Top

New Device for Biomass Monitoring in Shake Flask Culture

Development of an analytical unit based on the SFR Shake Flask Reader

J. Schmidt-Hager1, C. Ude1, M. Findeis2, G. T. John2, T. Scheper1, and S. Beutel1
1Institute of Technical Chemistry, Leibniz University of Hanover, Germany
2PreSens Precision Sensing GmbH, Regensburg, Germany

We developed a novel system for online biomass measurement in shake flasks based on near 360° scattered light detection. The optosensoric components were combined with the established PreSens SFR Shake Flask Reader technology, so biomass, pO2 and pH can be monitored non-invasively through the flask bottom. First application of the prototype SFR vario on Escherichia coli K12 and Kluyveromyces marxianus culture resulted in highly reproducible online biomass data, with relative errors matching those of standard offline OD measurements.

In biopharmaceutical production relevant cultivation parameters should be monitored throughout the whole culture period. In early stages of bioprocess development shake flasks are often applied, because results can be achieved with as less effort as possible running many experiments in parallel. In this small scale, however, it is difficult to monitor or control important cultivation parameters. Cell growth determination in shake flasks, for example, requires offline sampling. This is a major drawback as it increases the risk of contamination and the work load. Furthermore, these measurements can only be performed at distinct points in time, which makes them unsuitable for deeper process understanding. Now a new analytical unit - the SFR vario - has been developed for online monitoring of cell growth in shake flask culture. This method is based on detection of light scattered by particles inside the liquid. Correlation to optical density (OD) and cell dry weight (CDW) can be established by non-linear calibration models. The optoelectronic components used for cell growth determination were implemented in the established SFR Shake Flask Reader system to create a monitoring device which allows simultaneous pO2, pH and biomass monitoring in shake flask cultures. Initial evaluation tests with E. coli K12 as prokaryotic and K. marxianus as eukaryotic models were performed providing highly reproducible results.

Materials & Methods

A new additional sensor module for determination of scattered light with a scatter angle of about 360° was incorporated in the optics for pO2 and pH measurement of the SFR Shake Flask Reader (PreSens) (Fig. 1). The ray of light emitted by an LED is transmitted through the flask bottom and scattered by particles (cells) inside the culture medium; this scattered light is detected by a photodiode (Fig. 1B). During shaking a liquid sickle forms inside the shake flask (Fig. 1B), so a piezoelectric acceleration sensor was integrated to optimize the measuring cycle. It adjusts the time between measurements. The performance of this new analytical unit was demonstrated with two model microorganisms - E. coli K12 and K. marxianus. E. coli was cultivated in minimal medium containing glucose and lactose at 37 °C to follow diauxic growth. K. marxianus was cultivated in YM-medium with glucose monohydrate at 30 °C. Both microorganisms were cultured in 500 mL baffled shake flasks with integrated pO2 and pH sensors (SFS, PreSens) at 100 mL working volume and 150 rpm. All cultivations were carried out four times under same conditions. Three of the cultivations were used to generate a calibration model of scattered light intensity as a function of OD600. Therefore, offline samples were taken every 60 min under sterile conditions and OD600 was determined. The remaining cultivation was used for validation. Measurements of biomass, pH and pO2 were taken with a sampling interval of 15 s.

Online Biomass Monitoring

Before the analytical unit could be applied for biomass monitoring, a valid correlation between scattered light and biomass concentration explicitly for the monitored cell type in the respective cultivation conditions had to be deduced. Therefore, offline OD600 measurements of the samples were correlated with the respective online measured values. For E. coli K12 and K. marxianus this correlation can be best described by a simplified Bleasdale-Nelder function:

y = (a + b • x)(-1/c)                                  Eq. 1

where y is the OD600 value, x stands for the measured light intensity, and a - c are the respective median filtered parameter values determined for the specific cell type. In Figures 2 and 3 biomass measurements with the prototype SFR vario in E. coli K12 and K. marxianus culture are depicted. Additionally oxygen and pH measurements, and offline measured substrate concentrations are shown. With a sampling rate of 15 seconds a large amount of biomass measurement data was obtained giving a very good real-time overview of biomass development. At the beginning of the cultivations some measurement noise was observed in cultures of both microorganisms. This is caused by low starting cell densities and boundary layer reflection of the detection light beam. With cells growing and the culture broth becoming sufficiently turbid the biomass measurements stabled and started increasing. A median filter was applied to smooth the graphs. The interruptions in oxygen and pH measurements were caused by stopping the shaking movement and taking samples for offline measurements. In Figure 2 the diauxic growth of E. coli can be clearly determined from the measurement values provided by the prototype device. The period after glucose consumption is characterized by a metabolic shift form glucose to lactose, which can be observed in a small plateau in the biomass measurements. Moreover the exact match of stopped growth with rapidly increasing oxygen levels as well as a plateau in pH measurements could be recorded. The different growth phases of K. marxianus (Fig. 3) can also be seen in the online biomass data. In the first phase glucose is metabolized under aerobic conditions. When glucose in the medium becomes limiting, the cell metabolism switches to metabolize the products of the previous glucose consumption under high oxygen demand. The growth rate in the second phase is a lot lower than in the previous phase clearly visible in the biomass measurements; due to oxygen limitation the growth is linear in this phase. So with the prototype SFR vario we were able to follow biomass development inside shake flasks of eukaryotic and prokaryotic model microorganisms and gained measurement results, that have never been recorded like this before in shake flasks. With the data provided by the SFR vario all relevant information about culture conditions can be monitored without the need for offline analysis.

Conclusion

The functionality and potential of the new sensor system was successfully demonstrated by the online monitoring of different cell types. Biomass measurements with the SFR vario prototype can be applied to both eukaryotic and prokaryotic cell suspensions. The application range of the device for online measurement of pO2, pH and biomass is versatile. Individual calibration models set up for different cell types and culture conditions are delivering precise predictions of the biomass concentration. Developed calibration models for E. coli K12 and K. marxianus showed that high accuracy can be obtained with a relative error of less than 12 % compared to alternative offline techniques. By using this non-invasive sensor system, it is possible to get information about cell growth in shake flasks in real time. Further evaluations of the developed sensor system will be conducted to adapt it for more microorganisms and cell lines and widen the application spectrum.

Knowledge

Select by industry

Presens TV

Watch tutorials, webinars and informative videos about PreSens optical sensor systems.

All videos