### Generation of conditional Alkbh5-knockout mice

Alkbh5fl/+ mice with C57BL/6 background were generated by Cyagen (Suzhou, China) using /Cas-mediated genome engineering. Briefly, the Alkbh5 gene (NCBI Reference Sequence: NM_172943, Ensembl: ENSMUSG00000042650) is located on mouse chromosome 11. Exon 1 was selected as the conditional knockout region. To engineer the targeting vector, homologous arms and the Conditional knockout (CKO) region were generated by PCR using the BAC clone RP23-329 M3 as the template. Cas9, guide RNA, and the targeting vector were coinjected into fertilized eggs for CKO mouse production. The pups were genotyped by PCR followed by sequencing analysis.

Prx1-Cre transgenic mice were purchased from The Jackson Laboratory. We crossed Prx1-Cre mice with Alkbh5fl/+ mice to obtain Prx1-Cre; Alkbh5fl/+ mice as heterozygous conditional Alkbh5-knockout mice. By mating Prx1-Cre; Alkbh5 fl/+ male mice with Alkbh5fl/fl female mice, we obtained Prx1-Cre; Alkbh5fl/fl mice as homozygous conditional Alkbh5-knockout mice. The genotype of the transgenic mice was identified by PCR analyses of genomic DNA extracted from mouse tails. Primers for floxed Alkbh5-knockout allele genotyping were as follows: forward (5′-CAGGTTTGAAGTGGCCATAGTAGC-3′) and reverse (5′-GAGGCCAAGACAGGAGAATCAGAC-3′). Primers for Cre transgene genotyping were as follows: forward (5′-GCTCTGATGTTGGCAAAGGGGT-3′) and reverse (5′-AACATCTTCAGGTTCTGCGGG-3′). All mice were bred and maintained under specific pathogen-free conditions. All procedures involving animals were approved by the Animal Use and Care Committee of the Eighth Affiliated Hospital of Sun Yat-sen University.

### MSC isolation and expansion

This study was approved by the Ethics Committee of Sun Yat-sen Memorial Hospital of Sun Yat-sen University, Guangzhou, China. Recruited healthy volunteers aged 20–30 years were fully informed of the relevant precautions and potential risks of bone marrow extraction, and signed the informed consent form. Bone marrow was collected from the posterior superior iliac spine of the volunteers and the MSCs were separated by density gradient centrifugation. The MSCs were resuspended in Dulbecco’s modified Eagle’s medium (DMEM, 1000 mg/L glucose, Gibco) containing 10% fetal bovine serum (FBS, Hangzhou Sijiqing Biological Engineering Material Company, Limited) and placed in a cell incubator containing 5% CO2 at 37 °C. After 48 h, the medium was changed to remove nonadherent cells. After 3 days, the medium was changed once. When the cell density reached 80–90%, the cells were digested with trypsin containing 0.53 mM EDTA and re-seeded in new flasks at passage 1. Cells from passage 3 to passage 5 were used for experiments.

### Osteogenic differentiation induction

MSCs were seeded in a 12-well plate at a density of 0.6 × 105/well. After 12 h, when the cells adhered to the wells, the culture medium of the MSCs was changed to osteogenic differentiation medium consisting of DMEM (1000 mg/L glucose) with 10% FBS, 100 IU/mL penicillin, 100 IU/mL streptomycin, 0.1 μM dexamethasone, 10 mM β-glycerol phosphate, and 50 μM ascorbic acid (Sigma-Aldrich). The medium was changed every 3 days and continued to be induced to the desired time point.

### ARS staining and quantification

The original medium was discarded from the well plate, 500 μl of 4% paraformaldehyde was added to each well and the plate was placed at room temperature for 30 min to fix the cells. After paraformaldehyde fixation, the plate was rinsed twice with phosphate-buffered saline (PBS), 500 μl of 1% ARS (pH 4.3) was added, and staining was performed for 20 min. Then, PBS was added to rinse the plate and the central field of view of the well plate was selected under the microscope to be photographed. For ARS quantification, 10% cetylpyridinium chloride monohydrate (Sigma-Aldrich) was used to destain the cells for 1 h at room temperature. Then, 200 μl of the liquid was transferred to a 96-well plate and the spectrophotometric absorbance was measured at 562 nm.

### ALP staining and activity assay

ALP staining was performed according to the instructions of the BCIP/NBT Alkaline Phosphatase Color Development Kit (Beyotime Institute of Biotechnology). Briefly, MSCs were fixed in 4% paraformaldehyde for 15 min and then stained with mixture solution at 37 °C in the dark for 15 min. The central field of view of the well plate was selected to be photographed under the microscope.

For the ALP activity assay, an ALP assay kit (Nanjing Jiancheng Bioengineering Institute, China) was used. Then, 80 μl of radioimmunoprecipitation assay (RIPA) buffer was added to lyse cells in a well plate, followed by centrifugation at 14,000 r.p.m. for 10 min. A 96-well plate was prepared and 50 μl of A/1 solution and 50 μl of B/2 solution were added. Then, 30 μl of deionized water was added to the blank well, 30 μl of control solution was added to the standard well, and 30 μl of sample supernatant was added to the sample well. Then, the 96-well plate was placed in a 37 °C incubator for 15 min. After adding stop solution, the absorbance was measured at 405 nm. In addition, the protein concentration of the sample was determined by the Bicinchoninic acid (BCA) method, and finally, the ALP activity was calculated based on the measured ALP absorbance (optical density, OD) value and protein concentration; the unit of ALP activity was unit/g pro/15 min.

### Western blotting

RIPA solution containing phosphatase inhibitor and protease inhibitor was added to lyse the cells. After measuring the protein concentration by the BCA method, equal amounts of each sample were diluted in 5× sodium dodecyl sulfate (SDS) loading buffer and denatured by boiling. Protein lysates were separated by SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore). The membranes were blocked in 5% nonfat dry milk dissolved in TBST (150 mM NaCl, 50 mM Tris-HCl pH 7.5, and 0.05% Tween-20) at room temperature for 1 h and then incubated overnight with primary antibodies against ALKBH5 (1 : 1000, ab195377, Abcam), total AKT (1/5000, ab179463, Abcam), p-AKT (1 : 1000, ab38449, Abcam), SP7 (1 : 1000, ab209484, Abcam), RUNX2 (1 : 1000, ab23981, Abcam), PRMT6 (1 : 1000, 14641 S, Cell Signaling Technology), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 1 : 3000, AF0006, Beyotime) overnight at 4 °C. The protein signals were detected using chemiluminescent reagents (Millipore) according to the manufacturer’s instructions.

### Quantitative reverse-transcriptase PCR

Total RNA from MSCs was extracted using TRIzol (Invitrogen) reagent. The RNA concentration was measured with a NanoDrop 2000 (Thermo Fisher Scientific). cDNA was transcribed by using PrimeScript RT Master Mix (Takara). Quantitative reverse-transcriptase PCR was performed using SYBR Green Premix Ex Taq (TaKaRa) in a Light-Cycler® 480 PCR System (Roche). Relative gene expression was normalized to GAPDH expression using the 2−ΔΔCt method. The primers are listed in Supplementary Table 1.

### Cells transfection and chemical inhibition

The siRNAs were designed and synthesized by GenePharma (Shanghai, China). Three interference sequences were designed for the target gene. The interference efficiency was verified by qPCR and western blotting methods. Two sequences with a knockdown efficiency of >70% were selected for further experiments. Details of the sequences are shown in Supplementary Table 2. Transfection was carried out according to the instructions. After transfection, osteogenic induction medium was added to induce the induction of MSCs into osteoblasts. On the seventh day of osteoinduction, siRNA was transfected again for a second knockdown. The overexpression lentivirus was constructed by OBiO Technology (Shanghai, China). Mut-ALKBH5 contains a mutation of histidine to alanine at position 204. MSCs were incubated with the lentiviruses for 24 h at a multiplicity of infection of 30.

LY294002 and SC79 were purchased from Calbiochem and dissolved into dimethyl sulfoxide. MSCs were treated with 10 μM LY294002 and 4 µg/ml SC79 after siRNA or lentivirus transfection.

### m6A dot blot

Total RNA was extracted from the cells using TRIzol and the RNA concentration was determined by a NanoDrop 2000 (Thermo Fisher Scientific). The RNA concentration of different samples was adjusted to 250 and 500 ng/µl. Two microliters of RNA was spotted onto the nylon membrane (Sigma-Aldrich, GERPN1210B), followed by 1500 J ultraviolet crosslinking twice. One of the membranes was dyed in methylene blue solution for 30 min. After 30 min, the membrane was rinsed twice with ultrapure water and photographed. The other membrane was blocked with 5% nonfat dry milk dissolved in PBST (PBS with 0.1% Tween-20) for 1 h, followed by incubation with m6A antibody (1 : 1000, 202003, Sysy) for 14–16 h. Then, PBST was added to wash the membrane for 5 min, which was repeated three times. An horseradish peroxidase-conjugated AffiniPure goat anti-mouse IgG dilution (1 : 3000, BA1050, BOSTER) was added and the membrane was incubated at room temperature for 1 h. Then, PBST solution was added again to rinse the membrane three times. The protein signals were detected using chemiluminescent reagents (Millipore) according to the manufacturer’s instructions.

### m6A RNA methylation assay

This experiment was performed using the m6A RNA Methylation Assay Kit (ab185912, Abcam) according to the manual. In short, TRIzol was used to extract total RNA from the cells, the concentration was adjusted to 100 ng/μl after determining the RNA concentration, and 2 μl of RNA, 2 μl of negative control, and 2 μl of positive control were added to the well plate. Then, 100 μl of Developer Solution was added and the plate was incubated at room temperature for 6 min in the dark. Then, 100 μl of Stop Solution was added to stop the enzyme-linked reaction. The absorbance value was measured in a microplate reader (450 nm). The overall m6A methylation level was calculated according to the absorbance value and the formula used was as follows:

$${\mathrm{m}^6{\mathrm{A}}}\% = \frac{{\left( {\mathrm{Sample}}\;{\mathrm{OD}} – {\mathrm{NC}}\;{\mathrm{OD}} \right) \div {\mathrm{S}}}}{( {\mathrm{PC}}\;{\mathrm{OD}} – {\mathrm{NC}}\;{\mathrm{OD}}) \div {\mathrm{P}}} \times 100\%$$

where Sample OD is the OD value of sample well, NC OD is the OD value of negative control well, PC OD is the OD value of positive control well, S is total mass of RNA added to sample well, and P is the total mass of the positive control added to the positive control well.

### m6A-seq and RNA-seq

After using siRNA to interfere with the expression of ALKBH5 in MSCs, osteogenic differentiation was induced for 3 days. Total RNA was extracted using TRIzol reagent (Invitrogen). Approximately 50 µg of total RNA was subjected to isolation of poly(A) mRNA with poly-T oligo-attached magnetic beads (Invitrogen). After fragmentation, RNA was incubated with an m6A-specific antibody (No. 202003, Synaptic Systems, Germany) for immunoprecipitation. Then, eluted m6A-containing fragments and untreated input control fragments were converted to the final cDNA library in accordance with strand-specific library preparation by the dUTP method. Paired-end 2 × 150 bp sequencing was performed on an Illumina NovaSeq™ 6000 platform at LC-BIO Biotech, Ltd (Hangzhou, China).

### meRIP-qPCR and RIP-qPCR

The Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (Millipore) was used for the RIP assay. Briefly, 1 × 107 cells were collected by RIP lysis buffer and then incubated with RIP buffer containing magnetic beads conjugated to anti-m6A (202003, sysy) and anti-ALKBH5 (ABE547, Merck Millipore) antibodies. Purified rabbit IgG was used as a negative control. Then, immunoprecipitated RNAs were isolated and purified for qPCR analysis to detect the presence of the target mRNA.

### Nuclear and cytoplasmic fractionation

This experiment was performed according to the instructions of the PARIS™ Kit (AM1921, Thermo Fisher). The cells were collected after digestion with trypsin. Cell fractionation buffer was added to the cells and the cells were then centrifuged. Then, Cell Disruption Buffer was added to lyse the nuclei. Then, the mixture was transferred to a filter column, preheated elution solution was added to dissolve the RNA after centrifugation, and the RNA concentration was measured. After reverse transcription to cDNA, qPCR was used to detect the expression of the target gene.

### RNA decay assay

After MSCs were transfected with siRNA, they were added to osteogenic induction medium and cultured for 3 days. Subsequently, 5 μg/ml actinomycin D was added. Total RNA at 0, 1, 2, and 4 h was extracted using TRIzol reagent (Invitrogen) and the relative expression was detected by qPCR according to a previously described method39.

### µCT and histomorphometric analyses

We selected 6-month-old male mice for bone mass analysis. µCT was used to analyze the bone structure of the femur. The collected bone tissues were fixed in 4% polyoxymethylene for 2 days and then stored in 70% ethanol at 4 °C before being processed. To analyze the trabecular bone, images were acquired at an effective pixel size of 9.56 μm, a voltage of 80 kV, a current of 500 μA, and an exposure time of 1500 ms in each of the 360 rotational steps. The BV/TV, bone surface area/BV, Tb. Th, Tb. N, and trabecular spacing were analyzed according to the guidelines40. Two-dimensional and three-dimensional bone structure image slices were reconstructed.

### Immunofluorescence staining

MSCs were seeded on sterile glass coverslips and osteogenic differentiation was induced for 10 days. Cells were fixed in 4% paraformaldehyde for 30 min and then 0.1% Triton X-100 was added for 15 min at room temperature. Normal goat serum (5%) was used to block cells for 30 min. A primary antibody against collagen I (1 : 500, Abcam, ab34710) was added and the cells were incubated overnight at 4 °C. Then, anti-rabbit IgG (1 :5 00, Cell Signaling Technology, 4413) was added and incubated for 1 h at room temperature. 4′,6-Diamidino-2-phenylindole (DAPI) was used to counterstain the nuclei. Thereafter, the samples were viewed under a laser scanning confocal microscope at wavelengths of 555 nm (red, collagen I) and 405 nm (blue, DAPI).

### Statistical analysis

The experiments in this study were independently repeated at least three times. SPSS 18.0 software (SPSS, Chicago, IL, USA) was used to analyze the experimental data and the results are expressed as the mean ± SD. Student’s t-test and one-way analysis of variance followed by the Bonferroni test was performed for statistical analyses. The bone mass of mice in the CKO and wild-type (WT) groups was compared using an unpaired two-tailed Student’s t-test. PPP