Large-scale cell suspension culture technology opens up opportunities for numerous medical and bioengineering applications. For these purposes, scale-up of the culture system is paramount. For initial small-scale culture, a simple static suspension culture (SSC) is generally employed. However, cell sedimentation due to the lack of agitation limits the culture volume feasible for SSC. Thus, when scaling up, cell suspensions must be manually transferred from the culture flask to another vessel suitable for agitation, which increases the risk of contamination and human error. Ideally, the number of culture transfer steps should be kept to a minimum. The present study describes the fabrication of an ultrasonic suspension culture system that stirs cell suspensions with the use of acoustic streaming generated by ultrasound irradiation at a MHz frequency. This system was applied to 100-mL suspension cultures of Chinese hamster ovary cells—a volume ten-fold larger than that generally used. The cell proliferation rate in this system was 1.88/day when applying an input voltage of 40 V to the ultrasonic transducer, while that of the SSC was 1.14/day. Hence, the proposed method can extend the volume limit of static cell suspension cultures, thereby reducing the number of cell culture transfer steps.

Cell culture technology opens up opportunities for a wide variety of medical and bioengineering applications. Regenerative medicine, for example, employs a relatively large number of cultures to restore the normal function of damaged tissue. When using regenerative therapy for heart failure, ≥109 myocytes are required to compensate for the lost tissue (Jing et al., 2008). The production of biopharmaceuticals is another example of an application with high culture demands. More than 50% of the top-100 selling drugs are reported to be biotechnology products (Evaluate Pharma, 2019), indicating the importance of cell culture in drug development. To address worldwide concerns regarding food supply and environmental impact, cultured meat is a growing application of cell culture technology (Post, 2012). To generate one ton of cultured meat, 1013 cells are required (Post et al., 2020). Therefore, the mass production of cells is a critical parameter in these applications and will certainly be an important factor in many others yet to be developed.

For the mass production of cells, the scale-up of culture systems remains a big challenge. Suspension culture [or three-dimensional (3D) culture], in which the cells float in the culture medium, is generally used for mass cultures because of its higher space efficiency than adherent culture [or two-dimensional (2D) culture], in which the cells adhere to a cell-culture surface. To achieve suspension culture with a large number of cells, the size of the culture vessel needs to be enlarged. However, for the initial small-scale suspension cultures, the static suspension culture (SSC) method is typically employed in which cells are cultured statically in a culture flask (Clynes, 1998; Jackson et al., 1996; Zweigerdt et al., 2011). However, the volume of cell suspension feasible for SSC is quite limited due to the lack of agitation. As the quantity of cells in the culture flask becomes too large, cell proliferation is increasingly suppressed because of cell aggregation, an inhomogeneous supply of oxygen and nutrients in the medium, and buildup in the local concentration of toxic waste products around the cells. Therefore, the cell suspension must be transferred from the SSC culture flask to another vessel suitable for the agitation of cell suspensions.

A spinner flask is one such culture vessel that is widely used, as these flasks can efficiently stir the cell suspension and homogenize the contents of the culture medium. Although a spinner flask is effective for agitating cell suspensions, transferring a cell suspension from a small spinner flask to a larger one is required as the number of cells increases. For example, two transfer operations, first from the SSC to a 125-mL spinner flask and then from a 125-mL to a 1-L spinner flask, are required to obtain 109 cells. The higher the number of culture transfers, the higher the risk of human error and contamination, and this can ultimately lead to increased costs and reduced product quality.

It would, therefore, be ideal to reduce the number of culture transfer steps to a minimum. To reduce the number of cell transfers, the cell number and cell density capacity of a particular suspension culture process should be maximized for either SSC or suspension culture with medium agitation. The key challenge here is how to stir the cell suspension in the culture vessel efficiently such that a larger number of cells can be cultured in a flask with a similar volume to that conventionally used.

For stirring cell suspensions in a variety of flasks, ultrasound irradiation is a good candidate (Imashiro et al., 2019; Imashiro et al., 2020; Kurashina et al., 2017; Kurashina et al., 2019; Kuriyama et al., 2020; Nakao et al., 2018; Nakao et al., 2019; Tauchi et al., 2019; Terao et al., 2019). In particular, as a substitute for SSC, ultrasonic irradiation has been introduced to agitate cell suspensions in T25 culture flasks (Fujii et al., 2018). Flow of the culture medium is induced by acoustic streaming, which stirs the cell suspension in the culture flask. This method improves cell proliferation compared with SSC. However, using this approach, the volume limit of the cell suspension has been reported to remain the same as that of SSC because of the relatively low-frequency ultrasound employed. Therefore, in the present study, an ultrasonic suspension culture (USC) system was developed in which the cell suspension was effectively stirred in the culture flask through the action of acoustic streaming. Importantly, the proposed USC method extended the volume limit of the cell suspension used for SSC, and it may, therefore, contribute toward a reduction in the number of culture transfer steps used during the scaleup of suspension culture systems.

In this study, cell suspensions in the tissue culture flasks were stirred by acoustic streaming, which is the flow generated by converting acoustic wave energy into energy for fluid motion in the process of sound propagation (Mitome, 1998). The driving force of acoustic streaming per unit volume, F, is given by

F=αIc,
(1)

where I, c, and α represent the acoustic intensity, the speed of sound in liquid, and the attenuation coefficient of liquid, respectively (Lighthill, 1978). The attenuation coefficient of fluid, α, is given by

α=2π2f2νc343+μBμ+γ-1Pr,
(2)

where f, ν, μB, μ, γ, and Pr represent the frequency of ultrasound, the kinematic viscosity coefficient, the bulk viscosity coefficient, the shear viscosity coefficient, the ratio of specific heats, and the Prandtl number defined as the ratio of kinematic viscosity to thermal diffusivity, respectively (Blackstock, 2000). In addition, the acoustic intensity, I, is given by

I=prms2ρc,
(3)

where Prms and ρ represent the root-mean-square (rms) pressure and the density of the liquid, respectively (Blackstock, 2000). In this study, a piezoelectric transducer was used, and a sinusoidal alternating current (AC) voltage was applied to the transducer to generate ultrasound.

When the acoustic pressure, p, has a sinusoidal waveform, p=p0sin(2πft+φ), the rms pressure in the vertical direction relative to the surface of the ultrasonic transducer is given by

prms=p02=2πfρcA2,
(4)

where p0 and A represent the acoustic pressure amplitude and the vibration amplitude of the ultrasonic transducer, respectively (Kim et al., 2014). From Eqs. (1)–(4), the driving force of the acoustic streaming per unit volume, F, is proportional to f4A2. Therefore, the ultrasonic transducer was chosen as it could generate high-frequency ultrasound to generate stronger acoustic streaming. High-frequency ultrasound possibly causes temperature increase, which induces cell damage. This should be taken into account when irradiating ultrasound in the cell culture experiment.

Figure 1(A) provides an overview of the newly developed USC system. The USC system consists of a T75 flask (90075; TPP Techno Plastic Products AG, Trasadingen, Switzerland), a flask holder, an ultrasonic transducer, and a water tank. For the USC system, a piezoelectric transducer (C213; Fuji Ceramics Corporation, Shizuoka, Japan) was employed as the ultrasonic transducer, as it has a MHz-range resonance frequency with a thickness mode. The ultrasonic transducer was placed on the flask holder, and the flask was placed on top of the transducer on the flask holder. Ultrasound propagated from the ultrasonic transducer generated acoustic streaming and stirred the cell culture medium. To propagate the ultrasound efficiently and to inhibit heat generation, 1 L of water was placed in the water tank. For cell culture experiments, the flask was filled with 100 mL of the culture medium. Accordingly, the acoustic path length (i.e., the water level from the flask's bottom) was set at 32 mm.

FIG. 1.

(Color online) (A) Overview of the fabricated ultrasound suspension culture system. (B) Enlarged view of the bottom part of the flask and the coordinate system used for measurements.

FIG. 1.

(Color online) (A) Overview of the fabricated ultrasound suspension culture system. (B) Enlarged view of the bottom part of the flask and the coordinate system used for measurements.

Close modal

The vibration characteristics of the USC system were evaluated to determine the resonance frequency of the ultrasonic transducer, the vibration amplitude of the ultrasonic transducer, and the distribution of the vibration amplitude at the bottom surface of the flask. Note that for the measurements of the vibration amplitude, a flask containing 100 mL of water was placed on the flask holder in a water tank containing 1 L of water to replicate the conditions of the USC method in the suspension culture experiment. The resonance frequency of the ultrasonic transducer was evaluated by measuring the electric impedance of the ultrasonic transducer in the USC device with an impedance analyzer (FRA5097; NF Corporation, Kanagawa, Japan). The relationship between the input voltage and the vibration amplitude at the center of the ultrasonic transducer surface was measured with a laser Doppler vibrometer (LV-1800; Ono Sokki Co., Ltd., Kanagawa, Japan). The frequency of the input voltage was 1.007 MHz, and the amplitude of the input voltage was varied from 0 to 180 V at 20-V intervals. The distribution of the vibration amplitude on the bottom surface of the flask was also measured along the x and y axes [Fig. 1(B)]. For this measurement, the frequency and the amplitude of the input voltage were 1.007 MHz and 40 V, respectively. These measurements were conducted at points from −32 to 32 mm at 8 mm intervals along the x-axis and points from −8 to 8 mm along the y-axis relative to the center of the bottom of the flask.

Temperature variation of 100 mL of water subjected to ultrasonic exposure in the flask was measured with a temperature sensor (TR-71wf; T&D Corporation, Nagano, Japan). The measurements were conducted for 24 h while the system was running in a humidified 5% CO2 incubator (CPE-2201; Hirasawa Works, Inc., Tokyo, Japan) maintained at 36.5 °C. The frequency of the input voltage was 1.007 MHz, the amplitude was 40, 60, or 80 V, and the waveform of the ultrasonic irradiation was intermittent to inhibit heat generation. Regarding the ultrasonic waves, the exposure period was 7 s and the pause period was 10 s. For this measurement, the flask containing 100 mL of water was placed on the flask holder in a water tank containing 1 L of water. The temperature sensor was placed at x = 0 mm and z = 16 mm in the flask [cf. Fig. 1(B)].

The velocity of the flow in the flask generated by ultrasound irradiation was measured by particle image velocimetry (PIV). For this measurement, a flask containing 100 mL of culture medium was placed on the flask holder in a water tank containing 1.5 L of water. The frequency of the input voltage was 1.007 MHz and the amplitude of the input voltage was varied from 0 to 60 V at 20-V intervals. Note that the ultrasound was irradiated intermittently by repeating 7 s of exposure time and 10 s of a pause period. Fluorescent particles (FA-207; Sinloihi Co., Ltd., Kanagawa, Japan) with a diameter and density of 4 μm and 1300 kg/m3, respectively, were seeded in the culture medium as tracers. The particles are small enough to make negligible their slip (relative to the host liquid) caused by acoustic radiation force (Slama et al., 2017; Yamashita and Ando, 2019). To illuminate the particles, a PIV laser (G2000; Kato Koken Co., Kanagawa, Japan) of thickness at 1 mm and of wavelength at 532 nm was employed. The motion of the particles in the flask was recorded at 15 frames per second with 4 ms of exposure by high-speed camera units (VW-9000, VW-300M, and VWZ5; Keyence, Osaka, Japan). The field of view of the recording was set in the range from −32 to 32 mm along the x-axis and 0 to 32 mm along the z-axis [Fig. 1(B)]. The 2-min recording was performed 5 s before applying the irradiating ultrasound. PIV analysis was performed with the Flow Expert 2D2C software (Flow Expert 2D2C; Kato Koken Co., Kanagawa, Japan) by applying the direct cross-correlation method to the recorded images, which allowed the extraction of components of the flow velocity, w directed along the z-axis.

A Chinese hamster ovary (CHO) cell line (CHO-K1, RCB0403; Riken BioResource Center, Ibaraki, Japan) was used in the suspension culture experiments. CHO cells were selected as they are commonly used for protein production, and protein production is a major source of biopharmaceuticals. The CHO cells were first cultured on the bottom of a T25 flask (Nunc Cell Culture T25 EasYFlask; Thermo Fisher Scientific, Waltham, MA) in Ham's F-12 medium (α-MEM Ham's F-12; Wako, Tokyo, Japan) supplied with 10% fetal bovine serum (S1820; Biowest SAS, Nuaillé, France). Next, the CHO cells were suspended in protein-free CHO medium (HyQ PF-CHO MPS; GF Healthcare, Chicago, IL) supplemented with NaOH and L-glutamine in a 5% CO2 humidified-atmosphere incubator at 37 °C for the SSC to obtain a sufficient number of cells for the subsequent suspension culture experiment.

For the suspension culture experiment, 2.0 × 107 cells in 100 mL of medium were seeded in a T75 flask and cultured in the incubator under each of three conditions: conventional SSC, USC with an input voltage of 40 V (USC40V), and USC with an input voltage of 60 V (USC60V). The temperature of the atmosphere in the incubator was kept at 36.5 °C when using the USC method and at 37.0 °C for the SSC method, and the frequency of the AC input for the USC method was 1.007 MHz. For the USC groups, the waveform of the ultrasonic irradiation was intermittent with an exposure time and pause period of 7 and 10 s, respectively. After 24-h suspension culture, the number of dead cells was counted with an automated cell counter (TC20 Automated Cell Counter; Bio-Rad, Hercules, CA) using Trypan blue staining (Trypan Blue Solution 0.4%; Thermo Fisher Scientific, Waltham, MA). Additionally, a lactate dehydrogenase (LDH) assay was conducted to estimate the level of cell damage and death based on LDH release after 24 h of culture. For the LDH assay, 50 μL of the supernatant from each cell suspension was harvested and placed in a 96-well microtiter plate (92096; TPP Techno Plastic Products, AG, Trasadingen, Switzerland). Next, 50 μL of assay buffer provided with the Cytotoxicity LDH Assay Kit-WST (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) was added to each well, followed by incubation for 30 min. The absorbance of the samples was measured at 492 nm using a microtiter plate reader (Multiskan FC; Thermo Fisher Scientific, Waltham, MA).

Glucose consumption of the cells during the 24-h suspension culture under each condition was measured with the Glucose (HK) Assay Kit (GAHK20; Sigma-Aldrich, St. Louis, MO). Note that the glucose consumption per cell, C, was calculated with the following equation:

C=CtotalN24hNseed2,
(5)

where Ctotal, N24h, and Nseed represent the total glucose consumption after 24-h suspension culture, the number of cells after 24-h suspension culture, and the number of initially seeded cells, respectively.

Data were analyzed by two-way analysis of variance (ANOVA) to determine the impact of each culture condition. A p-value < 0.05 was considered statistically significant.

The relationship between the frequency of the AC input and the admittance of the ultrasound transducer with an input voltage of 1 Vp-p indicated that the resonance frequency of the ultrasound transducer was 1.007 MHz [Fig. 2(A)]. Thus, this frequency was used as the driving frequency of the USC device.

FIG. 2.

Vibration characteristics of the ultrasonic suspension culture system. (A) Relationship between admittance and the frequency of the AC input. (B) Relationship between the vibration amplitude at the center of the upper surface of the ultrasonic transducer and the input voltage. (C) Distribution of the vibration amplitude at the bottom surface of the flask. Data represent the mean ± standard deviation (SD), n = 3.

FIG. 2.

Vibration characteristics of the ultrasonic suspension culture system. (A) Relationship between admittance and the frequency of the AC input. (B) Relationship between the vibration amplitude at the center of the upper surface of the ultrasonic transducer and the input voltage. (C) Distribution of the vibration amplitude at the bottom surface of the flask. Data represent the mean ± standard deviation (SD), n = 3.

Close modal

The displacement amplitude at the upper surface of the ultrasound transducer was measured, as this parameter directly regulates the flow velocity of acoustic streaming. The relationship between the input voltage applied to the ultrasound transducer and the displacement amplitude demonstrated that the vibration displacement of the ultrasound transducer was proportional to the input voltage in the range from 0 to 100 V [Fig. 2(B)]. Thus, input voltages in the range of 0 to 100 V were used as the driving voltages of the USC device for subsequent experiments.

Measurement of the distribution of the vibration amplitude at the bottom surface of the flask showed that the vibration amplitudes in the area above the ultrasonic transducer were higher than those outside of this range [Fig. 2(C)]. Note that the diameter of the ultrasonic transducer was 40 mm. This result indicated that the ultrasound passed through the bottom of the flask and was effectively propagated into the flask.

Temperature variation during 24 h of ultrasound exposure was measured (Fig. 3). The relationship between the duration of ultrasound exposure and the temperature variation of water in a flask demonstrated that with an input voltage of 80 V, the temperature increased beyond 38 °C after 5 h from the beginning of ultrasound exposure. However, when using input voltages of 40 or 60 V, the temperature was kept below 38 °C. Since the temperature suitable for cell culture is 36 °C–38 °C (Wiklund, 2012), the input voltage of 80 V was not deemed suitable for 24-h cell culture with ultrasonic irradiation. Thus, for subsequent measurements, input voltages of up to 60 V only were used.

FIG. 3.

Temperature variation of water in a flask placed inside the ultrasonic suspension culture system. The flask was incubated for 24 h in a humidified 5% CO2 incubator at 36.5 °C with an applied voltage of 1.007 MHz and a voltage amplitude of 40, 60, or 80 V.

FIG. 3.

Temperature variation of water in a flask placed inside the ultrasonic suspension culture system. The flask was incubated for 24 h in a humidified 5% CO2 incubator at 36.5 °C with an applied voltage of 1.007 MHz and a voltage amplitude of 40, 60, or 80 V.

Close modal

To evaluate whether the cell cultures were being effectively mixed by the USC system, the flow velocity was measured using PIV analysis. The relationship between the duration of ultrasonic exposure and the z-component of flow velocity at the center of the recording range (x = 0, z = 16) suggested that when applying input voltages of 40 and 60 V, the flow velocity increased from the beginning of each ultrasonic pulse (7-s exposure period) and then decreased during the 10-s pause period [Fig. 4(A)]. Note that the ultrasound was irradiated intermittently by repeating a cycle of 7 s of exposure and 10 s of a pause period. Figures 4(B), 4(C), and 4(D) show the distributions of the z-components of the flow velocity at x = 0 mm, z =16 mm, and throughout the flow velocity field when applying different input voltages to the flask. Note that the data were obtained by analyzing the flow velocity at 74.93 s, which was the last time point before the fifth exposure period ended. The results indicated that while the liquid is essentially stagnant with an input voltage of 20 V, steady streaming was effectively generated with the higher voltage at 40 and 60 V and its maximum velocity was 13.94 and 42.97 μm/s, respectively. Thus, subsequent cell culture experiments were performed with input voltages of 40 and 60 V.

FIG. 4.

(Color online) Comparison of the distribution of the flow velocity when applying input voltages of 0, 20, 40, or 60 V. (A) Relationship between the flow velocity at the center of the recording range (x = 0, z = 16) and time. (B) Distribution of the flow velocity at x = 0 mm. (C) Distribution of the flow velocity at z = 16 mm. (D) Flow velocity field in the flask at different input voltages. Note that the data in (B)–(D) were obtained by analyzing the flow velocity at 74.93 s, which was the last time point before the fifth exposure period ended.

FIG. 4.

(Color online) Comparison of the distribution of the flow velocity when applying input voltages of 0, 20, 40, or 60 V. (A) Relationship between the flow velocity at the center of the recording range (x = 0, z = 16) and time. (B) Distribution of the flow velocity at x = 0 mm. (C) Distribution of the flow velocity at z = 16 mm. (D) Flow velocity field in the flask at different input voltages. Note that the data in (B)–(D) were obtained by analyzing the flow velocity at 74.93 s, which was the last time point before the fifth exposure period ended.

Close modal

Cell suspension culture was conducted using three different methods: conventional SSC, and the USC method with an input voltage of 40 V (USC40V) or 60 V (USC60V). The number of living cells observed in the USC40V system after 24 h of culture was significantly higher (p < 0.01) than that present in the other two systems, while there was no significant difference between SSC and USC60V [Fig. 5(A)]. Although there was no significant difference in the absolute number of dead cells between the three systems [Fig. 5(B)], the percentage of dead cells in the USC60V system tended to be higher than that in the other two systems. As another measure of cell damage and death, an LDH assay was performed after 24 h of culture and the results for the three culture methods were compared [Fig. 5(D)]. The absorbance of the USC60V culture tended to be higher than that of the others, indicating higher levels of cell damage and death. Thus, taken together, the results demonstrated that the USC40V system was the most conducive to cell proliferation.

FIG. 5.

Comparison of the proportion of live cells in 24-h cell suspension cultures prepared using three different methods: SSC, USC with an input voltage of 40 V (USC40V), and USC with an input voltage of 60 V (USC60V). (A) The number of living cells, (B) the number of dead cells, and (C) the percentage of dead cells were measured based on the TBE assay. (D) The absorbance of the cultures was measured to detect LDH release as an indicator of cell damage and death. Data represent the mean ± SD, n = 3, **p < 0.01.

FIG. 5.

Comparison of the proportion of live cells in 24-h cell suspension cultures prepared using three different methods: SSC, USC with an input voltage of 40 V (USC40V), and USC with an input voltage of 60 V (USC60V). (A) The number of living cells, (B) the number of dead cells, and (C) the percentage of dead cells were measured based on the TBE assay. (D) The absorbance of the cultures was measured to detect LDH release as an indicator of cell damage and death. Data represent the mean ± SD, n = 3, **p < 0.01.

Close modal

The glucose consumption of cells cultured with each method was analyzed and compared (Fig. 6). The glucose consumption of the SSC system was significantly higher (from p < 0.05 to p < 0.01) than that of the other two systems, with the USC40V system consuming the least glucose. A previous study demonstrated that the glucose consumption of cells increases in cultures with low oxygen levels (Lin et al., 1993). Thus, it is possible that more efficient homogenization of oxygen occurred in the USC systems compared with the SSC system, as a result of the lack of agitation in the latter.

FIG. 6.

Comparison of the glucose consumption of 24-h cultures prepared using three different methods: SSC, USC at 40 V (USC40V), and USC at 60 V (USC60V). Data represent the mean ± SD, n = 3, *p < 0.05, **p < 0.01.

FIG. 6.

Comparison of the glucose consumption of 24-h cultures prepared using three different methods: SSC, USC at 40 V (USC40V), and USC at 60 V (USC60V). Data represent the mean ± SD, n = 3, *p < 0.05, **p < 0.01.

Close modal

When scaling up a suspension culture system, culture transfer from general flasks to spinner flasks is required because of the volume limit of cell suspensions in traditional SSC. These transfer operations can increase the risk of contamination and human error. To address these issues, a method for inducing flow in the culture medium was developed here by applying ultrasonic irradiation at a MHz frequency. The proposed USC method developed in this study could contribute toward reducing the number of culture transfer steps, thereby helping to address the problems faced when scaling up suspension culture systems.

Evaluation of the mechanical characteristics of the USC system indicated that USC with an input voltage of 40 or 60 V could suspend the cells without resulting in heat-induced cell damage. When ultrasound propagates through the flask and culture medium, the temperature should increase as a result of the inevitable attenuation of the ultrasound. If the temperature exceeds the suitable range for cell culture (36 °C–38 °C), the cells may be damaged. Thus, the temperature variation of the culture medium was measured during 24-h ultrasound exposure and found that while the optimal temperature range for cell culture was maintained when applying the input voltages of 40 and 60 V, the temperature increased beyond 38 °C when applying an input voltage of 80 V. Additionally, while an input voltage of 20 V was not found to be sufficient to suspend the cells, 40 and 60 V were efficient in generating flow.

To confirm whether the drag forces induced by 40 and 60 V of input voltage could suspend the cells, numerical analysis was conducted (see  Appendix A). Based on this analysis, a flow velocity, w, of >5.372μm/s is required to suspend cells. Here, the flow velocities after 74.93 s of intermittent ultrasonic irradiation with 40 and 60 V of input voltage were found to be 13.94 and 42.97 μm/s, respectively. Because these flow velocities fulfill the above condition, USC with 40 and 60 V of input voltage has the potential to suspend cells without heat-induced damage.

When using the USC60V method, the number of living cells remaining after 24 h of culture was smaller than that observed for USC40V. There are two possible explanations for this phenomenon: an increased rate of cell death or a reduced rate of cell proliferation. Furthermore, how the cells die can be classified as either death from cell rupture or death from cell damage (no rupture). To evaluate cell death in the three culture systems, the Trypan blue dye exclusion (TBE) assay and LDH assay were conducted. With the TBE assay, dead cells are counted based on their cytoplasm being stained blue, which implies that dead cells that have ruptured cannot be counted with this method. In contrast, the LDH assay can estimate the extent of cell death and damage, including that of ruptured cells, by detecting LDH release. By comparing the results of these assays, the level of cell death could be fully evaluated. The USC60V method tended to induce cell death as indicated by both the TBE and LDH assays, while the USC40V method promoted cell proliferation. These findings suggested that the death of cells induced by the USC60V method was mainly through cell damage instead of cell rupture, and this damage was likely a result of the strength of the applied ultrasound irradiation.

Previous reports have indicated that ultrasound-induced stimulation of cells and tissue can be divided into four main factors: heat generation, acoustic radiation pressure, acoustic streaming, and acoustic cavitation (Dalecki, 2004; Haar, 2010; Wiklund, 2012). These factors might have induced cell death when using the USC60V method and we, therefore, examined each in turn.

In terms of heat generation, the culture temperature when applying an input voltage of 60 V was confirmed here to be suitable for cell culture. Next, to evaluate the effects of acoustic radiation pressure, the maximum acoustic radiation force of the traveling waves was calculated (see  Appendix B). This analysis indicated that the maximum acoustic radiation force of the traveling waves was 6.871 × 10−14 N in the USC system. In a previous study, cytoskeleton destruction was not induced even after exposure to 2.497 × 10−3 N of cyclic acoustic radiation force, which is much higher than that generated by the USC60V method (Zhang et al., 2012). Thus, it is unlikely that the acoustic radiation force in the USC system was sufficient to damage the cells. As for acoustic streaming, the maximum flow velocity in the USC60V system was not observed to exceed 2 mm/s. A previous study demonstrated that the flow velocity in spinner flasks, which are widely used for suspension culture, is approximately 10 mm/s (Ismadi et al., 2014; Liovic et al., 2012). Therefore, the level of acoustic streaming in the USC60V system was not sufficient to damage the cells. Acoustic cavitation is a vaporization process that may arise from pressure reduction below some critical values (Brennen, 2014; Leighton, 1994). In pure water, cavitation will be triggered under a very strong tension state, say, at −30 MPa (Herbert et al., 2006). However, the acoustic pressure amplitude in the culture medium when using the USC60V method can be predicted, based on Eq. (4), at 0.285 MPa. Here, the prediction is calculated with the maximum vibration amplitude at the bottom of the flask where the vibration amplitude is assumed to be proportional to the inputted voltage. This means that the liquid pressure can be negative (and thus below the saturated vapor pressure) under rarefaction cycles of the ultrasound, but will not reach the cavitation inception threshold in pure water. However, we speculate that cavitation can occur in our ultrasonic suspension culture system even with the low-intensity ultrasound irradiation, for heterogeneous cavitation nuclei (such as stabilized gas bubble nuclei of micron or even smaller) will exist in the cell suspensions saturated with the ambient air, thereby lowering the cavitation inception threshold. This speculation is supported by a previous study that cavitation occurred with 1 MHz ultrasound irradiation whose pressure amplitude is 0.169 MPa (Nguyen et al., 2017). In addition, cell media contains surfactants, which reduce the surface tension and make it easier to form gas bubbles (Rosen, 1979). It is likely that heterogeneous bubble nucleation and acoustic cavitation occur in our culture system under continuous cycles of the ultrasound irradiation, giving rise to cell damage. Acoustic cavitation can also induce decreased cell proliferation. When a cavitation bubble collapses adiabatically, the internal gas pressure and temperature get extremely high, which produces free radicals such as hydroxyl radicals (Milowska and Gabryelak, 2007; Ward et al., 1999). These radicals can cause DNA damage, which halts cell proliferation until the deoxyribose nucleic acid (DNA) damage is repaired (Boonstra and Post, 2004; Miller et al., 1995; Riesz and Kondo, 1992). From the above, the smaller number of cells produced with the USC60V method was concluded to be a result of acoustic cavitation.

To evaluate the efficiency of the USC40V method, the proliferation rate of cells cultured with the SSC and USC40V methods were compared. The proliferation rates of the SSC and USC40V methods were 1.14 and 1.88/day, respectively. Compared with the calculated proliferation rate of CHO-K1 cells based on the results of previous studies (Hayter et al., 1991; Nishijima et al., 2000), which range from 1.63 to 2.10/day, the proliferation rate of the SSC was low. Therefore, cell proliferation was suppressed in SSC when using 100 mL of cell suspension, indicating that 100 mL was in excess of the volume limit of cell suspension suitable for SSC. In contrast, the cell proliferation rate in the USC40V system was within the reported range. These findings indicate that the USC40V method could be used to extend the volume limit of the cell suspension used for SSC. In addition, the percentage of dead cells observed when using the USC40V method was 1%, and the absorbance measured with the LDH assay was low. These results clearly indicate that the USC40V method did not have a negative effect on cell viability. This implies that cavitation did not occur when using the USC40V method with which the acoustic pressure is lower and cavitation inception is thus less likely.

In this study, an ultrasonic suspension culture method was developed that employed a custom-fabricated system to generate acoustic streaming through the irradiation of high-frequency ultrasound. The results showed that this USC method could generate sufficient flow for cells to become suspended in the culture medium but without inducing cell damage. Using the optimized conditions determined here, the USC method could be used to produce suspension cultures that maintain cell proliferation even in a 100-mL culture volume—a volume at which cell proliferation is suppressed in conventional SSC. Thus, the USC method extends the current volume limit typically used for suspension cultures by approximately ten-fold. In other words, the result implies that until the volume of cell suspension reaches 100 mL, it is unnecessary to transfer cell suspension to a spinner flask. This contributes toward a reduction in the number of culture transfer steps. For example, to obtain 109 cells, the conventional scale-up process requires two transfer operations: first from a T-flask to a 125-mL spinner flask and then from a 125-mL to a 1-L spinner flask. On the other hand, using the USC method, the scale of suspension culture can be enlarged from small volume to 100 mL in one T-flask, and thus this scale-up process requires one transfer operation: from a T-flask to a 1-L spinner flask. This USC method could, therefore, contribute toward reducing the number of culture transfer steps and the attendant risks. It is possible that future optimization of additional parameters, such as the installation location of the ultrasonic transducer, will make it possible to perform even larger-scale ultrasonic suspension cultures. The advantages offered by this ultrasound suspension culture method will hopefully help to address the challenges currently faced when scaling up suspension culture systems.

Funding for this work was provided in part from a JSPS KAKENHI (Grant No. 20J00337 to C.I.). The authors have applied for a patent (PCT No. PCT/JP2018/025577) related to this manuscript.

To confirm whether the input voltages used in this study could suspend cells in the culture medium, the drag forces generated by medium flow were calculated when applying input voltages of 40 and 60 V. Note that these calculations were performed by assuming that a cell is a perfect sphere. When the Reynolds number of the flow generated by the USC method is low, the drag force on a cell generated by the medium flow, FD, is given by

FD=3πμmdw,
(A1)

where μm, d, and w represent the viscosity of the culture medium, the diameter of a single cell, and the flow velocity, respectively (White, 1979). When FD is larger than the sum of the forces of gravity and buoyancy, a cell is suspended by the flow. Based on the relationship between these forces and Eq. (1), to suspend the cells, the flow velocity, w, should fulfill the following condition:

w>cρm)g(πd3/6)3πμmd5.372μm/s,
(A2)

where ρc, ρm, g, μm, and d represent the density of the cells (1043 kg/m3) (Wolff and Pertoft, 1972), density of the culture medium (1010 kg/m3) (Fukuma et al., 2020), gravitational acceleration (9.810 m/s2), viscosity of the culture medium (1.350 × 10−3 N m/s) (Fukuma et al., 2020), and estimated diameter of a single cell (20.00 μm). From the results shown in Fig. 4(A), the flow velocities generated by input voltages of 40 and 60 V at 74.93 s, which was the last time point before the fifth exposure began, were 13.94 and 42.97 μm/s, respectively. Because these flow velocities fulfill the above condition, these calculations confirmed that USC with input voltages of 40 and 60 V has the potential to suspend cells without causing heat-induced cell damage.

The acoustic radiation force induced by the ultrasound was also calculated. Using King's equation (King, 1934), the acoustic radiation pressure of traveling waves, P, is given by

P=4πκ4R069+21ρcρm292+ρcρm2E,
(B1)

where κ, R0, ρc, ρm, and E represent the wavenumber of the ultrasound (κ=2πf/c4082/m), estimated cell radius (R0=10μm), density of the cells (ρc = 1043 kg m−3) (Wolff and Pertoft, 1972), density of the culture medium (ρm= 1010 kg/m) (Fukuma et al., 2020), and mean total energy density in the ultrasonic wave (E=2π2ρmf2A27.017J/m3). Note that the mean total energy density in the ultrasonic wave was calculated using 0.0297 μm as the maximum vibration amplitude at the bottom of the flask, A, which was an estimated value based on the assumption that the amplitude was proportional to the voltage. The calculation results indicated that the maximum acoustic radiation force of the traveling waves was 6.871 × 10−14 N.

1.
Blackstock
,
D. T.
(
2000
).
Fundamentals of Physical Acoustics
(
Wiley-Interscience
,
New York
).
2.
Boonstra
,
J.
, and
Post
,
J. A.
(
2004
). “
Molecular events associated with reactive oxygen species and cell cycle progression in mammalian cells
,”
Gene.
337
,
1
13
.
3.
Brennen
,
C. E.
(
2014
).
Cavitation and Bubble Dynamics
(
Cambridge University Press
,
New York)
, pp.
1
28
.
4.
Clynes
,
M.
(
1998
).
Animal Cell Culture Techniques
(
Springer
,
Berlin-Heidelberg, Germany)
, pp.
459
474
.
5.
Dalecki
,
D.
(
2004
). “
Mechanical bioeffects of ultrasound
,”
Ann. Rev. Biomed. Eng.
6
,
229
248
.
6.
Evaluate Pharma
(
2019
).
World Preview 2019, Outlook to 2024
(
Evaluate Ltd
.,
London
).
7.
Fujii
,
G.
,
Kurashina
,
Y.
,
Takahara
,
O.
,
Kodeki
,
K.
,
Morikawa
,
A.
,
Azuma
,
T.
, and
Takemura
,
K.
(
2018
). “
Cell suspension culture using acoustic flow
,” in
Proceedings of the 40th Annual International Conference IEEE Engineering in Medicine and Biology Society
, July 17–21, Honolulu, HI, p.
3040
.
8.
Fukuma
,
Y.
,
Inui
,
T.
,
Imashiro
,
C.
,
Kurashina
,
Y.
, and
Takemura
,
K.
(
2020
). “
Homogenization of initial cell distribution by secondary flow of medium improves cell culture efficiency
,”
PLoS One
15
,
e0235827
.
9.
Haar
,
G.
(
2010
). “
Ultrasound bioeffects and safety
,”
Proc. Inst. Mech. Eng. Part H
224
,
363
373
.
10.
Hayter
,
P. M.
,
Curling
,
E. M. A.
,
Baines
,
A. J.
,
Jenkins
,
N.
,
Salmon
,
I.
,
Strange
,
P. G.
, and
Bull
,
A. T.
(
1991
). “
Chinese hamster ovary cell growth and interferon production kinetics in stirred batch culture
,”
Appl. Microbiol. Biotechnol.
34
,
559
564
.
11.
Herbert
,
E.
,
Balibar
,
S.
, and
Caupin
,
F.
(
2006
). “
Cavitation pressure in water
,”
Phys. Rev. E
74
,
041603
.
12.
Imashiro
,
C.
,
Hirano
,
M.
,
Morikura
,
T.
,
Fukuma
,
Y.
,
Ohnuma
,
K.
,
Kurashina
,
Y.
,
Miyata
,
S.
, and
Takemura
,
K.
(
2020
). “
Detachment of cell sheets from clinically ubiquitous cell culture vessels by ultrasonic vibration
,”
Sci. Rep.
10
,
9468
.
13.
Imashiro
,
C.
,
Kurashina
,
Y.
,
Kuribara
,
T.
,
Hirano
,
M.
,
Totani
,
K.
, and
Takemura
,
K.
(
2019
). “
Cell patterning method on a clinically ubiquitous culture dish using acoustic pressure generated from resonance vibration of a disk-shaped ultrasonic transducer
,”
IEEE Trans. Biomed. Eng.
66
,
111
118
.
14.
Ismadi
,
M.
,
Hourigan
,
K.
, and
Fouras
,
A.
(
2014
). “
Experimental characterisation of fluid mechanics in a spinner flask bioreactor
,”
Processes
2
,
753
772
.
15.
Jackson
,
L. R.
,
Trudel
,
L. J.
,
Fox
,
J. G.
, and
Lipman
,
N. S.
(
1996
). “
Evaluation of hollow fiber bioreactors as an alternative to murine ascites production for small scale monoclonal antibody production
,”
J. Immunol. Methods
189
,
217
231
.
16.
Jing
,
D.
,
Parikh
,
A.
,
Canty
,
J. M.
, and
Tzanakakis
,
E. S.
(
2008
). “
Stem cells for heart cell therapies
,”
Tissue Eng. Part B
14
,
393
406
.
17.
Kim
,
K. H.
,
Franc
,
J. P.
, and
Karimi
,
A.
(
2014
).
Advanced Experimental and Numerical Techniques for Cavitation Erosion Prediction
(
Springer
,
New York
),
pp.
21
34
.
18.
King
,
L. V.
(
1934
). “
On the acoustic radiation pressure on spheres
,”
Proc. R. Soc. Lond.
147
,
212
240
.
19.
Kurashina
,
Y.
,
Hirano
,
M.
,
Imashiro
,
C.
,
Totani
,
K.
,
Komotori
,
J.
, and
Takemura
,
K.
(
2017
). “
Enzyme-free cell detachment mediated by resonance vibration with temperature modulation
,”
Biotechnol. Bioeng.
114
,
2279
2288
.
20.
Kurashina
,
Y.
,
Imashiro
,
C.
,
Hirano
,
M.
,
Kuribara
,
T.
,
Totani
,
K.
,
Ohnuma
,
K.
,
Friend
,
J.
, and
Takemura
,
K.
(
2019
). “
Enzyme-free release of adhered cells from standard culture dishes using intermittent ultrasonic traveling waves
,”
Commun. Biol.
2
,
393
.
21.
Kuriyama
,
T.
,
Fukuma
,
Y.
,
Imashiro
,
C.
,
Kabayama
,
K.
,
Kurashina
,
Y.
, and
Takemura
,
K.
(
2020
). “
Detachment of RAW264.7 macrophages from a culture dish using ultrasound excited by a Langevin transducer
,”
J. Biosci. Bioeng.
131
,
320
325
.
22.
Leighton
,
T. G.
(
1994
).
The Acoustic Bubble
(
Academic Press
,
New York)
, pp.
67
125
.
23.
Lighthill
,
S. J.
(
1978
). “
Acoustic streaming
,”
J. Sound Vib.
61
,
391
418
.
24.
Lin
,
A.
,
Kimura
,
R.
, and
Miller
,
W. M.
(
1993
). “
Production of tPA in recombinant CHO cells under oxygen-limited conditions
,”
Biotechnol. Bioeng.
42
,
339
350
.
25.
Liovic
,
P.
,
Šutalo
,
I. D.
,
Stewart
,
R.
,
Glattauer
,
V.
, and
Meagher
,
L.
(
2012
). “
Fluid flow and stresses on microcarriers in spinner flask bioreactors
,” in
Proceedings of the 9th International Conference on CFD Minerals and Process Industries
, December 10–12, Melbourne, Australia, pp.
1
6
.
26.
Miller
,
D. L.
,
Thomas
,
R. M.
, and
Buschbom
,
R. L.
(
1995
). “
Comet assay reveals DNA strand breaks induced by ultrasonic cavitation in vitro
,”
Ultrasound Med. Biol.
21
,
841
848
.
27.
Milowska
,
K.
, and
Gabryelak
,
T.
(
2007
). “
Reactive oxygen species and DNA damage after ultrasound exposure
,”
Biomol. Eng.
24
,
263
267
.
28.
Mitome
,
H.
(
1998
). “
The mechanism of generation of acoustic streaming
,”
Electron. Commun. Jpn. Part III: Fundam. Electron. Sci.
81
,
1
8
.
29.
Nakao
,
M.
,
Imashiro
,
C.
,
Kuribara
,
T.
,
Kurashina
,
Y.
,
Totani
,
K.
, and
Takemura
,
K.
(
2019
). “
Formation of large scaffold-free 3-D aggregates in a cell culture dish by ultrasound standing wave trapping
,”
Ultrasound Med. Biol.
45
,
1306
1315
.
30.
Nakao
,
M.
,
Kurashina
,
Y.
,
Imashiro
,
C.
, and
Takemura
,
K.
(
2018
). “
A method for collecting single cell suspensions using an ultrasonic pump
,”
IEEE Trans. Biomed. Eng.
65
,
224
231
.
31.
Nguyen
,
T. T.
,
Asakura
,
Y.
,
Koda
,
S.
, and
Yasuda
,
K.
(
2017
). “
Dependence of cavitation, chemical effect, and mechanical effect thresholds on ultrasonic frequency
,”
Ultrason. Sonochem.
39
,
301
306
.
32.
Nishijima
,
K.
,
Fujiki
,
T.
,
Kojima
,
H.
, and
Iijima
,
S.
(
2000
). “
The effects of cell adhesion on the growth and protein productivity of animal cells
,”
Cytotechnology
33
,
147
155
.
33.
Post
,
M. J.
(
2012
). “
Cultured meat from stem cells: Challenges and prospects
,”
Meat Sci.
92
,
297
301
.
34.
Post
,
M. J.
,
Levenberg
,
S.
,
Kaplan
,
D. L.
,
Genovese
,
N.
,
Fu
,
J.
,
Bryant
,
C. J.
,
Negowetti
,
N.
,
Verzijden
,
K.
, and
Moutsatsou
,
P.
(
2020
). “
Scientific, sustainability and regulatory challenges of cultured meat
,”
Nat. Food
1
,
403
415
.
35.
Riesz
,
P.
, and
Kondo
,
T.
(
1992
). “
Free radical formation induced by ultrasound and its biological implications
,”
Free Radical Biol. Med.
13
,
247
270
.
36.
Rosen
,
J. M.
(
1979
).
Surfactants and Interfacial Phenomena
(
John Wiley & Sons, Inc
.,
New York)
, pp.
208
242
.
37.
Slama
,
R. B. H.
,
Gilles
,
B.
,
Chiekeh
,
M. B.
, and
Béra
,
J. C.
(
2017
). “
PIV for the characterization of focused field induced acoustic streaming: Seeding particle choice evaluation
,”
Ultrasonics
76
,
217
226
.
38.
Tauchi
,
H.
,
Imashiro
,
C.
,
Kuribara
,
T.
,
Fujii
,
G.
,
Kurashina
,
Y.
,
Totani
,
K.
, and
Takemura
,
K.
(
2019
). “
Effective and intact cell detachment from a clinically ubiquitous culture flask by combining ultrasonic wave exposure and diluted trypsin
,”
Biotechnol. Bioprocess Eng.
24
,
536
543
.
39.
Terao
,
Y.
,
Kurashina
,
Y.
,
Tohyama
,
S.
,
Fukuma
,
Y.
,
Fukuda
,
K.
,
Fujita
,
J.
, and
Takemura
,
K.
(
2019
). “
An effective detachment system for human induced pluripotent stem cells cultured on multilayered cultivation substrates using resonance vibrations
,”
Sci. Rep.
9
,
15655
.
40.
Ward
,
M.
,
Wu
,
J.
, and
Chiu
,
J.
(
1999
). “
Ultrasound-induced cell lysis and sonoporation enhanced by contrast agents
,”
J. Acoust. Soc. Am.
105
,
2951
2957
.
41.
White
,
F. M.
(
1979
).
Fluid Mechanics
, 8th ed. (
McGraw-Hill Education
,
New York)
, pp.
407
454
.
42.
Wiklund
,
M.
(
2012
). “
Acoustofluidics 12: Biocompatibility and cell viability in microfluidic acoustic resonators
,”
Lab Chip
12
,
2018
2028
.
43.
Wolff
,
D. A.
, and
Pertoft
,
H.
(
1972
). “
Separation of Hela cells by colloidal silica density gradient centrifugation: I. Separation and partial synchrony of mitotic cells
,”
J. Cell Biol.
55
,
579
585
.
44.
Yamashita
,
T.
, and
Ando
,
K.
(
2019
). “
Low-intensity ultrasound induced cavitation and streaming in oxygen-supersaturated water: Role of cavitation bubbles as physical cleaning agents
,”
Ultrason. Sonochem.
52
,
268
279
.
45.
Zhang
,
S.
,
Cheng
,
J.
, and
Qin
,
Y.
(
2012
). “
Mechanobiological modulation of cytoskeleton and calcium influx in osteoblastic cells by short-term focused acoustic radiation force
,”
PLoS One
7
,
e38343
.
46.
Zweigerdt
,
R.
,
Olmer
,
R.
,
Singh
,
H.
,
Haverich
,
A.
, and
Martin
,
U.
(
2011
). “
Scalable expansion of human pluripotent stem cells in suspension culture
,”
Nat. Protoc.
6
,
689
700
.