September 1, 2016 (Vol. 36, No. 15)
Jolanda Meister Zurich University of Applied Sciences, School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology
Rüdiger W. Maschke Zurich University of Applied Sciences, School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology
Soren Werner University of Applied Sciences, Institute of Biotechnology, Zurich, Switzerland
Dr. G.T. John Director PreSens
Dieter Eibl, Ph.D. Ph.D. Zurich University of Applied Sciences, School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology
Culture Broth Viscosity Evaluated in Shake Flasks Studies
In biotechnology, the use of shake flasks is widespread due to their easy handling, especially in process design and optimization as well as in small-scale upstream processing and inoculum production.
Recently, geometrically optimized Thomson Optimum Growth™ shake flasks were introduced to the market, promoting higher yields with the same shaker footprint. Knowledge on different engineering parameters (including the oxygen mass transfer coefficient, the mixing time, or the power consumption) is essential to compare these new flasks with established shake flasks and to evaluate potential benefits.
In the present work, the oxygen mass transfer in Thomson Optimum Growth disposable shake flasks is quantified at 500 mL and 5 L using the PreSens Shake Flask Reader.
Erlenmeyer and Fernbach shake flasks are still the predominant shapes, despite their limitations (e.g. the relatively low filling volumes). Consequently, Thomson Instruments developed a new shake flask design:
Optimum Growth for cell cultures, allowing filling volumes of up to 60% of the nominal volume.
The Shake Flask Reader is built to measure the dissolved oxygen concentration and the pH value in shake flasks by means of single-use sensor spots that are attached to the inner surface of the shake flask. Besides a reduced contamination risk, another decisive advantage of the sensor spots (compared to conventional oxygen probes) is their small size, which does not change the flow behavior. The device can accommodate and measure up to nine shakes flasks (from 125–2,000 mL) simultaneously. The data is transmitted to a PC via a Bluetooth connection.
To attach the 5 L Thomson shake flask to the shake flask reader, a custom-built clamp and metal plate were used to reduce the large distance between the sensor in the flask and the reader (Figure 1). For the 500 mL flask, a standard clamp was used.
To determine the kLa value, the dynamic gassing-out method was used (according to the recommendations of the DECHEMA single-use technology expert group). In this procedure, nitrogen was passed directly into the bulk solution via tubes until an oxygen concentration below 20% was attained. Subsequently, the headspace was purged with air.
Afterwards, the lid was removed to focus on the oxygen transition from the gas to the liquid phase. Measurement was stopped at an oxygen concentration of 80%. The maximum and minimum dissolved oxygen values were scaled. Subsequently, the y-axis values were calculated and plotted in the range of 20–80% as a function of time. A linear trend line was inserted and the kLa value calculated. The experiments were performed with varying shaking speeds on two popular commercial shakers with the two most frequently used shaking diameters of 25 mm and 50 mm and different filling volumes. All experiments were conducted in triplicate.
Field of Application
The Optimum Growth flasks are intended for cultivating cell cultures of plant and animal origin up to 3 L filling volumes. During cultivation with plant cells, the viscosity increases due to high cell concentration and aggregation. Carboxymethylcellulose sodium salt (CMC) was used to simulate several viscosities (aqueous solutions of 4, 8, and 20 g L-1).
These concentrations correlate well with packed cell volumes (PCV) at different stages of plant cell cultivation. The viscosity of plant cell suspensions may reach values of 0.4 Pa s (equivalent to 70% PCV of a Nicotiana tabacum Bright-Yellow 2 cell suspension or 20 g L-1 CMC solution) and show shear thinning behavior.
As shown in Figure 2, the kLa increases with higher shaking rates. After a critical point (at higher shaking rates, e.g., ~200 rpm for a shaking amplitude of 25 mm), the effect reverses. Regarding process design, the determination of a suitable parameter combination is key to optimal cultivation results, especially if the cell growth is limited due to a high oxygen demand. Hence, the process parameters (e.g., aeration) have to be adapted to the specific needs.
Shaking at low rates (e.g., 80–160 rpm) and high filling volumes only ensures kLa values below 5h-1. In addition, increasing viscosity decreases the kLa value (Figure 2, A–D). Higher kLa values (between 50 and 60h-1) were achieved (Figure 2) with low filling volumes (1.2 –1.6 L) and medium to high shaking rates (140–240 rpm). To estimate the kLa value at different parameters, Equation 1 could be used in combination with the values in Table 1.
kLa = 10 (C+n·Cn+V·CV+CMC·CCMC+n2·Cnn+n·V·CnV)
The Thomson Optimum Growth shake flasks are geometrically modified and easy-to-operate cultivation systems. It was shown that very low shaking rates, high filling volumes, and high viscosities lead to kLa values lower than 5h-1. Average shaking rates and filling volumes are sufficient to reach high cell densities in Newtonian culture broths as this is the case for mammalian cell expansions.
When growing cells in non-Newtonian, high viscous culture broths, the resulting kLa values below 5h-1 may inhibit cell growth. Nevertheless, kLa values in the range of 50–60h-1 are possible, using high shaking rates (> 160 rpm) and low filling volumes (1.2 L).
However, at higher shaking rates the shear stress may become a further limiting factor. Using Equation 1, the kLa value can be approximated and assists researchers to avoid oxygen transfer limitations as well as to optimize the cultivation process.
Jolanda Meister, Rüdiger W. Maschke, Sören Werner, and Prof. Dieter Eibl are with the Zurich University of Applied Sciences, School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, and Dr. Gernot T. John (firstname.lastname@example.org) is with PreSens Precision Sensing. Website: www.presens.de.