Female, 7-week-old Sprague Dawley rats (n=20; 225–275 g) were purchased from the Experimental Animal Center of The Fourth Military Medical University (Xi'an, China). They were maintained under controlled conditions of 22–26°C, 12 h light/dark cycle and a relative humidity of 50–70%. Animals had access to food and water throughout the experiment. All studies were conducted in accordance with the standards of humane animal care described in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (20), using protocols approved by Zhengzhou University Institutional Animal Care and Research Advisory Committee (Zhengzhou, China). Low glucose Dulbecco's modified Eagle's medium (LG-DMEM) and trypsin were purchased from Gibco; Thermo Fisher Scientific, Inc. (Waltham, MA, USA), fetal bovine serum (FBS) was obtained from Hangzhou Sijiqing Biological Engineering Materials Co., Ltd. (Hangzhou, China). Penicillin and streptomycin sulfate were purchased from North China Pharmaceutical Group Co., Ltd. (Shijiazhuang, China), and Percoll separation medium was purchased from Pfizer, Inc. (New York, NY, USA). For flow cytometry, the mouse anti-rat monoclonal antibodies cluster of differentiation (CD)44-fluorescein isothiocyanate (FITC; cat. no. 561859), CD90-phycoerythrin (PE; cat. no. 551401), CD45-FITC (cat. no. 551401), and the isotype control antibodies mouse anti-rat IgG1-FITC (cat. no. 553892) and mouse IgG1-PE (cat. no. 554680) were obtained from BD Biosciences (Franklin Lakes, NJ, USA); the mouse anti-rat CD34-FITC monoclonal antibody (cat. no. sc-7324 FITC) was purchased from Santa Cruz Biotechnology, Inc., (Dallas, TX, USA). The Total Protein assay kit was purchased from Beyotime Institute of Biotechnology (Shanghai, China). The following antibodies were used for western blotting and immunofluorescence: Alexa Fluor^® 555 Phalloidin (cat. no. 8953; Cell Signaling Technology, Inc., Danvers, MA, USA); mouse anti-rat F-actin antibody (cat. no. ab205, Abcam, Cambridge, UK); mouse anti-rat GAPDH antibody (cat. no. SC-47724; Santa Cruz Biotechnology, Inc.); rabbit anti-rat tenascin-C antibody (cat. no. SC-20932; Santa Cruz Biotechnology, Inc.); goat anti-rabbit IgG-FITC (cat. no. SC-2012; Santa Cruz Biotechnology, Inc.); horseradish peroxidase-conjugated goat anti-mouse IgG secondary antibodies (cat. no. 1706516; Bio-Rad Laboratories, Inc., Hercules, CA, USA) and anti-rabbit IgG secondary antibodies (cat. no. 1706515; Bio-Rad Laboratories, Inc.). One-Step SYBR PrimeScript RT-PCR kit II was obtained from Takara Biotechnology Co., Ltd., (Dalian, China) for reverse transcription-quantitative polymerase chain reaction (RT-qPCR).

Isolation, culture and passage of rat BMSCs from 7-day-old Sprague Dawley rats were performed as previously described (21) following sacrifice of rats by cervical dislocation. Briefly, under sterile conditions, a syringe was used to flush out the bone marrow. Bone marrow cells were isolated with Percoll separation medium. The interface layer of mononuclear cells was washed twice with cold PBS and maintained in complete LG-DMEM (10% FBS; 100 U/ml penicillin and 100 mg/ml streptomycin sulfate). These primary cells were subcultured at 80–90% confluence. Fourth passage BMSCs were used for experimental analysis.

Flow cytometry was performed to determine the surface expression of CD44, CD90, CD45 and CD34 on rat BMSCs. On reaching 80–90% confluence, fourth passage BMSCs (1×10⁵ cells/ml) were stained with 1 µl CD44-FITC (1:1,000), CD90-PE (1:1,000), CD45-FITC (1:1,000), CD34-FITC (1:1,000), IgGl-FITC (1:1,000) or IgGl-PE (1:1,000) at room temperature in the dark for 30 min. The labeled cells were centrifuged at 208 × g for 5 min at 4°C and resuspended in PBS for analysis using a FACScan flow cytometer (BD Biosciences). CellQuest™ Pro Version 4.0 (BD Biosciences) was used for analysis. As described previously (4), the control cells were cultured in complete medium (LG-DMEM; 10% FBS; 100 U/ml penicillin and 100 mg/ml streptomycin sulfate) and the treatment group was cultured with osteogenic and adipogenic induction, respectively. After 12 and 14 days, cells were fixed with paraformaldehyde and stained with alizarin red or Oil-red-O solution for observation under microscope (22,23).

SB 431542 (a TGF-β receptor type I inhibitor), PD 98059 [a mitogen-activated protein kinase (MAPK) kinase (MEK-1) inhibitor] and SB 203580 (a p38 MAPK inhibitor) were purchased from EMD Millipore (Billerica, MA, USA). Stock solutions of SB 431542, PD 98059 and SB 203580 were prepared in dimethyl sulfoxide (DMSO). Neutralizing anti-TGF-β1 antibodies were obtained from R&D Systems, Inc. (Minneapolis, MN, USA). Human recombinant TGF-β1 was purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany) and prepared in bovine serum albumin (BSA; Sigma-Aldrich; Merck KGaA)-containing stock solutions according to the manufacturer's protocol. Growth factor and inhibitor concentrations used in assays were optimized based on the literature (24–26), our preliminary experiments and the manufacturer's data sheets, and the following final concentrations were selected: Neutralizing anti-TGF-β1 antibodies, 100 ng/ml; SB 431542, 10 µM; PD 98059, 10 µM; SB 203580, 3 µM; and TGF-β1, 5 ng/ml.

Rat pelvic ligament fibroblasts were prepared as previously described (21) and maintained in LG-DMEM containing 10% FBS, and 100 U/ml penicillin and 100 mg/ml streptomycin sulfate. Briefly, ligament was removed and immediately placed into serum-free Hanks' balanced salt solution (HBSS) and minced into 1 mm³, followed by treatment with 0.25% trypsin and 0.01% EDTA at 37°C for 2 days with gentle rotation. The supernatant, which contained the fibroblasts, was washed with HBSS and DMEM with 10% heat-inactivated FBS. The cell suspension was spun down twice to remove collagenase. The pellets were resuspended in growth medium and plated in tissue culture flasks. The medium was changed every other day. Prior to stretching, a small number of fourth passage fibroblasts were collected for immunofluorescence staining to confirm that they were fibroblasts. Passage 3 fibroblasts (3×10⁵/membrane) were seeded onto gelatin-coated silicone membranes and cultured for 24 h, following culture, cells were considered to be passage 4. Subsequently, 10% of the load transformation was exerted on cells with 1 Hz horizontal stretch stimulation for various durations (0, 3, 12 or 24 h) under 5% CO₂ and 37°C conditions. Cells were fixed in paraformaldehyde and stained with 5 µg/ml DAPI. Cell morphology was examined under a confocal microscope (Olympus Corporation, Tokyo, Japan). Prior to co-culture, ligament fibroblasts were subjected to mechanical stretch stimulation for 12 h and subsequently seeded into the upper chamber at 5×10⁴/well. BMSCs (2×10⁵/well) were seeded into the lower chamber of a 6-well Transwell plate. Following indirect co-culture for 3, 6 or 12 days, cells were collected for RT-qPCR and western blot analysis.

BMSCs (2×10⁵ cells/well in 2 ml) in the lower chamber and fibroblast cells (5×10⁴ cells/well in 2 ml) in the upper chamber were cultured in 6-well Transwell plates. Protein kinase inhibitors and growth factors were diluted with medium at double the final concentration stated previously: Neutralizing anti-TGF-β1 antibodies, 200 ng/ml; SB 431542, 20 µM; PD 98059, 20 µM; SB 203580, 6 µM; and TGF-β1, 10 ng/ml. The control group was treated with medium containing 0.1% DMSO. Control- or inhibitor-containing medium (2 ml) was placed into each well and incubated for 30 min. Subsequently, 2 ml medium with or without TGF-β1 was added to the upper and lower compartments of each well. Cells were collected at different time points following co-culture (3, 6 or 12 days), which was followed by lysis and total RNA extraction as described below.

TGF-β1 was detected using a commercially available ELISA kit (cat. no. MB100B, R&D Systems). Representative samples were analyzed for the presence of active TGF-β1; however, the concentration was below the detection limit of the assay. Therefore, all samples were activated by acidification to separate TGF-β1 from its binding proteins, allowing for measurement of total TGF-β1. For activation of samples containing FBS, 2.5 M acetic acid+10 M urea was used, and 1 M HCl was used for microdialysis samples (as recommended by the supplier). Activated samples were neutralized using 2.7 M NaOH + 1 M 4-(2-hydroxyethyl)-1-piperazinee-thanesulfonic acid (HEPES) for samples with FBS and 1.2 M NaOH+0.5 M HEPES for microdialysis (as recommended by the supplier). Samples were loaded onto ELISA plates immediately following neutralization. All samples were measured in duplicate.

RNA was isolated using TRIzol^® reagent (Takara Biotechnology Co., Ltd.). The RNA yield was determined by measuring the absorbance at a wavelength of 260 nm. RNA was subsequently reverse-transcribed to cDNA, and the cDNA was amplified by qPCR using the SYBR^®-Green reporter that was included in the kit used for qPCR and the following thermocycling conditions: Pre-denaturation for 2 min at 94°C; followed by 45 cycles of 30 sec at 60°C and 60 sec at 55°C. The specificity of the produced amplification product was confirmed by examination of dissociation reaction plots. Each sample was tested in triplicate, and samples obtained from 3 independent experiments were used for analysis of relative gene expression, normalization to GAPDH expression was performed using the ∆∆Cq method (27–29). The following primers were designed by Takara Biotechnology Co., Ltd.: GAPDH, 5′-GACATCAAGAAGGTGGTGAAGC-3′ (forward) and 5′-TGTCATTGAGAGCAATGCCAGC-3′ (reverse); and tenascin-C, 5′-ACCATGCTGAGATAGATGTTCCAAA-3′ (forward) and 5′-CTTGACAGCAGAAACACCAATCC-3′ (reverse).

Cells were lysed using 200 µl radioimmunoprecipitation assay lysis buffer (cat. no. R0278, Sigma-Aldrich; Merck KGaA). Equal quantities of proteins (20 µg) were loaded onto 10% Tris-Glycine gels following denaturation. The proteins were subsequently transferred to polyvinylidene difluoride membranes. Membranes were blocked with 5% milk in PBS containing 0.1% Tween-20 (PBST), followed by incubation with primary antibodies against F-actin (1:1,000), and GAPDH (1:1,000) at 4°C overnight. The membranes were washed with PBST and incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies (1:2,000). Following washing, protein bands were visualized using a Chemiluminescence Protein Detection kit (Pierce; Thermo Fisher Scientific, Inc.). Each experiment was repeated in triplicate. GAPDH served as the internal control.

Cells (fibroblasts and BMSCs) were cultured on coverslips in a 6-well plate. Following mechanical stress or co-culture, cells were fixed with 3% paraformaldehyde at 4°C in PBS containing 1 mM calcium chloride for 10 min and permeabilized by the use of 0.2% Triton X-100 at room temperature for 5 min. Subsequently, cells were incubated with blocking solution (containing 5% goat serum, 0.3% Triton X-100, and 0.1% BSA in PBS) at room temperature for 30 min. Following 3 rinses, slides were incubated with Alexa Fluor 555 Phalloidin (1:20) for 1 h at 37°C. Rinsed slides were incubated with anti-tenascin-C (1:500) primary antibody at 4°C overnight. After washing and the addition of goat anti-rabbit IgG-FITC (1:500) for 1 h at 37°C, nuclear staining was performed with DAPI (1:20,000; Sigma-Aldrich; Merck KGaA) for 5 min. Slices were subsequently mounted and observed under a fluorescent microscope.

Data are expressed as the mean ± standard deviation. Statistical significance was analyzed using one-way analysis of variance followed by Tukey's post hoc test, and was performed using SPSS software version 11.0 (SPSS, Inc., Chicago, IL, USA). P<0.05 was considered to indicate a statistically significant difference.

As described in our previous studies (21,30), scattered small BMSCs colonies formed, fused and stacked following several passages. Following 3–4 passages, cells exhibited more homogeneous morphology. These homogeneous cells were collected for expression profile analysis by flow cytometry. Cells expressed CD44 and CD90 (Fig. 1) and may be induced to differentiate into osteoblasts and adipogenic cells (data not shown).

Fibroblasts were subjected to 10% load and 1 Hz stretch optimized according to our previous studies (21,30) and preliminary experiments. Compared with the control group, this mechanical stretch elongated the cell body and increased shape index values at different time points (data not shown). Under a laser confocal scanning microscope, non-stretching rat ligament fibroblasts exhibited the typical morphology of polygonal cells with a thick F-actin fiber bundle (Fig. 2A). However, 24 h of mechanical stress decreased the F-actin staining intensity (Fig. 2B); the effect on F-actin protein expression levels was time-dependent, as demonstrated by western blotting (Fig. 2C).

Preliminary experiments (data not shown) revealed that mechanical stress significantly and directly increased tenascin-C expression in fibroblasts in a time-dependent manner. However, whether indirect co-culture with mechanically stretched ligament fibroblasts may regulate tenascin-C expression in BMSCs remains controversial. Therefore, the present study used an indirect co-culture system to investigate tenascin-C expression levels and pathway in BMSCs cultured with stretched fibroblasts. Following co-culture with stretched fibroblasts, tenascin-C mRNA expression levels in BMSCs increased ~1.5-fold (P<0.05) after 3 days, ~2.4-fold (P<0.05) after 6 days and 4- to 5-fold (P<0.05) after 12 days, compared with the control group (Fig. 3A). The data indicated that co-culture with mechanically stretched fibroblasts increased the tenascin-C mRNA expression levels in BMSCs in a time-dependent manner. In accordance with this, immunofluorescence staining revealed that tenascin-C expression was increased in BMSCs following co-culture with mechanically stimulated fibroblasts for 12 days, compared with control BMSCs (Fig. 3B).

It was subsequently investigated whether TGF-β treatment and co-culture with strained fibroblasts had synergistic effects. TGF-β is a paracrine growth factor mediating the cellular response to mechanical strain. This factor induces tenascin-C mRNA expression in fibroblasts (31,32). Therefore, it was hypothesized that mechanical stretching of fibroblasts may indirectly promote tenascin-C expression in BMSCs via the release of TGF-β1. To investigate this possibility, BMSCs were treated with exogenous TGF-β1 to determine whether this synergistically increased the effects of mechanical stretch on tenascin-C mRNA expression levels. With or without addition of TGF-β1, BMSCs were cultured with mechanically stretched fibroblasts for 6 days. As presented in Fig. 4, in the absence of TGF-β1, mRNA expression levels of tenascin-C in BMSCs co-cultured with mechanically stimulated fibroblasts were 2-fold greater (P<0.05) compared with control BMSCs co-cultured with fibroblasts without mechanical stimulation. The addition of 5 ng/ml TGF-β1 significantly increased tenascin-C mRNA expression levels ~3.8-fold (P<0.05) compared with BMSCs cultured in the absence of TGF-β1. Tenascin-C mRNA expression levels increased further following co-culture with mechanically stretched fibroblasts. These data suggested that the effects of TGF-β1 and indirect co-culture with strained fibroblasts on tenascin-C expression in BMSCs were synergistic.

TGF-β1 has been reported to be a potent growth factor with the capacity to induce tenascin-C expression in fibroblasts (33). If indirect co-culture with the stretched fibroblasts induces tenascin-C mRNA expression level in BMSCs via the paracrine release of TGF-β1, the effects of mechanical stimulation and TGF-β1 treatment may be mediated via an intracellular signaling pathway. To test this hypothesis, the effects of specific inhibitors on the increase of tenascin-C mRNA expression levels by TGF-β1 and co-culture with mechanically stretched fibroblasts were investigated. Treatment with 5 ng/ml TGF-β1 for 6 days led to a 2.4-fold increase in tenascin-C mRNA expression levels in BMSCs (Fig. 5). SB 431542 is a specific inhibitor of activin receptor-like kinase, the TGF-β receptor type I. As presented in Fig. 5, 10 µM SB 431542 inhibited the TGF-β1-dependent increase of tenascin-C mRNA expression levels by ~59%. A specific inhibitor of MEK-1, 10 µM PD 98059 significantly decreased tenascin-C mRNA expression levels in BMSCs by ~61%. A similar inhibition was observed following treatment with 3 µM SB 203580, a p38 MAPK inhibitor. These data indicate that activation of the MEK/p38 MAPK signaling pathway may contribute to the TGF-β1-dependent increase in tenascin-C expression.

TGF-β1-dependent induction of tenascin-C expression may require activation of MEK/p38 MAPK in BMSCs. In addition, mechanical stretching significantly induced TGF-β1 expression in stretched fibroblasts (Fig. 6A) and BMSCs co-cultured with stretched fibroblasts (Fig. 6B). Therefore, the involvement of the MEK/MAPK signaling pathway was investigated in the induction of tenascin-C expression in BMSCs by co-culture with mechanically stretched fibroblasts. As presented in Fig. 6C, the increase in tenascin-C mRNA expression levels induced by indirect co-culture with mechanically stretched fibroblasts was significantly inhibited by neutralizing anti-TGF-β1 antibodies and SB 431542. Similarly, PD 98059 and SB 203580 significantly attenuated the increase in tenascin-C expression levels in BMSCs. These results suggested that in the indirect co-culture system, the intracellular MEK/p38 MAPK signaling pathway is responsible for the upregulation of tenascin-C expression levels induced by mechanical stretch stimulation.