MFN2 knockdown promotes osteogenic differentiation of iPSC-MSCs through aerobic glycolysis mediated by the Wnt/β-catenin signaling pathway

Background Mitofusin-2 (MFN2) is a kind of GTPase that participates in the regulation of mitochondrial fusion, which is related to a variety of physiological and pathological processes, including energy metabolism, cell differentiation, and embryonic development. However, it remains unclear whether MFN2 is involved in the metabolism and osteogenic differentiation of mesenchymal stem cells (MSCs). Methods MFN2 knockdown (MFN2-KD) and MFN2-overexpressing (MFN2-OE) induced pluripotent stem cell-derived mesenchymal stem cells (iPSC-MSCs) were constructed by lentivirus. The commercial kits were utilized to detect the glycolysis and oxidative phosphorylation (OXPHOS) rate. Flow cytometry, Western blot, quantitative real-time polymerase chain reaction (qRT-PCR), RNA-seq, immunofluorescence, and immunoprecipitation were employed for phenotype and molecular mechanism assessment. Results We demonstrated that MFN2 and Wnt/β-catenin signaling pathway regulated glycolysis of iPSC-MSCs. The lack of MFN2 promoted the osteogenic differentiation of iPSC-MSCs, and aerobic glycolysis in the presence of sufficient oxygen, which increased glucose consumption and lactic acid production, as well as the glycolytic enzyme activity and gene expression. Inhibiting the Wnt/β-catenin signaling pathway normalized the enhanced glycolytic rate and osteogenic differentiation of MFN2-KD iPSC-MSCs. MFN2-OE iPSC-MSCs displayed the opposite phenotype. Conclusions Downregulating MFN2 promotes osteogenic differentiation of iPSC-MSCs through aerobic glycolysis mediated by the Wnt/β-catenin signaling pathway. Our research reveals the new function of MFN2 in regulating the osteogenic differentiation and energy metabolism of MSCs, which will provide a new therapeutic target and theoretical basis for alveolar bone repair and periodontal regenerative treatment. Supplementary Information The online version contains supplementary material available at 10.1186/s13287-022-02836-w.


Introduction
Periodontal regeneration treatment has been a hot spot in the field of periodontal research in recent years. Mesenchymal stem cells (MSCs) are considered to be the most promising resource in periodontal tissue regeneration [1]. The number of autologous MSCs is limited, and our previous study reprogrammed human gingival fibroblasts into iPSCs, which is an endless source of induced pluripotent stem cell-derived mesenchymal stem cells (iPSC-MSCs). Understanding the molecular mechanism of the osteogenic differentiation process and promoting the osteogenic differentiation efficiency of MSCs are very important for periodontal regeneration.
Aerobic glycolysis and oxidative phosphorylation (OXPHOS) are utilized during the proliferation of MSCs [2]. Some types of stem cells and rapidly proliferating cells such as cancer cells consume glucose mainly through glycolysis, rather than OXPHOS when oxygen is sufficient. This phenomenon of cellular aerobic glycolysis is similar to the Warburg effect [3]. Previous studies have found that the Warburg effect is critical for bone formation, which chooses glycolysis rather than OXPHOS as the main metabolic mode. It consumes a large amount of glucose and produces lactic acid as the main end product [4]. Another study demonstrates that MSCs make more use of OXPHOS in osteogenic culture conditions [5]. It is worth noting that it remains controversial about the main pathway of energy metabolism and its molecular mechanism of regulation during the osteogenic differentiation of MSCs.
Mitofusin 2 (MFN2) is encoded by the MFN2 gene located on chromosome 1 in the nuclear genome [6]. MFN2 is a multifunctional protein located on the mitochondrial outer membrane; in addition to mitochondrial fusion, it participates in various biological processes under physiological and pathological conditions, such as cell proliferation, autophagy, apoptosis, endoplasmic reticulum-mitochondrial connection, and stem cell differentiation [7]. Our previous studies showed that MFN2 knockdown promoted neural ectodermal differentiation of hESCs [8]. Several studies have shown that it also plays an important role in mitochondrial metabolism and function, mainly manifested in the effect on glucose oxidation and cellular respiration [9]. Studies have shown that the lack of MFN2 inhibits the expression of OXPHOS complexes, and its participation in mitochondrial metabolism is independent of mitochondrial fusion [10]. However, MFN2 knockdown significantly increases the expression of genes encoding glycolytic enzymes and glycolytic metabolites [11]. In the early stage of osteogenesis, the lack of Mfn2 enhances the osteogenic differentiation and cortical bone accumulation of mice in vitro and in vivo [12]. These findings indicate that MFN2 ablation may promote the conversion of glycolytic bioenergy to meet the energy requirements of specific cells such as osteoblasts.
Here we reported that MFN2 ablation promoted the osteogenic differentiation of iPSC-MSCs. MFN2 ablation increased glucose consumption and lactic acid production in the presence of sufficient oxygen, as well as the expression of glycolytic enzymes, which promotes the conversion of cell metabolism to aerobic glycolysis. Inhibiting the Wnt/β-catenin signaling pathway normalized the enhanced glycolysis and osteogenic differentiation of MFN2-KD iPSC-MSCs. MFN2 overexpressing iPSC-MSCs displayed the opposite phenotype. Our results reveal the new role of MFN2 in the osteogenic differentiation and energy metabolism of MSCs, which are expected to provide new ideas for alveolar bone repair and periodontal regeneration. medium (STEMCELL Technologies, Canada) and MesenCult ™ -ACF Plus medium (STEMCELL Technologies, Canada) from day 4-5. From day 6 to day 21, cells were passaged at approximately 80% confluent and cultured with MesenCult ™ -ACF Plus medium to get iPSC-MSCs. PDLSCs were cultured in DMEM containing 10% FBS (Gibco, USA) and 1% penicillin-streptomycin (Gibco, USA). Cells were cultured in a cell culture incubator (Thermo Fisher Scientific, USA) containing 5% CO 2 at 37 °C.

Construction of MFN2 knockdown (MFN2-KD) iPSC-MSCs by lentivirus
A pair of oligonucleotides encoding shRNA targeting against MFN2 (GGA AGA GCA CCG TGA TCA ATG) and NC (negative control, TTC TCC GAA CGT GTC ACG T) were cloned into GV307 plasmid (Genechem, China) with XhoI and EcoRI restriction sites. The plasmid was transformed to competent cells and screened by ampicillin-containing agar plates. The PCR product was identified and sequenced. pGag/Pol, pRev, pVSV-G, and transfection reagent (Genechem, China) were utilized to package the lentiviral particles and then infected the 293 T cells for 6 h. After being cultured for 48 h, the virus was concentrated with ultracentrifuge (Beckman Coulter, USA). For infection, fresh MesenCult ™ -ACF Plus medium containing 10 µl concentrated virus (10 8 TU/ml) and 8 μg/ml polybrene was added to the iPSC-MSCs with 50% confluence for 12 h. The positive cells were screened with 0.3 µg/ml puromycin (Thermo Fisher Scientific, USA) 48 h after infection.

Construction of MFN2 overexpressing (MFN2-OE) iPSC-MSCs by lentivirus
The linearized GV492 vector (Genechem, China) was obtained by digestion with BamHI and AgeI. The target gene fragment was amplified by PCR (primers with underlined homologous recombination sequence: MFN2 -1: AGG TCG ACT CTA GAG GAT CCCGC CAC CAT GTC  CCT GCT CTT CTC TCG; and MFN2-2: TCC TTG TAG  TCC ATACC TCT GCT GGG CTG CAG GTA CTG GTG  ). The linearized vector and target gene fragment were connected. Empty vector was utilized as negative control (NC). The following steps transformation, virus packaging, purification, and infection were the same as mentioned before.

Transmission electron microscopy
The iPSC-MSCs were fixed in 3% glutaraldehyde overnight, rinsed with PBS, and fixed in 1% osmium tetroxide for 3 h. Then the cells were dehydrated in gradient ethanol and acetone, embedded in epoxy resin and sliced. Images were obtained by an electron microscope (JEOL, Japan).

Cell counting kit-8 (CCK-8) assay
Cells were seeded in a 96-well plate with a density of 2000 cells/well, and 3 parallel wells were set for each group at each time point. After 0, 2, 4, 6, 8 days of culture, cells were added with 10 µl CCK-8 reagent (Dojindo, Japan), 100 µl DMEM, and incubated for 1 h at 37 °C. A microplate reader (Molecular Devices, USA) was used to measure the absorbance at 450 nm.

Cell cycle assay
The cells were trypsinized, centrifuged at 1000g for 3 min, rinsed with precooled PBS and centrifuged again, fixed in 70% precooled ethanol at 4 °C overnight. The cells were centrifuged again, washed twice with precooled PBS, incubated with RNase A and PI (Beyotime, China) at 37 °C in the dark for 30 min, and the cell cycle (DNA content) was detected by flow cytometry (BD, USA) and evaluated by the ModFit software.

Quantitative real-time polymerase chain reaction (qRT-PCR) analysis
Total RNA was extracted by Trizol reagent (Thermo Fisher Scientific, USA), and reverse transcription was performed using PrimeScript RT kit (Takara, Japan) according to the manufacturer's instructions. qRT-PCR was conducted using Power Up SYBR Green Master Mix (Thermo Fisher Scientific, USA) in CFX384 Touch Real-Time PCR Detection System (Bio-Rad, USA). β-Actin was employed as an internal control. The primer sequences are shown in Table 1.

RNA-sequencing (RNA-seq)
Total RNA of the WT and MFN2-KD iPSC-MSCs were collected. Magnetic beads with polythymine were used to enrich mRNA containing polyA tails. The enriched mRNA was fragmented and reverse-transcribed into double-stranded cDNA. The ligation product was amplified by PCR and heat-denatured to obtain a singlestranded product, which is circularized to obtain a DNA library. The library preparations were sequenced on the BGISEQ-500 platform after quality inspection. Raw reads obtained by sequencing were filtered to get clean reads, which were compared to reference genes. mRNA quantification and differential expression analysis were performed, and P-value < 0.05 was considered differentially expressed. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) enrichment analysis on differentially expressed genes were performed on the Multi-omics System of BGI Genomics. In short, for GO analysis, all candidate differentially expressed genes were mapped and calculated to the GO database. A hypergeometric test was used to find out significantly enriched GO terms compared to the background genes. For KEGG pathway analysis, a hypergeometric test was used to find out significantly enriched pathways compared with background genes. The results were corrected by multiple tests to obtain Q value, and Q value ≤ 0.05 was defined as significantly enriched.

Glycolysis (extracellular acidification) assay and extracellular oxygen consumption assay
According to the manufacturer's protocols, cells were seeded on a 96-well plate (5 × 10 4 cells/well). The glycolysis (basal glycolysis and glycolytic capacity) rate was measured by fluorescence intensity in a fluorescence microplate reader (Molecular Devices, USA) with Glycolysis Stress Test Complete Assay Kit (Abcam, USA). Extracellular oxygen consumption assay was performed using Mitochondrial Stress Test Complete Assay Kit (Abcam, USA). Basal glycolysis rate refers to the increase in extracellular acidification caused by the proton output of glycolysis. Glycolytic capacity refers to the maximum glycolytic capacity of cells while OXPHOS is blocked by oligomycin. Basal respiration rate refers to the oxygen consumption used to meet mitochondrial proton leakage and cellular ATP demand, indicating the energy demand under basal conditions. Maximal respiration rate refers to the stimulation of the mitochondrial respiratory chain to operate to the maximum extent after FCCP treatment, indicating the maximum respiration potential of the cell.

HK (hexokinase) and LDH (lactate dehydrogenase) activity assay
Cells were harvested, and the activity of HK and LDH was performed by Hexokinase Activity Assay Kit (Abcam, USA) and Lactate Dehydrogenase Assay Kit (Abcam, USA) according to the manufacturer's protocols. In brief, the cells were collected by trypsin and washed 3 times with precooled PBS, resuspended in precooled assay buffer, centrifuged at 12,000 rpm at 4 °C for 5 min to remove insoluble materials. The supernatant was collected, mixed with the reaction mixture, and incubated at room temperature in the dark for 40 min. The absorbance at 450 nm was measured by the microplate reader (BD, USA) and calculated the relative activity.

Lactic acid measurements and glucose consumption assay
According to the manufacturer's protocols, for lactic acid measurement, cells were lysed and assayed with the Lactic Acid Assay Kit (Nanjing Jiancheng, China). In brief, the cells were lysed by ultrasonic on ice, and the enzyme working solution and the color reagent were added and mixed. The absorbance at 530 nm was measured by the microplate reader (BD, USA), and the lactic acid concentration was calculated according to the standard curve. For glucose consumption measurement, the media cultured cells for 24 h were assayed using Glucose (HK) Assay Kit (Sigma-Aldrich, USA). In brief, the same amount of cells was seeded in the well plate, and the glucose determination reagent was added. The absorbance at 340 nm was measured by the microplate reader (BD, USA). The standardized glucose concentration and consumption were calculated based on the glucose standard solution and the amount of sample protein.

Immunoprecipitation (IP)
Immunoprecipitation was carried out by Immunoprecipitation Kit (Beyotime, China) according to the manufacturer's protocol. In brief, cells were lysed in prechilled cell lysate and centrifuged to remove cell debris. One part is used as input, and the other part is incubated with protein A/G magnetic beads bound to MFN2 antibody (1∶50, CST, USA) or IgG at 4 °C overnight. Input and proteins eluted from protein A/G magnetic beads were detected by Western blot.

Statistical analysis
All experiments were performed three times independently. The biological triplicate results were presented as means ± standard deviation using GraphPad Prism (ver. 6.01, GraphPad Software Inc, USA). The data were analyzed by one-way ANOVA for comparisons. P < 0.05 was considered statistically significant.

MFN2 knockdown promotes osteogenic differentiation of iPSC-MSCs
iPSCs were reprogrammed from human gingival fibroblasts in our previous studies, which highly expressed pluripotency markers (Additional file 1: Fig. S1). MFN2-KD and MFN2-OE iPSC-MSCs showed little change in morphology with the wild type (WT) iPSC-MSCs (Fig. 1A). Flow cytometry analysis demonstrated that the expression of surface markers conforms to ISCT criteria for MSCs (Additional file 2: Fig. S2, Additional file 3: Fig.  S3). MFN2 knockdown and overexpression efficiency in iPSC-MSCs were confirmed by qRT-PCR and Western blot (Fig. 1B-E). Furthermore, transmission electron microscopy showed that MFN2 knockdown increased the proportion of spherical mitochondria in iPSC-MSCs, indicating that mitochondrial fusion decreased, and MFN2 overexpressing increased the length of mitochondria (Additional file 4: Fig. S4A). MFN2 has no significant effect on the cell proliferation and cell cycle of iPSC-MSCs (Additional file 4: Fig. S4B-D). After 14-day culture in the osteogenic induction medium (IM), MFN2-KD iPSC-MSCs showed enhanced mRNA and protein expression of osteogenic differentiation markers RUNX2, OSX, ALP, and OCN (Fig. 1F, H). Alizarin red staining showed more mineralized nodules in the MFN2-KD group (Fig. 1I). MFN2-OE iPSC-MSCs showed the opposite phenotype (Fig. 1G-I). There is no obvious difference in adipogenic and chondrogenic differentiation ability between the three groups (Additional file 5:

Bioinformatics analysis of MFN2-KD compared to WT iPSC-MSCs
RNA-sequencing (RNA-seq) was performed to obtain the differential gene expression profile of WT and MFN2-KD iPSC-MSCs. There were 5629 upregulated and 5463 downregulated genes in MFN2-KD iPSC-MSCs compared with the WT iPSC-MSCs ( Fig. 2A), which showed that MFN2 conferred considerable changes in the transcriptome of the iPSC-MSCs. GO enrichment analysis showed that metabolic and developmental processes were enriched in the biological process (Fig. 2B).
The KEGG pathway analysis demonstrated the gene number enriched in the metabolism category (Fig. 2C).  This indicated that MFN2 may play an important role in the differentiation of iPSC-MSCs by regulating cell energy metabolism. Both analyses showed that the Wnt signaling pathway was enriched (Fig. 2D, E). Studies have shown that the ablation of MFN2 promotes the proliferation and migration of bladder cancer cells by activating the Wnt/β-catenin signaling pathway [16]. MFN2 promotes the NSC differentiation of hiPSCs by activating the Wnt/β-catenin signaling pathway [18]. Recent studies have shown that depletion of MFN2 in endothelial cells leads to inhibition of β-catenin sulfenylation, thereby activating the activity of the Wnt/β-catenin signaling pathway [19]. Therefore, we focused on the Wnt/βcatenin signaling pathway in the next section.

MFN2 knockdown enhances glycolysis of iPSC-MSCs
Glycolysis Stress Test Complete Assay Kit and Mitochondrial Stress Test Complete Assay Kit were utilized to determine whether MFN2 regulates the energy metabolism of iPSC-MSCs. The results showed that the basal glycolysis and glycolytic capacity in MFN2-KD iPSC-MSCs were significantly enhanced (Fig. 3A). We further determined the activity of key enzymes and mRNA expression levels in the glycolysis process and found that they were all significantly upregulated (Fig. 3C, D). Consistently, the concentration of intracellular lactic acid was increased (Fig. 3E). HIF-1α plays an important role in guiding the conversion of cell metabolism to glycolysis, which is required for increased glycolysis by MFN1/2 ablation [11]. Our result was consistent with it, and MFN2 knockdown induced the stabilization of HIF-1α (Additional file 6: Fig. S6A, B). There is no statistical difference in the effect of MFN2 on OXPHOS (Fig. 3I, J). In addition, MFN2-OE iPSC-MSCs showed the opposite phenotype (Fig. 3B, F, G, H). These results indicate that MFN2 knockdown enhances the glycolysis of iPSC-MSCs.

MFN2 knockdown promotes osteogenic differentiation through aerobic glycolysis via the Wnt/β-catenin signaling pathway
To further determine the role of MFN2 and the Wnt/βcatenin signaling pathway in the osteogenic differentiation and energy metabolism of iPSC-MSCs, we first determine the effect of MFN2 on the Wnt/β-catenin signaling pathway. Immunoprecipitation showed a binding relationship between MFN2 and β-catenin (Additional file 7: Fig. S7). After osteogenic induction with/without inhibitor/activator treatment for 24 h, Western blot results showed that MFN2 knockdown increased the expression level of β-catenin in iPSC-MSCs, which was normalized by XAV-939 (Fig. 5A). The expression level of β-catenin in MFN2-OE iPSC-MSCs decreased, which was rescued by SKL2001 (Fig. 5B). Therefore, MFN2 regulates the Wnt/β-catenin signaling pathway by directly binding to β-catenin. We next determined the effects of MFN2 and Wnt/ β-catenin signaling pathways on energy metabolism in the osteogenic differentiation of iPSC-MSCs. The results showed that MFN2 knockdown significantly increased glucose consumption, glycolysis level, intracellular lactic acid concentration, activity, and mRNA expression levels of key glycolysis enzymes and HIF-1α 14 days after osteogenic differentiation, while XAV-939 eliminated this effect (Fig. 5C, E, F, I, J and Additional file 6: Fig. S6C, D). SKL2001 rescued the reduced glycolysis level of MFN2-OE iPSC-MSCs (Fig. 5D, G, H, I, J and Additional file 6: Fig. S6E, F). XAV-939 and SKL2001 had no significant effect on OXPHOS during osteogenic differentiation (Fig. 5K, L).
Finally, we evaluated the effects of MFN2 and Wnt/ β-catenin signaling pathways on osteogenic differentiation. Consistent with changes in energy metabolism, we found that the increased expression of osteogenic differentiation markers (Fig. 6A, C) and mineralized nodules of MFN2-KD iPSC-MSCs (Fig. 6E) were normalized by XAV-939. The decreased osteogenic differentiation of MFN2-OE iPSC-MSCs was partially rescued by SKL2001 (Fig. 6B, D, E). We further used PDLSCs as a cell model and got consistent results as iPSC-MSCs (Figs. 7, 8 and 9). These results indicate that downregulated MFN2 promotes osteogenic differentiation of MSCs through aerobic glycolysis via the Wnt/β-catenin signaling pathway (Fig. 6F).

Discussion
MSCs are progenitor cells that can differentiate into osteoblasts, adipocytes, and chondrocytes [20]. Osteoblasts synthesize numerous extracellular matrix proteins and therefore have a high demand for materials and energy sources during the osteogenic differentiation of MSCs. There is controversy about the metabolic mode of MSCs' osteogenic differentiation. Recent studies have found that the OXPHOS is activated during the osteogenic differentiation of MSCs, as revealed by the upregulation of the oxygen consumption rate of mitochondria and respiratory enzymes, while glycolysis is reduced [21,22]. Other studies found that ATP production of hMSCs is more dependent on glycolysis compared with differentiated cells [23]. Inhibition of glycolytic enzymes in vitro impairs bone formation in mouse osteoblasts [24]. The activation of glycolysis by overexpression of HIF-1α leads to more osteoblasts and bone formation in mice [25]. This raises the question of why aerobic glycolysis, which produces less ATP, is chosen in osteoblasts. One possible explanation is that osteoblasts have important requirements that exceed ATP, aerobic glycolysis may be necessary to meet the specific metabolic needs of osteogenic differentiation. Another possible explanation is that the cell culture environment has sufficient energy sources, so inefficient ATP production is not a problem [3]. Studies have shown that osteogenic differentiation depends on changes in specific stages of energy acquisition while early osteogenic differentiation requires the synthesis of collagen and upregulation of OXPHOS. The increase in OXPHOS activity is accompanied by the increase in reactive oxygen species (ROS), which may lead to cell damage or lineage switching. Therefore, cell metabolism is converted to glycolysis during the maturation and mineralization of osteoblasts [26,27]. Our conclusion provides a compromise: Aerobic glycolysis is activated during iPSC-MSCs osteogenesis, and the OXPHOS level is unchanged. This indicates the metabolic plasticity of the osteogenic differentiation of MSCs in meeting cellular energy demands.
Mitochondria are important organelles for cells to synthesize ATP and critical for maintaining normal physiological functions. Mitochondria change their shape via GTPases (mitofusin-1 (MFN1), MFN2, optic atrophy-1 (OPA1), fission protein -1 (FIS1), and dynamin-related protein-1 (DRP1)) to respond to changes in cell energy demand and supply [28]. There are few studies on MFN2 and osteogenic differentiation. The lack of Mfn2 in the mouse osteoblast lineage enhances the formation of basal bone in vivo and in vitro by increasing the number of osteoblasts and osteogenic differentiation capability [12]. The expression of MFN2 and the endoplasmic reticulummitochondrial coupling increase in the inflammatory microenvironment, leading to impaired osteogenic differentiation of periodontal ligament stem cells [29]. This is consistent with our results. However, some studies have shown that Mfn2 increases during the differentiation of mouse osteoblasts, and Mfn2 knockdown reduces mitochondrial elongation and osteogenic differentiation [30]. Conflicting results may be due to differences in donor species, cell type, induction medium, and oxygen concentration. The depletion of MFN2 in different cell models changes the cellular metabolic profile, resulting in a decrease in cell oxygen consumption and OXPHOS level. The cells increase glucose uptake and turn to glycolysis to compensate for the decreased ATP production [7]. Our glycolysis results are consistent with it. But interestingly, our results showed that there is no significant difference in the effect of MFN2 on OXPHOS. Recent studies have shown that MFN1 regulates the metastasis of hepatocellular carcinoma through a metabolic shift from aerobic glycolysis to OXPHOS [31]. Knockdown of the other two mitochondrial fusion regulatory genes OPA1 or MFN1 attenuates OXPHOS and ATP production of tumor cells [32]. We speculate that there may be a certain compensatory effect between different mitochondrial fusion genes and the level of cellular OXPHOS, thereby preventing the aerobic metabolism and ATP production of cells from experiencing drastic fluctuations. Studies have shown that MFN2 plays an important role in regulating cell proliferation [7]. Interestingly, our study demonstrated that MFN2 has no significant effect on the cell proliferation and cell cycle of iPSC-MSCs, which may be the result of different cell types or mechanisms [33]. Therefore, MFN2 knockdown does not stimulate osteogenic differentiation by regulating cell proliferation.
In addition, we want to figure out the signaling pathway that mediates glycolysis. Although most signal pathways related to development have been shown to regulate osteogenic differentiation, their potential effects on cell metabolism have not been fully studied. The Wnt signaling pathway is critical in differentiation and development, which plays an important role in normal bone development and homeostasis [34]. Recent studies have shown that activation of the canonical Wnt/β-catenin signaling pathway rescues glycolytic and osteogenic differentiation disorders [35]. Wnt signaling pathway reprograms cell glucose metabolism during osteogenic differentiation. WNT3A induces aerobic glycolysis by increasing the levels of key glycolytic enzymes, thereby enhancing bone formation in vitro and in vivo [24]. We found that blocking the Wnt/β-catenin signaling pathway normalized the enhanced glycolysis and mineralization of MFN2-KD iPSC-MSCs, while the activation rescued the defects in MFN2-OE iPSC-MSCs. Interestingly, activating or inhibiting the Wnt/β-catenin signaling pathway has almost no significant effect on OXPHOS in all cells. However, a recent study shows that Wnt/β-catenin inhibition increases OXPHOS in colorectal cancer cells [36]. We think this may be due to the existence of other pathways coregulating OXPHOS during the osteogenic differentiation of iPSC-MSCs, or the existence of abnormal metabolic regulation mechanisms in cancer cells.
MFN2 may regulate osteogenic differentiation through multiple signaling pathways. MFN2 inhibits mTORC2 signaling by binding to its HR1 domain [37]. mTORC2 signal positively regulates bone development and  homeostasis [38]. MFN2 interacts with the Notch signaling pathway [39], which plays an important role in bone development and remodeling [40]. Studies have found that increasing OXPHOS by acetylation stabilizes β-catenin, thereby promoting the osteogenic differentiation of BMSCs, which represents a possible mechanism in which the interaction between MFN2 and Wnt/β-catenin signaling pathway is directly related to osteogenic differentiation [41]. There are few studies on the direct relationship between MFN2 and Wnt/β-catenin pathway. Recent studies have shown that MFN2 significantly interacts with β-catenin in endothelial cells. Sequence alignment analysis shows that MFN2 has a β-catenin binding motif between its HR domains. Although MFN2 is mainly located in the cytoplasm and mitochondria, both MFN2 and β-catenin can accumulate in the nucleus. Depletion of MFN2 leads to inhibition of β-catenin sulfenylation, thereby activating the transcriptional activity of β-catenin [19]. Similarly, our results showed that MFN2 can bind to β-catenin. The Wnt/βcatenin signaling pathway regulates the proliferation and migration of bladder cancer, which can be activated by the ablation of MFN2 [16]. This is consistent with our results. But another recent study found that MFN2 overexpression promotes the NSC differentiation of hiP-SCs by activating the Wnt/β-catenin signaling pathway [18]. This may be due to the interaction between MFN2 and Wnt/β-catenin signaling pathway may play different roles in different kinds of differentiation. Further work is needed to figure out the sites of action and mechanism between MFN2 and Wnt/β-catenin signaling pathway. Our work has some limitations. Firstly, we have not validated our conclusions in vivo, because there are certain differences in the microenvironment in vivo and in vitro, which may lead to differences in results. Secondly, fatty acids and glutamine are utilized as energy sources for cell metabolism in addition to glucose [42]. We only detected the changes in glucose uptake and did not further explore the possible roles and mechanisms of the other two raw materials in the osteogenic differentiation of iPSC-MSCs. Thirdly, MSCs can regulate the innate and adaptive immune system, which immunomodulatory activity is mainly mediated by paracrine factors [43]. Mitochondria are increasingly recognized as critical organelles for innate and adaptive immune signals. MFN2 is essential for cytokine production, antigen processing, and phagocytosis [44]. Further work needs to be done to get a better understanding of the function of MFN2 and serve the periodontal tissue regeneration.

Conclusions
In summary, our study clarified that the lack of MFN2 promoted the osteogenic differentiation of iPSC-MSCs, and increased glucose consumption and lactic acid production in the presence of sufficient oxygen, as well as the expression of glycolytic enzymes, which promoted aerobic glycolysis. Inhibition of the Wnt/β-catenin signaling pathway normalized the enhanced aerobic glycolysis and osteogenic differentiation of MFN2-KD iPSC-MSCs. Overexpression of MFN2 can lead to the opposite phenotype. This study identified MFN2 as a new regulator of MSCs' cell metabolism, which is expected to provide a new theoretical basis and therapeutic target for alveolar bone repair, periodontal regeneration, moreover, prevention and treatment of diseases related to osteogenic differentiation and abnormal bone formation.