Rhizopus oryzae. Fungal mycelia were immobilized on cotton cloth in the RFB for a prolonged period to study the fermentation kinetics and process stability. The pH and dissolved oxygen concentration (DO) were found to have significant effects on lactic acid productivity and yield, with pH 6 and 90% DO being the optimal conditions. A high lactic acid yield of 90% (w/w) and productivity of 2.5 g/L.h (467 g/h.m(2)) was obtained from glucose in fed-batch fermentation. When cornstarch was used as the substrate, the lactic acid yield was close to 100% (w/w) and the productivity was 1.65 g/L.h (300 g/h.m(2)). The highest concentration of lactic acid achieved in these fed-batch fermentations was 127 g/L. The immobilized-cells fermentation in the RFB gave a virtually cell-free fermentation broth and provided many advantages over conventional fermentation processes, especially those with freely suspended fungal cells. Without immobilization with the cotton cloth, mycelia grew everywhere in the fermentor and caused serious problems in reactor control and operation and consequently the fermentation was poor in lactic acid production. Oxygen transfer in the RFB was also studied and the volumetric oxygen transfer coefficients under various aeration and agitation conditions were determined and then used to estimate the oxygen transfer rate and uptake rate during the fermentation. The results showed that the oxygen uptake rate increased with increasing DO, indicating that oxygen transfer was limited by the diffusion inside the mycelial layer.     

The study of neural network-based controller for controlling dissolved oxygen concentration in a sequencing batch reactor

The design and development of the neural network (NN)-based controller performance for the activated sludge process in sequencing batch reactor (SBR) is presented in this paper. Here we give a comparative study of various neural network (NN)-based controllers such as the direct inverse control, internal model control (IMC) and hybrid NN control strategies to maintain the dissolved oxygen (DO) level of an activated sludge system by manipulating the air flow rate. The NN inverse model-based controller with the model-based scheme represents the controller, which relies solely upon the simple NN inverse model. In the IMC, both the forward and inverse models are used directly as elements within the feedback loop. The hybrid NN control consists of a basic NN controller in parallel with a proportional integral (PI) controller. Various simulation tests involving multiple set-point changes, disturbances rejection and noise effects were performed to review the performances of these various controllers. From the results it can be seen that hybrid controller gives the best results in tracking set-point changes under disturbances and noise effects.