Biomechanical Response of Cancer Stem Cells to Low-Intensity Ultrasound

Cancer has become the main cause of death in the world [1]. Although a variety of treatment strategies have dramatically increased the survival in cancer patients, drug resistance and metastasis are still major challenges to hinder cancer treatment [2]. Currently, evidences suggested that when a subset of cancer cells possesses stem cell-like features, including self-renewal, differentiation, tumorigenesis, and tumor heterogeneity, often referred to as cancer stem cells (CSCs) [3]. These cells have been considered to be the main reasons of drug resistance and metastasis of cancer [4]. Therefore, therapies that can produce the exciting therapeutic effects to these cells may eventually lead to cancer cures.

Nowadays, one of the therapeutic strategies is to inhibit the self-renewal ability of CSCs [5,6]. For example, the original properties of CSCs could be altered to promote the transformation of CSCs from the “stem” phenotype to “differentiated” phenotype by interfering stem cell signaling [7]. However, the internal regulation of directly inhibiting or activating the master regulators may affect the functions of normal stem cells and cause systemic toxicity. Except for regulating the functions of CSCs by internal channels, external regulations also play an important role in affecting the “stem” phenotype of CSCs [8,9]. The external regulations are mainly produced by mechanical stimulation, which possess the ability to regulate cell behavior and functions [10,11]. Rahmani et al. proved that electrical stimulation could contribute to mesenchymal stem cell differentiation [12]. Sun et al. also reported that shear stress could alter the biological features of liver CSCs to influence invasiveness [13]. Therefore, there may be an important association between mechanical stimulation and drug resistance and migration of CSCs.

Nowadays, a variety of mechanical stimuli have been used in targeting CSCs [14]. Low-intensity ultrasound (LIUS,<5 W/cm²) as a safe, noninvasive, and cost-effective mechanical stimulation [15,16], which can affect cell stiffness and interfere with the expression of cytoskeletal elements in cancer cells [17,18]. Basing on these changes in biomechanical properties, many investigations have shown that LIUS could enhance the anticancer effect of chemotherapeutic drugs (including doxorubicin, docetaxel, cisplatin) [19–21]. Also, it has been proven that LIUS could inhibit cell invasion and clone formation, thus showing great potential in cancer therapy [22]. However, the biomechanical relationships between LIUS and drug resistance and migration of CSCs have not been explored enough. Further exploration of these reasons may contribute to a new angle to understand the mechanics of cancer cells occurrence and provide a referenced basis for targeting CSCs in the clinical.

In this study, we aimed to unravel how LIUS, as a form of mechanical stimulation, affected the biomechanical properties to alter drug resistance and migration of CSCs. We believed that these findings would be helpful in providing suggestions for translational researches on cancer therapy.

Sphere-forming cells of ovarian cancer stem cells (OCSCs) were derived from the A2780 cells. A2780 cells were cultured in serum-free medium until the formation of spheroids. The spheroids cells were sorted by ALDEFLUOR assay and fluorescence-activated cell to obtain OCSCs in our previous works [23,24]. These OCSCs were used for follow-up experimental studies.

The ultrasound device of Sonvitro was applied in this research, which had been introduced in our previous study [25]. In brief, it had an ultrasonic transducer, a water tank and a control panel. The frequency of the ultrasonic probe was 1 MHz. The ultrasound parameters of different exposure times, duty cycles, and ultrasound intensity could be chosen through the control panel. In this study, the fixed ultrasonic parameters were 1 min, 20%. Furthermore, the ultrasound frequency of 1 MHz could induce the bio-effects in cancer stem cells, which could mediate the targeted gene into the cancer stem cells in our previous study [24,26]. Therefore, the ultrasound frequency of 1 MHz was chosen for this research. For ultrasound exposure, sterilized distilled water was filled between the transducer and the cell culture plate for ultrasound transmission.

The migration of OCSCs and A2780 cells was assessed by transwell insert with polycarbonate membranes of 8 μm pore size. Briefly, the A2780 cells and OCSCs were implanted in six-well plates. After LIUS exposure, the cells were collected with serum-free medium and seeded into the inner part of the inserts. The inserts were placed into a 24-well culture plate containing 600 μL medium with 1% FBS as the chemoattractant. After 6 h incubation at 37 °C in 5% CO₂, migratory cells would adhere to the membrane of the lower chamber which could ensure the accuracy of cells migration and all nonmigration cells were scraped from the upper surface of the membrane with a cotton swab, and migrated cells which remained on the bottom surface were stained with 0.05% crystal violet (Solarbio, Beijing, China) in phosphate buffered saline (PBS) for 30 min. Cells migration were assessed by randomly selecting 3 fields to count the number of the cells in each field.

OCSCs and A2780 cells in the exponential phase were collected, resuspended in complete culture medium at the required cell density, placed in culture dishes to 70% of confluence, and were then randomly divided into three groups: (1) 0.6 W/cm², (2) 1.0 W/cm², and (3) 1.4 W/cm². Every sample had three replicates. Cells in different groups were exposed to LIUS at different intensities mentioned above. Any bio-effect was supposed to be considered nonthermal in this study. Thus, the temperature increased inside the culture plates was measured before and after LIUS treatment with a digital thermometer, and no significant variation of temperature was detected (±1 °C). Subsequently, cells were incubated for additional time as specified and subjected to the following analysis. Cell viability was evaluated by the CCK-8 test (BestBio, Shanghai, China) at 2 h after different ultrasound intensity exposure.

The drug-resistance ability of A2780 cells and OCSCs was reflected by cell viability. First, A2780 cells and OCSCs were cultured in 24-well plates at a density of 2 × 10⁵ cells/well. Then, the cells were cultured in serum-free culture with LIUS exposure, and cisplatin (Sigma, St. Louis, MI) was added to the cells for 24 h. The cell viability of A2780 cells and OCSCs was detected by CCK8-assay detection kit following the manufacturer's instructions.

To observe the morphological changes of OCSCs, different ultrasound parameters were used to treat the OCSCs and A2780 cells. The cells were immediately fixed with 2.5% glutaraldehyde for 1 h and then washed twice by PBS. Cell samples were dehydrated with graded alcohol and dried with a freeze dryer. Finally, cell samples treated in the previous step were coated on the metal net, dried after gold plating, and were observed under a scanning electron microscope (SEM, S-3700N, Hitachi, Tokyo, Japan).

A2780 cells and OCSCs were implanted in six-well plates. After LIUS exposure, cells were fixed with 4% formaldehyde for 5 min and washed immediately with PBS for three times. The fixed cells were permeabilized with 0.1% TritonX-100 for 5 min and rewashed three times with PBS. After saturation with 5% bovine serum, the cells were incubated with FITC-phalloidin (BestBio, Shanghai, China) for 1 h at room temperature. The nuclei were labeled with 10 μg/ml 4,6-diamidino-2-phenylindole (Solarbio, Beijing, China) for 2 min. Finally, these cells were observed by confocal laser scanning microscope.

The A2780 cells and OCSCs were cultured onto glass coverslips at a density of 5 × 10⁴ cells/well. In order to evaluate the change of membrane stiffness after LIUS exposure, an atomic force microscope (Dimension Fastscan Bio, Germany) with a pyramid probe could be used to slowly contact with the cell surface for the measurement of cell stiffness. The data of the force–distance curves were recorded by nanoscope_analysis software. The elasticity modulus of these cells was calculated based on the Hertz model, which was described as the elastic deformation of two bodies in contact under load according to previous reports [27]. All measurements were performed at 37 °C, and cells had been cultured in PBS buffer. The local elasticity of these cells was quantitatively determined from the force–distance curves.

All data were expressed as mean±standard deviation. Statistical tests were analyzed by GraphPad prism 8.0. The data of each group were analyzed by One-way Analysis of Variance, and the significant data were analyzed between the two groups by the t-test. Significant differences were indicated as follows: *p < 0.05, **p < 0.01, and ***p < 0.001.

In our study, the monolayers of adherent A2780 cells were cultured in the presence of 10% FBS (Fig. 1(a)) and subsequently in tumor sphere medium. After 13 days, the adherent monolayer A2780 cells became suspended and spherical (Fig. 1(a)). In order to prove that the spherical cells originated from the A2780 cells, the whole evolution process of a single A2780 cell forming a sphere was recorded on day 1, 5, 7, 9, 11, and 13 (Fig. 1(b)). With the extension of culture time, the tumor spherical volume was obvious and displayed self-renewing.

Next, in order to verify whether LIUS could reverse the drug-resistance of CSCs, the cell viability between treatment with LIUS alone or LIUS combined with cisplatin was compared. Images in Fig. 2(a) showed that LIUS exposure inhibited A2780 cells and OCSCs viability in an intensity-dependent manner and OCSCs were more tolerant than A2780 cells. As shown in Fig. 2(b), the cell viability of A2780 cells and OCSCs decreased with the increase of ultrasound intensity and chemotherapy drug intervention. However, compared with A2780 cells, the OCSCs were more sensitive to chemotherapeutic drugs and the viability of OCSCs was lower under LIUS exposure (Table 1).

In order to verify whether LIUS stimulation could affect the migration of CSCs, the migration behavior was observed by transwell assay after LIUS exposure. Interestingly, these results showed that the migration ability of A2780 cells was higher than OCSCs, when the ultrasound intensity exceeded 0.6 W/cm² (Figs. 3(a) and 3(b)). Furthermore, compared to the control group, the migrated amounts of OCSCs decreased by 4.1%, 15.9%, and 26% when ultrasound intensity was 0.6, 1.0, and 1.4 W/cm², respectively (Fig. 3(c)).

In this experiment, the changes of cell morphology were observed by SEM after LIUS exposure. When A2780 cells and OCSCs were exposed to LIUS with an intensity of 0.6 W/cm², there were no significant morphological alterations compared with the control group. However, with ultrasound intensity increasing, the A2780 cells and OCSCs both showed different degrees of shrinkage of the cell membrane. OCSCs showed more significant morphological shrinkage to abnormal circular shape at an ultrasound intensity of 1.4 W/cm². Furthermore, there were many protuberances on the surface of OCSCs, which made the plasma membranes uneven, blurred the boundary and more pores formed (Figs. 4(a) and 4(b) and Table 3).

The dynamic changes of filament actin (F-actin) can affect the motility and stiffness of cells to some extent [28]. Therefore, it is worth studying whether LIUS could affect F-actin and produce a corresponding biomechanical response to A2780 cells and OCSCs. The FITC-conjugated phalloidin was chosen to stain the F-actin and observing the formation of F-actin by confocal microscopy. We found that the intensity of the fluorescence of FITC-phalloidin gradually decreased in A2780 cells when the ultrasound intensity was 0.6, 1.0, and 1.4 W/cm², respectively (Fig. 5(a)). However, LIUS intervention increased the intensity of the FITC-phalloidin in OCSCs (Fig. 5(b)).

Young's modulus is proportional to cell stiffness, so we utilized Young's modulus as the criterion to evaluate the variation of cell stiffness. After LIUS exposure, there was a significant difference in cell stiffness between A2780 cells and OCSCs. With the increase of ultrasound intensity, the stiffness of A2780 cells decreased in an ultrasound intensity-dependent manner, while the stiffness of OCSCs increased (Table 4). Furthermore, when ultrasound intensity was above 0.6 W/cm², the OCSCs were stiffer than A2780 cells. For example, the Young's modulus of OCSCs was higher by 1.76-fold than that of A2780 cells with 1.4 W/cm² ultrasound (Fig. 6).

In this paper, we focused on illustrated whether LIUS could affect the drug resistance and migration of CSCs, and to analyze the relationships between LIUS and CSCs from the perspective of biomechanical properties. Previous researches had demonstrated that LIUS could enhance drug delivery and promote stem cell regulation [29,30]. In addition, there were also reports showed that some mechanical forces could affect the behavior and function of CSCs by changing the biomechanical properties [31]. These results prompted us to confirm the influences of LIUS in CSCs. In our study, we found that the OCSCs had higher survival rates than that of A2780 cell after ultrasound exposure (Fig. 2(a)), which illustrated that LIUS could cause less lethal effects for OCSCs. However, the percentages of dead OCSCs increased more significantly than A2780 cells when LIUS combined with chemotherapy drug (Fig. 2(b)). As we all know, the CSCs usually had higher drug resistance than normal cancer cells due to the drug resistant genes and transporters overexpression [32]. The LIUS may lead to these associated genes and proteins had low expression in OCSCs, which made OCSCs more sensitive to chemotherapeutic drugs than A2780 cells. Moreover, we also found that the number of migratory OCSCs were gradually reducing with an ultrasound intensity increased (Fig. 3(b)). Although, the cells vitality could be affected under LIUS exposure, under the same exposure conditions of LIUS, the cells vitality of OCSCs was higher than that of A2780 cells, the amounts of OCSCs migration were less than A2780 cells. These results were consistent with previously reports [33,34]. We speculated that the abilities of drug resistance and migration of cancer cells were mainly focused on CSCs (Table 2). Hence, we compared the changes in biomechanical properties between A2780 cells and OCSCs after LIUS exposure to provide more information for the reversal of “stem” phenotype in CSCs.

Cell membrane is often regarded as the first barrier against various harmful stimuli and becomes an important target for ultrasound-induced action [35]. Cell membrane structure and function changes can affect drug penetration and cells migration [36,37]. Thus, exploring the relationships between LIUS and cell membrane may provide important information regarding the changes of drug resistance and migration of CSCs. In this study, when the ultrasound intensity was 1.0 W/cm² or 1.4 W/cm², it could be clearly observed that the cell membrane morphology of OCSCs was obviously shrunken, rough and pores compared to A2780 cells (Fig. 4). Jia et al. putted forward that the LIUS could improve drug uptake of cells and the membrane fluidity change was considered the main reason for this change. However, Guo et al. found that ultrasound could not only enhance the fluidity of cell membrane, but also reduce the expression level of ABCG2, which could decrease the drug dosage and significantly promote the stem-cell-like features changes of CSCs [38]. Meanwhile, some studies had shown that cell morphology could decide mesenchymal stem cells fate [39]. McBeath et al. used tensional forces to alter stem cell morphology, which could drive signaling cascades and lead to specific gene expression to affect differentiation [40]. But Li et al. considered that cell morphology damage would destruct the homeostasis of actin cytoskeleton, which resulted in the events adjustment, including metastasis, motility and differentiation [41]. It can be seen that the change of cell morphology may be a signal of differentiation of CSCs, which would promote us to further study.

Damage of the plasma membrane and cell morphology changes both can lead to adjustment of the cytoskeleton [42]. As an internal polymer to coordinate the cell and external environment, the cytoskeleton can change the cell shape and behavior characteristics [43]. F-actin is the main component of cytoskeleton, which can be isolated and recombined under external stimulation to reflect the mechanical properties of cells [44,45]. Some studies had pointed that the cytoskeleton could receive external signals of mechanical stimulation that accomplish complex behaviors such as lamellipodia formation, and migration [46]. Further, some key regulators such as RhoA, which is mainly involved in the remodeling of cytoskeleton and regulating cell stiffness and adhesion mechanisms to regulate cell migration [47]. In our study, we found that F-actin formation of OCSCs demonstrated a more significant increase compared to A2780 cells. With the ultrasound intensity increasing, the F-actin content of A2780 cells was reducing, while OCSCs showed the opposite result. (Fig. 5). Sun et al. obtained similar conclusion of salinomycin could increase the cytoskeleton formation in CSCs to suppress invasion [48].

Stiffness is an important mechanical feature to reflect the motility of cells, which is mainly determined by the cytoskeleton [49]. Basing on the current evidences, it is generally believed that cells with higher malignant degree are essentially softer than that have lower malignant degree [50], due to being soft is conducive to deformation and may increase the ability of migration. In our study, we observed that the A2780 cells had higher cell stiffness than OCSCs when the ultrasound intensity was 0.6 W/cm². However, the OCSCs stiffness got a rapidly promotion and exceeded that of A2780 cells after 1.0 and 1.4 W/cm² ultrasound exposure (Fig. 6). We considered that the change of cell stiffness may require appropriate ultrasound energy to affect them. For example, the “low” frequency ultrasound is more likely to formulate cavitational effects in cells due to the linear resonant, and the “medium” frequency ultrasound can lead to stronger thermal effects [51,52]. These ultrasonic effects can lead to different bio-effects in cells. Moreover, the changes of cytoskeleton and cell stiffness was not completely consistent. Some researches pointed that the category of actin filament and their particular organizations were identified as key factors of cell stiffness [53]. But the F-actin was often the focus when evaluating cell stiffness, and the spatial structure of cells was easy to be ignored. In brief, these results illustrated that LIUS could reduce the migration of CSCs by increasing the F-actin formation and cell stiffness in some extent. Furthermore, the different types of cells often showing different biosensitivity in same physical stimulus [54]. Mittelstein et al. demonstrated that LIUS could result in the cytoskeletal disruption, expression of apoptosis-related protein and the decrease of Young's modulus in some cancer cells, while the healthy cells still had a high Young's modulus [55]. Similar results were proved for cell stiffness and apoptosis when the mechanical stress of LIUS effect on the A2780 cells and OCSCs. Thus, it could be seen that the different changes of cell stiffness could be caused by nonmechanical stress and mechanical stress induced apoptosis.

Taken together, these studies revealed the regulatory role of biomechanics response in CSCs to LIUS and demonstrated that LIUS could promote the differentiation of CSCs to reduce drug resistance and migration due to the alteration of cell morphology, some intracellular regulators be activated to promote cytoskeleton formation and increase membrane stiffness. These founds had important implications in the utilization of LIUS in cancer research and therapy.

This work was supported by the National Key R&D Program of China (No. 2019YFE0110400), the National Natural Science Foundation of China (Nos. 81971621 and 82102087), the Key R&D Program of Hunan Province (No. 2021SK2035), and the Natural Science Foundation of Guangdong Province (Nos. 2021A1515011177, 2020A1515110628, and 2019A1515012212).

• National Key R&D Program of China (No. 2019YFE0110400; Funder ID: 10.13039/501100012166).

• National Natural Science Foundation of China (Nos. 81971621 and 82102087; Funder ID: 10.13039/501100001809).

• Key R&D Program of Hunan Province (No. 2021SK2035; Funder ID: 10.13039/501100002767).

• Natural Science Foundation of Guangdong Province (Nos. 2021A1515011177, 2020A1515110628, and 2019A1515012212; Funder ID: 10.13039/501100003453).

The authors attest that all data for this study are included in the paper. Data provided by a third party listed in Acknowledgements.