Applying vibration in early postmenopausal osteoporosis promotes osteogenic differentiation of bone marrow-derived mesenchymal stem cells and suppresses postmenopausal osteoporosis progression

We aimed to evaluate whether applying low magnitude vibration (LMV) in early postmenopausal osteoporosis (PMO) suppresses its progression, and to investigate underlying mechanisms. Rats were randomly divided into Sham (Sham-operated), Sham+V, OVX (ovariectomized), OVX+E2 (estradiol benzoate), OVX+V (LMV at 12–20 weeks postoperatively), and OVX+Vi (LMV at 1–20 weeks postoperatively) groups. LMV was applied for 20 min once daily for 5 days weekly. V rats were loaded with LMV at 12–20 weeks postoperatively. Vi rats were loaded with LMV at 1–20 weeks postoperatively. Estradiol (E2) rats were intramuscularly injected at 12–20 weeks postoperatively once daily for 3 days. The bone mineral densities (BMDs), biomechanical properties, and histomorphological parameters of tibiae were analyzed. In vitro, rat bone marrow-derived mesenchymal stem cells (rBMSCs) were subjected to LMV for 30 min daily for 5 days, or 17β-E2 with or without 1-day pretreatment of estrogen receptor (ER) inhibitor ICI 182,780 (ICI). The mRNA and protein expresion were performed. Data showed that LMV increased BMD, bone strength, and bone mass of rats, and the effects of Vi were stronger than those of E2. In vitro, LMV up-regulated the mRNA and protein expressions of Runx2, Osx, Col I, and OCN and down-regulated PPARγ, compared with E2. The effects of both LMV and E2 on rBMSCs were inhibited by ICI. Altogether, LMV in early PMO suppresses its progression, which is associated with osteogenic differentiation of rBMSCs via up-regulation of ERα and activation of the canonical Wnt pathway. LMV may therefore be superior to E2 for the suppression of PMO progression.


Introduction
Postmenopausal osteoporosis (PMO) usually occurs 5-10 years after menopause, and is mainly caused by a decrease in estrogen. It is a type of systemic skeletal disease characterized by osteopenia, deterioration of bone microarchitecture, increased bone fragility, and susceptibility to fracture [1]. With the aging global population, the incidence of PMO is rapidly rising. PMO seriously influences patients' health and quality of life, placing a heavy burden on patients' families and the national economy [2]. Osteoporotic fractures have become a public health problem. It is estimated that 84% of patients with osteoporotic fractures did Rats were weighed every 4 weeks during the experiment, and all animals received hypodermic injections of 10 mg/kg Calcein (Sigma, U.S.A.) on days 126, 127, 136, and 137 postoperatively to label bone forming surfaces. At the end of the experiment, rats were killed by cervical dislocation. Each uterus was extracted and weighed. The bilateral tibiae were cleaned off skin, muscle, and tendons. Left tibiae were wrapped in saline-soaked gauze and stored at −20 • C for BMD detection and three-point bending tests, while right tibiae were fixed overnight in 4% paraformaldehyde and then placed in 70% ethanol for histomorphometric analysis. All the experimental procedures were approved by the Ethics Committee of Sichuan University. All experiments were performed at Institute of Biomedical Engineering, West China School of Basic Medical Science and Forensic Medicine of Sichuan University.

BMD detection
The left tibiae were scanned with the VivaCT80 micro-computed tomography system (Scanco Medical, Switzerland). The scanning parameters were 70 kV and 113 μA. The samples were scanned with high resolution, and each layer was scanned at a thickness of 19 μm. After completing the overall scan of the tibia, we selected a 'region of interest' (ROI) at 1.5 mm below the upper growth plate, then 3 mm below the first point for 3D reconstruction. Then, the 3D structure images were automatically evaluated for BMD using the μCT Tomography software (V6.3, Scanco Medical) .

Bone biomechanical test
Biomechanical properties of left tibiae were evaluated by three-point bending test using a universal material test machine (AGIS-201, SHIMADZU, Japan). The two ends of the tibia were placed on a special bracket with a span of approximately 35 mm so that they did not move during the experiment. A cylindrical indenter moved down to the point of 5 mm from the proximal end of the tibia with a continuous displacement of 1 mm/min until the metaphyseal tibia broke. Then, the damaged part of the metaphyseal tibia was removed, and the rest of the diaphyseal tibia was placed on two cylindrical brackets with a span of 20 mm. A cylindrical indenter moved down to the midpoint of the diaphyseal tibia with a continuous displacement of 1 mm/min until the diaphyseal tibia broke. From the Force-Stroke curve, biomechanical parameters including elastic modulus, maximum stress, and failure stress were determined.

Bone histomorphometric analysis
Right tibiae were dehydrated by sequential ascending concentrations of ethanol and degreased by xylene. The metaphysis and diaphysis of the tibiae were embedded in methyl methacrylate (Sigma, U.S.A.), respectively. The embedded metaphyseal tibia specimens were cut longitudinally into sections of 5 μm along the sagittal plane using a hard tissue microtome (RM 2155, Leica, Germany). The sections were stained with Masson-Goldner Trichrome for measuring static parameters. Starting at the junction of the tibia and fibula, and toward the proximal end of the tibia, the diaphyseal tibiae were sawed transversely into 100-μm slices, and the second slices were taken and ground into 40-5 μm slices, observed with a fluorescence microscope (MGC30, Leica, Germany).
A semi-automatic image analysis system (OsteoMeasure, OsteoMetrics, U.S.A.) was used to quantitatively measure the secondary cancellous bone between 1 and 4 mm below the epiphyseal line of the metaphyseal tibia and the cortical bone of the diaphyseal tibia. rBMSCs isolation, culture, and grouping rBMSCs were obtained from 4-week-old female Sprague-Dawley rats (average body weight of 90 + − 10 g, the Laboratory Animal Center of Sichuan University, Chengdu, China). Briefly, the bone marrow of the tibias and femurs was flushed out with DMEM-LG basal medium (Gibco, U.S.A.), centrifuged, and resuspended with DMEM-LG complete medium containing 85% DMEM-LG medium, 15% fetal bovine serum (FBS) (Sigma, U.S.A.), and 1% penicillin-streptomycin solution (HyClone, U.S.A.), and was cultured in a humidified incubator with 5% CO 2 at 37 • C. The medium was replaced every 3 days, and adherent cells were passaged at 70-80% confluence. The third passage cells were divided into five groups for experiments: (1) control group, in which cells were cultured in DMEM-LG complete medium for 6 days; (2) E2 group, in which cells were cultured in DMEM-LG complete medium for 1 day, then cultured in DMEM-LG complete medium containing 100 nM 17β-E2 (Sigma, U.S.A.) for 5 days; (3) V group, in which cells were cultured in DMEM-LG complete medium for 1 day, and then subjected to vibration of 0.9×g, 45 Hz, and 30 min/day for 5 days; (4) ICI+E2 group, in which cells were cultured in DMEM-LG complete medium containing 50 ng/ml ICI 182,780 (ER inhibitor) (MCE, U.S.A.) for 1 day, then cultured in DMEM-LG complete medium containing 100 nM 17β-E2 for 5 days; and (5) ICI+V group, in which cells were cultured in DMEM-LG complete

Real-time quantitative RT-PCR
Total RNA was isolated from cells of each group using TRIzol reagent (Invitrogen, U.S.A.), and reverse transcribed to cDNA using PrimeScript™ RT reagent Kit (TaKaRa, China). A quantitative RT-PCR assay was performed on a CFX96™ Real-Time System (Bio-Rad, U.S.A.) using specific primers and SYBR ® Premix ExTaq™ (TaKaRa, China). Primers were designed as shown in Table 1. Relative expression values were calculated using the comparative threshold cycle (2 − C T ) [21]. GAPDH served as the housekeeping gene.

Western blotting
Cells were washed with iced PBS and lysed in RIPA lysis buffer (Beyotime, China). Whole cell proteins were quantified by the BCA Protein Assay Kit (Beyotime, China), separated by 10% SDS/PAGE, and then transferred on to PVDF membrane (Millipore, U.S.A.) and blocked with Tris-buffered saline buffer with Tween 20 (TBST) containing 5% non-fat milk powder. Membranes were incubated with primary antibodies overnight at 4 • C, washed with TBST, and incubated with secondary antibodies conjugated with horseradish peroxidase (Boster, China) for 2 h at room temperature. Primary antibodies used in the present study included those against β-actin, ERα, Wnt3a, and β-catenin (Proteintech, U.S.A.). Immunoreactive bands were visualized with an enhanced luminol-based chemiluminescent substrate (ECL) Kit (Beyotime, China) and quantified using ImageJ Plus software (National Institutes of Health, U.S.A.), normalized by β-actin.

Statistical analysis
All data were analyzed by SPSS 20.0 software (IBM Corp, Armonk, NY). Each test was repeated at least three times, and data were expressed as mean + − standard deviation (SD). Student's t test was used for comparison between two groups or one-way analysis of variance (ANOVA) for multiple groups, followed by Student-Newman-Keuls (S-N-K) test. P<0.05 was considered as statistically significant.

LMV increased the BMD in both Sham and OVX rats
The BMD of the left tibiae in ovariectomized osteoporosis (OVX) rats was significantly reduced by 26.7% as compared with the sham-operated (Sham) rats (P<0.01, Figure 1A,B). LMV (vibration from 12 to 20 weeks after the operation, V) significantly increased the BMD in both Sham and OVX rats, and the increase by V in OVX rats was significantly higher than by E2 (treated from 12 to 20 weeks after the operation). Interestingly, when loading the vibration after operation immediately (vibration from 1 to 20 weeks after ovariectomy, Vi), the BMD was significantly more than the V and E2 groups at the end time points (20 weeks, Figure 1), suggesting strong protection from bone loss by vibration, which was more effective during the early stage of osteoporosis.

Uterine weight unchanged after LMV and increased after E2
The body weight of the OVX rats was significantly increased compared with the Sham rats ( Figure 2A). After E2   treatment for 4 weeks (from 12 to 16 weeks postoperatively, the average body weight of OVX rats was significantly decreased 17.9% (P<0.05). Significant changes in body weight were not observed in rats loaded with V and Vi ( Figure  2A). Ovariectomy significantly resulted in 64.5% decrease in the uterus-to-body weight ratio (P<0.01, Figure 2B), with obvious atrophy of the uterus ( Figure 2C). The uterus-to-body weight ratio in OVX rats was significantly increased 555.6% (P<0.01), and uteri were enlarged by E2 treatment, but did not change by V and Vi ( Figure 2B,C). These results suggested that E2 reduced the body weight and enlarged the uteri in OVX rats, but V and Vi did not.

LMV improved biomechanical properties of the tibiae
The biomechanical properties of tibiae were analyzed by three-point bending test (Figure 3). The elastic modulus, maximum stress, and failure stress of the metaphyseal and diaphyseal tibiae were significantly reduced in OVX rats compared with Sham rats. The elastic modulus, maximum stress, and failure stress in OVX rats were significantly increased by V and Vi. Compared with E2, V and Vi loading resulted in a significant increase in the elastic modulus and maximum stress of tibiae and the failure stress of diaphyseal tibiae (Figure 3), suggesting that LMV improved the biomechanical properties of bones in OVX rats significantly compared with E2.  Figure 4B) and trabecular number (Tb.N, Figure 4C) of the OVX rats were significantly decreased 82.5% (P<0.01) and 85.6% (P<0.01), respectively, compared with the Sham rats, while the trabecular separation (Tb.Sp) was significantly increased 624.5% (P<0.01, Figure 4D). V loading significantly increased the %Tb.Ar 67.0% (P<0.05) and reduced the Tb.Sp 37.0% (P<0.05). Vi loading significantly reversed the changes in all three morphological parameters in the OVX rats. A significant role of E2 treatment was not observed. These results suggested that the application of LMV in early PMO could effectively reduce the loss of cancellous bone mass in the metaphyseal tibia.

Vibration promoted osteogenic differentiation but inhibited adipogenic differentiation of rBMSCs via ERα
rBMSCs were isolated and treated with vibration (V) and E2, respectively ( Figure 6). The qRT-PCR and Western blotting results showed that osteogenesis-related molecules including Runx2, Osx, Col I, and OCN in the E2 and V groups were significantly increased compared with control, and the effect in V group was more obvious (Figure 6A,B). The mRNA and protein expressions of adipogenesis-related molecule PPARγ in the V group were significantly decreased (P<0.05, Figure 6C,D). Moreover, osteoclastogenesis-related molecule RANKL was significantly decreased in the V group (P<0.05, Figure 6C,D) but did not significantly change by E2. Both E2 and V significantly increased the mRNA and protein expressions of osteoprotegerin (OPG) (Figure 6C,D). The ratio of RANKL/OPG mRNA and protein in the E2 and V groups were significantly decreased, and the effect in V group was more obvious (Figure 6C,D). These  results suggested that vibration promoted the osteogenic differentiation of rBMSCs, inhibited the adipogenic differentiation of rBMSCs, and regulated the osteoclastogenesis, while E2 promoted the osteogenic differentiation but did not inhibit the adipogenic differentiation of rBMSCs, indicating the vibration is a better treatment than E2.
To investigate the role of ERs, ERα was blocked by ER inhibitor ICI ( Figure 6). ICI significantly abolished the effect of vibration and E2 on the expression of osteogenesis-related genes, adipogenesis-related genes, and osteoclastogenesis-related genes, suggesting that vibration promoted the osteogenic differentiation of rBMSCs, inhibited the adipogenic differentiation of rBMSCs, and regulated the osteoclastogenesis via ERα.

Vibration up-regulated the expressions of ERα, Wnt3a, and β-catenin
The expression of ERα, Wnt3a, and β-catenin was detected by qRT-PCR ( Figure 7A) and Western blotting ( Figure  7B). E2 and V significantly increased the mRNA and protein expression of ERα, Wnt3a, and β-catenin in rBMSCs, and V showed a more obvious effect. Moreover, the ER inhibitor ICI suppressed the effects of E2 and V on the expression of ERα, Wnt3a, and β-catenin. These results suggested that vibration activated the canonical Wnt/β-catenin pathway in rBMSCs via ERα.

Discussion
Exercise therapy has been considered the safest and most effective way to strengthen bones to date, and is also an effective measure to prevent PMO, helping to minimize the side effects of treatment [16]. This study showed that LMV prevented PMO in rats, and that the application of LMV in early PMO was more effective than E2.
PMO previously relied on drug therapy. The drugs used for PMO treatment include two categories based on their roles in inhibiting bone resorption and promoting bone formation. Among the approved drugs, only teriparatide (human recombinant parathyroid hormone) promotes bone formation; the others, such as bisphosphonates, denosumab (RANKL monoclonal antibodies), and raloxifene (selective ER modulator) inhibit bone resorption, and cannot fundamentally solve the problem of reduced bone-forming ability in patients. Moreover, although these drugs can delay or stop bone absorption, they usually require long-term use, are expensive, and have many potential side effects and associations, including gastric ulcer, esophageal cancer, atrial fibrillation, atypical femoral fracture, and osteonecrosis of the jaw, making it difficult for patients to accept and adhere to long-term medication [22,23]. In this study, we confirmed that E2 increased the BMD in OVX rats, but induced hyperenlargement of the uterus. LMV not only increased the BMD, but also improved the biomechanical properties of bone in OVX rats without certain side effects, like E2.
High magnitude vibration (regardless of frequency) has been identified as a hazardous factor by the International Organization for Standardization (ISO) [24]. LMV with an acceleration of gravity of less than 1×g was usually used in therapy of osteoporosis. Zhou et al. [12] applied vibration of 0.3×g and 40 Hz for 30 min/12 h and 5 days/week to ovariectomized rats, and found after 12 weeks that vibration can enhance osseointegration by improving microstructure parameters surrounding implants. Li et al. [25] found that vibration of 0.25×g, 35 Hz, and 15 min/day for 8 weeks can up-regulate the expression of osteogenesis-related proteins in bone tissues of OVX rats. Qing et al. [26] detected that after 8 weeks of vibration at 0.3×g, 30 Hz, and 20 min/day, the deterioration of trabecular bone in OVX rats was ameliorated and the tibia trabecular BMD was significantly increased. However, Brouwers et al. [27] used a vibration of 0.3×g, 90 Hz on ovariectomized rats, twice a day for 20 min, 5 days a week for 6 weeks; no changes in structure and strength of tibiae were observed, and the bone structure in rats was still degraded. Xie et al. [28] concluded that vibration of 0.3×g, 30 Hz, and 20 min/day for 16 weeks might exacerbate trabecular bone loss in ovariectomized rat femurs. The different effects of vibration on OVX rats might be because of the tested parameters.
In the preliminary experiment, we loaded various parameters on the rats from 12 to 20 weeks including 0.9×g and 90 Hz, 0.9×g and 45 Hz, 0.3×g and 90 Hz, and 0.3×g and 45 Hz, and the results showed that 0.9×g and 45 Hz significantly promoted the bone formation (data now shown). Thus, in the present study, we selected an LMV of 0.9×g and 45 Hz for 20 min everyday, once a day and 5 days a week for 8 weeks (from 12 to 20 weeks postoperatively) or 19 weeks (from 1 to 20 weeks postoperatively). It was demonstrated that the estrogen was significantly elevated in the first 4 weeks after ovariectomy, and then was reduced significantly at 8 or 12 weeks after ovariectomy [29]. Therefore, we did not treat rats with E2 at the first week. We found that LMV increased BMD and bone strength and promoted bone formation of OVX rats significantly better than E2. Moreover, the application of LMV in early osteoporosis was more effective.
BMSCs are stress-sensitive cells that are constantly stimulated by the mechanical environment within the marrow cavity. They play important roles in osteogenic differentiation and bone formation during fracture healing [30]. We demonstrated that LMV promoted the differentiation of rBMSCs into osteoblasts and inhibited their differentiation into adipocytes. This is consistent with the results of Luu et al. [31]. E2 could only promote osteogenic differentiation, but could not inhibit adipogenic differentiation of rBMSCs. These results indicate that LMV is more effective than E2 in promoting the bone formation of rBMSCs.
BMSCs can also secrete OPG, which competitively binds to RANKL, causing RANKL to lose binding activity to RANK, thus inhibiting the generation of osteoclasts [32]. Therefore, the balance between RANKL and OPG determines the generation and activity of osteoclasts. In our study, LMV decreased the ratio of RANKL/OPG in rBMSCs, suggesting that LMV reduced the generation and activation of osteoclasts and reduced bone resorption. This is consistent with the results of Pichler et al. [33] and Lau et al. [34] on bone cells.
Our study found that LMV and E2 caused a significant increase in ERα expression in rBMSCs and promoted the expression of Wnt3a and β-catenin. It was previously reported that ER and Wnt3a can synergistically promote bone formation [35]. These LMV effects could be inhibited by ER inhibitor ICI. Furthermore, ICI inhibited the LMV-promoted differentiation of rBMSCs into osteoblasts. The estrogen pathways in LMV effects should be further studied using shRNA in the future. We did not observe a synergistic effect of E2 and vibration (data not shown).
In summary, LMV promotes osteogenic differentiation, but inhibits adipogenic differentiation of rBMSCs via the canonical Wnt pathway by up-regulating ERα. LMV promotes bone formation with no significant effects on body