Unraveling Key m6A Modification Regulators Signatures in Postmenopausal Osteoporosis through Bioinformatics and Experimental Verification

Abstract

Objective: Bone marrow mesenchymal stem cells (BMSCs) show significant potential for osteogenic differentiation. However, the underlying mechanisms of osteogenic capability in osteoporosis-derived BMSCs (OP-BMSCs) remain unclear. This study aims to explore the impact of YTHDF3 (YTH N6-methyladenosine RNA binding protein 3) on the osteogenic traits of OP-BMSCs and identify potential therapeutic targets to boost their bone formation ability.

Methods: We examined microarray datasets (GSE35956 and GSE35958) from the Gene Expression Omnibus (GEO) to identify potential m⁶A regulators in osteoporosis (OP). Employing differential, protein interaction, and machine learning analyses, we pinpointed critical hub genes linked to OP. We further probed the relationship between these genes and OP using single-cell analysis, immune infiltration assessment, and Mendelian randomization. Our in vivo and in vitro experiments validated the expression and functionality of the key hub gene.

Results: Differential analysis revealed seven key hub genes related to OP, with YTHDF3 as a central player, supported by protein interaction analysis and machine learning methodologies. Subsequent single-cell, immune infiltration, and Mendelian randomization studies consistently validated YTHDF3's significant link to osteoporosis. YTHDF3 levels are significantly reduced in femoral head tissue from postmenopausal osteoporosis (PMOP) patients and femoral bone tissue from PMOP mice. Additionally, silencing YTHDF3 in OP-BMSCs substantially impedes their proliferation and differentiation.

Conclusion: YTHDF3 may be implicated in the pathogenesis of OP by regulating the proliferation and osteogenic differentiation of OP-BMSCs.

Keywords: BMSCs; Osteogenic differentiation; Postmenopausal osteoporosis; Proliferation; YTHDF3.

FIGURE 1

Expression and correlation of 20 m⁶A regulators in osteoporosis. (A, B) The box plot and heatmap plot show the summary of 20 m⁶A regulators in bone marrow mesenchymal stem cells (BMSCs) from osteoporosis patients and non‐osteoporotic donors, and seven m⁶A regulators (ALKBH5, YTHDC2, YTHDF2, YTHDF3, FMR1, and HNRNPA2B) were significantly dysregulated. (C) ROC curves validated the performances of 7 m⁶A regulators for the prediction of OP in GSE35958 datasets. (D) The co‐expression network of 20 m⁶A regulators. *p < 0.05; **p < 0.01; ***p < 0.001.

FIGURE 2

Hub genes selection through machine learning: (A) A total of 101 prediction models were calculated through the leave‐one‐out cross‐validation (LOOCV) framework, and further calculated the AUC index of each model on all validation datasets; (B) Hub gene scoring chart, showing the average ranking of different features (Features) in different prediction models; X‐axis (average rank): Represents the average ranking of features in all models. Y‐axis (Feature): represents various features.

FIGURE 3

Construction of the protein‐protein interaction (PPI) network of 20 m⁶A regulators and hub genes related to Nomogram to assess clinical value. (A–D) The hub genes were extracted from the PPI network through Degree, EPC, MNC, and MCC algorithms of cytoHubba. (E) Four algorithms obtain a Venn diagram of common hub genes. (F) Decision curve analysis to evaluate the sensitivity and specificity of the hub genes related to Nomogram. (G) Calibration curves to assess the degree of similarity between the predicted and true results of hub genes related to Nomogram. (H) Nomogram demonstrates the prognostic value of YTHDF3, METTL3, METTL14, ALKBH5, and YTHDC2 for OP patients.

FIGURE 4

Correlation analysis between YTHDF3 and OP: (A) Single‐cell analysis, t‐SNE expression chart: Shows t‐SNE cell clustering based on the expression of the YTHDF3 gene. (B) t‐SNE clustering chart: This chart shows the t‐SNE clustering results of cells. Each color represents a different cell group. CH1, cluster of differentiation 1; CH2, cluster of differentiation 2; EMP, endothelial progenitor cell; IMP, immature myeloid progenitor; LCP, leukocyte common precursor; LMP, lymphoid myeloid progenitor; OB, osteoblast; Ocy, osteocyte; MALP, mesenchymal stem cell airway lineage precursor. (C) Immune infiltration correlation chart: Displays the correlation strength of different cell types with YTHDF3 expression. (D) MR test chart: Describes the SNP effect size estimates of different MR methods and their confidence intervals. Each point represents a specific SNP, and the lines show the corresponding confidence intervals. The three colored lines represent the average effect estimates of three different MR methods.

FIGURE 5

YTHDF3 expression in osteoporosis (OP) patients and OVX mice. (A) Representative images obtained by digital radiography imaging and staining with hematoxylin and eosin of femoral head tissues obtained from patients with OP. (B) The protein levels of PCNA, RUNX2, Osterix, and YTHDF3 in control and OP patients (n = 3). (C) 2D and 3D micro‐CT imaging were employed to analyze the distal femurs. BMD, BV/TV, Tb.th, Tb.Sp, and Tb.N were evaluated (n = 3). (D) Histological analysis of mouse distal femur tissue sections stained with H&E and Masson staining (n = 3). (E) The protein levels of PCNA, RUNX2, Osterix, and YTHDF3 in SHAM and OVX mice (n = 3). (F) IHC images showed the expression of RUNX2, Osterix, PCNA, and YTHDF3 in SHAM and OVX mice (n = 3). (G) Immunofluorescence staining for YTHDF3 and RUNX2 in femurs of sham and OVX mice, with DAPI (blue), YTHDF3 (red), RUNX2 (green), and merged images (n = 3). Data presented as means ± SD (**p < 0.01, ***p < 0.001).

FIGURE 6

Characterization of bone marrow‐derived mesenchymal stem cells (BMSCs). (A) Cell surface marker expression was assessed using flow cytometry. (B) Cell morphology of BMSCs at the second and third passages (P2, P3) as observed under light microscopy. (C, D) The osteogenic and adipogenic differentiation potential of BMSCs was evaluated by alizarin red staining (ARS) (C) and oil red O staining (D), respectively.

FIGURE 7

YTHDF3 regulates proliferation and osteogenic differentiation in OP‐BMSCs. (A, B) The mRNA and protein expression levels of YTHDF3 were assessed in SHAM and OP‐BMSCs using (A) qRT‐PCR and (B) Western blot analysis. (C–F) Knockdown and overexpression of YTHDF3 were conducted in OP‐BMSCs, and their efficiency was validated through qRT‐PCR and Western blot analysis. (G) Following osteogenic induction, ALP and ARS staining were conducted in both si‐NC and si‐YTHDF3 groups. (H) Following YTHDF3 knockdown, the protein expression levels of RUNX2, Osterix, and PCNA were diminished. (I) Following YTHDF3 overexpression in OP‐BMSCs, the protein expression level of RUNX2 exhibited augmentation. (J) The knockdown of YTHDF3 resulted in diminished proliferation activity of BMSCs, as evidenced by EdU staining. (K) Immunostaining analysis was conducted to assess the protein expression of YTHDF3, RUNX2, and PCNA. Data presented as means ± SD (**p < 0.01, ***p < 0.001, ns, no significance).

FIGURE 8

The schematic demonstrates YTHDF3's role in regulating the proliferation and osteogenic differentiation of BMSCs during postmenopausal osteoporosis. As osteoporosis progresses, YTHDF3 expression is reduced, consequently inhibiting BMSC proliferation and osteogenic differentiation through the downregulation of PCNA, RUNX2, and Osterix.