Pengbo Wang1, Lihua Yan2,*
1Shanghai Zhongqiao Vocational and Technical University, Shanghai, China, 201514
2The 13th Research Institute of China Electronics Technology Group Corporation, Shijiazhuang, China, 050051
*Corresponding author: yanlh86@163.com
Keywords: Autophagy, Hub genes, Bioinformatics, Protein-protein Interaction.
Abstract: Osteonecrosis of the femoral head (ONFH) progressively deteriorates the hip joint, and its molecular mechanisms remain incompletely understood, with autophagy’s role in its pathogenesis garnering significant interest. This work was designed to identify the hub genes and core pathways associated with autophagy in ONFH through bioinformatics analysis. Gene expression data from the GSE74089 dataset, comprising four disease samples and four control samples, were analyzed using the limma library in R to identify differentially expressed genes (DEGs). Autophagy-relevant genes were acquired from the Gene Ontology (GO) resource. The overlap between DEGs and autophagy-relevant genes generated a subset of autophagy-related differentially expressed genes (AT-DEGs). Enriched pathways were detected by Gene Set Enrichment Analysis (GSEA) and functionally annotated through the GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) approaches. The STRING platform was introduced to build a protein–protein interaction (PPI) network, and pivotal genes were subsequently screened by degree centrality within Cytoscape. The results indicated that among the 3,648 DEGs, 547 were autophagy-related, of which 106 showed significant differential expression. These genes clustered primarily in pathways related to neurodegenerative diseases, mTOR signaling, and TGF-β signaling. Notably, PIK3C3, UVRAG, ATG13, ATG12, RB1CC1, CASP3, FOXO3, BNIP3, ZFYVE1, and WIPI1 ranked among the key hub genes. This study identified key autophagy-related genes and signaling pathways implicated in ONFH pathogenesis, illuminating molecular mechanisms and therapeutic targets.
- Introduction
Osteonecrosis of the femoral head (ONFH) is a degenerative condition characterized by avascular necrosis due to compromised blood supply. It predominantly affects young and middle-aged adults, often leading to femoral head collapse, severe pain, and loss of joint function. As the condition progresses, medical intervention such as total hip replacement is often required to restore mobility and relieve pain. As the disease advances, surgical intervention such as total hip arthroplasty is frequently required to restore mobility and alleviate symptoms [1]. Based on etiology, ONFH is generally categorized into traumatic and non-traumatic subtypes. The traumatic subtype predominantly stems from proximal femoral fractures or dislocations, which interrupt femoral head perfusion, cause localized ischemia, and ultimately lead to structural deterioration.
In contrast, the non-traumatic subtype, which is more, is often associated with specific risk factors, including prolonged corticosteroid therapy, chronic alcohol abuse, and autoimmune disorders. These factors commonly contribute to vascular compromise and initiate osteonecrosis via ischemic pathways [2]. Among these, corticosteroid-induced ONFH has received increasing attention, particularly in the context of systemic inflammation, such as that observed in COVID-19 patients receiving high-dose steroids [3]. A typical case involved a 45-year-old patient who developed bilateral ONFH after 6 months of high-dose prednisone therapy for autoimmune hepatitis, presenting with severe hip pain and restricted mobility, ultimately requiring total hip arthroplasty [4]. Similarly, chronic alcohol abuse represents another critical risk factor, as exemplified by a 75-year-old man with a 50-year history of heavy alcohol consumption (approximately 1 kg daily) who progressed to advanced-stage ONFH with femoral head collapse, necessitating unilateral hip replacement due to intractable pain and functional impairment [5]. ONFH development is driven by multifactorial factors, including vascular disruption, abnormalities in lipid metabolism, and inflammatory responses which act synergistically to promote disease progression [6]. Current treatment strategies range from conservative approaches, such as physiotherapy and medications, to more invasive procedures, such as core decompression and total hip replacement. However, the outcomes of these treatments are often unpredictable. The underlying molecular mechanisms of ONFH are still poorly understood, highlighting the urgent need for further investigations into its biochemical and cellular basis [7].
Autophagy utilizes lysosomal activity to eliminate defective organelles and malformed polypeptides, thereby maintaining cellular homeostasis. Autophagy dysregulation drives multiple disease processes such as neurodegeneration, carcinogenesis, and metabolic disorders. [8]. Within bone biology, autophagy critically modulates osteoblast differentiation, governs osteoclast activity, and coordinates bone remodeling dynamics [9]. ONFH is characterized by ischemia, oxidative stress, and apoptosis of bone cells. Autophagy functioned as both a promoter and suppressor in this process. At early stages, autophagy can safeguard osteoblasts from glucocorticoid-induced damage by mitigating apoptosis and maintaining cell viability. Modulating autophagy via the PI3K/Akt/mTOR axis may mitigate glucocorticoid-induced bone thinning and apoptosis [10]. However, aberrant or excessive autophagic activity may lead to non-selective degradation of cellular components, resulting in autophagy-associated cell death and tissue dysfunction [11]. Recent evidence suggests that autophagy may influence ONFH pathogenesis by modulating cellular stress responses and inflammatory pathways [12]. Impaired autophagic flux has been associated with elevated oxidative stress and apoptosis in osteocytes, both of which are key contributors to bone necrosis. Elucidating autophagy’s pathological involvement during ONFH progression may uncover druggable pathways for the development of precision therapeutics.
Biochips, also known as microarrays, have become essential tools in contemporary biomedical research, especially for unraveling the genetic mechanisms underlying diseases. These high-throughput technologies allow for the simultaneous analysis of thousands of genes, offering a robust platform for genome-wide investigations. In this study, a microarray dataset was employed to identify autophagy-associated hub genes and critical pathways involved in ONFH, aiming to provide novel insights into autophagic dysregulation in this disorder’s development.
- Materials and Methods
2.1 Microarray Dataset
The GSE74089 microarray was retrieved from the publicly accessible Gene Expression Omnibus (GEO) repository [13]. This profile utilized the GPL13497 platform and included four disease samples alongside four control samples. All data within our research were derived from open-access repositories. Therefore, no additional ethical approval was required for the analyses performed.
2.2 Differential Expression Analysis
Differentially expressed genes (DEGs) between the ONFH and control groups were identified employing the limma pipeline, with statistical significance set at p < 0.05 and |log2FoldChange| > 1.5. A volcano plot and heat map were using the ggplot2 library to provide an intuitive visual representation of the DEGs. Subsequently, Autophagy-related genes were retrieved from the Gene Ontology (GO) database (https://www.geneontology.org/) by searching term “autophagy” and setting the organism option to “Homo sapiens”. The intersection of the genes obtained from the GO database and the genes from GSE74089 were identified as autophagy-related genes in ONFH. Differentially expressed genes within this intersection were termed autophagy-DEGs (AT-DEGs).
2.3 Gene Set Enrichment Analysis
To investigate ONFH pathophysiology, we performed pathway enrichment analysis using GSEA v4.4.0. The Normalized Enrichment Score (NES) was used to evaluate the enrichment of biological processes and identify candidate molecular pathways, with cutoff criteria set as false discovery rate (FDR) less than 25% and p value less than 0.05.
2.4 Gene Ontology Analysis
To characterize the functional relevance of AT-DEGs, GO enrichment was conducted using DAVID (https://davidbioinformatics.nih.gov/home.jsp), a functional annotation platform designed to help researchers interpret the biological meaning behind large gene lists. DAVID determines whether certain GO terms are statistically overrepresented in AT-DEGs compared with a reference gene list, thereby elucidating autophagy-associated terms in ONFH pathogenesis [14].
2.5 Kyoto Encyclopedia of Genes and Genomes Analysis
To pinpoint critical signaling routes enriched among AT-DEGs, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was conducted via Metascape (http://metascape.org/), an online platform integrating gene functional annotation and pathway discovery [15]. Metascape integrates KEGG pathway data to highlight pathways showing significant enrichment in AT-DEGs, thereby elucidating metabolic signaling pathways associated with autophagy in ONFH. Pathways meeting the significance criterion (adjusted p < 0.05) were retained.
2.6 Protein-Protein Interaction Network Analysis
To elucidate autophagy-regulatory interactomes, the STRING platform (https://string-db.org/) was adopted to generate a protein-protein interaction (PPI) network. STRING integrates experimentally validated interactions and computational predictions from multiple sources [16]. The network was established based on AT-DEGs and Cytoscape (v3.10.3) for visualization and topological analysis. The molecular complex detection (MCODE) toolkit identified densely connected sub-networks. Additionally, the CytoHubba toolkit ranked genes according to their degree of connectivity. The ten most-connected genes emerged as candidate key regulators in autophagy-related processes [17].
- Results
3.1 Differential Expression Analysis
Mining the GSE74089 dataset yielded 3,648 DEGs from 21,754 genes: 2,483 upregulated and 1,165 downregulated. Furthermore, 612 autophagy-related genes were screened through the GO repository, of which 547 matched the GSE74089 expression profile and 106 intersected with the DEGs. A volcano graph visualizing the DEGs is presented in Figure 1. The Venn diagram and color-coded matrix of the 106 AT-DEGs are illustrated in Figures 2 and Figures 3, respectively.
Figure 1. Volcano plot of DEGs using thresholds set at |log2FC|>1.5 and p < 0.05, with red circles marking upregulated genes and blue circles marking downregulated genes.
Figure 2. Venn diagram illustrating overlap between DEGs filtered from GSE74089 and autophagy-associated genes selected from the GO database.
Figure 3. Clustered heat map of the 106 AT-DEGs, with annotations for the 10 most central genes selected from subsequent analysis.
3.2 GSEA Analysis
The expression profile of 547 autophagy-associated genes were analyzed employing the GSEA software (v4.4.0), with the KEGG gene set database applied to evaluate the overall patterns across the expression profile. The cutoff criteria were defined as FDR < 0.25 and p < 0.05. As shown in Figures 4 to Figures 9, the enrichment plots demonstrate that the majority of upregulated genes were involved in ubiquitin-mediated proteolysis, spliceosome, renal cell carcinoma, cell cycle, and mTOR and TGF-β signaling pathways.
Figure 4. GSEA result of upregulated genes involved in ubiquitin-mediated proteolysis.
Figure 5. GSEA result of upregulated genes involved in spliceosome.
Figure 6. GSEA result of upregulated genes involved in renal cell carcinoma.
Figure 7. GSEA result of upregulated genes involved in mTOR signaling pathway.
Figure 8. GSEA result of upregulated genes involved in cell cycle.
Figure 9. GSEA result of upregulated genes involved in TGF-β signaling pathway.
3.3 GO Analysis
GO enrichment analysis of AT-DEGs revealed that the biological processes in patients were primarily associated with autophagy, positive regulation of autophagy, protein transport, regulation of autophagy, autophagosome assembly, DNA damage response, and macroautophagy. The top seven enriched cellular component included cytosol, cytoplasm, membrane, nucleoplasm, mitochondrion, extracellular exosome, and endoplasmic reticulum. For molecular function, the most prominent GO terms were protein binding, ATP binding, ubiquitin binding, as well as protein kinase, protein serine/threonine kinase, and protein serine kinase activities. These GO enrichment results are visualized in the bar plot as shown in Figure 10.
Figure 10. GO analysis highlighted enriched BP, CC, and MF terms among AT-DEGs.
Figure 11. Enriched KEGG pathways in AT-DEGs.
3.4 KEGG Analysis
KEGG pathway assessment revealed that the target genes were significantly enriched in pathways, including neurodegeneration-multiple diseases, amyotrophic lateral sclerosis, shigellosis, endocytosis, Alzheimer disease, longevity regulating pathway, and FoxO signaling pathway. These enriched pathways are visualized in the bubble plot shown in Figure 11.
3.5 PPI Network Approach
The PPI network comprised 82 nodes interconnected by 264 edges. MCODE plugin analysis identified two high-scoring modules (scores >5) that represent critical functional units. KEGG enrichment revealed the first module to be enriched in neurodegeneration-linked pathways —Alzheimer’s, Huntington’s, amyotrophic lateral sclerosis (ALS), and spinocerebellar ataxia. The second module showed enrichment in pathways associated with endocytosis and necroptosis. The two modules were presented in Figures 12 to Figures 15. Using the CytoHubba toolkit, degree centrality pinpointed the ten highest-ranking hub genes: PIK3C3, UVRAG, ATG13, ATG12, RB1CC1, CASP3, FOXO3, BNIP3, ZFYVE1, and WIPI1, as shown in Figure 16.
Figure 12. PPI network diagram of module I
Figure 13. Bubble plot of KEGG analysis for module I.
Figure 14. PPI network diagram of module II.
Figure 15. Bubble plot of KEGG analysis for module II.
Figure 16. PPI network diagram of AT-DEGs. Darker red color indicates a higher degree of nodes.
- Discussion
Differential expression analysis identified 3,648 DEGs, including 2,483 upregulated and 1,165 downregulated genes between ONFH patients and healthy controls. Among these, 106 identified as autophagy-related DEGs, underscoring autophagy as a critical regulatory mechanism in the pathogenesis of ONFH. Gene dysregulation has been reported in ischemic bone diseases, wherein hypoxia-inducible factors (HIFs) modulate autophagy to adapt to hypoxic stress [18]. These results align with studies linking autophagy activation to glucocorticoid-induced osteocyte survival, emphasizing its dual role in bone homeostasis [19]. The heat map in Figure 2(b) demonstrated distinct clustering patterns between ONFH and control samples, indicating differential gene expression patterns. Co-expression analysis highlighted pathways related to autophagy and apoptosis, and CASP3 upregulation correlates with enhanced osteocyte apoptosis in ONFH, consistent with the established role of caspase-3 as a hallmark of glucocorticoid-induced bone cell death [20]. Moreover, the differential expression of ATG12 and RB1CC1, key drivers of autophagosome formation, suggests a potential defect in autophagosome maturation during femoral head necrosis [21]. These findings are in accordance with histopathological observations reporting the accumulation of autophagic vacuoles in osteocytes of ONFH patients [22]. The GSEA approach showed that the upregulated autophagy-associated genes were significantly enriched in pathways including ubiquitin-mediated proteolysis, the mTOR signaling pathway, the spliceosome, and TGF-β signaling pathway. These pathways are intimately associated with cellular survival, proliferation, and metabolism. Among them, the mTOR signaling pathway serves as a central regulator of cell growth and metabolism and plays a pivotal role in autophagy regulation. Activated mTOR inhibits autophagy by phosphorylating components such as ATG13 and ULK1, whereas suppression of mTOR activity induces autophagy, enabling cells to adapt to stress conditions [23]. In ONFH, dysregulated of the mTOR pathway may lead to aberrant autophagic activity, potentially compromising bone cell viability and function. Similarly, the TGF-β signaling is essential for osteoblast differentiation and bone formation and also modulates osteoclast activity and bone resorption. Its interaction with autophagy contributes to skeletal disorders [24]. The crosstalk between autophagy and the key pathways may contribute to the etiology of ONFH. A deeper understanding of these molecular mechanisms will advance the autophagy-targeted therapies to prevent or slow disease progression [25,26].
GO analysis revealed AT-DEGs over-represented in autophagosome assembly and protein transport, both of which are essential for maintaining cellular homeostasis under oxidative stress conditions [27]. Enrichment in terms related to the DNA damage response further supports the role of autophagy in mitigating oxidative injury—an established driver of osteocyte apoptosis in ONFH [28]. Cellular component annotations, including mitochondria and endoplasmic reticulum are consistent with previous findings implicating mitochondrial dysfunction and endoplasmic reticulum stress in impaired autophagic flux during ONFH progression [29]. Moreover, molecular functions such as ATP binding and ubiquitin binding emphasize the energy-dependent nature of autophagy and its reliance on proteasomal coordination [30]. The mTOR signaling pathway, as mentioned earlier, integrates various cellular signals and serves as a master modulator of autophagy [31]. The FoxO signaling also emerged as a key regulatory axis, with FoxO transcription factors promoting autophagy by activating autophagy-related genes. As for ONFH, the triggering of FoxO signaling may represent a protective mechanism against cellular stress [32]. The identification of these pathways provides a foundation for future research on ONFH pathobiology and might may inform the development of novel therapeutic strategies targeting autophagy regulation [33].
KEGG results revealed target genes enriched in pathways related to neurodegeneration-multiple diseases, shigellosis, amyotrophic lateral sclerosis, endocytosis, Alzheimer disease, longevity regulating pathway, and FoxO signaling pathway. The neurodegeneration-multiple diseases pathway comprises various conditions like Alzheimer’s and Parkinson’s diseases, marked by misfolded-protein buildup and organelle failure. Autophagy removes these deleterious cellular components, thereby preserving neuronal and cellular homeostasis [34]. The endocytosis pathway, which governs the internalization and recycling of extracellular materials and membrane components, is functionally linked to autophagy and contributes to the regulation of intracellular trafficking and metabolic balance [35]. Additionally, the longevity regulating pathway suggests a potential link between autophagy activation and delayed bone degeneration [36]. The enrichment of these pathways highlights the multifaceted roles of autophagy-related genes in ONFH, implicating them in a range of cellular processes and signaling cascades that may contribute to disease onset and progression.
PPI network analysis identified two high-scoring modules, each representing a critical functional unit. The first showed over-representation in neurodegeneration-linked pathways—Alzheimer’s, amyotrophic lateral sclerosis (ALS), Huntington’s, and spinocerebellar ataxia. This enrichment implies that module I might contribute to the pathogenesis of neurodegenerative conditions, which are known to share common mechanisms with ONFH, particularly oxidative stress and mitochondrial impairment. Studies have demonstrated that these two factors significantly accelerate neurodegeneration. In Alzheimer’s disease, oxidative stress impairs mitochondrial function, thereby affecting neuronal survival [37]. Similarly, in amyotrophic lateral sclerosis, both oxidative stress and mitochondrial dysfunction critically drive its pathogenesis [38]. Given these mechanistic parallels, it is plausible that the genes identified in this module may also be implicated in the onset and progression of ONFH through similar oxidative and mitochondrial pathways. The second module showed enrichment in pathways associated with endocytosis and necroptosis. Endocytosis is a fundamental cellular process that regulates the uptake of nutrients and signaling molecules, while necroptosis is a type of necrosis triggered by cellular stress. The involvement of these pathways suggests that the interplay among autophagy, endocytosis, and necroptosis likely contributes critically to ONFH pathogenesis. pathogenesis of ONFH. Impaired endocytosis can aggregate extracellular matrix constituents and damage tissue. For example, in certain pathological contexts, defective endocytosis hinders the clearance of matrix proteins, promoting fibrosis and structural deterioration [39].
Similarly, necroptosis significantly contributes to the pathogenesis of various inflammatory conditions, and its activation within bone cells might exacerbate bone necrosis in ONFH. Research shows necroptosis can promote inflammatory mediator release that exacerbates local inflammation, and accelerate bone matrix degradation.
Research has demonstrated that necroptosis can promote the secretion of inflammatory mediators, thereby intensifying local inflammation and contributing to further degradation of the bone matrix [40]. Moreover, the crosstalk between necroptosis and other cellular processes such as autophagy and endocytosis may also have an impact on the progression of ONFH. Emerging evidence suggests that autophagy-endocytosis networks and necroptosis reciprocally regulate their activation and progression. This bidirectional interaction represents a complex regulatory network that could be a key contributor to the pathophysiological mechanisms underlying ONFH [41].
Using the CytoHubba toolkit, we identified the top 10 core genes: PIK3C3, UVRAG, ATG13, ATG12, RB1CC1, CASP3, FOXO3, BNIP3, ZFYVE1, and WIPI1. These genes are well-acknowledged functions in autophagy regulation. PIK3C3, also referred to as VPS34, a class III PI3-kinase, drives autophagosome formation [42]. UVRAG suppresses tumors while modulating autophagy and DNA damage response [43]. ATG13, a key constituent of the ATG1/ATG13/ATG17 complex, significantly influences the autophagy initiation [44]. ATG12 is a ubiquitin-like modifier covalently attached to ATG5, forming a complex crucial for autophagosome elongation [45]. RB1CC1, also known as FIP200, is part of the ULK1 complex, crucial for autophagy induction [46]. CASP3 is a key executioner caspase involved in apoptosis, and its connection to autophagy highlights the crosstalk between these two cellular processes. FOXO3 transcriptionally activates autophagy-related genes under oxidative stress, thereby contributing to cellular stress adaptation. BNIP3, a pro-apoptotic protein, can also activate autophagy in hypoxic conditions. ZFYVE1 is implicated in autophagosome maturation, and WIPI1 critically contributes to the early stages of autophagosome formation by binding to phosphatidylinositol 3-phosphate.
The identification of these hub genes provides potential targets for therapeutic intervention in ONFH. For example, targeted modulation of their expression or functional activity, as well as the signaling pathways in which they participate, may offer novel strategies to regulate autophagy and mitigate disease progression. However, further experimental validation is required to determine their specific pathogenic roles in ONFH and therapeutic potential.
- Conclusions
This study presents a comprehensive bioinformatic analysis of autophagy-associated genes and pathways within ONFH. Through the integration of multiple bioinformatics approaches, the identified AT-DEGs were primarily linked to pathways including neurodegenerative diseases, the mTOR and TGF-β signaling pathways. The top ten hub genes —PIK3C3, UVRAG, ATG13, ATG12, RB1CC1, CASP3, FOXO3, BNIP3, ZFYVE1, and WIPI1—were identified as central regulators within the autophagic network and are likely to play critical roles in ONFH pathogenesis. These findings enhance our insight into the molecular pathology underlying ONFH and provide potential treatment objectives for future clinical translation. Future research might focus on verifying the functions of hub genes through in vitro experiments (such as steroid-treated osteocyte models) to clarify their effects on autophagic activity and cell survival, or exploring the therapeutic potential of drugs targeting key pathways (such as mTOR) in ONFH, to provide experimental evidence for developing precision therapies.
Acknowledgements
The authors gratefully acknowledge the financial support from the Cyanine Blue Project (Grant No.2025ZQZR46).
References
[1] Igei T, Nakasone S, Takaesu M, et al. Characteristics of osteonecrosis of the femoral head after COVID-19 steroid treatment compared to steroid-associated osteonecrosis[J]. BMC Musculoskeletal Disorders, 2025, 26(1): 352.
[2] Li Z, Shao W, Lv X, et al. Advances in experimental models of osteonecrosis of the femoral head[J]. Journal of Orthopaedic Translation, 2023, 39: 88-99.
[3] Chen C, Zhao X, Luo Y, et al. Imbalanced T-cell subsets may facilitate the occurrence of osteonecrosis of the femoral head[J]. Journal of Inflammation Research, 2022, Volume 15: 4159-4169.
[4] Mont M A, Salem H S, Piuzzi N S, et al. Nontraumatic osteonecrosis of the femoral head: where do we stand today?: a 5-year update[J]. Journal of Bone and Joint Surgery, 2020, 102(12): 1084-1099.
[5] He Q, Pan G, Yu Z, et al. Novel rim plating technique for treatment of the inferior pole fracture of the patella[J]. Orthopaedic Surgery, 2021, 13(2): 651-658.
[6] Tan Z, Wang Y, Chen Y, et al. The dynamic feature of macrophage M1/M2 imbalance facilitates the progression of non-traumatic osteonecrosis of the femoral head[J]. Frontiers in Bioengineering and Biotechnology, 2022, 10: 912133.
[7] Ma M, Tan Z, Li W, et al. Osteoimmunology and osteonecrosis of the femoral head[J]. Bone Joint Research, 2022, 11((1)): 26-28.
[8] Zhong C, Xu H, Chen J, et al. Human umbilical cord mesenchymal stem cells prevent steroid-induced avascular necrosis of the femoral head by modulating cellular autophagy[J]. Biomedicines, 2024, 12(12): 2817.
[9] Liu G, Cao R, Liu Q, et al. M2 macrophages-derived exosomes for osteonecrosis of femoral head treatment: modulating neutrophil extracellular traps formation and endothelial phenotype transition[J]. Bone Research, 2025, 13(1): 42.
[10] Tang Y, Yue Z, Li G, et al. Effect of β‑ecdysterone on glucocorticoid‑induced apoptosis and autophagy in osteoblasts[J]. Molecular Medicine Reports, 2017, 17: 158-164.
[11] Sorice M. Can the interplay between autophagy and apoptosis be targeted as a novel therapy for Parkinson’s disease?[J]. Cells, 2022, 11(9): 1479.
[12] Tang L, Zhang W, Liao Y, et al. Autophagy: a double-edged sword in ischemia–reperfusion injury[J]. Cellular & Molecular Biology Letters, 2025, 30(1): 42.
[13] Gene Expression Omnibus. Available online: https://www.ncbi.nlm.nih.gov/geo/ (accessed on 6th May 2025)[DS]. 2025.
[14] Hinderer E W, Flight R M, Dubey R, et al. Advances in gene ontology utilization improve statistical power of annotation enrichment[J]. PLOS One, 2019, 14(8): e0220728.
[15] Zhou Y, Zhou B, Pache L, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets[J]. Nature Communications, 2019, 10(1): 1523.
[16] Szklarczyk D, Gable A L, Nastou K C, et al. The STRING database in 2021: customizable protein–protein networks, and functional characterization of user-uploaded gene/measurement sets[J]. Nucleic Acids Research, 2021, 49(D1): D605-D612.
[17] Chin C H, Chen S H, Wu H H, et al. cytoHubba: identifying hub objects and sub-networks from complex interactome[J]. BMC Systems Biology, 2014, 8(S4): S11.
[18] Semenza G L. Hypoxia-inducible factors in physiology and medicine[J]. Cell, 2012, 148(3): 399-408.
[19] Shen G, Ren H, Shang Q, et al. Autophagy as a target for glucocorticoid-induced osteoporosis therapy[J]. Cellular and Molecular Life Sciences, 2018, 75(15): 2683-2693.
[20] Zhang S, Liu Y, Liang Q. Low-dose dexamethasone affects osteoblast viability by inducing autophagy via intracellular ROS[J/OL]. Molecular Medicine Reports, 2018. DOI:10.3892/mmr.2018.8461.
[21] Mizushima N, Yoshimori T, Ohsumi Y. The role of atg proteins in autophagosome formation[J]. Annual Review of Cell and Developmental Biology, 2011, 27(1): 107-132.
[22] Xia X, Kar R, Gluhak-Heinrich J, et al. Glucocorticoid-induced autophagy in osteocytes[J]. Journal of Bone and Mineral Research, 2010, 25(11): 2479-2488.
[23] Paquette M, El-Houjeiri L, Pause A. mTOR pathways in cancer and autophagy[J]. Cancers, 2018, 10(1): 18.
[24] Chen P Y, Qin L, Simons M. TGFβ signaling pathways in human health and disease[J]. Frontiers in Molecular Biosciences, 2023, 10: 1113061.
[25] Mizushima N, Komatsu M. Autophagy: renovation of cells and tissues[J]. Cell, 2011, 147(4): 728-741.
[26] Rubinsztein D C, Codogno P, Levine B. Autophagy modulation as a potential therapeutic target for diverse diseases[J]. Nature Reviews Drug Discovery, 2012, 11(9): 709-730.
[27] Trojani M C, Santucci-Darmanin S, Breuil V, et al. Autophagy and bone diseases[J]. Joint, Bone, Spine, 2022, 89(3): 105301.
[28] Zhang J, Cao J, Liu Y, et al. Advances in the pathogenesis of steroid-associated osteonecrosis of the femoral head[J]. Biomolecules, 2024, 14(6): 667.
[29] He Y, Liu T, Peng X, 等. Molecular mechanism of mitochondrial autophagy mediating impaired energy metabolism leading to osteoporosis[J]. Biochimica et Biophysica Acta (BBA) – Molecular Basis of Disease, 2025, 1871(3): 167685.
[30] Wang J, Zhang Y, Cao J, et al. The role of autophagy in bone metabolism and clinical significance[J]. Autophagy, 2023, 19(9): 2409-2427.
[31] Wang Y, Zhang H. Regulation of Autophagy by mTOR Signaling Pathway[J]. Advances in Experimental Medicine and Biology, 2019, 1206: 67-83.
[32] Webb A E, Brunet A. FOXO transcription factors: key regulators of cellular quality control[J]. Trends in Biochemical Sciences, 2014, 39(4): 159-169.
[33] Mizushima N, Levine B, Cuervo A M, et al. Autophagy fights disease through cellular self-digestion[J]. Nature, 2008, 451(7182): 1069-1075.
[34] Mizushima N, Klionsky D J. Protein turnover via autophagy: implications for metabolism[J]. Annual Review of Nutrition, 2007, 27(1): 19-40.
[35] Lamb C A, Dooley H C, Tooze S A. Endocytosis and autophagy: shared machinery for degradation[J]. Bioessays, 2013, 35(1): 34-45.
[36] Wang X, Jiao L, He J, et al. Parathyroid hormone 1‑34 inhibits senescence in rat nucleus pulposus cells by activating autophagy via the m‑TOR pathway[J/OL]. Molecular Medicine Reports, 2018. DOI:10.3892/mmr.2018.9229.
[37] Khan S, Barve K H, Kumar M S. Recent advancements in pathogenesis, diagnostics and treatment of alzheimer’s disease[J]. Current Neuropharmacology, 2020, 18(11): 1106-1125.
[38] Cunha-Oliveira T, Montezinho L, Mendes C, et al. Oxidative stress in amyotrophic lateral sclerosis: pathophysiology and opportunities for pharmacological intervention[J]. Oxidative Medicine and Cellular Longevity, 2020, 2020: 1-29.
[39] Zhao X, Chen J, Sun H, et al. New insights into fibrosis from the ECM degradation perspective: the macrophage-MMP-ECM interaction[J]. Cell & Bioscience, 2022, 12(1): 117.
[40] Galluzzi L, Kepp O, Chan F K M, et al. Necroptosis: mechanisms and relevance to disease[J]. Annual Review of Pathology: Mechanisms of Disease, 2017, 12(1): 103-130.
[41] Papandreou M, Tavernarakis N. Crosstalk between endo/exocytosis and autophagy in health and disease[J]. Biotechnology Journal, 2020, 15(8): 1900267.
[42] Parzych K R, Klionsky D J. An overview of autophagy: morphology, mechanism, and regulation[J]. Antioxidants & Redox Signaling, 2014, 20(3): 460-473.
[43] Liang C, Lee J soo, Inn K S, et al. Beclin1-binding UVRAG targets the class C vps complex to coordinate autophagosome maturation and endocytic trafficking[J]. Nature Cell Biology, 2008, 10(7): 776-787.
[44] Davies C W, Stjepanovic G, Hurley J H. How the Atg1 complex assembles to initiate autophagy[J]. Autophagy, 2015, 11(1): 185-186.
[45] Walczak M, Martens S. Dissecting the role of the Atg12–Atg5-Atg16 complex during autophagosome formation[J]. Autophagy, 2013, 9(3): 424-425.
[46] Jung C H, Jun C B, Ro S H, et al. ULK-Atg13-FIP200 complexes mediate mTOR signaling to the autophagy machinery[J]. Molecular Biology of the Cell, 2009, 20(7): 1992-2003.