The calcium dynamics of human dental pulp stem cells stimulated with tricalcium silicate-based cements determine their differentiation and mineralization outcome

Calcium (Ca2+) signalling plays an indispensable role in dental pulp and dentin regeneration, but the Ca2+ responses of human dental pulp stem cells (hDPSCs) stimulated with tricalcium silicate-based (TCS-based) dental biomaterials remains largely unexplored. The objective of the present study was to identify and correlate extracellular Ca2+ concentration, intracellular Ca2+ dynamics, pH, cytotoxicity, gene expression and mineralization ability of human dental pulp stem cells (hDPSCs) stimulated with two different TCS-based biomaterials: Biodentine and ProRoot white MTA. The hDPSCs were exposed to the biomaterials, brought in contact with the overlaying medium, with subsequent measurements of extracellular Ca2+ and pH, and intracellular Ca2+ changes. Messenger RNA expression (BGLAP, TGF-β, MMP1 and BMP2), cytotoxicity (MTT and TUNEL) and mineralization potential (Alizarin red and Von Kossa staining) were then evaluated. Biodentine released significantly more Ca2+ in the α-MEM medium than ProRoot WMTA but this had no cytotoxic impact on hDPSCs. The larger Biodentine-linked Ca2+ release resulted in altered intracellular Ca2+ dynamics, which attained a higher maximum amplitude, faster rise time and increased area under the curve of the Ca2+ changes compared to ProRoot WMTA. Experiments with intracellular Ca2+ chelation, demonstrated that the biomaterial-triggered Ca2+ dynamics affected stem cell-related gene expression, cellular differentiation and mineralization potential. In conclusion, biomaterial-specific Ca2+ dynamics in hDPSCs determine differentiation and mineralization outcomes, with increased Ca2+ dynamics enhancing mineralization.

www.nature.com/scientificreports/ environment to promote healing and repair of soft and hard tissues 3,4 . While hydroxide ions create an alkaline environment responsible for antibacterial and anti-inflammatory activity, Ca 2+ ions play a comprehensive role as a vital intracellular second messenger that governs diverse cellular processes such as gene transcription, protein expression, cell proliferation, differentiation, apoptosis, and activation of excitatory cell types 5,6 . TCS-based cements release Ca 2+ in the microenvironment causing elevated extracellular Ca 2+ concentration and as a consequence transiently increase the intracellular Ca 2+ concentration due to calcium-sensing receptor (CaSR) activation 7,8 . The dynamic Ca 2+ dependent signalling system can affect numerous Ca 2+ sensitive enzymes that convert changes in extracellular Ca 2+ concentration and intracellular Ca 2+ dynamics into well-defined cell actions 9 . Analysis of the molecular cues embedded in the intracellular Ca 2+ dynamics will lead to better understanding of the intracellular signalling mechanisms involved in regulating the bioactivities of cells 10 . Intracellular Ca 2+ dynamics play a dual role by acting both as an initiator and mediator of stem cell differentiation. Thorough knowledge on the role of Ca 2+ dynamics in the differentiation of stem cells into a tissue-specific lineage may offer an alternative biotechnological approach to exploit the unique properties of stem cells 10 .
Calcium signalling is multifaceted and depends on the cell type 9 . Although Ca 2+ signalling plays an indispensable role in dental pulp and dentin regeneration, there is a lack of information on the Ca 2+ dynamics of human dental pulp stem cells (hDPSCs). The aim of the present study was to identify and correlate extracellular Ca 2+ concentration, intracellular Ca 2+ dynamics, pH, cytotoxicity, gene expression and mineralization ability of human dental pulp stem cells (hDPSCs) stimulated with two different TCS-based biomaterials; ProRoot WMTA and Biodentine. Our work shows significant differences in extracellular and intracellular Ca 2+ changes that link to distinct patterns of hDPSCs gene expression, cellular differentiation and mineralization potential.

Results
Calcium release and pH. Ca 2+ measurements in the cell medium, in response to exposure to TCS-based cements in the absence as well as presence of hDPSCs were performed (Fig. 8). In the absence of cells, Biodentine released significantly more Ca 2+ in the α-MEM medium than ProRoot WMTA (p < 0.001) (Fig. 1); no significant difference in pH was found between the groups (Supplementary Fig. 1). Interestingly, in the presence of hDPSCs in the culture dishes, the larger Biodentine-linked Ca 2+ release in the medium was reduced to the level observed with ProRoot WMTA (p < 0.001), indicating that the cells take up the extra Ca 2+ load provided by Biodentine (Fig. 1). Intracellular calcium. Ca 2+ imaging experiments to determine the cellular responses to biomaterial exposure and its associated changes in extracellular Ca 2+ were performed. Cells with acetylcholine (1 µM) were challenged to verify their responsiveness. Oscillatory changes in intracellular Ca 2+ that are typical for hDPSCs were found ( Supplementary Fig. 2). Subsequently, cells were exposed to TCS-based cements placed in an insert, which after baseline Ca 2+ recording, was lowered to contact the bathing solution under continuous Ca 2+ imaging (Fig. 2). Gross analysis of the intracellular Ca 2+ signal averaged over all cells in view, demonstrated a Ca 2+ increase characterized by a peak followed by recovery (Fig. 2).
Finer grained analysis of the Ca 2+ dynamics in individual cells, in response to TCS challenging with the two biomaterials were performed. This showed **that Biodentine produced a significantly higher maximum amplitude of the Ca 2+ transients (p < 0.0001) (Fig. 3A), a reduced time to reach the maximum (p < 0.0001) (Fig. 3B) and an increased area under the curve (p < 0.01) (Fig. 3C) as compared to ProRoot WMTA. The number of cellular Ca 2+ transients per cell dish did not differ significantly between the groups (Fig. 3D). On average, the number Figure 1. Changes in extracellular Ca 2+ ion concentration (mM) in α-MEM with and without hDPSCs after 30 min biomaterial exposure. One way ANOVA with Tukey post-hoc comparisons showed significantly higher Ca 2+ released by Biodentine than ProRoot WMTA (p < 0.001) when no cells were present. In the presence of hDPSCs, the larger Biodentine-triggered extracellular Ca 2+ increase was significantly reduced to the level observed with ProRoot WMTA (p < 0.001).  Fig. 3). www.nature.com/scientificreports/ Gene expression. Elevations in extracellular Ca 2+ concentrations 11 as well as intracellular Ca 2+ changes are well known to influence gene expression 12 . The response of four specific gene markers were tested: BGLAP, TGFβ, MMP1 and BMP2 to the two TCS-based biomaterials. Biodentine exposure for 1 day resulted in upregulation of BGLAP and TGF-β that was more than twice as large compared to ProRoot WMTA (Fig. 4A,B); only small differences were observed for MMP-1 (Fig. 4C). BMP2 was upregulated by Biodentine but ProRoot WMTA downregulated the gene (Fig. 4D), demonstrating again a pronounced difference between the two TCS-based cements. These experiments were repeated under conditions of intracellular Ca 2+ chelation by loading the cells with 1 µM BAPTA-AM which also present during the 1 day exposure to the TCS-based cements. Such treatment clearly suppressed the BGLAP and TGF-β responses observed with both biomaterials. By contrast, the small changes observed for MMP1 became more pronounced in BAPTA-AM treated cells. For BMP2, the effects were intermediate, with a reduction by one third of the Biodentine induced upregulation and a one third increased downregulation in response to ProRoot WMTA (Fig. 4D).
Experiments on cell death. Subsequently, it was verified whether cell survival/cell death was affected by any of the treatments used in the experiments on gene expression. Viability as measured with the MTT assay demonstrated that 1 day exposure to the TCS-based cements did not result in significant differences compared to control (Fig. 5A). However, BAPTA-AM treatment (as applied in the gene studies) significantly decreased the MTT signal (p < 0.01) (Fig. 5A). Further investigations with the TUNEL assay ( Fig. 5B) and live/dead staining ( Supplementary Fig. 4) did not reveal any difference in terms of cell death, suggesting that the BAPTA-AM induced decrease in MTT signal is most likely caused by suppression of mitochondrial metabolic activity resulting from the dampening effect of BAPTA-AM on intracellular Ca 2+ dynamics and not resulting from cell death 13 .
Mineralization assay. Further, the mineralization potential of TCS-based cements was verified. Therefore, hDPSCs cultured in the presence of TCS-based biomaterials were evaluated for 7, 14, 21 and 28 days. Our results show that Biodentine induced faster mineralization (14 days) compared to ProRoot WMTA (21 days). At 21 days, Biodentine showed extensive mineralization while ProRoot WMTA displayed scattered presence of mineral nodules comparable to that seen in the Biodentine group at 14 days (Fig. 6).

Discussion
TCS-based cements are primarily used in dentistry and their interaction with hDPSCs are of great interest 14 . Potential use of hDPSCs in tissue regeneration by virtue of their ability to differentiate into fibroblasts, osteoblasts, odontoblasts, adipocytes, neurogenic and myogenic tissues in vitro make them a favorable model system for studying Ca 2+ signals 15 . hDPSC derived odontoblasts play an active role in the transport and accumulation of Ca 2+ which leads to regeneration of the dentin-pulp complex 16 . In this study, custom prepared hDPSCs with www.nature.com/scientificreports/ 96% purity, extracted by an enzyme digestion method from healthy unerupted human third molars were used. Transwell membranes with a pore diameter of 0.4 µm were used to mimic the absence of direct contact between the biomaterial and hDPSCs while allowing for soluble compounds from the biomaterials to reach the cells, similar to clinical conditions 17 . TCS-based cements are known to release Ca 2+ in varying concentration, depending on the cement composition. In this study, Biodentine released significantly more Ca 2+ than ProRoot WMTA after 30 min in α-MEM medium (p < 0.001). This could be explained by the fact that despite both ProRoot WMTA and Biodentine contain TCS as their main ingredient, the structure and composition of the two TCS-based cements reveal certain differences that could contribute either directly or indirectly to the difference in Ca 2+ release. Compared to ProRoot WMTA, Biodentine has a smaller particle size, which is known to increase the Ca 2+ release 18 . In addition to TCS as the major ingredient, ProRoot WMTA contains dicalcium silicate and tricalcium aluminate 19 . These additional ingredients form minimal or no calcium hydroxide upon hydration leading to a lesser degree of Ca 2+ release. The fact that the Ca 2+ release was not accompanied by a corresponding rise in pH suggests that the source of this Ca 2+ is not only from the formation and dissolution of calcium hydroxide but possibly also from other calcium compounds being present. Biodentine uses calcium chloride as the liquid medium for hydration while ProRoot WMTA uses water. The use of calcium chloride could enhance the Ca 2+ release by accelerating the setting reaction, increasing calcium hydroxide formation and by Ca 2+ release from unreacted calcium chloride 20 .
The larger Ca 2+ release from Biodentine was not detectable in the presence of hDPSCs, indicating that Ca 2+ is taken-up by the cells. Cellular Ca 2+ uptake involves subsequent Ca 2+ binding to proteins and uptake into organelles like the endoplasmic reticulum (ER) and mitochondria 9,21 . This organelle-based Ca 2+ load may affect gene  Table 1 www.nature.com/scientificreports/ expression and viability/cell death. In fact, the ER-mitochondrial Ca 2+ axis is a major player and determinant of cell survival/cell death 22 and is also known to affect gene expression 23 .
Our results obtained from intracellular Ca 2+ measurements closely reflect the differences observed between the two TCS-based cements on extracellular Ca 2+ . The significantly larger Ca 2+ release from Biodentine correlates well with the significantly larger area under the curve of intracellular Ca 2+ changes (p < 0.01), the significantly higher maximum amplitude (p < 0.0001) and the significantly shorter time to maximum change (p < 0.0001) observed in Biodentine as compared to ProRoot WMTA (see Fig. 3). The larger Ca 2+ load associated with Biodentine may act through increased Ca 2+ entry into the cell, possibly via Cav1.2 L-type Ca 2+ channels 24,25 , ORAI1, an essential pore subunit of store-operated Ca 2+ entry (SOCE) channels in stem cells 26 or TRPM4 channels 27 . Increased Ca 2+ entry is followed by intracellular cycling between the cytoplasm, ER Ca 2+ stores and mitochondria in mesenchymal stem cells, thus leading to the observed Ca 2+ dynamics 28,29 .
Calcium dynamics can activate signalling pathways in both the nucleus and cytoplasm to induce gene expression by different pathways, that can function both as an inhibitor or activator of gene expression 30 . The frequency, duration and amplitude of Ca 2+ transients are essential for increasing the efficiency and specificity of gene expression 21 . There is a nonlinear relationship between gene transcription and intracellular Ca 2+ dynamics, that periodically exceeds the threshold for activation of gene transcription 31 . The Ca 2+ oscillation frequency hereby differentially controls the activation of different genes and direct cells to specific developmental pathways 32 . The present experiments did not reveal prominent oscillatory activity, with only 3 elevations occurring per cell over the 30 min recording period for both biomaterials. This makes it possible that amplitude rather than frequency is most important in the observed effects, as is the case in the slow oscillatory activity leading to oocyte www.nature.com/scientificreports/ activation 33,34 . In any case, the results revealed that Biodentine promoted mRNA expression of BGLAP, TGF-β, and BMP2 indicating that the increased Ca 2+ load and intracellular dynamics can enhance the expression of genes associated with odonto-/osteogenic differentiation necessary for successful regeneration of the dentinepulp complex 35,36 . Intracellular buffering with BAPTA-AM treatment inhibited the expression of BGLAP and TGF-β, thereby supporting an underlying role of Ca 2+ signalling in the upregulation of these genes. Furthermore, addition of BAPTA-AM reduced the upregulation of BMP2 but resulted in overexpression of MMP1. It is well known that an increase in extracellular Ca 2+ enhances the expression of BMP2 by activation of the Ca 2+ sensing receptor, elevation of intracellular Ca 2+ and stimulation of Ca 2+ /calmodulin-dependent nuclear factor of activated T cells (NFAT) signalling pathways 7 . Calcium plays a significant role in molecular processes responsible for mediating cell survival and death, including defence and programmed cell death mechanisms, such as cell cycle, apoptosis, and autophagy 37,38 . In the present study, combined MTT and TUNEL assays revealed absence of cytotoxicity after 1 day exposure to the biomaterials. Moreover, BAPTA-AM treatment caused suppression of mitochondrial metabolic activity causing a decrease in the MTT signal in all the groups. Inconsistent results regarding cell death were seen in the literature where Biodentine performed better 39 , similar [40][41][42][43] or worse 44 than MTA. Such contradictory results may be related to the type of target cells, method of cytotoxicity assessment, direct contact of cells with the materials and concentration of materials. Three-dimensional culture of dental pulp stem cells in direct contact to Biodentine and MTA revealed higher cell viability compared to ProRoot MTA after 1, 3 5 and 7 days 45 . In the study by Daltoe et al., serial dilutions (1:1, 1:10 and 1.100) of extracts from Biodentine and ProRoot MTA showed no significant differences in cell viability after 1 and 2 days in any of the concentrations tested 46 . The cytotoxic effect of Biodentine and ProRoot MTA on hDPSCs were concentration dependent as cell viability was higher at 1:100 concentration compared to 1:10 or 1:1 concentration. Similar results were obtained when hDPSCs were cultured in medium conditioned with Biodentine or ProRoot MTA 47 . When transwell membrane was used to avoid direct contact between the biomaterials and hDPSCs, there was no significant difference in cell viability and cell migration between ProRoot MTA and Biodentine 48 .
Further, the effects of TCS-based cements on the mineralization potential of hDPSCs were investigated using the Alizarin red and Von Kossa staining technique. Biodentine induced faster and increased mineralized nodule formation compared with the other groups, consistent with previous literature 42, 49,50 . The increased mineralization potential of Biodentine is in conformity with its larger material-linked Ca 2+ discharge, the larger intracellular www.nature.com/scientificreports/ Ca 2+ response and the upregulation of BMP2, BGLAP and TGF-β observed in our study. As Ca 2+ is deposited by osteoblasts and/or odontoblasts, the present results also highlight the osteogenic and/or odontogenic differentiation potential of hDPSCs in the presence of Biodentine and ProRoot WMTA. In agreement with the findings of the present study, previous literature suggests that in comparison to ProRoot MTA, Biodentine demonstrated significantly better cell survival and proliferation of osteoblasts and periodontal ligament cells 39,51 . Limitations to simulate the complex biological conditions of the clinical situation exist and hence the results obtained from the present study must be observed with caution for direct correlation with clinical scenarios. Calcium sensitive dyes are widely used for evaluating Ca 2+ signalling, but their applications have certain drawbacks. Special conditions are required for loading of the cells, bleaching may occur because of extended imaging periods and intracellular dye accumulation may increase cytoplasmic Ca 2+ buffering 52 .
It is essential to expand our knowledge on the numerous pathways by which Ca 2+ regulates cellular functions 9 . A typical characteristic of Ca 2+ signalling is the manner by which different Ca 2+ signals can be translated into specific cell functions depending on the type of signal 6 . The wide range of functions executed by Ca 2+ is attributed to the versatility in speed, amplitude, duration and spatiotemporal pattern of Ca 2+ signals as well as by interactions between Ca 2+ and other signalling pathways most likely mediated by different cellular processes 53 . Further research on the versatile patterns of Ca 2+ signals is essential for studying the mechanisms lying beyond the cellular functions 5 .

Conclusion
Our work demonstrates that hDPSCs take up the Ca 2+ released from TCS-based biomaterials without provoking cell death. The larger Biodentine-linked Ca 2+ load was reflected in altered intracellular Ca 2+ dynamics, which consequently resulted in differential gene expression, cellular differentiation and mineralization potential of hDPSCs stimulated with TCS-based cements.

Materials and methods
Isolation of stem cells. Human Dental Pulp Stem Cells (hDPSCs) were extracted from unerupted human third molars by enzyme digestion 54 . Written informed consent was collected from all patients and ethical approval was obtained from the Ethical Committee of University hospital, Ghent, Belgium according to laws of ICH Good Clinical Practice (GE11-LM-go-2006/57). The tooth crown was cleaned with iodine and 70% ethanol. The tooth was then cut with a bone cutter at the cemento-enamel junction to remove the pulp tissue and digested with type 1 collagenase and dispase. Cell suspension was cultured in a 25cm 2 flask in Alpha modified Eagles medium (α-MEM, Sigma-Aldrich, Overijse, Belgium) with 10% fetal bovine serum and antibiotics (100U/ml Pencillin and 100 mg/ml streptomycin) at 37 °C and 5% CO 2 . Flow cytometry analysis was performed to identify the purity of the stem cell culture obtained. Purity of the custom prepared stem cells were determined by mesenchymal stem cell markers CD90, and CD105. Custom prepared hDPSCs were grown to subconfluence and attained a purity of 96% (Fig. 7).
Cell culture. For the experiments, hDPSCs were seeded in a 24 well plate at a density of 40,000 cells/well (for Ca 2+ release, pH, cytotoxicity and mineralization assay), in a 6 well plate at a density of 5 × 10 5 cells/well (for qRT-PCR), or in 35 mm glass bottom dishes (MatTek Corporation, Massachusetts, USA) at a density of 1 × 10 5 (for live Ca 2+ fluorescence imaging). For intracellular buffering 1 µM of BAPTA-AM (Life technologies, California, USA) was added to α-MEM medium. All groups were maintained at 37 °C and 5% CO 2 for 1 day. Sample preparation. Biodentine (Septodont, Saint-Maur-des-Fossés, France) and ProRoot White MTA (WMTA) (Dentsply, Tulsa Dental, OK, USA) were mixed according to manufacturer's instructions and condensed in teflon moulds of height 1 mm and diameter 5 mm (for Ca 2+ release, pH, live Ca 2+ fluorescence imaging and cytotoxicity) or height 2 mm and diameter 8 mm (for qRT-PCR). Although the size of samples was different, the ratio of the exposed surface area of the sample to the volume of the surrounding medium (mm 2 /ml) was maintained as a constant in all experiments. The samples were allowed to set for 3 h in 100% relative humidity at 37 °C and sterilized by ultraviolet radiation for one hour. Samples were placed in transwell inserts of 0.4 µm pore size (Greiner bio-one, Kremsmünster, Austria) to avoid direct contact of the biomaterial with hDPSCs (Fig. 8). Positive control (without biomaterial), Biodentine and ProRoot WMTA groups were evaluated with/without BAPTA-AM loading. For cytotoxicity assay, qRT-PCR and mineralization assay, a sample size of n = 3/group were used while n = 6/group were used for Ca 2+ release, pH and live Ca 2+ fluorescence imaging experiments.
Calcium release and pH. qRT-PCR analysis. qRT-PCR analyses were done after 1 day. cDNA was produced and amplified using the Reverse Transcriptome kit (QuantiTect Reverse Transcription kit, Qiagen, Hilden, Germany). Target cDNA was amplified using specific primer pairs. qRT-PCR was performed using Sensimix SYBR No-ROX Kit (Bioline, London, UK) on Light cycler 480 System (Roche Life Science, Penzberg, Germany). Samples were normalized using qBasePlus (Biogazelle NV, Zwijnaarde, Belgium) against at least three of the following genes: Rpl13a, Eif4b, B2m, Actb, or Gapdh as described previously 55 . Details of specific primers used for gene expression analysis are provided in Table 1.   Mineralization assay. The cells were seeded on plastic coverslips of 13 mm diameter (Nunc Thermanox, ThermoFisher, Massachusetts, USA) in an adhesive 24 well plate. Osteogenic medium was prepared by supplementing standard culture medium with 10 mM β-glycerophosphate (Sigma-Aldrich), 100 µM L-ascorbic acid 2-phosphate (Sigma-Aldrich) and 100 nM dexamethasone (Sigma-Aldrich). Sample extracts were prepared at a non-toxic dilution as determined by MTT assay. For both Alizarin Red and Von Kossa staining, either pure extract with standard culture medium or extract supplemented with osteogenic medium was added to the cells. At 7, 14, 21 and 28 days, Alizarin Red (Alizarin red dye 1 g in 40 ml ultra-pure water, VWR, Oud Heverlee, Belgium) and Von Kossa (silver nitrate 1 g in 20 ml ultra-pure water, VWR, Oud Heverlee, Belgium) staining were performed. After rinsing with PBS, the cells were fixed with neutral buffered formaldehyde. The reaction was accelerated by adding formaldehyde-sodium carbonate solution (1 g in 15 ml ultra-pure water, VWR, Oud Heverlee, Belgium). Unreduced silver ions were removed by Farmer's solution (Sigma-Aldrich, Overijse, Belgium) containing 10% potassium ferrocyanide and 90% sodium thiosulfate. All images were captured with a microscope (Olympus BX51, Olympus, Tokyo, Japan) equipped with Xcellence software (Olympus, Tokyo, Japan).

Statistical analysis.
All data were subjected to statistical analysis by unpaired t-test with Welch's correction, analysis of variance (ANOVA) and individual comparisons were performed by Tukey post-hoc at a significance level of p < 0.05 using Statistical Package for Social Sciences (SPSS) v25.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism (version 6, GraphPad Software Inc., San Diego, CA, USA).
Conference presentation. Parts of this study has been presented at the 6th Belgian Symposium on Tissue Engineering, 2018.