Biotechnology Letters

, Volume 31, Issue 5, pp 665–669

Online monitoring of oxygen in spinner flasks


  • Rahul Ravi Deshpande
    • Biochemical Engineering InstituteSaarland University
    • Department of Biochemical and Chemical EngineeringUniversity of Dortmund
    • Biochemical Engineering InstituteSaarland University
Original Research Paper

DOI: 10.1007/s10529-009-9919-2

Cite this article as:
Deshpande, R.R. & Heinzle, E. Biotechnol Lett (2009) 31: 665. doi:10.1007/s10529-009-9919-2


We show the application of a novel optical on-line sensor fixed in spinner flasks for the online monitoring of dissolved O2 concentrations during mammalian cell growth. Using this sensor that requires only minute changes to the flask to be made, we could determine the volumetric O2 transfer coefficient as well as O2 consumption rates. Under normal growth conditions the cells did not undergo O2 limitation. Also, the transfer of O2 from the atmosphere to the spinner flasks is influenced by the use of screw caps. The on-line measurement was further applied to determine the O2 uptake rates which can then be used to monitor the metabolic state of the cells and also for online process monitoring.


Mammalian cell cultureOnline O2 monitoringOptical oxygen sensorOxygen transferSpinner flask


O2 plays an important role in the metabolism of mammalian cells where it is essential for the oxidative phosphorylation for the production of ATP. The metabolism of mammalian cells is highly influenced by the dissolved oxygen concentrations (Zupke et al. 1995). Dissolved oxygen concentration influences ATP production and specific substrate uptake rates (Eyer and Heinzle 1996). The importance of O2 uptake rates has been stressed by many researchers and its measurement allowed controlled feeding of glutamine to increase productivity (Eyer et al. 1995; Oeggerli et al. 1995). The most commonly used small scale culture devices for mammalian cells are microtiter plates. The transfer characteristics of oxygen in 96-well plates have been discussed previously (Deshpande and Heinzle 2004). In a laboratory or even pilot scale mammalian cells in suspension are generally grown in spinner flasks. This is the method of choice for suspension lines including hybridomas, CHO cells and attached lines that have been adapted to growth in suspension. Also attached cell lines can be scaled up with microcarrier beads and grown in spinner flasks. Spinner flasks are either plastic or glass bottles with a central magnetic stirrer shaft and side arms for the addition and removal of cells and medium, and gassing with CO2-enriched air. Inoculated spinner flasks are placed on a stirrer and incubated under the culture conditions appropriate for the cell line. Commercial spinner flasks are available in the range of 125–5,000 ml.

O2 transfer characteristics in spinner flasks, however, has not been greatly studied. Insect cells undergo O2 limitation under standard spinner flask cultivation conditions (Annathur et al. 2003). However, the O2 consumption of mammalian cells is much lower than insect cells and it has been assumed that suspension cells do not undergo O2 limitation in spinner flasks under normal operating conditions. For metabolic studies, it is essential that the cells are not stressed by O2 limitation. Recent developments resulted in fluorescence sensors that allow robust measurement of dissolved oxygen using internal referenced and measurement of fluorescence decay time (Klimant and Wolfbeis 1995; Kocincová et al. 2008). In this paper we studied O2 transfer in a spinner flask equipped with an optical O2 sensor for on-line monitoring using a dynamic method. We tested for potential oxygen transfer limitation by sterile caps. We monitored O2 consumption in spinner flasks using liquid phase balance.

Materials and methods

Cell line and culture conditions

Chinese Hamster Ovary (CHO) cell line T-CHO ATIII, obtained from Gesellschaft fuer Biotechnologische Forschung mbH (Braunschweig, Germany), was used. The cells produce recombinant Antithrombin III, which has clinical applications for its coagulation inhibitory activity. The cells are grown in suspension and are adapted to serum free medium. CHO-S-SFM II (GIBCO, Invitrogen Corporation, USA), a serum-free medium for CHO suspension cells, was used for the routine growth and maintenance of cells. The cells were grown in 250 ml spinner flasks (Techne, Staffordshire, United Kingdom) placed in an incubator (INCO2, Memmert GmbH + Co.KG, Schwabach, Germany) set at 37°C and 88% relative humidity with an overlay of 12% CO2. The spinner flasks were placed on a stirrer (Model 104-S, Techne, Staffordshire, United Kingdom) which has a maximum speed of 80 rpm. The cells were also adapted to grow in a protein and peptide free, chemically defined media (SMIF 6, Life Technologies, Karlsruhe, Germany).

Cell viability assays

Cells were counted using a Neubauer hemocytometer. Cell concentration and viability were determined using the Trypan Blue exclusion method as described earlier (Deshpande and Heinzle 2004).

Spinner flasks with O2 sensing

The optical sensing system employed is similar to a system described earlier for shake flask cultures (Wittmann et al. 2003). Briefly the system consists of three parts: (i) an optical sensor spot immobilized on the spinner flask bottom, (ii) a coaster placed below the flask containing an optical fiber, and (iii) a module for data processing and connecting to a PC. The spinner flask with sensor attached to the bottom was supplied by Pre-Sens GmbH (Regensburg, Germany). The sensor containing a fluorescent, O2-sensitive fluorescent dye (Klimant and Wolfbeis 1995) was placed in the centre of a 250 ml spinner flask to allow uninterrupted rotation of the shaft. The system uses a phase modulation technique to determine the luminescent decay time of the O2 sensor. The decay time is related to the actual dissolved O2 (DO) by the Stern–Volmer equation (Wittmann et al. 2003; Kocincová et al. 2008). The system uses a two-point calibration with water; zero dissolved O2 was achieved by purging water with N2 and 100% dissolved O2 by purging with air. The experiments were carried out in normal culture conditions of CHO cells. The volumetric gas–liquid mass transfer coefficient (kLa) was determined using a dynamic method as describe earlier (Dunn et al. 2003; Wittmann et al. 2003; Deshpande and Heinzle 2004). The flask was first gassed with N2 and after switching to air the change of dissolved O2 concentration was monitored on-line. O2 uptake rate measurement was made using a liquid phase balance together with the determined kLa value and the measured dissolved oxygen concentration (Dunn et al. 2003; Deshpande and Heinzle 2004).

Results and discussion

Characterization of O2 transfer in spinner flasks

The volumetric gas–liquid mass transfer coefficient (kLa) was calculated at two different spinner speeds at various volumes using a dynamic method. The speeds used were 40 and 80 rpm which is the maximum rate allowed with the used equipment. Figure 1 gives the plot depicting the relationship between various volumes at two different speeds with the gas–liquid mass transfer coefficient. At higher volumes the transfer at both the speeds almost reaches the same values of 0.02 min−1 thus giving a maximum O2 transfer rate of 0.28 mM h−1. Considering an average specific O2 uptake rate of 2 × 10−13 mol (O2) cell−1 h−1 and average cell density of 106 cells ml−1, this maximum transfer rate would be just enough to avoid O2 limitation. However, in rapid phase of cell growth, the specific O2 uptake rates of cells are much higher and hence the use of large volumes is not recommended for cell culture in spinner flasks.
Fig. 1

The relationship between the volumetric gas–liquid mass transfer coefficient (kLa) and volume at two different speeds. The kLa was calculated by dynamic method using gas–liquid mass transfer equation after removal of O2 from water by N2, followed by flushing the air space with air

Another interesting facet about cellular growth in spinner flasks is the use of the side arm. The main purpose of these arms is for the removal of cells or addition of media but also serves as gas-phase O2 supply. These side arms are usually provided with screw caps, and it is recommended to grow the cells without closing them tightly to avoid O2 limitation. It was decided to investigate the influence of the number of turns the screw cap of the side arms is open on the O2 transfer. The number of turns the caps takes to fully close was counted. O2 transfer characteristic was determined for the flask with no caps, flask with half opened caps and flask whose caps were turned only once after closing the cap completely. The flask was filled with 60 ml water and O2 is removed by purging with N2, placed into the incubator on the stirrer with 80 rpm with the desired rotation of cap and the rise of O2 monitored.

Figure 2 gives the rise of O2 under the three conditions. It is seen that the rise is the fastest in the case where there are no caps, and slowest in flask with a single turn. From these data it is clear that O2 supply to the culture was limited by O2 diffusion from outside into the reactor head space. However, the N2 was not replaced by purging air for the above study. The rate of O2 transfer would be higher if the N2 was replaced with air.
Fig. 2

The influence of the turns of the screw caps on the side arm of the spinner flasks on the transfer of O2. Flask with no cap (fully open), flask with half-opened cap (half open) and flask whose cap was turned only once after closing the cap completely (single turn) were investigated

Online monitoring of cell growth

Cells were grown in the spinner flasks with 60 ml medium and at a stirrer speed of 80 rpm. Cells were inoculated at a density of 2 × 105 cells ml−1 and monitored for 190 h (Fig. 3). In the first 50 h, dissolved O2 decreased due to increasing cell density to reach a steady-state concentration of about 30% air saturation that lasted until about 150 h. Then there was a rise in the dissolved O2 concentration probably indicating that the cells started to die during this phase. From the earlier determination of kLa and the measured dissolved O2 concentration we could directly calculate O2 uptake rate.
Fig. 3

Online monitoring of dissolved O2 along with O2 uptake rates with spinner flasks equipped with optical O2 sensor. The cells were grown in 60 ml medium at 80 rpm (kLa = 0.076 min−1). The O2 uptake rates are calculated using liquid phase oxygen balance

A second experiment was carried out where parallel cultures were grown, one in the normal media and the other in a modified medium where the concentrations of glucose and glutamine had been halved. Additionally samples were taken during the cell growth to determine the cell density. Figure 4 gives the O2 consumption of both the cell cultures along with the growth curve. It is seen that the O2 consumptions are very similar as are the growth curves. Moreover, none of the culture goes into O2 limitation. However, the process of sampling disturbs the measurements indicated by a rise in the dissolved O2 concentration after the sampling. It did, however, not lead to any O2 limitation during sampling that required removal of the flask from the incubator. Nonetheless the O2 concentrations reach equilibrium after some time. Hence it was seen that under normal cell culture conditions, the cells do not undergo O2 limitation when grown in spinner flasks.
Fig. 4

a Online monitoring of dissolved O2 with spinner flasks equipped with optical O2 sensor for parallel cell cultures. b The growth curve of the two cultures. The cell density was determined with Trypan Blue method. The cells were grown in a 60 ml at 80 rpm. The modified medium indicates medium with half the concentration of glucose and glutamine as in the normal medium. The sampling is indicated by arrows


In this paper, we show the application of novel spinner flasks for the online monitoring of dissolved oxygen concentrations during mammalian cell growth. Under normal growth conditions, the cells do not undergo O2 limitations. However, at high volumes there is a possibility of O2 limitation. Also, the transfer of oxygen from the atmosphere to the spinner flasks is influenced by the use of screw caps. This methodology can further be expanded to determine the O2 uptake rates, which can then be used to monitor the metabolic state of the cells and also for online process monitoring.

Copyright information

© Springer Science+Business Media B.V. 2009