Plant Cell Tiss Organ Cult DOI 10.1007/s11240-014-0448-x
ORIGINAL PAPER
Image-based analysis of cell-specific productivity for plant cell suspension cultures Heide Havenith • Nicole Raven • Stefano Di Fiore Rainer Fischer • Stefan Schillberg
•
Received: 28 November 2013 / Accepted: 5 February 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract More and more plant cell suspension cultures are regarded as an attractive alternative to mammalian cells as host organism for production of complex recombinant proteins. The most important advantages of the production platform are low costs, easy scalability and enhanced safety by complete lack of animal components in the cultivation media. In order to characterize, understand and control such systems accurately, it is important to determine the cellspecific productivity (Qp) of plant cell-based production platforms. Compared to many microbial and mammalian cells the morphology of plant cells is nonhomogeneous and the cells tend to form aggregates, therefore commercial cell counting systems are too unreliable to determine cell numbers in plant suspension cultures. We addressed this limitation by developing a novel cell counting method based on a combination of cell-staining and automated confocal fluorescence microscopy. This method allowed us, for the first time, to determine the cell-specific productivity of transgenic tobacco (Nicotiana tabacum cv. Bright Yellow-2) cell suspension cultures producing the human antibody M12. In the future this method will be a useful tool in the development of optimized plant cell-based production processes. H. Havenith N. Raven S. Di Fiore R. Fischer S. Schillberg (&) Fraunhofer Institute for Molecular Biology and Applied Ecology IME, Forckenbeckstrasse 6, 52074 Aachen, Germany e-mail:
[email protected] R. Fischer Institute for Molecular Biotechnology, RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany S. Schillberg Institute for Phytopathology and Applied Zoology, Justus-Liebig University Giessen, Heinrich-Buff-Ring 26-32, 35392 Giessen, Germany
Keywords Automated cell counting Growth rate determination Human monoclonal antibody Molecular farming OPERA Transgenic BY-2 cells
Introduction Plant cell suspension cultures have recently emerged as versatile production systems for recombinant proteins that cannot be expressed in prokaryotes (Fischer et al. 2004; van Dussen et al. 2013). Plant cells are safer and less expensive to cultivate than mammalian cells, and are easier to scale up (Boehm 2007; Schillberg et al. 2013). Cell suspension cultures also offer several advantages over whole plants, including shorter production cycles, a controlled environment and the lack of biotic and abiotic contaminants (Schillberg et al. 2013; Xu et al. 2011). Although many recombinant proteins have been produced in plant cell suspension cultures, there is currently no method available for the accurate determination of cell numbers. Plant cells often undergo morphological changes during cultivation due to the culture conditions (McDonald et al. 2001), and are prone to form aggregates several millimeters in diameter (Naill and Roberts 2004), making it difficult to separate individual cells properly for counting. Therefore standard parameters for measuring biomass— such as dry weight (DW), fresh weight (FW) or packed cell volume (PCV)—are not suitable for the determination of BY-2 cell numbers. Treatments that can produce single cells include the maceration of cell clusters with CrO3, enzymatic digestion of the middle lamella, and protoplast generation, but they are expensive and time consuming, many cells are lost during preparation, and the precise number of cells in culture cannot be determined (de Gunst et al. 1990; Nicoloso et al. 1994).
123
Plant Cell Tiss Organ Cult
Several cell counting protocols and instruments have been developed for homogeneous suspensions of mammalian and bacterial cells, but these methods tend to be unsuitable for plant cell suspension cultures. Efforts to establish counting systems for plant cells that do not rely on hemocytometry include focused beam reflectance measurement and flow sorting (Jeffers et al. 2003; Nicoloso et al. 1994). Lamboursain and Jolicoeur (2005) described an approach in which the total fluorescence signal from stained nuclei was analyzed in a microplate reader and the cell number was estimated by referring to standards with known cell numbers. However, the accuracy of this method was influenced by the cell cycle, which determines the size of the nucleus and the degree of chromatin condensation. Furthermore the standards were produced by generating single cells and counting them using a hemocytometer, with the inherent disadvantages already stated above. There is strong demand for an accurate method that describes the growth kinetics of plant cell suspension cultures, and we have therefore developed a new approach for cell counting which is compatible with highly-aggregated cells and does not rely on the preparation of standards. The method is based on a combination of nuclear labeling, automated fluorescence microscopy and image analysis. This allows the accurate estimation of cell-specific productivity in tobacco (Nicotiana tabacum cv. BY-2) cell suspension cultures producing recombinant proteins.
Materials and methods Generation of transgenic BY-2 suspension cells Agrobacterium tumefaciens strain GV3101:pMP90RK (Koncz and Schell 1986) was transformed by electroporation with the pTRAkc-MTAD vector containing the expression cassettes for the heavy and light chains of monoclonal antibody M12 (Raven et al. 2010) and used for the transformation of tobacco (Nicotiana tabacum) cv. Bright Yellow-2 (BY-2) wild type cells, followed by regeneration on selective medium (An 1985). Stably-transformed callus tissue was used to establish the cell suspension cultures. Maintenance of cell cultures Transgenic BY-2 cell suspension cultures (Nagata et al. 1992) were routinely cultivated in 50 ml MSN medium in 100-ml glass Erlenmeyer flasks under sterile conditions at 26 °C in the dark, with constant orbital agitation at 180 rpm. The optimized MSN medium described by Holland et al. (2010) is based on standard MS medium (Murashige and Skoog 1962) and comprised 0.43 % (w/v) basal MSMO salts (Duchefa, Haarlem, Netherlands), 3 %
123
(w/v) sucrose, 0.02 % (w/v) potassium dihydrogen phosphate, 0.01 % (w/v) myo-inositol, 1 mg/l thiamine and 0.2 mg/l 2,4-dichlorophenoxyaceticacid, supplemented with 1.01 % (w/v) sodium nitrate (final pH 5.8). Suspension cells were subcultured at 7-d intervals by transferring 5 % (v/v) of the existing culture into fresh liquid medium. Time course experiment and sample processing We inoculated 200 ml of fresh medium with 10 ml of an existing BY-2 cell suspension culture in 500-ml shake flasks and incubated as above. Two parallel cultures (MSN1 and MSN-2) were cultivated and three samples, one 10-ml and two 2-ml aliquots were taken from each culture on days 0, 3, 4, 5, 6, 7 and 10 post-subculture. We measured the packed cell volume (PCV), fresh weight (FW), dry weight (DW), antibody concentration and cell counts. Growth parameters were determined from the 10-ml aliquot as follows. After centrifugation (15 min, 4,0009g, room temperature) the packed cell volume (PCV) was determined according to the graduation of the centrifuge tube. The fresh weight (FW) was determined after separating the cells from the culture broth by vacuum tank filtration and weighing the biomass. The dry weight (DW) was determined by incubating the filtered cell material at 37 °C until a constant mass was obtained. One 2-ml aliquot served as the starting material for cell staining and microscopy (see below), and the other aliquot was used to quantify the recombinant antibody M12. The cells were first pelleted by centrifugation (20 min, 14,0009g, 4 °C) and the culture supernatant containing the secreted M12 antibody was transferred to a new vial. Samples were stored at -20 °C for subsequent en bloc analysis at the end of the time course experiment. Quantification of the recombinant protein The quantity of the M12 antibody in the spent medium was determined using an enzyme-linked immunosorbent assay (ELISA) as previously described (Kirchhoff 2012). Cell staining and microscopy The 2 ml BY-2 cell suspension cultures samples were diluted 1:20–1:160 fold in PBS (pH 7.4) according to their density, and 30 ll aliquots were transferred to 96-well lclear half-area plates (Greiner Bio-One GmbH, Frickenhausen, Germany). We fixed the cells in the same volume of 99.8 % (w/v) ethanol to allow complete nuclear staining with 1.4 lM DAPI. All samples were measured in technical triplicates. We acquired 20 images from each well, representing 8.82 mm2 surface area (i.e. 26 % coverage of the cavity), using the OPERA QEHS High Content
Plant Cell Tiss Organ Cult
Screening System (PerkinElmer, Hamburg, Germany) equipped with a 10 9 air objective with 0.4 numerical aperture. The autofocus was set to 27 lm at plate bottom. DAPI staining was detected using an excitation wavelength of 405 nm and a 450/50 nm band pass emission filter. Data processing and cell number evaluation Individual images were evaluated using the Acapella ‘‘Nuclei Counting’’ script (v2.0, PerkinElmer) adjusting the detection parameters to achieve the highest degree of cell recognition: Nuclei detection algorithm, G; Minimum nuclei distance, 1.5; Threshold adjustment, 1; Nuclear splitting adjustment, 6; Individual threshold adjustment, 0.7; Minimum nuclear area, 5; Minimum nuclear contrast, 0.01. Only those values showing linear dependency between the dilution factor and the cell number were selected for further evaluation. As stated above, 20 images per well were acquired, thereby covering 26 % of the total cavity area. The cell number in the suspension culture was calculated using the following equation: P20 1 N cell counts= ¼ ml Að%Þ VðmlÞ 1=D where N is the number of nuclei per image, A is the area covered by the images, V is the sample volume and D the dilution factor. Cell-specific productivity calculation The cell-specific productivity (Qp) is a function of growth rate (l) and productivity. The growth rate (l) was calculated in the exponential phase (72–168 h) as followed: lð1=hÞ ¼
lnN2 lnN1 t2 t1
where N is the cell counts and t is the cultivation time. The equation for the cell-specific productivity: P1 mg P2 mg ml ml Qp ðmg=cell=hÞ ¼ lð1=hÞ 1 1 N2 ml N1 ml where P is the volumetric antibody concentration in the cell culture supernatant.
propidium iodide. We also tested the labeling of all cell boundaries with calcofluor white. DAPI staining was the most suitable method because sample preparation was simpler than the other methods, and produced images with the best signal-to-noise ratio (data not shown). On each plate, the samples were applied at different dilutions in triplicate wells. The cells were fixed with ethanol and stained with DAPI. The fixation step is advantageous because it allows rapid penetration of the nuclear dye. Furthermore, the plant cells sediment to the bottom of the well so that after sufficient dilution, individual cells can be visualized effectively by confocal microscopy. The nuclei were imaged with the OPERA high-content screening system (Fig. 1). We acquired 20 images per well distributed in a grid pattern that covered 26 % of the area of each well. The images were analyzed using Acapella software with the Nuclei Counting script and an optimized parameter set (see materials and methods). By inspecting the analyzed images, we estimated a false negative detection rate of less than 5 % of counted events, e.g. where dividing nuclei were close together or where cell clustering caused nuclei to overlap in the image. In such cases the software counted only one nucleus instead of two (Fig. 1b) introducing a marginal error that did not significantly affect the accuracy of the method. Figure 2 shows different dilutions of a BY-2 sample plotted against the cell counts from the 20 images taken from each well. We observed a high linear correlation between the dilution of the suspension culture and the number of counted cells for up to 2,800 nuclei per well. The coefficient of determination was greater than 0.99, confirming the accuracy of the measurement. Dilutions with fewer than 1,500 nuclei were generally found to be most suitable for achieving good cell number estimates, because these dilutions tended to lack large clusters of cells pilling up at the well bottom. The cell number per well was extrapolated from the outcome of 20 images, taking into account that the images only covered 26 % of the total well area. For the sample in Fig. 2, 2,800 nuclei were counted in 20 images, indicating that the actual number of cells at the bottom of the well was approximately 10,700. Considering that the sample volume was 30 ll with a dilution factor of 1:4, the number of cells was therefore approximately 1.4 9 106 per ml culture volume. cell counts= ¼ 2; 800 1 ¼ 1:43 106 ml 0:26 0:03 0:25
Results Determination of cell-specific productivity Establishment of a cell counting method Tobacco BY-2 cells were distributed into 96-well microtiter plates and the nuclei were stained with DAPI or
The specific productivity of the BY-2 cell suspension culture was determined by carrying out a time course experiment using two parallel cultures of the transgenic
123
Plant Cell Tiss Organ Cult
Fig. 1 a Image of BY-2 cells stained with DAPI (1 of 20 images per well), acquired with the OPERA system using a 109 objective. b The same image analyzed with the Acapella software. Nuclei detected by
the software are surrounded by colored circles, with randomly assigned color codes
Fig. 2 Different dilutions of a BY-2 cell suspension culture plotted against cell counts in 20 pictures per well, representing 26 % of total well area (technical triplicates, n = 3). Solid line is the interpolated line corresponding to the equation shown. Dotted line shows the results for less diluted samples, where cells tend to form aggregates that are not well separated and result in a nonlinear outcome
Fig. 4 Time course of secreted M12 antibody concentration in the culture supernatant for two parallel cultures (MTAD#31.2) cultivated in MSN medium. The antibody concentration was determined by ELISA
Fig. 3 Time course of DW (solid line) and FW (dashed line) for two parallel cultures of the MTAD#31.2 cell line in MSN medium
123
BY-2 cell line (MTAD#31.2) producing the recombinant human antibody M12. We measured the packed cell volume (PCV), fresh weight (FW), dry weight (DW) (Fig. 3), antibody concentration (Fig. 4) and cell counts (Fig. 5). The PCV correlated strongly with the FW (data not shown). The FW reached 310 g/l by the end of cultivation, whereas the DW reached 15 g/l. The highest antibody concentration (*100 mg/l) was achieved 168 h post-subcultivation, when the cell number reached the maximum of *4.5 9 106 cells/ml. The cell growth rate was obtained from data collected during the exponential growth phase and was 0.034/h. The cell-specific productivity for the M12 antibody was correlated to concentrations of M12 and was estimated at 8.0 pg/cell/day for MSN-1 and 7.6 pg/cell/day for MSN-2.
Plant Cell Tiss Organ Cult
Fig. 5 Time course of cell counts for two parallel cultures (MTAD#31.2) cultivated in MSN medium
Discussion Many different counting methods are commercially available for bacterial and mammalian cells, but there is no equivalent system for plant cells despite the significant demand. Plant cells have emerged as a competitive platform for the production of recombinant proteins, so accurate cell counting is an important requirement. However, counting plant cells during cultivation is challenging because of their nonhomogeneous morphology and tendency to form aggregates. This is particularly important in the case of tobacco cells, which are widely used as an expression platform due to their high growth rate but whose morphology shows more extreme diversity compared to other plant suspension cells (McDonald et al. 2001). To address these challenges, previous research has focused on the development of suitable methods for cell counting. For example, Nicoloso et al. (1994) counted tobacco cells (cv. White Burley) by flow cytometry after the culture was treated to obtain single cells, whereas Lamboursain and Jolicoeur (2005) stained Eschscholzia californica cells and measured the total fluorescence in a microplate reader to estimate cell counts by comparison to an external standard. We developed a novel method for counting plant cells based on a combination of nuclear staining, automated fluorescence microscopy and image analysis. We used this method to count tobacco BY-2 cells and demonstrated a linear relationship between dilution and cell number, confirming the validity of our approach. The method benefited from a simple sample preparation technique that required neither the generation of single cells or protoplasts nor the use of an external standard, therefore saving time and costs. The use of the OPERA system and the Acapella Software for image analysis allowed the method to be automated, so that large numbers of nuclei could be counted more rapidly and accurately than manual counting methods. For
example, Tucker et al. (1994) showed that cell counts obtained by five people studying the same picture varied by more than 10 %. The time course experiment indicated a good correlation between the cell counts and DW. However the DW did not directly correlate to the biomass (FW) during the end of cultivation due to the strong uptake of water by the cells (Jeffers et al. 2003; Ullisch et al. 2012). This phenomenon is described as the ‘‘applesauce effect’’ and is related to the difference in osmolality between the media and the cells, which can often cause the cells to burst. The high water content increased the FW values until the end of cultivation, whereas the cell number and the DW values declined slightly due to cell bursting. The determined cell counts also corresponded to previously reported values (Lamboursain and Jolicoeur 2005; Nicoloso et al. 1994). Although our experiments focused on a single plant species, DAPI staining is suitable for all eukaryotic cells and the parameters for image analysis can be adapted to nuclei with different sizes and morphologies. We anticipate that our method should be transferrable to cell suspension cultures of any plant species and we have already obtained encouraging preliminary results using Arabidopsis thaliana cells (data not shown). The automated microscopy method we developed not only allows the detection of nuclei stained with DAPI but also provides additional opportunities based on the use of other excitation wavelengths. For the OPERA system, up to three fluorescence signals can be imaged simultaneously so that the counting of DAPI-stained nuclei can be combined e.g. with the measurement of GFP and/or DsRed fluorescence to determine transformation efficiencies and select high-performance clones (Kirchhoff et al. 2012). The new counting method for plant cell suspension cultures facilitates the detailed characterization of cell growth and is also a prerequisite for the determination of cell-specific productivity. In our study, the growth rate was found to be 0.034/h. This is equivalent to 0.82 cell divisions per day, similar to the results reported by Holland et al. (2013) for transgenic BY-2 cells (0.62 cell divisions per day) and Nagata et al. (1992) for wild-type BY-2 cells (0.9 cell divisions per day). The marginal differences between wild type and transgenic BY-2 cells may reflect the slightly lower growth rate of transformed cells, resulting from the metabolic burden of producing recombinant proteins (Jiang et al. 2006). The accumulation of antibody M12 reached a maximum of *100 mg/l in the culture medium under our experimental conditions. This value is in the medium range compared to other eukaryotic production systems but there are many opportunities to increase productivity, including medium and process optimization by switching to methods such as fed batch or continuous fermentation (Holland
123
Plant Cell Tiss Organ Cult
et al. 2013; Vasilev et al. 2013; Xu et al. 2011). The FW/ DW ratio indicated that the BY-2 line we tested had a high content of water. The industrial use of plant cell suspension cultures is primarily limited by a combination of economic and engineering factors (Kieran 2001), so reducing the cell water content e.g. by using high molar medium could help to improve the volumetric productivity. The cell-specific productivity for BY-2 cells secreting the model antibody M12 (8 pg/cell/day) indicated that plant cell suspension cultures are a competitive platform for the production of recombinant antibodies. In comparison, antibody yields of 25–40 pg/cell/day have been obtained with mammalian cell culture systems (Chartrain and Chu 2008). However, these levels were only achieved after time consuming and laborious selection of high producers that show increased transgene copy numbers and/or transcription rates. In conclusion, we have described for the first time a method that allows the accurate determination of the cellspecific productivity of plant cell suspension cultures. The method is easy to apply, can be adapted to any automated fluorescence image plate reader system (preferentially but not necessarily confocal microscopy) in combination with software for image analysis, and can be easily transferred to plant cell suspension cultures of other relevant species. Acknowledgments The authors wish to thank Dr. Flora Schuster for her skilled technical assistance with the preparation of plant cell cultures and Dr. Richard M Twyman and Holger Spiegel for critical reading of the manuscript. This research was funded by the European Union Seventh Framework Programme under Grant Agreement No. 227420 CoMoFarm.
References An GH (1985) High-efficiency transformation of cultured tobacco cells. Plant Physiol 79(2):568–570. doi:10.1104/Pp.79.2.568 Boehm R (2007) Bioproduction of therapeutic proteins in the 21st century and the role of plants and plant cells as production platforms. Ann NY Acad Sci 1102:121–134. doi:10.1196/annals. 1408.009 Chartrain M, Chu L (2008) Development and production of commercial therapeutic monoclonal antibodies in Mammalian cell expression systems: an overview of the current upstream technologies. Curr Pharm Biotechnol 9(6):447–467 de Gunst MCM, Harkes PAA, Val J, van Zwet WR, Libbenga KR (1990) Modelling the growth of a batch culture of plant cells: a corpuscular approach. Enzyme Microb Tech 12(1):61–71. doi:10.1016/0141-0229(90)90182-P Fischer R, Stoger E, Schillberg S, Christou P, Twyman RM (2004) Plant-based production of biopharmaceuticals. Curr Opin Plant Biol 7(2):152–158. doi:10.1016/j.pbi.2004.01.007 Holland T, Sack M, Rademacher T, Schmale K, Altmann F, Stadlmann J, Fischer R, Hellwig S (2010) Optimal nitrogen supply as a key to increased and sustained production of a monoclonal full-size antibody in BY-2 suspension culture. Biotechnol Bioeng 107(2):278–289 Holland T, Blessing D, Hellwig S, Sack M (2013) The in-line measurement of plant cell biomass using radio frequency
123
impedance spectroscopy as a component of process analytical technology. Biotechnol J 8(10):1231–1240. doi:10.1002/biot. 201300125 Jeffers P, Raposo S, Lima-Costa ME, Connolly P, Glennon B, Kieran PM (2003) Focussed beam reflectance measurement (FBRM) monitoring of particle size and morphology in suspension cultures of Morinda citrifolia and Centaurea calcitrapa. Biotechnol Lett 25(23):2023–2028 Jiang Z, Huang Y, Sharfstein ST (2006) Regulation of recombinant monoclonal antibody production in Chinese hamster ovary cells: a comparative study of gene copy number, mRNA level, and protein expression. Biotechnol Progr 22(1):313–318. doi:10. 1021/Bp0501524 Kieran PM (2001) Bioreactor design for plant cell suspension cultures. In: Cabral JMS, Mota M, Tramper J (eds) Multiphase bioreactor design. Taylor & Francis, London, pp 391–426 Kirchhoff J (2012) Generation of highly productive polyclonal and monoclonal tobacco suspension lines from a heterogeneous transgenic BY-2 culture through flow cytometric sorting. Dissertation, RWTH Aachen University Kirchhoff J, Raven N, Boes A, Roberts JL, Russell S, Treffenfeldt W, Fischer R, Schinkel H, Schiermeyer A, Schillberg S (2012) Monoclonal tobacco cell lines with enhanced recombinant protein yields can be generated from heterogeneous cell suspension cultures by flow sorting. Plant Biotechnol J 10(8):936–944. doi:10.1111/j.1467-7652.2012.00722.x Koncz C, Schell J (1986) The promoter of TL-DNA gene 5 controls the tissue-specific expression of chimaeric genes carried by a novel type of Agrobacterium binary vector. Molec Gen Genet 204(3):383–396. doi:10.1007/bf00331014 Lamboursain L, Jolicoeur M (2005) Determination of cell concentration in a plant cell suspension using a fluorescence microplate reader. Plant Cell Rep 23(10–11):665–672. doi:10.1007/s00299004-0899-3 McDonald KA, Jackman AP, Hurst S (2001) Characterization of plant suspension cultures using the focused beam reflectance technique. Biotechnol Lett 23(4):317–324. doi:10.1023/A:1005646 826204 Murashige T, Skoog F (1962) A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiol Plantarum 15(3):473–497. doi:10.1111/j.1399-3054.1962.tb08052.x Nagata T, Nemoto Y, Hasezawa S (1992) Tobacco BY-2 cell-line as the Hela-cell in the cell biology of higher-plants. Int Rev Cytol 132:1–30 Naill MC, Roberts SC (2004) Preparation of single cells from aggregated Taxus suspension cultures for population analysis. Biotechnol Bioeng 86(7):817–826. doi:10.1002/Bit.20083 Nicoloso FT, Val J, Vanderkeur M, Vaniren F, Kijne JW (1994) Flow-cytometric cell counting and DNA estimation for the study of plant-cell population-dynamics. Plant Cell Tiss Org 39(3):251–259. doi:10.1007/Bf00035978 Raven N, Schillberg S, Kirchhoff J, Bra¨ndli J, Imseng N, Eibl R (2010) Growth of BY-2 suspension cells and plantibody production in single-use bioreactors. In: Single-use technology in biopharmaceutical manufacture. Wiley, New York, pp 251–261. doi:10.1002/9780470909997.ch21 Schillberg S, Raven N, Fischer R, Twyman RM, Schiermeyer A (2013) Molecular farming of pharmaceutical proteins using plant suspension cell and tissue cultures. Curr Pharm Des 19(31): 5531–5542 Tucker KG, Chalder S, al-Rubeai M, Thomas CR (1994) Measurement of hybridoma cell number, viability, and morphology using fully automated image analysis. Enzyme Microb Technol 16(1): 29–35 Ullisch DA, Muller CA, Maibaum S, Kirchhoff J, Schiermeyer A, Schillberg S, Roberts JL, Treffenfeldt W, Buchs J (2012)
Plant Cell Tiss Organ Cult Comprehensive characterization of two different Nicotiana tabacum cell lines leads to doubled GFP and HA protein production by media optimization. J Biosci Bioeng 113(2):242–248. doi:10. 1016/j.jbiosc.2011.09.022 van Dussen L, Zimran A, Akkerman EM, Aerts JM, Petakov M, Elstein D, Rosenbaum H, Aviezer D, Brill-Almon E, Chertkoff R, Maas M, Hollak CE (2013) Taliglucerase alfa leads to favorable bone marrow responses in patients with type I Gaucher disease. Blood Cells Mol Dis 50(3):206–211. doi:10.1016/j. bcmd.2012.11.001
Vasilev N, Gro¨mping U, Lipperts A, Raven N, Fischer R, Schillberg S (2013) Optimization of BY-2 cell suspension culture medium for the production of a human antibody using a combination of fractional factorial designs and the response surface method. Plant Biotechnol J 11(7):867–874. doi:10. 1111/pbi.12079 Xu J, Ge X, Dolan MC (2011) Towards high-yield production of pharmaceutical proteins with plant cell suspension cultures. Biotechnol Adv 29(3):278–299. doi:10.1016/j.biotechadv.2011. 01.002
123