Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: J Cell Physiol. 2021 Feb 7;236(9):6391–6406. doi: 10.1002/jcp.30314

Mechanical loading attenuates breast cancer-associated bone metastasis in obese mice by regulating the bone marrow microenvironment

Menglu Huang a,b, Hong Liu c,d,e,f, Lei Zhu d,g, Xinle Li a,b, Jie Li a,b, Shuang Yang a,b, Daquan Liu a,b, Xiaomeng Song a, Hiroki Yokota h, Ping Zhang a,b,h,i,*
PMCID: PMC8222149  NIHMSID: NIHMS1672473  PMID: 33554336

Abstract

Breast cancer, a common malignancy for women, preferentially metastasizes to bone and obesity elevates the chance of its progression. While mechanical loading can suppress obesity and tumor-driven osteolysis, its effect on bone-metastasized obese mice has not been investigated. Here, we hypothesized that mechanical loading can lessen obesity-associated bone degradation in tumor-invaded bone by regulating the fate of bone marrow-derived cells. In this study, the effects of mechanical loading in obese mice were evaluated through X-ray imaging, histology, cytology, and molecular analyses. Tumor inoculation to the tibia elevated body fat composition, osteolytic lesions, and tibia destruction, and these pathologic changes were stimulated by the high-fat diet. However, mechanical loading markedly reduced these changes. It suppressed osteoclastogenesis by downregulating RANKL and cathepsin K and promoted osteogenesis, which was associated with the upregulation of OPG and downregulation of C/EBPα and PPARγ for adipogenic differentiation. Furthermore, it decreased the levels of tumorigenic genes such as Rac1, MMP9 and IL1β. In summary, this study demonstrates that although a high-fat diet aggravates bone metastases associated with breast cancer, mechanical loading significantly protected tumor-invaded bone by regulating the fate of bone marrow-derived cells. The current study suggests that mechanical loading can provide a non-invasive, palliative option for alleviating breast cancer-associated bone metastasis, in particular for obese patients.

Keywords: mechanical loading, breast cancer bone metastases, high-fat diet, PPARγ, RANKL

Introduction

Cancer is the second common cause of death worldwide. Among all cancers that affect women, breast cancer is the leading cause of death and its incidence of bone metastasis is estimated as 65–75% (Cheng et al., 2018; Pan et al., 2017). Obesity has been reported to be associated with higher morbidity (Picon-Ruiz et al., 2017), and it is plausible that adipose tissues may stimulate the progression of primary and metastasized tumors. However, the mechanism by which obesity aggravates breast cancer-associated bone homeostasis has not been fully elucidated.

The cross-talk between tumor cells and their microenvironment is crucial for tumor initiation and progression (Anderson et al., 2019). Bone metastases obstruct the normal bone remodeling process and aberrantly enhance osteoclast-mediated bone resorption (Hesse et al., 2019). It is reported that metabolically active bone marrow with abundant adipocytes can be the preferred site for metastasis (Morris and Edwards, 2016; Shin and Koo, 2020), and an increase in bone marrow adiposity may potentially lead to a decrease in the number of osteoblasts (Timaner et al., 2020). To develop a therapeutic option for breast cancer bone metastasis, this study aimed to understand the regulatory mechanism of the fate of bone marrow cells, including osteoclasts, osteoblasts, and adipocytes, in the bone microenvironment.

To prevent tumor-induced bone degradation, it is important to block RANKL-induced osteoclastogenesis by decreasing the activities of cathepsin K and matrix metalloproteinases that are mostly secreted by mature osteoclasts (Zhu et al., 2020). Peroxisome proliferator-activated receptor gamma (PPARγ) is a member of the nuclear receptor superfamily and plays an important role in lipid and bone metabolism (Xiong et al., 2019). It is also important to stimulate bone formation by osteoblasts. Adipogenic transcription factors including PPARγ and CCAAT/enhancer-binding protein alpha (C/EBPα) have been shown to suppress osteoblastogenesis (Fan et al., 2009; Kawai, 2013; Li et al., 2013).

Rac1, a member of the Rac subfamily of small GTPases, plays a role in the regulation of proliferation, differentiation, apoptosis, cell movement, and adhesion. Rac1 is reported to be overexpressed in many cancers including prostate, breast, stomach, lung cancers (Durand-Onayli et al., 2018; Seyfi et al., 2018). It is known to promote breast cancer metastasis (Tian et al., 2018). MMP9 has been linked to the proliferation and invasion of breast cancer cells (Yang et al., 2019). Interleukin-1 (IL-1) drives the growth and colonization of breast cancer cells in the bone marrow adipose tissue niche (Fan et al., 2018). In mice fed with a high-fat diet with breast cancer-associated bone metastases, the mechanism that links osteolysis (CTSK and RANKL), osteogenesis (OPG), bone marrow adipogenesis (PPARγ and C/EBPα) to bone metastases (Rac1, MMP9, and IL-1β) remains elusive.

Since bone is a mechano-sensitive organ, oscillatory compressive loads can be applied in a form of elbow, knee, and ankle loading. We demonstrate in the mouse model that mechanical loading is effective in alleviating breast cancer bone metastasis (Yang et al., 2019). It can prevent tumor growth and bone degradation by suppressing osteoclastogenesis and promoting osteogenesis. Our previous studies show that mechanical loading can improve obesity-associated metabolic alterations (Tan et al., 2018). Mechanical loading also reduces the formation, migration, and adhesion of osteoclasts in osteonecrosis, osteoarthritis, and osteoporosis (Li et al., 2019; Li et al., 2019; Li et al., 2016; Liu et al., 2015; Zheng et al., 2019). Furthermore, it significantly increases osteoblast differentiation and decreases adipocytes differentiation of the mesenchymal stem cells. Taken together, many lines of evidence allow us to hypothesize that mechanical loading can improve symptoms associated with bone metastasis from breast cancer in obese mice and inhibit tumor viability and invasion.

To test the hypothesis, we employed a mouse model of breast cancer–associated bone metastases in which tumor cells were injected into the tibia. In the evaluation of mechanical loading, we evaluated bone microstructure, tumor-bearing, bone resorption, and bone formation using X-ray imaging, and histological analysis. We also evaluated the molecular mechanism of mechanical loading focusing on the functions of osteoclasts, osteoblasts, and adipocytes, as well as migratory and invasive behaviors of breast cancer cells.

Materials and methods

Animals and material preparation

Four to six weeks of age female BALB/c mice were obtained from the Animal Center of Academy of Military Medical Sciences (Beijing, China). Upon arrival, animals were housed in cages and received measured amounts of food and water ad libitum. Housing facilities were maintained on a 12:12 h light-dark cycle, at 25°C with 40 to 50% relative humidity. All experiments were agreed by the Ethics Committee of Tianjin Medical University (Tianjin, China) and implemented the Guide for Care and Use of Laboratory Animals (National Institutes of Health, Bethesda, MD, USA).

The mouse mammary tumor cell line 4T1 and human mammary tumor cell line MDA-MB-231 were obtained from American Type Culture Collection (Manassas, VA, USA). Roswell Park Memorial Institute 1640 basic medium, fetal bovine serum, penicillin/streptomycin, nonessential amino acids, and 0.25% trypsin were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Murine receptor activator of nuclear factor Kappa-B ligand (RANKL) and murine macrophagecolony stimulating factor (M-CSF) were purchased from PeproTech (Rocky Hills, NC, USA). An ELISA kit for TRACP-5b was purchased from Immunodiagnostics System (Scottsdale, AZ, USA). An immunohistochemical staining kit and a 3, 39-dia-minobenzidine substrate kit were purchased from ZSGB Bio-technology (Beijing, China). Antibodies were purchased for IL-1β, MMP9, OPG, and β-actin from Abcam (Cambridge, MA, USA), PPARγ and C/EBPα from Cell Signaling Technology (Danvers, MA, USA), and cathepsin K from Proteintech Group (Rosemont, IL, USA). Other chemicals were purchased from MilliporeSigma (Bur-lington, MA, USA).

Experiment design

After acclimation for two weeks, seventy-five mice were randomly assigned into five groups: the sham control group fed with a low-fat diet (LFD: chow containing 10% of energy derived from fat, Beijing Huafu Kang Biological Co, Beijing, China, n=15), the model group fed with a low-fat diet (M; n=15), the model group fed with a low-fat diet and received loading (ML; n=15), the model group fed with a high-fat diet (HFD: chow with 60% energy derived from fat, Beijing Huafu Kang Biological Co, Beijing, China; HM, n=15), and the model group fed with a high-fat diet and received loading (HML, n=15). We injected 4T1 tumor cells into the tibia and induced tumor invasion as described previously (Figure 1a) (Yang et al., 2019).

Figure 1.

Figure 1.

Open in a new tab

The effect of loading on body weight and body composition with 4T1 cell intratibial injection. (a) Timeline of the study. (b) Intratibial injection in the proximal tibia (Bar = 1 cm). (c) Bilateral loading applied to the intratibial ankle (1 N at 5 Hz for 6 min/d) for 3 weeks (Bar = 2.5 mm). (d) Food of intake on week 2, 4, 6, 8, 10, 12, 13, 14 and 15. (e) Before sacrifice body weight and changes in body weight. (f) Before sacrifice fat mass and changes in fat mass. (g) Before sacrifice BMI and changes in BMI. #p < .05, and ###p < .001 vs sham group; *p < .05, **p < .01, and ***p < .001 (n = 10).

Body weight and body fat composition

The body weight of each mouse was measured weekly. Body compositions (BMI, fat, and lean mass) were determined using a composition analyzer (ImpediVet, Pinkenba, Qld, Australia). The food intake of each mouse was measured every day (Tan et al., 2018).

Animal model of breast cancer cell intra-tibial injection

4T1 cells were cultured in a complete 1640 medium with the addition of 10% fetal bovine serum, 1% penicillin/streptomycin, and 1% nonessential amino acid, and they were incubated at 37°C with 5% CO2. 4T1 cells were injected into the bilateral tibia as described previously (Yang et al., 2019). Mice were anesthetized by 1.5% isoflurane at a flow rate of 0.6-1.0 L/min. The hair was removed from the proximal tibia and the skin was disinfected with 70% alcohol. The cell suspension (1×104 cells suspended in 25 ml sterile PBS) was injected into the medullary cavity of the proximal tibia through the middle of the patellar ligament using a 29-gauge insulin syringe. The injection was conducted to the left and right tibia. The sham control group received PBS without tumor cells (Figure 1b).

Mechanical loading

One day after the injection of tumor cells, mice in the loading groups (ML and HML) were anesthetized and received daily mechanical loading using the procedure previously described (Yang et al., 2019). After finishing the loading to the left ankle, the same loading procedure was conducted to the right ankle. Loads were 1 N force at 5 Hz and given for 3 min for each ankle for six consecutive days per week. The sham control and non-loading groups (M and HM) were also placed in the apparatus without receiving any loads (Figure 1c). Three weeks after the injection, the measurement of BMD, BMI, fat and lean mass were conducted and animals were sacrificed. Bone marrow-derived cells were harvested from the ilia and femora, and the tibia was used for histology.

X-ray imaging

Osteolytic lesion area and abnormal bone remodeling could be visualized and assessed in vivo using an X-ray machine. The mice were sedated, anesthetized, and subjected to digital X-ray imaging of the middle region of the hind limbs using a Kodak In Vivo Imaging Systems (Carestream Health, Rochester, NY, USA). All images were acquired using the following parameters: 180 s exposure time; 2×2 binning; 60 s acquisition time; 120×120 mm field of view; and f 2.25 aperture stop (Yang et al., 2019).

Microcomputed tomography

Bone morphometric parameters were determined in the proximal tibia using a Scanco vivaCT 40 (Scanco Medical, Bassersdorf, Switzerland). Scans were taken with a source voltage of 70 kV and a source current of 110 μA. The resolution was set to 1k, and the rotation step was 0.5°. The regions of interest in the proximal tibia were distal to the growth plate with 250 cross-sectional slices. Three-dimensional images were reconstructed from CT slices, and morphometric parameters were determined. Measurements included tibial integrity (%, the number of intact tibias to the total number of tibias; the number of degraded tibias to the total number of tibias), fractional bone volume (BV/TV), trabecular thickness (Tb.Th; mm), trabecular number (Tb.N; 1/mm), and trabecular separation (Tb.Sp; mm) (Park et al., 2014; Yang et al., 2019).

Measurement of bone mineral density

Using peripheral dual-energy X-ray absorptiometry (pDEXA; PIXImus II, Lunar Corp, Madison, WI, USA), bone mineral density (BMD, g/cm2) of the bilateral femora and tibia was determined before the injection of tumor cells and sacrifice (version 1.47). We scanned the entire animal and conducted the region of interest analysis. The changes in BMD were determined and statistical analysis was conducted (Liu et al., 2015).

Histological assay and immunohistochemistry assay

Tibia samples were fixed in 10% neutral formaldehyde solution for 48 h and decalcified with 10% ethylenediaminetetraacetic acid (EDTA, pH 7.4) for 3 weeks. Samples were embedded in paraffin and cut into 5-μm-thick sections as we described (Yang et al., 2019). The ratio of tumor area/total tissue area (TuAr/Tar, in %) was analyzed using hematoxylin and eosin (H&E) staining. Tartrate resistant acid phosphatase (TRAP) staining was used to detect the percentage of osteoclasts (Li et al., 2016). MacNeal’s staining was used to determine the number of osteoblasts (Liu et al., 2017). All images were measured and analyzed using Cellsense Standard software (Olympus).

The expression levels of MMP9, IL-1β, and OPG were analyzed by immunohistochemistry. After the deparaffination and rehydration, antibodies were reacted at a 1:50 dilution overnight at 4°C. An immunohistochemical kit and 3, 3′-diaminobenzidine (DAB) substrate kit were used according to the manufacturer protocol. Histomorphometric measurements were conducted on the area of the proximal tibia. Quantitative analysis was conducted in a blinded fashion.

Quantification of serum biomarkers

The TRACP-5b, a bone resorption marker of bone turnover markers, was used as a marker for osteoclasts, as previously described (Xu et al., 2015; Yang et al., 2019). Serum was collected after sacrificed and selected biomarkers were evaluated by enzyme-linked immunosorbent assay (ELISA) according to manufacturers’ protocols.

Isolation of bone marrow-derived cells and osteoclast formation

Using harvested bone samples, we flushed the bone marrow of the iliac crests and femora with DMEM (containing 2% FBS). The cells were separated with Ficoll low-density gradient centrifugation. Bone marrow-derived cells were seeded onto 96-well plates at a density of 1×105 cells/well, and cultured in supplemented MEM-α with 10% FBS, 30 ng/ml M-CSF and 20 ng/ml RANKL for 3 days. On day 4, the medium was changed to include 10% FBS, 30 ng/ml M-CSF, and 60 ng/ml RANKL. Cells were cultured for 3 days, and osteoclasts were identified using a TRAP staining kit (Sigma-Aldrich). TRAP-positive multinuclear cells (more than 3 nuclei) were identified as matured osteoclasts, and images of every five fields were randomly photographed for analysis (Liu et al., 2017).

Osteoclast migration and adhesion

Pre-osteoclasts were seeded onto 6-well plates at a density of 2×106 cells/well, and cultured in supplemented with 10% FBS, 30 ng/ml M-CSF, and 20 ng/ml RANKL for 4 days. The migration of pre-osteoclasts was evaluated using a transwell assay (Yokota et al., 2013). The pre-osteoclasts (1×105 cells/well) were resuspended onto the upper chamber of transwell and allowed to migrate to the bottom chamber through an 8-μm polycarbonate filter coated with vitronectin (Takara Bio Inc., Otsu, Shiga, Japan). The medium in the upper chamber was replaced by MEM-α without serum, and MEM-α (consisting of 1% bovine serum albumin and 30 ng/ml M-CSF) were loaded onto the lower chamber. After 6 h incubation, the number of migrated pre-osteoclasts was counted. For the adhesion assay, osteoclast precursor cells (1×105 cells/well) were plated onto 96-well plates coated with 5 μg/ml vitronectin in α-MEM supplemented with 30 ng/ml M-CSF. After 30 min of incubation, cells were washed with PBS three times and fixed with 4% paraformaldehyde at room temperature for 10-15 min. Cells were stained with crystal violet, and the number of cells adherent to the bottom of plates was counted (Liu et al., 2017).

Osteoblast differentiation and mineralization assay

In an osteoblast differentiation assay, bone marrow-derived cells were cultured onto 6-well plates (2×106 cells/well) with the osteogenic differentiation medium, and the medium was changed every 2 days. On day 14, cells were stained using an alkaline phosphatase (ALP) staining kit (Sigma) and the percentages of ALP-positive cells were determined.

Using Alizarin Red S (ARS) staining for testing osteoblast mineralization, bone marrow-derived cells were cultured for 3 weeks. Cells were fixed in 4% paraformaldehyde for 15 min and stained with ARS for 30 min (2% of ARS dissolved in distilled water with the pH adjusted to 4.2) (Li et al., 2019).

Adipogenic induction and Oil-Red O stain assay

Bone marrow-derived cells were incubated onto 6-well plates (2×106 cells/well) using the induction medium consisting of 0.5 μM dexamethasone, 50 μM indomethacin, 5 μg/ml recombinant human insulin, 0.25 mM 3-isobutyl-1-methyl-xanthine and 10% FBS for 2 days. The medium was changed to include 5 μg/ml recombinant human insulin and 20% FBS, until lipid droplets appeared under the microscope (Olympus). Adipogenic differentiation was assayed by quantifying neutral lipid vacuoles stainable with Oil-Red O (Han et al., 2016). Cells were rinsed with PBS and fixed with 4% paraformaldehyde for 30 min. After washing with distilled water, cells were coated with 60% isopropanol. Then, cells were incubated with the Oil Red O working solution (diluted stock solution with water (3:2)) for 15 min, and the lipid droplets were detected microscopically (Sadeghi et al., 2020). To quantify the number of lipid droplets, the stain was dissolved with 100% isopropanol and the absorbance was measured at 490 nm.

Preparation of bone marrow mesenchymal stem cells-derived conditioned medium (BMSCs-CM)

Bone marrow-derived cells of passage 2-3 were used to prepare the BMSCs-CM. Upon 80% of confluence, the culture medium was removed and cells were washed with phosphate-buffered saline (PBS) before feeding with 4 ml of DMEM supplemented with 10% FBS and 1% penicillin-streptomycin. Cells were then incubated at 37°C in a humidified atmosphere with 5% CO2 for 24 h. The culture medium was collected, filtered with 0.20 μm sterile filters, and stored at −80°C (Su et al., 2015).

Cell viability assay

The MTT assay was used to evaluate cell viability. 4T1 or MDA-MB-231 cells were seeded in 96-well plates at a density of 1×104 cells/well with DMEM. Four hours later, cells were cultured in BMSCs-CMs isolated from four groups (M, ML, HM, and HML). After 48 h, the absorbance at 570 nm was detected using a μQuant universal microplate spectrophotometer (Bio-tek, Winooski, USA) (Pan et al., 2020).

Wound-healing assay of 4T1 and MDA-MB-231

In a wound-healing assay, 4T1 and MDA-MB-231 cells were cultured with control media in 6-well plates for 48 h until grown to 100% confluence. Using a 200-μl pipette tip, a scratch was made on a cell monolayer. The width of the scratch was measured at 0 h and 20 h, and the percentages of scratch recovery were calculated (Chen et al., 2019).

Migration of 4T1 and MDA-MB-231

MDA-MB-231 cells were cultured in complete DMEM medium with the addition of 10% FBS and 1% penicillin/streptomycin, and incubated at 37°C with 5% CO2. 4T1 and MDA-MB-231 cells were cultured in BMSCs-CM upon 100%. Cells (1×105 cells/well) were resuspended onto the upper chamber of transwell and allowed to migrate to the bottom chamber through an 8-μm polycarbonate filter coated with vitronectin (Takara Bio Inc., Otsu, Shiga, Japan). The medium in the upper chamber was replaced by MEM-α without serum, and MEM-α (consisting of 1% bovine serum albumin and 30 ng/ml M-CSF) were loaded onto the lower chamber. After 6 h incubation, the number of migrated cells was counted (Bhattacharya et al., 2020).

Western blot assay

We isolated proteins from tibia, bone-marrow-derived cells, 4T1 cells, and MDA-MB-231 cells. Bone tissue powders from the tibia or cells were lysed in a RIPA lysis buffer, which contained the inhibitors of proteases and phosphatases (Roche Diagnostics GmbH, Mannheim, Germany). The levels of cathepsin K, RANKL, Rac1, PPARγ, C/EBPα, and β-actin were determined by Western blotting as described previously (Yang et al., 2019). The samples were resolved in the sodium dodecyl sulfate-polyacrylamide gel and transferred onto the polyvinylidene difluoride membranes. The membranes with primary antibodies were incubated over-night at 4°C (cathepsin K 1:3000, RANKL 1:4000, Rac1 1:1000, PPARγ 1:1000, C/EBPα 1:1000 and β-actin 1:10000). The membranes were washed and incubated with horseradish peroxidase-conjugated secondary antibody (1:20000). Enhanced chemiluminescence was used to quantify protein levels. Image acquisition and analysis software (Bio-Rad, Hercules, CA, USA) was used to quantify band intensities.

Statistical analysis

Data were expressed as means ± SEM and analyzed with independent-sample Student’s t-test (for 2 groups), or one-way ANOVA (for more than 2 groups). For pair-wise comparisons, a post hoc test was conducted using Fisher’s protected least significant difference. Correlation analysis of parameters was performed by using the Pearson correlation coefficient test. Statistical significance was assumed at p < .05.

Results

Mechanical loading changed body weight and body fat composition

The mean body weight and the change in body weight were determined. Of note, food intake did not change significantly among the groups (Figure 1d). Compared to the sham group (no tumor), the M group (tumor inoculated) showed a prominent decrease and the HM group showed a significant increase in body weight (both p < .001). Furthermore, mechanical loading suppressed weight loss in the ML group (p < .05). By contrast, loading markedly reduced the weight gain by the high-fat diet (p < .001). Compared to the M group, the ML group decreased the change of the body weight (p < .05). Compared to the HM group, the HML group increased the change of the body weight (p < .01; Figure 1e). Collectively, mechanical loading suppressed tumor-driven weight loss as well as high-fat diet-linked weight gain.

We next evaluated the breakdown of weight alterations. Compared to the sham control, the M group presented a reduction in body fat and BMI (both p < .05). Loading in the ML group increased BMI (p < .05), and loading in the HML group reduced body fat and BMI (both p < .05). Also, loading in the ML group reduced the change of fat mass (p < .05), and loading in the HML group increased the change of fat mass and BMI (both p < .05; Figure 1f, g).

Mechanical loading improved osteolysis and cancellous bone microarchitecture

To analyze the effect of mechanical loading on bone remodeling, cancellous mass and bone architecture were evaluated using X-ray and micro CT imaging (Figure 2a&c). Compared to the sham control group, tumor inoculation caused osteolytic lesions and reduced trabecular bone. It also generated the cavitation of cortical bone and destroyed tibia integrity. The responses in the HM group were more severe (Figure 2b). On the contrary, mechanical loading markedly inhibited osteolysis and enhanced bone microarchitecture. We also examined any changes in BMD. Compared to the sham control, the M and HM groups showed a decrease in BMD (both p < .05). However, the 3-week application of mechanical loading suppressed this reduction (both p < .05; Figure 2d). Compared to the sham control, the M group was declined (all p < .05) and the HM group showed a significantly decrease in the BV/TV (p < .01), Tb.Th (p < .05), and Tb.N (p < .05). However, the loading groups (ML and HML) significantly enhanced BV/TV (both p < .05), Tb.Th (both p < .05), and Tb.N (both p < .05). In contrast to the BV/TV, Tb.Th, and Tb.N, Tb.Sp was increased in the M group and HM group, while it was decreased in the ML group and HML group (all p < .05; Figure 2e).

Figure 2.

Figure 2.

Open in a new tab

Mechanical loading inhibited obesity-induced fast tumor growth and metastasis to bone. (a) X-ray images. The arrows indicate the proximal region of the tibia (Bar = 2.5 mm). (b) Tibial integrity. (c) Micro CT images (Bar = 1.0 mm). (d) Changes in BMD. (e). Quantitative analysis of BV/TV (%), Tb.Th (mm), Tb.N (1/mm) and Tb.Sp (mm) in the proximal tibia (n=3). (f) Representative images of H&E stained tibial sections (Bar = 500 μm). The TuAr was indicated by the solid line, (g&h) Histomorphometric evaluation of the TuAr/Tar ratio. The expression of MMP9 and IL-1β detected by immunohistochemistry (Bar = 50 μm). The arrows indicate the positive signal. Quantification of IL-1β and MMP9 positive cells in the tibia, (i) Western blot images for Rac1 and quantification. ##p < .01 vs sham group; *p < .05, **p < .01, and ***p < .001 (n = 6).

Mechanical loading inhibited obesity-induced fast tumor growth and metastasis to bone

We examined the effect of mechanical loading on cancellous bone. Tumor burden (TuAr/Tar – tumor area ratio) was measured using H&E-stained sections (Figure 2f). Compared to the M group, the ML group showed a decrease in TuAr/Tar (p < .001), but the HM group presented an increase in TuAr/Tar (p < .05). Also, compared to the HM group, the HML group showed a reduction in TuAr/TAr (p < .01).

To further evaluate the effect of mechanical loading, we examined the expression of the selected metastasis-related factors, such as MMP9, IL-1β, and Rac1. Immunohistochemistry analysis showed that compared to the M group, positive cells of MMP9 (p < .001) and IL-1β (p < .01) in the ML group were significantly decreased by mechanical loading, while in the HM group was markedly increased (p < .05). Moreover, the HML group presented the reduction in MMP9 (p < .01) and IL-1β (p < .01) compared to the HM group (Figure 2g&h). Western bolt analysis illustrated that compared to the M group, the ML group decreased the level of Rac1 (p < .05). Moreover, compared to the HM group, the HML group also showed a reduction in Rac1 (p < .05; Figure 2i).

Mechanical loading suppressed abnormal development of osteoclasts

We next evaluated the effect of mechanical loading on cellular differentiation and molecular signaling. TRAP staining was conducted to determine the effect of mechanical loading on bone resorption. Compared to the M group, the ML group decreased the TRAP-positive cells (p < .05), and compared to the HM group, the HML group also showed the reduction in the TRAP-positive cells (p < .05; Figure 3a).

Figure 3.

Figure 3.

Open in a new tab

Mechanical loading suppressed abnormal development of osteoclasts. (a) TRAP staining for evaluation of bone resorption (Bar = 50 μm). Red staining showed TRAP-positive osteoclasts. Osteoclast surface (%) (b) TRAP staining for images of mature osteoclasts formation in vitro (Bar = 50 μm). Percentage of the matured osteoclasts area. (c) Crystal violet staining for stand of pre-osteoclasts adhesion (Bar =100 μm). Numbers of adherent pre-osteoclasts. (d) Crystal violet staining for stand of pre-osteoclasts migration (Bar = 100 μm). Number of migratory pre-osteoclasts. (e) Representative images of Western blot with mechanical loading. The levels of RANKL and cathepsin K were shown, (f) Serum TRACP-5b analyses. #p < .05, and ###p < .001 vs M group; *p < .05, **p < .01, and ***p < .001 (n = 6).

To evaluate whether mechanical loading affected osteoclast development in the M and HM groups in vitro, we performed the formation, migration, and adhesion assays using bone marrow-derived cells. In the HM group, the formation (p < .05), adhesion (p < .001) and migration (p < .001) of bone marrow-derived cells were stimulated compared to the M group. Those activities were significantly reduced in the loaded groups (all p < .001; Figure 3b–d).

We also examined the expression of RANKL and CTSK in the tibia. In the ML and HML groups, the levels of RANKL and CTSK were suppressed by mechanical loading compared to the M and HM groups (all p < .05; Figure 3e). Compared to the M and HM groups, there was a substantial decrease in the serum TRACP-5b level in the ML and HML groups (both p < .05; Figure 3f).

Mechanical loading significantly improved bone-forming osteoblasts

MacNeal’s staining in vivo was used to identify osteoblasts that were located on the trabecular surface of the tibia. ALP staining in vitro was used to evaluate the differentiation of osteoblasts from mesenchymal stem cells. Mechanical loading elevated the number of osteoblasts in the ML (p < .05) and HML (p < .05) groups compared to the M group and HM group (Figure 4a). The ALP-positive cells in the ML (p < .01) group were increased and the HM (p < .05) group showed significantly decreased compared to the M group. Furthermore, compared to the HM group, the ALP-positive cells were markedly elevated in the HML group (p < .001; Figure 4b). Alizarin Red S staining showed that tumor inoculation decreased the number of mineralized modules, whereas mechanical loading noteworthily elevated the mineralization of osteoblasts (Figure 4c).

Figure 4.

Figure 4.

Open in a new tab

Mechanical loading significantly improved bone-forming osteoblasts. (a) MacNeal’s staining for identifying osteoblasts (Bar = 50 μm). Osteoblasts, located on the trabecular surface. The number of osteoblasts was quantified. (b) ALP staining for stand of osteoblast differentiation in vitro (Bar =100 μm). The number of ALP+ cells. (c) Alizarin Red staining to test mineralization of osteoblasts. (d) Expression of OPG detected by immunohistochemistry (Bar = 50 μm). Quantification of OPG positive cells in the tibia. (e) Oil-Red O staining for test lipid droplets (Bar = 50 μm) and OD520. (f) Representative images of Western blot with mechanical loading. The levels of C/EBPα and PPARγ were shown. #p < .05 vs M group; *p < .05, **p < .01, and ***p < .001 (n = 6).

We also examined the level of the bone forming-related factor, OPG. Immunohistochemistry analysis showed that compared to the M group, the number of OPG positive cells in the ML group was increased by mechanical loading (p < .05), but the HM group showed its reduction (p < .05). Compared to the HM group, the HML group presented the improvement in OPG (p < .01; Figure 4d).

Mechanical loading inhibited adipogenic formation

In the adipogenic differentiation assay, compared to the M group, the ML presented no significant difference by mechanical loading, and the HM group showed an increase in lipid droplet formation (p < .05). The HML group inhibited lipid droplet formation compared to the HM group (p < .05; Figure 4e).

We also examined the expression of C/EBPα and PPARγ. Compared to the M and HM groups, the levels of C/EBPα and PPARγ in the ML (both p < .05) and HML (both p < .05) groups were significant decreased (Figure 4f).

The effect of the bone microenvironment on tumor cells

To investigate whether 4T1 cells were affected by the medium harvested from the culture of bone marrow-derived cells, we performed the cell viability, migration, and wound-healing assay. To examine whether mechanical loading has the same efficacy in human breast cancer cells, MDA-MB-231 cells were also used. In the 4T1 cell viability assay, compared to the M and HM groups, cell viability in the ML (p < .05) and HML (p < .05) groups were decreased by mechanical loading (Figure 5a). In the 4T1 cell migration assays, compared with the M group, the number of migrated cells was markedly increased in the HM group (p < .01), but the ML group showed a decrease by mechanical loading (p < .01). Compared to the HM group, loading-linked media caused a reduction in the migratory cells (p < .001; Figure 5c&e). In the 4T1 cell wound-healing assay, compared to the M group, the recovered cells were increased in the HM group (p < .05), but in the ML group, those cells were decreased by mechanical loading (p < .05). Compared to the HM group, loading-linked media caused a reduction in the number of recovered cells (p < .05; Figure 5g&i). In addition, MDA-MB-231 cells showed the same tendency in the cell viability (Figure 5b), migration (Figure 5d&f), and wound-healing assay (Figure 5h&j). We also examined the expression of the tumor metastatic related factor, Rac1. In 4T1 or MDA-MB-231 cells, compared to the M group, the level of Rac1 in the ML group was decreased (both p < .05), while the level of Rac1 in the HM group was increased (both p < .05). Compared to the HM group, however, the level of Rac1 in the HML group was remarkable reduction (both p < .05; Figure 5k).

Figure 5.

Figure 5.

Open in a new tab

The effect of bone microenvironment to tumor cells. (a) MTT assay for cell viability of 4T1. (b) MTT assay for cell viability of MDA-MB-231. (c&d) Crystal violet staining for stand of 4T1 and MDA-MB-23 lmigration (Bar = 100 μm), (e) Number of migratory 4T1. (f) Number of migratory MDA-MB-231. (g&h) Representative images of 4T1 and MDA-MB-231 wound-healing (Bar = 200 μm). (i) Percentages of 4T1 wound recovery. (j) Percentages of MDA-MB-231 wound recovery. (k) Representative images of Western blot with mechanical loading. The level of Rac1 was shown. #p < .05, and ##p < .01 vs M group *p < .05, **p < .01, and ***p < .001 (n = 3).

Correlations among osteoclasts, osteoblasts, and adipocytes in breast cancer bone metastasis

To quantitatively characterize the role of osteoclasts, osteoblasts, and adipocytes in the progression of breast cancer bone metastasis, the correlation analysis was conducted using the values such as the numbers of TPAP+ and ALP+ cells, lipid droplet OD values, tumor burden with BV/TV, and the numbers of osteoclasts and osteoblasts. First, the number of ALP+ cells in the tibia was negatively correlated with the number of multinucleated osteoclasts (r = −0.717; Figure 6a). Second, the lipid droplet OD values in the tibia were positively correlated with the number of multinucleated osteoclasts (r = 0.6; Figure 6b). Third, the lipid droplet OD values in the tibia were negatively correlated with the number of ALP+ cells (r = −0.664; Figure 6c). Fourth, the level of OPG was negatively correlated with that of RANKL (r = −0.793; Figure 6d). Fifth, the number of osteoclasts (N.Oc/BS) in the tibia was positively correlated with TuAr/Tar (r = 0.7; Figure 6e). Lastly, the number of osteoblasts (N.Ob/BS) was negatively correlated with TuAr/TAr (r = −0.658; Figure 6f).

Figure 6.

Figure 6.

Open in a new tab

Correlation analysis, a&b) Correlations between ALP+ cells (a), OD values (b) and TPAR+ multinucleated osteoclasts. (c) Correlations between OD values and ALP+ cells. (d) Correlations between OPG and RANKL. (e&f) Correlations between osteoclast number (Oc.S/BS) (e), osteoblast number (N.Ob/BS) (f) and TuAr/TAr. (g) Correlations among TPAR+ multinucleated osteoclasts, ALP+ cells, OD values and 4T cells viability. (h) Correlations between TPAR multinucleated osteoclasts, ALP+ cells, OD values and MDA-MB-2311 cells viability (n = 6).

Correlations in the bone microenvironment with tumor cells

Correlation analysis was also conducted to evaluate any linkage among the numbers of tumor cells, osteoclasts, osteoblasts, and adipocytes. First, the number of multinucleated osteoclasts in the tibia was positively correlated with the viability of 4T1 and MDA-MB-231 cells (r = 0.678, 0.746). Second, the number of ALP+ cells in the tibia was negatively correlated with the viability of 4T1 and MDA-MB-231 cells (r = −0.726,−0.788). Lastly, the lipid droplet OD values in the tibia were positively correlated with the viability of 4T1 and MDA-MB-231 cells (r = 0.629, 0.685; Figure 6g&h).

Discussion

The current study demonstrates that mechanical loading can alleviate osteolysis linked to bone metastases from breast cancer, by regulating the fate of bone marrow-derived cells in obese mice fed with a high-fat diet (Figure. 7). While the bone loss was notably elevated in tumor-inoculated obese mice with an increase in the number of osteoclasts and a decrease in the number of osteoblasts, daily mechanical loading to the ankle significantly protected the tibia from tumor-driven osteolysis. It promoted the proliferation and differentiation of osteoblasts while suppressing the development of osteoclasts and adipocytes from bone marrow-derived cells.

Figure 7.

Figure 7.

Open in a new tab

Proposed mechanism of mechanical loading’s action on breast cancer-associated bone metastasis in high-fat diet mice.

We observed that mechanical loading increased BMD and the structural integrity of tumor-invaded tibia, because of its role in altering the fate of bone marrow-derived cells. Our data indicated that loading suppressed the maturation, migration, and adhesion of osteoclasts, as well as the level of osteoclast markers such as CTSK and RANKL. Increasing evidence suggests that the RANKL/RANK signaling system is widely involved in the progression of breast cancer and breast cancer cells often express abundant RANK (Infante et al., 2019). The bone also provides a unique nutrient-rich microenvironment and attracts tumor cells (Maroni and Bendinelli, 2020). Breast cancer cells that migrate to bone express cytokines and growth factors and they can induce RANKL expression. The up-regulated RANKL activates osteoclastogenesis, followed by the turnover of the bone matrix with a release of growth factors and cytokines (Geerts et al., 2020; Wu et al., 2020). RANKL also induced the expression of tumorigenic genes such as Rac1, MMP9, and IL-β (Park et al., 2014; Rachner et al., 2009; Shirakawa et al., 2019).

We also observed that loading enhanced the formation and mineralization of osteoblasts and elevated the expression of OPG. Mechanical loading inhibited the differentiation of adipocytes and suppressed the lipogenesis-related factors C/EBPα and PPARγ. While the epidemiological link between obesity and breast cancer incidence and outcome is well established, the mechanism explaining how obesity promotes breast cancer metastases to the bone is unclear. In obese mice, the differentiation of adipocytes from bone marrow-derived MSCs was increased. During the differentiation of adipocytes, transcription factors, including C/EBPs and PPARγ, activate lipogenesis (Lim et al., 2015). PPARγ haploinsufficiency has been reported to induce osteoblastogenesis and increase bone mass (Akune et al., 2004). This finding suggests that impaired PPARγ signaling might shift the fate of MSCs in the bone marrow towards the osteoblast lineage. In addition to the potent adipogenic function of PPARγ in MSCs, an osteoclastogenic function has also emerged from the selective ablation of the PPARγ gene in mouse osteoclastic precursor cells (Takada et al., 2009).

The reduction in bone mass in mice fed with a high-fat diet presented an elevated level of PPARγ and suggested the possibility of the PPARγ-mediated promotion of adipogenesis and osteoclastogenesis. Mechanical loading reduced the levels of RANKL, C/EBPα and PPARγ. Loading-driven reduction in Rac1, MMP9, and IL1-β indicates that the loaded bone does not enhance the provision of nutrients for the progression of tumor cells.

Conditioned media were used to study cell-to-cell interactions via paracrine and endocrine signals. We postulate that mechanical loading effects in this study are mediated via fluid flow, molecular transport, or intramedullary pressure in the interstitial bone (Zhang et al., 2009; Zhang et al., 2008; Zhang et al., 2007), which ultimately alter interactions and fates of bone marrow-derived cells. We thus harvested conditioned media from bone marrow-derived cells and evaluated their effects on breast cancer cells. The result showed that the conditioned medium, related to the high-fat diet, promoted tumor growth, and increased tumor invasion. However, the conditioned medium, linked to mechanical loading, inhibited the growth and invasion of tumor cells. Rac1 is associated with breast cancer metastasis and osteoclast activity. Studies have shown that decreased Rac1 activity can lead to osteoclast dysfunction and inhibit bone resorption (Shin et al., 2020). Rac1 has been shown to have an important role in tumor cell migration (Liu et al., 2019). Furthermore, we observed that mechanical loading reduced the levels of tumorigenic genes such as Rac1.

In summary, we demonstrated using the mouse model that mechanical loading is effective in protecting bone from breast cancer metastasis. It suppressed osteoclastogenesis and adipogenesis, while promoting osteogenesis. Furthermore, it decreased the growth and invasion of tumor cells by regulating the bone microenvironment. The present study suggests that mechanical loading may have an advantage for protecting bone by remotely applying loads to the site of tumor-induced osteolysis.

ACKNOWLEDGMENTS

This work was supported by grants from the National Natural Science Foundation of China (81772405 and 81572100 to P. Zhang), Tianjin Science and Technology Program Fund (18PTZWHZ00100 to L. Zhu), Natural Science Foundation of Tianjin (18JCQNJC82200 to X. Li), and NIH (AR052144 to H. Yokota).

Footnotes

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

REFERENCES

  1. Akune T, Ohba S, Kamekura S, Yamaguchi M, Chung UI, Kubota N, Terauchi Y, Harada Y, Azuma Y, Nakamura K, Kadowaki T, Kawaguchi H. (2004). PPARγ insufficiency enhances osteogenesis through osteoblast formation from bone marrow progenitors. J Clin Invest 113(6):846–855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson RL, Balasas T, Callaghan J, Coombes RC, Evans J, Hall JA, Kinrade S, Jones D, Jones PS, Jones R, Marshall JF, Panico MB, Shaw JA, Steeg PS, Sullivan M, Tong W, Westwell AD, Ritchie JWA, Cancer Research UK, Cancer Therapeutics CRCAMWG. (2019). A framework for the development of effective anti-metastatic agents. Nat Rev Clin Oncol 16(3): 185–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bhattacharya S, Ghosh A, Maiti S, Ahir M, Debnath GH, Gupta P, Bhattacharjee M, Ghosh S, Chattopadhyay S, Mukherjee P, Adhikary A. (2020). Delivery of thymoquinone through hyaluronic acid-decorated mixed Pluronic® nanoparticles to attenuate angiogenesis and metastasis of triple-negative breast cancer. J Control Release 322:357–374. [DOI] [PubMed] [Google Scholar]
  4. Chen Y, Chen L, Zhang JY, Chen ZY, Liu TT, Zhang YY, Fu LY, Fan SQ, Zhang MQ, Gan SQ, Zhang NL, Shen XC. (2019). Oxymatrine reverses epithelial-mesenchymal transition in breast cancer cells by depressing αv β3 integrin/FAK/PI3K/Akt signaling activation Onco Targets Ther 12:6253–6265. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cheng Y, Yan Y, Gong J, Yang N, Nie S. (2018). Trends in incidence and mortality of female breast cancer during transition in Central China. Cancer Manag Res 10:6247–6255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Durand-Onayli V, Haslauer T, Harzschel A, Hartmann TN. (2018). Rac GTPases in hematological malignancies. Int J Mol Sci 19(12 ):4041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fan FS, Yang CF, Wang YF. (2018). Diffuse bone marrow metastasis as the initial presentation of an occult breast cancer. Case Rep Oncol Med 2018:2946409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Fan Q, Tang T, Zhang X, Dai K. (2009). The role of CCAAT/enhancer binding protein (C/EBP)-α in osteogenesis of C3H10T1/2 cells induced by BMP-2. J Cell Mol Med 13(8B):2489–2505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Geerts D, Chopra C, Connelly L. (2020). Osteoprotegerin: relationship to breast cancer risk and prognosis. Front Oncol 10:462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Han N, Li Z, Cai Z, Yan Z, Hua Y, Xu C. (2016). P-glycoprotein overexpression in bone marrow-derived multipotent stromal cells decreases the risk of steroid-induced osteonecrosis in the femoral head. J Cell Mol Med 20(11):2173–2182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hesse E, Schroder S, Brandt D, Pamperin J, Saito H, Taipaleenmaki H. (2019). Sclerostin inhibition alleviates breast cancer-induced bone metastases and muscle weakness. JCI Insight 5(9):e125543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Infante M, Fabi A, Cognetti F, Gorini S, Caprio M, Fabbri A. (2019). RANKL/RANK/OPG system beyond bone remodeling: involvement in breast cancer and clinical perspectives. J Exp Clin Cancer Res 38(1): 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kawai M (2013). Adipose tissue and bone: role of PPARγ in adipogenesis and osteogenesis Horm Mol Biol Clin Investig 15(3): 105–113. [DOI] [PubMed] [Google Scholar]
  14. Li J, Li X, Liu D, Hamamura K, Wan Q, Na S, Yokota H, Zhang P. (2019). eIF2α signaling regulates autophagy of osteoblasts and the development of osteoclasts in OVX mice. Cell Death Dis 10(12):921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Li J, Zhang N, Huang X, Xu J, Fernandes JC, Dai K, Zhang X. (2013). Dexamethasone shifts bone marrow stromal cells from osteoblasts to adipocytes by C/EBPα promoter methylation. Cell Death Dis 4(10):e832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Li X, Liu D, Li J, Yang S, Xu J, Yokota H, Zhang P. (2019). Wnt3a involved in the mechanical loading on improvement of bone remodeling and angiogenesis in a postmenopausal osteoporosis mouse model. FASEB J 33(8):8913–8924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Li X, Yang J, Liu D, Li J, Niu K, Feng S, Yokota H, Zhang P. (2016). Knee loading inhibits osteoclast lineage in a mouse model of osteoarthritis. Sci Rep 6:24668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lim H, Yeo E, Song E, Chang YH, Han BK, Choi HJ, Hwang J. (2015). Bioconversion of Citrus unshiu peel extracts with cytolase suppresses adipogenic activity in 3T3-L1 cells. Nutr Res Pract 9(6):599–605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Liu D, Li X, Li J, Yang J, Yokota H, Zhang P. (2015). Knee loading protects against osteonecrosis of the femoral head by enhancing vessel remodeling and bone healing. Bone 81:620–631. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu D, Zhang Y, Li X, Li J, Yang S, Xing X, Fan G, Yokota H, Zhang P. (2017). eIF2α signaling regulates ischemic osteonecrosis through endoplasmic reticulum stress. Sci Rep 7(1):5062. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu W, Chen G, Sun L, Zhang Y, Han J, Dai Y, He J, Shi S, Chen B. (2019). TUFT1 promotes triple negative breast cancer metastasis, sternness, and chemoresistance by up-regulating the Rac1/β-catenin pathway. Front Oncol 9:617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Maroni P, Bendinelli P. (2020). Bone, a secondary growth site of breast and prostate carcinomas: role of osteocytes. Cancers (Basel) 12(7):E1812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Morris EV, Edwards CM. (2016). The role of bone marrow adipocytes in bone metastasis. J Bone Oncol 5(3): 121–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pan L, Chen X, Fu S, Yu W, Li C, Wang T, Lo HW, Lin J. (2020). LLY17, a novel small molecule STAT3 inhibitor induces apoptosis and suppresses cell migration and tumor growth in triple-negative breast cancer. Breast Cancer Res Treat 181(1):31–41. [DOI] [PubMed] [Google Scholar]
  25. Pan R, Zhu M, Yu C, Lv J, Guo Y, Bian Z, Yang L, Chen Y, Hu Z, Chen Z, Li L, Shen H, China Kadoorie Biobank Collaborative G. (2017). Cancer incidence and mortality: A cohort study in China, 2008-2013. Int J Cancer 141(7):1315–1323. [DOI] [PubMed] [Google Scholar]
  26. Park SY, Kim HJ, Kim KR, Lee SK, Lee CK, Park KK, Chung WY. (2014). Betulinic acid, a bioactive pentacyclic triterpenoid, inhibits skeletal-related events induced by breast cancer bone metastases and treatment. Toxicol Appl Pharmacol 275(2): 152–162. [DOI] [PubMed] [Google Scholar]
  27. Picon-Ruiz M, Morata-Tarifa C, Valle-Goffin JJ, Friedman ER, Slingerland JM. (2017). Obesity and adverse breast cancer risk and outcome: Mechanistic insights and strategies for intervention. CA Cancer J Clin 67(5):378–397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Rachner TD, Benad P, Rauner M, Goettsch C, Singh SK, Schoppet M, Hofbauer LC. (2009). Osteoprotegerin production by breast cancer cells is suppressed by dexamethasone and confers resistance against TRAIL-induced apoptosis. J Cell Biochem 108(1): 106–116. [DOI] [PubMed] [Google Scholar]
  29. Sadeghi F, Vahednia E, Naderi Meshkin H, Kerachian MA. (2020). The effect of adrenocorticotropic hormone on α-2-macroglobulin in osteoblasts derived from human mesenchymal stem cells. J Cell Mol Med 24(8):4784–4790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Seyfi D, Behzad SB, Nabiuni M, Parivar K, Tahmaseb M, Amini E. (2018). Verbascoside Attenuates Rac-1 and HIF-1α Signaling Cascade in Colorectal Cancer Cells. Anti cancer Agents Med Chem 18(15):2149–2155. [DOI] [PubMed] [Google Scholar]
  31. Shin B, Won H, Adams DJ, Lee SK. (2020). CD55 regulates bone mass in mice by modulating RANKL-mediated Rac signaling and osteoclast function. J Bone Miner Res 35(1): 130–142. [DOI] [PubMed] [Google Scholar]
  32. Shin E, Koo JS. (2020). The role of adipokines and bone marrow adipocytes in breast cancer bone metastasis. Int J Mol Sci 21(14) :E4967. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Shirakawa J, Takegahara N, Kim H, Lee SH, Sato K, Yamagishi S, Choi Y. (2019). Flrt2 is involved in fine-tuning of osteoclast multinucleation. BMB Rep 52(8):514–519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Su X, Zhang L, Ye J, Yang L, Li Y, Wang Y. (2015). Bone marrow mesenchymal stem cells suppress ascitogenous hepatoma progression in BALB/c mouse through reducing myeloid-derived suppressor cells. Biomed Mater Eng 25(1 Suppl): 167–177. [DOI] [PubMed] [Google Scholar]
  35. Takada I, Kouzmenko AP, Kato S. (2009). Wnt and PPARγ signaling in osteoblastogenesis and adipogenesis. Nat Rev Rheumatol 5(8):442–447. [DOI] [PubMed] [Google Scholar]
  36. Tan N, Li X, Zhai L, Liu D, Li J, Yokota H, Zhang P. (2018). Effects of knee loading on obesity-related non-alcoholic fatty liver disease in an ovariectomized mouse model with high-fat diet. Hepatol Res 48(10):839–849. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Tian Y, Xu L, He Y, Xu X, Li K, Ma Y, Gao Y, Wei D, Wei L. (2018). Knockdown of RAC1 and VASP gene expression inhibits breast cancer cell migration. Oncol Lett 16(2):2151–2160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Timaner M, Tsai KK, Shaked Y. (2020). The multifaceted role of mesenchymal stem cells in cancer. Semin Cancer Biol 60:225–237. [DOI] [PubMed] [Google Scholar]
  39. Wu X, Li F, Dang L, Liang C, Lu A, Zhang G. (2020). RANKL/RANK system-based mechanism for breast cancer bone metastasis and related therapeutic strategies. Front Cell Dev Biol 8:76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Xiong Z, Luo P, Zhou J, Tan M. (2019). 15-Deoxy-Δ12,14-prostaglandin J2 as a potential regulator of bone metabolism via PPARγ-dependent and independent pathways: a review. DrugDesDevel Ther 13:1879–1888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Xu W, Neill T, Yang Y, Hu Z, Cleveland E, Wu Y, Hutten R, Xiao X, Stock SR, Shevrin D, Kaul K, Brendler C, Iozzo RV, Seth P. (2015). The systemic delivery of an oncolytic adenovirus expressing decorin inhibits bone metastasis in a mouse model of human prostate cancer. Gene Ther 22(3):247–256. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yang S, Liu H, Zhu L, Li X, Liu D, Song X, Yokota H, Zhang P. (2019). Ankle loading ameliorates bone loss from breast cancer-associated bone metastasis. FASEB J 33(10): 10742–10752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Yokota H, Hamamura K, Chen A, Dodge TR, Tanjung N, Abedinpoor A, Zhang P. (2013). Effects of salubrinal on development of osteoclasts and osteoblasts from bone marrow-derived cells. BMC Musculoskelet Disord 14:197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zhang P, Hamamura K, Yokota H, Malacinski GM. (2009). Potential applications of pulsating joint loading in sports medicine. Exerc Sport Sci Rev 37(1):52–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang P, Malacinski GM, Yokota H. (2008). Joint loading modality: its application to bone formation and fracture healing. Br J Sports Med 42(7):556–560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang P, Su M, Liu Y, Hsu A, Yokota H. (2007). Knee loading dynamically alters intramedullary pressure in mouse femora. Bone 40(2):538–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zheng W, Li X, Liu D, Li J, Yang S, Gao Z, Wang Z, Yokota H, Zhang P. (2019). Mechanical loading mitigates osteoarthritis symptoms by regulating endoplasmic reticulum stress and autophagy. FASEB J 33(3):4077–4088. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zhu L, Tang Y, Li XY, Keller ET, Yang J, Cho JS, Feinberg TY, Weiss SJ. (2020). Osteoclast-mediated bone resorption is controlled by a compensatory network of secreted and membrane-tethered metalloproteinases. Sci Transl Med 12(529):eaaw6143. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES