Bone remodeling stimulated by Wnt-mediated mitophagy regulated extracellular vesicles in subchondral bone contributes to osteoarthritis development

Abstract

Rationale: Osteoarthritis (OA) is increasingly understood as a disease involving not only cartilage degeneration but also pathological subchondral bone remodeling. The contribution of osteoblast (OB) heterogeneity and their secreted extracellular vesicles (EVs) to this process remains poorly characterized. This study aims to investigate how EVs from distinct OB subtypes modulate subchondral bone remodeling and contribute to OA progression. Methods: OB subtypes representing endothelial (EnOBs), stromal (StOBs), and mineralizing (MinOBs) stages were generated by time-controlled osteogenic induction of BMSCs. EVs were isolated from each OB subtype and characterized by TEM, Western blot, DLS, and miRNA profiling. Functional assays included osteogenic induction, angiogenesis, and cartilage degradation analyses in vitro. RNA-seq and qRT-PCR were used to identify relevant signaling pathways and miRNAs. In vivo effects of EVs were tested in a DMM-induced OA mouse model using intravenous injections, followed by histology, micro-CT, and immunostaining. Results: EVs derived from different OB subtypes exhibited distinct pro-osteogenic, pro-angiogenic, and cartilage-degrading effects. MinOB-derived EVs significantly enhanced osteogenic differentiation and mineralization, correlated with enrichment of calcium phosphate content and specific pro-osteogenic miRNAs. These EVs also carried amorphous calcium phosphate and mitochondrial content, linked to activated mitophagy. Wnt signaling dynamically regulated mitophagy and EV composition, particularly in MinOBs. In vivo, tail vein administration of OB-derived EVs exacerbated subchondral bone sclerosis and cartilage degradation in a time-dependent manner, with MinOB-EVs inducing the most pronounced pathological changes. Conclusions: OB-derived EVs exhibit subtype-dependent regulatory functions in subchondral bone remodeling, mediated by distinct miRNA profiles and mineral cargo shaped by Wnt-regulated mitophagy. These EVs actively participate in OA progression, and their effects vary with disease stage and route of administration. Targeting specific OB subtypes or modulating Wnt-mitophagy signaling may offer novel therapeutic strategies for stage-specific OA intervention.

Keywords: extracellular vesicles; mitophagy; osteoarthritis; osteoblast; subchondral bone.

Figure 1

Identification of OB subtypes and the functions to chondrocytes, BMSC and HUVECs. (A) GO enrichment analysis of osteoblast subtypes in subchondral bone from OA patients by scRNA-seq. (B-C) Western Blot and qRT-PCR analysis validated the osteogenic genes expressed in EnOBs, StOBs and MinOBs. (D-E) ARS and ALP staining of OB subtypes and relative quantitative analysis. Scale bar: 200 μm. (F) Schematic showing the strategy for chondrocytes, BMSC and HUVECs treated with culture medium (CM) form three OB subtype. (G-H) ARS and ALP staining of BMSC after treated with CM for 5 days and 9 days respectively and relative quantitative analysis. Scale bar: 200 μm. (I-J) Western blot and qRT-PCR were used to analyze the expression levels of RUNX2, Osterix, BMP-2, and osteocalcin (OCN) in Bone Marrow Stromal Cells (BMSC) treated with Conditioned Media (CM) and subjected to osteogenic induction over periods of 3 and 7 days. (K) Western blot analysis of the protein level in chondrocytes treated with CM for 24 h. (I) qRT-PCR analysis of the mRNA level of SOX9, Col-2, MMP13 and Adamts5. (M) Alcian blue staining of chondrocytes and ROS localization in chondrocytes treated with CM for 24 h. anova: 200 μm; 500 μm. (N) The formation of tubes by HUVECs was assessed following a 6-hour incubation on Matrigel. Images of migrating cells were captured through crystalline violet staining. Scale bar: 500 μm. (O) Statistical analysis of the numbers of formed tubes and migration rate of HUVECs. Differences in mean values were evaluated for statistical significance using Tukey's HSD multiple comparisons test at a significance level of p ≤ 0.05. The results, depicted graphically with representative images, are based on three independent replicates, and error bars represent the mean ± SD. Means sharing the same letters (ns) are not considered significantly different. The following significance levels were assigned: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. Quantitative data are presented as mean ± SD.

Figure 2

Osteoblast subtypes-derived extracellular vesicle promote osteogenesis. (A) A diagrammatic representation outlining the method used for extracting three different types of EVs. (B) Representative TEM images of three EVs. Scale bars: 100 nm. (C) Analysis of the size distribution of the EVs, as determined by DLS. (D) Identification of specific marker proteins present within the EVs. (E) EVs were endocytosed into the BMSCs after co-cultured for 6 h located by DID (red) and F-actin (green). Scale bar: 10 μm. (F-G) Western blot analysis and qRT-PCR analysis of the expression level of osteogenic markers in BMSC treated with EVs after osteogenic induced for 3 days. (H) ARS and ALP staining of BMSC after treated with EVs for 5 days and 9 days respectively. Scale bar: 200 μm. (I-J) Western blot analysis and qRT-PCR analysis of the expression level of osteogenic markers in BMSC co-cultured with GW4869-treated EVs after osteogenic induced for 3 days. (K-L) ARS and ALP staining of BMSC with GW4869-treated EVs for 5 days and 9 days respectively and relative quantitative analysis. Scale bar: 200 μm.

Figure 3

Functions of miRNAs in osteoblasts subtypes-derived EVs. (A) GO analysis of pathways and biological processes for differentially expressed miRNAs in three EVs. (B) KEGG analysis of pathways and biological processes for differentially expressed miRNAs in three EVs. (C) Venn analysis between differentially expressed miRNAs in three EVs. (D) Heatmaps for intersect 34 miRNAs in three EVs and unique 10 miRNAs in Min-EVs. (E) GO enrichment analysis suggests potential functions for the target genes of the 34 intersecting miRNAs. (F) GO enrichment analysis indicates possible functions of target genes associated with 10 miRNAs found in Min-EVs. (G) Analysis of KEGG pathways reveals the top 10 pathways linked to the target genes of the 34 intersecting miRNAs. (H) KEGG pathway analysis identifies the top 10 pathways associated with the target genes of 10 miRNAs in Min-EVs. (I) Bar plot showed the mRNA expression level 10 miRNAs in Min-EVs.

Figure 4

Intramitochondrial electron-dense amorphous calcium phosphate (ACP) granules contained in EVs. (A) TEM images show electron-dense granules in EVs. (B) SAED revealed the amorphous nature of granules. (C) Elemental mapping of granules in EVs demonstrated the presence of calcium and phosphorus (bar: 500 nm). (D) FT-IR analysis of ACP and three EVs. (E) TEM images show electron-dense granules in OB subtypes released EVs. (F) Flow cytometric analysis of Mitotracker Green-labeled mitochondria and DID-labeled cell membrane in EVs. (G) Western blotting analysis of mitochondrial maker proteins in EVs. (H) Confocal microscopy images depict various OBs subtypes marked with Mito-tracker Green (green) for mitochondria, Lyso-tracker Red (red) for lysosomes, and DAPI (blue) for nuclei. (n = 3). Scale bar: 10 μm. (I) ATP activity in three OB subtypes showed the activation of mitochondria. (J) High-magnification TEM images displayed intramitochondrial ACP granules (red arrow) within mitochondria (green arrow), and EDX elemental mapping confirmed the presence of calcium and phosphorus (bar: 1/50 nm).

Figure 5

Relationships between Wnt pathway and mitophagy in OB subtypes. (A) Genes shown by heat map of top 40 DEGs in three OB subtypes. (B-C) Relative mRNA expression level and protein level of Wnt proteins and autophagy-related genes showed the dynamic changes in Wnt pathway and autophagy activity. (D) KEGG analysis of unique different expressed genes in OBs subtypes. (E) TEM images and EDX elemental mapping show no electron-dense granules in Mdivi-1 and Wnt agonist-1 treated-EVs. (F) miRNA expression level of 10 miRNAs in Mdivi-1 and Wnt agonist-1 treated Min-EVs (n = 3). (G-H) The expression levels of osteogenic maker genes in BMSC after treated with Mdivi-1 and Wnt agonist-1 treated-EVs.

Figure 6

EVs aggravated the subchondral bone remodeling in DMM mice at late administration stage. (A) Schematic showing the strategy for DMM surgery and sample processing. EVs were injected into the tail vein every week after DMM surgery, and histological analysis was performed. H&E staining images (B), and SO&FG staining (C) after 8 weeks of DMM modeling revealed distinct patterns (scale bar: 200 μm, n = 6 per group). (D) Representative 3D reconstruction of the medial tibial plateau subchondral bone sagittal plane. Scale bar: 500 μm. n = 6 per group. (E) Immunohistochemical staining of OCN (green) and CD31 (red). (F) in subchondral bone were quantified and presented. n = 4 per group. Scale bar: 100 μm. (G) Quantification of articular cartilage histological scores on the tibial plateau using OARSI criteria. n = 6 per experimental group. (H) Using Micro-CT analysis, subchondral bone parameters such as subchondral bone volume to total tissue volume (BV/TV) and subchondral bone plate thickness (SBP.th) were evaluated in the medial tibial plateau. n = 6 per group. (I) Quantification of OCN positive cells/mm². n = 4 per group. Differences in mean values were evaluated for statistical significance using two-way ANOVA followed by Tukey's HSD multiple comparisons test at a significance level of p ≤ 0.05. The results, depicted graphically with representative images, are based on three independent replicates, and error bars represent the mean ± SD. Means sharing the same letters (ns) are not considered significantly different. The following significance levels were assigned: *p < 0.05, **p < 0.01, ***p < 0.001 and **** p < 0.0001. Quantitative data are presented as mean ± SD.

Figure 7

Osteoblasts subtye-derived EVs aggravated subchondral bone remodeling in OA progression. Regulated by Wnt pathway, osteoblasts were heterogeneous in osteogenic differentiation and the activation of mitophagy will provide calcium phosphate for further mineralization. EVs from three OB subtypes inheriting the heterogeneity could promote osteogenic with functional miRNAs and calcium phosphate. Osteoblasts subtypes-derived EVs could accelerate the subchondral bone remodeling in OA progression, resulting in bone sclerosis and cartilage degeneration. Created by Biorender.