Ultrasonics xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Ultrasonics journal homepage: www.elsevier.com/locate/ultras

Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags Yupeng Zhao a, Jida Xing b, James Z. Xing c,d, Woon T. Ang c, Jie Chen a,b,⇑ a

Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Alberta, Canada c IntelligentNano Inc., Edmonton, Alberta, Canada d Department of Laboratory Medicine & Pathology, University of Alberta, Canada b

a r t i c l e

i n f o

Article history: Received 26 December 2013 Received in revised form 19 April 2014 Accepted 27 April 2014 Available online xxxx Keywords: Low-intensity pulsed ultrasound (LIPUS) Chinese Hamster Ovary (CHO) cell CHO cell expression system Monoclonal antibody production Wavebag cell culture and Cell permeability

a b s t r a c t Many technologies, such as cell line screening and host cell engineering, culture media optimization and bioprocess optimization, have been proposed to increase monoclonal antibody (mAb) production in Chinese Hamster Ovary (CHO) cells. Unlike the existing biochemical approaches, we investigated stimulation using low-intensity pulsed ultrasound (LIPUS) as a purely physical approach, offering enhanced scalability, contamination control and cost-efficiency, while demonstrating significantly increased cell growth and antibody production. It was found that daily ultrasound treatments at 40 mW/cm2 for 5 min during cell culture increased the production of human anti-IL-8 antibody by more than 30% using 10 or 30 mL shake flasks. Further increasing the ultrasound dosage (either intensities or the treatment duration) did not appreciably increase cell growth or antibody production, however feeding the culture with additional highly-concentrated nutrients, glucose and amino acids (glutamine in this case), did further increase cell growth and antibody titer to 35%. Similar ultrasound treatments (40 mW/cm2, 5 min per day) when scaled up to larger volume wavebags, resulted in a 25% increase in antibody production. Increased antibody production can be attributed to both elevated cell count and the ultrasound stimulation. Theoretical study of underlying mechanisms was performed through the simulations of molecular dynamics using the AMBER software package, with results showing that LIPUS increases cell permeability. The significance of this study is that LIPUS, as a physical-based stimulation approach, can be externally applied to the cell culture without worrying about contamination. By combining with the existing technologies in antibody production, LIPUS can achieve additional mAb yields. Because it can be easily integrated with existing cell culture apparatuses, the technology is expected to be more acceptable by the end users. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction Monoclonal antibodies (mAbs) have great potential in biomedical research and the pharmaceutical industry due to their specific targeting abilities in a very broad range of clinical applications, such as treating inflammatory disorders, cancers and infectious diseases [1,2]. However, antibody therapies generally require sustained doses over long periods of time, indicating that large quantities of antibodies are needed. In order to meet market demands, it is desirable to establish highly productive and consistent manufacturing processes [3]. ⇑ Corresponding author at: Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada. Tel.: +1 (780) 492 9820. E-mail address: [email protected] (J. Chen).

The wavebag approach, originally developed for optimizing mammalian cell cultures, has become one of the leading technologies in protein production based on animal cell culture. Singh introduced the first widely accepted single use culture system in 1999 [4]. Mounted on a rocking platform, pre-sterilized wavebags can sustain cell densities in excess of 10 million cells/mL in recombinant protein production. One advantage of this technology is that it can produce recombinant protein in a large scale one-time-use wavebag without worrying about contamination. Another advantage is reducing capital costs by eliminating the need for large instruments, such as 10,000 L stainless-steel stirred-tank bioreactors. Wavebag bioreactors are useful tools for transferring the technology from a small scale (more research-oriented) to a large scale (more manufacturing-oriented). This technology has been used by many biotechnology companies for producing recombinant

http://dx.doi.org/10.1016/j.ultras.2014.04.025 0041-624X/Ó 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

2

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

proteins. Wave-motion bioreactors are also preferred in generating bulk material with large-scale transient transfections. Chinese Hamster Ovary (CHO) cells, one of the recombinant protein expression systems, have been extensively studied and developed, serving as a major stable platform for producing mAbs. To increase manufacturing capacity, a wide range of techniques have been applied to the CHO cell system, including cell line screening and host cell engineering [5–7], culture media optimization [8] and bioprocess optimization [9]. Some very useful reviews in this area, including Birch and Racher [10], Wurm [11], Rodrigues et al. [12] and Costa et al. [13], have been published. To further increase antibody production beyond existing limits, we developed a unique physical stimulation method in which low-intensity pulsed ultrasound (LIPUS) was applied to stimulate cell growth. Ultrasound is defined as a longitudinal mechanical wave with frequency exceeding 20 kHz, beyond the range of human hearing. LIPUS is a type of pulsed ultrasound that has a low intensity ranging from 0.02 to 0.2 W/cm2 (here cm2 refers to the ultrasound emitting area). In what follows, we will show how LIPUS can increase human anti-IL-8 mAb production in CHO cells using shake flasks or wavebags. The same design can be applied to produce other types of antibodies. 2. Materials and methods 2.1. CHO cells and media To demonstrate the function of LIPUS in increasing antibody production, a CHO cell line (ATCC#: CRL-12445) was purchased from the American Type Culture Collection (ATCC, Manassas, VA 20108, USA). This cell line is stable in terms of expressing recombinant human anti-IL-8. The culture medium (Hyclone, SFM4CHO, Cat #: SH30548.02) was purchased from Thermo Scientific (IA 52001, USA). Cells were cultivated and maintained in either 50 mL flasks (with 10–15 mL of media) or 125 mL flasks (with 30 mL of media). The cells were cultured in media undergoing mechanical oscillations, either shake flasks or rotational wavebag reactors, ensuring cells receive homogeneous exposure to ultrasound stimulation while improving oxygen exchange between CHO cells and their environment. The shaker was fixed in a CO2 incubator at 37 °C with 5% CO2. The shake flasks were purchased from Fisher Sci. (NH 03842 USA, Cat #: PBV125) and were composed of optically clear, durable polycarbonate. The flask is about 2–3 mm thick. In the wavebag experiment, a Cellbag™ disposable bioreactor (Pat #: CB500 ml 10-01, General Electric Healthcare, PA 15264, USA) was used, with each bag capable of holding 50–250 mL of media. Fresh mixed air (with 5% CO2) was added into the bag and exhausted air was vented out every day during the culture. A rotator (Model #: 4631, Thermo Sci. IA 52001, USA) was used to circularly oscillate the wave-bag. The rotator and wavebag were both placed in a CO2 incubator to keep the temperature at 37 °C. 2.2. Instrumentation LIPUS was generated using SonaCell™ (jointly developed by our lab and IntelligentNano Inc., Edmonton, Canada), whose electronic circuitry drives 5–6 ultrasound transducers arrayed on a platform (see Fig. 1). Cultured cells in either a flask or a wavebag reactor were placed on a SonaCell™ station for ultrasound stimulation. Under the control of SonaCell™ software, the SonaCell™ device automatically generated ultrasound with the designated power intensity needed for cell cultures. SonaCell™ operated at a frequency of 1.5 MHz and a pulse repetition rate of 1.0 kHz (or the pulse duty cycle used was 20% i.e. ‘pulse’ duration for 200 ls and

‘null’ duration for 800 ls). The intensity of the pulsed ultrasound was adjustable from 20 mW/cm2 up to 200 mW/cm2. Here the intensity refers to the rate at which energy passes through a unit area of an ultrasound transducer. The average intensity of an ultrasound beam is the total power in the ultrasound beam divided by the cross-sectional area of the beam. The intensity used in this article refers to the intensity of spatial peak temporal average, (Ispta). 40, 60 and 80 mW/cm2 (or ultrasound pressure amplitude of 268.76, 403.15, and 537.53 mPa) needed for cell stimulations were individually calibrated using an ultrasound power meter (Ohmic Instruments Company, Maryland, USA). In our experiment, we used a round transducer with a contact area of 3.5 cm2 (the diameter is 2.1 cm) and pulsed ultrasound with an average intensity from 40 mW/cm2 to 80 mW/cm2. For the wavebag stimulation, we used an array of ultrasound transducers to cover the bottom of cell wavebag because the base area of cell wavebag is much larger than individual transducers (see Fig. 2). 2.3. Setup of ultrasound stimulation The SonaCell™ and experimental setup are shown in Figs. 1 and 2. We calibrated the ultrasound intensities using the power meter prior to experiments. When ultrasound stimulation was applied, ultrasound gel was applied on top of the transducers. The flask was then placed above the transducers. The thickness of the gel was about 1–2 mm. To avoid any deviation in ultrasound intensity delivered to the cells, we agitated the culture medium during sonication. Interference between transmitted and reflected ultrasound which produce energetic standing waves render potentially adverse effects on CHO cells. However, in shake flask or wavebag configurations, our system features an open-ended design with acoustically transparent boundary conditions. A quarterwavelength impedance matching layer was adhered to the ultrasound-emitting side of the transducer to reduce power reflection; therefore ultrasound standing waves should not pose a major concern in the experiments. Ultrasound is well known for its excellent propagation through liquids. Because CHO cells were grown in liquid suspension, minimal attenuation was expected as ultrasound penetrated the media from the bottom of a shake flask, confirmed by the measurements using the power-meter and the hydrophone probe (Precision acoustics Inc. UK). 2.4. Ultrasound stimulation Fig. 1 shows how we performed the studies. The CHO cells in shake flasks were cultured on a shaker in a CO2 incubator. The cells were seeded in a flask containing antibody production medium. To stimulate the cells with LIPUS, the flasks were removed from the shaker and placed on a SonaCell™ transducer station. After ultrasound treatment, the flasks were placed back on the shaker and continued to be incubated at 37 °C. The ultrasound treatment started at 0 h and was performed every 24 h for either 5 or 10 min per day for 6 days. Because the ultrasound intensity of LIPUS is low and daily treatments are less than 10 min in duration, no temperature changes were observed using a high accuracy J/K input themocouple thermometer (Model #HH802U, OMEGA Engineering Inc. Connecticut, USA) in culture media during ultrasound treatment. Cell counting was performed every day and antibody production was measured using ELISA (Enzyme-linked Immunosorbent Assay). 2.5. Experimental setup Four sets of experiments were designed for this study with each individual experiment performed in triplicate.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

3

Fig. 1. The schematic diagram of SonaCell and its experimental setup for shake flask or wavebag.

Fig. 2. After the Chinese Hamster Ovary (CHO) cells were inoculated in a wavebag, the SonaCell was used to stimulate the cells in the wavebag through the ultrasound transducer placed underneath each flask. Ultrasound parameters were adjusted according to the desired conditions. After each stimulation, the wavebag was placed back in the incubator. An array of ultrasound transducers was used for the cell wavebag experiment. Electrical cables connected the station to the LIPUS generator. The inset graph shows the quantitative analysis of standard mAbs with ELISA assay and the linear calibration.

1. The first experiment investigated whether the ultrasound treatment affects CHO cell growth and antibody production in 50 mL shake flasks as an initial screening. 2. The second was designed to check whether stronger ultrasound treatment is needed during scale-up cell cultivations in 125 mL shake flasks. 3. The third experiment was performed to test whether nutrient feeding affects CHO cell growth and antibody production along with ultrasound treatment. Two methods were used: by adding original media and adding additional highlyconcentrated nutrient components (glucose and glutamine).

4. The last experiment was to test whether the ultrasound treatment can be scaled up and applied to wavebag cell cultures. The purpose of these experiments is to find good LIPUS stimulation and cell culture conditions, but we cannot claim these are the best conditions due to the limited number of trials. In this series of experiments, the CHO cell culture was treated by ultrasound once per day while the control did not receive any ultrasound treatment. The ultrasound treatment intensities (Ispta = 40, 60 or 80 mW/cm2) and the treatment durations (5 or 10 min per

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

4

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

treatment) were set in each experiment. The detailed information can be found in the ‘‘Results and Discussion’’ section. 2.6. Analysis 2.6.1. Cell count 1 mL of sample was taken to check cell viability and antibody titer every 24 h. Viable and dead cells were counted using a hemacytometer with trypan blue staining. The cell viability was calculated based on the number of viable cells and total number of cells (both viable and dead cells). Both chambers of the hemacytometer from the same sample were counted and the average was used in this study. 2.6.2. Antibody analysis 1 mL of culture sample was centrifuged and the supernatant was collected to determine the antibody titer. ELISA was used to determine or compare the antibody titer in a sample. The capture antibody was interleukin-8 (Gibco, PHC0881) and the detection antibody was mouse anti-human IgG1 horseradish peroxidase conjugate (Invitrogen, cat #: 982455). 3,30 ,5,50 -Tetramethylbenzidine (TMB) substrate was purchased from BD science (BD OptEiA, cat #: 555214). The standard monoclonal antibody (mAb) was titrated and accurately measured by ELISA in order to determine whether the optical density (OD) value obtained with ELISA was quantitatively correlated with the concentration of antibodies. The OD values determined by ELISA were then linearly correlated with the standard mAbs (see Fig. 2). The OD values for each antibody concentration were measured by three replicates of appropriate antibody concentrations to evaluate the precision in the ELISA. The ELISA was also assessed by using a real cell culture suspension containing the human anti-IL-8 mAbs, and similar correlation coefficients for both standard and the anti-IL-8 mAbs were obtained (data not shown). 2.6.3. Statistical analysis Experimental values were determined in three replicates. All values regarding measurement were expressed as means and standard deviations (SD). The pairwise comparison of either cell viability or absorption (for antibody test) values between the samples with and without LIPUS treatment was calculated using the t-test. p values less than 0.05 (or p < 0.05) were considered statistically significant. 3. Results and discussions 3.1. Test whether the ultrasound treatment affects CHO cell growth and antibody production Based on the literature studies [15–17] and our previous success [18], ultrasound at 1.5 MHz can promote cell growth. In this experiment we selected the same ultrasound frequency to test whether it generates any positive effect on the growth of CHO cell cultures and mAb production (testing ultrasound of different frequencies is worthy of further study, though not within the scope of this study). To achieve this goal, three ultrasound intensities (Ispta = 40, 60 and 80 mW/cm2) and two treatment durations (5 and 10 min) were chosen. Each 50 mL shake flask was seeded with 10 mL of working volume and a starting cell density of 4  105 cells/mL. One milliliter of culture sample was extracted every 24 h to check cell viability and antibody titer. The comprehensive results are presented as a bar diagram in the Appendix, displaying the full range of conditions and time-points. Here the averaged curve plots are presented for simplicity. Fig. 3 shows that viable cell density by varying LIPUS intensities or treatment time.

(a) By using 40 mW/cm2 but different treatment durations (5 min or 10 min) to stimulate CHO cells, both LIPUS treatments performed better than the control in increasing viable cell density. However, 40 mW/cm2 and 10 min worked better (refer to Fig. 3A). (b) Performing the same tests at 60 mW/cm2, 10-min treatments outperformed controls, however 5-min treatments were worse than controls after 72 h (refer to Fig. 3B). Fig. 3C and D shows the experimental results of using the same treatment duration (either 5 min or 10 min) but with different ultrasound intensities. The studies indicate that LIPUS promotes CHO cell growth during the first 96 h, especially when treated at 40 mW/cm2 for 5 min or 10 min per day, and 60 mW/cm2 for 10 min per day. However, after 96 h cell density started to drop (see Fig. 4A) due to nutrient depletion (refer to later discussion). LIPUS treatments under the other conditions, for example 80 mW/cm2 and 5 min or 10 min were also performed (see Fig. 4B). The number of cells treated by LIPUS increased, but no obvious dose–response effects were observed. We further discovered that the ultrasound affects the growth of CHO cells only within a certain range of LIPUS intensities (between 30 mW/cm2 and 100 mW/cm2). Lower intensity (less than 30 mW/cm2) had no measurable effect while higher intensities (greater than 100 mW/cm2) killed the cells (data not shown). Fig. 4C shows how LIPUS increased antibody production. With the stimulation of 40 mW/cm2 for 5 min per day, a significant increase in antibody production occurred at 96 h (p < 0.05). The maximum antibody production increase (about 30% above that of control cells) occurred at 96 h while the number of cells only increased 10% over the control under the same ultrasound conditions of 40 mW/cm2 and 5 min per day. This phenomenon suggests that  Ultrasound can increase cell numbers and enhance mAb production at the same time. However, the increase of antibody production does not necessarily correlate with the increase in the number of cells. For example, although 40 mW/cm2 and 5 min does not result in the most cell growth (refer to Fig. 4A), it yields the greatest increase in mAb production (about 30% increase). Evidently, sonication independently promotes both cell growth and antibody production. If mAb production is high, the corresponding viable cell density might not be the highest. On the other hand, if all of the gain is in cell number, mAb production most likely is not the highest. Therefore, the experimental designs are very much dependent on the goal of end users, either increasing mAb production or increasing viable cells. The LIPUS treatment then has to be tailored or customized accordingly. 3.2. Check whether stronger ultrasound treatment is needed during scale-up cell cultivations In our previous studies, we screened ultrasound conditions on a 50 mL shake flask with 10 mL working volume of cell culture by either (a) fixing the ultrasound intensity but varied treatment time or (b) fixing the ultrasound treatment duration while varying ultrasound intensity. The purpose of this experiment was to test whether a greater total ultrasound dose is needed to accommodate increased working volume (a 125 mL shake flash with 30 mL working volume). A hypothesis is, as cell density during cultivation increases, each cell receives less ultrasound treatment thus requiring higher total ultrasound doses to compensate. Here, ‘‘increasing ultrasound dose’’ is defined as either ‘‘increasing the ultrasound intensity’’ or ‘‘extending the treatment duration’’, or both. To vali-

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

5

Fig. 3. CHO cell growth with and without ultrasound treatment (here only a few selected conditions were plotted. Comprehensive results spanning all conditions and time points are presented as a bar graph in Appendix). (A) 40 mW/cm2 at 5 min or 10 min; (B) 60 mW/cm2 at 5 min or 10 min; (C) 5 min at 40 mW/cm2 and 60 mW/cm2; (D) 10 min at 40 mW/cm2 and 60 mW/cm2. The experiments started with 4  105 cells per mL at Day 0. The symbol ‘‘’’ indicates that a significant difference (p < 0.05) was shown when comparing the LIPUS treated cells with control. 1 mL of sample was removed from the 10 mL working volume at each time point to check cell viability and antibody titer.

date this hypothesis, cells were divided into four groups and the experiments were carried out as follows: (a) Cells did not receive any ultrasound treatments as the negative control. (b) Cells were treated by ultrasound at 40 mW/cm2 and 5 min as the unmodified dose baseline. (c) Cells were treated by ultrasound with an intensity of 40 mW/cm2 at day 1, 60 mW/cm2 at day 2, and 80 mW/ cm2 at day 3 with the duration of each treatment fixed at 5 min (the test was denoted as 40/60/80 mW and 5 min in Fig. 5A). (d) Cells were treated with a fixed intensity of 40 mW/cm2 but with an extended duration of treatment, 5 at day 1, 7.5 at day 2, then 10 min at day 3 (the test was denoted as 40 mW and 5/7.5/10 min in Fig. 5A). Fig. 5A shows the testing results (the working volume was 30 mL and the starting cell density was 5  105 cells/mL). It can be observed from Fig. 5A that ultrasound will help cells grow, especially at 40 mW/cm2 and 5 min of treatment per day, and especially after 72 h. The cell density was marginally lower at 96 h than that of the control and that of 80 mW/cm2 and 5 min per day due to nutrient (glucose and amino acid, glutamine in this case) exhaustion (see the later discussion). Extending the treatment duration (40 mW/cm2 and 5/7.5/10 min) or increasing

ultrasound intensity (40/60/80 mW/cm2 and 5 min), however, did not further increase cell growth. The results indicate that the hypothesis is wrong, and higher ultrasound strength is not necessary for scale-up cell cultivations. This finding is in line with what we have learned about LIPUS: that is, although the working volume increased, each cell still received the same amount of ultrasound treatment because ultrasound can penetrate cell culture media without much attenuation. Antibody production in the LIPUS treated sample (40 mW/cm2 and 5 min) at 72 h was the highest, which is about 30% higher than the control (see Fig. 5B). Although the ultrasound parameters (40 mW/cm2 and 5 min) used in both experiments (refer to Fig. 4C and Fig. 5B) are the same, the culture conditions are different:  The working volume was 10 mL and the starting cell numbers were 4  105 cells per mL at day 0 in Fig. 4C.  The working volume was scaled up to 30 mL and the starting cell numbers were increased to 5  105 cells per mL at day 0 in Fig. 5B. The antibody production increases for both experiments were about the same, or 30% increase. However, the time took to reach the production peak was shortened (from 96 h to 72 h) in the scale-up design. The possible hypothesis is that cells under LIPUS treatment can grow better in a large space with enough nutrients.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

6

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

Fig. 4. (A) CHO cell growth compared to the control. The ultrasound treatment conditions are 40 mW/cm2 for 5 min and 10 min, 60 mW/cm2 for 10 min. (B) CHO cell growth compared with the control. The ultrasound treatment conditions are 80 mW/cm2 for 5 min and 10 min. (C) Antibody expression of CHO cells with or without LIPUS stimulation (40 mW/cm2 for 5 min). The titer of the antibody is measured by ELISA. The experiments started with 4  105 cells per mL at Day 0. The symbol ‘‘’’ indicates that a significant difference (p < 0.05) was shown when comparing the LIPUS treated cells with control.

3.3. Test whether nutrient feeding or extra media affects CHO Cell growth and antibody production The purpose of this experiment was to test whether ultrasound helps cells take advantage of extra culture media added to further increase their growth. The culture started in a 30 mL shake flask (10 mL of working volume and 4  105 cells/mL). After taking two samples (1 mL at 24 h and another 1 mL at 48 h), 8 mL of media remained. 8 mL of fresh media was then added into each flask and culture continued. A parallel experiment was performed with exactly the same setup as above except without LIPUS stimulation. Viable cell density was calculated and the results are shown in Fig. 6A. The results indicate that the ultrasound can enhance cell growth during the first 48 h. After adding the extra 8 mL of media at 48 h, the cell concentration in both samples were diluted by two times, but the cell density at 72 h did not increase in both samples (with or without the LIPUS stimulation). The cells continually grew in both samples at 96 h and 120 h. By comparing the cell density at 72 h, 96 h and 120 h in the LIPUS treated sample with those of the untreated sample, the ultrasound treatment did not show a significant enhancement (p > 0.05). The ultrasound treatment no longer affected cell growth after adding extra media. In the previous study (see Fig. 4A), we observed that the cell density with 40 mW/cm2 and 5 min or 10 min, and 60 mW/cm2 and 10 min dropped after 96 h. We projected that this was caused by nutrient depletion. The purpose of this experiment was to test

how adding nutrients (feeding highly-concentrated nutrients e.g. glutamine and glucose) impacted cell growth. Unlike the previous design, this feeding strategy can provide cells with nutrients without increasing the working volume. During this experiment 300 lL of glutamine and glucose were added into both the control and the ultrasound treated flasks at 72 h and 120 h. After 120 h (the peak of cell growth), the viable cell density of ultrasound stimulated cells decreased and thus no more nutrients were added (refer to Fig. 6B). To evaluate if LIPUS stimulation could catalyze nutrient addition to promote cell growth in synergy, the number of living cells (viability) in the culture after adding the nutrients was measured (Fig. 6C). The results show that the ultrasound treatment can keep increasing cell growth during the first 120 h. However, the growth rate of ultrasound treated cells decreased even by adding the second nutrient supplement (the control reached its maximum density at 144 h). The reason why the cells cannot continue their growth after 120 h can be attributed to the limitations of the shake flask experimental setup. For instance, parameters, such as the pH and dissolved oxygen, cannot be controlled during the cultivation. In addition, elevated production of wastes such as NH4 and lactic acid accumulate to suppress cell growth. The highest antibody titer was at 120 h. The ultrasound treated flask produced 35% greater antibody yield than the control. The conclusion is that feeding the culture with additional nutrients can further increase cell growth, close to 4.0106 cell/mL at 120 h (Fig. 6B) vs. close to 3106 cell/mL at 96 h (Fig. 4A), and antibody titer, 35% increase at 120 h vs. 30% increase (Fig. 4C).

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

7

Fig. 5. CHO cell growth in 30 mL media in shake flask with and without ultrasound treatment (the control). (A) CHO cell culture was treated by stronger ultrasound conditions: either extending the treatment duration (40 mW/cm2 and 5/7.5/10 min) or increasing ultrasound intensity (40/60/80 mW/cm2 and 5 min). Control refers to CHO cell culture without ultrasound treatment. The experiment started with 5  105 cells per mL at Day 0. (B) Antibody production in the LIPUS treated sample (40 mW/cm2 and 5 min) at 72 h was the highest, which is about 30% higher than the control.

3.4. Test whether the ultrasound treatment can be scaled up and applied to wavebag cell cultures To test whether the LIPUS technology can be scaled up to an industrially viable technology for CHO production, we applied LIPUS to a wavebag reactor (refer to Fig. 2). It was again confirmed that ultrasound treatment could increase CHO cell growth in a wavebag within the first 120 h, after then the cell growth declined due to the nutrient exhaustion (refer to Fig. 6D). When comparing antibody titer, ultrasound yielded increases of 10% and 24% at 96 and 144 h, respectively. This finding indicated that cell growth does not directly correlate to the increase in antibody production. In other words, the increase of antibody production by individual cells can be attributed to LIPUS sonication. 3.5. Mechanisms why ultrasound can increase cell growth Cell membranes are composed of three parts: phospholipid bilayer, proteins and carbohydrates. Because the lipid bilayer structure is the main component of cell membrane and occupies the most area, for simplicity, we have only simulated how ultrasound impacts the lipid bilayer by using the molecular dynamics software — AMBER. This software package is well suited for investigating the molecular dynamics of cell membranes, proteins and nucleic acids and their interactions (please refer to http://ambermd.org/ for the detailed information). The simula-

tion results are shown in Fig. 7. The cell permeability increases as the amount of ultrasound stimulation increases (Here we define the amount of ultrasound stimulation = ultrasound intensityultrasound stimulation duration). According to the simulation results, high cell permeability leads to the increase in the number of cells due to better circulation and thus faster cell metabolism. This finding is in line with the experimental results in [14], which are shown in the Appendix. LIPUS actively enhances some stage(s) in the antibody secretion process. In addition to elevating cell permeability, ultrasound helps CHO cells grow at the molecular level. Different research groups worldwide have previously studied how ultrasound affects molecular interactions in living systems, though their frequency and duty cycle are different from ours. For instance, ultrasound (1.5 MHz at 30 mW/cm2) has been shown to enhance proliferation of human skin fibroblasts by activating the ERK1/2 signaling pathway [15]. LIPUS (1.5 MHz at 30 mW/cm2 for 20 min) was also reported to significantly activate cell growth-promoting tissue inhibitors of metalloproteinase-1 (TIMP-1) in the nucleus of pulposus cells [16]. Ultrasound (45 KHz, 1 MHz and combined frequency (45 KHz and 1 MHz) at 10 mW/cm2 to 75 mW/cm2 for 30 min) induced a dose-dependent response in b-catenin staining in both odontoblast and osteoblast model cell lines, which implicates the Wnt/b-catenin pathway as a possible mechanism for intracellular ultrasound transduction [17]. We will perform studies of LIPUS of 1.5 MHz on CHO cells at the molecular level, which will be the subject of another article.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

8

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

Fig. 6. (A) Cell growth with and without ultrasound treatment in shake flasks (with fresh media added after 48 h culture). The ultrasound treated conditions are 40 mW/cm2, 5 min per treated and once a day. Media was added at 48 h after sampling. Control refers to CHO cell culture without ultrasound treatment. (B) CHO cell growth in 30 mL media in shake flask with and without ultrasound treatment (the control). The ultrasound treatment conditions are 40 mW/cm2, 5 min per treatment and once a day. Concentrated nutrients (100  glutamine and glucose) were added at 72 h and 120 h after sampling at each time point. (C) CHO cell growth curve (viability) in 30 mL media in shake flask with and without (the control) ultrasound treatment. (D) CHO cell growth in a wavebag with and without ultrasound treatment. The ultrasound treated conditions are 40 mW/cm2, 5 min per treatment and once a day.

Fig. 7. The blue points are water molecules and yellow ones are lipid molecules of lipid bilayer. The simulation results are (a) without ultrasound, (b) ultrasound amount of 6  1015 J/cm2, (c) ultrasound amount of 1.8  1014 J/cm2 and (d) ultrasound amount of 3  1014 J/cm2. Here the amount of ultrasound stimulation = ultrasound intensityultrasound stimulation duration. As the amount of ultrasound stimulation increases, from (a) to (d), more water molecules pass through lipid bilayers, thus induce higher cell permeability.

4. Conclusion This work confirms that ultrasound treatment can increase antibody production in the CHO cell expression system. The greatest increase of 35% higher antibody production compared to the control was observed under the ultrasound treatment conditions of 40 mW/cm2, 5 min per treatment and one treatment per day. Increasing the ultrasound dose (either increasing treatment duration or treatment intensity) during cell cultivation is not required, which makes the ultrasound set up and operation easy and scal-

able. Adding concentrated nutrients can further enhance cell growth and increase cell density. The highest cell density does not correspond to the highest antibody production. As a result, the selection of ultrasound treatment conditions depends on whether the goal is to increase cell growth or antibody production. For future work, ultrasound with frequencies other than 1.5 MHz will be studied to see whether it can bring any positive effects on CHO growth and mAb production. In addition, the molecular mechanisms or signal pathways about how LIPUS stimulates CHO cell growth will be investigated.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025

Y. Zhao et al. / Ultrasonics xxx (2014) xxx–xxx

Acknowledgements We would like to acknowledge funding support from National Sciences and Engineering Research Council of Canada (NSERC) discovery grant and Collaborative Research and Development (CDR) grant. We would like to thank Yan Duan for performing the molecular dynamics simulation using AMBER. We would also like to thank Scott MacKay, Xiaoyan Yang and Ray Yang for providing valuable feedback in revising the manuscript. Appendix A CHO cell growth with and without ultrasound treatment presented as a bar chart (refer to Supplemental Fig. S1). After 96 h of cultivation, the ultrasound treated culture (40 mW/cm2 for 10 min and 60 mW/cm2 for 10 min) did not show any significant further growth except for the 80 mW/cm2 condition. The control samples caught up with the growth of the ultrasound treated samples. The reason for this is nutrient (glucose and amino acids) exhaustion because fast growth in the ultrasound treated culture consumed more nutrients during the first 96 h than the control. Appendix B. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ultras.2014. 04.025. References [1] H.E. Chadd, S.M. Chamow, Therapeutic antibody expression technology, Curr. Opin. Biotechnol. 12 (2001) 188–194. [2] J.M. Reichert, C.J. Rosensweig, L.B. Faden, M.C. Dewitz, Monoclonal antibody successes in the clinic, Nat. Biotechnol. 23 (2005) 1073–1078.

9

[3] F. Li, B. Lee, J.X. Zhou, T. Tressel, X. Yang, Current therapeutic antibody production and process optimization, Bioprocess. J. 5 (2006) 16–25. [4] V. Singh, Disposable bioreactor for cell culture using wave-induced agitation, Cytotechnology 30 (1999) 149–158. [5] K. Chen, Q. Liu, L. Xie, P.A. Sharp, D.I.C. Wang, Engineering of a mammalian cell line for reduction of lactate formation and high monoclonal antibody production, Biotechnol. Bioeng. 72 (2001) 55–61. [6] D.M. Dinnis, D.C. James, Engineering mammalian cell factories for improved recombinant monoclonal antibody production: lessons from nature?, Biotechnol Bioeng. 91 (2005) 180–189. [7] R.P. Nolan, K. Lee, Dynamic model for CHO cell engineering, J. Biotechnol. 158 (2012) 24–33. [8] L. Xie, D.I.C. Wang, High cell density and high monoclonal antibody production through medium design and rational control in a bioreactor, Biotechnol. Bioeng. 51 (1996) 725–729. [9] E. Jain, A. Kumar, Upstream processes in antibody production: evaluation of critical parameters, Biotechnol. Adv. 26 (2008) 46–72. [10] J.R. Birch, A.J. Racher, Antibody production, Adv. Drug Deliver Rev. 58 (2006) 671–685. [11] F.M. Wurm, Production of recombinant protein therapeutics in cultivated mammalian cells, Nat. Biotechnol. 22 (2004) 1393–1398. [12] M.E. Rodrigues, A.R. Costa, M. Henriques, J. Azeredo, R. Oliverira, Technological progresses in monoclonal antibody production systems, Biotechnol. Prog. 26 (2010) 332–351. [13] A.R. Costa, M.E. Rodrigues, M. Henriques, J. Azeredo, R. Oliveira, Guidelines to cell engineering for monoclonal antibody production, Eur. J. Pharm. Biopharm. 74 (2010) 127–138. [14] J. Xing, X.Y. Yang, P. Xu, W.T. Ang, J. Chen, Ultrasound enhanced monoclonal antibody production, Ultrasound Med. Biol. 38 (11) (2012) 1949–1957. [15] S. Zhou, A. Schmelz, T. Seufferlein, Y. Li, J. Zhao, M.G. Bachem, Molecular mechanisms of low intensity pulsed ultrasound in human skin fibroblasts, J. Biol. Chem. 279 (2004) 54463–54469. [16] H. Omi, J. Mochida, T. Iwashina, R. Katsuno, A. Hiyama, T. Watanabe, K. Serigano, S. Iwabuchi, D. Sakai, Low-intensity pulsed ultrasound stimulation enhances TIMP-1 in nucleus pulposus cells and MCP-1 in macrophages in the rat, J. Orthop. Res. 26 (2008) 865–871. [17] J.S. Man, Biological Effects Of Low Frequency Ultrasound On Bone And Tooth Cells, PhD Thesis, School of Dentistry, College of Medicine and Dentistry, The University of Birmingham, 2011. [18] P. Xu, H. Gul-Uludag, W. Ang, X.Y. Yang, M. Huang, L. Marquez-Curtis, L. McGann, A. Janowska-Wieczorek, J. Xing, E. Swanson, J. Chen, Low-intensity pulsed ultrasound-mediated stimulation of hematopoietic stem/progenitor cell viability, proliferation and differentiation in vitro, Biotechnol. Lett. 34 (10) (2012) 1965–1973.

Please cite this article in press as: Y. Zhao et al., Applications of low-intensity pulsed ultrasound to increase monoclonal antibody production in CHO cells using shake flasks or wavebags, Ultrasonics (2014), http://dx.doi.org/10.1016/j.ultras.2014.04.025