The Positive Effect of TET2 on the Osteogenic Differentiation of Human Adipose-Derived Mesenchymal Stem Cells

Abstract

This study aims to understand the possible effects of TET2 (ten–eleven translocation 2) on the osteogenic differentiation of human adipose-derived mesenchymal stem cells (ADSCs). The human ADSCs were transfected with TET2 siRNAs. The osteogenesis-related genes were detected by quantitative real-time reverse transcription PCR (qRT-PCR), and the osteogenic differentiation was evaluated by alkaline phosphatase (ALP) staining and Alizarin Red staining. TET2 and 5-hydroxymethylcytosine (5hmC) expressions were determined by western blotting and immunofluorescence staining. Meanwhile, wild-type (WT) and TET2-deficient (TET2^−/−) mice were selected to observe the alteration of biological characteristics in vivo. TET2 was significantly upregulated along with the osteogenic differentiation of human ADSCs. Compared with Blank group, TET2 siRNA-3 group showed apparent reductions in TET2, 5hmC, and osteogenesis-related genes, as well as decreases in mineralized nodules, ALP activity, and cell growth (all p < 0.05). Besides, Tet2^−/− mice had shorter femoral length, lower bone mineral density, and reduced bone volume to total volume (BV/TV) ratio relevant to WT mice; and meanwhile, the percentage of TUNEL-positive chondrocytes increased significantly with the decreased total collagen-positive area, and the distance between two markers of calcein narrowed with declined bone formation rate (BFR) and mineral apposition rate (all p < 0.05). Furthermore, Toluidine Blue staining presented the appreciable decrease of BFR/bone surface (BS) ratio, BFR/BV ratio, osteoblast number over bone perimeter (N.Oc/B.Pm), and osteoblast surface (Ob.S)/BS in Tet2^−/− mice (all p < 0.05). Taken together, TET2 upregulation was observed during the osteogenic differentiation of ADSCs, whereas TET2 inhibition may lead to reductions of osteogenesis-related genes and downexpression of 5hmC, which eventually plays a negative role in osteoporosis.

Introduction

Osteoporosis, a systemic bone disease featured by decreased bone mass and damaged bone microstructures, is greatly affecting the life quality of patients because of its high incidence rate and serious complications (Lee et al., 2016). The osteoporotic fracture would usually lead to the decreases in bone healing and bone regeneration, ultimately resulting in delayed healing, bone nonunion, and bone defect, which is a difficult medical problem puzzling the clinical work (Kolios et al., 2010). According to the findings by some previous studies, bone transplantation does have certain effects on repairing bone defect caused by osteoporosis, but this method still has some great limitations (Cao et al., 2012; Diwan et al., 2013).

As a prominent strategy in regenerative medicine, tissue engineering has become one of the most promising therapy for the bone defect repairing in the recent years (Keeney et al., 2016; Lappalainen et al., 2016), whereas adipose-derived mesenchymal stem cells (ADSCs), the multipotent cells with high self-renewal ability routinely isolated from adipose tissue, hold great promise in potential treatments for tissue regeneration (Narai et al., 2015; Zidi and Allaire, 2015). As a matter of fact, ADSCs can be induced to differentiate into osteoblasts under the stimulation of certain culture conditions (Liu et al., 2011). Owing to the features of abundant sources, simple sampling and multidirectional differentiation, ADSCs have become the ideal seed cells for bone tissue engineering (Zuk et al., 2001).

Therefore, knowing the molecular action of ADSC osteogenic differentiation provides a better understanding of the pathogenesis of bone-related diseases, such as osteoporosis, and may lead to the development of new strategies for therapies (Li et al., 2015).

It has been revealed that the biological processes of ADSCs would be controlled by various genes and epigenetic regulations (Ge et al., 2014). DNA methylation is an important epigenetic regulator, especially in the form of 5-methylcytosine (5mC), which has a critical role to play in the regulation of gene expression (Dan and Chen, 2016; Marinus and Casadesus, 2009). Of note, ten–eleven translocation (TET) protein can promote the conversion of 5mC to 5-hydroxymethylcytosine (5hmC), thereby regulating genome transcription and facilitating the normal development of mammals (Ko et al., 2013). In terms of TET2, a member of the TET family, exerts vital effects on numerous pathophysiological processes, including malignant tumors, neurodegenerative diseases, and aging (Mi et al., 2015; Zhang et al., 2014).

There was recent evidence demonstrating a close relation between TET2 and cell differentiation (Hon et al., 2014). For example, Chu et al. (2018) reported that mice with TET2 deletion led to the mild osteoporosis, which could regulate osteoclast differentiation by altering the expression of certain related genes. Besides, TET2 deletion could do harm to the self-renewal and differentiation of bone marrow mesenchymal stem cells, and eventually result in the significant reduction of bone mass, as reported in the study by Yang et al. (2018). More importantly, TET2 loss could also exert regulatory effect on the proliferation and osteogenic differentiation of bone marrow mesenchymal stromal cells (BMSCs) by negatively regulating the expression of 5mC (Li et al., 2018).

A previous finding also indicates that TET2 may play an important role during osteoblast differentiation (Silva et al., 2019). But the importance of TET2 in ADSCs has not been delineated. Therefore, we put forward a hypothesis that there may be some significant correlation between TET2 and the osteogenic differentiation of ADSCs. After isolating and culturing the human ADSCs, we investigated whether and how TET2 can affect the osteogenic differentiation of human ADSCs, with the purpose of providing an alternative perspective for the clinical treatment of osteoporosis-related bone defect.

Materials and Methods

Ethics statement

This study was approved by the Ethics Committee of Jining No. 1 People's Hospital. All animal experiments in this article were approved by the Ethics Committee of Laboratory Animals and in accordance with the Guide for the Care and Use of Laboratory Animals published by National Institute of Health (NIH) (Bayne, 1996). In addition, all tissue samples were collected from subjects after they signed the informed consent form before the study.

ADSCs isolation and culture

Adipose tissues were obtained from healthy female donators (with the age range of 25–35 years) without malignant tumors or metabolic diseases who underwent selective liposuction. After careful removal of fibrous tissues and blood vessels under sterile conditions, 10 g adipose tissue was obtained and cleaned with sterile phosphate-buffered saline (PBS) to remove contaminated debris and red blood cells. Then, the human ADSCs were isolated from adipose tissue samples and resuspended in 12 mL regular culture medium at a cell density of 2 × 10⁶ cells/L, which was placed in a 75-cm²culture flask until cells were subcultured to the third generation for subsequent experiments.

Next, ADSCs were inoculated into a six-well plate by 2 × 10⁵ cells/cm², and high-glucose Dulbecco's modified Eagle's medium (H-DMEM) was added for culturing. When cell confluence reached ∼80%, the osteogenic induction medium containing DMEM, 10% (v/v) fetal bovine serum, 1% (v/v) antibiotics, 100 nM dexamethasone, 10 mM β-glycerophosphate, and 0.2 mM l-ascorbic acid (all purchased from Sigma-Aldrich), was used to further induce differentiation, and changed once every 3 days. The cell morphological changes and late extracellular matrix (ECM) calcification were observed under an inverted microscope.

Transfection of siRNA

We performed siRNA knockdown experiments of the TET2 gene. Human ADSCs were transfected with three different predesigned siRNA oligonucleotides (TET2 siRNA-1, TET2 siRNA-2, TET2 siRNA-3) targeting TET2 or a nontargeting siRNA (control siRNA). The TET2 siRNAs and the nontargeting siRNA were all purchased from Dharmacon. Before performing transfection according to the instructions on the Lipofectamine™ 2000 Kit (Invitrogen), the density of human ADSCs was adjusted to 5 × 10⁵ cells/mL. After transfection of ADSCs lasted 24 hours, the transfection working solution was removed and then cells were washed with PBS buffer three times. The culture medium was changed into osteogenic induction medium for differentiation. Besides, the nontransfected cells were used as the Blank group.

Quantitative real-time reverse transcription PCR (qRT-PCR)

The cellular total RNA was extracted with TRIzol (Invitrogen) and the concentration of RNA was detected with a UV spectrophotometer. Then, RNA was reversely transcribed into cDNA using the PrimeScript RT Kit (RR014A; TaKaRa Biomedical Technology Co., Ltd., Beijing, China). An appropriate amount of cDNA was used as the template for polymerase chain reaction (PCR), and primers needed were designed with the software Primer5.0 (Table 1) and then synthesized by GenScript Biotech Corp. (Nanjing, China).

Table 1.

Sequences of Primers Used in Quantitative Real-Time Reverse Transcription PCR (qRT-PCR) in the Current Study

Gene		Sequence
ALP	Forward primer	5′-CCACGTCTTCACATTTGGTG-3′
ALP	Reverse primer	5′-AGACTGCGCCTGGTAGTTGT-3′
OCN	Forward primer	5′-CACTCCTCGCCCTATTGGC-3′
OCN	Reverse primer	5′-GCCTGGGTCTCTTCACTACCT-3′
Runx2	Forward primer	5′-AGGGACTATGGCGTCAAACA-3′
Runx2	Reverse primer	5′-GGCTCACGTCGCTCATCTT-3′
Col1	Forward primer	5′-AGACTGGCAACCTCAAGAAGT-3′
Col1	Reverse primer	5′-GATGGAGGGAGTTTACACGAAG-3′
OSX	Forward primer	5′-GGCACAAAGAAGCCATACA-3′
OSX	Reverse primer	5′-TAGGCAGGCAGTCAGAAGAG-3′
OPN	Forward primer	5′-GGAGGCCTCGGTGGACATTA-3′
OPN	Reverse primer	5′-CACTCCAATCGTCCCTACA-3′
TET1	Forward primer	5′-GAGCCTGTTCCTCGATGTGG-3′
TET1	Reverse primer	5′-CAAACCCACCTGAGGCTGTT-3′
TET2	Forward primer	5′-TGTTGTTGTCAGGGTGAGAATC-3′
TET2	Reverse primer	5′-TCTTGCTTCTGGCAAACTTACA-3′
GAPDH	Forward primer	5′-CTCAACTACATGGTCTACATGT-3′
GAPDH	Reverse primer	5′-CTTCCCATTCTCAGCCTTGACT-3′

The reaction of qRT-PCR was performed by following the procedures on the PCR Kit (KR011A1; Tiangen Biotech Co., Ltd., Beijing, China). The expression of mRNA was calculated by using 2^−ΔΔCt method and ΔCt = Ct_{target gene} − Ct_{internal reference gene}; ΔΔCt = ΔCt _{experiment group} − ΔCt _{control group} (Ayuk et al., 2016). The relative expression level of target genes was normalized to the expression of GAPDH.

Western blotting

Cells were washed with PBS three times, the whole protein lysate was added, and cells were split on ice for 10 minutes to obtain total proteins. Bicinchoninic acid was used for protein quantification and SDS-PAGE (sodium dodecyl sulfate–polyacrylamide gel electrophoresis) was used to transfer proteins to the nitrocellulose membrane. The proteins on the membrane were blocked for 60 minutes in 5% bovine serum albumin/tris-buffered saline with Tween-20 (TBST). Next, the membrane was incubated with primary antibodies TET1 (ab191698; Abcam), TET2 (ab124297; Abcam), and β-actin (ab8226; Abcam) at 4°C overnight.

The following day, horseradish peroxidase-labeled secondary antibodies were added for 2 hours of incubation at room temperature. Subsequently, the blots were visualized with an enhanced chemiluminescence reagent (ECL detection kit; Amersham Pharmacia Biotech). For graph analysis, ImageJ software was used to analyze the gray value of bands. β-Actin was used as a loading control to quantify the relative protein level.

Alkaline phosphatase staining and Alizarin Red staining

The Alkaline Phosphatase (ALP) Kit (Institute of Hematology, Chinese Academy of Medical Sciences) was utilized to perform ALP staining in strict compliance with the manufacturer's instructions. The activity of ALP was measured using the ALP Detection Kit (Nanjing Jiancheng Bioengineering Institute). Alizarin Red staining (all kits purchased from Nanjing Jiancheng Bioengineering Institute) was conducted as follows: The culture dish was rinsed with PBS three times, ethanol (95% v/v) was used for 10 minutes of fixation, and deionized water was used to rinse the dish three times. Then, the cells were treated with 0.1% Alizarin Red S (prepared with Tris-HCl, pH 8.3) for 30 minutes at 37°C, and positive cells were stained orange red. Distilled water was used two times to rinse and glycerol was used for sealing. A microscope was used for observation and picture taking.

Cell growth curve

The human ADSCs in each transfection group were digested and suspended in culture medium, which were seeded into a six-well plate by 1 × 10⁵ cell/well. At the 24 hours (first day), second day, third day, and fourth day after inoculation, cells were collected and counted to draw the cell growth curve.

Immunofluorescence staining

The culture medium was removed, cells were washed with PBS buffer three times (5 min/time), and fixed in 4% paraformaldehyde for 15 minutes at 37°C. Next, cells were washed again with PBS three times, and treated for 10 minutes in precooled methanol, which was placed in a refrigerator at −20°C for pretreatment in advance.

Thereafter, cells were incubated with primary antibody 5hmC (1:500, ab214728; Abcam) overnight at 4°C, then incubated in the presence of secondary antibody for 60 minutes at room temperature. The two antibodies were diluted 1:800 for AlexaFluor 594 and 1:400 for AlexaFluor 488, respectively. Finally, cells were washed again with PBS buffer and sealed with 50 glycerol per PBS, which were prepared for subsequent observation under a Nikon fluorescence microscope. Images were acquired and analyzed with the software Image ProPlus Version 4.0 and fluorescence intensity was measured.

Experimental animals

The TET2 knockout (TET2^−/−) mice were generated according to protocols reported previously (Li et al., 2011). Wild-type (WT) mice and TET2^−/− mice were anesthetized and killed by cervical vertebra dislocation and long bones (femur and tibia) were taken and fixed in 75% ethanol for X-ray and microcomputed tomography (CT) scanning. The two-dimensional images were used for three-dimensional reconstruction with the software Reconstruction Utility and the ratio of bone volume to total volume (BV/TV) ratio was calculated. Next, the bilateral femur and tibia were separated and fixed in periodate/lysine/paraformaldehyde (PLP) for 24 hours, followed by 2 weeks of decalcification in EDTA, dehydration with gradient alcohol, hyalinization in xylene, embedment with paraffin, and slicing of transverse sections (5 μm in thickness).

Sections of each group were used for total collagen staining by incubating with prepared supersaturated picric acid staining solution for 1 hour, followed by 1 minute of counterstaining with Hematoxylin, 10 seconds of differentiation in 1% hydrochloric ethanol, dehydration with gradient alcohol, hyalinization in xylene, and section mounting with neutral resin. At last, sections were observed under a fluorescence microscope (Olympus, Japan) and photographs were taken with DP70 CCD (Olympus). The quantitative analysis of images was conducted by using the software Northern Eclipse (Empix Imaging). Accordingly, TUNEL staining was performed based according to the manufacturer's instructions on the TUNEL Kit.

Calcein-AM/PI double staining and Toluidine blue staining

On the first day, mice were given calcein (10 μg/g body weight) by intraperitoneal injection. On the fifth day, those mice were given the second shot of calcein. On the seventh day, mice were killed by cervical dislocation and bilateral femur and tibia were separated for 24 hours of fixation in PLP. Hard tissue embedment and super-thin section slicing were performed. Sections were observed under a fluorescence microscope (Olympus) and photographed with DP70 CCD (Olympus). The images were quantified by using the software Northern Eclipse. Quantitative parameters included mineral apposition rate (MAR), which was calculated according to the formula: MAR = The widths of two calcein deposition lines/Interval of days between two injections (Shono et al., 2010). Sections were then stained with Toluidine Blue and analyzed with the software OsteoMeasure (OsteoMetrics, Inc.).

Statistical methods

Data were analyzed by using the statistical software SPSS 21.0. Measurement data were presented by mean ± standard deviation. The comparison among multiple groups was conducted by using one-way ANOVA and the intergroup comparison (post hoc test) among multiple groups was analyzed with Turkey's test. The difference between two groups was tested by Student's t-test. p < 0.05 indicated the statistical significance of differences.

Results

TET1 and TET2 expressions during osteogenic differentiation of human ADSCs

Under the optical microscope, the human ADSCs were spindle shaped (Fig. 1A) and confirmed to be the osteoblasts after differentiation induction at day 4, as shown by the analysis results of ALP staining (Fig. 1B). Alizarin Red staining demonstrated the existence of matrix mineralization at day 15 (Fig. 1C). Meanwhile, we detected TET1 and TET2 expressions on the 0, 3rd, 5th, 7th, 10th, and 14th day of osteogenic differentiation. qRT-PCR and western blotting analysis revealed that TET1 mRNA and protein levels were relatively unchanged during ADSCs differentiation. Meanwhile, TET2 mRNA and protein levels significantly increased during the osteogenic differentiation of human ADSCs in a time-dependent manner (all p < 0.05, Fig. 1D–H).

FIG. 1.

TET1 and TET2 expressions during osteogenic differentiation of human ADSCs. (A) The morphology of human ADSCs observed under the microscope; (B) ALP staining of ADSCs was performed at day 4 after differentiation induction; (C) Alizarin Red staining of ADSCs was performed to indicate mineral deposition at day 15 after differentiation induction; scale bars: 100 μm (A–C). (D) The protein expressions of TET1 and TET2 evaluated by western blotting during the osteogenic differentiation of human ADSCs; (E, F) the density of the bands of TET1 (E) and TET2 (F) was analyzed by ImageJ software; β-actin was used as an internal control to quantify the relative protein level; (G, H) qRT-PCR was performed to analyze the mRNA levels of TET1 (G) and TET2 (H) during the osteogenic differentiation of human ADSCs; data were normalized to GAPDH; *p < 0.05, compared with day 0. Results are from three independent experiments (mean ± SD). ADSC, adipose-derived mesenchymal stem cell; ALP, alkaline phosphatase; qRT-PCR, quantitative real-time reverse transcription PCR; SD, standard deviation; TET, ten–eleven translocation. Color images are available online.

Expressions of TET2 and 5hmC in the human ADSCs after transfection

As shown in Figure 2, cells in the TET2 siRNA-1, TET2 siRNA-2, and TET2 siRNA-3 groups had lower expressions of TET2 and 5hmC than Blank group (all p < 0.05), although there was no obvious difference between Control siRNA group and Blank group with respect to TET2 and 5hmC expressions (all p > 0.05). Furthermore, TET2 siRNA-3 showed the most silencing efficiency on human ADSCs among three TET2 siRNAs (p < 0.05). Then, TET2 siRNA-3 was used to transfect human ADSCs for further experiments.

FIG. 2.

The expressions of TET2 and 5hmC in human ADSCs after TET2 siRNAs transfection. (A) TET2 expression in the human ADSCs detected by western blotting; (B) comparison of TET2 protein expressions in the human ADSCs of different transfection group; β-actin was used as an internal control; (C) quantification of fluorescence intensities of 5hmC; (D) immunostaining for 5hmC in the human ADSCs; scale bars: 50 μm; *p < 0.05, compared with the Blank group; ^#p < 0.05, compared with the TET2 siRNA-1 group; ^&p < 0.05, compared with the TET2 siRNA-2 group. Results are from three independent experiments (mean ± SD). 5hmC, 5-hydroxymethylcytosine. Color images are available online.

Expressions of osteogenesis-related genes in human ADSCs after transfection

The technology of qRT-PCR was applied to detect the expression levels of osteogenesis-related genes in human ADSCs of different transfection groups. The results showed that the mRNA levels of ALP, OCN, Runx2, Col1, OSX, and OPN (osteopontin) were significantly decreased in the TET2 siRNA-3 group, but no observable differences were found regarding these osteogenesis-related genes from the Control siRNA group, when compared with the Blank group (all p > 0.05, Fig. 3).

FIG. 3.

The mRNA levels of osteogenesis-related genes (including ALP, OCN, Runx2, Col1, OSX, and OPN) in human ADSCs of each transfection group. *p < 0.05, compared with the Blank group. Results are from three independent experiments (mean ± SD). OPN, osteopontin.

Silencing TET2 inhibits osteogenic differentiation of ADSCs

According to the cell growth curve drawn based on the cell numbers, no differences were shown in the cell growth of human ADSCs from the Control siRNA group and the Blank group (all p > 0.05), whereas the cell growth declined obviously in those cells from the TET2 siRNA-3 group (all p < 0.05, Fig. 4A). As illustrated by Figure 4B–E, the human ADSCs in the TET2 siRNA-3 group exhibited light staining after ALP staining and Alizarin Red staining, with the reduced mineralized nodules and decreased ALP activity relevant to those cells in the Blank group (all p < 0.05). However, no significant difference was observed between Control siRNA group and Blank group in these aspects (all p > 0.05).

FIG. 4.

Silencing TET2 inhibits osteogenic differentiation of ADSCs. (A) The cell growth curve drawn based on the cell numbers on the 0, first, second, third, and fourth day after TET2 siRNA-3 transfection. (B) The osteogenic differentiation of ADSCs detected after ALP staining on day 4; (C) mineral deposition of ADSCs was detected by Alizarin Red staining on day 15 of osteogenic differentiation; (D) comparison of ALP activity in the human ADSCs from each transfection group; (E) comparison of the quantitative results of Alizarin Red staining in the human ADSCs from each transfection group. Scale bars: 200 μm (B, D). *p < 0.05, compared with the Blank group. Results are from three independent experiments (mean ± SD). Color images are available online.

Effects of TET2 deletion on the bone growth in vivo

The X-ray scanning and micro-CT scanning were performed to observe the changes in morphology and bone density in the left femur and tibia of WT mice and Tet2^−/− mice, and the total collagen staining and TUNEL staining were conducted to detect the effect of TET2 on bone growth. As shown in Figure 5, Tet2^−/− mice had shorter femur length, lower bone density, and significantly decreased BV/TV ratio than WT mice (all p < 0.05). TUNEL staining demonstrated that Tet2^−/− mice had significantly higher TUNEL-positive chondrocyte percentage than WT mice (p < 0.05), which suggested that TET2 deletion induced the apoptosis of chondrocyte and inhibited intrachondral osteogenesis. Total collagen staining presented the significant reduction of total collagen-positive area in Tet2^−/− mice (all p < 0.05).

FIG. 5.

Effects of TET2 deletion on the bone growth in vivo. (A, B) The X-ray scanning (A) and micro-CT scanning (B) performed to observe the femur of mice; (C) comparison of BV/TV ratio among different groups of mice; (D, E) the graph and quantitative results of total collagen staining in the upper tibia of mice; scale bars: 1 mm; (F, G) the ratio of apoptotic chondrocytes to TUNEL-positive chondrocytes in cartilage growth plate shown after TUNEL staining; scale bars: 50 μm. Data are presented as mean ± SD (n = 6). BV/TV, bone volume to total volume. Color images are available online.

Effects of TET2 deletion on the osteogenesis of mouse osteoblasts

To investigate whether TET2 deletion-induced osteoporosis was associated with the reduced osteogenesis of osteoblasts, we conducted Calcein-AM/PI double staining and the results were displayed in Figure 6. Compared with WT mice, Tet2^−/− mice had significantly narrowed distance between two fluorescence markers and apparently decreases in bone formation rate (BFR) and MAR (all p < 0.05). According to the results of Toluidine Blue staining of tibia, Tet2^−/− mice had smaller trabecular BV and decreased osteoid formation, and also showed significant reductions in the following indexes, including BFR/bone surface (BS), BFR/BV, osteoblast number over bone perimeter (N.Oc/B.Pm), and osteoblast surface (Ob.S)/BS (all p < 0.05).

FIG. 6.

Effects of TET2 deletion on the osteogenesis of mouse osteoblasts in vivo. (A) The distance between two fluorescence markers (distance between two arrow heads) observed under a fluorescence microscope after Calcein-AM/PI double staining; scale bars: 50 μm; (B) comparison of MAR between WT mice and Tet2^−/− mice; (C) the changes in TBV and osteoid formation revealed after Toluidine Blue staining; scale bars: 1 mm; (D–G) comparison of BFR/BV (D), BFR/BS (E) N.Oc/B.Pm (F) and Ob.S/BS (G) between WT mice and Tet2^−/− mice. Data are presented as mean ± SD (n = 6). BFR, bone formation rate; BS, bone surface; MAR, mineral apposition rate; N.Oc/B.Pm, osteoblast number over bone perimeter; Ob.S, osteoblast surface; TBV, trabecular bone volume; WT, wild-type. Color images are available online.

Discussion

In this study, TET2 expression was found gradually increased during the process of osteogenic differentiation of human ADSCs. Zhong et al. (2017) has pointed out a similar finding as ours that TET2 expression was upregulated during myogenic differentiation, whereas TET2 silencing could reduce myogenic differentiation by downregulating the expression of myogenic genes, such as myogenin, Myf 6, and Myomaker, which indicated the involvement of TET2 in the regulation of osteogenic differentiation of human ADSCs. As far as we know, DNA demethylation is an important process affecting the differentiation of stem cells and it constitutes a mechanism of promoting osteogenic differentiation by inducing the high expression of osteogenic marker genes (Arnsdorf et al., 2010; Chen et al., 2015; Delgado-Calle et al., 2013).

5hmC was known as the intermediate leading to DNA demethylation (Feldmann et al., 2013). Evidence stated that DNA methylation-modified enhancer region has the high content of 5hmC, and TET protein family was involved in the dynamic changes of 5hmC in the enhancer region (Hon et al., 2014). As a member of DNA hydroxylase family, TET proteins have been shown to mediate DNA demethylation process by oxidizing 5mC to 5hmC, thus it has good epigenetic modification characteristics to regulate gene expression and maintain cell surface markers (Breiling and Lyko, 2015; Yang et al., 2013). Collectively, we supposed that TET2 may involve in osteogenic differentiation through regulating DNA demethylation.

Next, we conducted transfection experiments in vitro and found that TET2 siRNA could significantly reduce the expression of osteogenesis-related genes (including ALP, OCN, Runx2, Col1, OSX, and OPN). In the meantime, ALP staining and Alizarin Red staining also suggested the reduction of the human ADSC osteogenic differentiation after silencing TET2. ALP is a specific marker of a mature osteoblast phenotype and it is one of the most common indicators used for the evaluation of the secretory function of osteoblasts (Metavarayuth et al., 2015). Meanwhile, RUNX2 is a transcription factor greatly involved in the regulation of osteoblast development, osteoblast differentiation, and bone formation, and it exerts its regulatory impact by inducing the expression of osteogenic differentiation-ending proteins, including bone sialoprotein, OCN, OSX, and OPN (Lee et al., 2010).

On the other hand, as the protein framework of bone tissues, Coll is of great implication for the maintenance of biomechanical properties and structural integrity of bone tissues, and it is also a useful marker to indicate the differentiation of osteoblasts toward matrix maturation (Nyman et al., 2011). Consistent with our results, Dimitrios Cakouros and his team also reported that TET2 and 5hmC expression levels went up substantially during BMSC osteogenic differentiation, and TET2 knockout by siRNA led to the significant decrease of osteogenic potential, appreciable reductions in RUNX2, BMP-2, OPN, and OCN, and shrinking number of mineralized nodules (Cakouros et al., 2019).

These findings indicated that TET2 silencing may inhibit osteogenic differentiation of human ADSCs by blocking the transcription of osteogenic genes. Actually, TET2 mutations, including deletion, insert, and frameshift mutation, tend to be accompanied with the significant reduction of total 5hmC, as described previously (Ko et al., 2010). Thus, we detected the expression levels of 5hmC in the human ADSCs from each group, and as a result, we found TET2 siRNA can inhibit both TET2 and 5hmC expressions. During the process of osteogenic differentiation, a large number of promoter regions related to osteogenic differentiation were demethylated to facilitate the gradual acquisition of osteoblast phenotypes by osteoblasts, and these promoter regions included Runx2, OCN, OPN, ALP, and Col1 (Ling et al., 2017; Zhang et al., 2011; Zhou et al., 2009).

Therefore, silencing TET2 may inhibit 5hmC to hinder the demethylation of osteogenic genes and downregulate the expression of osteogenesis-related genes, thus eventually resulting in the inhibition of osteogenic differentiation. Moreover, Van Etten et al. (2011) revealed that TET2 was linked to activation of the Hedgehog in chronic myeloid leukemia, suggesting various possible targets for new therapies. The Hedgehog signaling was reported to be essential in osteoblast generation from preosteoblasts (Marumoto et al., 2017).

In addition, Qu et al. (2018) have shown that TET2 deficiency led to the activation of AKT and ERK and resulted in stem cell factor-dependent clonal expansion of dysfunctional erythroid progenitors. Baroncelli et al. (2019) found that inhibition of PI3K signaling significantly reduced osteoblast viability and adhesion to ECM by decreasing the AKT activity. Considering the above studies, TET2 might also regulate osteogenic differentiation of ADSCs through Hedgehog or AKT pathways. But the potential molecular mechanisms still need to be investigated in future studies.

Furthermore, the animal experiments in vivo were also performed and we found mice with TET2 deletion showed significant decreases in bone density, BV/TV, and collagen area, but apparent increases in chondrocyte percentage, suggesting that TET2 deletion may cause premature osteoporosis. In line with our observation, the study of Chu et al. (2018) also demonstrated the mild osteoporosis of mice with the reduced 5hmC and the alterations of osteogenesis-related genes induced by TET2 loss.

Moreover, the Tet2^−/− mice presented osteoporosis phenotype in the work of Yang et al. (2018) and TET2 deletion could reduce the demethylation of P2rX7 promoters to inhibit the expression of osteogenesis-related genes. Additionally, Tet2^−/− mice in our investigation also showed significant decreases in MAR, BFR/BS, BFR/BV, N.Oc/B.Pm and Ob.S/BS, all of which were indicators to reflect the details of osteoblast-related activity from different aspects (Allen et al., 2006; Ciria-Recasens et al., 2005; Matsushima et al., 2003; Smith et al., 2005), and further verified the findings consistent with our in vitro experiments.

In conclusion, TET2 expression was up-regulated during the osteogenic differentiation of the human ADSCs. However, TET2 deletion could down-regulate the expressions of osteogenesis-related genes to hinder the osteogenic differentiation. Hence, this study may offer some experimental and theoretical basis in human ADSCs for finding a potential therapeutic strategy for patients with osteoporosis.

Footnotes

Acknowledgment

The authors thank all members for their helpful suggestions and comments on this work.

Author Disclosure Statement

The authors declare they have no conflicting financial interests.

Funding Information

The authors received no funding for this study.

References

Allen

M.R.

, Follet

, Khurana

, Sato

, and Burr

D.B.

(2006). Antiremodeling agents influence osteoblast activity differently in modeling and remodeling sites of canine rib. Calcif. Tissue Int. 79, 255–261.

Arnsdorf

E.J.

, Tummala

, Castillo

A.B.

, Zhang

, and Jacobs

C.R.

(2010). The epigenetic mechanism of mechanically induced osteogenic differentiation. J. Biomech. 43, 2881–2886.

Ayuk

S.M.

, Abrahamse

, and Houreld

N.N.

(2016). The role of photobiomodulation on gene expression of cell adhesion molecules in diabetic wounded fibroblasts in vitro. J. Photochem. Photobiol. B, 161, 368–374.

Baroncelli

, Fuhler

G.M.

, van de Peppel

, Zambuzzi

W.F.

, van Leeuwen

J.P.

, van der Eerden

B.C.J.

, and Peppelenbosch

M.P.

(2019). Human mesenchymal stromal cells in adhesion to cell-derived extracellular matrix and titanium: Comparative kinome profile analysis. J. Cell. Physiol. 234, 2984–2996.

Bayne

(1996). Revised Guide for the Care and Use of Laboratory Animals available. American Physiological Society. Physiologist, 199, 208–211.

Breiling

, and Lyko

(2015). Epigenetic regulatory functions of DNA modifications: 5-methylcytosine and beyond. Epigenetics Chromatin 8, 24.

Cakouros

, Hemming

, Gronthos

, Liu

, Zannettino

, Shi

, and Gronthos

(2019). Specific functions of TET1 and TET2 in regulating mesenchymal cell lineage determination. Epigenetics Chromatin, 12, 3.

Cao

, Liu

, Gan

, Fan

, Yang

, Zhang

, Tang

, and Dai

(2012). The use of autologous enriched bone marrow MSCs to enhance osteoporotic bone defect repair in long-term estrogen deficient goats. Biomaterials, 33, 5076–5084.

Chen

J.C.

, Chua

, Bellon

R.B.

, and Jacobs

C.R.

(2015). Epigenetic changes during mechanically induced osteogenic lineage commitment. J. Biomech. Eng. 137, 020902.

10.

Chu

, Zhao

, Sant

D.W.

, Zhu

, Greenblatt

S.M.

, Liu

, Wang

, Cao

, Tho

J.C.

, Chen

, Liu

, Zhang

, Maciejewski

J.P.

, Nimer

, Wang

, Yuan

, Yang

F.C.

, and Xu

(2018). Tet2 regulates osteoclast differentiation by interacting with Runx1 and maintaining genomic 5-hydroxymethylcytosine (5hmC). Genomics Proteomics Bioinformatics, 16, 172–186.

11.

Ciria-Recasens

, Perez-Edo

, Blanch-Rubio

, Marinoso

M.L.

, Benito-Ruiz

, Serrano

, and Carbonell-Abello

(2005). Bone histomorphometry in 22 male patients with normocalciuric idiopathic osteoporosis. Bone, 36, 926–930.

12.

Dan

, and Chen

(2016). Genetic studies on mammalian DNA methyltransferases. Adv. Exp. Med. Biol. 945, 123–150.

13.

Delgado-Calle

, Fernandez

A.F.

, Sainz

, Zarrabeitia

M.T.

, Sanudo

, Garcia-Renedo

, Perez-Nunez

M.I.

, Garcia-Ibarbia

, Fraga

M.F.

, and Riancho

J.A.

(2013). Genome-wide profiling of bone reveals differentially methylated regions in osteoporosis and osteoarthritis. Arthritis Rheum. 65, 197–205.

14.

Diwan

A.D.

, Leong

, Appleyard

, Bhargav

, Fang

Z.M.

, and Wei

(2013). Bone morphogenetic protein-7 accelerates fracture healing in osteoporotic rats. Indian J. Orthop. 47, 540–546.

15.

Feldmann

, Ivanek

, Murr

, Gaidatzis

, Burger

, and Schubeler

(2013). Transcription factor occupancy can mediate active turnover of DNA methylation at regulatory regions. PLoS Genet. 9, e1003994.

16.

, Liu

, Chen

, Zhang

, Lv

, Jin

, Jiang

, Shi

, and Zhou

(2014). The epigenetic promotion of osteogenic differentiation of human adipose-derived stem cells by the genetic and chemical blockade of histone demethylase LSD1. Biomaterials, 35, 6015–6025.

17.

Hon

G.C.

, Song

C.X.

, Du

, Jin

, Selvaraj

, Lee

A.Y.

, Yen

C.A.

, Ye

, Mao

S.Q.

, Wang

B.A.

, Kuan

, Edsall

L.E.

, Zhao

B.S.

, Xu

G.L.

, He

, and Ren

(2014). 5mC oxidation by Tet2 modulates enhancer activity and timing of transcriptome reprogramming during differentiation. Mol. Cell, 56, 286–297.

18.

Keeney

, Chung

M.T.

, Zielins

E.R.

, Paik

K.J.

, McArdle

, Morrison

S.D.

, Ransom

R.C.

, Barbhaiya

, Atashroo

, Jacobson

, Zare

R.N.

, Longaker

M.T.

, Wan

D.C.

, and Yang

(2016). Scaffold-mediated BMP-2 minicircle DNA delivery accelerated bone repair in a mouse critical-size calvarial defect model. J. Biomed. Mater. Res. A, 104, 2099–2107.

19.

, An

, Bandukwala

H.S.

, Chavez

, Aijo

, Pastor

W.A.

, Segal

M.F.

, Li

, Koh

K.P.

, Lahdesmaki

, Hogan

P.G.

, Aravind

, and Rao

(2013). Modulation of TET2 expression and 5-methylcytosine oxidation by the CXXC domain protein IDAX. Nature, 497, 122–126.

20.

, Huang

, Jankowska

A.M.

, Pape

U.J.

, Tahiliani

, Bandukwala

H.S.

, An

, Lamperti

E.D.

, Koh

K.P.

, Ganetzky

, Liu

X.S.

, Aravind

, Agarwal

, Maciejewski

J.P.

, and Rao

(2010). Impaired hydroxylation of 5-methylcytosine in myeloid cancers with mutant TET2. Nature, 468, 839–843.

21.

Kolios

, Hoerster

A.K.

, Sehmisch

, Malcherek

M.C.

, Rack

, Tezval

, Seidlova-Wuttke

, Wuttke

, Stuermer

K.M.

, and Stuermer

E.K.

(2010). Do estrogen and alendronate improve metaphyseal fracture healing when applied as osteoporosis prophylaxis?. Calcif. Tissue Int. 86, 23–32.

22.

Lappalainen

O.P.

, Karhula

, Haapea

, Kyllonen

, Haimi

, Miettinen

, Saarakkala

, Korpi

, Ylikontiola

L.P.

, Serlo

W.S.

, and Sándor

G.K.

(2016). Bone healing in rabbit calvarial critical-sized defects filled with stem cells and growth factors combined with granular or solid scaffolds. Childs Nerv. Syst. 32, 681–688.

23.

Lee

J.M.

, Libermann

T.A.

, and Cho

J.Y.

(2010). The synergistic regulatory effect of Runx2 and MEF transcription factors on osteoblast differentiation markers. J. Periodontal. Implant. Sci. 40, 39–44.

24.

Lee

Y.Y.

, Kim

H.B.

, Lee

J.W.

, Lee

G.M.

, Kim

S.Y.

, Hur

J.A.

, and Cho

H.C.

(2016). The association between urine albumin to creatinine ratio and osteoporosis in postmenopausal women with type 2 diabetes. J. Bone Metab. 23, 1–7.

25.

, Hu

, Han

, Liu

, Jing

, Tang

, Tian

, and Long

(2015). MiR-154-5p regulates osteogenic differentiation of adipose-derived mesenchymal stem cells under tensile stress through the Wnt/PCP pathway by targeting Wnt11. Bone, 78, 130–141.

26.

, Zhou

, Cao

, Liu

, Wang

, Chen

, Xing

, Chen

, Bai

, Yuan

, Cheng

, Xu

, Yang

F.C.

, and Zhao

(2018). TET2 loss dysregulates the behavior of bone marrow mesenchymal stromal cells and accelerates Tet2(-/-)-driven myeloid malignancy progression. Stem Cell Rep. 10, 166–179.

27.

, Cai

C.L.

, Wang

, Zhang

, Petersen

B.E.

, Yang

F.C.

, and Xu

(2011). Deletion of Tet2 in mice leads to dysregulated hematopoietic stem cells and subsequent development of myeloid malignancies. Blood, 118, 4509–4518.

28.

Ling

, Huang

, Islam

, Heruth

D.P.

, Li

, Zhang

L.Q.

, Li

D.Y.

, Hu

, and Ye

S.Q.

(2017). Epigenetic regulation of Runx2 transcription and osteoblast differentiation by nicotinamide phosphoribosyltransferase. Cell. Biosci. 7, 27.

29.

Liu

Y.P.

, Li

S.Z.

, Yuan

, Xia

, Yu

, Liu

, and Yu

G.R.

(2011). Infrapatellar fat pad may be with tendon repairing ability and closely related with the developing process of patella Baja. Med. Hypotheses, 77, 620–623.

30.

Marinus

M.G.

, and Casadesus

(2009). Roles of DNA adenine methylation in host-pathogen interactions: Mismatch repair, transcriptional regulation, and more. FEMS Microbiol. Rev. 33, 488–503.

31.

Marumoto

, Milani

, da Silva

R.A.

, da Costa Fernandes

C.J.

, Granjeiro

J.M.

, Ferreira

C.V.

, Peppelenbosch

M.P.

, and Zambuzzi

W.F.

(2017). Phosphoproteome analysis reveals a critical role for hedgehog signalling in osteoblast morphological transitions. Bone, 103, 55–63.

32.

Matsushima

, Torii

, Ozaki

, and Narama

(2003). Iron lactate-induced osteomalacia in association with osteoblast dynamics. Toxicol. Pathol. 31, 646–654.

33.

Metavarayuth

, Sitasuwan

, Luckanagul

J.A.

, Feng

, and Wang

(2015). Virus nanoparticles mediated osteogenic differentiation of bone derived mesenchymal stem cells. Adv. Sci. (Weinh) 2, 1500026.

34.

, Gao

, Dai

, Ma

, Xu

, and Jin

(2015). A novel function of TET2 in CNS: Sustaining neuronal survival. Int. J. Mol. Sci. 16, 21846–21857.

35.

Narai

, Katoh

, Inoue

, Taniguchi

, Kazuki

, Sato

, Kodani

, Ryoke

, and Oshimura

(2015). Construction of a luciferase reporter system to monitor osteogenic differentiation of mesenchymal stem cells by using a mammalian artificial chromosome vector. Yonago Acta Med. 58, 23–29.

36.

Nyman

J.S.

, Lynch

C.C.

, Perrien

D.S.

, Thiolloy

, O'Quinn

E.C.

, Patil

C.A.

, Bi

, Pharr

G.M.

, Mahadevan-Jansen

, and Mundy

G.R.

(2011). Differential effects between the loss of MMP-2 and MMP-9 on structural and tissue-level properties of bone. J. Bone Miner. Res. 26, 1252–1260.

37.

, Zhang

, Wang

, Li

, Huang

, Zhao

, Wu

, An

, Guo

, Hale

, Li

, Hillyer

C.D.

, Mohandas

, Liu

, Yazdanbakhsh

, Vinchi

, Chen

, Kang

, and An

(2018). TET2 deficiency leads to stem cell factor-dependent clonal expansion of dysfunctional erythroid progenitors. Blood, 132, 2406–2417.

38.

Shono

, Ueha

, Wang

, Abe

, Kurachi

, Matsuno

, Sugiyama

, Nagasawa

, Imamura

, and Matsushima

(2010). Bone marrow graft-versus-host disease: Early destruction of hematopoietic niche after MHC-mismatched hematopoietic stem cell transplantation. Blood, 115, 5401–5411.

39.

Silva

R.A.D.

, Fuhler

G.M.

, Janmaat

V.T.

, Fernandes

C.J.D

.C., and Peppelenbosch

M.P.

(2019). HOXA cluster gene expression during osteoblast differentiation involves epigenetic control. Bone, 125, 74–86.

40.

Smith

B.J.

, Lucas

E.A.

, Turner

R.T.

, Evans

G.L.

, Lerner

M.R.

, Brackett

D.J.

, Stoecker

B.J.

, and Arjmandi

B.H.

(2005). Vitamin E provides protection for bone in mature hindlimb unloaded male rats. Calcif. Tissue Int. 76, 272–279.

41.

Van Etten

R.A.

, Koschmieder

, Delhommeau

, Perrotti

, Holyoake

, Pardanani

, Mesa

, Green

, Ibrahim

A.R.

, Mughal

, Gale

R.P.

, and Goldman

(2011). The Ph-positive and Ph-negative myeloproliferative neoplasms: Some topical pre-clinical and clinical issues. Haematologica, 96, 590–601.

42.

Yang

, Liu

, Bai

, Zhang

J.Y.

, Ma

S.H.

, Liu

, Xu

Z.D.

, Zhu

H.G.

, Ling

Z.Q.

, Ye

, Guan

K.L.

, and Xiong

(2013). Tumor development is associated with decrease of TET gene expression and 5-methylcytosine hydroxylation. Oncogene, 32, 663–669.

43.

Yang

, Yu

, Kou

, Gao

, Chen

, Liu

, Zhou

, and Shi

(2018). Tet1 and Tet2 maintain mesenchymal stem cell homeostasis via demethylation of the P2rX7 promoter. Nat. Commun. 9, 2143.

44.

Zhang

Q.Q.

, Xu

M.Y.

, Qu

, Hu

J.J.

, Li

Z.H.

, Zhang

Q.D.

, and Lu

L.G.

(2014). TET3 mediates the activation of human hepatic stellate cells via modulating the expression of long non-coding RNA HIF1A-AS1. Int. J. Clin. Exp. Pathol. 7, 7744–7751.

45.

Zhang

R.P.

, Shao

J.Z.

, and Xiang

L.X.

(2011). GADD45A protein plays an essential role in active DNA demethylation during terminal osteogenic differentiation of adipose-derived mesenchymal stem cells. J. Biol. Chem. 286, 41083–41094.

46.

Zhong

, Wang

Q.Q.

, Li

J.W.

, Zhang

Y.M.

, An

X.R.

, and Hou

(2017). Ten-eleven translocation-2 (Tet2) is involved in myogenic differentiation of skeletal myoblast cells in vitro. Sci. Rep. 7, 43539.

47.

Zhou

G.S.

, Zhang

X.L.

, Wu

J.P.

, Zhang

R.P.

, Xiang

L.X.

, Dai

L.C.

, and Shao

J.Z.

(2009). 5-Azacytidine facilitates osteogenic gene expression and differentiation of mesenchymal stem cells by alteration in DNA methylation. Cytotechnology, 60, 11.

48.

Zidi

, and Allaire

(2015). Mechanical behavior of abdominal aorta aneurysm in rat model treated by cell therapy using mesenchymal stem cells. Biomech. Model. Mechanobiol. 14, 185–194.

49.

Zuk

P.A.

, Zhu

, Mizuno

, Huang

, Futrell

J.W.

, Katz

A.J.

, Benhaim

, Lorenz

H.P.

, and Hedrick

M.H.

(2001). Multilineage cells from human adipose tissue: Implications for cell-based therapies. Tissue Eng. 7, 211–228.