Enzyme and Microbial Technology 64–65 (2014) 44–51
Contents lists available at ScienceDirect
Enzyme and Microbial Technology journal homepage: www.elsevier.com/locate/emt
Production of monoclonal antibodies for breast cancer by HB8696 hybridoma cells using novel perfusion system Shweta Kamthan a , James Gomes b,1 , Pradip K. Roychoudhury a,∗ a b
Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India Kusuma School of Biological Sciences, Indian Institute of Technology, Hauz Khas, New Delhi 110016, India
a r t i c l e
i n f o
Article history: Received 7 March 2014 Received in revised form 24 May 2014 Accepted 7 July 2014 Available online 14 July 2014 Keywords: Breast cancer Monoclonal antibody production Silk spinfilter Stainless-steel spinfilter Filter clogging Extended perfusion operation
a b s t r a c t Perfusion culture using spinfilters have been used for the production of health-care products using mammalian cells culture. However, available spinfilters are either highly prone to clog and/or are disposable and hence affects product formation. To address these problems, a novel non-woven Bombyx mori silk screen based spinfilter module for clog-free extended perfusion culture of hybridoma cells has been designed. The module is versatile in nature and reusable, after autoclaving and replacement of used polymeric membrane. Its application for clog-free extended perfusion culture was demonstrated by comparative perfusion experiments of HB8696 cells with stainless-steel spinfilter. HB8696 cells produce monoclonal antibodies (MAbs) 520C9 active against breast cancer oncoprotein. Silk spinfilter was found to be less prone to clog with cells and debris owing to its negatively charged hydrophobic screen compared to the positively charged hydrophilic stainless-steel spinfilter. Therefore, it provides extended cell growth phase and production phase of up to 56 h and 40 h respectively and 57.4% increase in MAb productivity compared to the stainless-steel spinfilter. The effect of different perfusion rates on MAb production was studied and an optimal MAb productivity of 1.6 g L−1 day−1 was achieved. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Spinfilters have been primarily used for the perfusion culture of hybridoma cells due to their selective cell retention efficiency and low shear stress on cells [1–8]. But the primary drawbacks associated with existing spinfilter systems are either high filter clogging and/or disposability. For instance, positively charged hydrophilic stainless-steel spinfilters are highly prone to clog with negatively charged cells and cell debris [9–12] which lowers process productivity. To circumvent this problem, many strategies have been previously reported, such as the addition of deoxyribonuclease I to the culture medium [11], variation of perfusion rate, screen pore size and rotational speed of the spinfilter [13,14], application of neutral and hydrophobic filter screens [8,9,15] and use of ultrasonic
Abbreviations: MAb, monoclonal antibody; SS 316, stainless steel 316; DMEM, Dulbecco’s modified Eagle’s medium; FBS, foetal bovine serum; LDH, lactate dehydrogenase; IgG1 , immunoglobulin G subclass 1. ∗ Corresponding author at: Block 1, Room No. I-138, Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India. Tel.: +91 11 2659 1011; fax: +91 11 26582282. E-mail addresses: [email protected] (J. Gomes), [email protected] (P.K. Roychoudhury). 1 Tel.: +91 11 26591013. http://dx.doi.org/10.1016/j.enzmictec.2014.07.002 0141-0229/© 2014 Elsevier Inc. All rights reserved.
vibrations [7], etc. Considering the fact that neutral or negatively charged hydrophobic polymeric filter screens are less prone to clog, polymeric spinfilters such as ethylene-tetrafluoroethylene (ETFE), polyamide and polytetrafluoroethylene (PTFE) screen based spinfilter [9,10], polyester screen based spinfilter P [16], etc. were also designed. However, most of these spinfilters are disposable in nature. Therefore, for improving the process productivity during perfusion culture of hybridoma cells, a novel perfusion system has been designed. The biocompatibility of silk to mammalian cells [17] has been used as a selective advantage in designing a novel non-woven Bombyx mori silk membrane based reusable spinfilter module for clog-free extended perfusion operation of hybridoma cells. Since breast cancer is one of the most widespread invasive cancers worldwide and accounts for approximately 29% of all cancers in women [18], hybridoma cells produces MAb against breast cancer were used in this study as a model system. Comparative perfusion experiments of HB8696 hybridoma cells that produce MAb 520C9 active against breast cancer oncoprotein C-erbB2, were performed using stainless-steel spinfilter to demonstrate the applicability of silk spinfilter for clog-free extended perfusion operation. To enhance the MAb productivity, the perfusion rate for high MAb productivity by HB8696 hybridoma cells was also optimized while keeping the spinfilter rotation speed below the critical shear stress of cells. The
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
45
variation in perfusion rate was chosen as the primary parameter for improving process productivity because increase in perfusion rate reduces nutrient depletion and waste product accumulation that maximizes cell growth and productivity [19,20]. Nevertheless, due to cell wash out and high shear stress on cells, at high perfusion rates [20], this parameter needs to be optimized. 2. Materials and methods 2.1. Cell line and cultivation medium Two cell lines: the non-adherent mouse–mouse hybridoma cell line HB8696 (obtained from ATCC, LGC Promochem Ltd, Bangalore), which produces monoclonal antibody immunoglobulin G subclass 1 (IgG1) active against breast cancer oncoprotein C-erbB2 and the adherent human fibrosarcoma kidney cell line HT1080, obtained from NCCS, Pune) were used in this study. HB8696 cells were grown in Hybricare medium (ATCC, 46-X) and HT1080 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA). Both the mediums were supplemented with Penicillin–Streptomycin solution (5%, v/v; 100 IU/mL and 0.1 mg/mL, respectively; Sigma–Aldrich Corp., MO) and heat inactivated foetal bovine serum (FBS; 20%, v/v; Biological Industries, Israel). Cells were thawed from frozen vials and grown in 25 cm2 tissue culture flasks (Corning, USA) in a humidified 5% CO2 incubator controlled at 37 ◦ C. 2.2. Design of reusable spinfilter module The reusable spinfilter module for perfusion culture of hybridoma cells was constructed using stainless-steel (SS) 316 and Teflon (Indian patent application number 2509/DEL/2012). Its cylindrical polymeric membrane supporting component was made up of Teflon with fine holes drilled in it that permits continuous media exchange during perfusion culture. It has two detachable silicone rubber O-rings and a SS-316 clamp to hold the polymeric filter membrane in place. The cylindrical component was attached to a SS-316 base, having an SS-316 Allen screw attached to it, which helps in its positioning at desired depth on the motor shaft of the bioreactor (Fig. 1a and b). All the materials used in its construction were biocompatible to mammalian cells, impervious to cell culture media constituents, chemical and corrosionresistant, and withstand high temperature up to 121 ◦ C. Thus the designed spinfilter module is autoclavable and can be easily cleaned. The designed spinfilter module is reusable, as after completion of one perfusion run the same module can be used again after autoclaving and replacement of used polymeric membrane with the new one. The designed spinfilter module is versatile in operation, as any biocompatible polymeric membrane, selected according to the type and size of cells cultured, can be mounted over it as a filter screen. The module is presently designed for a 2-L bioreactor system, but it is amenable to scale-up to a higher volume (the data will be communicated shortly). 2.3. Characterization of silk membrane as filter screen For selecting a non-woven Bombyx mori silk membrane as filter screen for retention of hybridoma cells, the following surface properties (wettability, surface charge density and pore size) of three different non-woven Bombyx mori silks: Seri-DSS, Seri-FSS and BF27PV (obtained from Sericare, Bangalore) and their effects on cell adhesion and proliferation were evaluated: 2.3.1. Wettability Wettability of both surfaces of different silk membranes was evaluated by measuring their mean water contact angle value using contact angle goniometer (DSA100, Kruss GmbH, Germany). For each measurement, droplet of 3 L of Milli-Q water was placed on 3 cm × 3 cm strip of each membrane and the contact angle value was measured. Three measurements were done for each side of the membrane and the mean value was calculated 2.3.2. Surface charge density Surface charge density of each silk membrane was analyzed by measuring their zeta potential value using cylindrical cell of electrokinetic analyzer with Ag/AgCl disc electrodes (Anton Paar GmbH, Graz, Austria). For each measurement, an irregular plug of each silk membrane was placed between the two electrodes and the measuring fluid (0.001 mol/L KCl solution) was passed through them. For all the silk membranes, three zeta potential values were measured and the mean value was calculated. The pH was maintained by addition of 0.1 mol/L KOH or 0.1 mol/L HCl solution. 2.3.3. Pore size Pore size of each silk membrane was measured in triplicates using their three scanning electron micrographs and IMAGE-J software (National Institutes of Health, USA, version 1.41). The images were captured using three strips of each silk membrane (5 mm × 5 mm each) by scanning electron microscope (Evo50, Zeiss, UK).
Fig. 1. (a) 3D image of silk spinfilter generated by Autodesk 3ds max software. (A) Porous teflon membrane support, (B) shaft holding cone of stainless steel with Allen screw, (C) metallic clamp for fixing the membrane, (D) top and bottom view of spinfilter module, (E) top and bottom view of spinfilter module with silk. (b) Dimensions of different components of silk spinfilter. The components clockwise from the top left panel are porous teflon membrane support, stainless steel shaft holding cone, Allen screw for fixing the module and stainless steel clamp for holding the membrane. All the dimensions are in centimetres. 2.3.4. Cell attachment supporting nature of silk membranes The mammalian cell attachment supporting nature of each silk membrane (Seri-DSS, Seri-FSS and BF27PV taken separately) was tested by culturing adherent HT1080 cells on them in 24-well plates (Corning, USA). Growth medium used was DMEM supplemented with 20% heat inactivated FBS and 5% penicillin–streptomycin solution. For each of the three membranes, two 24-well plates were used. The two other 24-well plates without membrane were used as control. Together, these eight 24-well plates constituted a set for one experiment. Three sets of these experiments were performed. In each of the wells, 1 cm diameter circular piece of membrane was taken. The seeding density of 1.4 × 105 cells/mL was used for each well and incubated in a humidified 5% CO2 incubator at 37 ◦ C. After every 8 h, samples were drawn from three wells of each membrane-type and control plates by trypsinization using 0.25% trypsin (Gibco, USA). At each sampling instant viable cell densities were measured in triplicates. Along with this, for visualization of attached adherent HT1080 cells on different silk membranes, their scanning electron micrographs were also captured after 72 h using the Evo50 scanning electron microscope (Zeiss, UK). 2.3.5. Cell proliferation supporting nature of silk membranes Proliferation of mammalian cells in the presence of three different silk membranes (Seri-DSS, Seri-FSS and BF27PV taken separately) was tested by culturing non-adherent HB8696 hybridoma cells on them in 24-well plates in exactly the same manner as described in Section 2.3.4. Growth medium used was Hybricare 46-X medium supplemented with 20% heat inactivated FBS and 5% penicillin–streptomycin solution. The seeding density of 1.4 × 105 cells/mL was used
46
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 2. Water contact angle measurement of Seri-DSS silk membrane by contact angle goniometer. (A) Contact angle of upper side of membrane. (B) Contact angle of lower side of membrane.
for each well. All the plates were incubated in 5% CO2 incubator at 37 ◦ C. Here also, three sets of these experiments were performed and after every 8 h sampling was done from three wells of each membrane-type and control plates by collection of suspended medium from them. At each sampling instant, viable cell densities were measured in triplicates. 2.4. Determination of critical shear stress of HB8696 hybridoma cells For operation of bioreactor below critical shear stress of HB8696 hybridoma cells, the optimum spinfilter rotation speed was determined from the batch experiments performed in triplicates in Biostat B.Braun bioreactor. Growth medium used was Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. The initial spinfilter rotation speed was set at 10 rpm and then stepwise increase in spinfilter rotation speed was done after every 5 h by a factor of 5 rpm. The pH inside the bioreactor was maintained at 7.4 with the combined addition of 0.1 M sodium bicarbonate (NaHCO3 ; Merck, Germany) and introducing 5% CO2 into the headspace of the reactor. The cultivation temperature was maintained at 37 ◦ C. Pluronic F-68 (0.1%, w/v, Sigma–Aldrich Corp., MO) was added to the medium as antifoaming agent, when required. The dissolve oxygen (DO) concentration was maintained at 50% of air saturation with the control module of the reactor by adjusting the air/O2 /N2 ratio of the inlet gas. Working volume of the bioreactor was 750 mL. The seeding density used for each experiment was 3.0 × 105 cells/mL. Sampling was done after every 5 h and at each sampling instant viable cell density and lactate dehydrogenase (LDH) concentration inside the bioreactor [21] was measured in triplicates to study shear effect on the cells.
for high MAb productivity was determined by continuous perfusion experiments of HB8696 cells in triplicates in perfusion bioreactor (Biostat, B.Braun, Germany) at four different perfusion rates of 0.625 mL min−1 , 0.75 mL min−1 , 1.125 mL min−1 and 1.5 mL min−1 under the same set of conditions as mentioned in Section 2.6. Perfusion operation was started from 40 h, using two peristaltic pumps operated at equal flow rates. The growth medium used was Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. Working volume of reactor was 750 mL. The seeding density of 3.2 × 105 cells/mL was used in each experiment. Each set of experiment was performed in triplicates. Samples were drawn after every 8 h, and at each sampling instant viable cell densities and MAb concentrations were measured in triplicates. 2.8. Analytical methods Cell density (viable and dead) was measured by trypan blue assay using a haemocytometer. Glucose concentration was measured by Glucose kit (Biolab diagnostics). Glutamine concentration was measured by enzymatic assay [22]. Lactate concentration was measured by d-lactate colorimetric assay kit (Catalogue.K667100, Bio-vision, USA). Ammonia concentration was measured by Nessler’s reagent [23]. LDH was measured by LDH cytotoxicity assay kit (Item no.10008882, Cayman chemical company, USA). MAb content was analyzed by Indirect ELISA and concentrations were calculated by the standard curve of purified IgG1 . MAb was purified from the culture supernatant by protein-G affinity column chromatography using IgG purification kit (Bangalore Genie, India). 2.9. Statistical analysis
2.5. Growth profile of HB8696 hybridoma cells in batch culture The growth profile of HB8696 hybridoma cells was studied by performing their batch culture experiments in triplicates in Biostat B.Braun bioreactor using the Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. Operating conditions were pH 7.2, temperature 37 ◦ C and DO level 50%. Antifoaming agent was added, when required. The seeding density used was 3 × 105 cells/mL. Spinfilter rotation speed was set at 45 rpm (optimized value). Samples were drawn after every 8 h and at each sampling instant, the measurement of each variable (viable cell density, total cell density, percentage of viable and dead cells, and glucose, glutamine, lactate, ammonium, LDH and MAb concentrations) was carried out in triplicates. 2.6. Comparative perfusion experiments in silk spinfilter and stainless-steel spinfilter Comparative continuous perfusion experiments of HB8696 hybridoma cells were performed in perfusion bioreactor (Biostat, B.Braun, Germany) in triplicates using Stainless-steel spinfilter (pore size 20 m) and silk spinfilter. The operating conditions were maintained at pH of 7.2, cultivation temperature of 37 ◦ C and DO concentration at 50% of air saturation. Antifoaming agent, Pluronic F-68 was used, when required. Working volume of reactor was 750 mL. The seeding density used was 3.2 × 105 cells/mL. Spinfilter rotation speed was set at 45 rpm (optimized value). Perfusion operation was started from 40 h, when viable cell density was started to decrease in batch culture, using two peristaltic pumps at equal perfusion rate of 0.75 mL min−1 . Sampling during these experiments was done after every 8 h and at each sampling instant viable cell density, total cell density and MAb production were measured as the primary parameters for evaluating the performance while percentage retention of viable and dead cells, lactate, ammonium and LDH accumulation inside the bioreactor were measured to correlate the changes in the medium affecting viability and productivity. At each sampling instant all of these parameters were measured in triplicates. 2.7. Perfusion experiments at different perfusion rates To enhance the MAb productivity from hybridoma cells, the performance of silk spinfilter at different perfusion rates was studied. The spinfilter rotation speed was kept constant at an optimized value of 45 rpm. The optimized perfusion rate
All the data has been presented as the mean of three independent series of experiments, each with n = 3 per group, mean ± standard deviation. Analysis of variance (ANOVA) was performed using Stat-Ease Inc. software (USA).
3. Results and discussions In this section, the characterization of silk as spinfilter screen for clog-free extended perfusion operation of hybridoma cells and optimized perfusion strategy for enhancing MAb productivity against breast cancer using novel silk spinfilter are described as follows. 3.1. Characterization of non-woven Bombyx mori silk as spinfilter screen For selection of suitable silk as filter screen of the reusable spinfilter module, the surface properties of three different non-woven Bombyx mori silks, Seri-DSS, Seri-FSS and BF27PV and their effect on cell adhesion and proliferation were studied. It was reported that neutral or negatively charged hydrophobic membranes having pore size ranges between 10 and 20 m were less prone to clog compared to the positively charged hydrophilic membranes [9,10,13–16]. Thus, among the three different silks, the most negatively charged and hydrophobic one was selected as the suitable spinfilter screen. A comparison of surface properties (Table 1) showed that SeriDSS was the most hydrophobic (water contact angle: 107.7 ± 1.1◦ for the upper side and 102 ± 2.7◦ for lower side, Fig. 2) and negatively charged (zeta potential value −38 ± 0.5 mV) silk. Therefore, Seri-DSS silk did not support the attachment and growth of adherent HT1080 cells on its surface over the entire culture period
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
47
Table 1 Comparison of properties of Seri-FSS, Seri-DSS and BF27PV silk membranes. Properties
Seri-DSS silk
Seri-FSS silk
BF27PV silk
Source Cell adhesion Cell proliferation
Bombyx mori Does not support cell adhesion Maximum viable cell density of 1.1 ± 0.1 × 106 cells/mL achieved at 40 h 14.9 ± 1.8 m Negative −38 ± 0.5 mV Both sides hydrophobic Upper side 107.7 ± 1.1◦ Lower side 102 ± 2.7◦
Bombyx mori Supports cell adhesion Maximum viable cell density of 1.0 ± 0.1 × 106 cells/mL achieved at 40 h 15.4 ± 0.8 m Positive 2.0 ± 0.2 mV One side hydrophobic 101.5 ± 2.1◦ Other side hydrophilic 15.3 ± 1.7◦
Bombyx mori Support cell adhesion Maximum viable cell density of 1.0 ± 0.1 × 106 cells/mL achieved at 40 h 14.5 ± 1.1 m Positive 7.9 ± 0.2 mV Both sides hydrophilic Upper side 11.7 ± 1.0◦ Lower side 11.1 ± 1.3◦
Pore size Surface charge density Wettability
Fig. 3. Time profile of viable cell density (×106 cells/mL) of adherent HT1080 cells in control wells and Seri-DSS, Seri-FSS, Bf27PV silks containing wells of 24-well plates.
of 120 h, as shown in Fig. 3. Similarly, in the scanning electron micrographs also no cells were found to be attached on the SeriDSS silk after 72 h (Fig. 4). To rule out if non-attachment and non-proliferation was caused by the Seri-DSS silk, non-adherent HB8696 cells were also cultured over all the silk membranes. It was observed that the presence of circular pieces of silk membranes had no detrimental effect on cell proliferation compared to the control (Fig. 5). Therefore, it was concluded that Seri-DSS silk neither supported cell attachment nor adversely affected cell proliferation. Further, its average pore diameter of 14.9 ± 1.8 m (Fig. 6) was also between 10 and 20 m, suitable for effective retention of hybridoma cells. Due to all of these features, Seri-DSS silk was used as the filter screen of the designed spinfilter module.
Fig. 5. Time profile of viable cell density (×106 cells/mL) of non-adherent HB8696 cells in control wells and Seri-DSS, Seri-FSS, BF27PV silks containing wells of 24-well plates.
3.2. Batch culture for determination of critical shear stress of HB8696 hybridoma cells Batch culture of HB8696 hybridoma cells were performed in bioreactor at different spinfilter rotation speeds to determine the critical shear stress. From the experiments (Fig. 7), it was found that initially with an stepwise increase in spinfilter rotation speed from 10 to 45 rpm, by a factor of 5 after every 5 h, the viable cell density continuously increase up to 40 h and at 40 h, the maximum viable cell density of 1.5 ± 0.1 × 106 cells/mL was achieved. But after 40 h, due to further increase in spinfilter rotation speed above 45 rpm, viable cell density started to decreased and at 50 h, the decreased viable cell density of 0.6 ± 0.1 × 106 cells/mL was achieved. This shows that spinfilter rotation speed above 45 rpm is detrimental for
Fig. 4. Scanning electron micrographs of silk membranes after 72 h of adherent HT-1080 cells culture. (a) Seri-DSS (b) Seri-FSS and (c) BF27PV silk.
48
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 6. Pore size distribution of Seri-DSS silk membrane. (A–C) Scanning electron micrographs of three Seri-DSS silk membrane. (D) Pore size distribution curve of image A, B and C measured by Image-J software.
HB8696 cells and above this, cell lysis occurred. Therefore after 40 h, a sharp increase in LDH concentration up to 520 ± 18.6 U/mL was obtained. Therefore, for all the further batch and perfusion experiments of HB8696 cells, the spinfilter rotation speed of 45 rpm was used. 3.3. Growth profile of HB8696 hybridoma cells in batch culture
Fig. 7. Effect of different spinfilter rotation speeds on HB8696 hybridoma cells during batch culture in bioreactor. Here vcd is viable cell density (×106 cells/mL) and LDH is lactate dehydrogenase accumulation (U/mL).
The growth profile of HB8696 hybridoma cells during batch culture was studied (Fig. 8). The initial concentration of glucose and glutamine were 5.3 mM and 5.4 mM respectively. From the figures, it was found that viable cell density was increased up to 40 h and at 40 h; the maximum viable cell density of 1.5 ± 0.1 × 106 cells/mL was achieved. Beyond this, viable cell density started to decrease gradually due to the accumulation of lactate concentration above 23.3 ± 0.8 mM and ammonium concentration above 5.4 ± 0.3 mM, as observed in all the three batch experiments [20,24–26]. However, MAb production was continued and the maximum MAb concentration of 286.2 ± 41.3 mg/L was achieved at 96 h. Beyond 96 h, due to further accumulation
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 8. Time profiles of cell growth and product formation of non-adherent HB8696 cells during batch culture in Biostat B.Braun bioreactor. (A) States characterizing the growth of HB8696 cells. Here vcd is viable cell density (×106 cells/mL); tcd is total cell density (×106 cells/mL); %v is cell viability (%). (B) States characterizing the nutrient consumption and waste products accumulation by HB8696 cells. Here Glu is glucose concentration (mM); Gln is glutamine concentration (mM); Lac is lactate concentration (mM); Amm is ammonium concentration (mM). (C) States characterizing product formation of HB8696 cells. Here MAb is monoclonal antibody concentration (mg/L); %v is percentage of viable cells; %d is percentage of dead cells; LDH is lactate dehydrogenase concentration (U/mL).
of lactate above 30 ± 2.4 mM and ammonium above 9.2 ± 0.2 mM, a sharp decrease in viable cell density occurred marked with an increase in LDH concentration and commensurate decrease in MAb concentration.
49
Fig. 9. Comparative time profiles of cell growth and product formation of nonadherent HB8696 cells during perfusion culture in bioreactor using stainless steel (SS) and silk spinfilter. (A) States characterizing the growth of HB8696 cells. Here vcd is viable cell density (×107 cells/mL); tcd is total cell density (×107 cells/mL); %v is cell viability (%). (B) States characterizing waste products accumulation by HB8696 cells. Here LC is lactate concentration (mM); AC is ammonium concentration (mM). (C) States characterizing product formation of HB8696 cells. Here MAb is monoclonal antibody concentration (mg/L); %v is percentage of viable cells; %d is percentage of dead cells; LDH is lactate dehydrogenase concentration (U/mL).
From these results, it was concluded that to increase the MAb productivity, perfusion operation should be started from 40 h, so that the onset of inhibitory concentrations of lactate and ammonium could be delayed and the duration of cell growth phase and production phase extended.
50
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Table 2 Comparative viable cell densities and monoclonal antibody productivities of HB8696 cells during batch culture in bioreactor and perfusion culture using stainless steel spinfilter and silk spinfilter. HB8696 hybridoma cells Parameters
Time (h)
Bioreactor Batch
Stainless steel spinfilter Perfusion
Silk spinfilter
Viable cell density (107 cells/mL)
40 96 152
0.15 ± 0.1 0.07 ± 0.1 − 0.53 2.6 57.4%
0.5 ± 0.1 1.3 ± 0.1 1.1 ± 0.1 0.94 9.4
0.9 ± 0.1 2.1 ± 0.1 2.5 ± 0.1 1.48 14.8
Monoclonal antibody productivity (g L−1 day−1 ) Total monoclonal antibody produced (g) Increase in productivity from stainless steel to silk Values shown in bold are the maximum viable cell density achieved in each case.
Fig. 10. Time profiles of cell growth and product formation of non-adherent HB8696 hybridoma cells at different perfusion rates and constant spinfilter rotation speed of 45 rpm during perfusion culture using silk spinfilter. (A) Viable cell density (×107 cells/mL). (B) Monoclonal antibody concentration (mg/L).
3.4. Comparative study in silk spinfilter and stainless-steel spinfilter This comparative study demonstrated the application of silk spinfilter for clog-free extended perfusion operation of hybridoma cells (Fig. 9). It is clear from the figure that silk spinfilter provides the extended cell growth phase of 152 h and extended MAb production phase of 192 h, compared to the cell growth phase of 96 h and MAb production phase of 152 h achieved using stainless-steel spinfilter. Using silk spinfilter, a maximum viable cell density of 2.5 ± 0.1 × 107 cells/mL was achieved at 152 h. After this, the viable cell density start to decrease gradually due to the accumulation of lactate above 23.5 mM and ammonium above 5.4 mM, as observed at 40 h during batch culture of HB8696 cells. However, the production of MAb was still continued and at 192 h, the maximum MAb concentration of 723.9 ± 70.5 mg/L was achieved. The overall MAb productivity was 1.48 g L−1 day−1 . In comparison, when the Stainless-steel spinfilter was used, the first inhibitory level of lactate (23.5 mM) and ammonium (5.4 mM)
reached 56 h prior than the silk spinfilter and thus the maximum viable cell density of only 1.3 ± 0.1 × 107 cells/mL was achieved. Similarly, the second inhibitory level of lactate (>30 mM) and ammonium (>9.2 mM) crossed 40 h earlier as compared to the silk spinfilter, therefore the MAb concentration as well as MAb productivity of only 528.6 ± 24.2 mg/L and 0.94 g L−1 day−1 respectively was achieved. Therefore, it can be concluded that silk spinfilter performs much better than the Stainless-steel spinfilter (Table 2) and provides 57.4% increase in MAb productivity. This is due to the lower likelihood of hydrophobic and negatively charged Seri-DSS silk screen of the silk spinfilter to clog with negatively charged cells and cell debris as compared to the hydrophilic and positively charged stainless steel screen of stainless steel spinfilter. Due to the lower clogging tendency of silk spinfilter, while using it during the perfusion culture, the accumulation of inhibitory concentrations of toxic metabolites (lactate and ammonium) inside the bioreactor was delayed and thus the healthy environment for cell growth was maintained for longer period of time [20–26]. Therefore it provides extended duration of cell growth and
Table 3 Comparative maximum viable cell densities and maximum monoclonal antibody concentrations achieved during perfusion culture of HB8696 hybridoma cells using silk spinfilter at different perfusion rates.
a b
S. no.
Perfusion rate (mL min−1 )
Rotation speed (rpm)
Maximum viable cell density1 (107 cells/mL)
1 2 3 4
0.625 0.75 1.125 1.5
45 45 45 45
2.3 2.5 3.5 2.1
± ± ± ±
0.1 0.1 0.1 0.0
Time2 (h)
Maximum monoclonal antibody concentration3 (mg/L)
128 152 176 120
454 723.9 793 432
The maximum viable cell density ‘1’ (mean ± standard deviation) occurs at time ‘2’. The maximum monoclonal antibody concentration ‘3’ (mean ± standard deviation) occurs at time ‘4’.
± ± ± ±
25.5 70.5 22.2 22.2
Time4 (h) 152 192 200 152
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
production phase and enhanced the MAb productivity as compared to the conventional Stainless-steel spinfilter. 3.5. Effect of perfusion rates on monoclonal antibody productivity using silk spinfilter The study was carried out in a bioreactor at four different perfusion rates (0.625 mL min−1 , 0.75 mL min−1 , 1.125 mL min−1 and 1.5 mL min−1 ) using silk spinfilter for each set of experiment, sampling was carried out after every 8 h and the state-time profile of viable cell density and MAb concentration was constructed (Fig. 10). From the maximum value of the viable cell density and MAb concentration achieved in each case (Table 3), it was found that at a perfusion rate of 1.125 mL min−1 , the maximum viable cell density of 3.5 ± 0.1 × 107 cells/mL and maximum MAb concentration of 793 ± 22.2 mg/L was achieved. The MAb productivity achieved at these optimum operating conditions was 1.6 g L−1 day−1 . From this, it is concluded that the optimum perfusion rate for MAb production using silk spinfilter was 1.125 mL min−1 . 4. Conclusion The results show that the novel silk screen based spinfilter can be used for longer perfusion culture of hybridoma cells with enhanced MAb productivity. The negatively charged hydrophobic surface in the silk spinfilter reduces cell and debris adhesion on the filter screen and lowers its tendency to clog as compared to the positively charged hydrophilic stainless-steel spinfilter. Using silk spinfilter, 57.4% increase in overall MAb productivity, as compared to the conventional stainless-steel spinfilter was achieved. Author disclosure and contribution The idea of the research work was conceived by Prof. P.K. Roychoudhury. The experiments were designed by all the authors and performed by Shweta Kamthan. The manuscript was written jointly and approved by all the authors. Acknowledgement The authors gratefully acknowledge the funding received for this work from the Department of Biotechnology, Ministry of Science and Technology, Government of India (Grant no. BT/PR9154/PID/06/387/2007). References [1] Carter JB, Shevitz J. A brief history of perfusion biomanufacturing. BioProcess Int 2011;9(9):24–30.
51
[2] Vogel JH, Nguyen H, Giovannini R, Ignowski J, Garger S, Salgotra A, et al. A new large-scale manufacturing platform for complex biopharmaceuticals. Biotechnol Bioeng 2012;109(12):3049–58. [3] Warnock JN, Rubeai MA. Bioreactor systems for the production of biopharmaceuticals from animal cells. Biotechnol Appl Biochem 2006;45:1–12. [4] Warikoo V, Godawat R, Brower K, Jain S, Cummings D, Simons E, et al. Integrated continuous production of recombinant therapeutic proteins. Biotechnol Bioeng 2012;109(12):3018–29. [5] Castilho LR, Medronho RA. Cell retention devices for suspended cell perfusion cultures. Adv Biochem Eng Biotechnol 2002;74:129–69. [6] Kompala DS, Ozturk SS. Optimization of high cell density perfusion bioreactors. In: Ozturk SS, Hu WS, editors. Cell culture technology for pharmaceutical and cell-based therapies. New York: Taylor & Francis Group; 2006. p. 155–224. [7] Chetreanu FV, Ferreira LGF, Rabe R, Stockar UV, Marison IW. An on-line method for the reduction of fouling of spin-filters for animal cell perfusion cultures. J Biotechnol 2007;130:265–73. [8] Whitford WG. Single-use systems as principal components in bio-production. BioProcess Int 2010;8(11):34–42. [9] Avgerinos GC, Drapeau D, Socolow JS, Mao JI, Hsiao K, Broeze RJ. Spin filter perfusion system for high density cell culture: production of recombinant urinary type plasminogen activator in CHO cells. BioTechnology 1990;8:54–8. [10] Esclade LRJ, Carrel S, Péringer P. Influence of the screen material on the fouling of the spin filters. Biotechnol Bioeng 1991;38:159–68. [11] Mercille S, Johnson M, Lemieux R, Massie B. Filtration-based perfusion of hybridoma cultures in protein-free medium: reduction of membrane fouling by medium supplementation with DNase I. Biotechnol Bioeng 1994;43:833–46. [12] Ozturk SS. Comparison of product quality: disposable and stainless steel bioreactor. Berlin: BioProduction; 2007. [13] Deo YM, Mahadevan MD, Fuchs R. Practical considerations in operation and scale-up of spin-filter based bioreactors for monoclonal antibody production. Biotechnol Prog 1996;12:57–64. [14] Yabannavar VM, Singh V, Connelly NV. Mammalian cell retention in a spinfilter perfusion bioreactor. Biotechnol Bioeng 1992;40:925–33. [15] Nermen M, Nakhla G, Wan W. Comparative assessment of hydrophobic and hydrophilic membrane fouling in waste water applications. J Membr Sci 2009;339:93–9. [16] Sartorius. Spinfilter P. An innovative disposable spinfilter system for cell culture perfusion. Pharm Online Newsl 2005. [17] Altman GH, Diaz F, Jakuba C, Calabro T, Horan RL, Chen J, et al. Silk based biomaterials. Biomaterials 2003;24:401–16. [18] Stacy S. Cancer statistics report: deaths down 20% in two decades. Cancer Stat 2014;7(January). [19] Banik GG, Heath CA. Hybridoma growth and antibody production as a function of cell density and specific growth rate in perfusion culture. Biotechnol Bioeng 1995;48(3):289–300. [20] Dong H, Tang YJ, Ohashi R, Hamel JFP. A perfusion culture system using a stirred ceramic membrane reactor for hyperproduction of IgG (2a) monoclonal antibody by hybridoma cells. Biotechnol Prog 2005;21(1):140–7. [21] Ludwig A, Kretzmer G, Schugerl K. Determination of a critical shear stress level applied to adherent mammalian cells. Enzyme Microb Technol 1992;14:209–13. [22] Sowerby JM, Ottaway JH. The enzymic estimation of glutamate and glutamine. Biochem J 1966;99:246–52. [23] Morrison GR. Micro chemical determination of organic nitrogen with Nessler reagent. Anal Biochem 1971;43:527–32. [24] Krampe B, Al-Rubeai M. Cell death in mammalian cell culture: molecular mechanisms and cell line engineering strategies. Cytotechnology 2010;62(3):175–88. [25] Ozturk SS, Riley MR, Palsson BO. Effects of ammonia and lactate on hybridoma growth, metabolism and antibody production. Biotechnol Bioeng 1992;39:418–31. [26] Ozturk SS, Thrift JC, Blackie JD, Naveh D. Real-time monitoring and control of glucose and lactate concentrations in a mammalian cell perfusion reactor. Biotechnol Bioeng 1997;53(4):372–8.
Contents lists available at ScienceDirect
Enzyme and Microbial Technology journal homepage: www.elsevier.com/locate/emt
Production of monoclonal antibodies for breast cancer by HB8696 hybridoma cells using novel perfusion system Shweta Kamthan a , James Gomes b,1 , Pradip K. Roychoudhury a,∗ a b
Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India Kusuma School of Biological Sciences, Indian Institute of Technology, Hauz Khas, New Delhi 110016, India
a r t i c l e
i n f o
Article history: Received 7 March 2014 Received in revised form 24 May 2014 Accepted 7 July 2014 Available online 14 July 2014 Keywords: Breast cancer Monoclonal antibody production Silk spinfilter Stainless-steel spinfilter Filter clogging Extended perfusion operation
a b s t r a c t Perfusion culture using spinfilters have been used for the production of health-care products using mammalian cells culture. However, available spinfilters are either highly prone to clog and/or are disposable and hence affects product formation. To address these problems, a novel non-woven Bombyx mori silk screen based spinfilter module for clog-free extended perfusion culture of hybridoma cells has been designed. The module is versatile in nature and reusable, after autoclaving and replacement of used polymeric membrane. Its application for clog-free extended perfusion culture was demonstrated by comparative perfusion experiments of HB8696 cells with stainless-steel spinfilter. HB8696 cells produce monoclonal antibodies (MAbs) 520C9 active against breast cancer oncoprotein. Silk spinfilter was found to be less prone to clog with cells and debris owing to its negatively charged hydrophobic screen compared to the positively charged hydrophilic stainless-steel spinfilter. Therefore, it provides extended cell growth phase and production phase of up to 56 h and 40 h respectively and 57.4% increase in MAb productivity compared to the stainless-steel spinfilter. The effect of different perfusion rates on MAb production was studied and an optimal MAb productivity of 1.6 g L−1 day−1 was achieved. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Spinfilters have been primarily used for the perfusion culture of hybridoma cells due to their selective cell retention efficiency and low shear stress on cells [1–8]. But the primary drawbacks associated with existing spinfilter systems are either high filter clogging and/or disposability. For instance, positively charged hydrophilic stainless-steel spinfilters are highly prone to clog with negatively charged cells and cell debris [9–12] which lowers process productivity. To circumvent this problem, many strategies have been previously reported, such as the addition of deoxyribonuclease I to the culture medium [11], variation of perfusion rate, screen pore size and rotational speed of the spinfilter [13,14], application of neutral and hydrophobic filter screens [8,9,15] and use of ultrasonic
Abbreviations: MAb, monoclonal antibody; SS 316, stainless steel 316; DMEM, Dulbecco’s modified Eagle’s medium; FBS, foetal bovine serum; LDH, lactate dehydrogenase; IgG1 , immunoglobulin G subclass 1. ∗ Corresponding author at: Block 1, Room No. I-138, Department of Biochemical Engineering and Biotechnology, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India. Tel.: +91 11 2659 1011; fax: +91 11 26582282. E-mail addresses: [email protected] (J. Gomes), [email protected] (P.K. Roychoudhury). 1 Tel.: +91 11 26591013. http://dx.doi.org/10.1016/j.enzmictec.2014.07.002 0141-0229/© 2014 Elsevier Inc. All rights reserved.
vibrations [7], etc. Considering the fact that neutral or negatively charged hydrophobic polymeric filter screens are less prone to clog, polymeric spinfilters such as ethylene-tetrafluoroethylene (ETFE), polyamide and polytetrafluoroethylene (PTFE) screen based spinfilter [9,10], polyester screen based spinfilter P [16], etc. were also designed. However, most of these spinfilters are disposable in nature. Therefore, for improving the process productivity during perfusion culture of hybridoma cells, a novel perfusion system has been designed. The biocompatibility of silk to mammalian cells [17] has been used as a selective advantage in designing a novel non-woven Bombyx mori silk membrane based reusable spinfilter module for clog-free extended perfusion operation of hybridoma cells. Since breast cancer is one of the most widespread invasive cancers worldwide and accounts for approximately 29% of all cancers in women [18], hybridoma cells produces MAb against breast cancer were used in this study as a model system. Comparative perfusion experiments of HB8696 hybridoma cells that produce MAb 520C9 active against breast cancer oncoprotein C-erbB2, were performed using stainless-steel spinfilter to demonstrate the applicability of silk spinfilter for clog-free extended perfusion operation. To enhance the MAb productivity, the perfusion rate for high MAb productivity by HB8696 hybridoma cells was also optimized while keeping the spinfilter rotation speed below the critical shear stress of cells. The
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
45
variation in perfusion rate was chosen as the primary parameter for improving process productivity because increase in perfusion rate reduces nutrient depletion and waste product accumulation that maximizes cell growth and productivity [19,20]. Nevertheless, due to cell wash out and high shear stress on cells, at high perfusion rates [20], this parameter needs to be optimized. 2. Materials and methods 2.1. Cell line and cultivation medium Two cell lines: the non-adherent mouse–mouse hybridoma cell line HB8696 (obtained from ATCC, LGC Promochem Ltd, Bangalore), which produces monoclonal antibody immunoglobulin G subclass 1 (IgG1) active against breast cancer oncoprotein C-erbB2 and the adherent human fibrosarcoma kidney cell line HT1080, obtained from NCCS, Pune) were used in this study. HB8696 cells were grown in Hybricare medium (ATCC, 46-X) and HT1080 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA). Both the mediums were supplemented with Penicillin–Streptomycin solution (5%, v/v; 100 IU/mL and 0.1 mg/mL, respectively; Sigma–Aldrich Corp., MO) and heat inactivated foetal bovine serum (FBS; 20%, v/v; Biological Industries, Israel). Cells were thawed from frozen vials and grown in 25 cm2 tissue culture flasks (Corning, USA) in a humidified 5% CO2 incubator controlled at 37 ◦ C. 2.2. Design of reusable spinfilter module The reusable spinfilter module for perfusion culture of hybridoma cells was constructed using stainless-steel (SS) 316 and Teflon (Indian patent application number 2509/DEL/2012). Its cylindrical polymeric membrane supporting component was made up of Teflon with fine holes drilled in it that permits continuous media exchange during perfusion culture. It has two detachable silicone rubber O-rings and a SS-316 clamp to hold the polymeric filter membrane in place. The cylindrical component was attached to a SS-316 base, having an SS-316 Allen screw attached to it, which helps in its positioning at desired depth on the motor shaft of the bioreactor (Fig. 1a and b). All the materials used in its construction were biocompatible to mammalian cells, impervious to cell culture media constituents, chemical and corrosionresistant, and withstand high temperature up to 121 ◦ C. Thus the designed spinfilter module is autoclavable and can be easily cleaned. The designed spinfilter module is reusable, as after completion of one perfusion run the same module can be used again after autoclaving and replacement of used polymeric membrane with the new one. The designed spinfilter module is versatile in operation, as any biocompatible polymeric membrane, selected according to the type and size of cells cultured, can be mounted over it as a filter screen. The module is presently designed for a 2-L bioreactor system, but it is amenable to scale-up to a higher volume (the data will be communicated shortly). 2.3. Characterization of silk membrane as filter screen For selecting a non-woven Bombyx mori silk membrane as filter screen for retention of hybridoma cells, the following surface properties (wettability, surface charge density and pore size) of three different non-woven Bombyx mori silks: Seri-DSS, Seri-FSS and BF27PV (obtained from Sericare, Bangalore) and their effects on cell adhesion and proliferation were evaluated: 2.3.1. Wettability Wettability of both surfaces of different silk membranes was evaluated by measuring their mean water contact angle value using contact angle goniometer (DSA100, Kruss GmbH, Germany). For each measurement, droplet of 3 L of Milli-Q water was placed on 3 cm × 3 cm strip of each membrane and the contact angle value was measured. Three measurements were done for each side of the membrane and the mean value was calculated 2.3.2. Surface charge density Surface charge density of each silk membrane was analyzed by measuring their zeta potential value using cylindrical cell of electrokinetic analyzer with Ag/AgCl disc electrodes (Anton Paar GmbH, Graz, Austria). For each measurement, an irregular plug of each silk membrane was placed between the two electrodes and the measuring fluid (0.001 mol/L KCl solution) was passed through them. For all the silk membranes, three zeta potential values were measured and the mean value was calculated. The pH was maintained by addition of 0.1 mol/L KOH or 0.1 mol/L HCl solution. 2.3.3. Pore size Pore size of each silk membrane was measured in triplicates using their three scanning electron micrographs and IMAGE-J software (National Institutes of Health, USA, version 1.41). The images were captured using three strips of each silk membrane (5 mm × 5 mm each) by scanning electron microscope (Evo50, Zeiss, UK).
Fig. 1. (a) 3D image of silk spinfilter generated by Autodesk 3ds max software. (A) Porous teflon membrane support, (B) shaft holding cone of stainless steel with Allen screw, (C) metallic clamp for fixing the membrane, (D) top and bottom view of spinfilter module, (E) top and bottom view of spinfilter module with silk. (b) Dimensions of different components of silk spinfilter. The components clockwise from the top left panel are porous teflon membrane support, stainless steel shaft holding cone, Allen screw for fixing the module and stainless steel clamp for holding the membrane. All the dimensions are in centimetres. 2.3.4. Cell attachment supporting nature of silk membranes The mammalian cell attachment supporting nature of each silk membrane (Seri-DSS, Seri-FSS and BF27PV taken separately) was tested by culturing adherent HT1080 cells on them in 24-well plates (Corning, USA). Growth medium used was DMEM supplemented with 20% heat inactivated FBS and 5% penicillin–streptomycin solution. For each of the three membranes, two 24-well plates were used. The two other 24-well plates without membrane were used as control. Together, these eight 24-well plates constituted a set for one experiment. Three sets of these experiments were performed. In each of the wells, 1 cm diameter circular piece of membrane was taken. The seeding density of 1.4 × 105 cells/mL was used for each well and incubated in a humidified 5% CO2 incubator at 37 ◦ C. After every 8 h, samples were drawn from three wells of each membrane-type and control plates by trypsinization using 0.25% trypsin (Gibco, USA). At each sampling instant viable cell densities were measured in triplicates. Along with this, for visualization of attached adherent HT1080 cells on different silk membranes, their scanning electron micrographs were also captured after 72 h using the Evo50 scanning electron microscope (Zeiss, UK). 2.3.5. Cell proliferation supporting nature of silk membranes Proliferation of mammalian cells in the presence of three different silk membranes (Seri-DSS, Seri-FSS and BF27PV taken separately) was tested by culturing non-adherent HB8696 hybridoma cells on them in 24-well plates in exactly the same manner as described in Section 2.3.4. Growth medium used was Hybricare 46-X medium supplemented with 20% heat inactivated FBS and 5% penicillin–streptomycin solution. The seeding density of 1.4 × 105 cells/mL was used
46
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 2. Water contact angle measurement of Seri-DSS silk membrane by contact angle goniometer. (A) Contact angle of upper side of membrane. (B) Contact angle of lower side of membrane.
for each well. All the plates were incubated in 5% CO2 incubator at 37 ◦ C. Here also, three sets of these experiments were performed and after every 8 h sampling was done from three wells of each membrane-type and control plates by collection of suspended medium from them. At each sampling instant, viable cell densities were measured in triplicates. 2.4. Determination of critical shear stress of HB8696 hybridoma cells For operation of bioreactor below critical shear stress of HB8696 hybridoma cells, the optimum spinfilter rotation speed was determined from the batch experiments performed in triplicates in Biostat B.Braun bioreactor. Growth medium used was Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. The initial spinfilter rotation speed was set at 10 rpm and then stepwise increase in spinfilter rotation speed was done after every 5 h by a factor of 5 rpm. The pH inside the bioreactor was maintained at 7.4 with the combined addition of 0.1 M sodium bicarbonate (NaHCO3 ; Merck, Germany) and introducing 5% CO2 into the headspace of the reactor. The cultivation temperature was maintained at 37 ◦ C. Pluronic F-68 (0.1%, w/v, Sigma–Aldrich Corp., MO) was added to the medium as antifoaming agent, when required. The dissolve oxygen (DO) concentration was maintained at 50% of air saturation with the control module of the reactor by adjusting the air/O2 /N2 ratio of the inlet gas. Working volume of the bioreactor was 750 mL. The seeding density used for each experiment was 3.0 × 105 cells/mL. Sampling was done after every 5 h and at each sampling instant viable cell density and lactate dehydrogenase (LDH) concentration inside the bioreactor [21] was measured in triplicates to study shear effect on the cells.
for high MAb productivity was determined by continuous perfusion experiments of HB8696 cells in triplicates in perfusion bioreactor (Biostat, B.Braun, Germany) at four different perfusion rates of 0.625 mL min−1 , 0.75 mL min−1 , 1.125 mL min−1 and 1.5 mL min−1 under the same set of conditions as mentioned in Section 2.6. Perfusion operation was started from 40 h, using two peristaltic pumps operated at equal flow rates. The growth medium used was Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. Working volume of reactor was 750 mL. The seeding density of 3.2 × 105 cells/mL was used in each experiment. Each set of experiment was performed in triplicates. Samples were drawn after every 8 h, and at each sampling instant viable cell densities and MAb concentrations were measured in triplicates. 2.8. Analytical methods Cell density (viable and dead) was measured by trypan blue assay using a haemocytometer. Glucose concentration was measured by Glucose kit (Biolab diagnostics). Glutamine concentration was measured by enzymatic assay [22]. Lactate concentration was measured by d-lactate colorimetric assay kit (Catalogue.K667100, Bio-vision, USA). Ammonia concentration was measured by Nessler’s reagent [23]. LDH was measured by LDH cytotoxicity assay kit (Item no.10008882, Cayman chemical company, USA). MAb content was analyzed by Indirect ELISA and concentrations were calculated by the standard curve of purified IgG1 . MAb was purified from the culture supernatant by protein-G affinity column chromatography using IgG purification kit (Bangalore Genie, India). 2.9. Statistical analysis
2.5. Growth profile of HB8696 hybridoma cells in batch culture The growth profile of HB8696 hybridoma cells was studied by performing their batch culture experiments in triplicates in Biostat B.Braun bioreactor using the Hybricare 46-X medium supplemented with 20% FBS and 5% penicillin–streptomycin solution. Operating conditions were pH 7.2, temperature 37 ◦ C and DO level 50%. Antifoaming agent was added, when required. The seeding density used was 3 × 105 cells/mL. Spinfilter rotation speed was set at 45 rpm (optimized value). Samples were drawn after every 8 h and at each sampling instant, the measurement of each variable (viable cell density, total cell density, percentage of viable and dead cells, and glucose, glutamine, lactate, ammonium, LDH and MAb concentrations) was carried out in triplicates. 2.6. Comparative perfusion experiments in silk spinfilter and stainless-steel spinfilter Comparative continuous perfusion experiments of HB8696 hybridoma cells were performed in perfusion bioreactor (Biostat, B.Braun, Germany) in triplicates using Stainless-steel spinfilter (pore size 20 m) and silk spinfilter. The operating conditions were maintained at pH of 7.2, cultivation temperature of 37 ◦ C and DO concentration at 50% of air saturation. Antifoaming agent, Pluronic F-68 was used, when required. Working volume of reactor was 750 mL. The seeding density used was 3.2 × 105 cells/mL. Spinfilter rotation speed was set at 45 rpm (optimized value). Perfusion operation was started from 40 h, when viable cell density was started to decrease in batch culture, using two peristaltic pumps at equal perfusion rate of 0.75 mL min−1 . Sampling during these experiments was done after every 8 h and at each sampling instant viable cell density, total cell density and MAb production were measured as the primary parameters for evaluating the performance while percentage retention of viable and dead cells, lactate, ammonium and LDH accumulation inside the bioreactor were measured to correlate the changes in the medium affecting viability and productivity. At each sampling instant all of these parameters were measured in triplicates. 2.7. Perfusion experiments at different perfusion rates To enhance the MAb productivity from hybridoma cells, the performance of silk spinfilter at different perfusion rates was studied. The spinfilter rotation speed was kept constant at an optimized value of 45 rpm. The optimized perfusion rate
All the data has been presented as the mean of three independent series of experiments, each with n = 3 per group, mean ± standard deviation. Analysis of variance (ANOVA) was performed using Stat-Ease Inc. software (USA).
3. Results and discussions In this section, the characterization of silk as spinfilter screen for clog-free extended perfusion operation of hybridoma cells and optimized perfusion strategy for enhancing MAb productivity against breast cancer using novel silk spinfilter are described as follows. 3.1. Characterization of non-woven Bombyx mori silk as spinfilter screen For selection of suitable silk as filter screen of the reusable spinfilter module, the surface properties of three different non-woven Bombyx mori silks, Seri-DSS, Seri-FSS and BF27PV and their effect on cell adhesion and proliferation were studied. It was reported that neutral or negatively charged hydrophobic membranes having pore size ranges between 10 and 20 m were less prone to clog compared to the positively charged hydrophilic membranes [9,10,13–16]. Thus, among the three different silks, the most negatively charged and hydrophobic one was selected as the suitable spinfilter screen. A comparison of surface properties (Table 1) showed that SeriDSS was the most hydrophobic (water contact angle: 107.7 ± 1.1◦ for the upper side and 102 ± 2.7◦ for lower side, Fig. 2) and negatively charged (zeta potential value −38 ± 0.5 mV) silk. Therefore, Seri-DSS silk did not support the attachment and growth of adherent HT1080 cells on its surface over the entire culture period
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
47
Table 1 Comparison of properties of Seri-FSS, Seri-DSS and BF27PV silk membranes. Properties
Seri-DSS silk
Seri-FSS silk
BF27PV silk
Source Cell adhesion Cell proliferation
Bombyx mori Does not support cell adhesion Maximum viable cell density of 1.1 ± 0.1 × 106 cells/mL achieved at 40 h 14.9 ± 1.8 m Negative −38 ± 0.5 mV Both sides hydrophobic Upper side 107.7 ± 1.1◦ Lower side 102 ± 2.7◦
Bombyx mori Supports cell adhesion Maximum viable cell density of 1.0 ± 0.1 × 106 cells/mL achieved at 40 h 15.4 ± 0.8 m Positive 2.0 ± 0.2 mV One side hydrophobic 101.5 ± 2.1◦ Other side hydrophilic 15.3 ± 1.7◦
Bombyx mori Support cell adhesion Maximum viable cell density of 1.0 ± 0.1 × 106 cells/mL achieved at 40 h 14.5 ± 1.1 m Positive 7.9 ± 0.2 mV Both sides hydrophilic Upper side 11.7 ± 1.0◦ Lower side 11.1 ± 1.3◦
Pore size Surface charge density Wettability
Fig. 3. Time profile of viable cell density (×106 cells/mL) of adherent HT1080 cells in control wells and Seri-DSS, Seri-FSS, Bf27PV silks containing wells of 24-well plates.
of 120 h, as shown in Fig. 3. Similarly, in the scanning electron micrographs also no cells were found to be attached on the SeriDSS silk after 72 h (Fig. 4). To rule out if non-attachment and non-proliferation was caused by the Seri-DSS silk, non-adherent HB8696 cells were also cultured over all the silk membranes. It was observed that the presence of circular pieces of silk membranes had no detrimental effect on cell proliferation compared to the control (Fig. 5). Therefore, it was concluded that Seri-DSS silk neither supported cell attachment nor adversely affected cell proliferation. Further, its average pore diameter of 14.9 ± 1.8 m (Fig. 6) was also between 10 and 20 m, suitable for effective retention of hybridoma cells. Due to all of these features, Seri-DSS silk was used as the filter screen of the designed spinfilter module.
Fig. 5. Time profile of viable cell density (×106 cells/mL) of non-adherent HB8696 cells in control wells and Seri-DSS, Seri-FSS, BF27PV silks containing wells of 24-well plates.
3.2. Batch culture for determination of critical shear stress of HB8696 hybridoma cells Batch culture of HB8696 hybridoma cells were performed in bioreactor at different spinfilter rotation speeds to determine the critical shear stress. From the experiments (Fig. 7), it was found that initially with an stepwise increase in spinfilter rotation speed from 10 to 45 rpm, by a factor of 5 after every 5 h, the viable cell density continuously increase up to 40 h and at 40 h, the maximum viable cell density of 1.5 ± 0.1 × 106 cells/mL was achieved. But after 40 h, due to further increase in spinfilter rotation speed above 45 rpm, viable cell density started to decreased and at 50 h, the decreased viable cell density of 0.6 ± 0.1 × 106 cells/mL was achieved. This shows that spinfilter rotation speed above 45 rpm is detrimental for
Fig. 4. Scanning electron micrographs of silk membranes after 72 h of adherent HT-1080 cells culture. (a) Seri-DSS (b) Seri-FSS and (c) BF27PV silk.
48
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 6. Pore size distribution of Seri-DSS silk membrane. (A–C) Scanning electron micrographs of three Seri-DSS silk membrane. (D) Pore size distribution curve of image A, B and C measured by Image-J software.
HB8696 cells and above this, cell lysis occurred. Therefore after 40 h, a sharp increase in LDH concentration up to 520 ± 18.6 U/mL was obtained. Therefore, for all the further batch and perfusion experiments of HB8696 cells, the spinfilter rotation speed of 45 rpm was used. 3.3. Growth profile of HB8696 hybridoma cells in batch culture
Fig. 7. Effect of different spinfilter rotation speeds on HB8696 hybridoma cells during batch culture in bioreactor. Here vcd is viable cell density (×106 cells/mL) and LDH is lactate dehydrogenase accumulation (U/mL).
The growth profile of HB8696 hybridoma cells during batch culture was studied (Fig. 8). The initial concentration of glucose and glutamine were 5.3 mM and 5.4 mM respectively. From the figures, it was found that viable cell density was increased up to 40 h and at 40 h; the maximum viable cell density of 1.5 ± 0.1 × 106 cells/mL was achieved. Beyond this, viable cell density started to decrease gradually due to the accumulation of lactate concentration above 23.3 ± 0.8 mM and ammonium concentration above 5.4 ± 0.3 mM, as observed in all the three batch experiments [20,24–26]. However, MAb production was continued and the maximum MAb concentration of 286.2 ± 41.3 mg/L was achieved at 96 h. Beyond 96 h, due to further accumulation
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Fig. 8. Time profiles of cell growth and product formation of non-adherent HB8696 cells during batch culture in Biostat B.Braun bioreactor. (A) States characterizing the growth of HB8696 cells. Here vcd is viable cell density (×106 cells/mL); tcd is total cell density (×106 cells/mL); %v is cell viability (%). (B) States characterizing the nutrient consumption and waste products accumulation by HB8696 cells. Here Glu is glucose concentration (mM); Gln is glutamine concentration (mM); Lac is lactate concentration (mM); Amm is ammonium concentration (mM). (C) States characterizing product formation of HB8696 cells. Here MAb is monoclonal antibody concentration (mg/L); %v is percentage of viable cells; %d is percentage of dead cells; LDH is lactate dehydrogenase concentration (U/mL).
of lactate above 30 ± 2.4 mM and ammonium above 9.2 ± 0.2 mM, a sharp decrease in viable cell density occurred marked with an increase in LDH concentration and commensurate decrease in MAb concentration.
49
Fig. 9. Comparative time profiles of cell growth and product formation of nonadherent HB8696 cells during perfusion culture in bioreactor using stainless steel (SS) and silk spinfilter. (A) States characterizing the growth of HB8696 cells. Here vcd is viable cell density (×107 cells/mL); tcd is total cell density (×107 cells/mL); %v is cell viability (%). (B) States characterizing waste products accumulation by HB8696 cells. Here LC is lactate concentration (mM); AC is ammonium concentration (mM). (C) States characterizing product formation of HB8696 cells. Here MAb is monoclonal antibody concentration (mg/L); %v is percentage of viable cells; %d is percentage of dead cells; LDH is lactate dehydrogenase concentration (U/mL).
From these results, it was concluded that to increase the MAb productivity, perfusion operation should be started from 40 h, so that the onset of inhibitory concentrations of lactate and ammonium could be delayed and the duration of cell growth phase and production phase extended.
50
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
Table 2 Comparative viable cell densities and monoclonal antibody productivities of HB8696 cells during batch culture in bioreactor and perfusion culture using stainless steel spinfilter and silk spinfilter. HB8696 hybridoma cells Parameters
Time (h)
Bioreactor Batch
Stainless steel spinfilter Perfusion
Silk spinfilter
Viable cell density (107 cells/mL)
40 96 152
0.15 ± 0.1 0.07 ± 0.1 − 0.53 2.6 57.4%
0.5 ± 0.1 1.3 ± 0.1 1.1 ± 0.1 0.94 9.4
0.9 ± 0.1 2.1 ± 0.1 2.5 ± 0.1 1.48 14.8
Monoclonal antibody productivity (g L−1 day−1 ) Total monoclonal antibody produced (g) Increase in productivity from stainless steel to silk Values shown in bold are the maximum viable cell density achieved in each case.
Fig. 10. Time profiles of cell growth and product formation of non-adherent HB8696 hybridoma cells at different perfusion rates and constant spinfilter rotation speed of 45 rpm during perfusion culture using silk spinfilter. (A) Viable cell density (×107 cells/mL). (B) Monoclonal antibody concentration (mg/L).
3.4. Comparative study in silk spinfilter and stainless-steel spinfilter This comparative study demonstrated the application of silk spinfilter for clog-free extended perfusion operation of hybridoma cells (Fig. 9). It is clear from the figure that silk spinfilter provides the extended cell growth phase of 152 h and extended MAb production phase of 192 h, compared to the cell growth phase of 96 h and MAb production phase of 152 h achieved using stainless-steel spinfilter. Using silk spinfilter, a maximum viable cell density of 2.5 ± 0.1 × 107 cells/mL was achieved at 152 h. After this, the viable cell density start to decrease gradually due to the accumulation of lactate above 23.5 mM and ammonium above 5.4 mM, as observed at 40 h during batch culture of HB8696 cells. However, the production of MAb was still continued and at 192 h, the maximum MAb concentration of 723.9 ± 70.5 mg/L was achieved. The overall MAb productivity was 1.48 g L−1 day−1 . In comparison, when the Stainless-steel spinfilter was used, the first inhibitory level of lactate (23.5 mM) and ammonium (5.4 mM)
reached 56 h prior than the silk spinfilter and thus the maximum viable cell density of only 1.3 ± 0.1 × 107 cells/mL was achieved. Similarly, the second inhibitory level of lactate (>30 mM) and ammonium (>9.2 mM) crossed 40 h earlier as compared to the silk spinfilter, therefore the MAb concentration as well as MAb productivity of only 528.6 ± 24.2 mg/L and 0.94 g L−1 day−1 respectively was achieved. Therefore, it can be concluded that silk spinfilter performs much better than the Stainless-steel spinfilter (Table 2) and provides 57.4% increase in MAb productivity. This is due to the lower likelihood of hydrophobic and negatively charged Seri-DSS silk screen of the silk spinfilter to clog with negatively charged cells and cell debris as compared to the hydrophilic and positively charged stainless steel screen of stainless steel spinfilter. Due to the lower clogging tendency of silk spinfilter, while using it during the perfusion culture, the accumulation of inhibitory concentrations of toxic metabolites (lactate and ammonium) inside the bioreactor was delayed and thus the healthy environment for cell growth was maintained for longer period of time [20–26]. Therefore it provides extended duration of cell growth and
Table 3 Comparative maximum viable cell densities and maximum monoclonal antibody concentrations achieved during perfusion culture of HB8696 hybridoma cells using silk spinfilter at different perfusion rates.
a b
S. no.
Perfusion rate (mL min−1 )
Rotation speed (rpm)
Maximum viable cell density1 (107 cells/mL)
1 2 3 4
0.625 0.75 1.125 1.5
45 45 45 45
2.3 2.5 3.5 2.1
± ± ± ±
0.1 0.1 0.1 0.0
Time2 (h)
Maximum monoclonal antibody concentration3 (mg/L)
128 152 176 120
454 723.9 793 432
The maximum viable cell density ‘1’ (mean ± standard deviation) occurs at time ‘2’. The maximum monoclonal antibody concentration ‘3’ (mean ± standard deviation) occurs at time ‘4’.
± ± ± ±
25.5 70.5 22.2 22.2
Time4 (h) 152 192 200 152
S. Kamthan et al. / Enzyme and Microbial Technology 64–65 (2014) 44–51
production phase and enhanced the MAb productivity as compared to the conventional Stainless-steel spinfilter. 3.5. Effect of perfusion rates on monoclonal antibody productivity using silk spinfilter The study was carried out in a bioreactor at four different perfusion rates (0.625 mL min−1 , 0.75 mL min−1 , 1.125 mL min−1 and 1.5 mL min−1 ) using silk spinfilter for each set of experiment, sampling was carried out after every 8 h and the state-time profile of viable cell density and MAb concentration was constructed (Fig. 10). From the maximum value of the viable cell density and MAb concentration achieved in each case (Table 3), it was found that at a perfusion rate of 1.125 mL min−1 , the maximum viable cell density of 3.5 ± 0.1 × 107 cells/mL and maximum MAb concentration of 793 ± 22.2 mg/L was achieved. The MAb productivity achieved at these optimum operating conditions was 1.6 g L−1 day−1 . From this, it is concluded that the optimum perfusion rate for MAb production using silk spinfilter was 1.125 mL min−1 . 4. Conclusion The results show that the novel silk screen based spinfilter can be used for longer perfusion culture of hybridoma cells with enhanced MAb productivity. The negatively charged hydrophobic surface in the silk spinfilter reduces cell and debris adhesion on the filter screen and lowers its tendency to clog as compared to the positively charged hydrophilic stainless-steel spinfilter. Using silk spinfilter, 57.4% increase in overall MAb productivity, as compared to the conventional stainless-steel spinfilter was achieved. Author disclosure and contribution The idea of the research work was conceived by Prof. P.K. Roychoudhury. The experiments were designed by all the authors and performed by Shweta Kamthan. The manuscript was written jointly and approved by all the authors. Acknowledgement The authors gratefully acknowledge the funding received for this work from the Department of Biotechnology, Ministry of Science and Technology, Government of India (Grant no. BT/PR9154/PID/06/387/2007). References [1] Carter JB, Shevitz J. A brief history of perfusion biomanufacturing. BioProcess Int 2011;9(9):24–30.
51
[2] Vogel JH, Nguyen H, Giovannini R, Ignowski J, Garger S, Salgotra A, et al. A new large-scale manufacturing platform for complex biopharmaceuticals. Biotechnol Bioeng 2012;109(12):3049–58. [3] Warnock JN, Rubeai MA. Bioreactor systems for the production of biopharmaceuticals from animal cells. Biotechnol Appl Biochem 2006;45:1–12. [4] Warikoo V, Godawat R, Brower K, Jain S, Cummings D, Simons E, et al. Integrated continuous production of recombinant therapeutic proteins. Biotechnol Bioeng 2012;109(12):3018–29. [5] Castilho LR, Medronho RA. Cell retention devices for suspended cell perfusion cultures. Adv Biochem Eng Biotechnol 2002;74:129–69. [6] Kompala DS, Ozturk SS. Optimization of high cell density perfusion bioreactors. In: Ozturk SS, Hu WS, editors. Cell culture technology for pharmaceutical and cell-based therapies. New York: Taylor & Francis Group; 2006. p. 155–224. [7] Chetreanu FV, Ferreira LGF, Rabe R, Stockar UV, Marison IW. An on-line method for the reduction of fouling of spin-filters for animal cell perfusion cultures. J Biotechnol 2007;130:265–73. [8] Whitford WG. Single-use systems as principal components in bio-production. BioProcess Int 2010;8(11):34–42. [9] Avgerinos GC, Drapeau D, Socolow JS, Mao JI, Hsiao K, Broeze RJ. Spin filter perfusion system for high density cell culture: production of recombinant urinary type plasminogen activator in CHO cells. BioTechnology 1990;8:54–8. [10] Esclade LRJ, Carrel S, Péringer P. Influence of the screen material on the fouling of the spin filters. Biotechnol Bioeng 1991;38:159–68. [11] Mercille S, Johnson M, Lemieux R, Massie B. Filtration-based perfusion of hybridoma cultures in protein-free medium: reduction of membrane fouling by medium supplementation with DNase I. Biotechnol Bioeng 1994;43:833–46. [12] Ozturk SS. Comparison of product quality: disposable and stainless steel bioreactor. Berlin: BioProduction; 2007. [13] Deo YM, Mahadevan MD, Fuchs R. Practical considerations in operation and scale-up of spin-filter based bioreactors for monoclonal antibody production. Biotechnol Prog 1996;12:57–64. [14] Yabannavar VM, Singh V, Connelly NV. Mammalian cell retention in a spinfilter perfusion bioreactor. Biotechnol Bioeng 1992;40:925–33. [15] Nermen M, Nakhla G, Wan W. Comparative assessment of hydrophobic and hydrophilic membrane fouling in waste water applications. J Membr Sci 2009;339:93–9. [16] Sartorius. Spinfilter P. An innovative disposable spinfilter system for cell culture perfusion. Pharm Online Newsl 2005. [17] Altman GH, Diaz F, Jakuba C, Calabro T, Horan RL, Chen J, et al. Silk based biomaterials. Biomaterials 2003;24:401–16. [18] Stacy S. Cancer statistics report: deaths down 20% in two decades. Cancer Stat 2014;7(January). [19] Banik GG, Heath CA. Hybridoma growth and antibody production as a function of cell density and specific growth rate in perfusion culture. Biotechnol Bioeng 1995;48(3):289–300. [20] Dong H, Tang YJ, Ohashi R, Hamel JFP. A perfusion culture system using a stirred ceramic membrane reactor for hyperproduction of IgG (2a) monoclonal antibody by hybridoma cells. Biotechnol Prog 2005;21(1):140–7. [21] Ludwig A, Kretzmer G, Schugerl K. Determination of a critical shear stress level applied to adherent mammalian cells. Enzyme Microb Technol 1992;14:209–13. [22] Sowerby JM, Ottaway JH. The enzymic estimation of glutamate and glutamine. Biochem J 1966;99:246–52. [23] Morrison GR. Micro chemical determination of organic nitrogen with Nessler reagent. Anal Biochem 1971;43:527–32. [24] Krampe B, Al-Rubeai M. Cell death in mammalian cell culture: molecular mechanisms and cell line engineering strategies. Cytotechnology 2010;62(3):175–88. [25] Ozturk SS, Riley MR, Palsson BO. Effects of ammonia and lactate on hybridoma growth, metabolism and antibody production. Biotechnol Bioeng 1992;39:418–31. [26] Ozturk SS, Thrift JC, Blackie JD, Naveh D. Real-time monitoring and control of glucose and lactate concentrations in a mammalian cell perfusion reactor. Biotechnol Bioeng 1997;53(4):372–8.
