MEST Regulates the Stemness of Human Periodontal Ligament Stem Cells

Periodontal ligament (PDL) stem cells (PDLSCs) have been reported as a useful cell source for periodontal tissue regeneration. However, one of the issues is the difficulty of obtaining a sufficient number of PDLSCs for clinical application because very few PDLSCs can be isolated from PDL tissue of donors. Therefore, we aimed to identify a specific factor that converts human PDL cells into stem-like cells. In this study, microarray analysis comparing the gene profiles of human PDLSC lines (2-14 and 2-23) with those of a cell line with a low differentiation potential (2-52) identified the imprinted gene mesoderm-specific transcript (MEST). MEST was expressed in the cytoplasm of 2-23 cells. Knockdown of MEST by siRNA in 2-23 cells inhibited the expression of stem cell markers, such as CD105, CD146, p75NTR, N-cadherin, and NANOG; the proliferative potential; and multidifferentiation capacity for osteoblasts, adipocytes, and chondrocytes. On the other hand, overexpression of MEST in 2-52 cells enhanced the expression of stem cell markers and PDL-related markers and the multidifferentiation capacity. In addition, MEST-overexpressing 2-52 cells exhibited a change in morphology from a spindle shape to a stem cell-like round shape that was similar to 2-14 and 2-23 cell morphologies. These results suggest that MEST plays a critical role in the maintenance of stemness in PDLSCs and converts PDL cells into PDLSC-like cells. Therefore, this study indicates that MEST may be a therapeutic factor for periodontal tissue regeneration by inducing PDLSCs.


Introduction
The periodontal ligament (PDL) is a fiber-rich connective tissue located between the alveolar bone and cementum covering the tooth root, which plays important roles in tooth support as well as nutrition, protection from bacterial attack, sensory input for mastication, and homeostasis [1][2][3][4]. However, in most cases, severe damage to PDL tissue caused by deep caries, periodontitis, or trauma results in tooth loss because the current therapies have limited effects and it is difficult to regain complete regeneration [5].
Previous reports have indicated that human PDL tissue contains somatic stem cells [6]. These cells termed as PDL stem cells (PDLSCs) express not only mesenchymal stem cell (MSC) surface markers, such as CD105 and CD146 [6][7][8][9][10], but also various stem cell-related markers, such as p75NTR (the neural crest marker) [10,11], N-cadherin (the mesenchymal stem cell marker) [10], and NANOG (the embryonic stem cell marker) [11,12] and possess self-renewal properties [7,13]. PDLSCs also display a multidifferentiation capacity for osteoblasts, adipocytes, and chondrocytes in vitro similarly to MSCs [6,14] and possess the capacity to generate cementum-and PDL-like tissues in vivo [6]. Other studies have reported that transplantation of autologous PDLSCs into human and swine periodontal defects regenerates PDL tissue [15,16]. Thus, it has been considered that the use of PDLSCs in tissue engineering techniques may be a critical method for regenerative periodontal therapy. However, because the percentage of resident stem cells in PDL tissue is very low [17] and isolation of PDLSCs involves tooth extraction, it has been difficult to stably obtain sufficient PDLSCs for research and clinical applications.
Therefore, we considered that a method to address these issues is induction of stem cell populations from PDL cells. Previously, we showed that semaphorin 3A (Sema3A) induces MSC-like properties in human PDL cells [18]. Sema3A-overexpressing PDL cells exhibit an enhanced capacity to differentiate into both osteoblasts and adipocytes, but not chondrocytes, although not having increased expression of all MSC markers. Thus, we attempted to identify a factor in PDLSCs to induce MSC-like properties more effectively.
In this study, we aimed to identify such a factor by microarray analysis to compare gene profiles among three clonal cell lines with different properties. Among them, 2-14 and 2-23 cells strongly express MSC surface markers, such as CD105 and CD146, and possess multidifferentiation capacities for osteoblasts, adipocytes, and chondrocytes in vitro [9][10][11]. Conversely, another cell line, 2-52, expresses MSC surface markers less than 2-14 and 2-23 cells and exhibits a limited differentiation capacity [18]. We aimed to identify the factor that was more highly expressed in 2-14 and 2-23 cells than in 2-52 cells and examine whether this factor enables conversion of human PDL cells into stem-like cells.

Materials and Methods
2.1. Cell Culture. Clonal cell lines 2-14, 2-23, and 2-52 were obtained from a limiting dilution of a heterogeneous immortalized human PDL fibroblast line. The heterogeneous immortalized human PDL fibroblast line was generated by transduction with both simian virus 40 large T-antigen and human telomerase reverse transcriptase into a human PDL cell population which was isolated from the healthy premolars of a 21-year-old female who visited Kyushu University Hospital for extraction [19]. Cell line 1-17 was also an immortalized human PDL cell line that we established previously [20]. 2-14, 2-23, and 1-17 cells were reported as human PDLSC-like cells in our previous studies [9,10,20]. Cells maintained in alpha-minimum essential medium (α-MEM; Gibco-BRL, Grand Island, NY) containing 10% fetal bovine serum (FBS; Biowest, Nuaillé, France) (10% FBS/α-MEM) at 37°C in a humidified atmosphere with 5% CO 2 were used in all experiments. All procedures were performed in compliance with the Research Ethics Committee of the Faculty of Dentistry at Kyushu University.

Osteoblastic Differentiation Assay.
Cells were seeded at 2 × 10 4 cells per well in 24-well plates (Becton Dickinson Labware, Lincoln Park, NJ) in 10% FBS/α-MEM as control medium (CM) or CM containing 2 mM β-glycerophosphate (Sigma, St. Louis, MO), 50 mg/ml ascorbic acid (Nacalai Tesque, Kyoto, Japan), and 1 × 10 −7 M dexamethasone (Merck Millipore, Darmstadt, Germany) as osteoblastic differentiation medium (DM). Half of the medium was exchanged every 2 or 3 days. After 3 weeks of culture, the cells were fixed with 4% paraformaldehyde (PFA; Merck Millipore) and then washed with distilled water and stained with Alizarin red S as described previously [9]. The area of each Alizarin red Spositive region was imaged under a Biozero digital microscope (Keyence, Osaka, Japan). Total RNA was isolated from each culture after 3 days.

Chondrogenic Differentiation
Assay. Cell suspensions of 2:5 × 10 5 cells per 15 ml polypropylene tube (Thermo Fisher Scientific, Waltham, MA) were centrifuged at 150 × g for 5 min, and the cell pellets were cultivated in complete chondrogenic medium with 10 ng/ml recombinant TGF-β3 (R&D Systems, Minneapolis, MN) for 4 weeks as described previously [9]. After fixing with 4% PFA in phosphatebuffered saline (PBS), the cell pellets were embedded in paraffin and sectioned into 5 μm thick slices. Alcian blue staining was performed to identify a cartilaginous matrix. Total RNA was isolated from each culture after 3 days.
2.5. Flow Cytometric Analysis. The expression of cell surface antigens on 2-14, 2-23, and 2-52 cells was analyzed by flow cytometry. Cells (2 × 10 5 /tube) prepared as a single cell suspension by trypsin/EDTA digestion and resuspended in flow cytometry buffer (R&D Systems) were incubated with antibodies (10 mg/ml) specific for surface markers or isotype control antibodies (10 mg/ml) on ice for 45 minutes. Anti-CD105-PE and anti-CD146-PE antibodies (eBioscience, San Diego, CA) and mouse IgG-PE isotype control were used. The cells were washed with flow cytometry staining buffer and analyzed using an EC800 cell analyzer (Sony Biotechnology, Tokyo, Japan).
2.6. Semiquantitative RT-PCR. PCR was performed using Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA) in a PCR Thermal Cycler Dice (Takara Bio Inc., Shiga, Japan) under the following conditions: 94°C for 2 min and then the appropriate number of cycles at 94°C for 30 s, appropriate annealing temperature for 30 s, 72°C for 30 s, and finally 72°C for 7 min. Primer sequences, annealing temperatures, cycle number, and product sizes for Type-1 collagen (COL-1), Periodontal ligament-associated protein-1 (PLAP-1),   Table 2. β-Actin was used as an internal standard. Expression levels of the target genes were calculated using ΔΔCt values.
2.12. WST-1 Proliferation Assay. Cells were seeded in 96-well plates (Becton) at 3 × 10 3 cells/well in 10% FBS/α-MEM and cultured for 24 and 48 h. The cell proliferation rate was measured using the Premix WST-1 Cell Proliferation Assay Sys-tem (Takara), according to the manufacturer's instructions. Briefly, at the indicated time points, 10 μl WST-1 reagent was added to the culture medium of each well. After 1 h, 100 μl of supernatant was collected from each well, and the absorbance at 450 nm was measured using a microplate reader (ImmunoMini NJ-2300; System Instruments Co., Ltd., Tokyo, Japan).

Gene Transfection and Establishment of Transfected
Clones. The coding region of the human MEST cDNA (Gen-Bank accession no. NM_002402.4) was inserted into the pcDNA3.1/Hygro (+) (Invitrogen) expression vector. 2-52 cells were transfected with pcDNA3.1/Hygro (+) alone or with pcDNA3.1/Hygro (+) containing the coding region of    Stem Cells International the human MEST cDNA using Lipofectamine LTX and Plus Reagent (Invitrogen). These were termed "2-52_empty" and "2-52_MEST," respectively. 2-52_empty cells were regarded as controls. The candidate clones of 2-52_empty and 2-52_ MEST cells were established by selection with hygromycin (Invitrogen). The levels of MEST expression in the clones sta-bly expressing the MEST cDNA were assessed by quantitative RT-PCR and western blot analysis.
2.14. Statistical Analysis. All experiments were performed in triplicate or quadruplicate. Values are expressed as the mean ± standard deviation. Statistical analysis was performed using Compare1_Zscore Compare2_Zscore

Expression of MEST in Human PDLSC Lines.
Microarray analysis showed that the expression level of MEST in 2-23 cells was higher than that in 2-52 cells (Figure 2(a)). Quantitative RT-PCR demonstrated that MEST gene was highly expressed in not only 2-14 and 2-23 cells but also 1-17 cells (Figure 2(b)). In addition, western blot analyses revealed that the expression level of MEST protein in 2-23 cells was higher than that in 2-52 cells (Figure 2(c)). Moreover, immunofluorescence staining showed that MEST was expressed in 2-23 cells, and the expression level of MEST in 2-23 cells was     Stem Cells International higher than that in 2-52 cells (Figures 2(d) and 2(e)). These results indicate that MEST is highly expressed in human PDLSCs.

MEST Knockdown Suppresses the Expression of Stem Cell
Markers and the Proliferative Potential in 2-23 Cells. To evaluate the roles of MEST in human PDLSCs, we examined the effects of MEST knockdown by siRNA on stem cell-like properties, such as the expression of stem cell markers, the proliferative potential, and the multidifferentiation capacity. Quantitative RT-PCR and western blot analyses confirmed that the expression of MEST mRNA (Figure 3(a)) and protein (Figure 3(b)) in 2- 23 (Figure 3(c) and Supplemental Figure 1b). In addition, the expression of stem cell-related markers, such as p75NTR, N-cadherin, and NANOG, in 2-23 cells was also downregulated by MEST knockdown (Figure 3(d)). The effect of MEST knockdown on the proliferative potential of 2-23 cells was examined using the WST-1 proliferation assay. We found that MEST knockdown significantly decreased the proliferation rate of 2-23 cells (Figure 3(e)). In addition, we examined the gene Compared with 2-23_siCont cells, the expression level of both genes was downregulated in 2-23_siMEST cells (Figure 3(f)). Moreover, in contrast to 2-23_siMEST cells, immunofluorescence staining showed that 2-23_siCont cells were positive for Ki-67, a typical marker of proliferation (Figure 3(g)). These results suggest that MEST knockdown suppresses the proliferative potential of human PDLSCs.  The Alizarin red S-positive area of 2-23_siMEST cells cultured in osteoblastic differentiation medium was smaller than that of 2-23_siCont cells (Figure 4(a)). In addition, the expression levels of bone-related genes, such as ALP and Osterix (Figure 4(b)), were significantly lower in 2-23_siMEST cells cultured in osteoblastic differentiation medium than in 2-23_ siCont cells. Similarly, the Oil red O-positive area of 2-23_siM-EST cells in adipogenic differentiation medium was smaller than that of 2-23_siCont cells (Figure 4(c)), and the expression levels of adipocyte-related genes, such as PPARγ and C/EBPα (Figure 4(d)), were significantly lower in 2-23_siMEST cells than 2-23_siCont cells. Moreover, the Alcian blue-positive cartilaginous matrix of 2-23_siMEST cells in chondrogenic differentiation medium was smaller than that of 2-23_siCont cells (Figure 4(e)), and the expression levels of chondrocyte-related genes, such as COL-2 and Aggrecan (Figure 4(f)), were significantly lower in 2-23_siMEST cells than 2-23_siCont cells. These results suggest that MEST knockdown suppresses the multidifferentiation capacity of human PDLSCs.  (Figures 5(a) and 5(b)). We found that the expression of MSC surface markers, such as CD105 and CD146 ( Figure 5(c) and Supplemental Figure 1c), and stem cellrelated markers, such as p75NTR, N-cadherin, and NANOG ( Figure 5(d)), was higher in 2-52_MEST cells than in 2-52_ empty cells. In addition, the expression of PDL-related genes, such as COL-1, PLAP-1, and POSTN, was also higher in 2-52_MEST cells than in 2-52_empty cells ( Figure 5(e)). Moreover, the capacities for osteoblastic, adipogenic, and chondrogenic differentiation in 2-52_MEST cells were higher than those in 2-52_empty cells and almost the same as those in 2-23 cells (Figures 6(a)-6(f)). In addition, interestingly, 2-52_MEST cells exhibited a change in morphology from a spindle shape to a stem cell-like round shape similar to 2-23 cells (Figures 7(a)-7(d)). These results suggested that overexpression of MEST induces PDLSC-like properties in human PDL cells.

Discussion
In the present study, the loss-of-function analysis using siRNA showed that MEST plays a critical role in the maintenance of stemness in PDLSCs, whereas the gain-of-function analysis employing gene transfer revealed that MEST might convert PDL cells into PDLSC-like cells. This is the first report to demonstrate MEST as a candidate factor that regulates the stemness of human PDLSCs.
First, we identified factors expressed highly in human PDLSC lines by microarray analysis and focused on MEST. MEST is an imprinted gene that plays important roles in embryonic development [21]. Imprinted genes affect all aspects of stem cells from establishment to differentiation [22]. Previous reports have detected imprinted gene expression in adult somatic stem cells and shown that genomic imprinting may be a marker for pluripotency [23,24]. In our study, MEST was highly expressed in multiple human PDLSC lines including 2-14, 2-23, and 1-17. These findings demonstrated that imprinted genes including MEST might be important in human PDLSCs. Moreover, we found that knockdown of MEST suppressed several stem-like properties in human PDLSCs. This finding confirmed that MEST might play a critical role in the maintenance of stemness in PDLSCs. To date, no factors involved in regulating the stemness of PDLSCs have been reported. Our present study suggested that MEST might be a novel factor that regulates the stemness of PDLSCs. In addition, the findings indicated the possibility that MEST induces stem-like cells.
Previously, transplantation of autologous PDLSCs into human and swine periodontal defects has been performed to regenerate PDL tissue [15,16]. Furthermore, among other dental stem cells, it has been considered that PDLSCs are the most suitable cells for periodontal tissue regeneration therapy [25,26]. Therefore, we investigated a method to obtain a sufficient number of PDLSCs for clinical application. Thus far, as a method to acquire PDLSCs, it has been common to separate them from extracted teeth [6]. However, the number of PDLSCs obtained from the extracted teeth of one patient is extremely small [17]. Therefore, as an alternative method, we considered whether PDLSCs could be induced from nonstem cells which are present in the PDL cell population.
Urodele amphibians, such as newts, regenerate amputated limbs and tails by dedifferentiating mature cells into undifferentiated cells [27]. In addition, it has been reported that ceiling cultures induce dedifferentiation of mature adipocytes into MSC-like multipotent cells [28]. Other reports have shown that treatment with reversine induces dedifferentiation of muscle-committed cells into multipotent progenitor cells [29]. However, such findings have not yet been reported for PDL cells. In our recent study, we established a method to dedifferentiate human PDL cells into PDLSClike cells by transferring a gene that is highly expressed in human PDLSC lines. Previously, we found that Sema3A

13
Stem Cells International converts 2-52 cells into MSC-like cells which expressed MSC surface markers and exhibited the differentiation capacity for osteoblasts and adipocytes [18]. However, these cells did not express PDL-related genes. PDL-derived stem-like cells have been reported to express both stem cell-related markers and PDL-related genes [6,9,10,20,30]. In our present study, we induced PDLSC-like cells that expressed both stem cell-related markers and PDL-related genes by overexpression of the MEST gene. To date, there has been no report of PDLSCs expressing both stem cell-related markers and PDL-related genes by transferring a single gene. In addition, these cells exhibited a change in morphology from a spindle shape to a round shape that was similar to human PDLSC lines. These findings indicate that MEST might induce more suitable PDLSC-like cells for periodontal tissue regeneration therapy.
In summary, this study indicates that MEST might enable obtaining sufficient numbers of PDLSCs for clinical application and may be a novel factor with the potential to regenerate PDL tissue efficiently.

Conclusion
MEST, one of imprinted genes, is highly expressed in human PDLSC lines. Then, MEST knockdown suppresses the expression of stem cell markers, the proliferative potential, and the multidifferentiation capacity for osteoblasts, adipocytes, and chondrocytes in human PDLSC lines. On the other hand, MEST overexpression induces PDLSC-like properties, such as the expression of stem cell markers and PDLrelated markers, the multidifferentiation capacity, and the PDLSC-like morphology in human PDL cells. In summary, this study indicates that MEST plays an essential role in the maintenance of stemness in PDLSCs and has the potential to convert PDL cells into stem-like cells. Therefore, MEST may be a novel factor with the potential to regenerate PDL tissue efficiently.

Data Availability
All data used to support the findings of this study are included within the article.

Conflicts of Interest
The authors declare no conflict of interest.