Primary BMSCs were harvested from 5 male Sprague Dawley rats (age, 4–5 weeks; weight, 80–100 g) using previously described methods (2). The male Sprague-Dawley rats were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). The rats were maintained in a temperature controlled room (22±3°C) with a 12-h light/dark cycle and were allowed free access to drinking water and food. All procedures were approved by The Animal Research Committee of Zhongshan Hospital, Fudan University (Shanghai, China). Briefly, rats were sacrificed by cervical dislocation followed by soaking in 75% ethanol for 10 min. The tibias and femurs from both legs of each rat were dissected and the two ends of each bone were cut under aseptic conditions. The marrow was flushed out with rat Mesenchymal Stem Cell Growth medium (Cyagen Biosciences, Inc., Santa Clara, CA, USA) containing 10% fetal bovine serum, 1% penicillin/streptomycin and 1% glutamine. The cells were centrifuged at 500 × g for 5 min at 4°C, resuspended with Mesenchymal Stem Cell Growth medium, plated in 25 cm² culture flasks and incubated at 37°C in 5% CO₂. The medium was replaced every 3 days and non-adherent cells were removed. The cells reached ~80% confluence at 10–14 days of culture, and were subsequently dissociated using TrypLE Express (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and subcultured to passage 3 for further analysis. The cells were analyzed using an Olympus SZX12 (Olympus Corporation, Tokyo, Japan) inverted microscope equipped with a digital camera and connected to a PC using MagnaFire 2.0 camera software (Optronics, Goleta, CA, USA).

BMSC cell-surface markers were analyzed by flow cytometry. Briefly, cells at passage 3 with 80% confluence were trypsinized using TrypLE Express (Gibco; Thermo Fisher Scientific, Inc.), washed twice with PBS and centrifuged at 500 × g for 5 min at 4°C. Cells were diluted with stain buffer (BD Bioscience, Franklin Lakes, NJ, USA) and the cell number was determined by cell counting assay following the manufacturer's instructions (Precision Instrument Co. Ltd., Shanghai, China). Following counting, a 100 µl cell suspension containing 5×10⁵ cells was added to flow tubes, mixed with rat antibodies, including 1 µl FITC-labeled anti-CD29 (1:100; cat. no. 102205; BioLegend, Inc., San Diego, CA, USA), anti-CD45 (1:100; cat. no. 202205; BioLegend, Inc.) and anti-CD90 (1:100; cat. no. 206105; BioLegend, Inc.) and 2.5 µl phycoerythrin (PE) -labeled anti-CD34 (1:40; cat. no. ab187284; Abcam, Cambridge, UK), and incubated in dark for 40 min at 4°C. Following two washes with PBS, the stained cells were resuspended in 300 µl PBS and immediately analyzed using a BD FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA). Identification of rat BMSCs was performed in triplicate. The mouse immunoglobulin G1, κ was used as an isotype control (BioLegend, Inc.) and stain buffer was used as blank control. Data were analyzed using FlowJo software version 10 (Tree Star, Inc., Ashland, OR, USA).

Cells at passage 3 with 80% confluence were trypsinized using TrypLE Express and seeded into six plates at 1×10⁴ cells/cm² and incubated with Mesenchymal Stem Cell Growth medium at 37°C for 24 h. Following incubation, the growth medium was replaced with 2 ml osteo-induction medium was prepared as previously described (12) containing 10 mM Na β-glycerophosphate (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), 0.25 mM ^l-ascorbic acid (MP Biomedicals, LLC, Santa Ana, CA, USA) and 0, 5, 50, 500 or 5,000 nM FK506 (Sigma-Aldrich; Merck KGaA). The osteo-induction medium was replaced every other day; cells were collected at 3, 7 and 14 days for RT-qPCR analysis, and phenotype staining was done at 7 or 14 days. All biochemical assays were carried out at least three times.

ALP staining was performed following 7 days osteogenic stimulation. Briefly, cells were rinsed twice with PBS and fixed with 4% formaldehyde for 30 min at room temperature. The fixed cells were stained with 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium solution (Beyotime Institute of Biotechnology, Shanghai, China) for 30 min at room temperature and washed 3 times with PBS. ARS staining was used to assess mineral deposition produced at late-stage bone formation. Briefly, following 14 days osteogenic stimulation, cells were washed twice with PBS, fixed with 4% formaldehyde for 30 min at room temperature, washed twice with PBS and then stained with the Alizarin Red S Solution (Cyagen Biosciences, Inc., Chicago, USA) for 30 min at room temperature. Following ARS staining, the cells were washed 3 times with PBS. Images were captured using an Olympus SZX12 inverted microscope equipped with a digital camera and connected to a PC using MagnaFire 2.0 camera software.

According to the manufacturer's protocol, total RNA was extracted from 1×10⁵ control and FK506-induced BMSCs with TRIzol Reagent (Invitrogen, Carlsbad, USA), purified using the RNeasy Mini Kit (Qiagen, CA, USA) and quantitated by using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). RNA (500 ng) was reverse transcribed into cDNA using the PrimeScript^™ RT-PCR kit (Takara Bio, Inc., Otsu, Japan). The cDNA was amplified using the SYBR Premix Ex Taq II kit (Takara Bio, Inc.) with the GeneAmp-PCR system 7500 (Thermo Fisher Scientific, Inc.). Total RNA for miRNA analysis was isolated using a miRCute miRNA Isolation kit (Tiangen Biotech Co., Ltd., Beijing, China) and the expression of mature miRNA was determined by miRCute miRNA PCR Detection kit (Tiangen Biotech Co., Ltd.) with the GeneAmp-PCR system 7500. All primers used for RT-qPCR are listed in the Table I, and each sample was measured in triplicate. Gene expression results of mRNA or miRNA were evaluated by the 2^−ΔΔCq method (13) and normalized to GAPDH or U6, respectively.

miRNA microarray analysis was performed by KangChen Bio-tech Inc. (Shanghai, China). Briefly, total RNA was extracted from 1×10⁷ control and FK506-induced BMSCs using TRIzol reagent and purified with RNeasy Mini kit (Qiagen, Inc., Valencia, CA, USA), according to manufacturer's protocol. The quality and quantity of RNA were measured by using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA), and RNA integrity was determined using 1.5% gel electrophoresis. miRNAs were labeled with Hy3/Hy5 fluorescence dyes using the miRCURY LNA microRNA Array Power Labeling kit (Exiqon A/S, Vedbaek, Denmark) and hybridized in a Hybridization Chamber II (Ambion; Thermo Fisher Scientific, Inc.) using the miRCURY LNA microarray kit (Exiqon A/S) according to the manufacturer's protocol. Following hybridization, the microarray slides were processed with an Axon GenePix 4000B Microarray Scanner (Axon Instruments; Molecular Devices, LLC, Sunnyvale, CA, USA), and data analysis was performed using GenePix Pro software version 6.0 (Molecular Devices, LLC). The experiment was performed with 3 samples both in control and treatment group, and miRNAs were defined as differentially expressed if changes in expression levels were ≥2-fold than those of the controls.

Target genes of the differentially expressed miRNAs were predicted using miRBase (http://www.mirbase.org), miRanda (http://www.microrna.org), TargetScan (http://www.targetscan.org) and miRDB (http://mirdb.org/miRDB) databases. GO analysis was performed using the Database for Annotation, Visualization and Integrated Discovery (https://david.ncifcrf.gov) to determine the functions of the predicted target genes and to uncover the miRNA-target gene regulatory network based on the predicted biological processes and molecular functions. KEGG pathway analysis (http://www.genome.jp/kegg) was used to determine the potential pathways that the predicted target genes may be a part of. Fisher's exact test and the χ² test were used to classify the GO category and KEGG pathway, and the false discovery rate (FDR) (14) was calculated to correct the P-value. Thresholds of P<0.05 and FDR<0.05 were used to select significant GO categories. Fisher P-value stands for the enrichment P-value used Fisher's exact test and Fisher-P value<0.05 was used as a threshold to select significant KEGG pathways.

Statistical analysis of two groups was determined by unpaired Student's t-test. All data were represented as the mean ± standard error of the mean. P<0.05 was considered to indicate a statistically significant difference.

Following 3 days of primary culture, the isolated BMSCs were mostly spindle shaped, single or small colony size and grew at a low density. (Fig. 1A). Primary BMSCs cultured for 12 days replicated rapidly and reached ~80% confluence (Fig. 1B). Cells were expanded by successive subculture and exhibited colony-like growth at passage 3. BMSCs phenotypes were verified by flow cytometric analysis. The subcultured BMSCs at passage 3 were positive for MSCs markers CD29 (99.9%) and CD90 (99.3%), and negative for hematopoietic lineage markers CD34 (0.998%) and CD45 (0.064%) (Fig. 1C).

To detect the effects of FK506 treatment on the osteogenic differentiation of BMSCs, cells were collected at passage 3 and cultured in osteo-induction medium with a concentration of FK506 ranging between 5 and 5,000 nM. ALP and ARS staining were strongest in cells treated with 50 nM FK506 (Fig. 2A). RT-qPCR analysis of BMSCs treated with 50 nM FK506 revealed that the expression of Sp7, Runx2, and OPN were significantly increased following 7 days stimulation and OCN expressed significantly at 7 and 14 days. (P<0.05 vs. negative control; Fig. 2B). These results indicated that FK506 treatment has the potential to stimulate the BMSCs differentiation into osteoblasts at a concentration of 50 nM for 7 days of stimulation, and thus this concentration was used for further miRNA array experiments.

To examine the differential expression of miRNAs that are involved in FK506-induced osteodifferentiation of rat BMSCs, miRCURY LNA microarrays were used to profile miRNA expression 7 days following FK506 (50 nM) induction. Comparison with the control group identified a total of 56 miRNAs, including 17 that were upregulated and 39 that were downregulated (Tables II and III, respectively). RT-qPCR analysis was used to verify the microarray results and demonstrated that five miRNAs were upregulated (miR-106b-5p, miR-101b-3p, miR-193a-3p, miR-485-3p and miR-142-3p) and four were downregulated (miR-27a-3p, miR-207, miR-218a-2-3p and let-7a-5p) compared with untreated control cells (Fig. 3).

To investigate the specific roles of the nine identified differentially expressed miRNAs in the regulation of osteodifferentiation in BMSCs, target genes were predicted and subjected to GO term and KEGG pathway analysis. GO analysis demonstrated that the highly enriched GO terms targeted by these dysregulated miRNAs were involved in multiple biological processes such as cell differentiation, cell growth and cell migration (Tables IV and V). KEGG pathway analysis revealed the target genes to be highly enriched in 22 upregulated and 21 downregulated signaling pathways, including mitogen-activated protein kinase (MAPK) signaling pathway, calcium signaling pathway and Toll-like receptor signaling pathway (Fig. 4).

Osteodifferentiation is driven by a series of complex events that result in the activation of specific transcription factors. Of the predicted target genes and differentially regulated function analysis, only those closely related to osteoblast differentiation, ossification and signal transduction are listed in Table VI. RT-qPCR demonstrated that MAPK9, Smad5 and jagged 1 (Jag1), which are positive regulators of osteogenesis, were significantly upregulated in FK506-induced rat BMSCs undergoing osteogenic differentiation, compared with the untreated control (Fig. 5); whereas dual-specificity phosphatase 2 (Dusp2), Smad7 and BMP and activin membrane-bound inhibitor (Bambi), which are negative regulators of osteodifferentiation, were demonstrated to be significantly downregulated. The results suggested that FK506 treatment may induce osteogenic differentiation by regulating the expression of specific target genes of these differentially expressed miRNAs.