Alizarin Red S

EphrinB2 signaling enhances osteogenic/odontogenic differentiation of human dental pulp stem cells

A B S T R A C T
Objective: To investigate the role of the EphrinB2 signaling pathway in the osteogenesis/odontogenesis of human dental pulp stem cells (DPSCs).Design: The endogenous expression levels of EphrinB2 and its cognate receptors EphB2 and EphB4 in DPSCs were analyzed by qRT-PCR and Western blotting after 7, 14 and 21 days of osteogenic/odontogenic induction culture. Additionally, the phosphorylation of EphrinB2, EphB4 and ERK1/2 proteins at early time-points fol- lowing osteogenic induction, were also investigated by Western blots. Subsequently, we investigated whether supplementation of recombinant EphrinB2-Fc within the induction milieu can enhance the osteogenic/odon- togenic differentiation of DPSCs.Results: Endogenous gene and protein expression levels of EphrinB2, EphB2 and EphB4 were upregulated in induced versus non-induced DPSCs, over 21 days of osteogenic/odontogenic induction. Western blots showed increase in phosphorylated EphrinB2, EphB4 and ERK1/2 proteins at early time-points following osteogenic induction. Preliminary investigation of a concentration range (0, 0.5, 1 and 2 μg/ml) of recombinant EphrinB2-Fc within osteogenic induction media, showed that 0.5 μg/ml was optimal for enhancing the osteogenic/
odontogenic differentiation of DPSCs over a culture duration of 14 days. Subsequently, more comprehensive qRT-PCR analysis with 0.5 μg/ml EphrinB2-Fc revealed significant upregulation of several key osteogenic marker genes in treated versus untreated DPSCs after 21 days of osteogenic/odontogenic induction. By 7 days of osteogenic induction, DPSCs treated with 0.5 μg/ml EphrinB2-Fc exhibited significantly more calcium miner- alization (Alizarin red S staining) and alkaline phosphatase activity than the untreated control.Conclusions: EphrinB2 signaling plays a key role in the osteogenic/odontogenic differentiation of DPSCs.

1.Introduction
The endogenous adult stem cell niche within human dental pulp was first identified and characterized by the pioneering studies of Gronthos et al. (2002), Gronthos, Mankani, Brahim, Robey, and Shim (2009). These cells, universally referred to as dental pulp stem cells (DPSCs), have been shown to possess high proliferative potential, self- renewal capacity, and multi-lineage differentiation potential, both in vitro and in vivo (Gronthos et al., 2002, 2009; Huang, Gronthos, & Shi, 2009; Sharpe, 2016). DPSCs have been investigated in various tissue engineering and regenerative medicine applications, which is not re- stricted to just the field of dentistry (Nuti, Corallo, Chan, Ferrari, & Gerami-Naini, 2016; Tatullo, Marrelli, Shakesheff, & White, 2015).DPSCs play a key role in tooth homeostasis, by giving rise toodontoblasts that synthesize reparative dentin in response to pathological and inflammatory stimuli at tooth lesions. Although dental caries or physical trauma may result in damage or destruction of odontoblasts, reparative dentin can still be formed to protect the un- derlying pulp tissue. This regenerative process is mediated by the newly-differentiated odontoblasts that arise from DPSCs (Mitsiadis, Feki, Papaccio, & Catón, 2011). However, there are often insufficient numbers of endogenous DPSCs to repair major tooth lesions such as deep caries, whereby the dental pulp has been heavily damaged (Cao et al., 2015).Hence, a major challenge in clinical dentistry is to accelerate the process of DPSCs differentiation into functional odontoblasts with bioactive materials, which could in turn improve the amount, quality and functionality of the newly-formed reparative dentin.

The first step to achieving this objective, would be to gain a better understanding of the molecular signaling mechanisms involved in this process. Previous studies by our group and other researchers, have demon- strated that the signaling pathways mediated by EphrinB2 and its cognate receptor EphB2 and EphB4 play key roles in the osteogenic differentiation of osteoblasts (Arthur et al., 2013; Martin et al., 2010; Tonna et al., 2014; Zhao et al., 2006), mesenchymal stem cells (Pennisi, Ling, Li, Khan, & Barlogie, 2009; Toda, Yamamoto, Uyama, & Tabata, 2017; Zhang et al., 2015; Zhang et al., 2016) and periodontal ligament stem cells (PDLSCs) (Zhu et al., 2017). EphrinB2, EphB2 and EphB4 are transmembrane receptor tyrosine kinases (RTKs) that play key roles in cell fate decisions and cell migration/motility (Barquilla & Pasquale, 2015). A distinctive feature of the receptor-ligand interactions of these molecules is the simultaneous generation of bidirectional signals: for- ward signaling through the EphB2 and EphB4 receptor and reverse signaling through the EphrinB2 ligand (Nikolov, Xu, & Himanen, 2013). Because both the ligand (EphrinB2) and receptors (EphB2 and EphB4) are membrane-bound proteins, the forward and reverse signaling pathways mediated by these surface molecules need to be initiated by direct physical contact between cells.A number of previous studies have already shown that signalingpathways mediated by the closely-related EphrinB1 and EphrinB3 molecules play key roles in the mobilization of dental pulp stem cells for tooth repair, and the subsequent differentiation of these cells into functional odontoblasts (Arthur, Koblar, Shi, & Gronthos, 2009; Stokowski et al., 2007; Wang, Jong, Lin, & Shimizu, 2013). Never- theless, the role of EphrinB2-EphB4 signaling in the osteogenic/odon- togenic differentiation of dental pulp stem cells have not yet been in- vestigated, which would be the focus of this study.

2.Materials and methods
Unless otherwise stated, all culture media and supplements were purchased from Life Technologies (Carlsbad, CA, USA). All plastic labware were purchased from Becton-Dickinson (Franklin Lakes, NJ, USA), while all chemical reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA).Cryopreserved DPSCs utilized in this study, were derived from the human extracted third molars (18 and 25 years old), as described in our solution, 10−8 M dexamethasone, 10 mM β-glycerophosphate, and 50 μg/ml ascorbic acid for a duration of up to 21 days. Fresh induction medium was replaced every 3–4 days. For some experiments, re- combinant EphrinB2-EphrinB2-Fc protein (Cat No. 7397-EB-050, R&Dsystems, Minneapolis, MN, USA) was supplemented at concentrations of 0, 0.5, 1.0 and 2.0 μg/ml within the osteogenic/odontogenic induction culture milieu. In this case, the control is the same osteogenic medium without the supplementation of EphrinB2-FC. Additionally, there was another control whereby DPSCs were cultured for 21 days in the pre- sence and absence of 0.5 μg/ml EphrinB2-Fc, under non osteogenic-inducing conditions in normal culture medium.DPSCs were cultured for either 7, 14, or 21 days in normal or os- teogenic/odontogenic induction medium, and the RNeasy Plus Mini Kit (Qiagen, Valencia, CA, USA) was used to isolate total RNA, which was then reverse transcribed using the SuperScript VILO Master Mix (Life Technologies, Grand Island, NY, USA). qRT-PCR was ran utilizing the SYBR Select Master Mix (Applied Biosystems, Grand Island, NY, USA) on a StepOne Real-Time PCR System (Applied Biosystems, Grand Island, NY, USA). The following gene markers were analyzed: alkaline phosphatase (ALP), bone morphogenetic protein 2 (BMP2), bone sia- loprotein (BSP) osteocalcin (OCN), osteopontin (OPN), collagen type I (COL1), Runx2, dentin matrix protein 1 (DMP-1), dentin sialopho- sphoprotein (DSPP), Ephrin B2, EphB2, EphB4 and the internal control housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH).

The corresponding primer sequences of these gene markers are listed in Table 1. The amplification parameters utilized for qRT-PCR were as follows: 2 min at 50 °C, 20 s at 95 °C, and 40 cycles of 3 s at 95 °C followed by 30 s at 60 °C. The relative cycle threshold was de-termined using the 2−ΔΔ cycle threshold method and normalizedagainst the endogenous GAPDH gene. For osteogenic induction in the absence of exogenous EphrinB2-Fc, the qRT-PCR data was normalized to the day 0 time point before commencement of osteogenic induction. For all other experiments involving supplementation of exogenous EphrinB2-Fc, the qRT-PCR data was normalized with respect to the negative control without EphrinB2-Fc at the same time point. previous study (Zou et al., 2017). Altogether, there were three donors of DPSCs, and special care was taken to ensure that cells from only one donor was utilized for one particular set of experiment or analysis in this study, so as to avoid inter-batch variability in the experimental results. The ‘stemness’ of the isolated cells was validated by flow cy-tometry analysis of the expression of CD73, CD90, CD105, and CD45,together with multi-lineage differentiation assays (osteogenic/odonto- genic, adipogenic, and neurogenic induction), as previously described (Zou et al., 2017). Propagation of DPSCs was carried out through cul- ture in T75 culture flasks with alpha minimum essential medium (α- MEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1%(v/v) penicillin-streptomycin antibiotic solution.

The culture media was replaced every 4–5 days, and confluent monolayers were dissociated with 0.5% (w/v) trypsin-EDTA for routine passage. Cell culture was carried out at 37 °C within a humidified 5% CO2 incubator. DPSCs be- tween passages 5 to 10 were utilized for all experiments. DPSCs were seeded in 6-well plates and cultured in osteogenic/ odontogenic induction medium constituted of α-MEM supplementedwith 10% (v/v) FBS, 1% (v/v) penicillin-streptomycin antibiotic Cells were lysed in M-PER Mammalian Protein Extraction Reagent (Thermo Scientific, Rockford, IL, USA) combined with a protease in- hibitor cocktail (Thermo-Fisher Scientific, Carlsbad, CA, USA). After incubating for 30 min on ice and vortexing, the homogenized samples were centrifuged to remove insoluble material (14,000g for 15 min at 4 °C). The extracted cellular proteins were quantified with a bicincho- ninic acid assay kit (Pierce, Rockford, IL, USA) and then separated bysodium dodecyl sulfate–polyacrylamide gel electrophoresis using 10%Tris-Glycine gels (Bio-Rad, Hercules, CA, USA), followed by transfer onto a 0.45-μm Immun-Blot polyvinylidene fluoride membrane (GE Healthcare Life Sciences, Little Chalfont, UK). The blotted membranes were then blocked with PBS containing 5% (w/v) skimmed milk and 0.05% (v/v) Tween 20 at room temperature for 1 h before incubationwith appropriate primary and secondary antibodies.

Primary antibodies specific for EphrinB2 and phospho-EphrinB2 were obtained from Abcam (Cambridge, UK; Cat No. ab131536 and ab119323 respec- tively), while that for EphB2, EphB4 and beta-actin were obtained from Santa Cruz Biotechnologies (Santa Cruz, CA, USA; Cat No. sc-130752, sc-5536 & sc-47778 respectively). Primary antibodies specific for phospho-EphB4 were obtained from Signalway Antibody (College Park, MD, USA; Cat No. 12720). Primary antibodies specific for ERK1/2 and phospho ERK1/2, were obtained from Cell Signaling Technology(Danvers, MA, USA; Cat No. 4695 and 4377 respectively). The sec- ondary antibody utilized was horseradish peroxidase–conjugated goat anti-rabbit or anti-mouse immunoglobulin G (Cell Signaling Technology, Danvers, MA). Blots were visualized using Pierce ECL Western Blotting Substrate (Thermo-Fisher Scientific, Carlsbad, CA,USA). The dilutions of primary and secondary antibodies utilized for western blots ranged from 1:500 to 1:1000, depending on the manu- facturer’s recommendation for that particular antibody. The westernblots were performed in triplicates, and one representative gel for eachexperiment is presented in the results.The Alizarin red S assay was used to assess calcium mineralization within the DPSC cultures after 7, 14 and 21 days of osteogenic/odon- togenic induction. The DPSC cultures were initially fixed with 4% (w/v) formaldehyde at room temperature followed by a wash in deionized water, prior to being stained with 1% (w/v) alizarin red S(pH = 4.1–4.2) at room temperature. After another rinse in deionizedwater, the digital images of the stained mineralized nodules within the DPSC cultures were captured.

To quantify the calcium mineralization within each well of DPSC culture, 10% (w/v) sodium dodecyl sulfate solution was added, and the plates were incubated overnight at 37 °C. The lysates were then collected for absorbance readings at 405 nm with a SpectraMAX 340® microplate reader (Molecular Devices, Sunnyvale, CA, USA). The DNA contents of the dye extracts in 10% (w/v) SDS were quantified by the Quan-iT®Pico Green assay (Thermo-Fisher Scientific,Carlsbad, USA), as per the manufacturer’s instructions. The absorbancereadings of the dye extracts at 405 nm were then normalized with re- spect to cell number based on corresponding DNA content in the pre- sented results.For staining of alkaline phosphatase activity, the DPSC cultures were also fixed with 4% (w/v) formaldehyde, followed by rinsing in PBS, prior to being incubated with nitro-blue tetrazolium/5-bromo-4- chloro-3′-indolyphosphate (NBT/BCIP) substrate solution, which pro- duces a blue-colored insoluble product upon reaction with cellular al-kaline phosphatase. Additionally, alkaline phosphatase activity in whole cell lysates was quantified by the SensoLyte pNPP Alkaline Phosphatase Assay Kit (Cat No. AS-72146, AnaSpec Inc., Fremont, CA, USA), following the manufacturer’s protocol. Briefly, on day 7 of os- teogenic induction, the cells cultured on were lysed in 0.2% (v/v) Triton X-100 in deionized water. Subsequently, 50 μl cell lysate and 50 μl pNPP substrate solution were added into each well of a 96-well plate and incubated for 30 min at room temperature. Absorbance readings were read at 405 nm, and normalized with respect to cell number basedon DNA content quantified by the Quan-iT®Pico Green assay (Thermo- Fisher Scientific, Carlsbad, USA).Unless otherwise stated, all experiments were performed in tripli- cates for each group. The Student’s t-test was used to evaluate statis- tically significant differences between two groups, while for compar- ison between more than two groups, one-way Anova with Tukey’s post- hoc test was utilized. All statistical computations were carried out with the SPSS 19.0 Statistics Software (SPSS Inc, Chicago, IL). p values <.05 were considered statistically significant. 3.Results As can be seen in Fig. 1A, there were initially no significant dif- ferences in EphrinB2 gene expression by the induced DPSCs versus the uninduced control after 7 and 14 days of osteogenic/odontogenic dif- ferentiation. However, by day 21 of osteogenic/odontogenic induction, there was a significantly higher gene expression level of EphrinB2 by induced DPSCs versus the uninduced control (2.47 ± 0.41 versus 1.41 ± 0.09 folds w.r.t. day 0 respectively, p < 0.05). There was in- itially no significant difference in EphB2 gene expression on day 7 (Fig. 1B). However, by day 14 and 21 of osteogenic/odontogenic in- duction, there were significantly higher gene expression levels of EphB2 by induced DPSCs versus the uninduced control (day 14: 3.23 ± 0.69 versus 0.87 ± 0.23 folds w.r.t. day 0 respectively, p < 0.01; day 21:4.60 ± 0.48 versus 0.82 ± 0.10 folds w.r.t. day 0 respectively, p < 0.01) (Fig. 1B). EphB4 gene expression displayed a similar trend as EphB2, with initially no significant difference on day 7 (Fig. 1C). By day 14 and 21, however, there were significantly higher gene expression levels of EphB4 by induced DPSCs versus the uninduced control (day 14:2.37 ± 0.63 versus 0.79 ± 0.24 folds w.r.t. day 0 respectively,p < 0.01; day 21: 2.33 ± 0.17 versus 0.79 ± 0.24 folds w.r.t. day 0 respectively, p < 0.01) (Fig. 1C). The concomitant increase in osteo- genic marker gene expression by DPSCs over 21 days of osteogenic/ odontogenic induction, is shown in Supplementary Fig. 1.As can be seen in Fig. 2, the western blot results showed a similar trend as the gene expression data (Fig. 1), with elevated protein ex- pression levels of EphrinB2, EphB2 and EphB4 in the induced DPSCs versus the uninduced control, over 7, 14 and 21 days of osteogenic/ odontogenic differentiation. Nevertheless, differences in proteins ex- pression levels (Fig. 2) were much less pronounced, as compared to the gene expression data (Fig. 1). As can be seen in Fig. 3A & B, the western blot results showed that there were modest increases in the phosphorylation of both the EphrinB2 and EphB4 proteins respectively, at early time-points fol- lowing the initiation of osteogenic/adontogenic differentiation of DPSCs. Phospho-EphB2 could not however be detected at the early time-points following the initiation of odontogenic/osteogenic differ- entiation of DPSCs (data not shown).As can be seen in Fig. 4, the western blot results showed that protein expression levels of ERK1/2 in the induced DPSCs remained relatively constant over 120 min upon initiation of osteogenic/odontogenic in- duction. By contrast, the protein expression levels of phospho ERK1/2 gradually increased and peaked at the 30 min time-point, but subse- quently decreased afterwards.We performed an initial screening of the effects of varying con- centrations of recombinant EphrinB2-Fc (0, 0.5, 1.0 & 2.0 μg/ml) on the osteogenic/odontogenic differentiation of DPSCs over a duration of 14 days (Fig. 5). The expression of osteogenic marker genes ALP, BMP2,Col1, OCN and RUNX2 appear to be more optimal at 0.5 μg/ml of re- combinant EphrinB2-Fc, as compared to higher concentrations of1.0 μg/ml and 2.0 μg/ml. As seen in Fig. 5, the gene expression levels of Col1 and OCN were significantly lower at 1.0 μg/ml versus 0.5 μg/ml EphrinB2-Fc (p < 0.05); while at 2.0 μg/ml, the gene expression levels of ALP, Col1, OCN and RUNX2 were significantly lower than at 0.5 μg/ ml (p < 0.05). Hence, a concentration of 0.5 μg/ml of recombinant EphrinB2-Fc was utilized for further experiments. As seen in Fig. 6, after 21 days of osteogenic/odontogenic induction with 0.5 μg/ml of recombinant EphrinB2-Fc, the qRT-PCR results showed that there were significantly higher gene expression levels of BMP2, BSP, Col1, DMP1, OCN, OPN and RUNX2 by the EphrinB2- treated DPSCs versus the untreated control (1.69 ± 0.26, 2.83 ± 0.41,2.28 ± 0.14, 1.78 ± 0.41, 3.66 ± 0.71, 3.33 ± 0.71, 4.60 ± 0.43folds respectively w.r.t. the uninduced control, p < 0.05). There were however no significant differences in the gene expression levels of ALP and DSPP1 between the EphrinB2-treated DPSCs and the untreated control (p > 0.05). With regards to the day 7 time point, all of the middle and late osteogenic markers (BSP, DMP1, OCN, OPN, DSPP1) were not detectable by qRT-PCR.

Only some of the early or less-specific osteogenic markers can be detected such as ALP, BMP2, Col 1 and RUNX2, and the results are presented in Supplementary Fig. 2. There was significant upregulation of ALP and BMP2 expression by EphrinB2- Fc on day 7, while expression levels of Col1 and RUNX2 were un- changed. In the control study that investigated the effects of EphrinB2- Fc under non-osteogenic inducing culture conditions, the results (Sup- plementary Fig. 3) showed only a significant upregulation of BMP2, and a significant downregulation in ALP, while there were no significant changes in the gene expression levels of Col1 and RUNX2. The middle and late osteogenic/odontogenic markers such as BSP, DMP1, DSPP1, OPN and OCN could not be detected under non-osteogenic inducing culture conditions.As seen in Fig. 7A, there was more intense Alizarin red S staining for the EphrinB2-treated DPSCs versus the untreated control, after 7 days of osteogenic/odontogenic induction. Under light microscopy (Fig. 7B), there was observed to be more prominent mineralized nodules being formed by the EphrinB2-treated DPSCs versus the untreated control. This was further confirmed by quantification of the staining intensities by absorbance measurements (Fig. 7C), which showed a significant difference between the EphrinB2-treated DPSCs versus the untreated control (1.57 ± 0.04 versus 1.00 ± 0.06 respectively, p < 0.01). Nevertheless, there were no differences between the EphrinB2-treated DPSCs and the untreated control after 14 and 21 days of osteogenic/ odontogenic induction (data not shown). The EphrinB2-treated DPSCs also exhibited higher levels of alkaline phosphatase activity compared to the untreated control after 7 days of osteogenic/odontogenic induc- tion (Fig. 7D & E), which is consistent with the ALP gene expression data on day 14 (Fig. 5). 4.Discussion One of the promising strategies to augment the reparative process of diseased/damaged tooth would be the delivery of bioactive factors, either alone, or in combination with cells and/or scaffolds. Of particular interest would be bioactive factors that can promote the differentiation of DPSCs into functional odontoblast that synthesize reparative dentin. Commercially-available recombinant EphrinB2-Fc is one such pro- mising candidate factor, as the role of EphrinB2 and its cognate re- ceptor EphB2 and EphB4 in osteoblast function (Arthur et al., 2013; Martin et al., 2010; Tonna et al., 2014; Zhao et al., 2006) and osteo- genic differentiation of mesenchymal stem cells (MSCs) (Pennisi et al., 2009; Toda et al., 2017; Zhang et al., 2015; Zhang et al., 2016) have already been well-characterized by several studies. Because the odon- togenic lineage is closely-related to the osteogenic lineage, and DPSCs are a subtype of MSCs (Shi & Gronthos, 2003), it is thus plausible to hypothesize that EphrinB2 and its cognate receptor EphB2 and EphB4 play similar key roles in the odontogenic differentiation of DPSCs. Nevertheless to date, there has not yet been any study that has sys- tematically examined the role of these surface receptor molecules in the osteogenic/odontogenic differentiation of DPSCs. Hence, this study at- tempts to fill in this gap within the scientific literature. First of all, we analyzed the changes of endogenous expression le- vels of EphrinB2, EphB2 and EphB4 within DPSCs during 21 days of osteogenic/odontogenic induction in vitro. Our qRT-PCR results (Fig. 1) showed that there were indeed significant changes in the expression levels of EphrinB2 and its cognate receptors EphB2 and EphB4 during the course of induction culture, which was further corroborated by Western blot analysis of the corresponding protein expression levels. These thus validated that these surface receptor molecules do indeed have a role to play in the osteogenic/odontogenic differentiation pro- cess of DPSCs.Previously, Zhao et al. (2006) reported that the maintenance of bone homeostasis through osteoblast-osteoclast interaction was medi- ated through simultaneous forward and reverse signaling transduction through EphrinB2 and its receptor EphB4. Through gain- and loss-of- function experiments, it was demonstrated that reverse signaling through ephrinB2 expressed on osteoclast precursors suppressed os- teoclast differentiation by inhibiting the osteoclastogenic c-Fos-NFATc1 cascade. On the other hand, forward signaling through EphB4 ex- pressed on osteoblasts enhanced osteogenic differentiation. These re- sults were corroborated by the study of Martin et al. (2010) which showed that synthetic peptide antagonists blocking ephrinB2/EphB4 receptor interaction, and hence both forward and reverse signaling, were able to inhibit mineralization and expression of late osteogenic differentiation markers.Arthur et al. (2013) demonstrated that post-fracture bone repair was significantly faster in transgenic mice that overexpress EphB4 under the control of the collagen type 1 promoter (Col1-EphB4). Besides over- expression of EphB4, Tierney et al. (2013) demonstrated that re- combinant overexpression of EphrinB2 in MSCs can also enhance os- teogenic differentiation. Conversely, Tonna et al. (2014) demonstrated that ablation of ephrinB2 in murine osteoblasts inhibited expression of osteoblast differentiation markers, and that the same results were ob- tained with siRNA silencing of ephrinB2, but not the EphB4 gene. Ad- ditionally, Tonna et al. (2014) also showed that ablation of ephrinB2 in mice osteoblasts led to increased apoptosis of osteoblasts and osteocytes. In our previous study on the osteogenesis of PDLSCs (Zhu et al., 2017), we observed increased phosphorylation of EphrinB2 and EphB4 proteins during the early time-points following initiation of odontogenic/osteogenic differentiation of PDLSCs. Indeed, we obtained similar results in this study on DPSCs (Fig. 3). Additional Western blot analysis also demonstrated activation of the ERK1/2 signaling pathway through phosphorylation of the ERK1/2 protein at early time-points following initiation of osteogenic/odontogenic induction of DPSCs (Fig. 4), which is consistent with previous studies on the role of this signaling pathway in the osteogenesis of adult stem cells (Mizumachi et al., 2017; Murakami et al., 2017; Wang et al., 2016). The next step of the study attempted to optimize the concentration of recombinant EphrinB2-Fc for stimulating the osteogenic/odonto- genic differentiation of DPSCs in vitro. A range of different concentra- tions of recombinant EphrinB2-Fc was supplemented within the in- duction milieu (0, 0.5, 1.0 & 2.0 μg/ml), and the expression of osteogenic marker genes was examined after 14 days. The concentration range of between 0.5 to 2 μg/ml EphrinB2-Fc was examined in this study, because these were within the range of concentrations reported by previous studies that have investigated the effects of recombinant EphrinB2-Fc on the osteogenic differentiation of MSCs (Pennisi et al., 2009; Zhang et al., 2015, 2016), PDLSCs (Zhu et al., 2017), a pre-osteoblast cell line (Wang et al., 2017), and ST2 bone marrow stromal cell line (Li et al., 2015). It was observed that the expression of osteogenic marker genes ALP, BMP2, Col1, OCN and RUNX2 was optimal at a concentration of 0.5 μg/ ml of EphrinB2-Fc, whereas higher concentrations of 1.0 μg/ml and 2.0 μg/ml led to significant decreases in the expression of some marker genes such as ALP, Col1, OCN and RUNX2 (Fig. 5). Hence, further ex- periments utilized a concentration of 0.5 μg/ml EphrinB2-Fc. The observed decrease in the expression of some osteogenic marker genes at higher concentrations of EphrinB2-Fc could possibly arise from negative feedback mechanisms on the forward signaling pathways of EphB2 or EphB4 triggered by an excess of the EphrinB2 ligand. Kawano et al. (2012) previously reported a similar concentration-dependent negative feedback mechanism; with low dosages of EphrinB1 and EphrinB2 proteins co-stimulating murine T-cell proliferation, but higher concentrations exerting a strong inhibitory effect. To our knowledge, this is the first reported observation of a concentration- dependent feedback mechanism of EphrinB2 on the osteogenic/odon- togenic differentiation of DPSCs. We hypothesize that this could serve as a means of fine-tuning cell differentiation and lineage fate, according to the different microenvironments encountered by adult stem cells, which would have varying levels of EphrinB2 ligands. Subsequently, DPSCs were subjected to osteogenic/odontogenic induction in the presence of 0.5 μg/ml EphrinB2-Fc for 21 days, and more comprehensive qRT-PCR analysis revealed significant upregula tion of seven key osteogenic marker genes (BMP2, BSP, Col1, DMP1, OCN, OPN and RUNX2) in the treated DPSCs versus the untreated control (Fig. 6). Although DSPP1 expression by the treated DPSCs was higher than the untreated control, the difference was not statistically significant. By contrast, ALP expression levels were similar between the two groups after 21 days of osteogenic/odontogenic induction. This is consistent with peaking of ALP activity in DPSCS after 7 days of os- teogenic/odontogenic induction, followed by subsequent down- regulation, as previously reported (Yu et al., 2006). The enhancement of osteogenic/odontogenic differentiation of DPSCs in the presence of 0.5 μg/ml EphrinB2-Fc, was further corroborated by the Alizarin red S and ALP staining data on day 7 (Fig. 7). In the absence of osteogenic- inducing culture conditions, exogenous EphrinB2-Fc exerted only a marginal effect on osteogenic differentiation, with only one osteogenic marker gene BMP2 being upregulated (Supplementary Fig. 3). Our results are consistent with previous studies that demonstrated enhanced osteogenic differentiation of other cell types upon treatment with recombinant EphrinB2-Fc. Li et al. (2015) reported that ephrinB2- Fc significantly enhanced erythropoietin-mediated osteoblastic differ- entiation in ST2 bone marrow stromal cell line; while Wang et al. (2017) reported that activation of the ephrinB2-EphB4 signaling pathway by EphrinB2-Fc enhanced TNF-α-induced osteogenic differentiation of MC3T3-E1 murine pre-osteoblasts. Pennisi et al. (2009) and Zhang et al. (2015, 2016) demonstrated that the osteogenic differentiation of MSCs could be enhanced upon treatment with EphrinB2-Fc. Toda et al. (2017) immobilized recombinant EphrinB2-Fc within a novel hydrogel scaffold, which was used to enhance commit- ment of MSCs to the osteogenic lineage, as evidenced by upregulated expression of the upstream osteogenic marker RUNX2. Hence, this study demonstrated conclusively that signaling path- ways mediated by EphrinB2 and its cognate receptors play a key role in the osteogenic/odontogenic differentiation of DPSCs. In particular, EphrinB2 forward signaling enhances the osteogenesis/odontogenesis of DPSCs. Our results are consistent with previous studies on osteo- blasts, MSCs and Alizarin Red S PDLSCs.