In Vitro Characterization of Periodontal Ligament Stem Cells Derived from Supernumerary Teeth in Three-Dimensional Culture Method

Objective: The aim of this study was to compare the characteristics of periodontal ligament stem cells derived from supernumerary teeth (sPDLSCs), cultured using a three-dimensional (3D) method and a conventional two-dimensional (2D) method. Methods: The morphology, viability, and osteogenic differentiation of the cells were analyzed. In addition, gene expression was analyzed by RNA sequencing, to characterize the functional differences. Results: The diameter of the 3D-cultured sPDLSCs decreased over time, but the spheroid shape was maintained for 7 days. The osteogenic differentiation was similar in the 2D and 3D. The gene expression related to the extracellular matrix (7.3%), angiogenesis (5.6%), cell proliferation (4.6%), inflammatory response (3.7%), and cell migration (3.5%) differed (p < 0.05). Conclusions: Within the limitations of this study, sPDLSCs varied in formation and function, depending on the culture method. In future, it is necessary to study tissue engineering using the advantages of 3D culture and the fewer ethical problems of supernumerary teeth.


Introduction
Mesenchymal stem cells (MSCs) can be harvested from various human tissues, such as bone marrow, adipose tissue, and umbilical cord blood [1]. In addition, MSCs have been extracted from the periodontal ligaments of permanent teeth, primary teeth, and supernumerary teeth [2][3][4]. Periodontal ligament (PDL) represents a fibrous network, connecting the cementum of the tooth root and the alveolar bone [5]. It serves many functions, such as tooth support, nutrition, and protection. Periodontal ligament stem cells (PDLSCs) are easy to obtain and have a greater proliferation rate than bone marrow stem cells. PDLSCs are capable of differentiating into osteoblasts, cementoblasts, adipocytes, and chondrocytes, and can form a structure similar to periodontal ligaments in vitro [6][7][8][9].
In 2011, PDLSCs were obtained from a supernumerary tooth (sPDLSCs) [4]. The sPDLSCs show a greater colony-forming ability than bone marrow stem cells, and can differentiate into adipocytes and osteoblasts [10]. Expendable dental tissue, such as supernumerary teeth, may be a source of stem cells. Therefore, studies of the biological properties of PDLSCs obtained from various dental tissue are needed to optimize their regenerative effects.
The live/dead assay quantifies living and dead cells by measuring their esterase activity and plasma membrane integrity. The non-fluorescent dye calcein AM is converted into a fluorescent calcein that fluoresces in the presence of esterase in living cells, producing a uniform green fluorescence. Ethidium homodimer-1 enters dead cells through damaged membranes and binds with nucleic acids to produce red fluorescence.
The 2D-and 3D-cultured sPDLSCs were seeded at a density of 1.0 × 10 5 per well on a six-well plate and subjected to live/dead assay on days 3, 5, and 7. Images were obtained with a fluorescence microscope (IX71; Olympus, Tokyo, Japan). CCK-8 assay was performed on days 1, 3, 5, and 10, and the absorbance at 450 nm was measured with a Benchmark Plus multiplate spectrophotometer (Bio-Rad, Hercules, CA, USA).

Osteogenic Differentiation
After 2D or 3D culture, sPDLSCs were cultured in osteo-inductive medium (osteo group) or basal medium (control group) to confirm their ability to differentiate into bone cells. The osteo group was cultured for 2 weeks in osteo-inductive medium with 15% fetal bovine serum, 250 µL gentamycin reagent solution (final concentration 5 µg/mL), 5 mL filtered L-ascorbic acid (final concentration 10 mM), and 5 mL filtered dexamethasone. Alizarin red S (ARS) staining was performed to visualize calcium formation for evaluating differentiation. Cells were fixed in ice-cold 70% ethanol for 15 min, stained with ARS solution for 3 min, washed, and observed under an optical microscope. For quantitative evaluation of hard tissue formation, 10% cetylpyridinium chloride was added for 10 min, and the ARS stain was extracted and transferred to a 96-well plate. The
The live/dead assay quantifies living and dead cells by measuring their esterase activity and plasma membrane integrity. The non-fluorescent dye calcein AM is converted into a fluorescent calcein that fluoresces in the presence of esterase in living cells, producing a uniform green fluorescence. Ethidium homodimer-1 enters dead cells through damaged membranes and binds with nucleic acids to produce red fluorescence.
The 2D-and 3D-cultured sPDLSCs were seeded at a density of 1.0 × 10 5 per well on a six-well plate and subjected to live/dead assay on days 3, 5, and 7. Images were obtained with a fluorescence microscope (IX71; Olympus, Tokyo, Japan). CCK-8 assay was performed on days 1, 3, 5, and 10, and the absorbance at 450 nm was measured with a Benchmark Plus multiplate spectrophotometer (Bio-Rad, Hercules, CA, USA).

Osteogenic Differentiation
After 2D or 3D culture, sPDLSCs were cultured in osteo-inductive medium (osteo group) or basal medium (control group) to confirm their ability to differentiate into bone cells. The osteo group was cultured for 2 weeks in osteo-inductive medium with 15% fetal bovine serum, 250 µL gentamycin reagent solution (final concentration 5 µg/mL), 5 mL filtered L-ascorbic acid (final concentration 10 mM), and 5 mL filtered dexamethasone. Alizarin red S (ARS) staining was performed to visualize calcium formation for evaluating differentiation. Cells were fixed in ice-cold 70% ethanol for 15 min, stained with ARS solution for 3 min, washed, and observed under an optical microscope. For quantitative evaluation of hard tissue formation, 10% cetylpyridinium chloride was added for 10 min, and the ARS stain was extracted and transferred to a 96-well plate. The absorbance was measured at a wavelength of 562 nm with a Benchmark Plus multiplate spectrophotometer (Bio-Rad).

Library Preparation and Sequencing
Libraries of control and test RNA were constructed with the QuantSeq 3 mRNA-Seq library prep kit (Lexogen, Vienna, Austria) according to the manufacturer's instructions. In brief, 500 ng RNA was prepared, hybridized to an oligo-dT primer containing an Illuminacompatible sequence at its 5 end, and subjected to reverse transcription. After degradation of the RNA template, second-strand synthesis was initiated with a random primer containing an Illumina-compatible linker sequence at its 5 end. The double-stranded library was purified with magnetic beads to remove all reaction components. Next the library was amplified to add the complete adapter sequences required for cluster generation. The finished library was purified to remove PCR components. High-throughput sequencing was performed as single-end 75 bp sequencing using a NextSeq 500 sequencer (Illumina, San Diego, CA, USA).

Data Processing for the Identification of Differentially Expressed Genes
QuantSeq 3 mRNA-Seq reads were aligned with Bowtie 2. Bowtie 2 indices were generated from a genome assembly sequence or the representative transcript sequences for aligning the genome and transcriptome. The alignment file was used to assemble transcripts, estimate their abundance, and detect the differential expression of genes. Differentially expressed genes were identified based on counts from unique and multiple alignments using coverage in BEDtools. The read count data were processed based on the quantile normalization method with EdgeR in R (R Development Core Team, Vienna, Austria) with Bioconductor.

Statistical Analyses
Statistical analyses were performed with SPSS version 25.0 (IBM, Armonk, NY, USA). One-way analysis of variance (ANOVA) and the Scheffé post hoc test were performed to compare absorbance values. p < 0.05 was considered indicative of statistical significance.

Morphology
The sPDLSCs grown in the 2D culture exhibited a bipolar and stellate form, were attached to the plate, and increased in number over time (Figure 2A-D). The sPDLSCs grown in the 3D culture aggregated and became spheroid for about 24 h ( Figure 2E). Although the diameters of the spheroids decreased significantly over the first 5 days (75.3%), they maintained their shape ( Figure 2E-H).

Cell Viability
The absorbance of 2D-cultured sPDLSCs increased significantly over time, from 0.26 to 2.60 ( Figure 3). In particular, it increased exponentially on day 10 (p < 0.05). By contrast,

Cell Viability
The absorbance of 2D-cultured sPDLSCs increased significantly over time, from 0.26 to 2.60 ( Figure 3). In particular, it increased exponentially on day 10 (p < 0.05). By contrast, the absorbance of 3D-cultured sPDLSCs decreased over time, from 0.31 to 0.25. This decrease was significant from days 3 to 5 (p < 0.05), but not between days for the first 3 days (p > 0.05). At day 10, the absorbance increased slightly, but not significantly (p > 0.05).

Cell Viability
The absorbance of 2D-cultured sPDLSCs increased significantly over time, from 0.26 to 2.60 ( Figure 3). In particular, it increased exponentially on day 10 (p < 0.05). By contrast, the absorbance of 3D-cultured sPDLSCs decreased over time, from 0.31 to 0.25. This decrease was significant from days 3 to 5 (p < 0.05), but not between days for the first 3 days (p > 0.05). At day 10, the absorbance increased slightly, but not significantly (p > 0.05).
The 2D-cultured sPDLSCs did not show a significant decrease in viability; however, the cell viability decreased significantly in the center of the spheroids in the 3D culture ( Figure 4).

Figure 3.
Results of CCK-8 assays. The 2D-cultured sPDLSCs showed an increase in absorbance and 3D-cultured sPDLSCs showed a significant decrease in absorbance from day 1 to day 5. The absorbance of 3D-cultured sPDLSCs increased non-significantly on day 10. * p < 0.05; Scheffé post hoc test following one-way ANOVA. The 2D-cultured sPDLSCs did not show a significant decrease in viability; however, the cell viability decreased significantly in the center of the spheroids in the 3D culture ( Figure 4).

Stemness and Osteogenic Differentiation
ARS staining revealed significant extracellular calcium deposits in both the 2D and

Stemness and Osteogenic Differentiation
ARS staining revealed significant extracellular calcium deposits in both the 2D and 3D cultures, which was indicative of osteogenic differentiation ( Figure 5). Compared to the control group, the absorbance of the ARS stain extract was 2.3 in the 3D group, which was slightly higher than in the 2D group. Live/dead assays. The 2D-cultured sPDLSCs did not show a significant decrease in viability; however, the cell viability decreased significantly in the center of the spheroids in the 3D culture. (A-F) Live/dead assay results for 2D-cultured sPDLSCs at days 3, 5, and 7 (magnification, ×100). The number of cells increased, and no dead cells were observed. (G-L) Live/dead assay results for 3D-cultured sPDLSCs at days 3, 5, and 7 (magnification, ×100). The number of dead cells in the center of spheroids increased with decreasing spheroid diameter.

Stemness and Osteogenic Differentiation
ARS staining revealed significant extracellular calcium deposits in both the 2D and 3D cultures, which was indicative of osteogenic differentiation ( Figure 5). Compared to the control group, the absorbance of the ARS stain extract was 2.3 in the 3D group, which was slightly higher than in the 2D group.

Identification of Differentially Expressed Genes and Gene Ontology Analyses
The expression of 25,737 differentially expressed genes in the 2D-and 3D-cultured sPDLSCs was examined. Of these, 5664 genes, related to cell function, immunity, inflammation, and the extracellular matrix, were selected for further analyses. The expression of 89 genes (58.9%) was upregulated, and that of 62 genes (41.1%) was downregulated in the 3D culture compared to in the 2D culture. The genes up-or downregulated were related to the extracellular matrix (ECM; 7.3%), angiogenesis (5.6%), cell proliferation (4.6%), inflammatory response (3.7%), and cell migration (3.5%; Figure 6). The genes in the 3D-cultured sPDLSCs that showed at least a 50-fold increase or decrease in expression compared to the 2D-cultured sPDLSCs are listed in Tables 1 and 2.

Identification of Differentially Expressed Genes and Gene Ontology Analyses
The expression of 25,737 differentially expressed genes in the 2D-and 3D-cultured sPDLSCs was examined. Of these, 5664 genes, related to cell function, immunity, inflammation, and the extracellular matrix, were selected for further analyses. The expression of 89 genes (58.9%) was upregulated, and that of 62 genes (41.1%) was downregulated in the 3D culture compared to in the 2D culture. The genes up-or downregulated were related to the extracellular matrix (ECM; 7.3%), angiogenesis (5.6%), cell proliferation (4.6%), inflammatory response (3.7%), and cell migration (3.5%; Figure 6). The genes in the 3D-cultured sPDLSCs that showed at least a 50-fold increase or decrease in expression compared to the 2D-cultured sPDLSCs are listed in Tables 1 and 2.  Table 1. Genes upregulated more than 10-fold in 3D-compared to 2D-cultured sPDLSCs.

Gene Name
Fold Change Description Related Function Figure 6. Expression of filtered gene categories. Distribution and ratio of genes in 10 selected categories with 10-fold increased expression (p < 0.05).

Discussion
A scaffold-free 3D culture method, Stemfit 3D ® , was used for the 3D culture of sPDLSCs, and was compared to a traditional 2D culture. The analyses of morphology, viability, osteogenic differentiation, and gene expression confirmed that the functions and characteristics of the sPDLSCs varied depending on the culture method.
Stemfit 3D ® uses a non-adhesive plate, which is 600 µm in diameter, with a poly dimethylsiloxane-based concave micromold. Concave micromolds were used because they accelerate cell aggregation compared to other methods (plane, cylindrical), and the cells form spheroids of uniform size that are easily harvested [2]. Moreover, 3D culture methods that use low-adhesive plates have lower long-term stability than other methods, because of the difficulty limiting the size of the spheroids [21]. However, because Stemfit 3D ® can control the size of the spheroids, it shows improved stability over a longer period. Because of its economy, its ease of use, and the stability of the spheroids, Stemfit 3D ® is commonly used in stem cell cytology and studies of other types of cells [3,4].
The diameter of the sPDLSC spheroids in the 3D culture decreased (75.3%) over the first 5 days, and decreased more slowly thereafter. The initial reduction in diameter was caused by cell aggregation and organization, but the subsequent reduction appears to have been a result of restricted nutrients and oxygen. Several models for the transport of nutrients, oxygen, and waste in spheroids have been verified [8]. When the diffusion of oxygen is limited, avascular tissue forms, and inefficient transport results in the accumulation of metabolic waste and the formation of a necrotic core at the center of the spheroid [10,12,14]. Several studies on MSC spheroids, including ones by Hildebrandt et al. [14], who investigated human MSCs, and by Yamaguchi et al. [18], who evaluated rat MSCs, have reported that spheroids form within 1 day and that their diameters decrease over time. However, Lee et al. [19] reported that the diameter of human dental pulp stem cell spheroids, cultured on a six-well non-adhesive plate, increased over time. This discrepancy is due to the different culture methods used. Hildebrandt et al. [14] and Yamaguchi et al. [18] independently cultured MSCs in a small volume of medium, or used the scaffold method, whereas Lee et al. [19] used a relatively large volume of medium. It is possible that spheroid-spheroid interactions may increase the diameter of the spheroids, but further study is needed.
CCK-8 and live/dead assays were used to confirm the proliferative ability and viability of 3D-cultured sPDLSCs; these are frequently used in 2D culture experiments [22,23]. The CCK-8 assay revealed that the absorbance of the 2D culture increased over time, and that of 3D culture tended to decrease over time. This was probably due to continued proliferation of cells in the 2D environment, whereas in the 3D environment the differentiation rate decreased as a result of restricted nutrients and oxygen in the center of the spheroids [20]. In addition, in the 3D culture, the number of dead cells in the center of the spheroids increased slightly over time. Similarly, in one study, spheroids of human MSCs in pallet culture maintained a stable internal structure for one month. However, a necrotic area developed in the center of the spheroids, with a loss of proliferation, impaired structural stability, and decreased cell-to-cell contact at two months [20]. Although quantitatively evaluating 2D and 3D culture methods is difficult [24], these results support the notion that the characteristics of 3D-cultured cells are fundamentally different from those of 2D-cultured cells.
Three-dimensional culture improves osteogenic differentiation compared to 2D culture. In addition, the intercellular interactions of embryonic stem cells are enhanced in a 3D environment [25]. For example, progenitor cells derived from the salivary gland can differentiate into hepatocytic and pancreatic islet cell lineages only when cultured in a 3D environment [26]. For this reason, 3D culture has attracted attention in stem cell biology and oncology, as well as in dentistry [12]. For example, 3D-cultured dental pulp stem cells or gingiva/papilla-derived stem cells show higher ALP activity than those grown in 2D culture [3,6,7]. In addition, the expression of genes associated with bone formation, such as BMP2, RUNX2, and dentine sialophosphoprotein, is increased in 3D-cultured cells [19].
The expression of BMP2 was upregulated 112.8-fold in 3D culture compared to in 2D culture. However, compared to our previous studies, the expression of osteogenic differentiation factors, such as RUNX2 and BMP2, tended to be lower in sPDLSCs than in pPDLSCs. Therefore, sPDLSCs can differentiate in 2D and 3D cultures, but their osteogenic differentiation is lower than that of cells extracted from permanent teeth.
RNA sequencing was conducted to analyze the gene expression profiles of sPDLSCs grown in 2D and 3D cultures. Analyses of gene expression profiles by microarray or PCR are limited by the fact that only selected genes can be identified. Because RNA sequencing captures a wider range of active genes, it is possible to detect small changes in expression with a limited amount of transcript [27].
The initial coalescence of spheroids is caused by caspase-dependent IL-1 autocrine signaling, which upregulates EGR2 [16]. In this study, the 3D-cultured sPDLSCs showed upregulated expression of EGR2 (2.3-fold). In addition, expression of EGR1 and EGR3 increased 5.4-and 9.7-fold, respectively. Similarly, immunomodulators, such as TN-FAIP6 (24.5-fold), are reportedly upregulated in response to IL1 autocrine signaling in 3D spheroids [29,30]. In the center of spheroids, where the apoptosis zone is formed as a result of nutrient and oxygen limitation, hypoxia-and apoptosis-related genes, such as VEGFA and hypoxia-inducible factor (HIF), are upregulated. In this study, hypoxia-associated genes, such as VEGFA (6.1-fold) and HIF3A (2.0-fold), and apoptosis-related genes, such as IL24 (156-fold) and COMP (SFRP2), were highly upregulated in the spheroids. In addition, the expression of genes related to angiogenesis, such as PGF (341-fold) and PTGS2 (195.6-fold), was highly upregulated. These results are consistent with the observation that over time, the apoptotic zone of the spheroid increased and the surface proliferation zone decreased.
This study has several limitations. First, the decreased diameter and reduced proliferative ability of the 3D-cultured sPDLSCs may have been a result of their inherent characteristics or the morphologic characteristics of the spheroid, which hamper the supply of nutrients and oxygen, and the discharge of waste products. Our results are compatible with those of other studies that used 3D pallet cultures, low-adhesive plates, or scaffolds. In this regard, it may be necessary to compare other 3D methods, such as a bioreactor. Second, it may not have been optimal to use assays optimized for a 2D environment to evaluate 3D-cultured cells. For example, the CCK-8 assay involves measuring the color change caused by the dehydrogenation of water-soluble tetrazolium salts by living cells. Therefore, it is not ideal for comparing the absorbance of 2D and 3D cultures quantitatively, because the number of cells contacting the solution differs. It is difficult, or impossible, to find protocols and assays optimized for 3D-cultured sPDLSCs, so further research is needed to establish optimized methods for analyzing 3D-cultured cells. Third, this was a pilot study of 2D-and 3D-cultured sPDLSCs, and so the sample was small. Fourth, several prior studies on spheroid morphology involved observations over only 5-7 days. In the previous work of this research team, pPDLSCs were observed for 20 days to confirm their long-term stability [9]. However, the long-term stability of sPDLSC spheroids was not evaluated in this study. Therefore, further investigation of the long-term stability of sPDLSC spheroids is needed.
We found that 3D-cultured sPDLSCs exhibit a different gene expression profile, morphology, and physiology compared to identical 2D-cultured cells. Our results contribute to the development of optimized methods for the 3D culture of dental stem cells.

Conclusions
Supernumerary teeth, as expendable dental tissue, may be ethically less restrictive for clinical use than permanent teeth. Thus, they could be a good source of stem cells. Within the limitations of this study, we confirmed that the function of sPDLSCs varies depending on the culture method used. In addition, 3D-cultured sPDLSCs have greater stemness than 2D-cultured sPDLSCs. Our findings accelerate the development of regenerative medicine using MSCs derived from expendable tissue, including supernumerary teeth. Further studies that aim to enable the clinical application of 3D-cultured sPDLSCs are warranted.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are openly available.