hBMSCs were purchased from the American Type Culture Collection (Manassas, VA, USA) and maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 12% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) at 37°C in a humidified atmosphere containing 5% CO₂. The present study was approved by the Ethics Committee of Xiamen Stomatological Hospital (Xiamen, China).

A BCA protein assay kit, enhanced chemiluminescent ultra-sensitive luminescence solution, mouse anti-GAPDH monoclonal antibodies (cat. no. AF0006) and horseradish peroxidase-labeled goat anti-rabbit immunoglobulin G (H+L) antibodies (cat. no. A0208) were purchased from Beyotime Institute of Biotechnology (Beijing, China). Peroxidase-conjugated goat anti-mouse IgG (H+L) antibodies (cat. no. ZB-2305) were purchase from OriGene Technologies, Inc. (Rockville, MD, USA). Rabbit anti-dentin sialophosphoprotein (DSPP; cat. no. ab216892), osteocalcin (OCN; cat. no. ab13420), Runx2 (cat. no. ab76956), RhoA (cat. no. ab54835) and ROCK (cat. no. ab45171) polyclonal antibodies, anti-cluster of differentiation (CD)29-phycoerythrin (PE) (cat. no. ab93759), CD90-PE (cat. no. ab93758), CD105-PE (cat. no. ab11414), CD146-PE (cat. no. ab75769), CD34-PE (cat. no. ab81289) and CD45-PE (cat. no. ab10558) antibodies were purchased from were purchased from Abcam (Cambridge, MA, USA). Hank's balanced salt solution (HBSS) and radioimmunoprecipitation assay (RIPA) lysis buffer were from Shanghai Solarbio Science & Technology Co., Ltd. (Shanghai, China). Collagenase I and neutral collagenase II were obtained from Gibco (Thermo Fisher Scientific, Inc.) and a TRIzol kit and one-step reverse transcription (RT)-polymerase chain reaction (PCR) kit were obtained from Invitrogen (HiFiScript cDNA synthesis kit; cat no. CW2569M; CWBIO, Beijing, China). C3 exoenzyme was purchased from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany). The flow cytometer FACScalibur was supplied by BD Biosciences (San Jose, CA, USA). Evolution^™ 220 UV-Visible spectrophotometer was from Thermo Fisher Scientific, Inc. and electrophoresis apparatus, electrophoresis transfer apparatus and the gel imaging system (ChemiDocTM XRS) were obtained from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). The fluorescence quantitative PCR instrument LightCycler96 was purchased from Roche Diagnostics (Basel, Switzerland) and the PCR thermal cycler PTC-200 was purchased from (Bio-Rad Laboratories, Inc.).

hDPSCs were isolated as described previously (22,23). Impacted third molars were collected from 29 healthy young (aged 19–29) male volunteers; the pulps were extracted under sterile conditions. The volunteers who possessed healthy teeth tissue, and no periodontal or periapical diseases were enrolled in the present study at Xiamen Stomatological Hospital (Xiamen, China) between July and December 2015. Patients who had previously had extraction were excluded. All volunteers provided their prior written, informed consent. The pulps were washed with HBSS, and digested with 3% collagenase I and 4% neutral collagenase II for 1 h at 37°C, and filtered through a 70 µm mesh to generate a single cell suspension. The suspension was centrifuged at 500 × g at 25°C for 10 min, resuspended in DMEM supplemented 15% FBS, inoculated into wells in 96-well plates and cultured at 37°C in an atmosphere containing 5% CO₂. The culture medium was refreshed every 3 days and subcultured when the confluency reached 80–90%.

hDPSCs at the third passage were digested with 0.25% EDTA-trypsin at 37°C for 2 min when the confluency reached 80–90% to obtain a single cell suspension. The cells were adjusted to a density of 1×10⁶ cells/ml and incubated with CD29-PE, CD90-PE, CD105-PE, CD146-PE, CD34-PE and CD45-PE antibodies (all 1:100) for 1 h at room temperature and analyzed by performing flow cytometry using a flow cytometer. The cells were blocked in 1% bovine serum albumin (BSA; CWBIO) for 30 min at 37°C and washed with PBS prior to the analysis. The data were analysed using BD CellQuest^™ software (version 5.1; BD Biosciences).

hDPSCs and hBMSCs at the third passage were inoculated to the wells of 6-well plates, combined with 100 µl mineralization induction solution (10 nmol/l dexamethasone, 10 mmol/l β-glycerophosphate and 50 µg/ml vitamin C) and α-minimum essential medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 100 ml/l FBS, and cultured at 37°C. Cells were harvested at day 0, 7, 14 and 21 for analysis of DSPP, OCN and Runx2 levels using western blot analysis.

hDPSCs (1×10⁶) at the third passage were inoculated into the wells of 6-well plates and C3 exoenzyme at a final concentration of 50 µg/ml was added. Following 48 h of incubation at 37°C, cells were harvested and assessed for the expression of RhoA, ROCK, Runx2, alkaline phosphatase (ALP), collagen type I (Col I), Msh homeobox 2 (MSX2), distal-less homeobox 2 (DLX2), DSPP, dentin matrix protein-1 (DMP-1), bone sialoprotein (BSP) and OCN at the protein or mRNA level. Cells cultured in DMEM and not treated with C3 exoenzyme were used as control.

Total RNA was extracted hDPSCs and hBMSCs using a TRIzol kit, according to the manufacturer's protocols. The purity and quantity of RNA were determined using a UV spectrophotometer. The RT reaction was performed with 200 ng RNA in a total volume of 10 µl containing 2 µl (2 µg) RNA, 2 µl dNTP (500 µg/ml) and 2 µl MgCl₂ (50 mM), and the resulting complementary DNA was quantified using a one-step PCR kit, according to the manufacturer's protocols, and the primers listed in Table I in a total volume of 25 µl. Human GADPH (Hs03929097_g1), was used as an internal control. The cycling conditions performed were the following: Pre-denaturation at 95°C for 5 min, followed by 40 cycles of 30 sec at 95°C, 30 sec at 60°C and 30 sec at 72°C. qPCR was performed using 30 cycles of denaturation at 95°C for 30 sec, annealing at 51°C for 20 sec and extension at 75°C for 30 sec. Experiments were performed in triplicate and the mean value was calculated for each case. The data were assessed using RQ Manager software (version 1.2.1.; Applied Biosystems; Thermo Fisher Scientific, Inc.). Relative expression was calculated by using the 2^−ΔΔCq method and obtaining the fold change value, according to previously described protocol (17).

Following different treatments, hDPSCs and hBMSCs were lysed with RIPA lysis buffer that contained protease and phosphatase inhibitors. The supernatants were collected following centrifugation at 7,000 × g at 4°C for 20 min. The protein was quantified using the BCA method, 20 µg protein was loaded per lane and separated using 12% SDS-PAGE, transferred to a PVDF membrane. The membranes were blocked with 2% BSA (Beyotime Institute of Biotechnology) at room temperature for 1 h, then the target proteins were detected by incubating the membrane with primary antibodies against DSPP, osteocalcin, Runx2, RhoA, ROCK and GAPDH (all 1:100) at 4°C overnight, and horseradish peroxidase-labeled goat anti-rabbit secondary and peroxidase-conjugated goat anti-mouse antibodies (both 1:2,000) at 37°C for 1 h. Visualization was achieved with a chemiluminescence kit. The intensity of blot signals was quantitated using Quantity one analysis software (version 4.62; Bio-Rad Laboratories, Inc.).

All data are expressed as the mean ± standard error of the mean obtained from at least three independent experiments. Statistical comparisons between groups were assessed by using the Student's t-test. P<0.05 was considered to indicate a statistically significant difference. Data were analyzed using SPSS software (version 10.1; SPSS, Inc., Chicago, IL, USA).

As indicated in Fig. 1, the primary hDPSCs were spindle-shaped and fibrous-like after 3 days of culture in vitro. The culture reached 80% confluence in 7–8 days. As indicated in Fig. 2, flow cytometry analysis revealed that >95% of the cells were positive for MSC markers CD29, CD90, CD105, and for endothelial cell marker CD146, whereas <5% were positive for hematopoietic stem cell markers CD34 and CD45. These results suggested that the purity of the hDPSCs was high enough for the subsequent experiments.

The differentiation of hDPSCs into odontoblast was investigated by determining the protein expression levels of DSPP. Results indicated that, following treatment with mineralization induction solution, the protein expression levels of DSPP in hDPSCs were increased in a time-dependent manner, and at 7, 14 and 21 days, expression levels were significantly increased compared with 0 days (P<0.01; Fig. 3). These findings suggested that odontoblast differentiation may have been induced in the hDPSCs.

The protein expression levels of OCN in hBMSCs cultured in mineralization solution were also examined in the present study. The data demonstrated that the protein expression levels of OCN were significantly upregulated at 14 and 21 days after induction compared with 0 days (P<0.01; Fig. 4), which indicated that mineralization may have induced hBMSCs to differentiate into osteoblasts.

In hDPSCs, increased incubation time in mineralization medium resulted in increased protein expression levels of Runx2 until day 5, after which the expression levels declined. Furthermore, significantly increased expression levels of Runx2 protein were indicated at 3, 5, 7 and 14 days in culture compared with 0 days in culture (P<0.01; Fig. 5A). Conversely, the expression of Runx2 protein in hBMSCs was increased in a time-dependent manner, with a significant increase in expression indicated at 5 (P<0.05), 7 (P<0.01), 14 (P<0.01) and 21 (P<0.01) days compared with 0 days (Fig. 5B).

To examine whether the Rho/ROCK signaling pathway is involved in regulating Runx2 expression, hDPSCs were treated with the pathway inhibitor, C3 exoenzyme. Results indicated that compared with the control, hDPSCs treated with the inhibitor exhibited significantly decreased protein expression levels of RhoA, ROCK and Runx2 (P<0.01; Fig. 6A). Furthermore, following mineralization induction, Runx2 protein expression levels in the inhibitor-treated hDPSCs were significantly decreased at 7 and 14 days compared with that in the control (P<0.01; Fig. 6B).

The present study evaluated the impact of the Rho/ROCK signaling pathway on the mRNA expression levels of early and late mineralization genes. Compared with the control, the mRNA expression levels of early mineralization genes ALP, Col I, MSX2 and DLX2 were not significantly different in inhibitor-treated hDPSCs (Fig. 7A). However, the mRNA expression levels of late mineralization genes, DSPP and DMP-1, were significantly upregulated in inhibitor-treated hDPSCs, whereas BSP and OCN were significantly downregulated compared with the controls (P<0.01; Fig. 7B).