Combination of resveratrol and 5‐azacytydine improves osteogenesis of metabolic syndrome mesenchymal stem cells

Abstract Endocrine disorders have become more and more frequently diagnosed in humans and animals. In horses, equine metabolic syndrome (EMS) is characterized by insulin resistance, hyperleptinemia, hyperinsulinemia, inflammation and usually by pathological obesity. Due to an increased inflammatory response in the adipose tissue, cytophysiological properties of adipose derived stem cells (ASC) have been impaired, which strongly limits their therapeutic potential. Excessive accumulation of reactive oxygen species, mitochondria deterioration and accelerated ageing of those cells affect their multipotency and restrict the effectiveness of the differentiation process. In the present study, we have treated ASC isolated from EMS individuals with a combination of 5‐azacytydine (AZA) and resveratrol (RES) in order to reverse their aged phenotype and enhance osteogenic differentiation. Using SEM and confocal microscope, cell morphology, matrix mineralization and mitochondrial dynamics were assessed. Furthermore, we investigated the expression of osteogenic‐related genes with RT‐PCR. We also investigated the role of autophagy during differentiation and silenced PARKIN expression with siRNA. Obtained results indicated that AZA/RES significantly enhanced early osteogenesis of ASC derived from EMS animals. Increased matrix mineralization, RUNX‐2, collagen type I and osteopontin levels were noted. Furthermore, we proved that AZA/RES exerts its beneficial effects by modulating autophagy and mitochondrial dynamics through PARKIN and RUNX‐2 activity.

abundant infiltration of adipose tissue by pro-inflammatory (M1) macrophages and CD4 + T lymphocytes, combined with adipocytes' hypertrophy, induces its dysfunction, characterized by increased IR, hypoxia and enhanced apoptosis. [6][7][8] Furthermore, excessive accumulation of reactive oxygen species (ROS), nitric oxide (NO), protein kinase C activity, with a simultaneous decrease in superoxide dismutase (SOD) activity, which provides antioxidant defence, ultimately leads to the development of cardiovascular diseases in humans and can cause laminitis in horses. [9][10][11] Additionally, a growing body of evidence suggests that in addition to inflammation, excessive oxidative stress (OS), ie ROS generated by mitochondria (MTs), plays a critical role in the development of obesity-related diseases as well as degradation processes. 6,12 Moreover, ectopic accumulation of lipids promotes lipotoxicity, which in turn impairs cellular functions not only of adipocytes, but also of other adipose tissue components, causing IR, apoptosis and inflammation. Microenvironment, combined with OS and inflammation in adipose tissues of EMS horses, is recognized as one of the most important factors that contributes to accelerated senescence and ageing. 1 Both inflammation and progressive ageing of adipose tissue are not without significance for adipose derived stem cells (ASCs) that reside within this tissue.
Adipose-derived mesenchymal stromal stem cells are increasingly often recognized as a therapeutic source of stem cells and recently have been extensively used in veterinary practice. 13 Clinical trials in humans have already been established for the intravenous administration of ASCs in autoimmune and inflammatory disorders, such as multiple sclerosis and arthritis. 14 The growing interest in ASCs' clinical applications results from their unique immunomodulatory and antiinflammatory effects as well as self-renewal potential. ASCs express specific surface markers, including CD90 + , CD105 + and CD44 + , and they do not express CD45 À . Moreover, ASCs have the ability to differentiate into adipocytes, myocytes, chondrocytes and osteoblasts, which underlines their potential utility in future cell-based therapies.
The pro-regenerative properties of ASCs are explained by their paracrine and autocrine activities based on the secretion of membranederived extracellular vesicles (ExMVs), which are known to play a critical role in intracellular signalling. 15,16 ExMVs were demonstrated to contain a broad range of growth factors, including vascular endothelial growth factors, fibroblast growth factors and transforming growth factor-b-all of which are crucial in the treatment of MetS. 17 Moreover, mesenchymal stem cells (MSCs) were shown to improve metabolic control in experimental models of type 2 diabetes (T2D), as measured by enhanced insulin secretion, improved insulin sensitivity and increased number of islet cells in the pancreas. 18 Therefore, they are a promising tool also in the field of endocrinology. RES is a well-known natural, polyphenolic compound and autophagy activator. What is more, RES has been thoroughly documented as an immunomodulatory, anti-inflammatory and antioxidant agent. 19,21,22 It has been shown that RES has the ability to prolong lifespan, increase insulin sensitivity and protect against age-related disorders, including T2D in mice fed a high-fat diet. 23 Moreover, it was found that RES activated AMPK/PGC-1a signalling, improved mitochondrial biogenesis and dynamics, reduced IGF-1 levels and inhibited apoptosis by stimulating autophagy in a mouse model of diabetic nephropathy. 22 Finally, a study conducted by Shakibaei et al showed that RES promoted osteogenic differentiation in human and mouse MSCs through Sirt1/Runx2 activation. Interesting data regarding the involvement of RES in mediating MT removal have been published recently. 23 It was shown that RES induced Pink and Parkin expression in senescent cardiomyocytes to degrade impaired MT. It has been speculated that the activation of these proteins may be a potential mechanism by which RES reverses the impairment of aged cells. ASC senescence and ageing, which is associated with DNA methylation, can be diminished by the application of DNA methyltransferase (DNMT) inhibitors, such as 5-azacitidine (5-AZA), which inhibits the methylation pattern of specific gene regions, while activating associated genes. 5-AZA is chemically an analogue of cytidine, commonly used in the treatment of acute myelogenous leukaemia. 24 However, it was shown in human-aged ASCs that 5-AZA reversed their aged phenotype, increased proliferative activity and improved osteogenic differentiation potential, while it decreased the accumulation of OS factors and DNA methylation status. 25 Moreover, a study conducted by Seeliger et al 26 revealed that AZA improved metabolic and enzymatic activity in hepatocyte-like cells.
Previous studies have shown that ASCs isolated from EMS horses (ASC EMS ) suffered from reduced proliferative activity, progressive apoptosis, ageing and senescence, which eventually impaired their multipotency and osteogenic differentiation potential. [27][28][29] It was shown that ASCs EMS exhibited reduced secretion of extracellular microvesicels (ExMVs) rich in bone morphogenetic protein 2 (BMP-2), collagen type II (COll-2) or osteocalcin (OCN)-all fundamental for ontogenesis. Deterioration of differentiation potential was caused by dysfunction of mitochondrial metabolism and dynamics. ASCs EMS were characterized by down-regulation of genes related to selective mitophagy and mitochondrial biogenesis, ie PINK, PARKIN, PGC1a and PDK4. Accumulation of damaged MT triggers their segregation from the mitochondrial network and targets these organelles for autophagic degradation in a process that requires Parkin-dependent ubiquitination of mitochondrial proteins. Ex vivo induction of the proper elimination of dysfunctional MT in the course of osteogenic differentiation may be essential for cell survival, extracellular matrix formation and maintenance of "stemness." Therefore, the strategy to search for pharmaceutical intervention that would improve mitochondrial metabolism and dynamics, combined with reduced senescence, apoptosis, OS and ageing, during osteogenic differentiation in ASCs EMS seems to be reasonable. What is more, recent discoveries underlined the requirement for ex vivo ASC EMS "recovery," as it has been demonstrated that allogeneic MSCs may affect the immune response in donor animals, and that allogeneic administration of ASCs is not as safe as previously thought.
The aim of the present study was to investigate whether ex vivo treatment of ASC EMS with a combination of AZA/RES would reverse aged phenotype and promote osteogenic differentiation potential in ASC EMS . For this purpose, mitochondrial metabolism and dynamics, autophagy, OS and senescence of ASC EMS cultured under osteogenic differentiation conditions were investigated. Finally, we analysed protein and mRNA levels of osteogenesis master regulators, including Runx, BMP-2, OCN and osteopontin.

| MATERIALS AND METHODS
All reagents used in this experiment were purchased from Sigma-Aldrich (Poland), unless indicated otherwise.

| Qualification of horses
Animals were age-matched (mixed sex, 9-14 years; mean AE SD, 11.2 AE 1.7 years) and assigned into 2 groups: EMS (n = 5, 2 female, 3 male) and healthy horses (n = 5, 2 female, 3 male). Detailed characterization of animals used in this experiment is shown in Table 1.
Qualification of animals was based on (i) extensive interviews with owners, (ii) measurement of body weight, (iii) estimation of body condition score (BCS) and cresty neck scoring system (CNS), (iv) palpation and visual assessment of the hoof capsule, (v) X-ray examination, (vi) resting insulin levels, (vii) combined glucose-insulin test (CGIT) and (viii) LEP concentration as described previously. 7

| Cell isolation
Approximately, 2 g of subcutaneous adipose tissue was harvested from the horses' tail base following the standard surgical procedure and ethical standards. Next, tissue samples were placed in sterile Hank's balanced salt solution (HBSS) and transported immediately to the laboratory. Cells were isolated under aseptic conditions following the previously described protocol by Marycz et al 29 Briefly, specimens were extensively washed 3 times with HBSS, cut into small pieces and minced. Tissue was then digested in collagenase type I solution (1 mg/mL) for 40 minutes at 37°C. Following digestion, samples were centrifuged (12009g, 10 minutes) and the supernatant was discarded. Remaining cell pellets were re-suspended in culture medium and transferred to a culture flask. In order to perform the experiments, cells were passaged 3 times.

| Cell culture
During the experiment, cells were cultured under aseptic and constant conditions in an incubator (37°C, 5% CO 2 and 95% humidity).
Culture medium consisted of DMEM with 4500 mg/L glucose supplemented with 10% fetal bovine serum (FBS) and 1% of penicillinstreptomycin (PS). Culture medium was changed every 3 days. Cells were passaged after reaching 90% confluence using trypsin solution (TrypLE TM Express; Life Technologies).

| Immunophenotyping and multipotency assay
The expression of the following surface antigens: CD44, CD45,

| Proliferation rate
Growth kinetics of ASCs was examined using a resazurin assay kit (TOX8), following the manufacturer's instructions. To perform the assay, cells were plated in 24-well plates at an initial concentration of 2 9 10 4 per well. Culture media were replaced with media containing 10% of resazurin dye and cells were incubated in a CO 2 incubator, 37°C for 2 hours. Next, supernatants were transferred to 96well plates and subjected to spectrophotometric assay (Epoch; Bio-Tek). The absorbance of the supernatants was measured at a wavelength of 600 nm for resazurin, and 690 nm reference wavelength.

| SEM, TEM and FIB-SEM analysis
Detailed morphology of cells was evaluated using scanning electron microscope (Zeiss EVO LS15). First, cells were fixed in 4% PFA and washed with HBSS 3 times. Next, specimens were dehydrated in a graded ethanol series (from 50% to 100%). Air-dried samples were sputtered with gold (ScanCoat 6; Oxford), placed in a microscope chamber and observed using a SE1 detector, at 10 kV of filament's tension. SEM was equipped with a BRUCKER energy dispersive X- Cells were analysed using FIB-SEM using a previously described method. 27 Briefly, samples were fixed with 2.5% glutaraldehyde in 0.1 mol/L cacodylate buffer and incubated on ice in 2% osmium tetroxide/1.5% potassium ferricyanide for 1 hour. After incubation, cells were dehydrated in graded ethanol series and infiltrated with resin (Agar Low Viscosity Resin Kit, Agar, UK). Resin blocks with recovered cell-containing surfaces were mounted on microscope stubs with a carbon tape, maintaining the horizontal position of the recovered surfaces, and sputtered with 20 nm layer of gold (Leica EM ACE600). Next, specimens were placed in a field-emission cross-beam electron/ion microscope (Zeiss Auriga 60). Prior to focused ion beam (FIB) milling, the areas containing chosen cells were covered with~50 nm protective layer of platinum. The imaging was performed using SE2 detector at 2 kV of electron beam voltage. Using obtained photographs, mitochondrial net was further reconstructed in Imaris Software. with live imaging chamber) and analysed using ImageJ software.

| Confocal imaging
Based on representative photographs, mitochondrial net was visualized in microP software. 30

| Quantification of ALP levels
In order to examine extracellular ALP activity, an Alkaline Phosphatase Colorimetric Assay Kit (Abcam) was used in accordance with the manufacturer's protocol. In the assay, the p-nitrophenyl phosphate was used as a phosphatase substrate. Obtained product was measured at 405 nm wavelength (Epoch; BioTek). The amount of pNP was obtained by sample readings applied to a standard curve.
ALP activity was calculated using the following formula: ALP activity added to well (mL); T-reaction time).

| Oxidative stress factors and senescence
Nitric oxide concentration was assessed using reagent kit (Life Technologies) while SOD activity was measured using a SOD Assay kit

| siRNA transfection
To knock down PARKIN expression in ASC EMS , a PARKIN small interfering RNA (siRNA) construct was purchased and tested for knockdown efficiency. ASC were seeded at an initial density of  Technologies). After 72 hours, total RNA was isolated from cells using TRI Reagent.

| Reverse-transcription polymerase chain reaction (RT-PCR)
In order to isolate total RNA, cells were homogenized with 1 mL of TRI Reagent. Total RNA was isolated according to method described by Chomczynski and Sacchi. 31 The quantity and quality of received genetic material were estimated using a nanospectrophotometer

| Co-culture of ASC with RAW 264.7
RAW 264.7 were counted, adjusted to a density of 1 9 10 6 cells/ mL and seeded onto a 24-well plate in a volume of 500 lL/well. was added to culture wells. After 24 hours of co-culture, media were collected for analysis of the macrophages' secretory activity, and the cells were lysed by adding TRI Reagent.

| Statistical analysis
All experiments were performed at least in 3 replicates. Differences between experimental groups were estimated using unpaired student's t test. Statistical analysis was conducted with GraphPad Prism 5 Software (La Jolla, USA). Differences with probability of P < .05 were considered significant. Statistical significance is indicated as asterisk (*) when comparing the result to ASC CTRL , and as hashtag (#) when comparing to ASC EMS .

| Immunophenotyping and multipotency assay
In order to characterize surface marker expression, cells were anal- No differences were noted between ASC EMS and ASC EMS AZA/RES .

However, incorporation of BrdU in ASC EMS at D1 decreased in com-
parison to ASC CTRL while increased at D3 and D5 ( Figure 2E, P < .01 and P < .05, respectively).

| Assessment of bone mineralization
Using SEM with EDX and CzBSD detectors, we evaluated the effectiveness of mineralized matrix formation ( Figure 4A). CzBSD detector allowed for the visualization of organic compounds (brighter signal refers to more organic compounds), whereas EDX was used for the evaluation of Ca and P content. Results obtained as photographs were further quantified. The number of bone nodules was decreased in ASC EMS in comparison to the control group ( Figure 4B, P < .001); however, AZA/RES treatment significantly increased nodules' number (P < .05). Analogous phenomenon was observed in nodules' size ( Figure 4C). Ca and P amounts were significantly down-regulated in ASC EMS ( Figure 4D, P < .001 and P < .01, respectively). However, Ca levels were increased in ASC EMS AZA/RES (P < .01). Notably, AZA/ RES treatment did not affect P concentration. Furthermore, the ratio of Ca and P was calculated ( Figure 4E). Robust accumulation of Ca in comparison to P was observed in ASC EMS (P < .01), while AZA/ RES treatment resulted in a decreased ratio of Ca and P (P < .01).

| Accumulation of oxidative stress factors
In order to evaluate mitochondrial membrane potential (MMP) with a flow cytometer, cells were subjected to analysis after day 1 of differentiation to avoid machine breakdown. Representative plots are shown in Figure 5A. MMP was significantly down-regulated in ASC EMS ( Figure 5B, P < .01), although AZA/RES treatment resulted in increased MMP (P < .001). Antioxidative capacity of cells was estimated by the evaluation of SOD activity. No differences were noted between control and EMS groups, whereas F I G U R E 1 Immunophenotyping and multipotency assay. In order to confirm the multipotent character of isolated cells, they were analysed for CD44, CD45 and CD90 surface antigens' expression with flow cytometer (A). Furthermore, cells were differentiated into chondrogenic, osteogenic and adipogenic lineage (B). Effectiveness of differentiation was confirmed by specific staining ASC EMS AZA/RES was characterized by an enhanced SOD activity ( Figure 5C, P < .01). Accumulation of NO was increased in ASC EMS ( Figure 5D, P < .01), although AZA/RES treatment diminished its levels (P < .01). Analogous phenomenon was observed in the measurement of ROS levels as shown in Figure 5E.

| Senescence and apoptosis
For visualization of senescence in cultures, staining for b-galactosidase was performed ( Figure 6A). All formed bone nodules were characterized by dye accumulation. Furthermore, the greatest number of dead cells was noted in the control group as indicated in Figure 6A. We further examined the expression of apoptotic-related genes during osteogenesis. We found increased mRNA levels of p53 in ASC EMS in comparison to ASC CTRL ( Figure 6B). No differences were noted between ASC EMS and ASC EMS AZA/RES. Expression of p21 was up-regulated in ASC from EMS horses ( Figure 6C, P < .01); however, AZA/RES treatment reduced its levels in comparison to untreated cells (P < .05). Similar phenomenon was noted in the case of caspase-3 ( Figure 6D) and caspase-9 ( Figure 6E) expression.
F I G U R E 2 Proliferation of cells pretreated with AZA/RES in control and osteogenic conditions. Cells were cultured in standard medium supplemented with AZA/RES for 24 h and subjected to alamar blue assay in order to investigate their proliferation rate (A). Untreated cells served as a control group (CTRL and EMS). Furthermore, PDT was assessed as well (B). Proliferation rate (C) PDT (D) and BrdU (E) established for cells cultured in osteogenic medium. Results expressed as mean AE SD. Statistical significance indicated as asterisk (*) when comparing the result to ASC CTRL , and as hashtag (#) when comparing to ASC EMS. #P < .05, ##P < .01, *P < .05, ***P < .001

| Evaluation of ER stress
As it was shown that ER stress is involved in the progression of osteogenesis, we decided to investigate ER ultrastructure and expression of genes related to ER stress pathway. TEM images indicated that ER underwent severe ultrastructural changes in ASC CTRL and ASC EMS AZA/RES. In those groups swelling of the ER lumen was observed. Expression of PERK was significantly increased in EMS group in comparison to healthy, control cells ( Figure 7B, P < .001).
Furthermore, PERK expression was markedly increased after AZA/ RES treatment (P < .001) in comparison to untreated cells. No differences were noted in the expression of CHOP between ASC CTRL and ASC EMS ( Figure 7C). However, its expression was up-regulated in ASC EMS AZA/RES in comparison to untreated cells (P < .05).
F I G U R E 3 Evaluation of osteogenesis effectiveness. After day 5 of osteogenesis, cells were captured using inverted, light microscope (A) and the formation of a mineralized matrix was visualized. Using RT-PCR we analysed the expression of osteogenic-related genes including RUNX 2 (B), collagen I (C) and BMP-2 (D). Secretion of OPN was tested with ELISA assay (E). ALP activity (F) was reduced in both EMS groups as indicated (F). As IL-10 plays a role in osteogenic differentiation, we evaluated its amount with ELISA (G). Furthermore, using RTR-PCR the expression of miR-451 was tested (H). Results expressed as mean AE SD. Statistical significance indicated as asterisk (*) when comparing the result to ASC EMS , and as hashtag (#) when comparing to ASC CTRL. ##P < .01, ###P < .001, *P < .05, **P < .01, ***P < . The obtained results revealed increased MT activity in the control group which supports the data received with JC-1 assay. SQSTM expression was increased in ASC EMS ( Figure 8B, P < .01) and AZA/ RES treatment augmented its expression even more (P < .01). Similarly, the expression of Beclin 3 ( Figure 8C) was also up-regulated in those groups (P < .01 and P < .05, respectively). Interestingly, no differences were noted in the expression of LC3 ( Figure 8D).

LAMP-2 expression was up-regulated in ASC EMS in comparison to
ASC CTRL ; however, no differences were found between ASC EMS and ASC EMS AZA/RES. For that reason, we performed western blot for LAMP-2 as well. Signal intensity was the greatest in ASC EMS AZA/RES ( Figure 8F) .

| Assessment of mitochondrial dynamics
Mitochondrial morphology was evaluated using TEM microscope and Imaris analysis ( Figure 9A Western blot for MFN indicated its increased synthesis in ASC CTRL ( Figure 9E). Furthermore, the shape of mitochondrial net was visualized by MitoRed ( Figure 9F). Obtained photographs were subjected to microP software analysis ( Figure 9G). In accordance with the data generated by the software, ASC EMS and ASC EMS AZA/RES were characterized by decreased MT number per cell (Figure 9H). Moreover, the number of globular MT was diminished in ASC EMS in comparison to ASC CTRL ( Figure 9I, P < .05). Treatment of ASC EMS with AZA/RES decreased globular organelles' number (P < .05). An opposite trend was observed in the amount of tubular MT ( Figure 9J). Their number was increased in ASC EMS in comparison to ASC CTRL (P < .05). However, ASC EMS preconditioned with AZA/RES were characterized by increased tubular organelles in comparison to untreated cells (P < .05). as indicated in Figure 11C. Furthermore, TNF-a secretion was analysed with ELISA assay ( Figure 11D). Increased amount of TNF-a was noted in ASC EMS in comparison to the control group. However, AZA/ RES treatment significantly down-regulated the TNF-a level (P < .001). Using spectrophotometric assay, the amount of NO was investigated; however, no statistically significant differences were noted in the investigated groups ( Figure 11E).

| Epigenetic changes during differentiation
In order to investigate epigenetic alternations in cells, we analysed with flow cytometer 5-mC and histone H3 accumulation. No

| DISCUSSION
The present study has demonstrated that ex vivo treatment of ASC EMS with a combination of AZA/RES improved osteogenic differentiation of these cells in comparison to non-treated cells. F I G U R E 1 0 Evaluation of mitophagy during osteogenic differentiation in control and AZA/RES treated cells. Using TEM-FIB, cells were sliced every 25 nm to visualize the formation of mitophagosomes (A, indicated with red arrows). Using RT-PCR the expression of mitophagy related genes: PINK (B) and PARKIN (C) was investigated. Moreover, we assessed the expression of PGC1-a as a marker for mitochondrial biogenesis (D). Amount of protein PARKIN was visualized with western blot (E). To confirm that increased mitophagy is responsible for the beneficial effects of AZA/RES, using siRNA expression of PARKIN was silenced, and expression of RUNX-2 (F), Coll-1 (G) and OPN (H) was investigated. Results expressed as mean AE SD. Statistical significance indicated as asterisk (*) when comparing the result to ASC EMS , and as hashtag (#) when comparing to ASC CTRL . ##P < .01, ###P < .001, *P < .05, ***P < .001 osteogenic differentiation potential and extracellular matrix formation. 28 Oxidative stress is recognized as a major factor contributing to the deterioration of ASC multipotency, isolated from patients with several conditions, including obesity, diabetes and ageing. 20,21 Our previous research has shown that autophagy in ASC EMS serves as a protective mechanism that allows to maintain cellular homeostasis and stemness. We have confirmed increased mitophagy in ASC EMS treated with AZA/RES using RT-PCR, TEM-FIB and confocal microscopy. In this study, we observed the overexpression of SQSTM1 . TET-2 protein levels were visualized using western blot (D). Moreover, the expression of DNMT-1 (E), TET-2 (F) and TET-3 (G) was tested by RT-PCR analysis. Results expressed as mean AE SD. Statistical significance indicated as asterisk (*) when comparing the result to ASC EMS , and as hashtag (#) when comparing to ASC CTRL . #P < .05, ###P < .001, *P < .05, ***P < .001 humans ASCs, which was associated with increased expression of osteogenic markers. We observed in the current study that the