Enhancement of Mesenchymal Stem Cell-Driven Bone Regeneration by Resveratrol-Mediated SOX2 Regulation

Mesenchymal stem cells (MSCs) are an attractive cell source for regenerative medicine. However, MSCs age rapidly during long-term ex vivo culture and lose their therapeutic potential before they reach effective cell doses (ECD) for cell therapy. Thus, a prerequisite for effective MSC therapy is the development of cell culture methods to preserve the therapeutic potential during long-term ex vivo cultivation. Resveratrol (RSV) has been highlighted as a therapeutic candidate for bone disease. Although RSV treatment has beneficial effects on bone-forming cells, in vivo studies are lacking. The current study showed that long-term (6 weeks from primary culture date)-cultured MSCs with RSV induction retained their proliferative and differentiation potential despite reaching ECD. The mechanism of RSV action depends entirely on the SIRT1-SOX2 axis in MSC culture. In a rat calvarial defect model, RSV induction significantly improved bone regeneration after MSC transplantation. This study demonstrated an example of efficient MSC therapy for treating bone defects by providing a new strategy using the plant polyphenol RSV.

Resveratrol (RSV) is a naturally occurring polyphenolic compound with beneficial biological effects such as antiaging [9], antioxidant [10], and bone protective activities [11], indicating its potential as a therapeutic candidate. We previously optimized methods for culturing MSCs in vitro with RSV such as the concentration and duration of RSV treatment and procedure for passaging the cells [12]. In the study, we further evaluated the safety and efficacy of using MSCs cultured with RSV for MSC therapy. Although many studies have reported the effects of RSV in MSCs [13][14][15], no study has confirmed the efficacy of RSV in bone healing in vivo. Therefore, the aim of this study was to examine RSV application for MSC-based therapy in the preclinical setting, particularly in bone regeneration. In this study, we confirmed that long-term treatment with 1-μM RSV retained the proliferative and differentiation potential of MSCs during prolonged ex vivo culture until effective cell doses (ECD) were reached for cell therapy. We designated the MSCs reaching ECD (1 x 10 7 cells) as 6w-MSCs, since it took about 6 weeks to acquire the ECD cell number. Transplantation of 6w-MSCs with RSV induction clearly enhanced the bone regeneration potential in a rat model with calvarial defects. Therefore, MSCs expanded via cell culture with RSV induction are an efficient cell source for stem cell therapy in regenerative medicine.

Isolation, cultivation, and RSV treatment
Bone marrow aspirates were obtained from the posterior iliac crests of six adult donors with the approval of institutional review board (IRB) of Yonsei University College of Medicine (IRB No. 4-2017-0232), and written informed consent was obtained from all patients. All methods were performed according to relevant guidelines and regulations of the institution. MSCs isolated from bone marrow were selected and cultured in accordance with published protocols [16], and their characteristics (i.e., positivity for CD90 and CD105, and negativity for CD34 and CD45) were confirmed by flow cytometry. MSC characteristics of the harvested cells were confirmed in our previous reports [16]. All MSCs used in this study were isolated and cultured individually.
Each experimental result was confirmed in at least three donor samples as well as triplicates. MSCs were maintained in low-glucose Dulbecco's modified Eagle's medium (DMEM-LG; Invitrogen, Grand Island, NY, USA) that was supplemented with 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA) and 1% antibioticantimycotic solution (Invitrogen) at 37 °C in 5% CO2 atmosphere. The cells were grown to 80% confluence, and then detached by incubation with 0.25% trypsin ⁄EDTA (Invitrogen) centrifuged at 188 g for 3 min. The cells collected by centrifugation were re-plated at density of 2 x 10 5 cells, and subcultivation was continued until the number of cells reached 1 x 10 7 cells. The concentration of RSV (Sigma-Aldrich, St. Louis, MO, USA) used (1 µM) was selected according to a previous study [12]. MSC cultivation was continued without freezing or terminating the culture until the number of cells reached 1x10 7 cells. Cells were subcultured every 3 days, and then counted each time. At each subculturing session, 2 x 10 5 MSCs were seeded onto new 10-cm 2 culture dishes. The number of MSCs was recorded every time the cells were subcultured, and the numbers were accumulated until we had approximately 1x10 7 cells; after that, the cell culture was terminated. The period lasted for about 6 weeks. The passage number, which reached 1 x 10 7 cells, can be somewhat different for each MSC derived from different donors. Therefore, the number of each passage requiring 1 x 10 7 cells has been recorded in each figure legend.

Colony-forming unit fibroblast assay
MSCs were seeded at 1 × 10 3 cells in 100-mm culture dishes and maintained in DMEM-LG supplemented with 20% fetal bovine serum for 12 days. The cells were fixed in 1:1 acetone:methanol, stained with 20% crystal violet solution (Merck, Darmstadt, Germany) for 10 min in the dark, and washed with distilled water. The colonyforming ability of the stained cells was evaluated for three donors in triplicate.

Cell proliferation assay
Proliferative capacity of 6w-MSCs was examined using EZ-Cytox Kit (Daeil Lab Service, Seoul, Korea). The 6w-MSCs were seeded in 12-well culture plate at density of 1 × 10 4 cells per well. DMEM-low glucose containing 10% FBS medium was used for maintaining cells for 7 days, and culture media were replaced every 2 days during assay periods. To measure levels of cell proliferation, cells were washed with PBS, and 20 μl of EZ-Cytox (tetrazolium salts) solution was added to each well and incubated at 37 °C for 4 h. After incubation, the conditioned media were transferred to 96-well plate. Absorbance was measured at 450 nm. All samples were tested in triplicate (n =3).

Osteogenic and adipogenic differentiation
The materials and methods for MSC differentiation, staining, and quantitative analysis were described previously [16]. Briefly, MSCs were seeded at 8 X 10 4 cells per well in 12-well culture plates. For osteogenic differentiation, cells were maintained for 10 days in osteogenic medium [DMEM-low glucose medium containing 10% FBS, 1% antibiotic-antimycotic solution, 100 nM dexamethasone (Sigma), 10 mM βglycerophosphate (Sigma), and 50 μg per ml ascorbic acid (Gibco)]. Alizarin red S staining was employed to determine osteogenic differentiation activity. For alizarin red S staining, after being fixed in ice-cold 70% ethanol, 1 ml of freshly prepared 3% alizarin red S solution (wt/vol) (Sigma) was added, then incubated in the dark for 30 minutes. For quantitative analysis of alizarin red S, absorbance was detected at 595 nm after destaining with 10% cetylpyridinium chloride monohydrate (Sigma) for 30 minutes. For adipogenic differentiation, cells were maintained for 10 days in adipogenic medium [DMEMlow glucose medium containing 10% FBS, 1% antibioticantimycotic solution, 1 μM dexamethasone, 0.5 mM isobutyltethylxanthin (Sigma), 5 μg ⁄ ml insulin (Gibco), and 200 lM indomethasin (Sigma)]. Oil red O staining was employed to determine adipogenic differentiation activity. To stain lipid droplets by oil red O, after being fixed in 10% neutral buffered formalin, 1 ml of 0.18% oil red O solution (Sigma) was added and incubated for 30 minutes. For quantitative analysis of oil red O-stained cells, absorbance was detected at 500 nm after destaining with 100% isopropanol for 30 minutes.

Immunoprecipitation
Cell lysates from MSCs were prepared with nondenaturing lysis buffer, as previously described [16]. Lysates were incubated with protein A/G agarose beads (Santa Cruz Biotechnology) and antibodies against acetylated lysine (Cell Signaling Technology) and SOX2 (Abcam). Beads conjugated with both lysates and antibodies were collected after centrifugation and washed three times with lysis buffer. The complexes were released from beads by boiling with 2× SDS sample dye, and then western blotting was performed. All membranes were incubated with antibody against SOX2 (Abcam) for 12 h, followed by incubation with HRP-conjugated secondary antibody for 1 h.

Cell cycle analysis
Single-cell suspension of differentiated cells was prepared as previously described [17]. MSCs treated with ethanol (vehicle) or RSV were harvested by incubation with 0.25% trypsin/EDTA and washed twice in PBS. Cells from each group (1 × 10 6 ) were fixed in ice-cold 70% ethanol for 30 min, stained with 50 µg/ml propidium iodide (Sigma-Aldrich) containing 100 µg/ml RNase A (Sigma-Aldrich) for 15 min at 4 °C, and then analyzed using FACSCalibur™ instrument (BD Biosciences, San Jose, CA, USA) to detect the cell cycle distribution. All samples used were tested in triplicate (n = 3).

Senescence-associated-β-galactosidase assay (SA-β-gal assay)
The SA-β-gal assay was performed using a cellular senescence assay kit (Millipore, Billerica, MA, USA) following the manufacturer's instructions. Cells were washed with PBS, and then were fixed for 10 min at room temperature with 1X fixing solution. After w ashing with D.W, cells were stained with prepared 1X SA-β-gal detection solution over 4h in the dark at 37 °C incubator without CO2. The number of positive cells was quantified by Image J software, version 1.41 (National Institutes of Health, Bethesda, MA, USA).

Rat calvarial defects
All animal use procedures were conducted with the approval of the Institutional Animal Care and Use Committee of Yonsei University College of Medicine (Approval number: IACUC-2016-0099). Twelve-weekold male Sprague-Dawley rats were anesthetized by an intraperitoneal injection of Zoretile (30 mg/kg body weight) and xylazine (10 mg/kg body weight). After shaving the hair on the head, a longitudinal incision was made in the skull, and then critical-sized calvarial bone defects with a diameter of 8 mm were created using a trephine bur. The defects were irrigated with saline and MSCs (1 × 10 6 /defect) mixed with fibrin glue were implanted into the defects, followed by suturing of the soft tissue. For pain relief, rats were administered a subcutaneous injection of meloxicam (0.2 mg/kg body weight). At 4-8 weeks after operation, all rats tested were sacrificed and the skull was harvested for micro-computed tomography (µCT) and histological analysis (n = 10 per a group).

µCT analysis
After fixation in 10% formalin for 5 days, the skulls were scanned with a high-resolution µCT instrument (SkyScan 1076; Bruker, Billerica, MA, USA) to quantitatively evaluate calvarial bone regeneration at the defect site as described previously [18]. Briefly, the imagery was reconstructed and analyzed using NRecon v1.6.6.0 and CTAn v1.13.2.1 (Bruker), respectively. Threedimensional model visualization software CTVol v2.0 (Bruker) was used to analyze calvarial bone regeneration. The settings for X-ray source were 70 kVp voltage and 140 mA current, and 0.5-mm-thick aluminum filter was used for beam induration. Pixel size was 18 mm, exposure time was 1475 ms, and rotation step was 0.5°, with a complete rotation through 360°.

Histological analysis and immunohistochemistry
The skulls were fixed for 1 week in 10% formalin and then embedded in paraffin. The paraffin-embedded sections were deparaffinized, rehydrated, and washed three times with PBS, and then the sections were used to evaluate tissue repair in the damaged regions. Histological analysis and immunohistochemistry were performed as previously described [18]. Briefly, tissue samples were sliced at a thickness of 4 µm and stained with hematoxylin and eosin (H&E) to observe new bone formation or incubated with human anti-vimentin antibodies (Santa Cruz Biotechnology, 1:100 dilution) to confirm whether the regenerated bone tissues were derived from human origin. Human vimentin was detected with a secondary goat antimouse IgG-horseradish peroxidase antibody (GenDEPOT, Katy, TX, USA) and 3,3′-diaminobenzidine (Vector Laboratories, Burlingame, CA, USA). The stained samples were observed using VS120 virtual microscope (Olympus, Tokyo, Japan), and sample images were analyzed using OlyVIA 2.5 software (Olympus). For immunofluorescence, paraffin-embedded tissue sections with a thickness of 4 µm were deparaffinized, Phycoerythrin-conjugated goat anti-rabbit secondary antibodies (Santa Cruz Biotechnology) and FITC-conjugated goat anti-mouse secondary antibodies were used to visualize the primary antibodies. All primary and secondary antibodies were applied at dilutions of 1:100 and 1:5000, respectively. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI, Sigma). Images were examined using Zeiss LSM 700 confocal laser scanning microscope (ZEN 2012 software; Carl Zeiss Micro Imaging GmbH, Jena, Germany).

Chromosomal abnormality assay
To analyze chromosome instability, 20 cells at metaphase per sample were counted after staining. For karyotyping, at least 10 more cells at metaphase were analyzed and images were captured. GTG-(Giemsa-Trypsin)-banding with 525 bands of resolution was performed by Samkwang medical laboratories (www.smlab.co.kr/).

In vivo tumorigenicity assay
To evaluate in vivo tumorigenicity, MSC, RSV-MSC and NKM-74 cells (5 × 10 6 ) were suspended in 200 µL of PBS containing 20% Matrigel (Sigma), and subcutaneously injected into Balb/C/nu/nu mice (Orient Bio, Inc., Seongnam, Korea). One or five months after implantation, tumor dimensions were determined with a caliper, and the volumes were calculated using the following formula: Tumor volume = π/6 × length × width × height [19]. Length represents the largest tumor diameter, and width represents the perpendicular tumor diameter.

Statistical analyses
To determine the differences between two groups, Student's t-tests were performed. For more than two groups, one-way analysis of variance was performed if normality tests passed, followed by Tukey's multiple comparison tests for all pairs of groups. The data are presented as the mean ± standard deviation. All experiments were conducted in triplicate. GraphPad Prism software (version 6.0; GraphPad, La Jolla, CA, USA) was used for statistical analysis. Values of p < 0.05 were considered statistically significant.

RSV allows mass production of MSCs with effective cell doses (ECD)
The marrow mixture was isolated and seeded onto a culture plate to allow the MSCs to attach the plastic culture dish. Attached cells were divided into groups treated with 0.05% vehicle or 1 μM RSV and cultured until vehicle-treated MSCs reached an ECD (1 x 10 7 cells) (Fig. 1A). There have been many clinical trials using MSCs to treat small or critical bone defects [20]. A minimum number of cells is required to obtain a successful therapeutic effect following MSC implantation. To date, there is no standard ECD for treating bone defects, but a minimum of 1 x 10 7 cells appears to be required for treating bone defects (See supplementary Table 1). Thus, we grew control MSCs treated with vehicle to 1 x 10 7 cells, and then compared the obtained cell amounts with RSV-treated MSCs. In the absence of RSV, the cell number reached more than 1 × 10 7 at about 6 weeks, whereas in the presence of RSV, the cell number was approximately 1 x 10 7 at about 4 weeks. Indeed, at 6 weeks, the number of MSCs cultured in the presence of RSV (5.82 ± 0.92 × 10 7 ) was approximately 4-fold higher than that of MSCs cultured in the absence of RSV (1.31 ± 0.29 × 10 7 ) (Fig. 1B, supplementary Table  2). We designated these cells as 6w-MSCs (MSCs with an effective cell dose for bone regeneration) and used the cells for further experiments. It took about 2 weeks to obtain a sufficient number of young MSCs for further experiments. Although these young MSCs were not suitable for in vivo experiments due to the limited number of cells, they were sufficient for further in vitro experiments. Therefore, we also designated them as 2w-MSCs (not effective dose for in vivo bone regeneration, but appropriate cell number for in vitro test), and used the cells as control groups of 6w-MSCs. We next compared the degree of cellular senescence between vehicle-and RSV-treated 6w-MSCs by SA-β-galactosidase (SA-β-gal) staining. RSV alleviated the level of cellular senescence compared to that of its counterpart (Fig. 1C). In addition, the proportion of bromodeoxyuridine-positive cells was higher in 6w-MSCs with RSV induction than in its counterpart (Fig. 1D). Moreover, the cell populations of control 6w-MSCs in S phase were smaller than those of 6w-MSCs with RSV induction (Fig. 1E). The expression of cell cycle-related genes CCNA, CCND, CDK2, and CCNE, which are associated with S phase [21,22], was higher in 6w-MSCs with RSV induction compared to in control 6w-MSCs (Supplementary Fig. 1). However, the results of SA-β-gal assay and cell cycle analysis cannot fully explain the higher rate of growth by RSV treatment in MSC culture. To address this phenomenon more clearly, we checked the protein level of CASPASE-3, To confirm acetylated-lysine and SOX2, each protein was immunoprecipitated using antibodies targeting each protein followed by western blotting using an anti-SOX2 antibody (n = 3, in triplicate per donor). All study groups were treated with M G132 (10 µM), a proteasome inhibitor. M SCs of 4th passage were used for immunoprecipitation. (E) Immunocytochemistry was conducted to observe cellular localization of SOX2 from the nucleus to the cytoplasm, following SIRT1 knockdown in the presence of RSV (n = 3, in triplicate per donor). To inhibit proteasomal degradation of SOX2, all study groups were treated with M G132 (10 µM ). Nucleus was stained with DAPI, and images were captured by confocal microscopy. Red arrowhead indicates nuclear exports of SOX2 protein. Scale bar = 50 μm. M SCs of 4th passage were used for immunocytochemistry. (F) Efficiency of SIRT1 knockdown or SOX2 overexpression was confirmed by western blot analysis in M SCs (n = 3, in triplicate per donor). M SCs of 4th passage were used for this western blot. (G) Colony-forming cells were detected by CV staining, and (H) the number was counted by three observers (lower panel) (n = 3, in triplicate per donor). *, p < 0.05. Alizarin red S (I) and oil red O (J) staining was conducted to compare mineralization and accumulation of lipid droplets. (K) Quantitative analysis for mineralization was measured at 595 nm absorbance (lower panel). *, p < 0.05. Lipid droplets were quantified with ImageJ software (n = 3, in triplicate per donor). Scale bar = 60 µm. *, p < 0.05. M SCs of 12th passage were used for CV, alizarin red S, and oil red O stains. which is known as a family of endoproteases that provide critical links in cell regulatory networks controlling the cell cycle, cell proliferation, and anti-apoptotic capacity of MSCs [23]. We also checked the protein levels of markers that are related to the cell cycle and potential cellular senescence, such as P53, P21, and P16 [24]. In our results, RSV treatment prevented 6w-MSCs from cleavage of CASPASE-3 protein as well as accumulation of cell cycle inhibitors (Fig. 1F, 1G), indicating that there were more proliferating and healthy cells in 6w-MSCs with RSV induction than in control 6w-MSCs.

6w-MSC with RSV induction retain self-renewal and multipotency over multiple passages
The colony-forming unit fibroblast assay showed that 6w-MSCs with RSV induction retained their colony-forming abilities compared to control 6w-MSCs ( Fig. 2A). To compare multipotency, we induced osteogenic and adipogenic differentiation of MSCs. As expected, 6w-MSCs with RSV induction maintained their potential for osteogenic and adipogenic differentiation. Alizarin red S staining showed that osteogenic differentiation was significantly more pronounced in 6w-MSCs with RSV induction compared to in control 6w-MSCs (Fig. 2B). The mRNA levels of RUNX2 and Collagen Type I Alpha 1 (COL1A1) were also highly expressed in 6w-MSCs treated with RSV (Fig. 2C). The 6w-MSCs with RSV induction also maintained their ability to differentiate into the adipogenic lineage. Oil red O staining showed that 6w-MSCs treated with RSV formed lipid droplets more frequently than control 6w-MSCs upon adipogenic induction (Fig. 2D). The mRNA levels of PPARγ and Adiponectin (APN) were also highly expressed in 6w-MSCs treated with RSV (Fig. 2E). To test a possibility that SOX2 overexpression solely regulated the selfrenewal and multipotency of 6w-MSCs independently of RSV-mediated signaling, we compared the effect of single SOX2 overexpression with its counterpart (pEGFP-C1 vector only) with or without RSV. SOX2 overexpression enhanced the proliferative capacity of vector control MSCs, and RSV synergistically increased the proliferative capacity of MSCs overexpressing SOX2 (supplementary Fig. 2A and 2B). Likewise, RSV enhanced the osteogenic and adipogenic potentials of 6w-MSCs overexpressing SOX2 (supplementary Fig. 2C-2F), which implies that RSV and SOX2 may have a synergistic effect in MSC self-renewal and multipotency. These data suggest that 6w-MSCs with RSV induction retain their self-renewal capacity and multipotency during ex vivo expansion over multiple passages.

SIRT1 and SOX2 are required for 6w-MSCs with RSV induction to maintain stemness
We previously reported that SIRT1 positively regulated SOX2 to maintain the stemness of early passage-MSCs and RSV-induced enhancement of MSC proliferation and differentiation was largely dependent on the presence of SOX2 [16]. However, we did not determine whether RSVmediated SOX2 overexpression depends on SIRT1 in the study (Fig. 3A). To clarify whether RSV-mediated SOX2 activation requires SIRT1, SIRT1 was knocked down in RSV-treated MSCs and then western blot analysis was performed. The results showed that the SOX2 protein level increased by RSV treatment was significantly reduced in the SIRT1 RNAi group, while the levels of OCT4 and NANOG, known to be important regulators for maintaining MSC stemness, were not altered by RSV treatment or SIRT1 RNAi (Fig. 3B, 3C). SIRT1 physically interacts with SOX2 protein to deacetylate and prevent nuclear export of the protein [16]. This mechanism is wellconserved in mouse somatic reprogramming [25]. We confirmed that RSV induced deacetylation of SOX2 and SIRT1 RNAi abolished the RSV-mediated effect on SOX2 protein (Fig. 3D). Additionally, SIRT1 RNAi induced nuclear export of SOX2 under RSV stimulation (Fig. 3E). To confirm whether RSV-mediated SOX2 activation requires SIRT1 as an intermediator and is important for enhancing MSC stemness, we used the pEGFP-C1 SOX2 vector to express SOX2-GFP in the presence of RSV and SIRT1 RNAi (Fig. 3F). As a result, RSV enhanced MSC self-renewal and multi-potency, while SIRT1 RNAi completely abolished the RSV effects ( Fig. 3G-3K). As expected, SOX2 overexpression rescued the effects of SIRT1 RNAi in the presence of RSV (Fig. 3G-3K). These results indicate that SIRT1 acts as an intermediator of RSV-mediated SOX2 activation, and the SIRT1-SOX2 axis is critical for the RSV-mediated benefit in maintaining MSC stemness.

RSV induction preserves nuclear levels of SIRT1 and SOX2 proteins during 6w-MSC proliferation and differentiation
SOX2 plays a critical role in the improvement of cell proliferation and multipotency, and preserving SOX2 expression in MSCs can be a useful method for maintaining MSC population with stemness for long-term culture periods [26][27][28][29]. To confirm whether RSVmediated maintenance of the self-renewal and multipotency of 6w-MSCs are due to the preservation of SIRT1-mediated SOX2 expression, we performed immunocytochemistry against SIRT1 and SOX2 during the proliferation and differentiation of 6w-MSCs. The 6w-MSCs were maintained in vehicle-or RSV-containing basal growth medium for 10 days, and then immunostaining against SIRT1 and SOX2 was performed to compare the levels of both proteins. Our results showed that in 6w-MSCs grown with vehicle-containing medium, the levels of SIRT1 and SOX2 were much weaker than in those grown with RSV-containing medium on Day 1, and both proteins were almost not detected from Day 4 to Day 10. In 6w-MSCs grown with RSV-containing medium, the levels of SIRT1 and SOX2 were much higher than those grown with vehicle-containing medium on Day 1, and although the levels gradually decreased over 10 days, both protein levels were still maintained from Day 4 to Day 10 (Fig. 4A). Likewise, the levels of SIRT1 and SOX2 were maintained higher during osteogenic and adipogenic differentiation when 6w-MSCs were cultured in RSV-containing differentiation media (Fig. 4B, 4C). These results indicate that MSCs with high differentiation potential can be obtained for clinical purposes, since RSV treatment allows MSCs to maintain high levels of SIRT1 and SOX2 during long-term in vitro cultivation.

RSV induction enhances 6w-MSC-driven bone regeneration
To address the clinical potential of RSV in MSC-driven bone regeneration, we investigated the regenerative potential of 6w-MSCs transplanted in a rat calvarial defect model. Critical size calvarial defects (8 mm diameter) were created and covered with fibrin glue mixed with the cells. At 4 weeks after implantation, micro-computed tomography analysis showed that bone regeneration was slightly increased in both 6w-MSCs with or without RSV induction (Supplementary Fig. 3). Eight weeks post-surgery, 6w-MSCs with RSV induction showed dramatically enhanced bone regeneration displayed by a better bone volume filled with mineralized bone compared to vehicle-treated MSCs (Fig. 5A, 5B). At 8 weeks post-transplantation, the bone bridge nearly filled with the defect in 6w-MSCs with RSV induction, whereas partially regenerated bone was observed in 6w -MSCs treated with vehicle (Fig. 5C). Immune response by allogenic cell transplantation to the recipient can be an important issue in clinical trials. Bone marrow-derived MSCs are known to have immune suppressive effects.
Aging and Disease • Volume 10, Number 4, August 2019 Some reports have shown that MSCs suppress alloreactive T cells depending on cell doses [30], and do not elicit lymphocyte proliferation response [31]. In addition, MSCs can modulate immune responses at multiple levels by inhibiting the generation and function of monocyte-derived dendritic cells [32]. Recent studies showed no obvious immune response and inflammatory cells in animal tissues grafted with human MSCs [33,34], thereby inducing successful tissue regeneration by MSCs without inflammation. In the current study, we also did not observe immune rejection responses, such as macrophage production, in rat calvarial tissues xenografted by human 6w-MSCs. Our results are in line with those of previous studies, indicating that MSCs have immune-modulatory effects in xenograft models. To confirm whether newly regenerated bone is derived from MSCs implanted in vivo, we performed immunehistochemistry using an antibody against human-specific vimentin, a marker of mesenchymal-derived cells because transplanted MSCs were of human origin [35]. Positive staining for human vimentin revealed that newly regenerated bone originated in implanted human MSCs (Fig. 5D). To explain this discrepancy on why 6w-MSCs treated with RSV had a greater bone regenerative property than that of control 6w-MSCs, we performed immunostaining against each SIRT1, SOX2, RUNX2, and OSTEOCALCIN (OCN), and they were cross-stained with human VIMENIN. RUNX2 is a major transcription factor of osteogenic differentiation, and OCN plays an essential role for the maturation of mineral formation [36]. Our results showed that in 6w-MSCs treated with RSV, human VIMENTIN-positive cells still retained SIRT1 and SOX2. In addition, levels of RUNX2 and OCN in RSVtreated 6w-MSC group were higher than those of control group and were positive to human VIMENTIN (Fig. 5E). Human VIMENTIN was not detected in groups with no human MSCs (supplementary Fig. 4). These results indicate that implanted human 6w-MSCs directly contribute to bone regeneration. To test the safety of longterm culture 6w-MSCs with RSV induction for MSC therapy, we analyzed the surface markers of MSCs, identified by the International Society for Cell Therapy [37], using flow cytometry in the absence or presence of RSV. All 6w-MSCs stably expressed CD90 and CD105, which are positive surface markers of MSCs, whereas the cells were negative for CD34 and CD45 ( Supplementary  Fig. 5), indicating that RSV induction did not alter the surface antigen profiles of MSCs. For in vivo tumorigenicity, 6w-MSCs with or without RSV induction were subcutaneously implanted into nude mice. We also injected MKN-74 cells, which are gastric adenocarcinoma cells, into nude mice as a positive control. As a result, MKN-74 adenocarcinoma cells formed tumors with a volume of 2.8 ± 1.2 cm 3 by 28 days, whereas 6w-MSCs with or without RSV induction did not form visible tumors after 150 days (Fig. 5F, 5G). In the field of stem cell-based regenerative medicine, genetic instability that appears in stem cells during ex vivo manipulation such as in vitro expansion, culture condition, and transgenics [38], thus resulting in tumorigenesis after in vivo implantation of stem cells, is a major limitation to application of this method [39]. Therefore, we tested the chromosomal instability of MSCs in a G-banding karyotyping assay to analyze chromosomal aberrations. Both 6w-MSCs contained the first 22 pairs of chromosomes, known as autosomes, and sex chromosomes (XY). Additionally, there were no structural abnormalities in these chromosomes (Fig. 5F). These results indicate that RSV treatment does not lead to chromosome instability and tumor formation in MSCs, indicating that 6w -MSCs with RSV induction are safe for clinical application. In conclusion, we demonstrated the possibility of using 6w -MSCs with RSV induction for stem cell therapy which showed sufficient efficacy and safety for clinical application. There are several potential benefits to using RSV as an additive to MSC cultures, including the faster acquisition of a sufficient number of cells, easy treatment procedure, high efficacy of differentiation, and no genotoxicity. This method overcomes the major limitations to clinical application of MSCs in tissue regeneration.

DISCUSSION
MSCs are used for skeletal tissue engineering, but they typically lose their stem cell characteristics during in vitro expansion, which is a necessary step to obtain a sufficient number of cells for disease treatment [6,7,40]. For the development of stem cell therapy, it is important for MSCs to maintain their stem cell features during longterm in vitro expansion. In addition, genotoxicity must be considered because it is possible for stem cells maintained in vitro to generate tumors when they are transplanted in vivo [41]. Herein, we report the possibility that MSCs expanded in culture with RSV can be used for stem cell therapy and have sufficient efficacy and safety for clinical application. There are several potential benefits to using RSV as an additive for MSC cultures, including the faster acquisition of a sufficient number of cells, an easy treatment procedure, a high efficacy of differentiation, and no genotoxicity. These results can eliminate major obstacles for the clinical application of MSCs in tissue regeneration. RSV is a natural polyphenol compound that is well-known as a potent SIRT1 activator and has been reported to have beneficial effects against aging and various human diseases including cancer [41,42]. Our previous studies revealed that SIRT1, which functions as an NAD + -dependent lysine deacetylase, prevents SOX2 against degradation by deacetylating it, thus maintaining the stemness of MSCs after extended passaging [16]. Indeed, RSV could retain the self-renewal and multilineage differentiation capacities of human MSCs beyond passage 10 through a SIRT1-dependent pathway [12,16]. Additionally, SIRT1 indirectly regulates the expression of OCT4 and NANOG, which is normally suppressed by P53, by deacetylating P53 in human embryonic stem cells and cancer stem cells [43,44], while RSV was reported to reduce the protein level of P53 through inducing its deacetylation in adult stem cells [45]. In agreement with these reports, Figure 3b showed that the protein level of SOX2 and NANOG were almost entirely retained in RSV-treated MSCs. However, the protein of OCT4 did not significantly change. Previous studies have reported that OCT4 is dispensable for the maintenance of the selfrenewal ability in somatic cells [46] and is independent on maintaining stemness of MSCs during their in vitro expansion [29], suggesting that OCT4 is less likely to be regulated during the long-term passaging of MSCs in culture. Based on these results, RSV treatment can retain the protein abundance of SOX2 through a SIRT1dependent pathway and can also retain that of NANOG via the regulation of P53, but it does not affect the decline in the abundance of OCT4 during long-term passaging.
Although it is a well-known SIRT1 activator, RSV also regulates diverse other proteins and signaling pathways in addition to SIRT1 [47]. In adipocytes, RSV modulates the PI3K/AKT and MAPK/ERK pathways in a SIRT1-independent manner, thus reducing the number and lipid accumulation of the adipocytes [48]. In addition, RSV was reported to induce cancer cell death via the ERK1/2-P53 axis [49]. It has been reported that the PI3K/AKT and ERK signaling pathways upregulate RUNX2 expression, finally resulting in tooth development and regeneration [50], which means that these signaling pathways are also involved in bone regeneration. In other words, the enhanced bone regeneration capacity of 6w-MSCs with RSV induction could be associated with the regulation of other proteins and pathways by RSV in addition to its effect on SIRT1. In the present study, we did not confirm whether RSV treatment regulates other pathways such as PI3K, AKT, and ERK during the osteogenic differentiation of MSCs. However, our previous study revealed that the osteogenic differentiation potential of MSCs treated with RSV was abolished in SIRT1-deficient cells [12]. A recent study suggested that RSV stabilizes SIRT1/peptide interactions in a substratespecific manner [51], which means that RSV can directly bind to SIRT1 protein. In a previous study, we demonstrated that SIRT1 physically interacts with SOX2, and the direct interaction stabilizes SOX2 protein by deacetylating lysine residues in order to block the nuclear export and ubiquitination of SOX2 protein [16].
Collectively, RSV-activated signaling pathway is shown as direct interaction of RSV-SIRT1-SOX2 complex. In current study, we showed that RSV could not enhance the proliferation and differentiation potentials of MSCs without SIRT1 (Fig. 3G-3K). Likewise, our previous study also showed that RSV cannot enhance the proliferation and differentiation potentials of MSCs without SOX2 [16]; therefore, RSV-mediated enhancement of MSC function is shown to be directly involved with SIRT1-SOX2. In this study, we also confirmed the effect of RSV on maintaining multi-lineage differentiation potential of MSCs is mediated by SIRT1-SOX2 axis, suggesting that the effect of RSV on bone regeneration may rely on the presence of SIRT1.
The relationship between self-renewal, osteogenic, and adipogenic capacities of MSCs has been found to be antagonistic and competitive in various MSC-related studies. For this reason, many researchers may believe that enhancing the self-renewal capacity by some regulators or molecules enables suppression of stem cell differentiation, which means that undifferentiated state (no loss of differentiation potentials) of stem cells can be achieved by maintaining the self-renewal capacity of stem cells. Since MSC osteogenesis and adipogenesis are also shown to be competitive in differentiation condition, it can be thought that enhancing the osteogenic capacity of MSCs by some regulators or molecules enables suppression of the adipogenic potential of MSCs. However, various factors or proteins may be involved in the RSV-mediated enhancement of MSC self-renewal and differentiation potentials. RSV can mainly activate SIRT1 to maintain the self-renewal and differentiation capacities in human MSCs [16]. The SIRT1 activated by RSV can affect MSC self-renewal-related regulators, such as SOX2 [16] and FOXO3A [52], and MSC osteogenesis-related regulators, such as RUNX2 [53] and Wnt/β-catenin [54]. Some studies suggested that RSV decreased adipocyte differentiation and lipid accumulation in 3T3-L1 cells preadipocytes [55,56], resulting in loss of body weight [57]. Collectively, it has been shown that RSV positively regulates MSC self-renewal and osteogenesis while negatively regulating MSC adipogenesis. However, the current study showed that RSV enhanced MSC adipogenic potential as well as the self-renewal and osteogenic potentials. These results can be explained by the fact that RSV-activated SIRT1 works in the cells by removing acetyl groups from protein substrates related to aging processes of MSCs in order to delay or prevent cellular aging. Delaying or preventing MSC aging by applying RSV in vitro enables gathering of healthy MSCs, which have self-renewal and multi-differentiation potentials during long-term cultivation. This means that in vitro RSV treatment maintains healthy MSCs over longterm cultivation, thereby resulting in better proliferation and differentiation potentials compared to the cells cultured without RSV. It can also be thought that RSV suppresses in vivo adipogenesis of MSCs if proper materials with RSV can be applied into in vivo models; however, we could not confirm the in vivo effect of RSV on MSC adipogenesis as this study aimed to confirm whether RSV-treated MSCs had higher therapeutic potential when cells were transplanted into in vivo bone defect model. In this study, RSV effect in vivo refers to the bone regenerative capacity of MSCs pre-treated with RSV for long-term duration (6w-MSCs) in a rat calvarial defect model. We wanted to confirm whether bone regenerative potential of MSCs after the in vitro long-term cultivation for obtaining MSCs with effective cell doses could be preserved when the cells were cultured in RSVcontaining culture medium. It is important to gather the number of cells with effective dosage for MSC therapy; however, after passaging during long-term culture, most of the bone marrow-derived MSCs lose their potential to differentiate into multiple lineages as well as self-renewal capacity. Although RSV treatment is known to activate MSC proliferative and differentiation potentials, there is still not enough evidence to confirm whether RSV-treated MSCs are indeed effective in tissue regeneration when transplanted into in vivo disease models. Although RSV was treated with MSC cultures in this study, we believe it is possible for RSV to be injected into some disease models if appropriate materials are available for bone defects. This study provides a possibility that natural small molecules, such as RSV, can be employed in MSC culture in order to enhance therapeutic potential of cells. Therefore, we suggest that using small molecules would offer more benefits than using growth factors to enhance MSC stemness.
Comprehensively, our study demonstrates that RSV treatment can enhance proliferation of MSCs and maintain differentiation potential during long-term expansion in vitro, and transplantation of MSCs pretreated with RSV improves bone regeneration in rat calvarial defect model. Although we transplanted MSCs mixed with fibrin glue for MSC delivery in vivo, examining alternative materials to provide the best environments for cell adherence, differentiation into cartilage, and effective delivery may facilitate more effective tissue regeneration.