Biological Mechanisms of Paeonoside in the Differentiation of Pre-Osteoblasts and the Formation of Mineralized Nodules

Paeonia suffruticosa is a magnificent and long-lived woody plant that has traditionally been used to treat various diseases including inflammatory, neurological, cancer, and cardiovascular diseases. In the present study, we demonstrated the biological mechanisms of paeonoside (PASI) isolated from the dried roots of P. suffruticosa in pre-osteoblasts. Herein, we found that PASI has no cytotoxic effects on pre-osteoblasts. Migration assay showed that PASI promoted wound healing and transmigration in osteoblast differentiation. PASI increased early osteoblast differentiation and mineralized nodule formation. In addition, PASI enhanced the expression of Wnt3a and bone morphogenetic protein 2 (BMP2) and activated their downstream molecules, Smad1/5/8 and β-catenin, leading to increases in runt-related transcription factor 2 (RUNX2) expression during osteoblast differentiation. Furthermore, PASI-mediated osteoblast differentiation was attenuated by inhibiting the BMP2 and Wnt3a pathways, which was accompanied by reduction in the expression of RUNX2 in the nucleus. Taken together, our findings provide evidence that PASI enhances osteoblast differentiation and mineralized nodules by regulating RUNX2 expression through the BMP2 and Wnt3a pathways, suggesting a potential role for PASI targeting osteoblasts to treat bone diseases including osteoporosis and periodontitis.


Introduction
Bone tissue is dynamically and continuously renewed through various cellular events in bone cells including osteoblasts, osteocytes, and osteoclasts to maintain a healthy skeleton [1,2]. Specifically, osteoblasts play an important role in bone development, formation, and regeneration via the synthesis and release of bone proteins and the mineralization of organic bone matrix [2,3]. Malfunction of the physiological and dynamic process pathologically induces bone diseases including osteoporosis and periodontal disease [4].
Bone morphogenetic proteins (BMPs) and Wnts stimulate various intracellular signaling and play crucial roles in multiple stages such as mesoderm patterning, osteoblast differentiation, bone formation, and craniofacial and limb development [5][6][7]. BMPs are multi-functional growth factors belonging to the transforming growth factor-beta superfamily. Among them, BMP2 is the first BMP to have been well-studied in the BMP2-Smad1/5/8 pathway and approved by the Food and Drug Administration for its use in bone diseases [8]. Wnt3a/β-catenin signaling is a well-studied pathway to regulate the proteasomal degradation of β-catenin [9]. The BMP2 and Wnt3a signaling pathways induce osteoblast differentiation by regulating runt-related transcription factor 2 (RUNX2) that is an important transcription factor [8,10,11].
Paeonia suffruticosa has been used as a medicinal plant for thousands of year and has beneficial effects in the treatment of various diseases including cancer, cardiovascular, inflammatory, and neurological diseases [12]. Paeonoside (PASI) is a bioactive compound identified in P. suffruticosa [13,14]. PASI possesses anti-diabetic activities by glycogen synthesis and glucose uptake through AMPK activation in HepG2 cells and contributes to glucose homeostasis [15]. It was also reported as a protective compound against sepsisinduced lethality [16]. Our group recently reported that another bioactive compound, paeonolide, identified from P. suffruticosa regulates core-binding factor subunit alpha 1 in bone-forming cells [17]. However, the biological effects and mechanisms of PASI have not yet been demonstrated in osteoblasts.
In the present study, we isolated PASI from the roots of P. suffruticosa and investigated the biological activities and underlying mechanism of PASI in osteoblast differentiation and mineralized nodule formation in pre-osteoblast MC3T3-E1 cells.

PASI Does Not Affect Cell Toxicity in Pre-Osteoblasts
The 1 H and 13 C nuclear magnetic resonance (NMR) spectra, high-performance liquid chromatography (HPLC) chromatogram, and chemical structure of 99.9% purity PASI from the dried roots of P. suffruticosa are shown in Figure 1A-C. At concentrations ranging from 0.1 to 100 µM, PASI was applied to pre-osteoblast MC3T3-E1 cells for 24 h to determine cytotoxic potential. No cytotoxic effects of PASI were measured by using an 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay ( Figure 1D). The doses of PASI below 100 µM were chosen in all subsequent experiments.
Paeonia suffruticosa has been used as a medicinal plant for thousands of year and has beneficial effects in the treatment of various diseases including cancer, cardiovascular, inflammatory, and neurological diseases [12]. Paeonoside (PASI) is a bioactive compound identified in P. suffruticosa [13,14]. PASI possesses anti-diabetic activities by glycogen synthesis and glucose uptake through AMPK activation in HepG2 cells and contributes to glucose homeostasis [15]. It was also reported as a protective compound against sepsis-induced lethality [16]. Our group recently reported that another bioactive compound, paeonolide, identified from P. suffruticosa regulates core-binding factor subunit alpha 1 in bone-forming cells [17]. However, the biological effects and mechanisms of PASI have not yet been demonstrated in osteoblasts.
In the present study, we isolated PASI from the roots of P. suffruticosa and investigated the biological activities and underlying mechanism of PASI in osteoblast differentiation and mineralized nodule formation in pre-osteoblast MC3T3-E1 cells.

PASI Does Not Affect Cell Toxicity in Pre-Osteoblasts
The 1 H and 13 C nuclear magnetic resonance (NMR) spectra, high-performance liquid chromatography (HPLC) chromatogram, and chemical structure of 99.9% purity PASI from the dried roots of P. suffruticosa are shown in Figure 1A-C. At concentrations ranging from 0.1 to 100 μM, PASI was applied to pre-osteoblast MC3T3-E1 cells for 24 h to determine cytotoxic potential. No cytotoxic effects of PASI were measured by using an 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) assay ( Figure  1D). The doses of PASI below 100 μM were chosen in all subsequent experiments.

PASI Enhances Cell Migration during Differentiation of Pre-Osteoblasts
To investigate whether PASI affects cell migration, we induced the differentiation of pre-osteoblasts using an osteogenic supplement (OS) medium in the presence or absence of PASI for 24 h, after which cell migration was observed using wound healing assay. The results showed that 1-30 µM PASI significantly recovered the wound area, compared to OS medium alone in a dose-dependent manner (Figure 2A,B). Next, we used a Boyden chamber assay to validate the effect of PASI. The assay showed that 1-30 µM PASI significantly promoted transmigration in a dose-dependent manner ( Figure 2C,D).
HPLC chromatogram (B), and chemical structure (C). (D) Cells were treated with 0.1-100 μM PASI for 24 h. Cell viability (%) was measured by using an MTT assay. Data are expressed as the mean ± S.E.M. of experiments.

PASI Enhances Cell Migration during Differentiation of Pre-Osteoblasts
To investigate whether PASI affects cell migration, we induced the differentiation of pre-osteoblasts using an osteogenic supplement (OS) medium in the presence or absence of PASI for 24 h, after which cell migration was observed using wound healing assay. The results showed that 1-30 μM PASI significantly recovered the wound area, compared to OS medium alone in a dose-dependent manner (Figure 2A,B). Next, we used a Boyden chamber assay to validate the effect of PASI. The assay showed that 1-30 μM PASI significantly promoted transmigration in a dose-dependent manner ( Figure  2C,D).

PASI Promotes ALP Staining and Activity during Differentiation of Pre-Osteoblasts
Next, we stimulated osteoblast differentiation in OS with 1-30 µM PASI for seven days to examine the biological activities of PASI in early osteoblast differentiation. Osteoblast differentiation was measured by using alkaline phosphatase (ALP) staining as an early osteoblast differentiation marker. The results showed that 1-30 µM PASI promoted ALP staining in a dose-dependent manner ( Figure 3A, upper). Under the same conditions, we subsequently performed an ALP enzymatic activity assay, and the activities were detected using a spectrophotometer. The results showed that 1-30 µM PASI significantly promoted the ALP activity in a dose-dependent manner ( Figure 3A, bottom). The increased ALPpositive cells in response to 1-30 µM PASI treatment were detected using a light microscope ( Figure 3B).

PASI Promotes ALP Staining and Activity during Differentiation of Pre-Osteoblasts
Next, we stimulated osteoblast differentiation in OS with 1-30 μM PASI for seven days to examine the biological activities of PASI in early osteoblast differentiation. Osteoblast differentiation was measured by using alkaline phosphatase (ALP) staining as an early osteoblast differentiation marker. The results showed that 1-30 μM PASI promoted ALP staining in a dose-dependent manner ( Figure 3A, upper). Under the same conditions, we subsequently performed an ALP enzymatic activity assay, and the activities were detected using a spectrophotometer. The results showed that 1-30 μM PASI significantly promoted the ALP activity in a dose-dependent manner ( Figure 3A, bottom). The increased ALP-positive cells in response to 1-30 μM PASI treatment were detected using a light microscope ( Figure 3B).

PASI Promotes ARS Staining during Differentiation of Pre-Osteoblasts
PASI-mediated osteoblast differentiation was further assessed using Alizarin red S (ARS) staining to investigate the biological activities of PASI in late osteoblast differentiation. Osteoblast differentiation was induced in OS with 1-30 μM PASI for 14 days, after which the formation of mineralized nodules was detected using a scanner. The results showed that 1-30 μM PASI potentiated mineralized nodule formation in a dose-dependent manner ( Figure 4A, upper). PASI-mediated late osteoblast differentiation was confirmed by the quantification of ARS stains ( Figure 4A, bottom). The mineralized nodules were observed and validated under a light microscope ( Figure 4B).

PASI Promotes ARS Staining during Differentiation of Pre-Osteoblasts
PASI-mediated osteoblast differentiation was further assessed using Alizarin red S (ARS) staining to investigate the biological activities of PASI in late osteoblast differentiation. Osteoblast differentiation was induced in OS with 1-30 µM PASI for 14 days, after which the formation of mineralized nodules was detected using a scanner. The results showed that 1-30 µM PASI potentiated mineralized nodule formation in a dose-dependent manner ( Figure 4A, upper). PASI-mediated late osteoblast differentiation was confirmed by the quantification of ARS stains ( Figure 4A, bottom). The mineralized nodules were observed and validated under a light microscope ( Figure 4B).

PASI-Mediated BMP2 and Wnt3a/β-Catenin Signaling Enhances RUNX2 Expression during Differentiation of Pre-Osteoblasts
To clarify the effect on osteoblast differentiation of PASI-mediated BMP2 and Wnt3a/βcatenin signaling, 10 µM PASI was treated with Noggin and PKF118-310 (PKF), which are inhibitors against BMP2 and Wnt3a/β-catenin signaling, respectively. The pretreatment of Noggin and PKF significantly attenuated the increased RUNX2 expression in response to 10 µM PASI during osteoblast differentiation ( Figure 6A,B). Immunofluorescence observation showed that Noggin and PKF blocked the RUNX2 expression increased by 10 µM PASI in the nucleus during osteoblast differentiation ( Figure 6C).

PASI-Stimulated BMP2 and Wnt3a/β-Catenin Signaling Promotes Osteoblast Differentiation
We further investigated whether PASI potentiated early and late osteoblast differentiation by stimulating BMP2 and Wnt3a/β-catenin signaling. The pretreatment of Noggin and PKF significantly attenuated PASI-mediated ALP staining and activity during early osteoblast differentiation ( Figure 7A,B). PASI stimulated the mineralized nodule formation during late osteoblast differentiation, whereas the pretreatment of Noggin and PKF significantly abolished the PASI-mediated effects ( Figure 7C,D).

PASI-Stimulated BMP2 and Wnt3a/β-Catenin Signaling Promotes Osteoblast Differentiation
We further investigated whether PASI potentiated early and late osteoblast differentiation by stimulating BMP2 and Wnt3a/β-catenin signaling. The pretreatment of Noggin and PKF significantly attenuated PASI-mediated ALP staining and activity during early osteoblast differentiation ( Figure 7A,B). PASI stimulated the mineralized nodule formation during late osteoblast differentiation, whereas the pretreatment of Noggin and PKF significantly abolished the PASI-mediated effects ( Figure 7C,D)
The migration into specific niches from the bone marrow, circulating blood, and periosteum is required for bone regeneration and repair [32][33][34]. Organic bone matrix and mineralization are achieved during differentiation of pre-osteoblasts [6,35,36]. In the present study, our data demonstrated that PASI promotes migration, ALP activity and its expression, and mineralized nodules. It was reported that ALP expression is increased in immature osteoblasts, and calcium deposition and mineralization are induced by mature osteoblasts [37][38][39]. Therefore, these data suggest that PASI promotes bone regeneration and repair through the migration, differentiation, and maturation of osteoblasts.
BMP2 and Wnt3a regulate osteoblast differentiation and bone formation [40][41][42]. BMP2 activates Smad1/5/8 proteins by binding to BMP receptors, induces the translocation of Smad1/5/8 proteins from the cytosol to the nucleus, and then Smad1/5/8 and Smad4 complexes regulate gene transcription in the nucleus [43]. Wnt3a interacts with Frizzled and LRP5/6 receptors, inactivates the GSK3β protein, stabilizes cytoplasmic β-catenin, and induces its nuclear translocation, leading to the regulation of gene expression in the nucleus [44]. Consequently, the signaling pathways promote the transcription of the RUNX2 gene for osteoblast differentiation and maturation [45][46][47][48]. In the present study, PASI enhanced BMP2 expression and Smad1/5/8 phosphorylation, as well as promoted Wnt3a expression and β-catenin stabilization, leading to RUNX2 expression. In addition, we demonstrated the nuclear expression of RUNX2 through PASI-stimulated BMP2 and Wnt3a signaling using Noggin and PKF. Noggin interacts with and antagonizes the action of BMP2 [49]. PKF disrupts β-catenin-TCF and inhibits β-catenin-regulated transcription activity [50,51]. It was reported that the BMP2 and Wnt3a/β-catenin signaling pathways are convergent to RUNX2 expression during osteoblast differentiation [52][53][54]. Previous studies also showed that the genetic modification of RUNX2 caused impaired bone formation and gene expression [55]. These results suggest that PASI-mediated RUNX2 upregulation leads to osteoblast differentiation and bone formation through the functional cross-talk between BMP2 and Wnt3a/β-catenin signaling.
Diet habits affect bone health and are important for reducing the risk of osteoporosis [56]. A dietary intake of calcium, vitamin D, and proteins also helps improve bone strength through bone mineralization such as hydroxyapatite crystals [57]. A soy-based diet that contain isoflavones such as soy milk and Tofu is thought to contain estrogen-like compounds that may help inhibit bone loss and prevent osteoporosis because estrogen hormones help protect bone loss against postmenopausal osteoporosis [58]. It was reported that the combination of soy isoflavones and milk basic protein preserves bone mineral density, and the combination of resveratrol which is found in red wine, red grape skins, and peanuts also prevents bone loss [59,60]. In the present study, we demonstrated that PASI promotes osteoblast differentiation and mineralization. Thus, the combination of PASI and adequate dietary intake could be a strategic approach for postmenopausal osteoporosis.
Osteoblasts secrete the receptor activator of nuclear factor kappa B ligand (RANKL), which is a key osteoclastogenic cytokine. In addition, osteoblasts mainly synthesize oteoprotegerin (OPG) as a circulating decoy receptor of RANKL [61]. The soluble RANKL/OPG system controls the differentiation and function of osteoclasts, thereby clinically determining bone status [62]. Growing evidence suggests that epigenetic modification plays a critical role in the formation of normal bone and the pathogenesis of bone diseases [63]. Epigenetic modification affects gene expression without alterations in genomic DNA [64].
Thus, two concerns of the RANKL/OPG system and epigenetic modification need to be explored to provide additional information on PASI-mediated osteoblast differentiation in future studies.
Currently, many pretreatment methods such as chemical, physical-chemical, and biological methods are being explored [65,66]. It was reported that hydrogen peroxide presoaking prior to ammonia fiber expansion pretreatment had a high advantage in agriculture [66]. Thus, any potential approaches to pretreat and extract PASI from the dried roots of P. suffruticosa including presoaking, liquid ammonia pretreatment, and others should be also considered.
In conclusion, we provide novel evidence that PASI from P. suffruticosa has biological effects as a regulator of RUNX2 by facilitating BMP2 and Wnt3a/β-catenin signaling ( Figure 8). Traditional natural compounds have received interest in bone diseases since the compounds have been used in various diseases for centuries with safety and efficiency [18,25,26,67,68]. Thus, our findings suggest that PASI as an anabolic compound may be developed and used to prevent and alleviate bone diseases by increasing osteoblast differentiation and bone formation.
Osteoblasts secrete the receptor activator of nuclear factor kappa B ligand (RANKL), which is a key osteoclastogenic cytokine. In addition, osteoblasts mainly synthesize oteoprotegerin (OPG) as a circulating decoy receptor of RANKL [61]. The soluble RANKL/OPG system controls the differentiation and function of osteoclasts, thereby clinically determining bone status [62]. Growing evidence suggests that epigenetic modification plays a critical role in the formation of normal bone and the pathogenesis of bone diseases [63]. Epigenetic modification affects gene expression without alterations in genomic DNA [64]. Thus, two concerns of the RANKL/OPG system and epigenetic modification need to be explored to provide additional information on PASI-mediated osteoblast differentiation in future studies.
Currently, many pretreatment methods such as chemical, physical-chemical, and biological methods are being explored [65,66]. It was reported that hydrogen peroxide presoaking prior to ammonia fiber expansion pretreatment had a high advantage in agriculture [66]. Thus, any potential approaches to pretreat and extract PASI from the dried roots of P. suffruticosa including presoaking, liquid ammonia pretreatment, and others should be also considered.
In conclusion, we provide novel evidence that PASI from P. suffruticosa has biological effects as a regulator of RUNX2 by facilitating BMP2 and Wnt3a/β-catenin signaling ( Figure 8). Traditional natural compounds have received interest in bone diseases since the compounds have been used in various diseases for centuries with safety and efficiency [18,25,26,67,68]. Thus, our findings suggest that PASI as an anabolic compound may be developed and used to prevent and alleviate bone diseases by increasing osteoblast differentiation and bone formation.

Differentiation of Pre-Osteoblast MC3T3-E1
Differentiation was induced as previously described [17]. Briefly, the cells were incubated in OS containing 50 µg/mL L-AA and 10 mM β-GP, and fresh medium was replaced every two days.

MTT Assay
Cell viability was measured using an MTT assay to detect NADH-dependent dehydrogenase activity as previously described [17].

Migration Assays
Wound-healing assay and Boyden chamber assay were carried out as previously described [17].

ALP Activity and Staining Assay
Osteoblast differentiation was induced using OS containing with PASI for seven days. ALP activity and staining assay were performed as previously described [17].

ARS Staining and Quantification
Osteoblast differentiation was induced using OS containing PASI for 14 days, and ARS staining and quantification were carried out as previously described [17].

Western Blot Analysis
Western blot analysis was carried out as previously described [69].

Immunofluorescence Analysis
Immunofluorescence was carried out as previously described [17].

Statistical Analysis
Data were analyzed using the Prism Version 5 program (GraphPad Software, Inc., San Diego, CA, USA). All numeric values are expressed as the mean ± S.E.M. Differences were considered statistically significant when * p < 0.05.