Many methods for culturing 3D spheroids have been studied for decades. Conventional methods are represented by a method using magnetic field, hanging drop, aggrewell, and rotating cells [1–4]. The spheroids formed by these methods have been known to improve intercellular interactions, proliferative capacity, and gene expression [1, 2]. However, the conventional methods have two limitations. One is that the formation time of spheroids in the culture process takes at least 2 days [5–7, 9, 30, 31]. In addition, a complicated device or a labor-intensive process is required for the culture process. In a case of hanging drop method, setting for incubation is necessary drop by drop [1, 5]. Methods that utilize electromagnetic fields or use rotation of cells require the arrangement of complex devices for culturing cells. And it is necessary to fine-tune the rotational speed, arrangement of cells [6–9]. To overcome these limitations, we used an anti-gravity bioreactor using acoustic levitation method.
The anti-gravity bioreactor was designed based on the levitation by a one-dimensional standing wave from a single acoustic source. Therefore, the bioreactor has a cylindrical shape with a piezoelectric transducer as a sound source at the bottom and a reflector at the top. Since the natural frequency of the transducer and the density of the cell medium are fixed constants, the only variable for the formation of an acoustic standing wave is the spacing between the reflector and the actuator. In theory, a standing wave occurs if the spacing is an integer multiple of half the acoustic wavelength. In consideration of the volume of medium required for cell culture, the appropriate spacing for the acoustic standing wave was estimated within 45 to 50 mm. through computational calculation. As a result of the calculation, the most uniform and largest amplitude standing wave was formed at a spacing of 47.3 mm.
Given the optimal acoustic standing wave conditions, it was calculated how fast single cells were trapped; the trapping of cells into the node was completed within 1 ms. This is the first step in the formation of spheroids, and it can be understood as the effect of bringing the cells injected into the bioreactor close to each other near the nodes. Meanwhile, the cells trapped in the nodes grow into spheroids over time, however their size is limited by the dimensional size of a given acoustic standing wave. Theoretically, in order to predict the size of the spheroid allowed in the acoustic standing wave of the present work, the size of the particle in which the acoustic levitation can be stably maintained was calculated. Among the particles of 100 to 500 µm diameter set in the calculation, it was found that particles with 300 µm diameter or larger did not maintain acoustic levitation, and those with less than 300 µm were levitated as stable.
Based on these theoretical calculation results, the actual anti-gravity bioreactor was manufactured. Since temperature maintenance is important during cell culture, the material of the holder of the piezoelectric transducer, which is a heat source, is made of aluminum, and the temperature can be controlled by using a Peltier cooler at the bottom. In addition, the power source of the piezoelectric transducer’s driving circuit was supplied from the outside of the incubator to minimize the volume of the bioreactor.
11.4 ml medium was used to match the previously optimized reflector and transducer distance 47.3 mm. The number of hMSCs optimized by subsequent experiments was 1 × 106 cells and the incubation time was 12 hr. 12V is supplied for the operation of the anti-gravity bioreactor (Fig. 2E). The transducer generates heat during operation [32, 33]. This heat raises the temperature of the medium. A fan was attached to the bottom of the transducer to avoid temperature increase during cell culture in this study. The optimal temperature without damage in the cell culture was around 37 oC [34]. Anti-gravity bioreactor increased to 36 oC within 1 hr of operation. And most of the incubation time was maintained at around 37 oC (Figs. 2F and G). Therefore, there was no damage to cells due to temperature during use of the device. During operation, the acoustic levitation causes the cells to be cultured in a row in the center (Fig. 2H).
hMSCs were cultured for 12 or 24 hr, which is shorter than the conventional methods. Incompletely filled spheroids had a hollow part in the middle or were very large, so the inside was less aggregated. The size of the spheroids differed depending on the culture method rather than the incubation time. When cultured by hanging drop, the size was significantly larger than that of the spheroids cultured in the anti-gravity bioreactor (Fig. 3C). The size of the spheroids cultured with hanging drop was significantly larger because the inside was not filled like the previous optical images. In the low magnification SEM image, the middle of spheroid was not completely filled in the hanging drop group like the previous data. In the SEM image at high magnification, many hMSCs were not completely aggregated on the surface of the spheroid. In the spheroids cultured in the anti-gravity bioreactor, hMSCs were completely aggregated at both low and high magnifications (Fig. 3D). After harvesting, it was confirmed whether additional aggregation of hMSCs occurred in additional culture. In the Sph-HD, the hMSCs were compressed into the empty spaces inside and the size was reduced after 12 hr. However, the size of Sph-AG group did not change after the spheroids were completely aggregated even before additional incubation (Figs. 3E and F). Sph-AG were formed within 12 hr, but the Sph-HD were not complete spheroids until 12 hr. Gap junction between cells and intercellular adhesion were evaluated with Cx43, F-actin, and E-cadherin. The expression of Cx43 was significantly elevated only in spheroids cultured in the anti-gravity bioreactor (Fig. 3G). Elevated Cx43 expression leads to reinforcement of gap junctions. The distribution of F-actin was confirmed by phalloidin staining, and the expression of E-cadherin was confirmed by immunostaining. The colors of phalloidin and E-cadherin appeared bright in the entire spheroid cultured in the anti-gravity bioreactor (Figs. 3H and I). The increase in F-actin distribution and E-cadherin expression is a phenomenon that appears as results of increased cell-cell adhesion. Therefore, the cell-cell adhesion was relatively well established in the anti-gravity bioreactor group compared to the other group. H&E staining images also confirmed that the arrangement of hMSCs was more compact in the anti-gravity bioreactor group (Fig. 3J). In anti-gravity group, all factors related to the density of spheroids were increased.
When cultured in the anti-gravity bioreactor for 12 hr, 95.33% of the hMSCs were used as spheroids, and when cultured for 24 hr, 94.25% or cells were used (Fig. 4A). There was no difference in the formation efficiency of the cells used as spheroids compared to the conventional spheroid studies. To investigate whether the acoustic levitation exhibited the cytotoxicity, we confirmed the viability of hMSCs by several methods. The expression of Caspase-3, a pro-apoptosis factor, was checked to determine the damage to hMSCs. There was no difference in the expression of Caspase-3 among all groups and there was no toxicity induced by the acoustic levitation (Fig. 4B). The expression of Caspase-3 and Bcl-2, a pro-apoptosis factor and anti-apoptosis factor was also confirmed by western blot [35, 36]. The Caspase-3 and BCL-2 data showed by western blot also supported the absence of toxicity (Fig. 4C). EB+ cells were not detected, and there was no dead signal in TUNEL assay (Figs. 4D and E).
There were genes which expression was changed in the spheroids formed in the anti-gravity bioreactor faster than the conventional methods. VEGF, an angiogenic factor frequently secreted from spheroids, was secreted more in the anti-gravity bioreactor group than in other groups (Fig. 5A). In addition, other factors related to angiogenesis, IGF-1, and ANGPT2, were also significantly increased in the anti-gravity bioreactor group (Figs. 5B and C). Although the formation time was short, an increase in angiogenic factor, which is one of the characteristics of spheroids, was confirmed. It is also confirmed that the expression levels of p16 and p21 were decreased, and PCNA was increased in Sph-AG group, indicating that the cells in Sph-AG might secrete paracrine factors continuously with maintaining proliferation and avoiding senescence (Figs. 5D-G). Further study is needed to reveal the mechanisms of the upregulation of these factors, which may be attributed to the rapid formation of cell-cell interaction in the Sph-AG group.
The treatment effects of spheroids were confirmed by the injection into hindlimbs of mice after surgery. Therapeutic efficacy was classified into 4 categories with the naked eyes: limb salvage, toe necrosis, foot necrosis, and limb loss [37]. In the NT and 2D culture groups, there the most ofmice underwent limb loss. In contrast, the Sph-HD and Sph-AG groups blocked the progression of limb loss (Figs. 6A and B). Furthermore, some of the mice in the Sph-AG group showed limb salvage after the injection. Inflammation and muscle degeneration were observed in all groups except the Sph-AG group. However, the muscles were regenerating like the normal group in the Sph-AG group (Fig. 6C). Muscle degeneration was confirmed in H&E staining. Inflammation was confirmed by purple in H&E staining, blue in Masson’s trichrome staining and red in picro sirius red staining. The gene and protein expression levels of CD31 and α-SMA, which are factors involved in vascular regeneration, were also increased in the Sph-AG group compared to no treatment group (Figs. 6D-G). Taken together, upregulated paracrine factors in Sph-AG group elicited enhanced therapeutic efficacy in mouse hindlimb ischemia model.