The results part is divided in three sections. The first section presents the data of the permittivity measurements (ΔεFOGALE) and compares the obtained off- and on-line data. The second part shows the calculated fC and compares it with the cell size measurements and in a third and final part data from duplicate runs are compared to demonstrate the reproducibility of the employed system.
On-line permittivity data, oxygen uptake rate and off-line biovolume measurements
Figure A displays data for a representative baculovirus-infected Sf-9 culture whereby virus addition took place during the exponential growth phase at 2.2 × 106 viable cells mL−1 after about 50 h cultivation time (time of infection, TOI). The cell counts given by the Vi-CELL™ system and the hemacytometer are depicted in the graph. Whereas the total cell count is similar for both measurements, the viable cell count (Vi-CELL) shows lower counts than the hemacytometer. This was the case in particular after virus infection. It is in agreement with our observation that the Vi-CELL™ system underestimates cell viability of non-infected and especially baculovirus-infected Sf-9 cells. Possible reasons for this observation might be the higher concentration of Trypan Blue (0.4% (w/v)) compared to our standard hemacytometer method (0.05% (w/v)) and/or the different and potentially harsher mixing procedure by the automated system.
Fig. 2 (A) Representative pattern of a baculovirus-infected Sf-9 culture (fermentation 1): Hemacytometer counts (total cell count (●), viable cell count (), viability (hema) (), Vi-CELL cell counts (total cell count (■), viable (more ...)
After infection the permittivity is increasing even faster than during the exponential growth phase whereas the viable and total cell counts remain nearly constant until about 100 h cultivation time. Quite noticeable is a plateau region for the permittivity from about 12 to 20 h post-infection (Fig. A and B marked by black arrow) followed by a subsequent further increase in signal.
After 100 h cultivation time the maximum of the permittivity is reached and viability is strongly decreasing (Fig. A and B). The total cell count remains constant until the end of the fermentation at about 250 h. During the whole time course nearly no cell debris was generated and even stained cells kept optically intact membranes (microscopic observation).
The cell size data (CASY®1, Vi-CELL™) and the viability (hema) for this run are shown in Fig. B. After infection, the mean cell diameter is increasing by 20% or 3–4 μm whereas the permittivity increases by about 80%. This is in agreement with the equation of Schwan as the diameter contributes to the permittivity to the power of three. Furthermore, the plateau in the permittivity is accompanied by a temporary leveling off of cell size during this period (Fig. A). The two different methods of measuring the cell diameter give quite similar results although in the later stages of the culture the Vi-CELL™ mean diameter is not decreasing in the same way as the CASY®1 mean diameter. Generally, it could be observed by microscopic observation that the cell size change in the later stages of the fermentation does not correlate with the extensive reduction measured by the CASY®1 system.
Fig. 3 (A) Permittivity (solid line), mean cell diameter (Vi-CELL) (×), mean cell diameter (CASY) (●) and oxygen uptake rate (·, dash dotted line) for the plateau region of fermentation 1. (B) Permittivity (solid line), total biovolume (more ...)
To explain the differences in the cell size determinations observed for the two methods a closer look into the underlying techniques is needed. On the one hand the CASY®
1 system is based on the principle of electrical resistance measurement. This method takes into account the integrity of the cell membrane in the sense that a disrupted membrane or even an increase in membrane permeability leads to a higher conductivity of the cytoplasm and will result in a smaller measured cell size for the respective cell (Winkelmeier et al. 1993
). On the other hand the Vi-CELL™ system employs optical analysis for the measurement of cell diameter which should not be influenced by changes in membrane integrity. This might explain why the patterns for the two mean cell diameters (Fig. B) are very similar in the first 100 h of the fermentation and start to deviate after the maximum in the permittivity. In particular the membrane permeability of infected Sf-9 cells changes in the later stages of the culture when viability is decreasing and produced virus gets released into the suspending medium (Zeiser et al. 2000
As mentioned in the theoretical part, the Fogale BIOMASS SYSTEM® measures the membrane enclosed volume fraction or biovolume in a cell suspension. Hence, permittivity increases with increasing cell diameter although the total cell count remains constant.
The CASY®1-Counter calculates a value which gives the volume of all particles (cells) measured in a sample which is in this article referred to as the total biovolume (CASY). This value correlates well with the permittivity (Fig. C). The Vi-CELL™ system allows the measurement of the cell diameter of viable cells by separating stained from unstained (viable) cells. The measured total biovolume (Vi-CELL), i.e. the number of all cells counted by the system multiplied with their respective diameter, was calculated by assuming that all cells are spherical. The viable biovolume (Vi-CELL) was calculated accordingly by computing the number of all viable cells and the viable cell diameters. Due to the observation that the Vi-CELL™ system seems to underestimate the viable cell count of infected Sf-9 cells in our system we used the hemacytometer cell counts for the respective calculation of the biovolume corresponding to the different populations of viable (viable biovolume (Vi-CELL) and total cells (total biovolume (Vi-CELL)). Packed cell volume (PCV) measurements were used as an additional estimation for the total biovolume (PCV).
Figure C presents the good correlation of all these values and the permittivity until a certain point in time after infection. The viable biovolume (Vi-CELL) and the total biovolume (CASY) are both in good agreement with the on-line permittivity measurement over the whole time course of the cultivation.
Whereas the viable biovolume (Vi-CELL) decreases faster than the permittivity after the drop in viability, the total biovolume (CASY) shows a smaller decrease compared to the permittivity. The values for the total biovolume (Vi-CELL) as well as the total biovolume (PCV) show a very different pattern. After the maximum value both biovolume parameters are decreasing much less than the other three values. Interestingly, the total biovolume (CASY) and the viable biovolume (Vi-CELL) correlate in the best way with the permittivity over the whole time course of the cultivation. The permittivity is according to the equation of Schwan influenced by the changes in membrane state being a function of the CM
. Although CM
was originally reported to be a biological constant, recent studies reported changes with decreasing viability (Fehrenbach et al. 1992
) and changes in the physiological state (Noll and Biselli 1998
; Ansorge et al. 2007
). This may explain the correlation with the total biovolume (CASY) as the two measurement techniques are both dependent on cell membrane properties. The good correlation with the viable biovolume (Vi-CELL) was expected since permittivity measurements take per definition only viable cells into account. However, stained cells still contribute to the signal which has been reported in different publications for various cell lines and yeast (Davey et al. 1993
; Guan et al. 1998
; Ducommun et al. 2002
; Cannizzaro et al. 2003
; Ansorge et al. 2007
). In addition, we observed in this system only a marginal cell disintegration that might have been the result of using a bench top surface-aerated bioreactor system. Thus any hydrodynamic stress caused by bursting bubbles was eliminated. In conclusion, the difference in between the viable biovolume (Vi-CELL) and the permittivity after 100 h cultivation time was most likely caused by the presence of stained cells with membranes that are still contributing to the signal.
An interesting point in the time course is the plateau region in the permittivity that is observed at ca. 12–20 h post-infection. It was apparent in both performed fermentations although it was not as obvious in the second one (Fig. A). Figures A and B represent an enlarged view of the fermentation course around the time of infection highlighting the plateau region of fermentation 1. The plateau can also be observed with all the other off-line measurements. Figure B represents the correlation of the total biovolume (PCV) and the total biovolume (CASY) with the permittivity. The cell size values also level off at that time and the oxygen uptake rate reaches a maximum of ca. 6 mmol 10E−9
(Fig. A). These values are in agreement with the data of other authors (Schopf et al. 1990
; Schmid 1996
; Schmid et al. 1994
). The plateau region for the permittivity signal has likewise been described by others (Zeiser et al. 1999
). Zeiser et al. further demonstrated that the plateau in the permittivity correlates to the maximum in the CO2
evolution rate (Zeiser et al. 2000
). Furthermore, these authors and others observed that this time point corresponds to the first release and the production of budded virus (Ooi and Miller 1988
; Wong et al. 1994
). The appearance of the plateau before 20 h post-infection was interpreted as a good indicator for synchronous infection which can be assumed here as we used a comparatively high MOI of ~10. With respect to the oxygen uptake rate our findings are also in agreement with the literature data stating that the maximal oxygen consumption after baculoviral infection is observed at 12–20 h post-infection (Kamen et al. 1996
; Schmid 1996
Fig. 5 (A) Permittivity (fermentation 1 (□, dash dotted line), fermentation 2 (·)) and characteristic frequency (fermentation 1: dotted line, fermentation 2: dashed line) over time for two duplicate fermentations. (B) Correlation of permittivity (more ...)
Monitoring the infection process by scanning permittivity measurements
As mentioned in the theoretical part the multi-frequency measurement allowed the calculation of the fC
. Figure displays the fC
data for the first fermentation run. Until the time of infection the fC
increases from ~0.7 to 0.9 MHz. This is in line with the simplified equation for the fC
since the cell diameter is decreasing. The values for fC
are not shown for the first few hours of cultivation because their calculation is subject to large errors due to the presence of insufficient biomass at the beginning of the culture. After viral infection fC
starts to decrease with the cell diameter increasing substantially. This holds true until the time of the plateau region in the permittivity. Although the cell size is still increasing after the plateau region, we observe a minimum in fC
followed by a subsequent increase from that time until the end of the culture. From then on, the trend in fC
is not in agreement with the cell diameter data anymore. From viral infection until plateau region however, the change in fC
is in agreement with the change in cell size following the respective equation, i.e. the increase in cell size is reflected by the decreasing fC
. Assuming that this is the time of first virus release as reported by Zeiser et al. (1999
) major changes of the cell membrane properties should arise that are also affecting fC
. This might explain the changes of fC
after the plateau phase that are not correlating with the cell diameter. In fact, what can be expected are either changes in CM
or/and the intracellular conductivity (σc
Permittivity (·), characteristic frequency (dashed line) and cell size (CASY) (●) over time (fermentation 1)