Abstract
Aims
This study explored the shared genetic traits and molecular interactions between postmenopausal osteoporosis (POMP) and sarcopenia, both of which substantially degrade elderly health and quality of life. We hypothesized that these motor system diseases overlap in pathophysiology and regulatory mechanisms.
Methods
We analyzed microarray data from the Gene Expression Omnibus (GEO) database using weighted gene co-expression network analysis (WGCNA), machine learning, and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis to identify common genetic factors between POMP and sarcopenia. Further validation was done via differential gene expression in a new cohort. Single-cell analysis identified high expression cell subsets, with mononuclear macrophages in osteoporosis and muscle stem cells in sarcopenia, among others. A competitive endogenous RNA network suggested regulatory elements for these genes.
Results
Signal transducer and activator of transcription 3 (STAT3) was notably expressed in both conditions. Single-cell analysis pinpointed specific cells with high STAT3 expression, and microRNA (miRNA)-125a-5p emerged as a potential regulator. Experiments confirmed the crucial role of STAT3 in osteoclast differentiation and muscle proliferation.
Conclusion
STAT3 has emerged as a key gene in both POMP and sarcopenia. This insight positions STAT3 as a potential common therapeutic target, possibly improving management strategies for these age-related diseases.
Cite this article: Bone Joint Res 2024;13(8):411–426.
Article focus
-
Identify common genetic characteristics between postmenopausal osteoporosis and sarcopenia using bioinformatics approaches.
-
Validate whether signal transducer and activator of transcription 3 (STAT3) is a shared key gene implicated in both diseases with differential expression analysis and cell experiments.
-
Explore the regulatory mechanisms of STAT3 by constructing a competing endogenous RNA (ceRNA) network.
Key messages
-
STAT3 was identified as a highly expressed gene common to both osteoporosis and sarcopenia through weighted gene coexpression network analysis.
-
STAT3 plays pivotal roles in osteoclast differentiation and muscle cell proliferation, as confirmed by knockdown and overexpression cell studies.
-
MicroRNA (miRNA)-125a-5p was predicted as a regulatory factor of STAT3 and could be a therapeutic target. Single-cell analysis revealed specific cell subsets with high STAT3 expression in each disease, providing a basis for experimental validation.
Strengths and limitations
-
Multiple bioinformatics approaches systematically analyzed datasets from Gene Expression Omnibus (GEO) to identify STAT3.
-
Findings were validated with differential expression analysis in a separate cohort and experimental cell studies. ceRNA network provided insights into potential regulatory mechanisms of STAT3.
-
This study was limited to computational analysis of existing datasets, so further mechanistic studies are needed. Specific signalling pathways involving STAT3 require further elucidation.
Introduction
Postmenopausal osteoporosis (POMP) is the most common motor system disease in postmenopausal women and is characterized by decreased bone mass and bone microstructure disorders, which exist in more than 50% of postmenopausal women.1 Sarcopenia is a kind of motor system disease related to ageing that mainly manifests as a decrease in muscle mass and muscle strength. Sarcopenia increases the risk of falls and fractures in the elderly, and affects quality of life and self-care ability.2 To date, dual-energy X-ray absorptiometry has been used to assess POMP and sarcopenia.3 Overall, 68% of women diagnosed with sarcopenia suffer from osteoporosis.4 Coin et al5 reported a positive correlation between bone mineral density (BMD) and muscle mass; that is, patients with low BMD also have less muscle mass. The term “osteosarcopenia” has been used in patients with both sarcopenia and osteoporosis. Therefore, these two diseases often do not exist independently, and there is a close relationship between them.
An increasing number of studies have found that bone and muscle, as two important parts of the motor system, are regulated by a variety of common factors.6 Muscle cells and bone cells both come from mesenchymal progenitor cells, at the earliest stage of cell development.7 After maturity, the two are not only neighbours, but also have endocrine regulatory hormones, molecular signalling pathways, and common therapeutic drugs, which affect their respective metabolisms.8 There is a close relationship between muscle and bone. On the one hand, the mechanical force caused by muscle contraction affects bone growth, bone geometry, and BMD; on the other hand, muscle and bone can regulate each other’s synthesis and catabolism by secreting muscle factors and bone factors.9 As age increases and activity decreases, the levels of these hormones and cytokines decrease, resulting in the weakening of bone and muscle anabolism and an increase in catabolism, muscle hypertrophy/atrophy, and BMD; these losses occur in parallel, leading to osteoporosis and sarcopenia.10
However, the existing treatments for POMP or sarcopenia are often aimed at only one disease. The therapeutic effect of osteoporosis therapy is often most obvious within one year of treatment, and 10% to 15% of patients who receive treatment do not experience an increase in bone mass.11 For sarcopenia, the best treatment is exercise, but its therapeutic effect is not substantial and there is a lack of effective drug treamtents and treatment targets.12 This phenomenon may be because the pathogeneses of these two diseases are linked, so the treatment of one of the diseases alone cannot achieve good results. The two diseases influence each other, and it may be better to treat them simultaneously. At present, some research progress has been made on these molecular pathways. Muscle secretion of irisin may also be a new common therapeutic target for osteoporosis and sarcopenia.13 Vitamin D, as a drug to prevent osteoporotic fracture, can effectively treat sarcopenia.14 Therefore, the two diseases may have common targets, and drugs designed for these targets may alleviate the symptoms of both diseases simultaneously.
In this study, we analyzed microarray data from the Gene Expression Omnibus (GEO) database using weighted gene co-expression network analysis (WGCNA) to identify co-expressed gene modules related to POMP and sarcopenia. We then employed the machine-learning approach random forest to screen for key genes. Differential gene expression analysis was performed to validate the findings in independent datasets. Single-cell RNA sequencing (scRNA-seq) analysis was used to pinpoint the cell type-specific expression patterns of the key gene. Finally, we constructed a competing endogenous RNA (ceRNA) network to explore the potential regulatory mechanism.
Methods
An overall research and design flowchart is shown in Supplementary Figure a.
Data sources and processing
We selected and downloaded datasets related to sarcopenia and osteoporosis from the GEO database (Supplementary Table i). For bulk RNA-seq analysis, we used datasets GSE1428, GSE136344, GSE7158, and GSE56814. These datasets were derived from human skeletal muscle and blood monocytes, with both female and mixed sex samples. For scRNA-seq analysis, we used datasets GSE172410 and GSE147287, which were derived from mouse skeletal muscle and human mesenchymal stem cells, respectively. The sample sizes for control and disease groups are listed in Supplementary Table i.
Data processing was executed using R (version 4.1.3; R Foundation for Statistical Computing, Austria), utilizing packages such as affy,15 lumi,16 and sva17 for normalization, quality assessment, and batch effect removal. Annotation of the gene expression matrix was further refined using the clusterProfiler package.18
Gene network and pathway analysis
We applied WGCNA19 to datasets GSE1428 and GSE56814 for analyzing gene co-expression patterns related to osteoporosis and sarcopenia. Top genes were selected based on standard deviation, followed by sample clustering and module partitioning. Modules were then subjected to Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis to elucidate their disease associations.20 Concurrently, we constructed a protein-protein interaction (PPI) network of key genes using the STRING database21 and identified hub genes via the cytoHubba plugin.22
The overlapping genes from the positive correlation modules of GSE1428 and GSE56814 were defined as gene set 1 (GS1), which was considered to be closely related to the pathogenesis of sarcopenia and POMP.
Machine learning and differential expression analysis
GSE136344 and GSE7158 were used as validation datasets to confirm the results obtained from the primary analysis of GSE1428 and GSE56814. The overlapping differentially expressed genes between GSE136344 and GSE7158 were defined as gene set 2 (GS2), which was used for validation purposes. Random forest analysis was employed to discern disease-related genes.23 Differential gene expression analysis was conducted using the limma package,24 and the diagnostic potential of these genes was assessed through receiver operating characteristic (ROC) curves generated with the pROC package.25
Single-cell analysis and ceRNA network construction
Single-cell analysis was performed using Seurat,26 Harmony,27 and other packages to pinpoint specific cell clusters expressing target genes. ceRNA network construction involved predicting microRNA (miRNA) and long noncoding RNA (lncRNA) targets related to signal transducer and activator of transcription 3 (STAT3) using miRanda, miRDB, and TargetScan,28 with the network visualized in Cytoscape software.29,30
Cell culture
RAW264.7 (ATCC, CL-0190) cells and C2C12 (ATCC, CRL-1774) cells were purchased from Procell (China). RAW264.7 cells were cultured in Minimum Essential Medium Alpha (C3064-0500; VivaCell, China), 10% fetal bovine serum (FBS) (BI.04-001-1ACS; Biological Industries, Israel), 1% penicillin/streptomycin, and 5% carbon dioxide at 37°C. C2C12 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (C11995500BT; Gibco, Thermo Fisher Scientific, USA), 10% FBS (BI.04-001-1ACS; Biological Industries), 1% penicillin/streptomycin, and 5% carbon dioxide at 37°C.
Cell transfection
STAT3-specific small interfering RNA (siRNA) and overexpression plasmids were obtained from General Biol (China). Lipofectamine 2000 (Invitrogen, Thermo Fisher Scientific) and Opti-MEM (Gibco, 31958-062) were used for siRNA transfection. When the osteoclast progenitor cells or myoblasts reached 40% confluence, siRNA was transfected at a concentration of 50 nM in the presence of transfection reagents. Lipofectamine 2000 and Opti-MEM were used for plasmid transfection. When the osteoclast progenitor cells or myoblasts reached 70% confluence, 10 µg of plasmid was used for transfection in the presence of transfection reagent. The protein was extracted 48 hours after transfection.
Reverse transcription-quantitative polymerase chain reaction
The total RNA of osteoclast progenitor cells or myoblasts was extracted by an RNA Extraction Kit (9767; Takara Bio, Japan), and the extracted RNA was reverse transcribed into complementary DNA (cDNA) by a Prime Script RT kit (RR047A; Takara Bio). Finally, qPCR was performed using TB Green (RR820A; Takara Bio) and a LightCycler480 system (Roche Diagnostics, Switzerland). The sequences of primers used are listed in Supplementary Table ii.
Protein extraction and western blot analysis
RAW264.7 cells and C2C12 cells were lysed with RIPA lysis buffer (P0013B; Beyotime, China) and phenylmethanesulfonyl fluoride (ST505; Beyotime), and then incubated for 30 minutes on ice. After ultrasonic crushing, the lysate was centrifuged at 14,000 rpm for 30 minutes. The supernatant was taken for protein quantification. RIPA and protein loading buffer (SIANT-BIO, China) were added to set the protein concentration to 3 μg/µl. The samples were boiled for five minutes at 100°C and frozen at -80°C.
A 10% PAGE Gel Fast Preparation Kit (PG212; Epizyme Biomedical Technology, China) was used to prepare electrophoretic gel, 8 μl of protein (24 μg) was added to each lane, electrophoresis was performed at 80 V, and when the Triple Color Prestained Protein Ladder band was clear, the voltage was adjusted to 120 V to continue electrophoresis of different molecular weight proteins for 30 minutes. The membrane was then rotated under a 180 mA current for one hour. The antibodies used were STAT3 (10253-2-AP, rabbit polyclonal; Proteintech, China), β-actin (66009-1-Ig, rabbit polyclonal; Proteintech), matrix metalloproteinase-9 (MMP9) (AF5228, rabbit polyclonal; Affinity, China), cathepsin K (CTSK) (DF6614, rabbit polyclonal; Affinity), tumour necrosis factor receptor-associated protein 1 (TRAP1) (DF7073, rabbit polyclonal; Affinity), FBXO32 (DF7075, rabbit polyclonal; Affinity), Myogenin (DF8273, rabbit polyclonal; Affinity), and TRIM63 (DF7187, rabbit polyclonal; Affinity).
Statistical analysis
All the data were measured at least three times, are expressed as mean and SD, and were statistically analyzed by GraphPad Prism 9 (GraphPad Software). Two-way analysis of variance (ANOVA) was used to analyze the results of reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and western blot analysis, and the comparison between groups was carried out by comparing the results of Šídák’s multiple comparisons test.
For the KEGG pathway enrichment analyses, a hypergeometric test was used to assess the statistical significance of gene enrichment in specific pathways, with p-values adjusted for multiple testing using the Benjamini-Hochberg procedure.
Weighted gene co-expression network analysis (WGCNA) was performed to identify modules of co-expressed genes. Soft thresholds were chosen based on scale independence and mean connectivity. Module-clinical relationship heatmaps were used to identify positively correlated modules, and independent-samples t-test was employed to determine the statistical significance of these correlations.
Weighted gene co-expression network analysis (WGCNA) was performed on the GSE56814 and GSE1428 datasets to identify modules of co-expressed genes. Soft thresholds were chosen based on scale independence and mean connectivity. Module-clinical relationship heatmaps were used to identify positively correlated modules, and independent-samples t-test was employed to determine the statistical significance of these correlations.
Statistical significance was set at p < 0.05.
Results
Data download
We paired GSE56814 and GSE1428 as the test queues for WGCNA, and GSE136344 and GSE7158 as the validation queues for differential analysis.
WGCNA
Based on scale independence and mean connectivity, the GSE56814 soft threshold was 20 (R^2 = 0.81) (Figure 1a), and the GSE1428 dataset soft threshold was 16 (R^2 = 0.9) (Figure 1b). In the observation module-clinical relationship heatmap, five modules were identified in GSE56814 species, in which turquoise was the only positive correlation module, and the correlation coefficient was 0.56 (p = 0.002, independent-samples t-test), including 1,832 genes (Figures 1c and 1d). Seven modules were identified in GSE1428 species, in which turquoise and brown were positive correlation modules, and the correlation coefficients were 0.62 (p = 0.003) and 0.66 (p = 0.006, both independent-samples t-test), respectively, including 2,129 genes (Figures 1c and 1d).
Fig. 1
Weighted gene co-expression network analysis (WGCNA). a) Scale-free network test GSE56814. b) Scale-free network verifies GSE1428. c) Topological overlap matrix (TOM) diagram GSE56814. d) GSE56814 of clinical character heat map. e) TOM diagram GSE1428. f) GSE1428 of clinical character heat map.
KEGG enrichment analysis of modular intersecting genes
KEGG enrichment analysis of GS1 (defined in the Methods section) (Supplementary Table iii), which consists of 273 overlapping genes from the positive correlation modules of GSE1428 and GSE56814, showed that it was enriched in many pathways, such as ‘cysteine and methionine metabolism’, ‘glutathione metabolism’, and ‘oxytocin signalling pathway’ (Supplementary Figure ba). The oxytocin signalling pathway may be associated with osteoporosis and sarcopenia in the occurrence and development of these two diseases.
Construction of the PPI network and screening of hub genes
We used the STRING database to construct a PPI network of GS1 (Figure 2a). Edge percolation component (EPC) is a method based on network connectivity that calculates the edge penetration component of each node under different thresholds, and evaluates the importance of each node according to its rate of change (Figure 2b). Maximum clique centrality (MCC) is a method based on maximum clique, which calculates the maximum number of cliques to which each node belongs and evaluates the importance of nodes according to their relative position (Figure 2c). Closeness is a method based on the shortest path, which calculates the mean distance from each node to all other nodes and evaluates the importance of the node according to its reciprocal (Figure 2d). Radiality is a method based on the shortest path and network diameter. We used cytoHubba to analyze the common gene expression matrix in the WGCNA positive expression module of osteoporosis and sarcopenia. We found some common genes that were rated as important nodes in all four algorithms. These genes are ACTB, TPI1, STAT3, and JUN.
Fig. 2
Protein network construction and hub gene screening. a) Protein-protein interaction network map. b) Hub gene map edge percolation component. c) Hub gene map maximum clique centrality. d) Hub gene map Closeness.
Screening hub genes by machine learning
First, the expression matrix of GS1 in the dataset GSE56814 was extracted, and the STAT3 and TRIM22 genes (seed = 4705) (Figure 3a) were obtained by the random forest method. Similarly, the expression matrix of GS1 in the dataset GSE1428 was extracted, and two genes, STAT3 and JUN (seed = 4185) (Figure 3b), were obtained by the random forest method.
Fig. 3
Machine learning screening hub gene. a) Random forest graph GSE56814. b) Random forest graph GSE1428. c) Random forest graph variable importance GSE56814. d) Random forest graph variable importance GSE1428.
The error graph data distribution of the GSE56814 datasets and GSE1428 datasets shows an obvious downward trend, indicating that with the increase in the number of trees, the classification error rate decreases gradually. The data have no outliers. When the number of trees is approximately 280 and 260, the error rate reaches the lowest point and then basically remains stable.
Variable importance maps are a kind of graph used to show the relative importance of different variables in the model. The horizontal axis is the average Gini reduction value, and the vertical axis is the name of the gene (Figures 3c and 3d). By observing the variable importance maps of GSE1428 and GSE56814, we found that the average Gini reduction of the STAT3 gene was the largest in the two images, so STAT3 was the most important gene shared by GS1 for the two diseases.
Verification of common genes
KEGG enrichment analysis of the validation GS2 (defined in the Methods section), which contains five overlapping differentially expressed genes (STAT3, JUN, TRIM22, ALOX5AP, and SLC25A6) from datasets GSE136344 and GSE7158, showed that it was enriched in pathways such as ‘thyroid cancer’, ‘bladder cancer’, and ‘oxytocin signalling pathway’. Notably, the ‘oxytocin signalling pathway’ was also enriched in GS1, supporting its potential involvement in the pathogenesis of POMP and sarcopenia (Supplementary Figure bb).
We found the common STAT3 gene from the genes screened by cytoHubba and random forest, and discovered that there was a significant difference between the disease group and the control group by difference analysis. In the box diagram of GSE7158 as the osteoporosis verification set, the expression of the STAT3 gene in the disease group was higher than that in the control group (p < 0.001, independent-samples t-test) (Figure 4a). In the quadrant map of GSE136433 as the verification set of osteoporosis, the expression of the STAT3 gene in the disease group was higher than that in the control group (p < 0.001, independent-samples t-test) (Figure 4b).
Fig. 4
Verification result. a) Validation set signal transducer and activator of transcription 3 (STAT3) difference analysis chart GSE7158. b) Validation set STAT3 difference analysis chart GSE136433. c) Validation set STAT3 receiver operating characteristic (ROC) curve GSE7158. d) Validation set STAT3 ROC curve GSE136433. All p-values were calculated using independent-samples t-test. AUC, area under the curve.
We used STAT3 as the input gene, GSE7158 and GSE136344 as the verification set, and ROC curves to evaluate the accuracy of the STAT3 gene in the diagnosis of osteoporosis and sarcopenia. The results show that our model achieves high prediction accuracy on the verification set (GSE7158: AUC 0.923, 95% CI 0.798 to 0.994; GSE136433: AUC 0.917, 95% CI 0.773 to 1.000), indicating that the STAT3 gene has a very high accuracy in the distinction between osteoporosis and sarcopenia (Figures 4c and 4d).
Results of single-cell analysis
We used reduced-dimensional clustering and uniform manifold approximation and projection (UMAP) visualization to transform thousands of cells in the database into a planar 2D structure, which is convenient for us to observe cell clustering. Then, STAT3 was extracted to generate the expression density map of the gene in single-cell data. By examining the annotated UMAP map of osteoporosis, we found that all the cells were divided into seven subgroups: macrophages, D14 monocytes, CD16 monocytes, granulocyte-monocyte progenitor cells (GMPs), M2 macrophages, monocytes, and monocyte progenitor cells. STAT3 was highly expressed in macrophages and GMPs (Figures 5a and 5b). Similarly, ageing muscle cells were divided into eight subgroups: skeletal muscle stem cells (MuSCs), fibro-adipogenic progenitors (FAPs), immune cells, myeloid cells, pericytes, skeletal muscle cells, smooth muscle cells, and stromal cells. STAT3 was highly expressed in skeletal MuSCs and FAPs (Figures 5c and 5d).
Fig. 5
Single-cell analysis. a) Uniform manifold approximation and projection (UMAP) dimensionality reduction map (osteoporosis). b) Signal transducer and activator of transcription 3 (STAT3) density map (osteoporosis). c) UMAP dimensionality reduction map (sarcopenia). d) STAT3 density map (sarcopenia).
ceRNA network construction
Through the prediction tools miRanda, miRDB, and TargetScan, we built and visualized the STAT3-centric miRNA target network. Upregulation of miR-18a-3p has been found to accelerate osteoporosis and may affect immune-mediated myopathy by inhibiting the GRIA1 gene.31 However, miR-29a-3p plays a key role in low BMD disease and its drug treatment.32 In particular, the upregulation of miR-125a-5p can promote osteoclast differentiation, while downregulation has the opposite effect (Figure 6).33
Fig. 6
Competing endogenous RNA (ceRNA) network diagram. miR, microRNA; STAT3, signal transducer and activator of transcription 3.
STAT3 silencing inhibits osteoclast progenitor cell differentiation and myoblast proliferation
To explore whether the STAT3 gene can affect bone metabolism and bone formation by affecting osteoclast differentiation of mononuclear macrophages, we used siRNA to perform targeted silencing of the STAT3 gene in mononuclear macrophages. The RT-qPCR results showed that the messenger RNA (mRNA) expression of osteoclast differentiation-promoting genes MMP9, CTSK, and TRAP1 in the silenced group was significantly lower than that in the control group (Figure 7a). The western blot results were consistent with those of quantitative PCR (qPCR), suggesting that STAT3 silencing reduced the expression of osteoclast differentiation genes and proteins (Figures 7b and 7c) at both the mRNA and protein levels. These results suggest that STAT3 may affect bone metabolism by promoting the differentiation of mononuclear macrophages into osteoclasts, thus promoting bone resorption and reducing bone formation.
Fig. 7
RAW264.7 osteoclast progenitor cells and C2C12 myoblasts were transfected with signal transducer and activator of transcription 3 (STAT3) small interfering RNA (siRNA). a) The expression of STAT3 and matrix metalloproteinase-9 (MMP9) was determined by reverse transcription-quantitative polymerase chain reaction (RT-qPCR). STAT3 messenger RNA (mRNA) in the treatment group was significantly decreased by siRNA, and MMP9 mRNA was also significantly lower than that in the control group. b) Western blot expressing protein in osteoclast progenitor cells. After 48 hours of siRNA silencing, the expression of STAT3 and MMP9 was significantly decreased. c) The gray value of western blot of osteoclast precursor cells was statistically analyzed. d) The expression of STAT3 and TRIM63 was determined by RT-qPCR. In the treatment group, STAT3 mRNA was significantly decreased by siRNA, while MMP9 mRNA was significantly higher than that in the control group. e) Western blot expressing proteins in myoblasts. After 48 hours of siRNA silencing, the expression of STAT3 was significantly decreased and the expression of Trim63 was significantly increased. f) The western blot gray values of myoblasts were statistically analyzed. **p < 0.01, ***p < 0.001, ****p < 0.0001 (Šídák’s multiple comparisons test). The number of samples in each set of experimental data is 3. CSTK, cytokine suppressive anti-inflammatory kinase; Ctrl, control; FBXO32, F-box only protein 32; MyoG, myogenin; STAT3, signal transducer and activator of transcription 3; si-STAT3, small interfering RNA targeting STAT3; TNF, tumour necrosis factor, TRAP1, TNF receptor-associated protein 1; TRIM63, tripartite motif containing 63.
We administered the same treatment to myoblasts, and the RT-qPCR and western blot results showed that the expression of muscle atrophy markers TRIM63 and FBXO32 in the silenced group was higher than that in the control group, while the expression of muscle marker protein MyoG decreased in the silenced group (Figures 7d to 7f) compared with the control group. These results suggest that silencing of the STAT3 gene in myoblasts can promote the expression of muscle atrophy markers, and that STAT3 has a positive effect on myoblast proliferation.
Overexpression of STAT3 promotes the differentiation of osteoclast progenitor cells but inhibits myoblast proliferation
To further verify the role of STAT3 in osteoclast progenitor cell differentiation, we used plasmid transfection to overexpress the STAT3 gene in osteoclasts. In osteoclast progenitor cells, the RT-qPCR and western blot results showed that overexpression of the STAT3 gene also induced high expression of the osteoclast differentiation genes MMP9, CTSK, and TRAP1 (Figures 8a to 8c). These results suggest that the high expression of STAT3 in POMP aggravates the development of POMP by promoting the differentiation of osteoclast precursors into osteoclasts and promoting bone resorption.
Fig. 8
Plasmid was transfected into RAW264.7 osteoclast progenitor cells and C2C12 myoblasts. a) The expression of signal transducer and activator of transcription 3 (STAT3) and matrix metalloproteinase-9 (MMP9) was determined by reverse transcription-quantitative polymerase chain reaction (RT-qPCR). The genes of STAT3 and MMP9 in the treatment group were significantly overexpressed by the plasmid compared with the control group. b) Western blot of protein expressed by osteoclast progenitor cells; 48 hours after overexpression of plasmid, the expression of STAT3 and MMP9 protein increased. c) Statistical analysis of western blot gray value of osteoclast progenitor cells. d) The expression of STAT3 and TRIM63 was detected by RT-qPCR. The genes of STAT3 and TRIM63 in the treatment group were significantly overexpressed by the plasmid compared with the control group. e) Western blot of protein expressed by myoblasts; 48 hours after overexpression of plasmid, the expression of STAT3 and TRIM63 protein increased. f) Statistical analysis of western blot gray value of myoblasts. **p < 0.01, ***p < 0.001, ****p < 0.0001 (Šídák’s multiple comparisons test). The number of samples in each set of experimental data is 3. CSTK, cytokine suppressive anti-inflammatory kinase; Ctrl, control; FBXO32, F-box only protein 32; MyoG, myogenin; STAT3, Ov-STAT3, overexpression of STAT3; STAT3, signal transducer and activator of transcription 3; TNF, tumour necrosis factor, TRAP1, TNF receptor-associated protein 1; TRIM63, tripartite motif containing 63.
Similarly, the STAT3 gene was overexpressed in myoblasts, and the RT-qPCR and western blot results suggested that overexpression of the STAT3 gene promoted the expression of muscle atrophy markers TRIM63 and FBXO32, and inhibited the expression of MyoG (Figures 8d to 8f). This finding suggests that the high expression of STAT3 in sarcopenia may promote the expression of muscular dystrophy factor and lead to muscle mass loss.
Discussion
To find the common ground between the two diseases, we took three steps. First, we found the shared biological pathways and co-expressed genes of the two diseases by analyzing the information in the bulk RNA-seq dataset. Then, the pathways and genes we found were validated in another bulk RNA-seq dataset. We finally confirmed that STAT3 is a common highly expressed gene in both diseases. To further confirm the distribution of STAT3 cell subsets in both diseases, we verified the genes at the single-cell level and found that STAT3 is highly expressed in mononuclear macrophages, MuSCs, and FAPs. Finally, through the construction of a ceRNA network, we found that miRNA-125a-5p is the regulatory factor of STAT3. Based on the results of our bioinformatics analysis, experimental results, and existing research progress, STAT3 is believed to be the key co-pathogenic gene of POMP and sarcopenia. Designing drugs targeting STAT3 itself or its regulatory factor might have a better therapeutic effect on both diseases than treating one of them alone.
STAT3 is a signal sensor and transcriptional activator that plays an important role in the regulation of cell proliferation, apoptosis, and differentiation.34 Yang et al35 noted that the bone mass of mice with osteoclast-specific STAT3 deficiency was significantly higher than that of the control group. Some scholars have found that IL-6 causes septicaemia-induced muscle mass loss through the gp130/JAK2/STAT3 pathway, and muscle mass loss is alleviated after the use of this channel inhibitor.36 The pathogenesis of osteoporosis combined with sarcopenia is not clear, but our study found that STAT3 may be a common pathogenic target of both diseases and play a very important role in both diseases.
RANKL/RANK is an important regulatory axis of muscle and bone metabolism. RANKL is recognized as an important molecule that affects osteoclast differentiation, together with macrophage colony-stimulating factor (M-CSF).37 STAT3 is an important activation signal molecule of the RANKL/RANK pathway. It has been proven that the activation of the JAK2/STAT3 axis by interleukin (IL)-6 to promote the expression of RANKL is an important step in osteoclast differentiation of mononuclear macrophages.38 Overactivated osteoclasts lead to bone resorption, resulting in osteopenia and osteoporosis.39 Furthermore, our KEGG enrichment analysis also highlighted the oxytocin signalling pathway as over-represented in both the gene modules related to POMP and sarcopenia. Oxytocin has been reported to have a dual effect on bone metabolism. On one hand, it promotes osteoblast differentiation and function, leading to increased bone formation.40-43 Oxytocin stimulates the differentiation of osteoblasts and upregulates the expression of genes involved in osteoblast differentiation, resulting in enhanced bone mineralization and formation.42,43 On the other hand, oxytocin has a complex effect on osteoclasts. It can stimulate osteoclast formation, both directly by activating specific signalling pathways and indirectly by upregulating RANKL production by osteoblasts. However, oxytocin also inhibits bone resorption activity of mature osteoclasts by triggering calcium release and nitric oxide synthesis.43 This suggests that oxytocin plays a crucial role in regulating the balance between bone formation and resorption. Dysregulation of oxytocin signalling may contribute to the development of osteoporosis by disrupting this delicate balance. The involvement of the oxytocin pathway in both POMP and sarcopenia gene modules indicates its potential as a molecular link between muscle and bone pathologies, which warrants further investigation.
Through single-cell analysis of the POMP dataset, we determined that the STAT3 gene was highly expressed in mononuclear macrophages, while overexpression and knockdown of STAT3 influenced the activity of macrophages via MMP9, CTSK, and TRAP1. Therefore, we speculate that STAT3 is involved in osteoporosis by promoting osteoclast differentiation of macrophages.
STAT3 also plays an irreplaceable role in sarcopenia due to various causes, including ageing, physical, poor nutrition, chronic diseases, hormonal changes, inflammation, oxidative stress, neurodegenerative processes, genetics, and medication. However, there are contradictions about the role of STAT3 in muscle. Some studies have found that STAT3 can promote muscle regeneration by activating satellite cells, and the activation of STAT3 can also directly promote the function of mitochondria in skeletal MuSCs.44,45 These studies have shown that STAT3 plays a positive role in muscle mass and repair by acting on muscle cells or muscle helper cells. Moreover, a large number of studies have published different views, such as the belief that the activation of the JAK2/STAT3 pathway mediated by IL-6 can lead to muscle atrophy, and that STAT3 inhibition can reduce muscle atrophy caused by cancer cachexia.46 Excessive activation of STAT3 in muscle cells in diabetes or chronic kidney disease can lead to muscle loss, and targeted knockout of STAT3 or inhibition of STAT3 expression in muscle can improve muscle quality.47 We confirmed the high expression of the STAT3 gene in skeletal MuSCs and FAPs in the disease group by single-cell analysis, suggesting that STAT3 may be closely related to sarcopenia. We silenced and overexpressed STAT3 in myoblasts, the muscle atrophy markers TRIM63 and FBXO32 increased, and the expression of MyoG decreased. This phenomenon suggests that there is a certain expression range of STAT3, and too high or too low levels of STAT3 will affect the physiological function of muscle.
Given these two completely different viewpoints, we speculate that STAT3 will have different effects when the muscles are in different states. When the body is in a steady state, STAT3, as an important signalling molecule, is directly involved in muscle synthesis and repair. Under unstable conditions such as chronic injury and inflammation caused by disease, as well as insulin resistance, excessive activation of STAT3 leads to muscle atrophy.36 FAPs are activated when muscle injury occurs and interact with MuSCs to promote muscle repair. A 2018 study clarified in detail that FAPs continue to activate STAT3 and secrete a large amount of IL-6 in muscular atrophy models, resulting in muscle atrophy and fibrosis, which can be alleviated in mice with STAT3 signal knockout in FAPs.48
Bone and muscle are interrelated, and the interaction between them is called the bone-muscle axis. Although sarcopenia and osteoporosis were initially considered to be separate diseases, recent studies have shown that bones and muscles are spatially and metabolically connected.49 This interaction, known as skeletal and muscular comorbidity, indicates a synergistic effect between low BMD and muscle atrophy and low function (Figure 9).50 Osteoporosis is closely related to chronic inflammation and oxidative damage. In the case of osteoporosis, the bone microenvironment changes, which may affect adjacent muscle tissue. Our results show that macrophages highly express STAT3 during the pathogenesis of osteoporosis, which in turn activates the RANK/RANKL pathway, promotes bone resorption, and aggravates osteoporosis. This aggravation of osteoporosis may further exacerbate chronic inflammation, creating a vicious circle. A chronic inflammatory environment can damage not only bone tissue but also muscle cells. In this environment, FAPs and MuSCs may be activated in an unstable state. Abnormally activated FAPs overexpress STAT3 and secrete a large amount of IL-6, which leads to muscle atrophy and fibrosis and further reduces the muscle mass of patients.48 Sarcopenia not only affects the ability to exercise of patients, but also may have a negative impact on bone health. Muscle is not only a motor organ but also an important endocrine organ. Muscles can secrete various cytokines (such as irisin and insulin-like growth factor 1 (IGF-1)) to regulate bone homeostasis.9 In the state of muscle atrophy, the secretion of muscle factors decreases, which aggravates osteoporosis.51 Moreover, mechanical stimulation between muscle and bone is a key factor in maintaining BMD.52 When muscle mass decreases, mechanical stimulation to bones also decreases, which may lead to a decrease in BMD. In addition, sarcopenia may lead to increased chronic inflammation and oxidative stress, both of which are associated with bone metabolic disorders and increased bone resorption.53
Fig. 9
Proposed molecular mechanisms linking postmenopausal osteoporosis (POMP) and sarcopenia. Chronic inflammation and comorbidity lead to a bidirectional interaction between bone and muscle, known as the bone-muscle axis. In POMP, single-cell RNA sequencing (scRNA-seq) analysis reveals high signal transducer and activator of transcription 3 (STAT3) expression in macrophages, which activates the RANKL/RANK pathway and promotes osteoclast differentiation, leading to increased bone resorption and dyskinesia/fracture. In sarcopenia, scRNA-seq analysis shows elevated STAT3 expression in muscle stem cells and fibro-adipogenic progenitors (FAPs). Abnormally activated FAPs overexpress STAT3 and secrete interleukin (IL)-6, resulting in muscle atrophy and fibrosis. The competing endogenous RNA (ceRNA) network analysis identifies microRNA (miRNA)-125a-5p as a potential regulator of STAT3, suggesting its role in modulating the pathogenesis of both diseases. The comorbidity and interaction between POMP and sarcopenia exacerbate each other, leading to a non-steady-state condition. Targeting the shared molecular pathways, such as the STAT3/miRNA-125a-5p axis, may provide a novel therapeutic strategy for managing these age-related musculoskeletal disorders.
As mentioned above, osteoporosis and sarcopenia exacerbate each other, so treating one of these diseases alone cannot break this vicious circle, and it is necessary to find a target to treat both diseases. Our study found a common therapeutic target, STAT3, but how to regulate the expression of this key gene is still a problem. Therefore, we identified the upstream regulatory factor of STAT3, miRNA-125a-5p, by constructing a ceRNA network. miRNA-125a-5p/STAT3 plays a regulatory role in the occurrence and development of osteoporosis and sarcopenia. Previous studies have shown that miR-125a-5p enhances the level of autophagy by targeting STAT3.54 This finding confirms that miRNA-125a-5p can regulate pathophysiological activity by targeting STAT3. Therefore, we speculate that miRNA-125a-5p mediates the expression of RANKL through targeted regulation of STAT3 in macrophages and affects the process of osteoporosis. Moreover, through targeted regulation of STAT3 in MuSCs and FAPs, it mediates the IL-6 signalling pathway and affects the progression of sarcopenia. Therefore, miRNA-125a-5p, as an important regulator of STAT3, may be a potential therapeutic target for the development of drugs to treat both osteoporosis and sarcopenia.
Comorbidity is characterized by a common underlying mechanism and the need for co-treatments.55 There is a common target for concurrent treatment of comorbid diseases, which has implications for the prevention and treatment of disorders. Osteoporosis and sarcopenia are also comorbidities in bones and muscles, and effective therapeutic approaches are lacking. In this study, miRNA-125a-5p was identified as a potential concurrent treatment for osteoporosis and sarcopenia via interaction with the mRNA of STAT3. These findings provide a more comprehensive perspective and help us to understand the complex relationship between osteoporosis and sarcopenia. Future research should focus on the common management and treatment of common diseases, while medications focusing on common targets of the comorbidity of osteoporosis and sarcopenia may bring about better therapeutic effects.
References
1. Chen K , Lv Z-T , Cheng P , et al. Boldine ameliorates estrogen deficiency-induced bone loss via inhibiting bone resorption . Front Pharmacol . 2018 ; 9 : 1046 . Crossref PubMed Google Scholar
2. Fielding RA , Vellas B , Evans WJ , et al. Sarcopenia: an undiagnosed condition in older adults. Current consensus definition: prevalence, etiology, and consequences. International working group on sarcopenia . J Am Med Dir Assoc . 2011 ; 12 ( 4 ): 249 – 256 . Crossref PubMed Google Scholar
3. Buckinx F , Landi F , Cesari M , et al. Pitfalls in the measurement of muscle mass: a need for a reference standard . J Cachexia Sarcopenia Muscle . 2018 ; 9 ( 2 ): 269 – 278 . Crossref PubMed Google Scholar
4. Hernández-Álvarez D , Mena-Montes B , Toledo-Pérez R , et al. Long-term moderate exercise combined with metformin treatment induces an hormetic response that prevents strength and muscle mass loss in old female wistar rats . Oxid Med Cell Longev . 2019 ; 2019 : 3428543 . Crossref PubMed Google Scholar
5. Coin A , Perissinotto E , Enzi G , et al. Predictors of low bone mineral density in the elderly: the role of dietary intake, nutritional status and sarcopenia . Eur J Clin Nutr . 2008 ; 62 ( 6 ): 802 – 809 . Crossref PubMed Google Scholar
6. Ellman R , Spatz J , Cloutier A , Palme R , Christiansen BA , Bouxsein ML . Partial reductions in mechanical loading yield proportional changes in bone density, bone architecture, and muscle mass . J Bone Miner Res . 2013 ; 28 ( 4 ): 875 – 885 . Crossref PubMed Google Scholar
7. Batsis JA , Villareal DT . Sarcopenic obesity in older adults: aetiology, epidemiology and treatment strategies . Nat Rev Endocrinol . 2018 ; 14 ( 9 ): 513 – 537 . Crossref PubMed Google Scholar
8. Edwards MH , Dennison EM , Aihie Sayer A , Fielding R , Cooper C . Osteoporosis and sarcopenia in older age . Bone . 2015 ; 80 : 126 – 130 . Crossref PubMed Google Scholar
9. Greco EA , Pietschmann P , Migliaccio S . Osteoporosis and sarcopenia increase frailty syndrome in the elderly . Front Endocrinol (Lausanne) . 2019 ; 10 : 255 . Crossref PubMed Google Scholar
10. Dufresne SS , Dumont NA , Boulanger-Piette A , et al. Muscle RANK is a key regulator of Ca2+ storage, SERCA activity, and function of fast-twitch skeletal muscles . Am J Physiol Cell Physiol . 2016 ; 310 ( 8 ): C663 – 72 . Crossref PubMed Google Scholar
11. Roux C , Briot K . How long should we treat? Osteoporos Int . 2014 ; 25 ( 6 ): 1659 – 1666 . Crossref PubMed Google Scholar
12. Hsu W-B , Lin S-J , Hung J-S , et al. Effect of resistance training on satellite cells in old mice - a transcriptome study: implications for sarcopenia . Bone Joint Res . 2022 ; 11 ( 2 ): 121 – 133 . Crossref PubMed Google Scholar
13. Colaianni G , Cinti S , Colucci S , Grano M . Irisin and musculoskeletal health . Ann N Y Acad Sci . 2017 ; 1402 ( 1 ): 5 – 9 . Crossref PubMed Google Scholar
14. Abe S , Kashii M , Shimada T , et al. Relationship between distal radius fracture severity and 25-hydroxyvitamin-D level among perimenopausal and postmenopausal women . Bone Jt Open . 2022 ; 3 ( 3 ): 261 – 267 . Crossref PubMed Google Scholar
15. Gautier L , Cope L , Bolstad BM , Irizarry RA . affy--analysis of Affymetrix GeneChip data at the probe level . Bioinformatics . 2004 ; 20 ( 3 ): 307 – 315 . Crossref PubMed Google Scholar
16. Du P , Kibbe WA , Lin SM . lumi: a pipeline for processing Illumina microarray . Bioinformatics . 2008 ; 24 ( 13 ): 1547 – 1548 . Crossref PubMed Google Scholar
17. Leek JT , Johnson WE , Parker HS , Jaffe AE , Storey JD . The sva package for removing batch effects and other unwanted variation in high-throughput experiments . Bioinformatics . 2012 ; 28 ( 6 ): 882 – 883 . Crossref PubMed Google Scholar
18. Yu G , Wang L-G , Han Y , He Q-Y . clusterProfiler: an R package for comparing biological themes among gene clusters . OMICS . 2012 ; 16 ( 5 ): 284 – 287 . Crossref PubMed Google Scholar
19. Yuan W , Yang M , Zhu Y . Development and validation of a gene signature predicting the risk of postmenopausal osteoporosis . Bone Joint Res . 2022 ; 11 ( 8 ): 548 – 560 . Crossref PubMed Google Scholar
20. Xiang C , Liao Y , Chen Z , et al. Network pharmacology and molecular docking to elucidate the potential mechanism of ligusticum chuanxiong against osteoarthritis . Front Pharmacol . 2022 ; 13 : 854215 . Crossref PubMed Google Scholar
21. No authors listed . STRING: functional protein association networks . 2023 . https://string-db.org/ ( date last accessed 29 July 2024 ). Google Scholar
22. Chin C-H , Chen S-H , Wu H-H , Ho C-W , Ko M-T , Lin C-Y . cytoHubba: identifying hub objects and sub-networks from complex interactome . BMC Syst Biol . 2014 ; 8 Suppl 4 ( Suppl 4 ): S11 . Crossref PubMed Google Scholar
23. Hu J , Szymczak S . A review on longitudinal data analysis with random forest . Brief Bioinform . 2023 ; 24 ( 2 ): bbad002 . Crossref PubMed Google Scholar
24. Ritchie ME , Phipson B , Wu D , et al. limma powers differential expression analyses for RNA-sequencing and microarray studies . Nucleic Acids Res . 2015 ; 43 ( 7 ): e47 . Crossref PubMed Google Scholar
25. Robin X , Turck N , Hainard A , et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves . BMC Bioinformatics . 2011 ; 12 : 77 . Crossref PubMed Google Scholar
26. Hao Y , Hao S , Andersen-Nissen E , et al. Integrated analysis of multimodal single-cell data . Cell . 2021 ; 184 ( 13 ): 3573 – 3587 . Crossref PubMed Google Scholar
27. Korsunsky I , Millard N , Fan J , et al. Fast, sensitive and accurate integration of single-cell data with Harmony . Nat Methods . 2019 ; 16 ( 12 ): 1289 – 1296 . Crossref PubMed Google Scholar
28. Min H , Yoon S . Got target? Computational methods for microRNA target prediction and their extension . Exp Mol Med . 2010 ; 42 ( 4 ): 233 – 244 . Crossref PubMed Google Scholar
29. Pliner HA , Shendure J , Trapnell C . Supervised classification enables rapid annotation of cell atlases . Nat Methods . 2019 ; 16 ( 10 ): 983 – 986 . Crossref PubMed Google Scholar
30. Alquicira-Hernandez J , Powell JE . Nebulosa recovers single-cell gene expression signals by kernel density estimation . Bioinformatics . 2021 ; 37 ( 16 ): 2485 – 2487 . Crossref PubMed Google Scholar
31. Zhao M , Dong J , Liao Y , et al. MicroRNA miR-18a-3p promotes osteoporosis and possibly contributes to spinal fracture by inhibiting the glutamate AMPA receptor subunit 1 gene (GRIA1) . Bioengineered . 2022 ; 13 ( 1 ): 370 – 382 . Crossref PubMed Google Scholar
32. Kyei B , Odame E , Li L , et al. Knockdown of CDR1as decreases differentiation of goat skeletal muscle satellite cells via upregulating miR-27a-3p to inhibit ANGPT1 . Genes (Basel) . 2022 ; 13 ( 4 ): 663 . Crossref PubMed Google Scholar
33. Sun L , Lian JX , Meng S . MiR-125a-5p promotes osteoclastogenesis by targeting TNFRSF1B . Cell Mol Biol Lett . 2019 ; 24 : 23 . Crossref PubMed Google Scholar
34. Lee SW , Ahn YY , Kim YS , et al. The immunohistochemical expression of STAT3, Bcl-xL, and MMP-2 proteins in colon adenoma and adenocarcinoma . Gut Liver . 2012 ; 6 ( 1 ): 45 – 51 . Crossref PubMed Google Scholar
35. Yang Y , Chung MR , Zhou S , et al. STAT3 controls osteoclast differentiation and bone homeostasis by regulating NFATc1 transcription . J Biol Chem . 2019 ; 294 ( 42 ): 15395 – 15407 . Crossref PubMed Google Scholar
36. Zanders L , Kny M , Hahn A , et al. Sepsis induces interleukin 6, gp130/JAK2/STAT3, and muscle wasting . J Cachexia Sarcopenia Muscle . 2022 ; 13 ( 1 ): 713 – 727 . Crossref PubMed Google Scholar
37. Kim MH , Choi LY , Chung JY , Kim E-J , Yang WM . Auraptene ameliorates osteoporosis by inhibiting RANKL/NFATc1 pathway-mediated bone resorption based on network pharmacology and experimental evaluation . Bone Joint Res . 2022 ; 11 ( 5 ): 304 – 316 . Crossref PubMed Google Scholar
38. Harmer D , Falank C , Reagan MR . Interleukin-6 interweaves the bone marrow microenvironment, bone loss, and multiple myeloma . Front Endocrinol (Lausanne) . 2018 ; 9 : 788 . Crossref PubMed Google Scholar
39. Wang X , Jiang W , Pan K , Tao L , Zhu Y . Melatonin induces RAW264.7 cell apoptosis via the BMAL1/ROS/MAPK-p38 pathway to improve postmenopausal osteoporosis . Bone Joint Res . 2023 ; 12 ( 11 ): 677 – 690 . Crossref PubMed Google Scholar
40. Feixiang L , Yanchen F , Xiang L , et al. The mechanism of oxytocin and its receptors in regulating cells in bone metabolism . Front Pharmacol . 2023 ; 14 : 1171732 . Crossref PubMed Google Scholar
41. Breuil V , Trojani M-C , Ez-Zoubir A . Oxytocin and bone: review and perspectives . Int J Mol Sci . 2021 ; 22 ( 16 ): 8551 . Crossref PubMed Google Scholar
42. Colaianni G , Sun L , Zaidi M , Zallone A . The “love hormone” oxytocin regulates the loss and gain of the fat-bone relationship . Front Endocrinol (Lausanne) . 2015 ; 6 : 79 . Crossref PubMed Google Scholar
43. Colaianni G , Tamma R , Di Benedetto A , et al. The oxytocin-bone axis . J Neuroendocrinol . 2014 ; 26 ( 2 ): 53 – 57 . Crossref PubMed Google Scholar
44. Blankenbach KV , Schwalm S , Pfeilschifter J , Meyer Zu Heringdorf D . Sphingosine-1-phosphate receptor-2 antagonists: therapeutic potential and potential risks . Front Pharmacol . 2016 ; 7 : 167 . Crossref PubMed Google Scholar
45. Sala D , Cunningham TJ , Stec MJ , et al. The Stat3-Fam3a axis promotes muscle stem cell myogenic lineage progression by inducing mitochondrial respiration . Nat Commun . 2019 ; 10 ( 1 ): 1796 . Crossref PubMed Google Scholar
46. Ma JF , Sanchez BJ , Hall DT , Tremblay A-M , Di Marco S , Gallouzi I-E . STAT3 promotes IFNγ/TNFα-induced muscle wasting in an NF-κB-dependent and IL-6-independent manner . EMBO Mol Med . 2017 ; 9 ( 5 ): 622 – 637 . Crossref PubMed Google Scholar
47. Zhang L , Pan J , Dong Y , et al. Stat3 activation links a C/EBPδ to myostatin pathway to stimulate loss of muscle mass . Cell Metab . 2013 ; 18 ( 3 ): 368 – 379 . Crossref PubMed Google Scholar
48. Madaro L , Passafaro M , Sala D , et al. Denervation-activated STAT3-IL-6 signalling in fibro-adipogenic progenitors promotes myofibres atrophy and fibrosis . Nat Cell Biol . 2018 ; 20 ( 8 ): 917 – 927 . Crossref PubMed Google Scholar
49. Chalhoub D , Boudreau R , Greenspan S , et al. Associations between lean mass, muscle strength and power, and skeletal size, density and strength in older men . J Bone Miner Res . 2018 ; 33 ( 9 ): 1612 – 1621 . Crossref PubMed Google Scholar
50. Kirk B , Harrison SL , Zanker J , et al. Interactions between HR-pQCT bone density and D3 Cr muscle mass (or HR-pQCT bone structure and HR-pQCT muscle density) in predicting fractures: the Osteoporotic Fractures in Men Study . J Bone Miner Res . 2023 ; 38 ( 9 ): 1245 – 1257 . Crossref PubMed Google Scholar
51. Hansen M , Rubinsztein DC , Walker DW . Autophagy as a promoter of longevity: insights from model organisms . Nat Rev Mol Cell Biol . 2018 ; 19 ( 9 ): 579 – 593 . Crossref PubMed Google Scholar
52. Plotkin LI , Bellido T . Osteocytic signalling pathways as therapeutic targets for bone fragility . Nat Rev Endocrinol . 2016 ; 12 ( 10 ): 593 – 605 . Crossref PubMed Google Scholar
53. Pérez-Baos S , Prieto-Potin I , Román-Blas JA , Sánchez-Pernaute O , Largo R , Herrero-Beaumont G . Mediators and patterns of muscle loss in chronic systemic inflammation . Front Physiol . 2018 ; 9 : 409 . Crossref PubMed Google Scholar
54. Kumar P , Kumawat RK , Uttam V , et al. The imminent role of microRNAs in salivary adenoid cystic carcinoma . Transl Oncol . 2023 ; 27 : 101573 . Crossref PubMed Google Scholar
55. Mukonzo J , Aklillu E , Marconi V , Schinazi RF . Potential drug-drug interactions between antiretroviral therapy and treatment regimens for multi-drug resistant tuberculosis: Implications for HIV care of MDR-TB co-infected individuals . Int J Infect Dis . 2019 ; 83 : 98 – 101 . Crossref PubMed Google Scholar
Author contributions
D. Liu: Data curation, Formal analysis, Methodology, Software, Writing – original draft
K. Wang: Investigation, Writing – original draft
J. Wang: Data curation, Methodology
F. Cao: Data curation, Investigation, Software
L. Tao: Funding acquisition, Project administration, Resources, Writing – review & editing
Funding statement
The authors disclose receipt of the following financial or material support for the research, authorship, and/or publication of this article: support from the Liaoning Province Natural Fund (2022-YGJC-58) and Shenyang City Natural Fund (22-321-32-08).
ICMJE COI statement
All authors report support from the Liaoning Province Natural Fund (2022-YGJC-58) and Shenyang City Natural Fund (22-321-32-08), related to this study.
The authors declare that they have no known competing financial interests or personal relationships, which may affect the work reported in this article.
Data sharing
The data for this study are publicly available at: https://www.ncbi.nlm.nih.gov/gds/.
Acknowledgements
We thank the authors of the GSE1428, GSE136344, GSE172410, GSE7158, GSE56814, and GSE147287 datasets for their contribution.
Open access funding
The authors report that they received open access funding for their manuscript from the Liaoning Province Natural Fund (grant number 2022-YGJC-58) and the Shenyang City Natural Fund (grant number 22-321-32-08).
Supplementary material
Figure a provides an overall research and design flowchart. Table i lists datasets related to sarcopenia and osteoporosis, including sample sizes for control and disease groups, obtained from the Gene Expression Omnibus (GEO) database. Detailed RNA extraction and quantitative polymerase chain reaction (qPCR) methodologies are described, with specific primer sequences listed in Table ii. Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis results for gene set 1 (GS1) are presented in Table iii, highlighting key pathways such as cysteine and methionine metabolism, and the oxytocin signalling pathway, further explored in Figures ba and bb.
© 2024 Liu et al. This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0) licence, which permits the copying and redistribution of the work only, and provided the original author and source are credited. See https://creativecommons.org/licenses/by-nc-nd/4.0/