METTL3‐m6 A methylase regulates the osteogenic potential of bone marrow mesenchymal stem cells in osteoporotic rats via the Wnt signalling pathway

Abstract Objectives Bone marrow mesenchymal stem cells (BMSCs) hold a high osteogenic differentiation potential, but the mechanisms that control the osteogenic ability of BMSCs from osteoporosis (OP‐BMSCs) need further research. The purpose of this experiment is to discuss the osteogenic effect of Mettl3 on OP‐BMSCs and explore new therapeutic target that can enhance the bone formation ability of OP‐BMSCs. Materials and Methods The bilateral ovariectomy (OVX) method was used to establish the SD rat OP model. Dot blots were used to reveal the different methylation levels of BMSCs and OP‐BMSCs. Lentiviral‐mediated overexpression of Mettl3 was applied in OP‐BMSCs. QPCR and WB detected the molecular changes of osteogenic‐related factors and Wnt signalling pathway in vitro experiment. The staining of calcium nodules and alkaline phosphatase detected the osteogenic ability of OP‐BMSCs. Micro‐CT and histological examination evaluated the osteogenesis of Mettl3 in OP rats in vivo. Results The OP rat model was successfully established by OVX. Methylation levels and osteogenic potential of OP‐BMSCs were decreased in OP‐BMSCs. In vitro experiment, overexpression of Mettl3 could upregulate the osteogenic‐related factors and activate the Wnt signalling pathway in OP‐BMSCs. However, osteogenesis of OP‐BMSCs was weakened by treatment with the canonical Wnt inhibitor Dickkopf‐1. Micro‐CT showed that the Mettl3(+) group had an increased amount of new bone formation at 8 weeks. Moreover, the results of histological staining were the same as the micro‐CT results. Conclusions Taken together, the methylation levels and osteogenic potential of OP‐BMSCs were decreased in OP‐BMSCs. In vitro and in vivo studies, overexpression of Mettl3 could partially rescue the decreased bone formation ability of OP‐BMSCs by the canonical Wnt signalling pathway. Therefore, Mettl3 may be a key targeted gene for bone generation and therapy of bone defects in OP patients.

Mettl3 could partially rescue the decreased bone formation ability of OP-BMSCs by the canonical Wnt signalling pathway. Therefore, Mettl3 may be a key targeted gene for bone generation and therapy of bone defects in OP patients.

| INTRODUCTION
As the worldwide population is rapidly aging, osteoporosis (OP) could become a global health-care challenge. 1,2 OP has low bone mineral density and structural degradation of bones and increases the risk of fracture in OP patients. 1,[3][4][5] Postmenopausal women suffer from OP due to the rapid decline in oestrogen levels. 6,7 At the same time, the bone loss caused by menopause usually precedes age-related bone loss, so postmenopausal women may develop osteoporotic fractures earlier than men of the same age. 1,[8][9][10] Treatments for osteoporotic fractures include drug, non-drug and stem cell therapies. [11][12][13] However, medical treatment has the disadvantages of not reversing the existing bone loss and potentially causing serious side effects. 14 Therefore, stem cell therapy is a research hotspot amongst treatments for OP-related bone defects. 15,16 Bone marrow mesenchymal stem cells (BMSCs) are one of the most effective mesenchymal stem cell types for treating OP. 17 Pino et al. found enhanced adipogenic differentiation and weakened osteogenic differentiation of BMSCs in postmenopausal OP patients (OP-BMSCs), 18 resulting in an imbalance between osteogenic and adipogenic differentiation, which disrupts the activity and microenvironment of OP-BMSCs. 19 Therefore, it is essential to find ways to reactivate the bone formation of OP-BMSCs and inactivate the formation of adipocytes. This issue is a key and difficult aspect of treating osteoporotic fractures with autologous OP-BMSCs transplantation.
Since its discovery in the 1970s, N6-methyl-adenosine (m 6 A) has been recognized as the most common internal chemical modification of mRNAs in eukaryotes. The m 6 A methyltransferase complex includes Wilms tumour 1-associating protein (WTAP), methyltransferase-like 3 (METTL3), methyltransferase-like 14 (METTL14) and m 6 A demethylase, including fat mass and obesity-associated protein (FTO) and ALKB homologue 5 (ALKBH5). 20,21 M 6 A is dynamically regulated by methylases and demethylases and controls cellular processes, and the translation of mRNAs involved in cell metabolism, cell growth and disease development. 22 It is well known that the m 6 A modification has multiple biological functions. Furthermore, METTL3 is an important m 6 A methylase that can catalyse the conversion of adenosine to m 6 A. METTL3 is a 70-kDa protein containing methylation-active catalytic residues. 23 The expression ratios of the components of the m 6 A methyltransferase complex vary greatly amongst tissues or cell types, indicating that they have different biological functions and methylation activities. Studies have shown that METTL3 deficiency decreases m 6 A levels, attenuating the normal lineage and disrupting cell cycle progression. 24,25 However, the role of Mettl3 in bone homeostasis is little known.
The Wnt signalling pathway is a pivotal regulator of bone differentiation, development and homeostasis of BMSCs. [26][27][28] Todd et al.
found that Wnt16 was significantly reduced in the BMSCs of ovariectomized (OVX) mice. 29 Wnt signalling pathway activation can enhance Wnt-related genes, restore the osteogenic potential of BMSCs and attenuate bone loss. 30 Similarly, Dickkopf-1 (DKK1) can reduce the osteogenic potential of BMSCs and cause bone loss. 31,32 Previous studies have shown that Mettl3 could regulate the Wnt/β-catenin pathway. 33,34 However, the regulatory mechanism of Mettl3 influences the bone differentiation of OP-BMSCs through the canonical Wnt pathway is unknown.
The experiment was to establish an OP model in Sprague Dawley (SD) rats with the OVX method, detect Mettl3 expression in OP-BMSCs from the OP model and control SD rats at molecular levels and then assess the effects of lentiviral-mediated Mettl3 overexpression. Finally, the Wnt signalling pathway was blocked by DKK1.
We investigated the mechanism of Mettl3 on the osteogenesis of OP-BMSCs on the transcriptional level by overexpressing Mettl3 and explored further methods of enhancing the osteogenic capacity of OP-BMSCs. Moreover, we examined the potential of Mettl3 to be a therapeutic target for OP-related bone defects.

| Micro-computed tomography analysis, Haematoxylin and eosin staining and Masson staining
Femurs were dissected from the OVX and control rats 3 months after OVX. The entire femur was scanned using an instrument (SCANCO Medical). Then, femurs samples were stained by haematoxylin and eosin (H&E) and Masson reagents after the decalcification was completed.

| Isolation and culture of BMSCs and OP-BMSCs
A sterile syringe filled in culture medium contained 10% foetal bovine serum (FBS; Schaumburg) was used to repeatedly flush the bone marrow cavity of the femur until the bone marrow cavity turned white and no bone marrow tissue remained. Finally, the culture medium containing bone marrow tissue was cultured in 25 cm 2 culture flasks. The liquid was 50% changed on the first 3 days and then completely changed on day 4. The cells were passaged to third passage for experiments.

| Dot blot of m 6 A RNA modification levels
Briefly, total RNA was collected by TRIzol reagent (Ambion), then a Dynabeads ® mRNA Purification Kit (Thermo Fisher Scientific) was used to extract mRNA, which was heat extracted at 95 C for 3 min.
Then, 2 μl of mRNA was directly dropped onto a Hybond-N+ (Sigma-Aldrich) membrane optimized for nucleic acid transfer. The spotted mRNA was crosslinked to the membrane using a Stratalinker 2400 UV Crosslinker in the autocrosslink mode (1200 μJ [Â100]; 25-50 s). PBS was used to rinse the membrane for 5 min to wash away unbound mRNA. Anti-m 6 A antibody (Synaptic Systems) was incubated with the membrane overnight at 4 C after blocking the bands. Next, the membrane was incubated with goat anti-rabbit IgG-HRP (Signalway Antibody) for 1 h, then submerged in chemiluminescence liquid, and an ECL chemiluminescence detection system (iBrightCL1000, Singapore) was used to obtain images.

| Statistical Analysis
The representative data are provided as mean ± standard deviation (SD). All statistical analyses were performed using SPSS 19.0 software (IBM). It is statistically significant at p < 0.05.  Figure 1E). These results confirmed that OP-BMSCs had self-renewal capacity and were of mesenchymal lineages.

| Decreased osteogenic differentiation potential in OP-BMSCs
We next assessed the bone formation differences of BMSCs and OP-BMSCs by culturing them in osteogenic induction medium. After induction, ALP staining showed reduced ALP activity in OP-BMSCs.
( Figure 2A,C). Alizarin red staining showed that calcium nodules had formed in both cell types, but OP-BMSCs had fewer calcium nodules than BMSCs did ( Figure 2B,C). These results tested that OP-BMSCs had impaired osteogenic ability. Next, we used immunofluorescence staining to detect the level of RUNX2 and OPN in OP-BMSCs and BMSCs. These results showed decreased nuclear RUNX2 expression in OP-BMSCs compared with BMSCs ( Figure 2D,F). Moreover, there was reduced cytoplasmic OPN staining in OP-BMSCs compared with BMSCs ( Figure 2E,F).

| Downregulation of m 6 A methylation level and Wnt signalling pathway in OP-BMSCs
We detected m 6 A methylation in OP-BMSCs and BMSCs using an β-Catenin, Opn and Runx2. Therefore, there was also decreased osteogenic ability ( Figure 4B). Similar results were found after osteogenic induction for 5 days ( Figure 4C). there was more intense ALP staining than after 3 days ( Figure 5B,F). In summary, the methylase METTL3 had a positive bone-related regulatory effect of OP-BMSCs through the canonical Wnt signalling pathway.

| Overexpression of Mettl3 upregulated the osteogenic ability of OP-BMSCs in vivo
In vivo, we implanted BCP with OP-BMSCs into critical-sized calvarial defects of OP model rats. First, the CCK-8 showed that OP-BMSCs could grow and proliferate on BCP ( Figure 6A). Next, we used DAPI staining to confirm that OP-BMSCs were seeded on BCP. The results showed DAPI-stained nuclei on BCP seeded with OP-BMSCs, whilst there was no staining on BCP without OP-BMSCs ( Figure 6B). Furthermore, we used SEM to observe the morphology and growth of OP-BMSCs on BCP. The results showed that OP-BMSCs adhered to BCP after co-culture for 3 days ( Figure 6C). Finally, BCP seeded with OP-BMSCs was transplanted to the critical-sized calvarial defects of OP rats.

| DISCUSSION
OP is a chronic systemic bone disease in postmenopausal women that has become a serious public health problem. 35 Mettl3 could rescue the impaired osteogenic ability of OP-BMSCs.
These results were consistent with in vitro experiments.
In summary, the methylation levels and osteogenic potential of OP-BMSCs were decreased in OP-BMSCs. Overexpression of Mettl3 could promote the osteogenic potential of OP-BMSCs by activating the Wnt signalling pathway. Mettl3 has the potential to be an important regulatory gene related to the osteogenesis of OP-BMSCs, providing a new direction for the treatment of bone defects in OP patients.