Bioprocess and Biosystems Engineering

, Volume 37, Issue 9, pp 1799–1808 | Cite as

Viability characterization of Taxus chinensis plant cell suspension cultures by rapid colorimetric- and image analysis-based techniques

  • Thomas Wucherpfennig
  • Annika Schulz
  • Jaime Arturo Pimentel
  • Gabriel Corkidi
  • Dominik Sieblitz
  • Matthias Pump
  • Gilbert Gorr
  • Kai Schütte
  • Christoph Wittmann
  • Rainer Krull
Original Paper


For the commercially established process of paclitaxel production with Taxus chinensis plant cell culture, the size of plant cell aggregates and phenotypic changes in coloration during cultivation have long been acknowledged as intangible parameters. So far, the variability of aggregates and coloration of cells are challenging parameters for any viability assay. The aim of this study was to investigate simple and non-toxic methods for viability determination of Taxus cultures in order to provide a practicable, rapid, robust and reproducible way to sample large amounts of material. A further goal was to examine whether Taxus aggregate cell coloration is related to general cell viability and might be exploited by microscopy and image analysis to gain easy access to general cell viability. The Alamar Blue assay was found to be exceptionally eligible for viability estimation. Moreover, aggregate coloration, as a morphologic attribute, was quantified by image analysis and found to be a good and traceable indicator of T. chinensis viability.


Taxus chinensis plant cell culture Plant cell aggregates Viability Aggregate coloration Alamar Blue assay Image analysis 


Commercial systems for cultivation of plant cells were established more than 25 years ago. Significant advancements have been made in understanding metabolite production in large-scale plant cell cultures, but so far controlling variability in product accumulation has often been neglected in favor of improving yield [1]. For the industrially established process of paclitaxel production by submerse cultivation of Taxus chinensis plant cells, a reliable and applicable way to measure the important process parameter of the plant aggregate size with <100 μm up to over 2 mm was established and evaluated by laser diffraction technique [2]; an alternative way to measure aggregate size had previously been established by Kolewe et al. [3]. The morphology of these aggregates was of some importance, because aggregate size has been proposed to correlate with productivity [4, 5], whereas aggregate coloration was reported to depend on cell age and viability [6]. However, no clear connection between aggregate coloration and viability has been shown so far.

For processes with T. chinensis plant cells, no viability estimating techniques fit for industrial application have so far been established, besides measurements of cell fresh weights and cell dry weights. A reduction of cell growth, however, does not imply a reduction of viability [7] and zero growth is not necessarily an indicator of poor viability. Since plant cells in suspension mainly stay connected after division and grow in aggregates, the determination of the number of cells is virtually impossible. Viability assays are based either on the physical properties or biochemical characteristics of the cells. The detection methods are based on the properties of the cells, such as endocytosis, enzymatic activity, the integrity of the cell membrane, or proliferation. There are generally different ways to measure the viability of cells. It is possible to obtain information on cell viability by observation of the cytoplasmic flow; membrane integrity can be determined by electrolyte flow, plasmolysis and dye exclusion or retention [8]. Furthermore, the biochemical activity can be measured as in protein synthesis, tetrazolium chloride reduction, DNA and RNA synthesis and fluorescein-di-acetate staining [8]. Microscopic measurements, based on the color of cells incubated with Evans Blue or Trypan Blue, are possible. Dead cells have a disrupted membrane, and can therefore take up the dye [9]. Currently, such tests are popular for mammalian cell culture, as they allow easy automation.

A major breakthrough in the field of biochemistry-based viability tests was achieved by the development of assays based on the activity of the electron transport chain [10]. Such assays based on water-soluble, colorless tetrazolium salts such as 3-(4.5-dimethylthiazol-2-yl)-2.5-diphenyl tetrazolium bromide (MTT) and 2.3.5-triphenyltetrazolium chloride (TTC) are popular for mammalian and plant cell cultures. When entering a living cell, TTC is reduced by the dehydrogenase activity of the mitochondrial electron transport chain to form insoluble red formazan product. Therefore, dead and living cells can be distinguished by color. For quantification, the red precipitate is extracted from the tissue with ethanol, and the absorbance of the extract can be determined spectrophotometrically [11]. The TTC has already been shown to be suited for the viability determination of Taxus cultures [12, 13]. Like the TTC test, the MTT assay is based on the dehydrogenase activity of the mitochondrial electron transport chain. Here, the yellow-colored tetrazole of MTT is converted to water-insoluble purple formazan. To dissolve the water-insoluble product, organic solvents such as DMSO or isopropanol are required. The intensity of the blue color can be directly correlated to the number of living cells [14]. All tetrazolium salts are generally toxic because the formazan crystals that are produced from reduction of the salts must be solubilized with DMSO or HCl/isopropanol which lead to cell death [15, 16]. However, in addition to the toxic character of the applied chemical viability assays like the MTT-based one require multiple steps. Therefore, a more simple and non-toxic method would be favorable for standard applications in terms of monitoring industrial processes as well as for process development.

Recently, the Alamar Blue (AB) dye has gained popularity as a very simple and versatile way of measuring cell proliferation and cytotoxicity [17]. AB is a non-toxic redox indicator, which can be detected by both fluorescence and by color change. It contains resazurin and resorufin as oxidation–reduction indicators. The oxidized blue, non-fluorescent form of resazurin is converted by reduction into a pink fluorescent dye resorufin [16]. The reduction of AB occurs intracellularly involving reductases and the mitochondrial electron transport chain [9]. AB acts as an electron acceptor between cytochrome a,a3 and the final reduction step of O2, and does not interfere with the function of the respiratory chain [9]. Other tetrazolium salt-based agents like MTT and TTC have a more negative redox potential than the components of the electron transport chain, and thus consume energy in form of ATP during the reduction. AB on the other hand possesses a more positive redox potential than FMNH2, FADH2, NADH, NADPH and the cytochromes. Thus, a reduction is possible without additional energy requirement [9]. The ‘resazurin reduction test’ has been used for about 50 years to monitor bacterial and yeast contamination of milk, and also for assessing semen quality [17]. The AB assay has so far been used for the determination of the viability of bacteria, mycobacteria, fungi and several human cell lines [18, 19, 20, 21, 22]. It was also used to assess the viability of tomato plant cells [9].

There has been an ongoing speculation, whether specific aggregate coloration might be related to cell viability, because such a link might be easily exploited for easy and rapid viability estimation. Specific aggregate colorations have previously been used for characterization of calli, as they are often an indicator on the specific potential of the callus [23]. Callus color may change under varying medium composition [24]. Thus, the color of suspended cells varies, which can be used for evaluation purposes [23]. Especially in suspension cultures for pigment production the color of cells is directly related to their quality [25]. Taxus suspension cultures have furthermore been reported to change appearance in color after exposure to light [6]. Paclitaxel production in calli depends on morphology and age [26]. The callus color was also previously found to correlate with variability in cell growth. Light brown calli showed a higher growth rate than dark and aggregated calli [27]. However, calli were found to produce more paclitaxel at an old age with brown coloration, than at a younger age without the brown stain [28, 29, 30]. The intensification of darkening at the stationary growth phase may be related to increased biosynthesis of phenolic compounds [31].

The aim of the present study was to elucidate the relationship between T. chinensis aggregate coloration and cell viability. To this end the AB assay had to be established as a non-toxic, rapid, robust and reproducible viability test for cultures of T. chinensis, because no viability assays for industrial application had been previously established. Consequently viability data gained from aggregate coloration through microscopy and automatic image analysis were compared and verified by the AB assay.

Materials and methods

Conditions for cultivation of Taxus chinensis

Gamborg B-5 Basal Medium [32] was used for cultivation of T. chinensis plant cells described by Wucherpfennig et al. [2] in 1L polycarbonate Erlenmeyer flasks with vent caps (Corning, USA), filled with 400 mL medium. Cultures were incubated for 240 h at 120 min−1 on a rotary shaker (Certomat BS-1/50 mm, Sartorius, Göttingen, Germany). Cultivations were carried out in triplicate. The growth temperature was 25 ± 0.1 °C.

Alamar Blue assay for T. chinensis viability determination

Since the AB assay has so far not been used for T. chinensis plant cells, it was necessary to adjust the assay for analysis of Taxus cell aggregates. The incubation time, the age of the culture used for generating the calibration curve, lighting conditions and type of incubation were identified to significantly influence measurement results and reproducibility of the test. These factors were gradually optimized. Byth et al. [9] found that common plant cell culture media and buffers reduce the AB reagent (CellTiter-Blue Cell Viability Assay Kit, Promega, Madison, USA). Therefore, it was necessary to filter cells (Miracloth filter, pore size 25 μm, Calbiochem, Germany) and rinse them in 10 mL deionized water; 50 mg of Taxus cells was suspended in 200 μL, pH 7.2 phosphate buffer. AB reagent was used in a final concentration of 10 % (v/v) within the suspension. After incubation (see "Results and discussion") the cells were disrupted for 60 s in an ultrasonic bath and centrifuged at 12,000 min−1 for 10 min (Biofuge pico, Heraeus, Thermo Fisher Scientific, Waltham, USA) at room temperature. Before fluorescence measurement done at room temperature as well, the samples were shaken in the fluorescence reader (Fluoroskan Ascent, Thermo Scientific) for 10 s at 180 min−1 and measured at 544 nm excitation and 590 nm emission wavelength, respectively. For evaluation of optimal incubation of the assay time resorufin absorbance was measured at 540 and 620 nm (SPECORD® 200, Analytik Jena AG, Jena, Germany). Since the oxidized and reduced spectra are overlapping the equation introduced in Byth et al. [9] was used for calculation of the reduction value (%).

To compare different methods of controlled eradication of plant cells, 50 mL of cell suspension was taken from a shaken culture. The cells were filtered through a water jet pump with a suction filter and the filter cake was subsequently rinsed with 10 mL distilled water; 250 mg wet biomass was each weighed and given into 1.5 mL tubes. For the eradication experiments by UV light, freezing and heating to 40 °C respectively, 900 μL of phosphate buffer was introduced into the reaction vessels. For eradication by ethanol, 450 μL phosphate buffer and 450 μL of 70 % ethanol were used. To exterminate cells using a saline solution, 250 mg wet biomass was mixed with 900 μL, 5 % NaCl (w/w) solution. All samples were incubated for 1 h at room temperature.

To determine the validity of the AB assay for T. chinensis suspension cultures, a comparison with the MTT assay, already established for plant cells [33], was conducted. For the preparation of the MTT reagent, 5 mg thiazolyl blue tetrazolium bromide powder (Sigma-Aldrich, St. Louis, USA) was dissolved in 1 mL deionized water. For performing the MTT assay, the cell culture suspension was filtered, rinsed and re-suspended in buffer as described for the AB assay. Each 100 μL of this suspension was transferred to a 96-well plate using a wide-nozzle pipette. After the addition of 10 μL of MTT reagent to each well the plates were incubated in a photometer (Sunrise microplate reader, Tecan, Maennedorf, Switzerland) at 37 °C for 2 h in the dark. After incubation, 100 μL of a DMSO/acidic isopropanol mixture 1:1 (v/v) was added to the wells. The samples were further incubated at 37 °C and measured every 30 min at 450 nm.

Microscopy and image analysis of T. chinensis samples

Aggregate brightness and coloration was monitored offline using an all-inclusive, all-digital, inverse microscope (EVOS xl, AMG, Bothell, WA, USA). Approximately 20 pictures with a magnification of 4× were taken, obtaining at the very least 400 analyzable single plant cell aggregates for statistical robustness. Aggregate coloration was evaluated by means of an automated image analysis procedure implemented within the Image-Pro 7.0 (Media Cybernetics, USA) macrobuilder. In the beginning of the image analysis process, a background flattening procedure (for bright background, 40 pixels as typical feature size) was applied in order to obtain a homogeneous background. A Gaussian filter (kernel size 7, strength 10) was applied to the whole set of images in order to eliminate high frequency components which may produce small artifacts and rough aggregate contours. To make the procedure usable for pictures of different microscopes with varying illumination, gray-level thresholds for the healthy and dead aggregates are not selected automatically. For a given set of micrographs (5 typically averaged) with the same illumination the operator has to manually select the areas he assesses to be healthy and those he deems unhealthy. The program uses these thresholds termed (‘LiveThres’ and ‘DeadThres’) for a whole set of images to obtain the respective total areas in order to evaluate the ratio of viable to dead aggregate area and therewith the viability. Figure 1 shows a typical, processed image. This image analysis routine basically combines the power and quick assessment of the operator with the superior data evaluation of a computer. The manual step in the procedure assures, furthermore that the program will be adaptable to other cells and conditions.
Fig. 1

Evaluation of dark and gray area for estimation of viability. The healthy area is depicted in red, dead and unhealthy areas are dark with a white border

Mechanical stress experiments

Mechanical stress experiments were carried out in a stirred tank glass reactor (3 L round-bottom bioreactor, Applikon, The Netherlands) with a reactor diameter of 130 mm and a working liquid volume of 800 mL (liquid height = 60 mm). The reactor included three baffles (width of baffles = 13 mm, gap between baffle and reactor wall = 9 mm) and one InterMIG stirrer (see Fig. 2, stirrer clearance from the reactor bottom = 30 mm). The stirrer was operated with 11.6 revolutions per second which corresponds to a stirrer tip velocity of 2.7 m s−1. To determine the effect of stirrer-induced stress on T. chinensis cells through application of the AB assay and image analysis, 500 mL of plant cell suspension was transferred after 5 days of shake flask cultivation (fresh cell weight of 75 g L−1) into the bioreactor. In order not to interfere with the performance of the stirrer, the cultivation broth was filled up with sodium chloride solution having the same osmolality as that of the culture medium of around 150 mosmol kg−1 to rule out any osmotic stress, to a volume of 800 mL and a diluted fresh cell weight of about 40 g L−1. After mechanical stress exposure, 7.5 mL of the plant cell suspension was transferred into a 15 mL Falcon tube by the use of a rinsed (5 mL of distilled water) wide-nozzle pipette. Then the sample was filtered (Miracloth filter, pore size 25 μm, Calbiochem, Germany) until a constant fresh weight was achieved. For AB and MTT assay, the cells were rinsed with deionized water and re-suspended in phosphate buffer. For microscopy and image analysis, the cells were rinsed with 10 mL distilled water in a petri dish (plastic, 9 cm, Omnilab, Germany), and re-suspended in a 150 mosmol kg−1 NaCl solution.
Fig. 2

InterMIG stirrer used for mechanical stress experiment: impeller diameter (outside) = 75 mm and axis diameter = 8 mm, impeller height = 12 mm, impeller blade width = 11.5 mm

Results and discussion

Development of a dye-based method for assessment of Taxus chinensis plant cell viability

The AB assay is a one-step, extremely simple, reproducible, economical, and non-toxic procedure to evaluate cell proliferation and survival [34]. The AB assay offers various advantages in comparison to other more established assays, including technical simplicity, freedom from radioisotopes, versatility in detection, no extraction, and excellent reproducibility and sensitivity [15]. Further advantages of the AB assay include the water solubility of the formazan product, economic efficiency and minimal toxicity [9]. To establish the AB assay for T. chinensis cultures, a calibration graph with a known composition of viable and dead cells and corresponding fluorescences had to be established. To this end, Taxus cells were eradicated in a controlled manner by incubation for 1 min in liquid nitrogen. Subsequently cells were gently defrosted in warm (37 °C) water and then blended with living cells to create the desired ratios of living cells. The proportions of living cells were normalized and arranged between 0 (all dead) and 1 (highest fluorescence measured, presumably all living).

The incubation time was varied from 1 up to 3 h (Fig. 3). Only an incubation time of 1.5 and 1.75 h produced reasonable reduction values and the characteristic sigmoidal trend [9]. An incubation time of 1 h was obviously not long enough for the cellular metabolism to reduce the resazurin to resorufin, whereas some degradation processes must have taken place after 3 h of cultivation. Thus, an incubation time of 1.5 h was used.
Fig. 3

Evaluation of optimal incubation time of the AB assay for determination of T. chinensis cell viability. For quantification, the absorbance was measured spectrophotometrically. Values are means for three replicates. The reduction value (%) was calculated according to Byth et al. [9] using the following formula: \(\text{Reduction} = \frac{{\varepsilon_{{\text{oxid}\; 620 \text{nm} }} \left( {\text{sample} \; A_{{540\,\;\text{nm} }} } \right) - \varepsilon_{{\text{oxid} \; 540 \text{nm} }} \left( {\text{sample} \; A_{{620\;\text{nm} }} } \right)}}{{\varepsilon_{{\text{red} \; 540 \text{nm} }} \left( {\text{ox} . \text{control} \; A_{{620\;\text{nm} }} } \right) - \varepsilon_{{\text{red} \; 620 \text{nm} }} \left( {\text{ox} . \text{control} \; A_{{540\;\text{nm} }} } \right)}}\)

To establish standardized and reproducible calibration graphs, an optimal viability of the cells used for the calibration had to be ensured. Therefore, T. chinensis cells of several culture ages were used to identify an optimal viability of the culture. Figure 4a shows the maximal measured relative fluorescence in dependence of culture age. With approximately 700 relative units the highest relative fluorescence was measured after 5 days of cultivation, showing that T. chinensis cell had the highest viability after 5 days of batch cultivation. The resulting calibration curve of the sample cultivated for 5 days is depicted in Fig. 4b. The fluorescence values of the various percentages of living cells rise steadily from 120 over the entire range to around 700 fluorescence units. Fluorescence ranges observed for this assay procedure can vary between cell cultures derived from different species, i.e., for tomato cell culture, showing a much faster growth performance than Taxus cell cultures. Significantly higher fluorescence units of up to 5,500 have been achieved [9]; these cells however had a much faster doubling time. Also exposure to light during incubation is essential and might have some impact on the fluorescence read out, too. To ensure highest possible reproducibility, samples in this study were incubated under artificial light sources and under absence of direct sunlight. The presence of aggregates is presumed to cause reduced uptake and release rates, which in the current case might lead to reduced fluorescence values. It is known that larger aggregates usually exhibit microenvironments with differing nutrient and oxygen concentrations [35]. However, these microenvironments are difficult to measure and most available data is based on theoretical predictions. Some data exists of oxygen diffusion models in dependence on aggregate size [35, 36], with some evidence of lower oxygen consumption by larger aggregates [37]. Kessler et al. [38] established a model predicting the critical diameter, defined as the maximum diameter for unlimited respiration throughout the aggregate, to be around 3.8 mm well above common plant cell aggregate sizes. In this study the mean aggregate diameter of T. chinensis aggregates was generally not larger than around 500 μm depending on the method of measurement as discussed by Wucherpfennig et al. [2]. In this range from 0.1 to 0.5 mm aggregates of Tripterygium wilfordii were also recently observed to be not diffusion limited, which was consistent with the observed growth kinetics [39]. A diffusion limitation and therefore a significant influence of the aggregate size on the performed assay within the observed size range seem therefore highly unlikely.
Fig. 4

a Maximum spectrofluorometrically measured relative fluorescence is depicted over age of culture in days, b regression curve at a culture age of 5 days and an incubation time of 1.5 h. Values are means for three replicates (R2 = 0.99)

To further improve the reproducibility of the assay, the samples were mixed during the entire incubation period in an overhead rotator (Revolver Tube Mixer, Labnet International, Inc. Edison, USA), to make sure that aggregates were suspended at all times. The rotating incubation was a reason for the generally high reproducibility of the assay, with deviations of less than 1 % between three measurements.

To compare results of the AB assay with a more established assay, the MTT assay was conducted. Like the AB assay, this viability test is based on the reducing properties of the mitochondrial electron transport chain. The water-insoluble purple product formazan was extracted with DMSO, and determined by spectrophotometry. Results for the AB and MTT assay for a 5 days pre-cultivated sample are depicted in Fig. 5. In comparison with the MTT method, the AB assay provides more reproducible results with fewer measurement errors. The MTT assay also yields relatively low absorption values between 0.01 and 0.2. However, these values can be expected, as the MTT assay usually produces absorption values between 0 and 1.5, depending on the cell line being examined [10, 14]. Furthermore, the AB assay is performed in a single step and is non-toxic, compared to the rather time consuming toxic multi step MTT assay.
Fig. 5

Regression curve for the AB (black) and MTT assay (green). Values are means for four replicates

Viability of T. chinensis plant cell suspension cultures using image analysis technique

A large grayish brown area is distinctive for unhealthy or dead cells, since aggregates with low viability are often observed to have a large dark area, whereas healthy aggregates show only a small or no dark area. Any browning or reddening of the cells or the medium indicates a problem with the culture [6, 40]. Using image analytic techniques it is possible to determine the ratio of the healthy to the dark and presumably dead cell area in microscopic pictures, therefore it should be possible to determine the percentage of dead cells and thus the viability. This can be done as long as the procedure is calibrated with known ratios of live/dead cells. For this purpose, T. chinensis cells were eradicated in the same controlled manner as for the AB assay calibration. Figure 6 shows the correlation between the dark cell area measured by image analysis and the percentage of prepared living cells. A clear linear relationship is apparent (R2 = 0.99), substantiating the assumption that dark cell coloration is related to viability. The significant reproducible behavior between cell coloration and viability makes it possible to estimate cell viability from coloration appearance alone. Although the link between aggregate cell coloration and viability has not been investigated previously, image analytic methods have been used for handling and analysis of plant cell culture before for example for identification of cells which begin to differentiate in very large aggregates [13, 41]. Pepin and colleagues showed that hue, saturation and intensity characteristics of plant cells permitted the establishment of a direct relationship between of the colored product anthocyanin and aggregate size in the V. pahalae production system [36]. However, most of these approaches display a lack of automation and a rather small number of samples. In the present study a sufficiently large number of aggregates are analyzed in a completely automated procedure, therefore industrial applicability is ensured.
Fig. 6

Relationship (R2 = 0.99) between dark cell coloration as determined by automatic image analysis and T. chinensis cell viability measured with the AB assay. Values are means for three replicates of 5-day-old shaking flask samples

Comparison of dye- and image analysis-based viability methods

The general suitability and responsiveness of a viability assay is influenced by a number of factors including culture medium constituents, pH, cell density, incubation temperature, test compound and assay reagent stability for time-course monitoring, dosage and exposure time to the test compound, and linearity of the assay [16]. To compare responsiveness and results of the image based viability estimation to the AB assay, Taxus samples were exposed to various viability impairing physical and chemical conditions including UV-light, ethanol, sodium chloride, heat and freezing. Subsequently, both viability estimating methods were conducted, to verify whether both methods produce similar results (Fig. 7).
Fig. 7

Comparison of the determination of viable T. chinensis cells by AB assay (a) and automatic image analysis (b) after perturbation of cells to various physiological properties, RT, standard; UV, exposure to UV-light; EtOH, addition of ethanol (70 %); NaCl, addition of sodium chloride (5 %); 40 °C and freezing, influence of temperature

The hypothesis that the viability of the plant cell culture correlated with the browning of the cells could be confirmed by the comparison of the image analysis and the AB assay. In general, both methods show similar results and are comparable. In comparison to the biochemical assay, the image analysis technique underestimates viability somewhat (Fig. 7b), variation of absolute values is also dependent on standardization. The presence of UV-light or NaCl did not lead to a significant reduction in viability, as the percentage of living cells measured with the AB assay remained at around 90 % and approximately at 80 % measured by image analysis. The addition of 70 % ethanol in a ratio of 1:1 (v/v) resulted in a significant reduction in cell viability to around 60 %. An increase in temperature to 40 °C led to a decrease in viability to around 80 % measured by the AB assay, and about 60 % as measured by image analysis. Freezing led to an eradication of all cells, following the AB assay and produced a rest viability of 10 %, measured by image analysis.

The effects of perturbed, sub-optimal cultivation conditions, like extreme pH and temperature, high inorganic salt concentrations or hydrodynamic stress, on various kinds of living cells in suspension have already been studied extensively [42]. Plant cells have commonly been regarded as extraordinarily shear sensitive because of their relatively large size, their rigid cell wall, and their large vacuoles [42]. Since plant cells often grow as aggregates in relatively large structures, they can be easily mechanically disintegrated by turbulent eddies, of the same or smaller size than the aggregates, which are introduced by stirring and gassing into the cultivation broth [42, 43]. Therefore, plant cells are very sensitive to hydrodynamic stress [42, 43, 44, 45]. This process parameter might therefore be one of the main viability impairing factors of large-scale plant cell culture. Thus, many large-scale plant cell bioreactor designs for minimization of hydrodynamic forces resulting from mechanical agitation have been suggested over the years [46]. To investigate the influence of mechanical stresses within a stirred tank bioreactor, T. chinensis cells were exposed to an InterMIG stirrer. Established viability tests were used to study the effect of mechanical stresses on cell viability. Figure 8 shows micrographs of T. chinensis cells after 3 and 48 h of exposure to hydrodynamic stress induced by the InterMIG stirrer. Initially, lots of opaque and translucent cells with some brownish areas are apparent, indicating a decent viability. However, after 48 h of stirring-induced stress, a clear disintegrating effect on the smaller aggregates can be observed. By the shear forces of the InterMIG stirrer, the plant cell aggregates are decomposed into smaller groups of cells. Furthermore, a significant increase of the familiar brownish color of Taxus cells during stress can be seen. These qualitative results correlate very well with the established AB assay and image analytic assessment. Figure 9 shows the cell viability analysis over a cultivation period of 48 h using the AB assay and image analytic method. The AB test showed approximately constant cell viability greater than 70 % for about 22 h. Subsequently, a continuous decrease in cell viability down to 10 % is obvious. The cell viability was also quantified by image analysis. The results show a similar trend with some discrepancies for high and low values. The transition from a fairly good viability over 70 %, to a gradual decrease down to 25 % starts around 22 h, the same time as the AB assay results indicate. Generally both methods lead to analogous results and are well suited for viability estimation of T. chinensis cells.
Fig. 8

Micrographs of T. chinensis cells after 3 h (a) and 48 h (b) stress exposure. The scale corresponds to 1,000 μm

Fig. 9

Determination of cell viability in shear stress experiments with InterMIG stirrer over time: AB assay (a) and image analysis (b). Values are means for three replicates


Production of metabolites via plant cell suspension culture is a renewable, environmentally friendly, and economically feasible alternative for extraction from whole plant material [46]. Advancements have been made in understanding metabolite production in large-scale plant cell cultures, but so far there is still a high optimization potential to increase final product titer. The introduced AB assay was found to be exceptionally eligible for viability estimation in industrial processes. Moreover, aggregate coloration, as a brightness attribute, was identified as a good indicator of viability using image analysis methods. Both methods provide useful techniques for determining the viability at different physico-chemical perturbations on T. chinensis cell aggregates.

By optimizing the process parameters incubation time, place, type and age of the cell culture used for generating the calibration curve, higher fluorescence values and more reproducible results for the AB assay could be achieved. By image analysis, the percentage of dead cells could be reliably measured over a grayscale balance between the living and dead cell areas. Both methods of viability determination were compared and led to similar conclusions. However, compared to the AB assay the image analysis technique exhibited higher error values while discrepancies between differently treated samples were less significant.

In total, both AB assay and image analysis offer significant advantages to already established viability assays in plant cells. The AB assay is not destructive or generates toxic samples. It offers many advantages over tetrazolium salts and omits their incompatibility issues [16]. It also has a higher sensitivity at lower percentage viability. The flexibility in the quantification method—fluoro- or spectrophotometrically—is also advantageous [47]. It could be observed in the experiments carried out, that the deviations of triplicates of the AB assay were significantly lower than the error values in the MTT test.

For application, the two viability assays were successfully introduced to determine the influence of different physico-chemical perturbations, such as UV, chemical additives, salt load, high and low temperature and stirrer-induced shear stress, on the viability of T. chinensis suspension cultures. The AB assay was found to be exceptional eligible for viability estimation in industrial processes. Moreover, aggregate coloration, as a brightness attribute, was identified to be clearly related to T. chinensis plant cell viability. Image analysis of culture broth photographs evaluating the brownish color of the aggregates provides an easy and fast way for viability estimation especially well suited for industrial application.

The setup used in this study provided a good way to quantify the influence of hydrodynamic stress within a stirred tank bioreactor on plant cell culture viability. Generally, hydrodynamic stresses lead to a decline not only in viability, but in aggregate size as well. An application to measure aggregate size of plant cells using the technique of laser diffraction was recently introduced by Wucherpfennig et al. [2]. This system could under non-growth conditions (suspension of aggregates in sodium chloride solution) also be used to study various stirrers and reactor geometries for their induced shear stress and effect on plant aggregate integrity. A similar approach was taken by Eslahpazir et al. [48] using pellets of A. niger. Usually the disintegration effect of mechanical stress and power input is measured using a shear sensitive clay polymer floc system [49, 50]. However, the employment of real biological aggregates has obvious advantages like analogous particle rigidity, size and form. Thus, comminution of native Taxus aggregates might provide further insight about shear susceptibility of biological samples and might complement and expand the results of this study.


  1. 1.
    Roberts SC (2007) Production and engineering of terpenoids in plant cell culture. Nat Chem Biol 3(7):387–395Google Scholar
  2. 2.
    Wucherpfennig T, Schilling J, Sieblitz D, Pump M, Schütte K, Wittmann C, Krull R (2012) Improved assessment of aggregate size in Taxus plant cell suspension cultures using laser diffraction. Eng Life Sci 12(6):595–602. doi:10.1002/elsc.201200039 CrossRefGoogle Scholar
  3. 3.
    Kolewe ME, Henson MA, Roberts SC (2010) Characterization of aggregate size in Taxus suspension cell culture. Plant Cell Rep 29(5):485–494CrossRefGoogle Scholar
  4. 4.
    Kolewe ME, Henson MA, Roberts SC (2011) Analysis of aggregate size as a process variable affecting paclitaxel accumulation in Taxus suspension cultures. Biotechnol Progr 27:1365–1372. doi:10.1002/btpr.655 CrossRefGoogle Scholar
  5. 5.
    Patil R, Kolewe M, Roberts S (2013) Cellular aggregation is a key parameter associated with long term variability in paclitaxel accumulation in Taxus suspension cultures. Plant Cell Tiss Org 112(3):303–310. doi:10.1007/s11240-012-0237-3 CrossRefGoogle Scholar
  6. 6.
    Hirasuna TJ, Pestchanker LJ, Srinivasan V, Shuler ML (1996) Taxol production in suspension cultures of Taxus baccata. Plant Cell Tiss Org 44(2):95–102. doi:10.1007/bf00048185 CrossRefGoogle Scholar
  7. 7.
    Zilkah S, Gressel J (1978) The estimation of cell death in suspension cultures evoked by phytotoxic compounds: differences among techniques. Plant Sci Lett 12(3–4):305–315. doi:10.1016/0304-4211(78)90083-4 CrossRefGoogle Scholar
  8. 8.
    Duncan D, Widholm J (1990) Measurements of viability suitable for plant tissue cultures. In: Pollard J, Walker J (eds) Plant cell and tissue culture vol. 6. Methods in molecular biology™. Humana Press, New York, pp 29–37. doi:10.1385/0-89603-161-6:29 CrossRefGoogle Scholar
  9. 9.
    Byth H-A, McHunu BI, Dubery IA, Bornman L (2001) Assessment of a simple, non-toxic alamar blue cell survival assay to monitor tomato cell viability. Phytochemical Anal 12(5):340–346. doi:10.1002/pca.595 CrossRefGoogle Scholar
  10. 10.
    Berridge MV, Tan AS (1993) Characterization of the cellular reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT): subcellular localization, substrate dependence, and involvement of mitochondrial electron transport in MTT reduction. Arch Biochem Biophys 303(2):474–482CrossRefGoogle Scholar
  11. 11.
    Towill LE, Mazur P (1975) Studies on the reduction of 2,3,5-triphenyltetrazolium chloride as a viability assay for plant tissue cultures. Can J Bot 53(11):1097–1102. doi:10.1139/b75-129 CrossRefGoogle Scholar
  12. 12.
    Iborra JL, Guardiola J, Montaner S, Cánovas M, Manjón A (1992) 2,3,5-triphenyltetrazolium chloride as a viability assay for immobilized plant cells. Biotech Tech 6(4):319–322. doi:10.1007/bf02439319 CrossRefGoogle Scholar
  13. 13.
    Jianfeng X, Zhiguo S, Pusun F (1998) Suspension culture of compact callus aggregate of Rhodiola sachalinensis for improved salidroside production. Enzyme Microb Technol 23(1–2):20–27. doi:10.1016/s0141-0229(98)00011-8 CrossRefGoogle Scholar
  14. 14.
    Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 65(1–2):55–63. doi:10.1016/0022-1759(83)90303-4 CrossRefGoogle Scholar
  15. 15.
    Nociari MM, Shalev A, Benias P, Russo C (1998) A novel one-step, highly sensitive fluorometric assay to evaluate cell-mediated cytotoxicity. J Immunol Methods 213(2):157–167. doi:10.1016/s0022-1759(98)00028-3 CrossRefGoogle Scholar
  16. 16.
    Rampersad SN (2012) Multiple applications of Alamar Blue as an indicator of metabolic function and cellular health in cell viability bioassays. Sensors 12(9):12347–12360CrossRefGoogle Scholar
  17. 17.
    O’Brien J, Wilson I, Orton T, Pognan F (2000) Investigation of the Alamar Blue (resazurin) fluorescent dye for the assessment of mammalian cell cytotoxicity. Eur J Biochem 267(17):5421–5426. doi:10.1046/j.1432-1327.2000.01606.x CrossRefGoogle Scholar
  18. 18.
    Baker C, Banerjee S, Tenover F (1994) Evaluation of Alamar colorimetric MIC method for antimicrobial susceptibility testing of gram-negative bacteria. J Clin Microbiol 32(5):1261–1267Google Scholar
  19. 19.
    Collins L, Franzblau S (1997) Microplate Alamarblue assay versus BACTEC 460 system for high-throughput screening of compounds against Mycobacterium tuberculosis and Mycobacterium avium. Antimicrob Agents Chemother 41(5):1004–1009Google Scholar
  20. 20.
    Larson EM, Doughman DJ, Gregerson DS, Obritsch WF (1997) A new, simple, nonradioactive, nontoxic in vitro assay to monitor corneal endothelial cell viability. Invest Ophthalmol Vis Sci 38(10):1929–1933Google Scholar
  21. 21.
    Pfaller M, Grant C, Morthland V, Rhine-Chalborg J (1994) Comparative evaluation of alternative methods for broth dilution susceptibility testing of fluconazole against Candida albicans. J Clin Microbiol 32(2):506–509Google Scholar
  22. 22.
    White M, DiCaprio M, Greenberg D (1996) Assessment of neuronal viability with Alamar blue in cortical and granule cell cultures. J Neurosci Methods 70(2):195–200CrossRefGoogle Scholar
  23. 23.
    Ibaraki Y, Kenji K (2001) Application of image analysis to plant cell suspension cultures. Comput Electron Agr 30(1–3):193–203CrossRefGoogle Scholar
  24. 24.
    Remotti PC, Löffler HJM (1995) Callus induction and plant regeneration from gladiolus. Plant Cell Tiss Org 42(2):171–178. doi:10.1007/bf00034235 CrossRefGoogle Scholar
  25. 25.
    Smith MAL, Reid JF, Hansen AC, Li Z, Madhavi DL (1995) Non-destructive machine vision analysis of pigment-producing cell cultures. J Biotechnol 40(1):1–11. doi:10.1016/0168-1656(95)00025-l CrossRefGoogle Scholar
  26. 26.
    Malik S, Cusidó RM, Mirjalili MH, Moyano E, Palazón J, Bonfill M (2011) Production of the anticancer drug taxol in Taxus baccata suspension cultures: a review. Process Biochem 46(1):23–34CrossRefGoogle Scholar
  27. 27.
    Gibson D, Ketchum R, Vance N, Christen A (1993) Initiation and growth of cell lines of Taxus brevifolia (Pacific yew). Plant Cell Rep 12(9):479–482. doi:10.1007/bf00236091 CrossRefGoogle Scholar
  28. 28.
    Bruňáková K, Babincová Z, Čellárová E (2004) Selection of callus cultures of Taxus baccata L. as a potential source of paclitaxel production. Eng Life Sci 4(5):465–469. doi:10.1002/elsc.200420050 CrossRefGoogle Scholar
  29. 29.
    Bruňáková K, Babincová Z, Čellárová E (2005) Production of taxanes in callus and suspension cultures of Taxus baccata L. In: Hvoslef-Eide A, Preil W (eds) Liquid culture systems for in vitro plant propagation. Springer, Netherlands, pp 567–574. doi:10.1007/1-4020-3200-5_43 CrossRefGoogle Scholar
  30. 30.
    Wickremesinhe ERM, Arteea RN (1993) Taxus callus cultures: initiation, growth optimization, characterization and taxol production. Plant Cell Tiss Org 35(2):181–193. doi:10.1007/bf00032968 CrossRefGoogle Scholar
  31. 31.
    Fett-Neto AG, Zhang WY, Dicosmo F (1994) Kinetics of taxol production, growth, and nutrient uptake in cell suspensions of Taxus cuspidata. Biotechnol Bioeng 44(2):205–210. doi:10.1002/bit.260440209 CrossRefGoogle Scholar
  32. 32.
    Gamborg O, Miller R, Ojima K (1968) Nutrient requirements of suspension cultures of soybean root cells. Exp Cell Res 50:151–158CrossRefGoogle Scholar
  33. 33.
    Kim BJ, Gibson DM, Shuler ML (2005) Relationship of viability and apoptosis to taxol production in Taxus sp. suspension cultures elicited with methyl jasmonate. Biotechnol Progr 21(3):700–707. doi:10.1021/bp050016z CrossRefGoogle Scholar
  34. 34.
    Voytik-Harbin S, Brightman A, Waisner B, Lamar C, Badylak S (1998) Application and evaluation of the Alamar Blue assay for cell growth and survival of fibroblasts. Vitro Cell Dev Biol Anim 34(3):239–246. doi:10.1007/s11626-998-0130-x CrossRefGoogle Scholar
  35. 35.
    Hulst AC, Meyer MMT, Breteler H, Tramper J (1989) Effect of aggregate size in cell cultures of Tagetes patula on thiophene production and cell growth. Appl Microbiol Biotechnol 30(1):18–25CrossRefGoogle Scholar
  36. 36.
    Pépin M, Smith M, Reid J (1999) Application of imaging tools to plant cell culture: relationship between plant cell aggregation and flavonoid production. Vitro Cell Dev-Pl 35(4):290–295. doi:10.1007/s11627-999-0036-7 Google Scholar
  37. 37.
    Zhong J–J, Pan Z-W, Wang Z-Y, Wu J, Chen F, Takagi M, Yoshida T (2002) Effect of mixing time on taxoid production using suspension cultures of Taxus chinensis in a centrifugal impeller bioreactor. J Biosci Bioeng 94(3):244–250. doi:10.1016/s1389-1723(02)80157-8 CrossRefGoogle Scholar
  38. 38.
    Keßler M, ten Hoopen HJG, Furusaki S (1999) The effect of the aggregate size on the production of ajmalicine and tryptamine in Catharanthus roseus suspension culture. Enzyme Microb Technol 24(5–6):308–315. doi:10.1016/S0141-0229(98)00121-5 CrossRefGoogle Scholar
  39. 39.
    Miao G-p, Zhu C-s, Feng J-t, Han J, Song X-W, Zhang X (2013) Aggregate cell suspension cultures of Tripterygium wilfordii Hook. f. for triptolide, wilforgine, and wilforine production. Plant Cell Tiss Org 112(1):109–116. doi:10.1007/s11240-012-0211-0 CrossRefGoogle Scholar
  40. 40.
    Wickremesinhe ERM, Arteca RN (1994) Taxus cell suspension cultures: optimizing growth and production of Taxol. J Plant Physiol 144(2):183–188. doi:10.1016/S0176-1617(11)80541-9 Google Scholar
  41. 41.
    Ellis DD, Zeldin EL, Brodhagen M, Russin WA, McCown BH (1996) Taxol1 production in nodule cultures of Taxus. J Nat Prod 59(3):246–250. doi:10.1021/np960104g CrossRefGoogle Scholar
  42. 42.
    Meijer JJ, ten Hoopen HJG, Luyben KCAM, Libbenga KR (1993) Effects of hydrodynamic stress on cultured plant cells: a literature survey. Enzyme Microb Technol 15(3):234–238. doi:10.1016/0141-0229(93)90143-p CrossRefGoogle Scholar
  43. 43.
    Kieran PM, MacLoughlin PF, Malone DM (1997) Plant cell suspension cultures: some engineering considerations. J Biotechnol 59(1–2):39–52. doi:10.1016/s0168-1656(97)00163-6 CrossRefGoogle Scholar
  44. 44.
    Dunlop EH, Namdev PK, Rosenberg MZ (1994) Effect of fluid shear forces on plant cell suspensions. Chem Eng Sci 49(14):2263–2276. doi:10.1016/0009-2509(94)e0043-p CrossRefGoogle Scholar
  45. 45.
    Joshi JB, Elias CB, Patole MS (1996) Role of hydrodynamic shear in the cultivation of animal, plant and microbial cells. Chem Eng J and the Biochem Eng J 62(2):121–141. doi:10.1016/0923-0467(95)03062-x Google Scholar
  46. 46.
    Kolewe ME, Gaurav V, Roberts SC (2008) Pharmaceutically active natural product synthesis and supply via plant cell culture technology. Mol Pharm 5(2):243–256. doi:10.1021/mp7001494 CrossRefGoogle Scholar
  47. 47.
    White S, McIntyre M, Berry DR, McNeil B (2002) The autolysis of industrial filamentous fungi. Crit Rev Biotechnol 22(1):1–14CrossRefGoogle Scholar
  48. 48.
    Eslahpazir Esfandabadi M, Wucherpfennig T, Krull R (2012) Agitation induced mechanical stress in stirred tank bioreactors-linking CFD simulations to fungal morphology. J Chem Eng Jpn 45(9):742–748CrossRefGoogle Scholar
  49. 49.
    Mahnke EU, Büscher K, Hempel DC (2000) A novel approach for the determination of mechanical stresses in gas-liquid reactors. Chem Eng Technol 23(6):509–513CrossRefGoogle Scholar
  50. 50.
    Pilz RD, Hempel DC (2005) Mechanical stress on suspended particles in two- and three-phase airlift loop reactors and bubble columns. Chem Eng Sci 60(22):6004–6012. doi:10.1016/j.ces.2005.04.001 CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Thomas Wucherpfennig
    • 1
  • Annika Schulz
    • 1
  • Jaime Arturo Pimentel
    • 2
  • Gabriel Corkidi
    • 2
  • Dominik Sieblitz
    • 3
  • Matthias Pump
    • 3
  • Gilbert Gorr
    • 3
  • Kai Schütte
    • 3
  • Christoph Wittmann
    • 1
  • Rainer Krull
    • 1
  1. 1.Institute of Biochemical EngineeringTechnische Universität BraunschweigBrunswickGermany
  2. 2.Image Analysis Laboratory, Institute of BiotechnologyUniversidad Nacional Autónoma de México (UNAM)CuernavacaMexico
  3. 3.Phyton Biotech GmbHAhrensburgGermany

Personalised recommendations