Low-dose strontium stimulates osteogenesis but high-dose doses cause apoptosis in human adipose-derived stem cells via regulation of the ERK1/2 signaling pathway

Background Strontium is a widely used anti-osteoporotic agent due to its dual effects on inhibiting bone resorption and stimulating bone formation. Thus, we studied the dose response of strontium on osteo-inductive efficiency in human adipose-derived stem cells (hASCs). Method Qualitative alkaline phosphatase (ALP) staining, quantitative ALP activity, Alizarin Red staining, real-time polymerase chain reaction and Western blot were used to investigate the in vitro effects of a range of strontium concentrations on hASC osteogenesis and associated signaling pathways. Results In vitro work revealed that strontium (25–500 μM) promoted osteogenic differentiation of hASCs according to ALP activity, extracellular calcium deposition, and expression of osteogenic genes such as runt-related transcription factor 2, ALP, collagen-1, and osteocalcin. However, osteogenic differentiation of hASCs was significantly inhibited with higher doses of strontium (1000–3000 μM). These latter doses of strontium promoted apoptosis, and phosphorylation of ERK1/2 signaling was increased and accompanied by the downregulation of Bcl-2 and increased phosphorylation of BAX. The inhibition of ERK1/2 decreased apoptosis in hASCs. Conclusion Lower concentrations of strontium facilitate osteogenic differentiation of hASCs up to a point; higher doses cause apoptosis of hASCs, with activation of the ERK1/2 signaling pathway contributing to this process.


Background
Osteoporosis is a common disease with reduced bone density, and as many as half of all postmenopausal women will have an osteoporosis-related fracture during their lifetime [1]. Anti-osteoporotic drugs, including bisphosphonates and estrogen, and selective receptor-modulating drugs reduce fractures by reducing the number, activity, and lifespan of osteoclasts [2]. Compared to these antiosteoporosis drugs, strontium ranelate (SrRan) has been shown to exert its function via promoting bone formation and simultaneously inhibiting bone resorption [3]; this was shown in several large clinical trials confirming the decreased risk of vertebral fractures in postmenopausal women with osteoporosis. SrRan may improve bone formation due to the stimulation of bone marrow stromal cell (BMSC) osteogenic differentiation. Saidak et al. [4] reported that SrRan counteracts age-related switches in BMSCs from osteoblasts to adipocytes via Wnt5a and NFATc/Maf signaling.
Mesenchymal stem cells (MSCs) are multipotent cells that have the potential for regeneration of damaged tissues in bone disease. Adipose-derived MSCs (ASCs) have several advantages, including immunoprivilege, higher yield at harvest, rapid expansion, and greater genetic stability during long-term culture [5,6]. Evidence suggests that the administration of ASCs accelerates fracture healing and bone regeneration by direct stimulation of osteogenic differentiation or by secretion of high concentrations of cytokines.
Ca, P, Si-containing bioactive ceramic dissolution products, and SrCaPO 4 (52.12% Sr) enhance the proliferation and osteogenic differentiation of rabbit ASCs [7,8]. Another study [9,10] indicated that Sr-hardystonite-gahnite (Sr-HT-gahnite), a highly porous and biocompatible Srzinc containing bioceramic (7.7 mg/L Sr ion), can induce osteogenesis of rat ASCs in vitro and enhance vascularized bone regeneration in vivo. However, SrRan is implicated in venous thromboembolism (VTE) and the risk of nonfatal myocardial infarction in women with postmenopausal osteoporosis [11]. Moreover, SrRan has been associated with significantly greater risk of death and cardiovascular risk compared with other osteoporotic drugs [12]. Thus, we should study the safety of SrRan with respect to osteogenic differentiation of ASCs.
Mitogen-activated protein kinases (MAPKs) regulate various physiological processes, such as cell growth, differentiation, and apoptosis [13]. Extracellular signal-related kinase (ERK)1/2 is a member of the MAPK family and is involved in the regulation of the cellular response to apoptotic-promoting signals [14]. Studies suggest that aberrant ERK activation can promote apoptosis under certain conditions, and our work suggests that osteogenic differentiation of human ASCs (hASCs) is mediated by ERK activation [15]. However, how ERK signaling functions during osteogenic differentiation of ASCs stimulated by SrRan is unclear. Thus, we studied the effect of SrRan over a range of doses on osteogenic differentiation of hASCs and assessed SrRan cytotoxicity.

Patients and ethics approval
Fresh human lipoaspirates were obtained from five healthy individuals (average age of 35 years) who received abdominal liposuction at the Department of Plastic and Reconstructive Surgery in Xinjiang Medical University First Affiliated Hospital (Xinjiang, China). Before surgery, all the patients offered written informed consent, and the protocols for human tissue handling were approved by Xinjiang Medical University First Affiliated Hospital.

Cell culture
Isolation and culture of hASCs was performed according to a previous work [16]. In brief, after washing with an equal volume of 0.1 M phosphate-buffered saline (PBS; pH 7.4), lipoaspirates were digested with 0.075% collagenase type I (Washington Biochemical Corp.) at 37°C for 60 min. The digested samples were centrifuged at 1200 × g for 10 min to obtain a high-density stromal vascular fraction (SVF), and they were resuspended in low-glucose Dulbecco's modified Eagle's medium (LG-DMEM) containing 10% fetal bovine serum (FBS), 100 mg/mL streptomycin, and 100 U/mL penicillin (growth medium (GM)). They were plated at 4 × 10 4 cells/cm 2 in 100-mm culture dishes (Falcon). When they reached 70-80% confluence, the cells were passaged and hASCs before the third passage were used in the following study.

Alkaline phosphatase staining and activity assay
At day 10 after induction, alkaline phosphatase (ALP) staining was performed using a leukocyte AP staining kit (System Biosciences). Briefly, the cells were fixed with formalin for 10 min, followed by incubation with BCIP/NBT working solution for 30 min in the dark at room temperature. After rinsing twice with deionized water, the cells were stained with naphthol AS-MX phosphate for 30 min in the dark. Excess dye was removed by washing with PBS three times. Positive staining of ALP was observed under a light microscope and quantified with Image-Pro Plus 5.0 software. ALP activity was measured using p-nitrophenylphosphate as the substrate. Absorbance at 405 nm was measured and the protein in cell lysates was measured using a Bradford assay at 595 nm on a microplate spectrophotometer (Bio-Rad). ALP activity was normalized according to the total protein.

Accumulated calcium assay
On day 21 of osteogenic differentiation, the medium was removed and the cells were fixed with 70% ice-cold ethanol for 1 h, followed by incubation in 40 mM Alizarin Red S (Sigma) at pH 4.2 for 30 min at room temperature. After rinsing with fresh PBS and drying at room temperature, the calcium deposits were quantified using a Quanti-Chrom Calcium Assay Kit (BioAssay Systems). A comparison with an Alizarin Red S dye standard curve was used to evaluate the calcium deposition.

Oil Red O staining
After 10 days of adipogenic induction, hASCs were stained with Oil Red O to measure the accumulation of intracellular lipid droplets. Briefly, the cells were fixed with 10% formalin for 30 min, washed in PBS, and stained with a 0.6% (w/v) Oil Red O solution for 15 min at room temperature. After rinsing with deionized water, fat droplet staining was photographed under a phasecontrast microscope (Olympus, Tokyo, Japan).

Live/dead viability staining
Live and dead cells were measured using an assay kit (Invitrogen) according to the manufacturer's instructions. After incubation with SrRan for 24 and 48 h, hASCs were stained with the live/dead reagent (ethidium homodimer and calcein-AM) at 37°C for 30 min. hASC viability was analyzed under a fluorescent microscope (Nikon, Tokyo, Japan).

TUNEL assay
To measure apoptotic hASCs, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was used with an In Situ Cell Death Detection Kit (Roche, Mannheim, Germany). Cultured hASCs were fixed with 4% paraformaldehyde for 1 h at room temperature, washed three times with PBS, and then permeabilized in 0.1% Triton X-100. Fixed cells were exposed to freshly prepared TUNEL reaction mixture for 1 h in the dark. Cells were then washed with PBS and observed under a fluorescent microscope (Nikon, Tokyo, Japan). Chondrogenic differentiation was verified with immunofluorescent staining for collagen type II (f). ALP activity of hASCs at indicated days after incubation with osteogenic differentiation medium (g) Table 1 Sequences of primers and real-time polymerase chain reaction conditions Gene Primer sequences ALP alkaline phosphatase, Cbfα1 core-binding factor α1, COL I collagen-1, OCN osteocalcin   Total RNA was isolated using Trizol reagent (Invitrogen) and 2 μg of total RNA was reverse-transcribed into 20 μl of cDNA. qRT-PCR analysis was performed using the Step One Plus Real-Time PCR System (Applied Biosystems). Primer sequences are shown in Table 1.

Western blot
For Western blotting, extracted proteins were resolved with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore). After incubation in 5% bovine serum albumin (BSA) at room temperature, the membranes were incubated with primary antibody at 4°C overnight. Horseradish peroxidase (HRP)-conjugated immunoglobulin G (IgG) was used as the secondary antibody. Immunostained protein bands were visualized with an ECL kit (CWBIO, Beijing, China). Protein was quantified and normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH).

Statistical analysis
Data are presented as means ± standard deviation (SD) from three experiments, and each was an average of the triplicate experiments. SPSS 16.0 software was used for statistical analysis. A paired Student's t test was used to compare groups (p < 0.05 was considered to be statistically significant).

Characterization of hASCs
In the in vitro experiments, hASCs were morphologically homogeneous and fibroblast-like ( Fig. 1a and b). Multiple lineage differentiation of hASCs was confirmed by osteogenic, chondrogenic, and adipogenic differentiation ( Fig. 1c-f ). As shown in Fig. 1g, ALP activity was significantly elevated on days 10 and 14 as indicated by ALP staining and quantitative assays. consistent with the quantitative ALP activity data (Fig. 2b), which shows that ALP activity was suppressed at 25, 100, and 250 μM and was enhanced at 500 μM after 4, 7, and 14 days. Alizarin Red S staining showed that, at 25 μM calcium deposition was significantly increased at 14 and 21 days. At 100, 250, and 500 μM, calcium deposition was reduced ( Fig. 3a and b). qRT-PCR showed that the expression of ALP and runt-related transcription factor 2 (RUNX2) was significantly reduced at 25, 100, and 250 μM SrRan, and elevated at 500 μM Fig. 6 a Morphological features of hASCs exposed to SrRan. hASCs were treated with SrRan in GM and OM for 3 days. b hASCs exposed to SrRan in GM and OM, with cell viability measured with a CCK-8 assay. # p < 0.01 and * p < 0.05 compared with the control group a b Fig. 5 a Morphological features of hASCs exposed to SrRan at different time points. b hASCs were treated with SrRan (1000 μM) in GM and OM at the indicated days and viability was measured with a CCK-8 assay. GM growth medium, OM osteogenic medium, Sr strontium compared with OM at 10 and 14 days ( Fig. 4a and b). In contrast, the expression of collagen-1 (COL-I) and osteocalcin (OCN) was slightly upregulated at 25 μM and was significantly suppressed at 100, 250, and 500 μM.
The ERK1/2 signaling pathway is involved in high-dose SrRan-induced apoptosis of hASCs Western blotting revealed that treatment with SrRan at 1000 μM in OM downregulated Bcl-2 and increased phosphorylation of BAX in hASCs at 8 and 12 h after induction. Moreover, Fig. 9 shows that 1000 μM SrRan treatment induced phosphorylation of ERK as early as 8 and 12 h. The phosphorylation of p38 and JNK was not observed. Consistent with TUNEL and live/dead viability staining data, the apoptosis of hASCs treated with high-

Discussion
Data show that strontium promotes osteogenesis of mature osteocytes and undifferentiated MSCs in a dosedependent manner at concentrations from 10 to 100,000 μM [17,18]. However, the effects of low-and high-dose SrRan on osteogenic differentiation of hASCs has not been reported previously. Thus, we investigated the influence of low-dose SrRan on osteogenesis of hASCs at the concentrations used in previous reports [9]. We found that early osteogenic marker expression and ALP activity was increased at 500 μM SrRan, but it augmented late osteogenic gene expression and increased calcium deposition at 25 μM. This result is consistent with Guo et al. [19] who reported that 250-1000 μM SrRan caused osteo-induction with the greatest effect being achieved with 250-500 μM SrRan. A previous study indicated that significant increases in osteogenic gene expression occurred in MSCs cultured in OM containing 100 μM SrRan [20]. Meanwhile, previous studies showed that enhanced osteogenesis occurred in PA20-h5 cell lines treated with SrRan at 50 and 400 μM [9]. Canalis et al. [21] reported that 10 μM SrRan increased bone formation in rat calvarial cells, and another study showed a significant increase in the osteogenic marker RUNX2 in human MSCs after treatment with SrRan from 2.4-240 μM [22]. Studies have also shown that osteo-induction was achieved at 300 μM SrRan in pre-osteoblastic U-33 cells [20]. This is probably because SrRan has a biphasic effect on bone regeneration; dosedependent effects of SrRan at different time points of hASC osteogenesis suggests a complexity of interactions of the effect of cations on bone regeneration. This is in agreement with data from Nardone et al. [9] who showed that ALP production occurred at 400 μM SrRan and calcium deposit formation occurred at lower concentrations (2.5-50 μM). Another study suggested that SrRan may inhibit calcium deposition in human preadipocytes, indicating that alterations of physicochemical properties in the structure of hydroxyapatite crystals impedes its formation [23]. Accordingly, our results demonstrated that SrRan (25-500 μM) promotes hASC osteogenesis in a dose-and time-dependent manner.
However, studies have also shown the opposite results, indicating that SrRan optimally simulates osteogenesis at 1000 μM or greater [24,25]. The expression of RUNX-2 and ALP increased with SrRan greater than 500 μM, and therefore we studied the influence of SrRan at high concentrations on the osteogenesis of hASCs and whether high-dose SrRan elicited any adverse effects. We found that the growth of hASCs was significantly suppressed by SrRan at 1000, 1500, and 2000 μM, and that number of dead hASCs significantly increased at these doses. The TUNEL assay revealed that hASCs exposed to SrRan at these three doses for 48 h had more apoptotic cells. Furthermore, TEM analysis showed typical apoptotic ultramicrostructural changes in hASCs treated with 1000 μM SrRan for 48 h. To our knowledge, this is the first study to show that high-dose SrRan (>1 mM) causes apoptosis of hASCs under osteogenic differentiation. Of note, some studies show that SrRan at high concentrations promotes osteogenic differentiation of MSCs. Fromigué et al. [26] reported that 3000 μM SrCl 2 had the greatest osteogenic induction of MC3T3-E1 cells when combined with SrRan. Peng et al. [27] demonstrated that 4000 μM SrRan increased the expression of RUNX2 and OCN in human MSCs. These differences may be explained by different cell lineages which respond differently to SrRan at similar concentrations [9,24,25]. Zhu et al. [20] showed that 300 μM SrRan enhanced osteogenesis of rat MSCs and pre-osteoblastic U-33 cells but it had no obvious effect on OB-6 cells. Furthermore, target genes associated with osteogenesis are variably activated by SrRan in the same cells. SrRan at 0.1-1 mM significantly upregulated BSP and OCN expression, but RUNX2 expression was not changed in osteoblast-like OB-6 cells. However, with the stimulation of SrRan at similar concentrations, the expression of RUNX2 was increased in bone marrow MSCs and OCN expression was not affected. ERK1/2, a member of the MAPK family, is a confirmed pro-survival factor that participates in the regulation of various cell responses, including proliferation, migration, differentiation, and death [28]. ERK activation is noted in apoptosis caused by diverse stimuli [29,30]. In our previous work, we reported ERK signaling pathway-mediated osteogenic differentiation of hASCs and adipogenic transition initiated by dexamethasone [16]. Here, we report that with the increased apoptosis of hASCs due to higher doses (≥1 mM) of SrRan, ERK was activated and the inhibition of ERK activation using PD98059 reversed hASC apoptosis. Phosphorylation of Fig. 9 SrRan activated MAPK pathways mediated by ERK, JNK, p38, and upregulated the expression of Bax and Bcl-2 in hASCs. a SrRan 1000 μM treatment downregulated Bcl-2 protein, and increased the expression of BAX and phosphorylation of P53in hASCs at 8 and 12 h. b The expression of total and phosphorylated ERK, p38, and JNK was measured by Western blotting in hASCs treated with SrRan 1000 μM at the indicated times. Phosphorylation of ERK was significantly increased at 8 and 12 h. No change occurred with phosphorylation of p38 and JNK after 0, 2, 4, 8, 12, and 24 h. c Results of densitometric scans of blots. Data are means ± SEM for three independent experiments p38 and JNK did not occur and an inhibitor of p38 and JNK had no effect on hASC apoptosis. Thus, the activation of ERK signaling is responsible for high-dose SrRaninduced apoptosis of hASCs and this was not changed by the p38 and JNK pathways. Bcl-2 is a principal family of proteins that protects cells from apoptosis, and Bax induces apoptosis [31]. Thus, the balance between the proand anti-apoptotic Bcl-2 family of proteins influences apoptosis [32]. We found that higher doses of SrRan downregulated Bcl-2 and that the expression of BAX was increased in hASCs. Apoptosis can be initiated by intrinsic or extrinsic pathways and p53 is central to the intrinsic apoptotic pathway [33]. Thus, the phosphorylation of p53 increased in hASCs exposed to high-dose SrRan for 8 and 12 h, suggesting that SrRan-induced apoptosis of hASCs may be modulated by p53 signaling.

Conclusions
We suggest that low-dose SrRan enhances hASC osteogenic differentiation and higher doses causes hASC Fig. 10 ERK signaling contributed to SrRan-induced hASC apoptosis. The apoptosis of hASCs was measured by live/dead staining, AO/EB staining, and TUNEL assays in hASCs treated with strontium 1000 μM at 48 h. The hASC apoptosis caused by high-dose SrRan treatment was greatly attenuated by ERK-specific inhibitors (PD98059); however, JNK (SP600125) and p38 (SB203580) inhibitors showed little effects on hASC apoptosis induced by high-dose SrRan apoptosis via activation of the ERK signaling pathway and subsequent Bcl-2 family-mediated apoptosis. This information may lay the foundation for developing SrRancontaining scaffolds for use in bone tissue engineering.  Fig. 11 Diagram of high-dose SrRan-induced hASC apoptosis by ERK/p53-mediated pathways and subsequent Bcl-2 family-mediated mitochondrial apoptosis. SrRan activates ERK phosphorylation, which may increase phosphorylation of its target gene P53. Activation of P53, in turn, downregulates predicted target genes, anti-apoptotic protein, and Bcl-2, and upregulates the expression of the pro-apoptotic protein BAX in hASCs in OM, and causes mitochondrial apoptosis