Visomitin Attenuates Pathological Bone Loss by Reprogramming Osteoclast Metabolism via the STAT3/LDHB Axis

Abstract

A persistently substantial energy demand and metabolic reprogramming endure throughout the entire course of osteoclastogenesis, accompanied by an intensified oxidative stress. Hence, balancing cellular energy metabolism and maintaining redox homeostasis offer potential for coordinating osteoclastogenesis and bone loss in pathological conditions. In the present study, we have discovered Visomitin, a novel antioxidant that specifically targets mitochondria, which efficiently decreases intracellular reactive oxygen species (ROS) levels, inhibits osteoclastogenesis, and impairs the function of bone resorption. Mechanistically, Visomitin directly targets signal transducer and activator of transcription 3 (STAT3), leading to the inhibition of its transcriptional activity and modulation of lactate dehydrogenase B (LDHB) expression levels, consequently triggering metabolic reprogramming and exerting antagonistic effects on osteoclasts. Furthermore, administration of Visomitin demonstrates marked protective effects against pathological bone loss in vivo. Given its established clinical safety profile in ophthalmologic applications, Visomitin emerges as a promising anti-resorptive agent for clinical translation. This study also unveils the STAT3/LDHB axis as a critical nexus linking mitochondrial redox regulation to osteoclast metabolism, providing a novel therapeutic strategy for osteoclast-driven bone diseases.

Fig. 1.

Visomitin diminishes the intracellular ROS levels and attenuates osteoclastogenesis. (A) The relative antioxidant capacity of indicated antioxidants at 300 nm was evaluated in the ABTS system (n = 3). (B and C) Evaluation and quantification of the intracellular ROS levels of BMMs exposed to H₂O₂ after pretreatment of a range of antioxidants (300 nm) using flow cytometry (n = 3). (D and E) Detection and quantification of the mean fluorescence intensity (MFI) of DCFH-DA probe in BMMs following treatment with either RANKL or Visomitin; scale bars, 100 μm (n = 5). (F and G) Detection and quantification of the MFI of MitoSOX probe in BMMs following treatment with either RANKL or Visomitin; scale bars, 50 μm (n = 5). (H) BMMs were treated with different dosages of Visomitin and subjected to in vitro osteoclast differentiation. Representative images of TRAP staining were shown. Scale bars, 50 μm. (I) Quantification of TRAP⁺ multinuclear cells per well in panel (A) (n = 3). (J) BMMs were subjected to in vitro osteoclast differentiation and treated with 300 nm Visomitin at specified stages. Representative images of TRAP staining were shown. Scale bars, 50 μm. (K) Quantification of TRAP⁺ multinuclear cells per well in panel (C) (n = 3). (L) Representative images of wheat germ agglutinin (WGA) staining in osteoclasts treated with or without Visomitin. Scale bars, 10 μm. (M) Quantification of bone pit depth in panel (I) (n = 12). (N) Representative SEM images of bone slice resorption pits. Scale bars, 10 μm. (O) Quantification of bone resorption pit area (n = 6). Data are mean ± SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.

Fig. 2.

Administration of Visomitin alleviates pathological bone loss in vivo. (A) Representative 3D micro-CT images of the calvaria from mice subjected to either sham or LPS injection, followed by treatment with PBS or Visomitin. Scale bars, 2 mm. (B) Quantification of BV/TV (%) in panel (A) (n = 5). (C) Representative TRAP and DHE staining of the calvaria from designated groups. Scale bars, 200 μm and 100 μm, respectively. (D to F) Quantification of N.Oc/BS (mm⁻¹), Oc.S/BS (%), and relative DHE MFI in panel (C) (n = 5). (G) Representative 3D micro-CT images of the femurs from mice subjected to either sham or OVX operation, followed by treatment with PBS or Visomitin. Scale bars, 500 and 200 mm, respectively. (H) Quantification of BV/TV (%), Tb.N (mm⁻¹), Tb.Th (mm), and Tb.Sp (mm) in panel (G) (n = 6). (I) Representative TRAP and DHE staining of the femurs from designated groups. Scale bars, 50 μm. (J to L) Quantification of N.Oc/BS (mm⁻¹), Oc.S/BS (%), and relative DHE MFI in panel (I) (n = 6).Data are mean ±SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.

Fig. 3.

Visomitin attenuates the activation of RANKL-RANK signaling pathways. BMMs treated with Visomitin or PBS were subjected to osteoclast differentiation, followed by Transcriptome RNA-seq. Genes with |log₂FC| > 1, P < 0.05, and TPM > 0.5 are designated as differentially expressed genes (DEGs) (n = 3). (A) The heatmap illustrating the gene expression profiles derived from RNA-seq data. (B) Analysis of RNA-seq data for cell and tissue specificity utilizing the PaGenBase database. (C) The volcano plot illustrating the gene expression profiles derived from RNA-seq data. (D) KEGG enrichment analysis of DEGs obtained from RNA-seq data. (E) GO enrichment analysis of down-regulated genes (Log₂FC < −1 and P < 0.05) obtained from RNA-seq data. (F) GSEA of Gene Ontology Biological Processes in RNA-seq data. (G) Network of enrich terms derived from RNA-seq data utilizing the Metascape database. (H) Representative immunoblots illustrating the effects of Visomitin on the activation of RANKL-RANK signaling pathways, including NF-κB, MAPK, and AKT pathways (n = 3). Data are mean ± SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.

Fig. 4.

Visomitin reprograms energy metabolism during osteoclastogenesis. (A) GSEA analysis of Reactome and KEGG pathways in RNA-seq data. (B) The heatmap depicting the gene expression profiles of the 2 pathways presented in panel (A). (C) Representative JC-1 staining of BMMs treated with MCSF, RANKL, or Visomitin as indicated. (D) Quantification of JC-1 staining as presented in panel (C) (n = 5). (E) Intracellular ATP levels following treatment with MCSF, RANKL, or Visomitin as indicated (n = 5). (F) Representative immunoblots illustrating the expression levels of proteins within the aforementioned 2 pathways following Visomitin treatment (n = 3). (G and H) Representative Seahorse graphs for oxygen consumption rate (OCR) or extracellular acidification rate (ECAR) (n = 3). (I) Quantification of the basal respiration, maximal respiration, and ATP production from OCR data. (J) Quantification of the non-glycolytic acidification, glycolysis, and glycolytic capacity from ECAR data. Data are mean ± SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.

Fig. 5.

Visomitin regulates osteoclastogenesis through the LDHB–lactate axis. (A) The heatmap depicting the gene expression profiles of metabolic pathways. (B) Representative immunoblots depicting the protein levels of LDHA and LDHB following Visomitin treatment (n = 3). (C) Immunofluorescence staining of LDHB in BMMs subjected to osteoclast differentiation, with or without Visomitin treatment; scale bars, 20 μm. (D) Quantification of the relative LDHB MFI in panel (C) (n = 6). (E) BMMs infected with either the vector or LDHB-overexpressing adenovirus were differentiated into osteoclasts in the presence or absence of Visomitin treatment. Representative images of TRAP staining were shown. Scale bars, 50 μm. (F) Quantification of TRAP⁺ multinuclear cells per well in panel (E) (n = 3). (G) The classification of metabolites detected through metabolomics. (H and I) The volcano plot and heatmap illustrating the metabolite affinity profiles derived from metabolomics. (J and K) Enrichment analysis of differential metabolites detected by metabolomics using SMPDB and KEGG databases. Data are mean ± SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.

Fig. 6.

STAT3 functions as a direct target of Visomitin to modulate LDHB transcription. (A) The potential transcription factors (TFs) for LDHB were predicted using the KnockTF, ENCODE, and ChIP_Atlas databases. (B) The potential targets of Visomitin were predicted using the SuperPRED database. (C and D) The mRNA and protein expression levels of LDHB under Stattic treatment (n = 3). (E) The thermal stability of FLAG-STAT3 under Visomitin treatment was detected using WB (n = 3). (F) The stability of FLAG-STAT3 in the presence of protease following treatment with Visomitin (0, 75, 150, and 300 nmol/l) was detected using WB (n = 3). (G) Three-dimensional image of molecular docking between Visomitin and STAT3. (H) Representative immunoblots for the indicated nuclear, phosphorylated, or total proteins following treatment with Visomitin (0, 75, 150, and 300 nm) (n = 3). (I) Representative Immunofluorescence images of STAT3 in BMMs after treatment with RANKL or Visomitin (0, 75, 150, and 300 nm) as indicated. Scale bars, 20 μm. (J) Quantification of Pearson’s correlation coefficient between STAT3 and DAPI in panel (I) (n = 3). (K) STAT3 ChIP assay of LDHB promoter region. (L) Quantification of the binding affinity between STAT3 and the LDHB promoter (n = 3). Data are mean ± SD; *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant.