Enhanced tumor cell killing by ultrasound after microtubule depolymerization

Abstract Recent studies show that tumor cells are vulnerable to mechanical stresses and undergo calcium‐dependent apoptosis (mechanoptosis) with mechanical perturbation by low‐frequency ultrasound alone. To determine if tumor cells are particularly sensitive to mechanical stress in certain phases of the cell cycle, inhibitors of the cell‐cycle phases are tested for effects on mechanoptosis. Most inhibitors show no significant effect, but inhibitors of mitosis that cause microtubule depolymerization increase the mechanoptosis. Surprisingly, ultrasound treatment also disrupts microtubules independent of inhibitors in tumor cells but not in normal cells. Ultrasound causes calcium entry through mechanosensitive Piezo1 channels that disrupts microtubules via calpain protease activation. Myosin IIA contractility is required for ultrasound‐mediated mechanoptosis and microtubule disruption enhances myosin IIA contractility through activation of GEF‐H1 and RhoA pathway. Further, ultrasound promotes contractility‐dependent Piezo1 expression and localization to the peripheral adhesions where activated Piezo1 allows calcium entry to continue feedback loop. Thus, the synergistic action of ultrasound and nanomolar concentrations of microtubule depolymerizing agents can enhance tumor therapies.


| INTRODUCTION
Recent findings highlighted the substantial differences between the mechanical properties of tumor and normal cells indicating altered mechanosensing as an important feature of tumor cells. [1][2][3] Detailed studies showed that many tumor cells from different tissue origin lack rigidity sensor that is required for proper substrate rigidity sensing. 3 The rigidity sensors contain several cytoskeletal mechanosensory pro- majority of tumor cells lack the rigidity sensors. This has a practical consequence that the treatments inhibiting transformed cell growth inhibit the tumor cell growth. In studies of tumor cell growth, it has been observed that the loss of mechanosensing causes tumor cells to be damaged by mechanical stresses resulting in growth inhibition and apoptosis. 4,5 For instance, periodic exercise or stretching of tumors in an animal model leads to tumor regression. 6,7 Moreover, NIH National Cancer Institute lists seven different types of cancer that are inhibited by exercise and another eight tumor types where data are suggestive. 8 Further, our latest studies of stretch-or ultrasound (US)-mediated tumor cell apoptosis (mechanoptosis) establish the role of calcium-activated calpains in inducing a mitochondrial apoptotic pathway. 9,10 Other US treatments of tumors rely on the activation of multiple pathways to cause cell apoptosis and necrosis. For instance, highintensity focused US (HIFU), 11 high-intensity pulsed US, 12 and low-intensity pulsed US 13 have been used to ablate tumors. Moreover, US has been used in combination with other treatments like hyperthermia, chemotherapy, and sonodynamic therapy to enhance the efficacy of tumor treatment. [14][15][16] In terms of mechanism of tumor cell killing, US activates oxidative stress, mitochondrial damage, and DNA damage, which stimulates an apoptotic pathway. 17,18 However, there are health concerns due to hyperthermia causing damage to healthy cells surrounding the target and thus, these US methods have found limited clinical use. Similarly, severe side effects are associated with conventional chemo-and radiotherapy, which severely compromises the patient's quality of life. Therefore, in recent years, a premium has been placed on the strategies that minimize the side effects and interestingly, there are no discernable side effects of the low-frequency US.
We observed that low-frequency US causes selective tumor cell apoptosis (mechanoptosis) in vitro and in vivo without damaging normal cells. 9 To further understand the mechanoptosis process, we tested whether cancer cells are more vulnerable to US damage in certain phases of the cell cycle. 19,20 Tumor cells were pre-treated with cytostatic concentrations of cell-cycle phase inhibitors and then exposed to low-frequency US. Surprisingly, only inhibitors of mitosis increased the US-mediated apoptosis. When the concentration dependence was measured, we found that tumor cells treated with nanomolar concentrations of microtubule depolymerizing agents

| MDAs enhance US-mediated tumor cell apoptosis
Recently, we observed that many tumor cells from different tissue origin were vulnerable to mechanical stresses and exhibited mechanoptosis upon mechanical activation. 5,10 We hypothesized that such mechanoptosis was dependent on the cell-cycle phase and cells were more vulnerable to mechanical stresses when present in a particular cellcycle phase. To test the hypothesis, MDA-MB-231, a highly metastatic breast carcinoma cell line, was selected. Tumor cells plated on matrigel were treated with concentrations of pharmacological inhibitors that arrest and synchronize the majority cells in specific cell-cycle phases ( Figure S1 and Table S1). Cells grown on matrigel were subjected to low-frequency US (33 kHz) for 2 h with 7.7 mW/cm 2 power intensity and a 50% duty cycle. US-treated tumor cells showed significantly higher apoptosis (13% vs. 6%) compared to the non-treated cells (Figure 1 To determine if MDAs and US treatment affected normal cells, primary human foreskin fibroblasts (HFF) were exposed to US after being treated with MDAs. Interestingly, there was a negligible apoptosis in both control and treated samples ( Figure S4), indicating that normal cells were resistant to the combination of MDA and US treatment. Altogether, results showed that US-mediated tumor cell apoptosis was enhanced using cytostatic concentration of MDAs without damaging normal cells.      The US-based apoptosis of tumor cells is consistent with several studies, which report that tumor cells are sensitive to external mechanical stresses. [5][6][7]12 In a process that may be related, there is growing evidence that physical exercise inhibits tumor growth in mice and humans. [6][7][8] There is currently no clear understanding of how mechanical stresses are causing an increase in cytoplasmic calcium that activates tumor cell apoptosis. The power intensity of US used here is significantly low (7.7 mW/cm 2 ) and does not cause mechanical damage to normal cells/tissues. Further, similar expression levels of Piezo1 are observed in many matched pairs of tumor and normal cells, 10 but it serves different roles in the tumor and normal cells.

| US disrupts microtubules via calpain activation in tumor cells
Transformed tumor cells differ from normal cells in many aspects, since the levels of over 700 mRNAs are altered when normal cells become transformed. 3 The "transformed cell state" may be related to the "activated wound-healing state" and hence tumor growth is likely the result of a failure of wound-healing cells to revert to the normal growth. 29 Tumor cells appear softer; however, they exert higher forces on the matrices and exhibit the transformed growth on soft  Based upon these findings, we proposed that US-mediated mechanical forces and microtubule disruptors have a synergistic role in augmenting the tumor cells apoptosis (Figure 6). US-mediated mechanical forces activate Piezo1 to allow calcium uptake inside the cell. As described previously, calcium-activated calpains trigger apoptosis through a mitochondria-dependent pathway. 10 Concomitantly, calpains disrupt microtubule assembly to release GEF-H1, which in turn enhances myosin contractility through RhoA activation. Myosin contractility promotes Piezo1 expression and its localization to the peripheral adhesions. Elevated levels of Piezo1 at cell periphery further cause increased calcium influx to continue the feedback loop. In addition, MDAs disrupt microtubules to further activate RhoA pathway to enhance the apoptosis.
Since clinically approved MDA, vincristine is widely used to treat multiple sarcoma (rhabdomyosarcoma and Kaposi's sarcoma) 35

| Surface coating with matrigel
Ninty-six well plate surface was coated with thin layer of matrigel (Corning) according to manufacturer's protocol. Briefly, matrigel (Corning) diluted with cold PBS (1:2) (33 μl) was quickly added on prechilled surface of a single well of 96 well plate. Well plates were then incubated at 37 C for 24 h to solidify matrigel before performing cell seeding experiments.
Calpain siRNAs were fabricated by cloning facility MBI, Singapore, as described before. 10

| Apoptosis assay
To identify cell apoptosis, Annexin V-Alexa Fluor 488 or Annexin V-Alexa Fluor 594 conjugates (Thermofisher) were used according to the manufacturer's protocol. Assay was performed 12 h after US treatment.

| US treatment
The details of customize-built US device fabrication were mentioned previously. 9 Prior to ultrasonication, cells were treated with the cellcycle inhibitors for 16 h with cytostatic concentrations as stated in Table S1. Drug containing culture medium was then replaced with fresh culture medium. Well plates/dishes were quickly sealed with parafilm to avoid possible contamination and water entry during US treatment Samples were then placed 8 cm above transducer in the US tank, which was mounted in the incubator with 37 C. Tank was filled with DI water to half submerge the samples. Samples were then treated with 33 kHz frequency for 2 h with 50% duty cycle. Samples were returned to the standard incubator for overnight incubation to determine the apoptosis level on following day.
Images were acquired using either wide-field Olympus live-EZ microscope equipped with photometrics CoolSNAP K4 camera or W1 live-SR spinning disk microscope equipped with photometrics Prime 95B sCMOS camera.

| Pharmacological inhibitor study
Pharmacological inhibitors of cell cycle were used as follow: Binimetinib (1 μM

| Microtubules analysis
Assessment of extent of microtubule network disruption upon US treatment was manually quantified per 10 μm 2 area, taking into consideration an entire uninterrupted strand extending toward cell periphery as one. Strands forming any cross-linked intersections were scored as two. Strands observed as diffused structures with fluorescent dots undergoing severe microtubule disruption were scored as zero.

| Statistical analyses
GraphPad Prism version 6 was used to plot and analyze the data.
Statistical analysis was carried out using one-way ANOVA unless otherwise stated, considering p-value <0.05 as statistically significant.
The quantitative data represented as bar graphs or box-and-whisker plots were shown as mean ± standard deviation. The whiskers in boxand-whisker plots extend from the minimum to maximum values, the box extends from the 25th to the 75th percentile, and the line within the box represents the median.