The mechanism of atorvastatin-induced sarcopenia elucidated based on network toxicology, single-cell RNA sequencing, and bulk transcriptomics data

The mechanism of atorvastatin-induced sarcopenia elucidated based on network toxicology, single-cell RNA sequencing, and bulk transcriptomics data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The mechanism of atorvastatin-induced sarcopenia elucidated based on network toxicology, single-cell RNA sequencing, and bulk transcriptomics data Fei Zhao, Yue Zhao, Liangling Cai, Yuanhuan Wei, Fang Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9058922/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Sarcopenia (SA) significantly affects the quality of life in the elderly. Atorvastatin (ATS), a lipid-lowering drug, may cause myopathy as a side effect. This study aimed to predict key genes involved in ATS-induced SA using network. Methods Public databases were used to obtain single-cell RNA sequencing (scRNA-seq) data, transcriptome data, and targets for SA and ATS toxicity. Toxicity prediction was conducted, and candidate genes were identified through differentially expressed genes (DEGs) associated with SA, as well as ATS toxicity targets. Machine learning and gene expression analyses were used to identify key genes, followed by functional enrichment and immune infiltration analysis. Molecular docking and dynamics simulations assessed the binding affinity between ATS and key genes. Key cell types were identified through scRNA-seq analysis. Results The toxicity prediction indicated that ATS exhibited relatively low toxicity in experimental animals. A total of 16 candidate genes were identified from the intersection of 2,402 DEGs, 101 toxicity targets, and 7,859 targets. BACE1 and PDE4A were selected as key genes. Functional enrichment analysis revealed their association with Parkinson's disease, oxidative phosphorylation, Alzheimer's disease, and Huntington's disease. Immune infiltration analysis showed that BACE1 and PDE4A were positively correlated with effector memory CD4 T cells. Molecular docking confirmed strong binding affinity between ATS and key genes, forming stable complexes. Myeloid cells were identified as key cells involved in SA. Conclusions This study provides insights into the molecular mechanisms of ATS-induced SA and supports strategies for preventing ATS-induced myopathy. Sarcopenia Atorvastatin Network toxicology Single-cell RNA sequencing Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Background Sarcopenia (SA) is a degenerative disease characterized by muscle strength decline, decreased muscle mass, muscle fiber atrophy, and overall impairment of physical function [ 1 ]. Aging is the primary risk factor, with people over 50 losing approximately 1% of muscle mass annually [ 2 ]. SA prevalence reaches 14%-33% in individuals aged 65 and above, escalating to 50%-60% in those over 80, representing a major public health concern [ 3 ]. This condition not only causes muscle weakness and functional decline but also significantly increases the risks of falls, fractures, and mortality [ 4 ]. SA pathogenesis involves multifactorial interactions, including aging, neuromuscular dysfunction, hormonal changes, increased pro-inflammatory cytokines, chronic physical inactivity, and nutritional imbalances, all of which accelerate disease progression [ 5 ]. Secondary sarcopenia caused by other diseases or medications is often more preventable than primary forms [ 6 ]. Current treatment strategies encompass non-pharmacological approaches such as resistance exercise and adequate intake of protein, vitamin D, antioxidants, and long-chain polyunsaturated fatty acids. Regarding pharmacological interventions, although drugs like growth hormone, selective androgen receptor modulators, and myostatin inhibitors have demonstrated some efficacy in clinical trials, the FDA has not approved any specific therapeutic agents to date [ 7 ]. However, drug-related sarcopenia remains insufficiently investigated, emphasizing the importance of exploring SA's multiple causes and interactive mechanisms for developing effective prevention and treatment strategies. Atorvastatin (ATS) is a widely prescribed lipid-lowering medication that selectively inhibits 3-hydroxy-3-methylglutaryl coenzyme A reductase, a key enzyme in hepatic cholesterol biosynthesis, thereby reducing plasma cholesterol, low-density lipoprotein (LDL), and triglyceride levels [ 8 ]. As a representative statin drug extensively used in elderly patients, ATS carries significant clinical relevance for drug-related sarcopenia, which is particularly prominent in elderly patients with polypharmacy [ 6 ]. The most common adverse effect of ATS is drug-induced myopathy, presenting with diverse clinical manifestations including muscle weakness, myalgia, muscle tenderness, spasms, and arthralgia, with or without elevated serum creatine kinase levels. In rare cases, patients may develop rhabdomyolysis, a serious and potentially life-threatening condition [ 9 ]. Clinical studies indicate that approximately 10%-25% of statin-treated patients experience muscle-related problems [ 10 ]. However, the specific molecular mechanisms by which ATS induces sarcopenia development remain insufficiently understood, with no systematic studies elucidating these pathways, representing an important area requiring further investigation. Network toxicology is an emerging discipline developed based on network pharmacology and network biology. By integrating bioinformatics, big data analysis, and multi-omics techniques, it constructs complex network relationships among chemicals, toxic effects, and targets. This approach enables systematic analysis of the mechanisms of multi-component, multi-target toxicants, predicts molecular pathways involved in disease induction, and demonstrates rapid, convenient, and comprehensive advantages in toxicity mechanism research [ 11 , 12 ]. Selecting the network toxicology method helps to predict key targets and pathways related to ATS-induced sarcopenia from a systemic perspective, providing a theoretical basis for subsequent experimental validation. Single-cell RNA sequencing (scRNA-seq) technology investigates the transcriptomic features of individual cells, revealing gene expression differences among cells and overcoming the limitations of traditional RNA sequencing that obscure cell heterogeneity. This technology has unique advantages in identifying cell types, cell subpopulations, and state transitions, offering an important tool for understanding cellular heterogeneity and communication networks [ 13 , 14 ]. Using scRNA-seq allows validation of key genes predicted by network toxicology at the single-cell level and reveals the specific effects of ATS on different muscle cell subpopulations. Integrating network toxicology with scRNA-seq enables a complete research chain from macro-level predictions to micro-level validations, facilitating systematic prediction of the molecular mechanisms of ATS-induced sarcopenia and verifying key targets at the single-cell level, thereby providing a comprehensive and precise research strategy for in-depth understanding of drug-induced sarcopenia pathogenesis. In this study, we identify the toxic key genes of ATS leading to SA pathogenesis by transcriptome combined with network toxicology, use a series of bioinformatics methods to explore the specific mechanism of key genes in the development of SA, and then explore the expression of key genes in key cells based on single-cell data, providing new theoretical support and basis for the pathogenesis of SA. 2. Methods 2.1 Data collection The data were obtained from the GEO database ( https://www.ncbi.nlm.nih.gov/geo/ ). The GSE1428 dataset (Sequencing platform: GPL96) was employed as a training set, which included 12 samples of skeletal muscle tissue from SA patients and 10 control skeletal muscle tissue samples. The GSE38718 dataset (Sequencing platform: GPL570) served as a validation set, which comprised 8 samples of skeletal muscle tissue from SA patients and 14 control skeletal muscle tissue samples. In addition, the scRNA-seq data were obtained from the CNGB Nucleotide Sequence Archive (CNP0004394, CNP0004395, CNP0004494, and CNP0004495) [ 15 ]). All processed data were available at the Human Muscle Aging Cell Atlas database ( https://db.cngb.org/cdcp/hlma/ ). The scRNA-seq data samples primarily consisted of 4 skeletal muscle tissue samples from SA and 3 control skeletal muscle tissue samples. 2.2 Differential expression analysis and collection of SA-related targets To identify DEGs between the SA and control groups, DEGs were carried out by employing the “limma” (v 3.54.0) [ 16 ] based on a training set for both groups of samples (|log 2 FC| > 0.5, P < 0.05). Moreover, the “ggplot2” (v 3.4.1) [ 17 ] was employed to construct a volcano plot of DEGs, and the top 10 genes (sorted by |log 2 FC| from high to low) with significant up- and down-regulation differences were labeled, and then the heatmap of these genes was drawn using the “ComplexHeatmap” (v 2.14.0) [ 18 ]. To identify the targets associated with SA, the CTD database ( https://ctdbase.org/ ) was searched using the keyword “sarcopenia”, which was documented as the target for SA. 2.3 Prediction of ATS toxicity and identification of toxicity targets In order to understand the toxicity information of ATS, the SMILES molecular structure of ATS compounds was first obtained in the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ). Based on this SMILES, the toxicity of ATS was predicted using the ProTox 3.0 database ( https://tox.charite.de ). Finally, a search was conducted in the Swiss Target Database ( http://swisstargetprediction.ch/ ) using the SMILES notation of ATS. To ensure data accuracy, the collected information was consolidated, duplicates were removed, and the remaining entries were classified as toxic targets of ATS. 2.4 Identification, functional enrichment, and PPI network of candidate genes To identify candidate genes associated with toxicity in ATS that contributed to the development of SA, the “ggvenn” (v 1.7.3) [ 19 ] was employed to determine the intersection of DEGs, toxic targets, and SA-related targets. These genes were subsequently documented as candidate genes for further analysis. Subsequently, to explore the biological pathways of candidate genes and their functional enrichment, candidate genes were analyzed by GO and KEGG using “clusterProfiler” (v 4.2.2) [ 20 ] (P 0.4 by “Cytoscape” (v 3.9.1) [ 21 ]. Next, candidate genes with interactions in the PPI network were selected for further study. 2.5 Network construction of compound-target-pathway-disease In order to understand the connections between the candidate genes, compounds, pathways, and diseases, we integrated the candidate genes, KEGG-enriched pathways (specifically, the top 10 pathways with the highest number of associated genes), and SA diseases. We utilized “Cytoscape” (v 3.9.1) to visualize the compound-target-pathway-disease network diagram, illustrating the complex regulatory network involved in the development of ATS-induced SA. 2.6 Machine learning and expression validation In order to further refine the selection of candidate genes, we performed LASSO and SVM-RFE analyses on the training set. The “glmnet” (v 4.1-4) [ 22 ] was employed to perform LASSO regression of candidate genes. The optimal lambda value was identified using 5-fold cross-validation to screen for LASSO genes. A SVM-RFE algorithm of candidate genes was conducted using the “e1071” (v 1.7–13) [ 23 ], which was employed to screen for the most optimal SVM-RFE genes based on 5-fold cross-validation. Furthermore, the “ggvenn” (v 1.7.3) was applied to obtain the intersection genes between 2 algorithm genes, which were defined as candidate key genes. An investigation was conducted to compare the expression of candidate key genes in the SA and control groups, utilizing the GSE1428 and GSE38718 datasets. The Wilcoxon test was employed to analyze the datasets. Genes exhibiting consistent expression trends and significant inter-group differences (P < 0.05) in both datasets were identified as key genes. 2.7 GSEA The GSEA was carried out with the objective of elucidating the biological functions of key genes throughout the progression of SA. The initial step in the research was to calculate the Spearman correlation coefficients between the various key genes and every single gene across the entire range of samples from the training set, with the “psych” (v 2.2.9) [ 24 ] being utilized for the purpose. Following that, the sorted genes were then arranged in descending order according to their respective Spearman correlation coefficients, with this ranking then being utilized as the gene set that was tested in further analyses. Meanwhile, “c2.cp.kegg_legacy.v2024.1.Hs.symbols” was retrieved from the MSigDB ( https://www.gsea-msigdb.org/gsea/msigdb ). Subsequently, the GSEA was performed employing the “clusterProfiler” (v 4.2.2), with a threshold of |NES| > 1 and adjust P < 0.05. 2.8 Immune infiltration analysis According to the ssGSEA algorithm, the GSVA package (v 1.46.0) [ 25 ] was utilized to determine the infiltration scores of 28 kinds of immune cells [ 26 ] between SA and control in the training set. Next, the Wilcoxon test was applied to make a comparison of the differences in immune-infiltrating cells between the 2 groups (P 0.3, P < 0.05). 2.9 Molecular docking To further assess the binding capacity of ATS with key genes and to enhance the accuracy of the interaction network between ATS and these genes, molecular docking was conducted for each key gene in conjunction with ATS. The structure file of the ATS was sourced from the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ), and the protein data for the key genes were extracted from the PDB database ( http://www.rcsb.org ). Subsequently, molecular docking was performed using the CB-Dock2 website ( https://cadd.labshare.cn/cb-dock2/php/index.php ). The binding activity was deemed to be satisfactory if the affinity was lower than − 5.0 kcal/mol. 2.10 Molecular dynamics simulation (MDS) MDS lasting 100 nanoseconds were conducted using the “GROMACS” (v 2024.2) ( https://manual.gromacs.org ) to evaluate the binding stability of the receptor-ligand complex. Parameter and topology files for proteins and small molecule ligands were generated using the AMBER14SB force field and the AMBER GAFF force field, respectively. Periodic boundary conditions were implemented, and the system was neutralized with NaCl at a concentration of 0.15 mol/L. To minimize the energy of the system, the steepest descent method was employed. The pre-balancing procedure was carried out in 2 stages. In the initial stage, a simulation was conducted using an NVT ensemble at 300 K (temperature) for 100 ps to stabilize the system's temperature. In the subsequent stage, a simulation was performed using an NPT ensemble at 1 bar (pressure) for 100 ps to stabilize the system's pressure. The analysis of the simulation included the root mean square deviation (RMSD), root mean square fluctuation (RMSF), hydrogen bonding, and energy analyses. 2.11 Treatment of scRNA-seq data In the scRNA-seq dataset, the “Seurat” (v 5.0.1) [ 27 ] was utilized for quality control (quality control criteria: 1,000 < nFeature_RNA < 4,000, 3 < nCount_RNA < 15,000, percent.mt < 5%). Then the FindVariableFeatures function was used to extract the top 2,000 HVGs for subsequent analysis. The data were scaled using the ScaleData function. Statistically significant principal components were identified using the JackStrawPlot function. PCA was conducted on the top 2,000 HVGs. Subsequently, from the results obtained via the JackStraw function, the top 50 principal components (PCs) were chosen (P < 0.05). Subsequently, the resolution parameter was configured to 0.1, and an unsupervised clustering analysis was carried out on every cell within the dataset by utilizing the FindClusters function and FindNeighbors function. Then, the outcomes of the clustering were visualized through the application of the RunTSNE function. Cell types of the different clusters were identified after dimensionality reduction and clustering, employing the reference [ 15 ]. The annotation results were visualized on a t-SNE plot, and the expression of marker genes across different cell types was depicted. The utilization of histograms was essential for visualizing cell cluster types and their corresponding percentages within both SA and control samples. To gain insights into the biological functions of the annotated cells involved, the functional enrichment of annotated cells was explored through the pathways function in the “ReactomeGSA” (v 1.12.0) [ 28 ]. 2.12 Selection of key cells To further identify key cells within the scRNA-seq dataset, t-SNE maps were generated using the “Seurat” (v 5.0.1) to illustrate the distribution of key genes across the annotated cells. Subsequently, the Wilcoxon test was employed to compare the expression differences of key genes between the SA and control groups across the annotated cells (P < 0.05). Cells exhibiting significant differences were designated as key cells. The DoRothEA algorithm is a crucial tool for analyzing cellular gene regulation, primarily focusing on the interactions between transcription factors (TFs) and their target genes. Utilizing the “viper” package (v 1.34.0) [ 29 ], all TFs in key cells were identified, and the differences in activity between the SA and control groups were compared. 2.13 Cell communication and pseudotime analyses In order to characterize the interactions among annotated cells, as well as between key cells and other cells in the scRNA-seq dataset, the “CellChat” (v 1.6.1) [ 30 ] was employed to analyze cell communication. This analysis illustrated the network of interactions among various cell types. Subsequently, “CellChat” (v 1.6.1) was utilized to identify ligand-receptor (L-R) interactions between key cells and other cells. Meanwhile, to identify the differentiation process of key cells and the expression of prognostic genes at different differentiation stages, the RunTSNE function was used to perform dimensionality reduction and clustering on the key cells (resolution = 0.1). Subsequently, the cells were re-clustered into subclusters based on the marker genes identified in the existing literature [ 31 ]. Then, the “Monocle” (v 2.22.0) [ 32 ] was used for pseudotime analysis. The expression changes of key genes throughout the pseudotime progression were also analyzed. 2.14 Statistical analysis Bioinformatics analyses were conducted utilizing the R language (v 4.2.2). A Wilcoxon test was employed to compare the disparities between the 2 groups. A P-value of less than 0.05 was deemed statistically significant. 3. Results 3.1 The 16 candidate genes of SA were associated with toxicity in ATS Firstly, SMILES notation and molecular structure for ATS were obtained from the PubChem database (Fig. 1 a). The toxicity prediction result indicated that the LD50 of ATS was 5,000 mg/kg, placing it in toxicity class 5. This classification suggested that the compound exhibited relatively low toxicity in experimental animals. The confidence level of the prediction was 67.38%, indicating a moderate degree of reliability in the results. Furthermore, ATS demonstrated a higher potential for neurotoxicity, nephrotoxicity, and respiratory toxicity compared to other types of toxicity (Fig. 1 b). The 102 toxic targets of ATS were identified by the Swiss Target database ( Additional file 1 ). Differential expression analysis showed that there were 2,402 DEGs between the SA group and the control group. Among them, 29 genes in the SA group were identified as upregulated genes, and 2,373 genes were identified as downregulated genes (Fig. 1 c-d). A total of 7,859 targets for SA were identified by the CTD database ( Additional file 2 ). Moreover, a total of 16 shared genes between DEGs, toxic targets, and targets for SA were identified and selected as candidate genes (Fig. 1 e). The enrichment analysis of the 16 candidate genes revealed associations with 549 GO terms, including response to lipopolysaccharide, cell cortex, and metalloendopeptidase activity, among others (Fig. 1 f, Additional file 3 ). Moreover, the KEGG enrichment analysis of the candidate genes revealed that a total of 96 KEGG pathways were enriched, for instance relaxin signaling pathway, the IL-17 signaling pathway, and lipid and atherosclerosis (Fig. 1 g, Additional file 4 ). Next, a PPI network was constructed, comprising 20 interactions associated with 16 candidate genes (Fig. 1 h). In this network, MMP9, NOS2, MMP13, and MAPK8 had frequent protein-level interactions with other genes. In the “compound-target-pathway-disease” network, it was possible to visualize the connections between candidate genes, compounds, pathways, and diseases. Notably, 10 candidate genes (MMP1, MMP13, MAPK8, NOS2, GRB2, PIK3CB, CASP3, IKBKE, BACE1, RARB) were associated with compounds, diseases, and enriched pathways (Fig. 1 i). 3.2 The BACE1 and PDE4A were identified as key genes For the results of the LASSO algorithm, when the minimum lambda value was 0.1472063, a total of 5 LASSO genes were identified among the candidate genes, namely FOLR1, MMP13, BACE1, NOS2, and PDE4A (Fig. 2 a-b). When applying the SVM-RFE algorithm, the model demonstrated the highest prediction accuracy with an optimal variable count of 7. This configuration enabled the selection of 7 SVM-RFE genes from the candidate genes: FOLR1, PDE4A, MMP13, NOS2, PIK3CB, BACE1, and MMP12 (Fig. 2 c). Subsequently, we intersected the genes obtained from the mentioned above machine learning algorithms and finally found that FOLR1, PDE4A, MMP13, NOS2, and BACE1 could be indicated by all the algorithms, meaning that these genes could be used as candidate key genes (Fig. 2 d). Subsequently, according to the GSE1428 and GSE38718 datasets, BACE1 and PDE4A exhibited markedly low expression levels in the SA group (Fig. 2 e-f). Consequently, 2 key genes—BACE1 and PDE4A—were selected for subsequent analysis. 3.3 Exploring the function of key genes and immune infiltration of SA GSEA analysis observed that pathways associated with the expression of BACE1 and PDE4A were 8, respectively ( Additional file 5 ). Notably, biomarkers were co-enriched in several common pathways, such as parkinsons disease, oxidative phosphorylation, alzheimers disease, and Huntington's disease (Fig. 3 a-b). The distribution of immunity scores for the 28 immune cell types in the SA and control groups was illustrated in the thermal map, which displayed varying degrees of infiltration for each immune cell type (Fig. 3 c). Differences in immune cell infiltration between the 2 groups revealed that the 4 immune cell types exhibited significant variations, including CD56dim natural killer cells, effector memory CD4 T cells, neutrophils, and regulatory T cells, all of which were notably less abundant in the samples from the SA group (Fig. 3 d). In addition, the association analysis of biomarkers with the most differentially infiltrated immune cell types revealed predominantly significant positive correlations (Fig. 3 e). BACE1 and PDE4A were positively correlated with effector memory CD4 T cells (r > 0.40, P < 0.05). These results suggested that BACE1 and PDE4A might have the same roles in regulating immune cell infiltration and function, potentially influencing the pathogenesis and immune regulation in SA. 3.4 Detection of ATS's ability to bind to key genes According to the molecular docking results, ATS demonstrated superior binding affinity to key genes BACE1 and PDE4A, with binding energies of less than − 8 kcal/mol ( Additional file 6 ). The conformation indicated that lifitegrast interacted with the residues of the key genes BACE1 and PDE4A through hydrogen bonds (Fig. 4 a-b). To validate the reliability of molecular docking, MDS was implemented in this study. The proteins reached a steady state within the time frame of 0 ns to 50 ns. This suggested that the interactions between BACE1-ATS and PDE4A-ATS exhibited a relatively stable binding conformation (Fig. 4 c). The findings from the RMSF analysis indicated that the amino acids of the protein exhibited flexibility and maintained stable binding interactions with the small molecule drug ligands throughout the duration of the simulation (Fig. 4 d). As can be seen in Fig. 4 e, the number of hydrogen bonds fluctuated periodically throughout the duration of the simulation, indicating a dynamic equilibrium at the binding interface. In addition, the energy of the proteins exhibited less fluctuation throughout the duration of the simulation and remained in a steady state (Fig. 4 f). The above analysis demonstrated that the binding of BACE1-ATS and PDE4A-ATS was relatively stable throughout the simulation process. 3.5 The 9 cell types were annotated in the SA scRNA-seq data First, quality control processing was performed on the raw data of the scRNA-seq data for subsequent analysis. Before quality control, the initial count comprised 79,649 cells and 48,355 genes. Subsequent to quality control, the cell count was refined to 72,378, while the gene count remained constant at 48,355 ( Figure S1 a ). Second, the top 2,000 HVGs were identified, and the top 10 genes with the greatest variability included APOD, DCN, and CXCL14 ( Figure S1 b ). After PC of 30, the significance decreased, and the curve in the PC scree plot flattened at PC = 30 ( Figure S1 c ). Therefore, we chose the top 30 PCs for additional analysis. After that, cells were partitioned into 22 cell clusters ( Figure S1 d ). Relying on the expression of marker genes ( Figure S1 e ), the 22 cell clusters were annotated into 9 cell types: endothelial cells (EC), fibroblast activation protein (FAP), lymphocytes, mast cells, muscle satellite cells (MuSC), myeloid cells, smooth muscle cells (SMC), fibroblast-like, and type II cells ( Figure S1 f ). The percentages of FAP, SMC, fibroblast-like, type II cells, and mast cells were higher in the SA samples than in the control samples, whereas the percentages of MuSC, EC, and myeloid cells were greater in the control group than in the SA group ( Figure S1 g ). Functional enrichment analysis revealed type II cells and mast cells were predominantly enriched in pathways related to “synthesis of Hepoxilins (HX) and Trioxilins (TrX)” and “Reactions specific to the hybrid N-glycan synthesis pathway”, and myeloid cells exhibited obviously enrichment in pathways such as the “metabolism of ingested MeSeO2H into MeSeH”, suggesting their involvement in substance metabolism ( Figure S1 h ). 3.6 Myeloid cells were identified as key cells First, 2 key genes, BACE1 and PDE4A, were expressed and distributed across 9 annotated cell types (Fig. 5 a). It was further discovered that there was a significant difference in the expression of the BACE1 gene between the SA and control groups in FAP, MuSC, and myeloid cells (P < 0.05) (Fig. 5 b). In contrast, the PDE4A gene exhibited significant differences in only 2 groups of samples within myeloid cells (Fig. 5 c). We identified TFs in myeloid cells that exhibited differing levels of activity between the SA and control groups (Fig. 5 d). For instance, ZEB2, HBP1, FOXP1, and HHEX demonstrated high activity in the control group, whereas TCF12, MEF2A, TEAD1, and HOXB13 showed increased activity in the SA group. This suggested that these TFs might have distinct regulatory roles in various physiological or pathological conditions. 3.7 Myeloid cells exhibited strong interactions with other cells in samples from the SA group A detailed analysis of cell communication revealed that, overall, the number of interactions among the 9 annotated cell types was increased, and the strength of these interactions was increased in the SA samples compared to the control samples (Fig. 6 a-b). Specifically, the interactions of myeloid cells with fibroblast-like, EC, and mast cells were diminished, while interactions with FAP were increased in control samples compared to samples from patients with SA (Fig. 6 c-d). Furthermore, the LGALS9-CD45 signaling axis was identified as a key mediator in the interactions between myeloid cells and lymphocytes in the SA group, while the LGALS9-CD45 signaling axis was crucial for interactions among myeloid cells in the control group (Fig. 6 e-f). These findings suggested distinct signaling mechanisms underlying cell communication in different disease contexts. 3.8 Key genes exhibited distinct expression trends during myeloid cells differentiation trajectories To further assess the cellular differentiation trajectory of the myeloid cells, they were first re-clustered into different cell subclusters, with myeloid cells divided into 7 subclusters (Fig. 7 a). Relying on the expression of marker genes, the 7 subclusters were annotated into 4 myeloid cell subclusters: type 2 conventional dendritic cells (cDC2), cDC3, macrophages, and monocytes (Fig. 7 b-c). Pseudotime analysis of myeloid cell differentiation uncovered a progressive trajectory from an early (dark blue) to a more mature (light blue) state, and the 4 subclusters could be roughly divided into 5 differentiation states, with macrophages representing the earliest stages of differentiation (Fig. 7 d). During cell differentiation, the expression of the key gene BACE1 remained stable during the pre-differentiation stage, increased during the middle stage (monocytes), and subsequently decreased before stabilizing in the late stage. In contrast, the expression of the key gene PDE4A initially decreased during the early stage of cell differentiation (macrophages) and then increased during the middle and late stages (monocytes, cDC2, cDC3) (Fig. 7 e). The results indicated that BACE1 and PDE4A were strongly correlated with the progression of the SA. 4. Discussion SA is a degenerative disease that seriously affects the prognosis and quality of life of elderly patients. Aging is the main pathogenic factor, along with secondary causes caused by other diseases or medications. ATS is one of the representative drugs of the statin class and is widely used in elderly patients, with the most common side effect being drug-induced myopathy. This study identified toxic key genes involved in ATS-induced SA pathogenesis through transcriptomics combined with network toxicology, ultimately finding two key genes, BACE1 and PDE4A. Further analysis was conducted on the biological pathways involving these key genes, their relationship with immune infiltration, and the binding capacity between ATS and the key genes. Additionally, single-cell analysis revealed significant differential expression of these key genes in myeloid cells and analyzed their expression trends during myeloid cell differentiation, aiming to provide new directions for SA treatment[ 1 , 4 , 6 , 8 ]. BACE1 and PDE4A, two key genes, showed significantly down-regulated expression in both the training set and validation set of sarcopenic groups, a finding that warrants attention. Currently, there are no relevant reports on the role of these two genes in the pathogenesis of sarcopenia, which suggests that they may represent a new molecular mechanism of ATS-induced sarcopenia. Beta-site amyloid precursor protein cleaving enzyme 1 (BACE1), located on human chromosome 11, is one of the chromosomes most frequently associated with human diseases. Studies have reported that downregulation of BACE1 induces an increase in Glypican-1(GPC-1) expression, while upregulation of GPC-1 concurrently enhances the expression of endothelial nitric oxide synthase (eNOS). Inhibition of BACE1 promotes an increase in eNOS expression by upregulating GPC-1 [ 33 ]. Under exposure to risk factors such as hypertension, hyperlipidemia, diabetes, and aging, the upregulation of eNOS protein expression may lead to a relative deficiency in L-arginine or reduced Nicotinamide Adenine Dinucleotide Phosphate (NADPH) supply, promoting eNOS uncoupling, resulting in increased superoxide anions and peroxynitrite, causing oxidative stress [ 34 ]. Chronic low-grade inflammation caused by oxidative stress has been shown to be harmful to human skeletal muscle[ 35 , 36 ]. Normally, skeletal muscle proteins are balanced, continuously degraded and resynthesized; however, in aging and the associated increase in oxidative stress, this balance is disrupted, promoting the occurrence of sarcopenia [ 37 , 38 ].In summary, although BACE1 inhibition upregulates eNOS expression, this pathway may ultimately lead to a dysregulation of skeletal muscle protein balance and promote the occurrence of sarcopenia through the cascade reaction of oxidative stress and chronic inflammation, especially when multiple risk factors are present. cAMP-specific phosphodiesterase 4A (PDE4A) plays an important role in regulating intracellular cyclic adenosine monophosphate (cAMP) levels. PDE4A can degrade cAMP [ 39 ], and its downregulation leads to significant accumulation of cAMP. Abnormally elevated cAMP levels can affect muscle function and sarcopenia development through two main pathways. In the neuromuscular function pathway, excess cAMP inhibits inhibitory signals of γ-aminobutyric acid ergic(GABAergic) neurons, resulting in weakened calcium ion elevation triggered by GABA release, thereby affecting neural signal transmission [ 40 ]. The neuromuscular junctions (NMJs), as critical structures transmitting nerve signals to muscles, when dysfunctional, can cause muscles to lose normal innervation, subsequently promoting muscle atrophy [ 41 , 42 ]. Regarding myogenic differentiation, increased cAMP levels and excessive activation of cAMP-dependent protein kinase (PKA) directly suppress the myogenic differentiation process [ 43 ]. This inhibitory effect may be achieved by downregulating the activity of muscle-specific transcription factors Myf-5 and MyoD, both of which are basic helix-loop-helix (bHLH) proteins that are key regulators of myogenic cell differentiation and muscle-specific gene transcription activation. Therefore, PDE4A downregulation may synergistically promote the development of sarcopenia through these interconnected pathways, affecting both neural regulation and muscle regeneration. The results of gene set enrichment analysis (GSEA) showed that the two key genes BACE1 and PDE4A are jointly enriched in the oxidative phosphorylation (OXPHOS) pathway and various neurodegenerative disease-related pathways, including Parkinson's disease, Alzheimer's disease (AD), Huntington's disease, and others. These enriched pathways may be associated with the pathogenesis of sarcopenia (SA). Oxidative phosphorylation, as the core metabolic pathway of mitochondria, and neurodegenerative diseases, through mechanisms such as neuro-muscular disjunction, both play important roles in the development of sarcopenia. Mitochondrial dysfunction, inflammatory response, neuro-muscular disjunction, and insulin resistance are considered critical links in the sarcopenia signaling network, with mitochondrial dysfunction regarded as a central regulatory hub [ 44 – 47 ]. As a metabolic hub for energy production, mitochondria generate ATP for skeletal muscle contraction through oxidative phosphorylation, and a decline in nicotinamide adenine dinucleotide (NAD + ) levels is related to many age-related diseases, including cognitive decline, metabolic diseases, SA, and weakness [ 48 ]. Studies have found that humans of different races with SA show reduced mitochondrial oxidative capacity and NAD + biosynthesis; lowering NAD + levels through ablation of nicotinamide phosphoribosyltransferase (NAMPT)-mediated NAD salvage may lead to age-related muscle degeneration in adult mice [ 49 , 50 ]. It is hypothesized that BACE1 and PDE4 influence mitochondrial function by participating in oxidative phosphorylation reactions, thereby affecting the pathogenesis of SA, leading to AST. Additionally, in the context of neurodegenerative disease pathways, the results of gene set enrichment analysis in this study indicate that key genes BACE1 and PDE4A are co-enriched in pathways related to neurodegenerative diseases such as Parkinson's disease, Alzheimer's disease, and Huntington's disease, suggesting that these genes may regulate common pathological mechanisms between myasthenia and nervous system disorders. Although these neurodegenerative diseases present with various clinical features, they all influence muscle function through shared mechanisms such as neuro-muscular denervation, neuroinflammation, and protein misfolding [ 51 ]. The study found that abnormal proteins in Alzheimer's, Parkinson's, and Huntington's diseases (tau protein, alpha-synuclein, and huntingtin protein, respectively) have similar damaging capabilities when invading brain cells. Once inside, these proteins damage or rupture vesicle membranes, allowing proteins to infiltrate the cytoplasm and cause functional impairments. The protein aggregates associated with these three diseases lead to a similar vesicle damage mechanism [ 52 ]. Similarly, protein aggregation and membrane structural damage are also observed in the pathogenesis of myasthenia. Research indicates that in elderly human skeletal muscle, 43% of 515 insoluble protein aggregates show a more than 1.5-fold variation in abundance across different age groups, with 15% showing significant differences, implying that insoluble protein aggregates are especially susceptible to aging and may play a pathogenic role similar to that in neurodegenerative diseases in myasthenia [ 53 ]. These findings support the idea that BACE1 and PDE4A may play critical roles in neurodegenerative diseases and myasthenia by regulating protein homeostasis and membrane integrity. The immune infiltration results showed significant differences in immune cells between the SA group and the control, including CD56dim natural killer cells, Effector memory CD4 T cells, neutrophils, and regulatory T cells. Among them, Effector memory CD4 T cells are significantly positively correlated with two key genes, BACE1 and PDE4A. Effector memory CD4 T cells (TEM CD4 cells), with mitochondrial metabolism, are crucial for the exit of T cells from the quiescent state [ 54 ]. Quiescent CD4 T cells in peripheral blood exist as naive cells, effector memory cells, and central memory cells [ 55 ]. TEM CD4 cells are antigen-experienced cells that can produce rapid and robust responses upon re-encountering the same antigen. Mitochondria, functioning as the "energy factory" and "reactive oxygen species (ROS) source" in muscle cells, play a core role in the development of sarcopenia when their function is abnormal. Studies have shown that mitochondrial ATP and ROS metabolism are critical in platelet-regulated TEM CD4 cell responses, with PF4 being the main mediator of platelet-regulated TEM CD4 cell responses. TEM CD4 cells respond to PF4 binding to their CXCR3 receptor by increasing expression of mitochondrial transcription factor A (TFAM), enhancing mitochondrial biogenesis, and increasing production of ATP and ROS. These changes, in turn, stimulate the expression of T-bet and FoxP3, promoting the TEM CD4 cell immune response [ 56 ]. Meanwhile, ROS accumulation can activate the JNK pathway, amplifying inflammatory signals and atrophic processes [ 57 – 59 ]. Both TEM CD4 cells and mitochondria collaboratively promote progressive declines in muscle mass and strength by disrupting muscle energy homeostasis, triggering oxidative stress, and activating catabolic pathways. In summary, BACE1 and PDE4A not only co-enriched in the oxidative phosphorylation pathway in pathway enrichment analysis, but also showed a significant positive correlation with TEM CD4 cells, suggesting that these two genes may influence the immune-metabolic network of TEM CD4 cells by regulating mitochondrial oxidative phosphorylation processes. Subsequently, they jointly drive the development and progression of sarcopenia through oxidative stress and inflammatory responses mediated by ROS. Molecular docking and molecular dynamics results show that key genes BACE1 and PDE4A have strong binding activity. Molecular dynamics simulations further validate the reasonableness and reliability of the docking results. The good molecular binding activity and the stability of the docking results suggest that ATS plays a certain role in inducing the occurrence and development of SA, and how to prevent the further development of SA while patients are treated with ATS drugs remains to be studied. Single-cell data analysis revealed that myeloid cells are key cells in the process of ATS-induced SA occurrence and development, and myeloid cells are further divided into subgroups such as macrophages. Fundamental changes in the immune system of aged animals occur at an early stage of hematopoietic development, with differentiated hematopoietic stem cells (HSCs) almost generating all blood cells; in aging, lymphoid potential decreases and shifts toward myeloid lineage [ 60 ]. This age-related bone marrow bias is reflected in the changing number of mature bone marrow cells in circulation, with the number of myeloid dendritic cell subpopulations gradually decreasing with age, accompanied by a reduction in CD34 + progenitors and an increase in circulating monocytes, indicating a partial disruption of the entire antigen-presenting cell differentiation process in the elderly [ 61 ]. Some studies report that transplanting young bone marrow cells into elderly recipients can prevent sarcopenia and age-related changes in muscle fiber phenotype, while transplanting aging bone marrow cells into young animals reduces satellite cell numbers and promotes their transition into a fibrotic phenotype. Aging of bone marrow cells promotes a decline in satellite cell number and function, leading to sarcopenia. Myeloid cell subpopulations, such as macrophages, play a crucial regulatory role in muscle growth, including modulating inflammatory responses, promoting tissue repair, and affecting muscle stem cell functions [ 62 – 64 ]. Macrophages are rapidly activated during muscle activity or injury [ 65 ], and their induced polarization states can stimulate satellite cell proliferation and differentiation, promote tissue angiogenesis, thereby accelerating muscle regeneration [ 64 , 66 ]. Notably, statins have been shown to possess specific regulatory abilities on myeloid cells, capable of targeting pathogenic myeloid cells with KRAS¹²ᴰ mutations by inhibiting the mevalonate pathway, blocking KRAS isoprenylation, and subsequently downregulating the PI3K-AKT pathway, achieving precise intervention without affecting normal myeloid or other immune cells [ 67 ]. Further single-cell differentiation trajectory analysis revealed the dynamic expression patterns of BACE1 and PDE4A during the differentiation process of myeloid cells: BACE1 significantly increased at the middle stage of differentiation (monocytes stage), while PDE4A gradually increased in the later stages of differentiation (monocytes, cDC2, cDC3 stages). This temporal and stage-specific expression pattern suggests that these two genes may play a synergistic regulatory role at different stages of myeloid cell differentiation. Combining the selective targeting ability of atorvastatin on myeloid cells, the differential expression of BACE1 and PDE4A may affect the differentiation and maturation process of monocytes into macrophages, interfere with the macrophage polarization state required for muscle repair, and thus play a key role in the pathogenesis of ATS-induced sarcopenia. This study identified the toxic key genes responsible for ATS-induced SA occurrence through transcriptome combined with network toxicology. A series of bioinformatics methods were used to explore the specific mechanisms by which these key genes influence the development of SA, and the expression of these genes in key cells was analyzed based on single-cell data to provide new theoretical support and evidence for the pathogenesis of SA caused by AST. However, there are some limitations in this study, as the specific molecular mechanisms of these key genes in SA have not yet been thoroughly investigated. Therefore, future research should focus on exploring the functions of these genes and their potential applications in treatment to further validate their clinical value. 5. Conclusions This study systematically elucidated the molecular mechanisms of atorvastatin-induced sarcopenia by integrating network toxicology, transcriptomics, and single-cell RNA sequencing. BACE1 and PDE4A were identified as key toxic genes, with their downregulation potentially contributing to sarcopenia pathogenesis through oxidative stress, immune dysregulation, and mitochondrial dysfunction. Myeloid cells were highlighted as critical cellular players, with dynamic expression patterns of these genes during cell differentiation. These findings provide novel insights into drug-induced sarcopenia and offer potential therapeutic targets for preventing atorvastatin-related myopathy. List of abbreviations Abbreviation Full Term SA Sarcopenia ATS Atorvastatin scRNA-seq single-cell RNA sequencing DEGs differentially expressed genes PDE4A cAMP-specific phosphodiesterase 4A BACE1 Beta-site amyloid precursor protein cleaving enzyme 1 LDL low-density lipoprotein FDA Food and Drug Administration PPI Protein-Protein Interaction GO Gene Ontology KEGG Kyoto Encyclopedia of Genes and Genomes LASSO Least Absolute Shrinkage and Selection Operator SVM-RFE Support Vector Machine-Recursive Feature Elimination GSEA Gene Set Enrichment Analysis NES Normalized Enrichment Score ssGSEA single-sample Gene Set Enrichment Analysis GSVA Gene Set Variation Analysis MDS Molecular Dynamics Simulation RMSD Root Mean Square Deviation RMSF Root Mean Square Fluctuation HVGs Highly Variable Genes PCA Principal Component Analysis TF Transcription Factor L-R Ligand-Receptor EC Endothelial Cells FAP Fibroblast Activation Protein MuSC Muscle Satellite Cells SMC Smooth Muscle Cells cDC2 type 2 conventional dendritic cells GPC-1 Glypican-1 eNOS endothelial nitric oxide synthase NADPH Nicotinamide Adenine Dinucleotide Phosphate cAMP cyclic adenosine monophosphate PKA cAMP-dependent protein kinase bHLH basic helix-loop-helix OXPHOS Oxidative Phosphorylation TEM CD4 cells Effector Memory CD4 T cells ROS Reactive Oxygen Species TFAM mitochondrial transcription factor A JNK c-Jun N-terminal Kinase HSCs Hematopoietic Stem Cells GEO Gene Expression Omnibus CTD Comparative Toxicogenomics Database STRING Search Tool for the Retrieval of Interacting Genes/Proteins Cytoscape Cytoscape Software MSigDB Molecular Signatures Database PDB Protein Data Bank CB-Dock2 CB-Dock2 Web Server GROMACS GROningen MAchine for Chemical Simulation AMBER Assisted Model Building with Energy Refinement GAFF General Amber Force Field NVT Canonical Ensemble (Constant Number of particles, Volume, Temperature) NPT Isothermal-Isobaric Ensemble (Constant Number of particles, Pressure, Temperature) Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Availability of data and materials The datasets analysed during the current study are available in the GEO database (GSE1428 and GSE38718, http://www.ncbi.nlm.nih.gov/geo/) and the CNGB Nucleotide Sequence Archive database (CNP0004394, CNP0004395, CNP0004494, and CNP0004495, https://db.cngb.org/cdcp/hlma/). Competing interests The authors declare that they have no competing interests. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Authors' contributions Wang Fang: Conceptualization, Data curation, Validation, Visualization, Writing–original draft, Writing–review & editing. Zhao Fei: Data curation, Validation, Visualization, Writing–review & editing. Zhao Yue: Data curation, Validation, Visualization, Writing–review & editing. Cai Liangling: Visualization, Writing–review & editing. Wei Yuanhuan: Supervision, Writing–review & editing. Acknowledgements We would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research. Special thanks to the following authors: Wang Fang, Zhao Fei, Zhao Yue, Cai Liangling, Wei Yuanhuan.In conclusion, we extend our thanks to everyone who has supported and assisted us along the way. Without your support, this research would not have been possible. References Li X, Wu C, Lu X, Wang L. Predictive models of sarcopenia based on inflammation and pyroptosis-related genes. Front Genet. 2024;15:1491577. Perez K, Ciotlos S, McGirr J, Limbad C, Doi R, Nederveen JP, Nilsson MI, Winer DA, Evans W, Tarnopolsky M, et al. Single nuclei profiling identifies cell specific markers of skeletal muscle aging, frailty, and senescence. Aging. 2022;14(23):9393–422. Huang S, Xiang C, Song Y. Identification of the shared gene signatures and pathways between sarcopenia and type 2 diabetes mellitus. PLoS ONE. 2022;17(3):e0265221. Zhang X, Zhu G, Zhang F, Yu D, Jia X, Ma B, Chen W, Cai X, Mao L, Zhuang C, et al. Identification of a novel immune-related transcriptional regulatory network in sarcopenia. BMC Geriatr. 2023;23(1):463. Damluji AA, Alfaraidhy M, AlHajri N, Rohant NN, Kumar M, Al Malouf C, Bahrainy S, Ji Kwak M, Batchelor WB, Forman DE, et al. Sarcopenia and Cardiovascular Diseases. Circulation. 2023;147(20):1534–53. Kuzuya M. Drug-related sarcopenia as a secondary sarcopenia. Geriatr Gerontol Int. 2024;24(2):195–203. Cho MR, Lee S, Song SK. A Review of Sarcopenia Pathophysiology, Diagnosis, Treatment and Future Direction. J Korean Med Sci. 2022;37(18):e146. Reig-López J, García-Arieta A, Mangas-Sanjuán V, Merino-Sanjuán M. Current Evidence, Challenges, and Opportunities of Physiologically Based Pharmacokinetic Models of Atorvastatin for Decision Making. Pharmaceutics 2021, 13(5). Reig-López J, Merino-Sanjuan M, García-Arieta A, Mangas-Sanjuán V. A physiologically based pharmacokinetic model for open acid and lactone forms of atorvastatin and metabolites to assess the drug-gene interaction with SLCO1B1 polymorphisms. Biomed pharmacotherapy = Biomedecine pharmacotherapie. 2022;156:113914. Ganga HV, Slim HB, Thompson PD. A systematic review of statin-induced muscle problems in clinical trials. Am Heart J. 2014;168(1):6–15. Zeng Z, Hu J, Xiao G, Liu Y, Jia D, Wu G, Xie C, Li S, Bi X. Integrating network toxicology and molecular docking to explore the toxicity of the environmental pollutant butyl hydroxyanisole: An example of induction of chronic urticaria. Heliyon. 2024;10(15):e35409. Zhou GL, Su SL, Yu L, Shang EX, Hua YQ, Yu H, Duan JA. Exploring the liver toxicity mechanism of Tripterygium wilfordii extract based on metabolomics, network pharmacological analysis and experimental validation. J Ethnopharmacol. 2025;337(Pt 2):118888. Wang Z, Ding H, Zou Q. Identifying cell types to interpret scRNA-seq data: how, why and more possibilities. Brief Funct Genomics. 2020;19(4):286–91. Cheng C, Chen W, Jin H, Chen X. A Review of Single-Cell RNA-Seq Annotation, Integration, and Cell-Cell Communication. Cells 2023, 12(15). Lai Y, Ramírez-Pardo I, Isern J, An J, Perdiguero E, Serrano AL, Li J, García-Domínguez E, Segalés J, Guo P, et al. Multimodal cell atlas of the ageing human skeletal muscle. Nature. 2024;629(8010):154–64. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. Gustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinf (Oxford England). 2022;38(15):3844–6. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinf (Oxford England). 2016;32(18):2847–9. Chen H, Boutros PC. VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R. BMC Bioinformatics. 2011;12:35. Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innov (Cambridge (Mass). 2021;2(3):100141. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw. 2010;33(1):1–22. Shi H, Yuan X, Liu G, Fan W. Identifying and Validating GSTM5 as an Immunogenic Gene in Diabetic Foot Ulcer Using Bioinformatics and Machine Learning. J Inflamm Res. 2023;16:6241–56. Robles-Jimenez LE, Aranda-Aguirre E, Castelan-Ortega OA, Shettino-Bermudez BS, Ortiz-Salinas R, Miranda M, Li X, Angeles-Hernandez JC, Vargas-Bello-Pérez E, Gonzalez-Ronquillo M. Worldwide Traceability of Antibiotic Residues from Livestock in Wastewater and Soil: A Systematic Review. Animals: open access J MDPI 2021, 12(1). Hänzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7. Liu ZY, Huang RH. Integrating single-cell RNA-sequencing and bulk RNA-sequencing data to explore the role of mitophagy-related genes in prostate cancer. Heliyon. 2024;10(9):e30766. Satija R, Farrell JA, Gennert D, Schier AF, Regev A. Spatial reconstruction of single-cell gene expression data. Nat Biotechnol. 2015;33(5):495–502. Griss J, Viteri G, Sidiropoulos K, Nguyen V, Fabregat A, Hermjakob H. ReactomeGSA - Efficient Multi-Omics Comparative Pathway Analysis. Mol Cell Proteom. 2020;19(12):2115–25. Cornwell M, Vangala M, Taing L, Herbert Z, Köster J, Li B, Sun H, Li T, Zhang J, Qiu X et al. VIPER: Visualization Pipeline for RNA-seq, a Snakemake workflow for efficient and complete RNA-seq analysis. BMC Bioinformatics 2018, 19(1):135. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088. Smalley I, Chen Z, Phadke M, Li J, Yu X, Wyatt C, Evernden B, Messina JL, Sarnaik A, Sondak VK, et al. Single-Cell Characterization of the Immune Microenvironment of Melanoma Brain and Leptomeningeal Metastases. Clin cancer research: official J Am Association Cancer Res. 2021;27(14):4109–25. Hong B, Li Y, Yang R, Dai S, Zhan Y, Zhang WB, Dong R. Single-cell transcriptional profiling reveals heterogeneity and developmental trajectories of Ewing sarcoma. J Cancer Res Clin Oncol. 2022;148(12):3267–80. He T, d'Uscio LV, Sun R, Santhanam AVR, Katusic ZS. Inactivation of BACE1 increases expression of endothelial nitric oxide synthase in cerebrovascular endothelium. J Cereb blood flow metabolism: official J Int Soc Cereb Blood Flow Metabolism. 2022;42(10):1920–32. Katusic ZS. Vascular endothelial dysfunction: does tetrahydrobiopterin play a role? Am J Physiol Heart Circ Physiol. 2001;281(3):H981–986. Howard C, Ferrucci L, Sun K, Fried LP, Walston J, Varadhan R, Guralnik JM, Semba RD. Oxidative protein damage is associated with poor grip strength among older women living in the community. Journal of applied physiology (Bethesda, Md : 1985) 2007, 103(1):17–20. Siu PM, Pistilli EE, Alway SE. Age-dependent increase in oxidative stress in gastrocnemius muscle with unloading. J Appl Physiol (Bethesda Md: 1985). 2008;105(6):1695–705. Koopman R, van Loon LJ. Aging, exercise, and muscle protein metabolism. J Appl Physiol (Bethesda Md: 1985). 2009;106(6):2040–8. Meng SJ, Yu LJ. Oxidative stress, molecular inflammation and sarcopenia. Int J Mol Sci. 2010;11(4):1509–26. Francis SH, Blount MA, Corbin JD. Mammalian cyclic nucleotide phosphodiesterases: molecular mechanisms and physiological functions. Physiol Rev. 2011;91(2):651–90. Obrietan K, van den Pol AN. GABA activity mediating cytosolic Ca2 + rises in developing neurons is modulated by cAMP-dependent signal transduction. J neuroscience: official J Soc Neurosci. 1997;17(12):4785–99. Arnold WD, Clark BC. Neuromuscular junction transmission failure in aging and sarcopenia: The nexus of the neurological and muscular systems. Ageing Res Rev. 2023;89:101966. Bao Z, Cui C, Chow SK, Qin L, Wong RMY, Cheung WH. AChRs Degeneration at NMJ in Aging-Associated Sarcopenia-A Systematic Review. Front Aging Neurosci. 2020;12:597811. Winter B, Braun T, Arnold HH. cAMP-dependent protein kinase represses myogenic differentiation and the activity of the muscle-specific helix-loop-helix transcription factors Myf-5 and MyoD. J Biol Chem. 1993;268(13):9869–78. Yang YF, Yang W, Liao ZY, Wu YX, Fan Z, Guo A, Yu J, Chen QN, Wu JH, Zhou J, et al. MICU3 regulates mitochondrial Ca(2+)-dependent antioxidant response in skeletal muscle aging. Cell Death Dis. 2021;12(12):1115. Xu H, Ranjit R, Richardson A, Van Remmen H. Muscle mitochondrial catalase expression prevents neuromuscular junction disruption, atrophy, and weakness in a mouse model of accelerated sarcopenia. J cachexia sarcopenia muscle. 2021;12(6):1582–96. González-Hedström D, Priego T, Amor S, de la Fuente-Fernández M, Martín AI, López-Calderón A, Inarejos-García AM, García-Villalón ÁL, Granado M. Olive Leaf Extract Supplementation to Old Wistar Rats Attenuates Aging-Induced Sarcopenia and Increases Insulin Sensitivity in Adipose Tissue and Skeletal Muscle. Antioxid (Basel Switzerland) 2021, 10(5). Karam C, Yi J, Xiao Y, Dhakal K, Zhang L, Li X, Manno C, Xu J, Li K, Cheng H, et al. Absence of physiological Ca(2+) transients is an initial trigger for mitochondrial dysfunction in skeletal muscle following denervation. Skelet Muscle. 2017;7(1):6. Covarrubias AJ, Perrone R, Grozio A, Verdin E. NAD(+) metabolism and its roles in cellular processes during ageing. Nat Rev Mol Cell Biol. 2021;22(2):119–41. Migliavacca E, Tay SKH, Patel HP, Sonntag T, Civiletto G, McFarlane C, Forrester T, Barton SJ, Leow MK, Antoun E, et al. Mitochondrial oxidative capacity and NAD(+) biosynthesis are reduced in human sarcopenia across ethnicities. Nat Commun. 2019;10(1):5808. Frederick DW, Loro E, Liu L, Davila A Jr., Chellappa K, Silverman IM, Quinn WJ 3rd, Gosai SJ, Tichy ED, Davis JG, et al. Loss of NAD Homeostasis Leads to Progressive and Reversible Degeneration of Skeletal Muscle. Cell Metabol. 2016;24(2):269–82. Ciurea AV, Mohan AG, Covache-Busuioc RA, Costin HP, Glavan LA, Corlatescu AD, Saceleanu VM. Unraveling Molecular and Genetic Insights into Neurodegenerative Diseases: Advances in Understanding Alzheimer's, Parkinson's, and Huntington's Diseases and Amyotrophic Lateral Sclerosis. Int J Mol Sci 2023, 24(13). In: Parkinson’s Disease: Pathogenesis and Clinical Aspects. Edited by Stoker TB, Greenland JC. Brisbane (AU): Codon Publications Copyright © 2018 Codon Publications.; 2018. Kedlian VR, Wang Y, Liu T, Chen X, Bolt L, Tudor C, Shen Z, Fasouli ES, Prigmore E, Kleshchevnikov V, et al. Human skeletal muscle aging atlas. Nat aging. 2024;4(5):727–44. Tan H, Yang K, Li Y, Shaw TI, Wang Y, Blanco DB, Wang X, Cho JH, Wang H, Rankin S, et al. Integrative Proteomics and Phosphoproteomics Profiling Reveals Dynamic Signaling Networks and Bioenergetics Pathways Underlying T Cell Activation. Immunity. 2017;46(3):488–503. Seder RA, Ahmed R. Similarities and differences in CD4 + and CD8 + effector and memory T cell generation. Nat Immunol. 2003;4(9):835–42. Tan S, Li S, Min Y, Gisterå A, Moruzzi N, Zhang J, Sun Y, Andersson J, Malmström RE, Wang M, et al. Platelet factor 4 enhances CD4(+) T effector memory cell responses via Akt-PGC1α-TFAM signaling-mediated mitochondrial biogenesis. J Thromb haemostasis: JTH. 2020;18(10):2685–700. Dimeloe S, Mehling M, Frick C, Loeliger J, Bantug GR, Sauder U, Fischer M, Belle R, Develioglu L, Tay S, et al. The Immune-Metabolic Basis of Effector Memory CD4 + T Cell Function under Hypoxic Conditions. J Immunol (Baltimore Md: 1950). 2016;196(1):106–14. Chen Y, Ye Y, Krauß PL, Löwe P, Pfeiffenberger M, Damerau A, Ehlers L, Buttgereit T, Hoff P, Buttgereit F, et al. Age-related increase of mitochondrial content in human memory CD4 + T cells contributes to ROS-mediated increased expression of proinflammatory cytokines. Front Immunol. 2022;13:911050. Hayashi T, Kato N, Furudoi K, Hayashi I, Kyoizumi S, Yoshida K, Kusunoki Y, Furukawa K, Imaizumi M, Hida A, et al. Early-life atomic-bomb irradiation accelerates immunological aging and elevates immune-related intracellular reactive oxygen species. Aging Cell. 2023;22(10):e13940. Dorshkind K, Höfer T, Montecino-Rodriguez E, Pioli PD, Rodewald HR. Do haematopoietic stem cells age? Nat Rev Immunol. 2020;20(3):196–202. Della Bella S, Bierti L, Presicce P, Arienti R, Valenti M, Saresella M, Vergani C, Villa ML. Peripheral blood dendritic cells and monocytes are differently regulated in the elderly. Clin Immunol (Orlando Fla). 2007;122(2):220–8. Hernandez-Torres F, Matias-Valiente L, Alzas-Gomez V, Aranega AE. Macrophages in the Context of Muscle Regeneration and Duchenne Muscular Dystrophy. International journal of molecular sciences 2024, 25(19). Minari ALA, Thomatieli-Santos RV. From skeletal muscle damage and regeneration to the hypertrophy induced by exercise: what is the role of different macrophage subsets? Am J Physiol Regul Integr Comp Physiol. 2022;322(1):R41–54. Saclier M, Cuvellier S, Magnan M, Mounier R, Chazaud B. Monocyte/macrophage interactions with myogenic precursor cells during skeletal muscle regeneration. FEBS J. 2013;280(17):4118–30. Tidball JG. Regulation of muscle growth and regeneration by the immune system. Nat Rev Immunol. 2017;17(3):165–78. Shen J, Zhu X, Liu H. MiR-483 induces senescence of human adipose-derived mesenchymal stem cells through IGF1 inhibition. Aging. 2020;12(15):15756–70. Kamata T, Giblett S, Pritchard C. KRAS(G12D) expression in lung-resident myeloid cells promotes pulmonary LCH-like neoplasm sensitive to statin treatment. Blood. 2017;130(4):514–26. Table 1 Table 1 The binding energy of molecular docking. Key genes drug Total Score BACE1 ATS -9.3 Kcal/mol PDE4A ATS -8.9 Kcal/mol Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.csv Additionalfile5.csv Additionalfile4.csv Additionalfile2.csv Additionalfile3.csv FigureS1.tif Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9058922","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":602745099,"identity":"419e4235-f0b5-4b74-b990-dd4cfcf87d29","order_by":0,"name":"Fei Zhao","email":"","orcid":"","institution":"Shenzhen Nanshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Fei","middleName":"","lastName":"Zhao","suffix":""},{"id":602745100,"identity":"d0f611e8-5134-4c62-a46a-1e0bc50505c0","order_by":1,"name":"Yue Zhao","email":"","orcid":"","institution":"Shenzhen Nanshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Zhao","suffix":""},{"id":602745102,"identity":"8bba1740-666e-4428-820e-9b8bbc160137","order_by":2,"name":"Liangling Cai","email":"","orcid":"","institution":"Shenzhen Nanshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Liangling","middleName":"","lastName":"Cai","suffix":""},{"id":602745103,"identity":"942b1f5f-6bf9-4938-9b54-d0565364d43f","order_by":3,"name":"Yuanhuan Wei","email":"","orcid":"","institution":"Shenzhen Nanshan Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yuanhuan","middleName":"","lastName":"Wei","suffix":""},{"id":602745105,"identity":"6beca33b-e766-494e-ae52-2022dfb8576b","order_by":4,"name":"Fang Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxElEQVRIiWNgGAWjYBACPgYGgwMMPDU8/MzMBx8QpYUNrEXmmJxkO1uyAdFaGBhsmI0NzvOYCRCnRSJ544EfOWyJmw8zmDEw1NhEE6ElreBgzxmZxG2HGdIeMBxLy20grCXH4ABvDxtIy3EDxobDxGk5+Pcfc+LmZsY2CaK1HObhAXqfmZmNSC08zwoOy/Ack5M4zMZskECMX/jZkzd/fAOKyv7zHx98qLEhrIVBIAGJk4BDEZo1B4hSNgpGwSgYBSMZAABFTTxtrl8YmQAAAABJRU5ErkJggg==","orcid":"","institution":"Shenzhen Nanshan Hospital","correspondingAuthor":true,"prefix":"","firstName":"Fang","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2026-03-07 13:54:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9058922/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9058922/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104288142,"identity":"2f14400b-a8b2-401a-ac94-f124d4655661","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":26548887,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of candidate genes and network construction.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) Molecular structure of atorvastatin (ATS). (b) Toxicity prediction radar plot of ATS. (c) Volcano plot of differentially expressed genes (DEGs) in sarcopenia (SA). (d) Heatmap of the top 10 upregulated and downregulated DEGs. (e) Venn diagram showing the 16 candidate genes from the intersection of DEGs, ATS toxicity targets, and SA-related targets. (f) Gene Ontology (GO) enrichment analysis of the 16 candidate genes. (g) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the candidate genes. (h) Protein-protein interaction (PPI) network of the candidate genes. (i) Compound-target-pathway-disease network illustrating the relationships between ATS, candidate genes, KEGG pathways, and SA.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/9ab29a8d151781fb4b5468fb.png"},{"id":104288140,"identity":"4059f75f-fb6b-4655-aa71-3b34c6dcd893","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":11005563,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification and validation of key genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) Least absolute shrinkage and selection operator (LASSO) regression analysis of candidate genes. (c) Support vector machine-recursive feature elimination (SVM-RFE) algorithm to identify optimal feature genes. (d) Venn diagram showing the intersection of genes from LASSO and SVM-RFE algorithms, identifying five candidate key genes. (e) Expression validation of candidate key genes in the GSE1428 training set. (f) Expression validation of candidate key genes in the GSE38718 validation set. BACE1 and PDE4A were significantly downregulated in SA samples in both datasets and were selected as key genes. *p \u0026lt; 0.05, **p \u0026lt; 0.01, ***p \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/668315bd8a285f648ab50b2f.png"},{"id":104405382,"identity":"3870aaa6-5289-4e56-bf44-e64517010603","added_by":"auto","created_at":"2026-03-11 12:22:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":15036523,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFunctional enrichment and immune infiltration analysis of key genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) Gene set enrichment analysis (GSEA) plots showing pathways co-enriched with BACE1 and PDE4A, including Parkinson's disease, oxidative phosphorylation, Alzheimer's disease, and Huntington's disease. (c) Heatmap of the infiltration levels of 28 immune cell types in SA and control samples. (d) Boxplot displaying the four immune cell types with significantly different infiltration scores between SA and control groups. (e) Correlation analysis between the expression of key genes (BACE1, PDE4A) and the abundance of significantly different immune cells, showing a positive correlation with effector memory CD4 T cells. *p \u0026lt; 0.05, **p \u0026lt; 0.01.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/ad4f750e42e948516c71cfe7.png"},{"id":104288146,"identity":"bf509394-097f-4fae-9479-5795efb09282","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":25980176,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMolecular docking and dynamics simulations of ATS with key genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) Molecular docking models showing the binding conformation and hydrogen bond interactions of ATS with BACE1 and PDE4A. (c) Root mean square deviation (RMSD) analysis of the BACE1-ATS and PDE4A-ATS complexes over 100 ns of molecular dynamics simulation, indicating stable binding. (d) Root mean square fluctuation (RMSF) analysis of protein residues in the complexes. (e) Number of hydrogen bonds formed between the protein and ligand throughout the simulation. (f) Energy analysis of the protein-ligand complexes during the simulation, showing minimal fluctuation and a steady state.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/7b99345d0afd3206de47e639.png"},{"id":104288144,"identity":"008336bf-9a2f-4a95-a74f-35eb2f29ae1b","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":10941791,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of key cells via single-cell RNA sequencing analysis.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) t-distributed stochastic neighbor embedding (t-SNE) plot showing the distribution of key genes BACE1 and PDE4A across nine annotated cell types in skeletal muscle. (b) Violin plot comparing BACE1 expression between SA and control groups in each cell type. Significant differences were found in FAP, MuSC, and myeloid cells. (c) Violin plot comparing PDE4A expression between SA and control groups in each cell type, with a significant difference observed only in myeloid cells. (d) Heatmap showing the differential activity of transcription factors (TFs) in myeloid cells between SA and control groups. *p \u0026lt; 0.05, **p \u0026lt; 0.01, ***p \u0026lt; 0.001. EC: endothelial cells; FAP: fibroblast activation protein; MuSC: muscle satellite cells; SMC: smooth muscle cells.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/f57552e403448a436f1ed666.png"},{"id":104404882,"identity":"9fa5f0cd-319a-482b-9812-ad2c8e8a6777","added_by":"auto","created_at":"2026-03-11 12:21:16","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":14597282,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCell-cell communication analysis.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a-b) Circle plots illustrating the number and strength of intercellular interactions among the nine annotated cell types in control and SA samples. (c-d) Differential interaction changes showing that myeloid cells have altered communication patterns with other cells in SA compared to control. (e-f) Ligand-receptor pair analysis highlighting the specific signaling axes, such as LGALS9-CD45, involved in myeloid cell interactions with other cell types in control and SA groups.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/eeb5a72997e4a4c93e54871a.png"},{"id":104405085,"identity":"c302902b-990a-4f7f-93d7-db051d61bfbf","added_by":"auto","created_at":"2026-03-11 12:21:43","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":15644054,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePseudotime analysis of myeloid cell differentiation.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(a) t-SNE plot showing the re-clustering of myeloid cells into seven subclusters. (b) Heatmap of marker gene expression used to annotate the myeloid subclusters. (c) t-SNE plot of the four annotated myeloid cell subclusters: macrophages, monocytes, cDC2, and cDC3. (d) Pseudotime trajectory of myeloid cell differentiation, with cells colored by pseudotime (dark blue: early; light blue: late). (e) Expression dynamics of BACE1 and PDE4A along the pseudotime trajectory, showing distinct expression patterns during differentiation. cDC2: type 2 conventional dendritic cells; cDC3: type 3 conventional dendritic cells.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/ed3d9dcc06ad86084809fd91.png"},{"id":104316533,"identity":"2c5357eb-30bc-4847-ba43-a59c19a14b11","added_by":"auto","created_at":"2026-03-10 12:13:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":54509008,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/a37d2574-d425-4bf8-8545-4abc7b5fbecb.pdf"},{"id":104288137,"identity":"9f9e3909-8ef3-4cf8-83b8-39fa9e4c8314","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"csv","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":704,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile1.csv","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/af1ae8a05665e21c4a97ce74.csv"},{"id":104779684,"identity":"b7c81dfd-71fe-46cd-acd5-07f438ae8ea6","added_by":"auto","created_at":"2026-03-17 07:44:31","extension":"csv","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":7280,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile5.csv","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/f55ff8ee34bac6332a47dd8f.csv"},{"id":104779842,"identity":"b7d43683-0bc4-414e-9043-03c1291e1402","added_by":"auto","created_at":"2026-03-17 07:46:40","extension":"csv","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":13915,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile4.csv","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/9f2201af89a767e5a42f7bb0.csv"},{"id":104288148,"identity":"549029a5-2f66-44ba-b057-f0dcad2129c8","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"csv","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":58140,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile2.csv","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/84446127afdb7c95ccdddf18.csv"},{"id":104288139,"identity":"514b6f80-d65a-41a5-8531-4e6a15294a51","added_by":"auto","created_at":"2026-03-10 06:03:45","extension":"csv","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":86536,"visible":true,"origin":"","legend":"","description":"","filename":"Additionalfile3.csv","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/5f657c2e00394c5c1dae3335.csv"},{"id":104405256,"identity":"9211fc5b-9493-4325-87f1-f0f6952eead3","added_by":"auto","created_at":"2026-03-11 12:22:17","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":8385000,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.tif","url":"https://assets-eu.researchsquare.com/files/rs-9058922/v1/4dff2de41b6e3cabc014d869.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"The mechanism of atorvastatin-induced sarcopenia elucidated based on network toxicology, single-cell RNA sequencing, and bulk transcriptomics data","fulltext":[{"header":"1. Background","content":"\u003cp\u003eSarcopenia (SA) is a degenerative disease characterized by muscle strength decline, decreased muscle mass, muscle fiber atrophy, and overall impairment of physical function [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Aging is the primary risk factor, with people over 50 losing approximately 1% of muscle mass annually [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. SA prevalence reaches 14%-33% in individuals aged 65 and above, escalating to 50%-60% in those over 80, representing a major public health concern [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. This condition not only causes muscle weakness and functional decline but also significantly increases the risks of falls, fractures, and mortality [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. SA pathogenesis involves multifactorial interactions, including aging, neuromuscular dysfunction, hormonal changes, increased pro-inflammatory cytokines, chronic physical inactivity, and nutritional imbalances, all of which accelerate disease progression [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Secondary sarcopenia caused by other diseases or medications is often more preventable than primary forms [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Current treatment strategies encompass non-pharmacological approaches such as resistance exercise and adequate intake of protein, vitamin D, antioxidants, and long-chain polyunsaturated fatty acids. Regarding pharmacological interventions, although drugs like growth hormone, selective androgen receptor modulators, and myostatin inhibitors have demonstrated some efficacy in clinical trials, the FDA has not approved any specific therapeutic agents to date [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. However, drug-related sarcopenia remains insufficiently investigated, emphasizing the importance of exploring SA's multiple causes and interactive mechanisms for developing effective prevention and treatment strategies.\u003c/p\u003e \u003cp\u003eAtorvastatin (ATS) is a widely prescribed lipid-lowering medication that selectively inhibits 3-hydroxy-3-methylglutaryl coenzyme A reductase, a key enzyme in hepatic cholesterol biosynthesis, thereby reducing plasma cholesterol, low-density lipoprotein (LDL), and triglyceride levels [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. As a representative statin drug extensively used in elderly patients, ATS carries significant clinical relevance for drug-related sarcopenia, which is particularly prominent in elderly patients with polypharmacy [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe most common adverse effect of ATS is drug-induced myopathy, presenting with diverse clinical manifestations including muscle weakness, myalgia, muscle tenderness, spasms, and arthralgia, with or without elevated serum creatine kinase levels. In rare cases, patients may develop rhabdomyolysis, a serious and potentially life-threatening condition [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Clinical studies indicate that approximately 10%-25% of statin-treated patients experience muscle-related problems [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. However, the specific molecular mechanisms by which ATS induces sarcopenia development remain insufficiently understood, with no systematic studies elucidating these pathways, representing an important area requiring further investigation.\u003c/p\u003e \u003cp\u003eNetwork toxicology is an emerging discipline developed based on network pharmacology and network biology. By integrating bioinformatics, big data analysis, and multi-omics techniques, it constructs complex network relationships among chemicals, toxic effects, and targets. This approach enables systematic analysis of the mechanisms of multi-component, multi-target toxicants, predicts molecular pathways involved in disease induction, and demonstrates rapid, convenient, and comprehensive advantages in toxicity mechanism research [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Selecting the network toxicology method helps to predict key targets and pathways related to ATS-induced sarcopenia from a systemic perspective, providing a theoretical basis for subsequent experimental validation. Single-cell RNA sequencing (scRNA-seq) technology investigates the transcriptomic features of individual cells, revealing gene expression differences among cells and overcoming the limitations of traditional RNA sequencing that obscure cell heterogeneity. This technology has unique advantages in identifying cell types, cell subpopulations, and state transitions, offering an important tool for understanding cellular heterogeneity and communication networks [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Using scRNA-seq allows validation of key genes predicted by network toxicology at the single-cell level and reveals the specific effects of ATS on different muscle cell subpopulations. Integrating network toxicology with scRNA-seq enables a complete research chain from macro-level predictions to micro-level validations, facilitating systematic prediction of the molecular mechanisms of ATS-induced sarcopenia and verifying key targets at the single-cell level, thereby providing a comprehensive and precise research strategy for in-depth understanding of drug-induced sarcopenia pathogenesis.\u003c/p\u003e \u003cp\u003eIn this study, we identify the toxic key genes of ATS leading to SA pathogenesis by transcriptome combined with network toxicology, use a series of bioinformatics methods to explore the specific mechanism of key genes in the development of SA, and then explore the expression of key genes in key cells based on single-cell data, providing new theoretical support and basis for the pathogenesis of SA.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data collection\u003c/h2\u003e \u003cp\u003eThe data were obtained from the GEO database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncbi.nlm.nih.gov/geo/\u003c/span\u003e\u003cspan address=\"https://www.ncbi.nlm.nih.gov/geo/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The GSE1428 dataset (Sequencing platform: GPL96) was employed as a training set, which included 12 samples of skeletal muscle tissue from SA patients and 10 control skeletal muscle tissue samples. The GSE38718 dataset (Sequencing platform: GPL570) served as a validation set, which comprised 8 samples of skeletal muscle tissue from SA patients and 14 control skeletal muscle tissue samples. In addition, the scRNA-seq data were obtained from the CNGB Nucleotide Sequence Archive (CNP0004394, CNP0004395, CNP0004494, and CNP0004495) [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]). All processed data were available at the Human Muscle Aging Cell Atlas database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://db.cngb.org/cdcp/hlma/\u003c/span\u003e\u003cspan address=\"https://db.cngb.org/cdcp/hlma/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The scRNA-seq data samples primarily consisted of 4 skeletal muscle tissue samples from SA and 3 control skeletal muscle tissue samples.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Differential expression analysis and collection of SA-related targets\u003c/h2\u003e \u003cp\u003eTo identify DEGs between the SA and control groups, DEGs were carried out by employing the \u0026ldquo;limma\u0026rdquo; (v 3.54.0) [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] based on a training set for both groups of samples (|log\u003csub\u003e2\u003c/sub\u003e FC| \u0026gt; 0.5, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Moreover, the \u0026ldquo;ggplot2\u0026rdquo; (v 3.4.1) [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] was employed to construct a volcano plot of DEGs, and the top 10 genes (sorted by |log\u003csub\u003e2\u003c/sub\u003e FC| from high to low) with significant up- and down-regulation differences were labeled, and then the heatmap of these genes was drawn using the \u0026ldquo;ComplexHeatmap\u0026rdquo; (v 2.14.0) [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. To identify the targets associated with SA, the CTD database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ctdbase.org/\u003c/span\u003e\u003cspan address=\"https://ctdbase.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was searched using the keyword \u0026ldquo;sarcopenia\u0026rdquo;, which was documented as the target for SA.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Prediction of ATS toxicity and identification of toxicity targets\u003c/h2\u003e \u003cp\u003eIn order to understand the toxicity information of ATS, the SMILES molecular structure of ATS compounds was first obtained in the PubChem database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubchem.ncbi.nlm.nih.gov/\u003c/span\u003e\u003cspan address=\"https://pubchem.ncbi.nlm.nih.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Based on this SMILES, the toxicity of ATS was predicted using the ProTox 3.0 database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://tox.charite.de\u003c/span\u003e\u003cspan address=\"https://tox.charite.de\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Finally, a search was conducted in the Swiss Target Database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://swisstargetprediction.ch/\u003c/span\u003e\u003cspan address=\"http://swisstargetprediction.ch/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) using the SMILES notation of ATS. To ensure data accuracy, the collected information was consolidated, duplicates were removed, and the remaining entries were classified as toxic targets of ATS.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Identification, functional enrichment, and PPI network of candidate genes\u003c/h2\u003e \u003cp\u003eTo identify candidate genes associated with toxicity in ATS that contributed to the development of SA, the \u0026ldquo;ggvenn\u0026rdquo; (v 1.7.3) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] was employed to determine the intersection of DEGs, toxic targets, and SA-related targets. These genes were subsequently documented as candidate genes for further analysis. Subsequently, to explore the biological pathways of candidate genes and their functional enrichment, candidate genes were analyzed by GO and KEGG using \u0026ldquo;clusterProfiler\u0026rdquo; (v 4.2.2) [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Moreover, the candidate genes were uploaded into the STRING database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://string-db.org/\u003c/span\u003e\u003cspan address=\"https://string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) with a confidence score\u0026thinsp;\u0026gt;\u0026thinsp;0.4 by \u0026ldquo;Cytoscape\u0026rdquo; (v 3.9.1) [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Next, candidate genes with interactions in the PPI network were selected for further study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Network construction of compound-target-pathway-disease\u003c/h2\u003e \u003cp\u003eIn order to understand the connections between the candidate genes, compounds, pathways, and diseases, we integrated the candidate genes, KEGG-enriched pathways (specifically, the top 10 pathways with the highest number of associated genes), and SA diseases. We utilized \u0026ldquo;Cytoscape\u0026rdquo; (v 3.9.1) to visualize the compound-target-pathway-disease network diagram, illustrating the complex regulatory network involved in the development of ATS-induced SA.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Machine learning and expression validation\u003c/h2\u003e \u003cp\u003eIn order to further refine the selection of candidate genes, we performed LASSO and SVM-RFE analyses on the training set. The \u0026ldquo;glmnet\u0026rdquo; (v 4.1-4) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] was employed to perform LASSO regression of candidate genes. The optimal lambda value was identified using 5-fold cross-validation to screen for LASSO genes. A SVM-RFE algorithm of candidate genes was conducted using the \u0026ldquo;e1071\u0026rdquo; (v 1.7\u0026ndash;13) [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], which was employed to screen for the most optimal SVM-RFE genes based on 5-fold cross-validation. Furthermore, the \u0026ldquo;ggvenn\u0026rdquo; (v 1.7.3) was applied to obtain the intersection genes between 2 algorithm genes, which were defined as candidate key genes.\u003c/p\u003e \u003cp\u003eAn investigation was conducted to compare the expression of candidate key genes in the SA and control groups, utilizing the GSE1428 and GSE38718 datasets. The Wilcoxon test was employed to analyze the datasets. Genes exhibiting consistent expression trends and significant inter-group differences (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in both datasets were identified as key genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 GSEA\u003c/h2\u003e \u003cp\u003eThe GSEA was carried out with the objective of elucidating the biological functions of key genes throughout the progression of SA. The initial step in the research was to calculate the Spearman correlation coefficients between the various key genes and every single gene across the entire range of samples from the training set, with the \u0026ldquo;psych\u0026rdquo; (v 2.2.9) [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e] being utilized for the purpose. Following that, the sorted genes were then arranged in descending order according to their respective Spearman correlation coefficients, with this ranking then being utilized as the gene set that was tested in further analyses. Meanwhile, \u0026ldquo;c2.cp.kegg_legacy.v2024.1.Hs.symbols\u0026rdquo; was retrieved from the MSigDB (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.gsea-msigdb.org/gsea/msigdb\u003c/span\u003e\u003cspan address=\"https://www.gsea-msigdb.org/gsea/msigdb\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Subsequently, the GSEA was performed employing the \u0026ldquo;clusterProfiler\u0026rdquo; (v 4.2.2), with a threshold of |NES| \u0026gt; 1 and adjust P\u0026thinsp;\u0026lt;\u0026thinsp;0.05.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Immune infiltration analysis\u003c/h2\u003e \u003cp\u003eAccording to the ssGSEA algorithm, the GSVA package (v 1.46.0) [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] was utilized to determine the infiltration scores of 28 kinds of immune cells [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] between SA and control in the training set. Next, the Wilcoxon test was applied to make a comparison of the differences in immune-infiltrating cells between the 2 groups (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Subsequently, the \u0026ldquo;psych\u0026rdquo; (v 2.2.9) was used to analyze the correlations between the key genes and the significantly different immune cells (|correlation (r)| \u0026gt; 0.3, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.9 Molecular docking\u003c/h2\u003e \u003cp\u003eTo further assess the binding capacity of ATS with key genes and to enhance the accuracy of the interaction network between ATS and these genes, molecular docking was conducted for each key gene in conjunction with ATS. The structure file of the ATS was sourced from the PubChem database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubchem.ncbi.nlm.nih.gov/\u003c/span\u003e\u003cspan address=\"https://pubchem.ncbi.nlm.nih.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), and the protein data for the key genes were extracted from the PDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.rcsb.org\u003c/span\u003e\u003cspan address=\"http://www.rcsb.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Subsequently, molecular docking was performed using the CB-Dock2 website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cadd.labshare.cn/cb-dock2/php/index.php\u003c/span\u003e\u003cspan address=\"https://cadd.labshare.cn/cb-dock2/php/index.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The binding activity was deemed to be satisfactory if the affinity was lower than \u0026minus;\u0026thinsp;5.0 kcal/mol.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.10 Molecular dynamics simulation (MDS)\u003c/h2\u003e \u003cp\u003eMDS lasting 100 nanoseconds were conducted using the \u0026ldquo;GROMACS\u0026rdquo; (v 2024.2) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://manual.gromacs.org\u003c/span\u003e\u003cspan address=\"https://manual.gromacs.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to evaluate the binding stability of the receptor-ligand complex. Parameter and topology files for proteins and small molecule ligands were generated using the AMBER14SB force field and the AMBER GAFF force field, respectively. Periodic boundary conditions were implemented, and the system was neutralized with NaCl at a concentration of 0.15 mol/L. To minimize the energy of the system, the steepest descent method was employed. The pre-balancing procedure was carried out in 2 stages. In the initial stage, a simulation was conducted using an NVT ensemble at 300 K (temperature) for 100 ps to stabilize the system's temperature. In the subsequent stage, a simulation was performed using an NPT ensemble at 1 bar (pressure) for 100 ps to stabilize the system's pressure. The analysis of the simulation included the root mean square deviation (RMSD), root mean square fluctuation (RMSF), hydrogen bonding, and energy analyses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e2.11 Treatment of scRNA-seq data\u003c/h2\u003e \u003cp\u003eIn the scRNA-seq dataset, the \u0026ldquo;Seurat\u0026rdquo; (v 5.0.1) [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] was utilized for quality control (quality control criteria: 1,000\u0026thinsp;\u0026lt;\u0026thinsp;nFeature_RNA\u0026thinsp;\u0026lt;\u0026thinsp;4,000, 3\u0026thinsp;\u0026lt;\u0026thinsp;nCount_RNA\u0026thinsp;\u0026lt;\u0026thinsp;15,000, percent.mt\u0026thinsp;\u0026lt;\u0026thinsp;5%). Then the FindVariableFeatures function was used to extract the top 2,000 HVGs for subsequent analysis. The data were scaled using the ScaleData function. Statistically significant principal components were identified using the JackStrawPlot function. PCA was conducted on the top 2,000 HVGs. Subsequently, from the results obtained via the JackStraw function, the top 50 principal components (PCs) were chosen (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Subsequently, the resolution parameter was configured to 0.1, and an unsupervised clustering analysis was carried out on every cell within the dataset by utilizing the FindClusters function and FindNeighbors function. Then, the outcomes of the clustering were visualized through the application of the RunTSNE function. Cell types of the different clusters were identified after dimensionality reduction and clustering, employing the reference [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The annotation results were visualized on a t-SNE plot, and the expression of marker genes across different cell types was depicted. The utilization of histograms was essential for visualizing cell cluster types and their corresponding percentages within both SA and control samples. To gain insights into the biological functions of the annotated cells involved, the functional enrichment of annotated cells was explored through the pathways function in the \u0026ldquo;ReactomeGSA\u0026rdquo; (v 1.12.0) [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e2.12 Selection of key cells\u003c/h2\u003e \u003cp\u003eTo further identify key cells within the scRNA-seq dataset, t-SNE maps were generated using the \u0026ldquo;Seurat\u0026rdquo; (v 5.0.1) to illustrate the distribution of key genes across the annotated cells. Subsequently, the Wilcoxon test was employed to compare the expression differences of key genes between the SA and control groups across the annotated cells (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Cells exhibiting significant differences were designated as key cells.\u003c/p\u003e \u003cp\u003eThe DoRothEA algorithm is a crucial tool for analyzing cellular gene regulation, primarily focusing on the interactions between transcription factors (TFs) and their target genes. Utilizing the \u0026ldquo;viper\u0026rdquo; package (v 1.34.0) [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e], all TFs in key cells were identified, and the differences in activity between the SA and control groups were compared.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e2.13 Cell communication and pseudotime analyses\u003c/h2\u003e \u003cp\u003eIn order to characterize the interactions among annotated cells, as well as between key cells and other cells in the scRNA-seq dataset, the \u0026ldquo;CellChat\u0026rdquo; (v 1.6.1) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e] was employed to analyze cell communication. This analysis illustrated the network of interactions among various cell types. Subsequently, \u0026ldquo;CellChat\u0026rdquo; (v 1.6.1) was utilized to identify ligand-receptor (L-R) interactions between key cells and other cells. Meanwhile, to identify the differentiation process of key cells and the expression of prognostic genes at different differentiation stages, the RunTSNE function was used to perform dimensionality reduction and clustering on the key cells (resolution\u0026thinsp;=\u0026thinsp;0.1). Subsequently, the cells were re-clustered into subclusters based on the marker genes identified in the existing literature [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Then, the \u0026ldquo;Monocle\u0026rdquo; (v 2.22.0) [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e] was used for pseudotime analysis. The expression changes of key genes throughout the pseudotime progression were also analyzed.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e2.14 Statistical analysis\u003c/h2\u003e \u003cp\u003eBioinformatics analyses were conducted utilizing the R language (v 4.2.2). A Wilcoxon test was employed to compare the disparities between the 2 groups. A P-value of less than 0.05 was deemed statistically significant.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.1 The 16 candidate genes of SA were associated with toxicity in ATS\u003c/h2\u003e \u003cp\u003eFirstly, SMILES notation and molecular structure for ATS were obtained from the PubChem database (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The toxicity prediction result indicated that the LD50 of ATS was 5,000 mg/kg, placing it in toxicity class 5. This classification suggested that the compound exhibited relatively low toxicity in experimental animals. The confidence level of the prediction was 67.38%, indicating a moderate degree of reliability in the results. Furthermore, ATS demonstrated a higher potential for neurotoxicity, nephrotoxicity, and respiratory toxicity compared to other types of toxicity (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). The 102 toxic targets of ATS were identified by the Swiss Target database (\u003cb\u003eAdditional file 1\u003c/b\u003e). Differential expression analysis showed that there were 2,402 DEGs between the SA group and the control group. Among them, 29 genes in the SA group were identified as upregulated genes, and 2,373 genes were identified as downregulated genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec-d). A total of 7,859 targets for SA were identified by the CTD database (\u003cb\u003eAdditional file 2\u003c/b\u003e). Moreover, a total of 16 shared genes between DEGs, toxic targets, and targets for SA were identified and selected as candidate genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). The enrichment analysis of the 16 candidate genes revealed associations with 549 GO terms, including response to lipopolysaccharide, cell cortex, and metalloendopeptidase activity, among others (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef, \u003cb\u003eAdditional file 3\u003c/b\u003e). Moreover, the KEGG enrichment analysis of the candidate genes revealed that a total of 96 KEGG pathways were enriched, for instance relaxin signaling pathway, the IL-17 signaling pathway, and lipid and atherosclerosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg, \u003cb\u003eAdditional file 4\u003c/b\u003e). Next, a PPI network was constructed, comprising 20 interactions associated with 16 candidate genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh). In this network, MMP9, NOS2, MMP13, and MAPK8 had frequent protein-level interactions with other genes. In the \u0026ldquo;compound-target-pathway-disease\u0026rdquo; network, it was possible to visualize the connections between candidate genes, compounds, pathways, and diseases. Notably, 10 candidate genes (MMP1, MMP13, MAPK8, NOS2, GRB2, PIK3CB, CASP3, IKBKE, BACE1, RARB) were associated with compounds, diseases, and enriched pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ei).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e3.2 The BACE1 and PDE4A were identified as key genes\u003c/h2\u003e \u003cp\u003eFor the results of the LASSO algorithm, when the minimum lambda value was 0.1472063, a total of 5 LASSO genes were identified among the candidate genes, namely FOLR1, MMP13, BACE1, NOS2, and PDE4A (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea-b). When applying the SVM-RFE algorithm, the model demonstrated the highest prediction accuracy with an optimal variable count of 7. This configuration enabled the selection of 7 SVM-RFE genes from the candidate genes: FOLR1, PDE4A, MMP13, NOS2, PIK3CB, BACE1, and MMP12 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). Subsequently, we intersected the genes obtained from the mentioned above machine learning algorithms and finally found that FOLR1, PDE4A, MMP13, NOS2, and BACE1 could be indicated by all the algorithms, meaning that these genes could be used as candidate key genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Subsequently, according to the GSE1428 and GSE38718 datasets, BACE1 and PDE4A exhibited markedly low expression levels in the SA group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee-f). Consequently, 2 key genes\u0026mdash;BACE1 and PDE4A\u0026mdash;were selected for subsequent analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Exploring the function of key genes and immune infiltration of SA\u003c/h2\u003e \u003cp\u003eGSEA analysis observed that pathways associated with the expression of BACE1 and PDE4A were 8, respectively (\u003cb\u003eAdditional file 5\u003c/b\u003e). Notably, biomarkers were co-enriched in several common pathways, such as parkinsons disease, oxidative phosphorylation, alzheimers disease, and Huntington's disease (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-b).\u003c/p\u003e \u003cp\u003eThe distribution of immunity scores for the 28 immune cell types in the SA and control groups was illustrated in the thermal map, which displayed varying degrees of infiltration for each immune cell type (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). Differences in immune cell infiltration between the 2 groups revealed that the 4 immune cell types exhibited significant variations, including CD56dim natural killer cells, effector memory CD4 T cells, neutrophils, and regulatory T cells, all of which were notably less abundant in the samples from the SA group (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed). In addition, the association analysis of biomarkers with the most differentially infiltrated immune cell types revealed predominantly significant positive correlations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee). BACE1 and PDE4A were positively correlated with effector memory CD4 T cells (r\u0026thinsp;\u0026gt;\u0026thinsp;0.40, P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). These results suggested that BACE1 and PDE4A might have the same roles in regulating immune cell infiltration and function, potentially influencing the pathogenesis and immune regulation in SA.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Detection of ATS's ability to bind to key genes\u003c/h2\u003e \u003cp\u003eAccording to the molecular docking results, ATS demonstrated superior binding affinity to key genes BACE1 and PDE4A, with binding energies of less than \u0026minus;\u0026thinsp;8 kcal/mol (\u003cb\u003eAdditional file 6\u003c/b\u003e). The conformation indicated that lifitegrast interacted with the residues of the key genes BACE1 and PDE4A through hydrogen bonds (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea-b). To validate the reliability of molecular docking, MDS was implemented in this study. The proteins reached a steady state within the time frame of 0 ns to 50 ns. This suggested that the interactions between BACE1-ATS and PDE4A-ATS exhibited a relatively stable binding conformation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec). The findings from the RMSF analysis indicated that the amino acids of the protein exhibited flexibility and maintained stable binding interactions with the small molecule drug ligands throughout the duration of the simulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed). As can be seen in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee, the number of hydrogen bonds fluctuated periodically throughout the duration of the simulation, indicating a dynamic equilibrium at the binding interface. In addition, the energy of the proteins exhibited less fluctuation throughout the duration of the simulation and remained in a steady state (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef). The above analysis demonstrated that the binding of BACE1-ATS and PDE4A-ATS was relatively stable throughout the simulation process.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e3.5 The 9 cell types were annotated in the SA scRNA-seq data\u003c/h2\u003e \u003cp\u003eFirst, quality control processing was performed on the raw data of the scRNA-seq data for subsequent analysis. Before quality control, the initial count comprised 79,649 cells and 48,355 genes. Subsequent to quality control, the cell count was refined to 72,378, while the gene count remained constant at 48,355 (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea\u003c/b\u003e). Second, the top 2,000 HVGs were identified, and the top 10 genes with the greatest variability included APOD, DCN, and CXCL14 (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb\u003c/b\u003e). After PC of 30, the significance decreased, and the curve in the PC scree plot flattened at PC\u0026thinsp;=\u0026thinsp;30 (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ec\u003c/b\u003e). Therefore, we chose the top 30 PCs for additional analysis. After that, cells were partitioned into 22 cell clusters (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ed\u003c/b\u003e). Relying on the expression of marker genes (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ee\u003c/b\u003e), the 22 cell clusters were annotated into 9 cell types: endothelial cells (EC), fibroblast activation protein (FAP), lymphocytes, mast cells, muscle satellite cells (MuSC), myeloid cells, smooth muscle cells (SMC), fibroblast-like, and type II cells (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ef\u003c/b\u003e). The percentages of FAP, SMC, fibroblast-like, type II cells, and mast cells were higher in the SA samples than in the control samples, whereas the percentages of MuSC, EC, and myeloid cells were greater in the control group than in the SA group (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eg\u003c/b\u003e). Functional enrichment analysis revealed type II cells and mast cells were predominantly enriched in pathways related to \u0026ldquo;synthesis of Hepoxilins (HX) and Trioxilins (TrX)\u0026rdquo; and \u0026ldquo;Reactions specific to the hybrid N-glycan synthesis pathway\u0026rdquo;, and myeloid cells exhibited obviously enrichment in pathways such as the \u0026ldquo;metabolism of ingested MeSeO2H into MeSeH\u0026rdquo;, suggesting their involvement in substance metabolism (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eh\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e3.6 Myeloid cells were identified as key cells\u003c/h2\u003e \u003cp\u003eFirst, 2 key genes, BACE1 and PDE4A, were expressed and distributed across 9 annotated cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). It was further discovered that there was a significant difference in the expression of the BACE1 gene between the SA and control groups in FAP, MuSC, and myeloid cells (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). In contrast, the PDE4A gene exhibited significant differences in only 2 groups of samples within myeloid cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec). We identified TFs in myeloid cells that exhibited differing levels of activity between the SA and control groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). For instance, ZEB2, HBP1, FOXP1, and HHEX demonstrated high activity in the control group, whereas TCF12, MEF2A, TEAD1, and HOXB13 showed increased activity in the SA group. This suggested that these TFs might have distinct regulatory roles in various physiological or pathological conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e3.7 Myeloid cells exhibited strong interactions with other cells in samples from the SA group\u003c/h2\u003e \u003cp\u003eA detailed analysis of cell communication revealed that, overall, the number of interactions among the 9 annotated cell types was increased, and the strength of these interactions was increased in the SA samples compared to the control samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea-b). Specifically, the interactions of myeloid cells with fibroblast-like, EC, and mast cells were diminished, while interactions with FAP were increased in control samples compared to samples from patients with SA (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ec-d). Furthermore, the LGALS9-CD45 signaling axis was identified as a key mediator in the interactions between myeloid cells and lymphocytes in the SA group, while the LGALS9-CD45 signaling axis was crucial for interactions among myeloid cells in the control group (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ee-f). These findings suggested distinct signaling mechanisms underlying cell communication in different disease contexts.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e3.8 Key genes exhibited distinct expression trends during myeloid cells differentiation trajectories\u003c/h2\u003e \u003cp\u003eTo further assess the cellular differentiation trajectory of the myeloid cells, they were first re-clustered into different cell subclusters, with myeloid cells divided into 7 subclusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea). Relying on the expression of marker genes, the 7 subclusters were annotated into 4 myeloid cell subclusters: type 2 conventional dendritic cells (cDC2), cDC3, macrophages, and monocytes (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb-c). Pseudotime analysis of myeloid cell differentiation uncovered a progressive trajectory from an early (dark blue) to a more mature (light blue) state, and the 4 subclusters could be roughly divided into 5 differentiation states, with macrophages representing the earliest stages of differentiation (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ed). During cell differentiation, the expression of the key gene BACE1 remained stable during the pre-differentiation stage, increased during the middle stage (monocytes), and subsequently decreased before stabilizing in the late stage. In contrast, the expression of the key gene PDE4A initially decreased during the early stage of cell differentiation (macrophages) and then increased during the middle and late stages (monocytes, cDC2, cDC3) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee). The results indicated that BACE1 and PDE4A were strongly correlated with the progression of the SA.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eSA is a degenerative disease that seriously affects the prognosis and quality of life of elderly patients. Aging is the main pathogenic factor, along with secondary causes caused by other diseases or medications. ATS is one of the representative drugs of the statin class and is widely used in elderly patients, with the most common side effect being drug-induced myopathy. This study identified toxic key genes involved in ATS-induced SA pathogenesis through transcriptomics combined with network toxicology, ultimately finding two key genes, BACE1 and PDE4A. Further analysis was conducted on the biological pathways involving these key genes, their relationship with immune infiltration, and the binding capacity between ATS and the key genes. Additionally, single-cell analysis revealed significant differential expression of these key genes in myeloid cells and analyzed their expression trends during myeloid cell differentiation, aiming to provide new directions for SA treatment[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBACE1 and PDE4A, two key genes, showed significantly down-regulated expression in both the training set and validation set of sarcopenic groups, a finding that warrants attention. Currently, there are no relevant reports on the role of these two genes in the pathogenesis of sarcopenia, which suggests that they may represent a new molecular mechanism of ATS-induced sarcopenia.\u003c/p\u003e \u003cp\u003eBeta-site amyloid precursor protein cleaving enzyme 1 (BACE1), located on human chromosome 11, is one of the chromosomes most frequently associated with human diseases. Studies have reported that downregulation of BACE1 induces an increase in Glypican-1(GPC-1) expression, while upregulation of GPC-1 concurrently enhances the expression of endothelial nitric oxide synthase (eNOS). Inhibition of BACE1 promotes an increase in eNOS expression by upregulating GPC-1 [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Under exposure to risk factors such as hypertension, hyperlipidemia, diabetes, and aging, the upregulation of eNOS protein expression may lead to a relative deficiency in L-arginine or reduced Nicotinamide Adenine Dinucleotide Phosphate (NADPH) supply, promoting eNOS uncoupling, resulting in increased superoxide anions and peroxynitrite, causing oxidative stress [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Chronic low-grade inflammation caused by oxidative stress has been shown to be harmful to human skeletal muscle[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. Normally, skeletal muscle proteins are balanced, continuously degraded and resynthesized; however, in aging and the associated increase in oxidative stress, this balance is disrupted, promoting the occurrence of sarcopenia [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].In summary, although BACE1 inhibition upregulates eNOS expression, this pathway may ultimately lead to a dysregulation of skeletal muscle protein balance and promote the occurrence of sarcopenia through the cascade reaction of oxidative stress and chronic inflammation, especially when multiple risk factors are present.\u003c/p\u003e \u003cp\u003ecAMP-specific phosphodiesterase 4A (PDE4A) plays an important role in regulating intracellular cyclic adenosine monophosphate (cAMP) levels. PDE4A can degrade cAMP [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e], and its downregulation leads to significant accumulation of cAMP. Abnormally elevated cAMP levels can affect muscle function and sarcopenia development through two main pathways. In the neuromuscular function pathway, excess cAMP inhibits inhibitory signals of γ-aminobutyric acid ergic(GABAergic) neurons, resulting in weakened calcium ion elevation triggered by GABA release, thereby affecting neural signal transmission [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The neuromuscular junctions (NMJs), as critical structures transmitting nerve signals to muscles, when dysfunctional, can cause muscles to lose normal innervation, subsequently promoting muscle atrophy [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Regarding myogenic differentiation, increased cAMP levels and excessive activation of cAMP-dependent protein kinase (PKA) directly suppress the myogenic differentiation process [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. This inhibitory effect may be achieved by downregulating the activity of muscle-specific transcription factors Myf-5 and MyoD, both of which are basic helix-loop-helix (bHLH) proteins that are key regulators of myogenic cell differentiation and muscle-specific gene transcription activation. Therefore, PDE4A downregulation may synergistically promote the development of sarcopenia through these interconnected pathways, affecting both neural regulation and muscle regeneration.\u003c/p\u003e \u003cp\u003eThe results of gene set enrichment analysis (GSEA) showed that the two key genes BACE1 and PDE4A are jointly enriched in the oxidative phosphorylation (OXPHOS) pathway and various neurodegenerative disease-related pathways, including Parkinson's disease, Alzheimer's disease (AD), Huntington's disease, and others. These enriched pathways may be associated with the pathogenesis of sarcopenia (SA). Oxidative phosphorylation, as the core metabolic pathway of mitochondria, and neurodegenerative diseases, through mechanisms such as neuro-muscular disjunction, both play important roles in the development of sarcopenia. Mitochondrial dysfunction, inflammatory response, neuro-muscular disjunction, and insulin resistance are considered critical links in the sarcopenia signaling network, with mitochondrial dysfunction regarded as a central regulatory hub [\u003cspan additionalcitationids=\"CR45 CR46\" citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. As a metabolic hub for energy production, mitochondria generate ATP for skeletal muscle contraction through oxidative phosphorylation, and a decline in nicotinamide adenine dinucleotide (NAD\u003csup\u003e+\u003c/sup\u003e) levels is related to many age-related diseases, including cognitive decline, metabolic diseases, SA, and weakness [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Studies have found that humans of different races with SA show reduced mitochondrial oxidative capacity and NAD\u003csup\u003e+\u003c/sup\u003e biosynthesis; lowering NAD\u003csup\u003e+\u003c/sup\u003e levels through ablation of nicotinamide phosphoribosyltransferase (NAMPT)-mediated NAD salvage may lead to age-related muscle degeneration in adult mice [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. It is hypothesized that BACE1 and PDE4 influence mitochondrial function by participating in oxidative phosphorylation reactions, thereby affecting the pathogenesis of SA, leading to AST. Additionally, in the context of neurodegenerative disease pathways, the results of gene set enrichment analysis in this study indicate that key genes BACE1 and PDE4A are co-enriched in pathways related to neurodegenerative diseases such as Parkinson's disease, Alzheimer's disease, and Huntington's disease, suggesting that these genes may regulate common pathological mechanisms between myasthenia and nervous system disorders. Although these neurodegenerative diseases present with various clinical features, they all influence muscle function through shared mechanisms such as neuro-muscular denervation, neuroinflammation, and protein misfolding [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. The study found that abnormal proteins in Alzheimer's, Parkinson's, and Huntington's diseases (tau protein, alpha-synuclein, and huntingtin protein, respectively) have similar damaging capabilities when invading brain cells. Once inside, these proteins damage or rupture vesicle membranes, allowing proteins to infiltrate the cytoplasm and cause functional impairments. The protein aggregates associated with these three diseases lead to a similar vesicle damage mechanism [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Similarly, protein aggregation and membrane structural damage are also observed in the pathogenesis of myasthenia. Research indicates that in elderly human skeletal muscle, 43% of 515 insoluble protein aggregates show a more than 1.5-fold variation in abundance across different age groups, with 15% showing significant differences, implying that insoluble protein aggregates are especially susceptible to aging and may play a pathogenic role similar to that in neurodegenerative diseases in myasthenia [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e]. These findings support the idea that BACE1 and PDE4A may play critical roles in neurodegenerative diseases and myasthenia by regulating protein homeostasis and membrane integrity.\u003c/p\u003e \u003cp\u003eThe immune infiltration results showed significant differences in immune cells between the SA group and the control, including CD56dim natural killer cells, Effector memory CD4 T cells, neutrophils, and regulatory T cells. Among them, Effector memory CD4 T cells are significantly positively correlated with two key genes, BACE1 and PDE4A. Effector memory CD4 T cells (TEM CD4 cells), with mitochondrial metabolism, are crucial for the exit of T cells from the quiescent state [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. Quiescent CD4 T cells in peripheral blood exist as naive cells, effector memory cells, and central memory cells [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. TEM CD4 cells are antigen-experienced cells that can produce rapid and robust responses upon re-encountering the same antigen. Mitochondria, functioning as the \"energy factory\" and \"reactive oxygen species (ROS) source\" in muscle cells, play a core role in the development of sarcopenia when their function is abnormal. Studies have shown that mitochondrial ATP and ROS metabolism are critical in platelet-regulated TEM CD4 cell responses, with PF4 being the main mediator of platelet-regulated TEM CD4 cell responses. TEM CD4 cells respond to PF4 binding to their CXCR3 receptor by increasing expression of mitochondrial transcription factor A (TFAM), enhancing mitochondrial biogenesis, and increasing production of ATP and ROS. These changes, in turn, stimulate the expression of T-bet and FoxP3, promoting the TEM CD4 cell immune response [\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e]. Meanwhile, ROS accumulation can activate the JNK pathway, amplifying inflammatory signals and atrophic processes [\u003cspan additionalcitationids=\"CR58\" citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]. Both TEM CD4 cells and mitochondria collaboratively promote progressive declines in muscle mass and strength by disrupting muscle energy homeostasis, triggering oxidative stress, and activating catabolic pathways. In summary, BACE1 and PDE4A not only co-enriched in the oxidative phosphorylation pathway in pathway enrichment analysis, but also showed a significant positive correlation with TEM CD4 cells, suggesting that these two genes may influence the immune-metabolic network of TEM CD4 cells by regulating mitochondrial oxidative phosphorylation processes. Subsequently, they jointly drive the development and progression of sarcopenia through oxidative stress and inflammatory responses mediated by ROS.\u003c/p\u003e \u003cp\u003eMolecular docking and molecular dynamics results show that key genes BACE1 and PDE4A have strong binding activity. Molecular dynamics simulations further validate the reasonableness and reliability of the docking results. The good molecular binding activity and the stability of the docking results suggest that ATS plays a certain role in inducing the occurrence and development of SA, and how to prevent the further development of SA while patients are treated with ATS drugs remains to be studied.\u003c/p\u003e \u003cp\u003eSingle-cell data analysis revealed that myeloid cells are key cells in the process of ATS-induced SA occurrence and development, and myeloid cells are further divided into subgroups such as macrophages. Fundamental changes in the immune system of aged animals occur at an early stage of hematopoietic development, with differentiated hematopoietic stem cells (HSCs) almost generating all blood cells; in aging, lymphoid potential decreases and shifts toward myeloid lineage [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. This age-related bone marrow bias is reflected in the changing number of mature bone marrow cells in circulation, with the number of myeloid dendritic cell subpopulations gradually decreasing with age, accompanied by a reduction in CD34\u0026thinsp;+\u0026thinsp;progenitors and an increase in circulating monocytes, indicating a partial disruption of the entire antigen-presenting cell differentiation process in the elderly [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. Some studies report that transplanting young bone marrow cells into elderly recipients can prevent sarcopenia and age-related changes in muscle fiber phenotype, while transplanting aging bone marrow cells into young animals reduces satellite cell numbers and promotes their transition into a fibrotic phenotype. Aging of bone marrow cells promotes a decline in satellite cell number and function, leading to sarcopenia. Myeloid cell subpopulations, such as macrophages, play a crucial regulatory role in muscle growth, including modulating inflammatory responses, promoting tissue repair, and affecting muscle stem cell functions [\u003cspan additionalcitationids=\"CR63\" citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Macrophages are rapidly activated during muscle activity or injury [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e], and their induced polarization states can stimulate satellite cell proliferation and differentiation, promote tissue angiogenesis, thereby accelerating muscle regeneration [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e, \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. Notably, statins have been shown to possess specific regulatory abilities on myeloid cells, capable of targeting pathogenic myeloid cells with KRAS\u0026sup1;\u0026sup2;ᴰ mutations by inhibiting the mevalonate pathway, blocking KRAS isoprenylation, and subsequently downregulating the PI3K-AKT pathway, achieving precise intervention without affecting normal myeloid or other immune cells [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e]. Further single-cell differentiation trajectory analysis revealed the dynamic expression patterns of BACE1 and PDE4A during the differentiation process of myeloid cells: BACE1 significantly increased at the middle stage of differentiation (monocytes stage), while PDE4A gradually increased in the later stages of differentiation (monocytes, cDC2, cDC3 stages). This temporal and stage-specific expression pattern suggests that these two genes may play a synergistic regulatory role at different stages of myeloid cell differentiation. Combining the selective targeting ability of atorvastatin on myeloid cells, the differential expression of BACE1 and PDE4A may affect the differentiation and maturation process of monocytes into macrophages, interfere with the macrophage polarization state required for muscle repair, and thus play a key role in the pathogenesis of ATS-induced sarcopenia.\u003c/p\u003e \u003cp\u003eThis study identified the toxic key genes responsible for ATS-induced SA occurrence through transcriptome combined with network toxicology. A series of bioinformatics methods were used to explore the specific mechanisms by which these key genes influence the development of SA, and the expression of these genes in key cells was analyzed based on single-cell data to provide new theoretical support and evidence for the pathogenesis of SA caused by AST. However, there are some limitations in this study, as the specific molecular mechanisms of these key genes in SA have not yet been thoroughly investigated. Therefore, future research should focus on exploring the functions of these genes and their potential applications in treatment to further validate their clinical value.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study systematically elucidated the molecular mechanisms of atorvastatin-induced sarcopenia by integrating network toxicology, transcriptomics, and single-cell RNA sequencing. BACE1 and PDE4A were identified as key toxic genes, with their downregulation potentially contributing to sarcopenia pathogenesis through oxidative stress, immune dysregulation, and mitochondrial dysfunction. Myeloid cells were highlighted as critical cellular players, with dynamic expression patterns of these genes during cell differentiation. These findings provide novel insights into drug-induced sarcopenia and offer potential therapeutic targets for preventing atorvastatin-related myopathy.\u003c/p\u003e"},{"header":"List of abbreviations","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAbbreviation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eFull Term\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eSA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eSarcopenia\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eATS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eAtorvastatin\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003escRNA-seq\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003esingle-cell RNA sequencing\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eDEGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003edifferentially expressed genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ePDE4A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ecAMP-specific phosphodiesterase 4A\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eBACE1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eBeta-site amyloid precursor protein cleaving enzyme 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eLDL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003elow-density lipoprotein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eFDA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eFood and Drug Administration\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ePPI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eProtein-Protein Interaction\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGene Ontology\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eKEGG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eLASSO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eLeast Absolute Shrinkage and Selection Operator\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eSVM-RFE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eSupport Vector Machine-Recursive Feature Elimination\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGSEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eNES\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eNormalized Enrichment Score\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003essGSEA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003esingle-sample Gene Set Enrichment Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGSVA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGene Set Variation Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eMDS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eMolecular Dynamics Simulation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eRMSD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eRoot Mean Square Deviation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eRMSF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eRoot Mean Square Fluctuation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eHVGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eHighly Variable Genes\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ePCA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ePrincipal Component Analysis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eTF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eTranscription Factor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eL-R\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eLigand-Receptor\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eEndothelial Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eFAP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eFibroblast Activation Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eMuSC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eMuscle Satellite Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eSMC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eSmooth Muscle Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ecDC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003etype 2 conventional dendritic cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGPC-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGlypican-1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eeNOS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eendothelial nitric oxide synthase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eNADPH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eNicotinamide Adenine Dinucleotide Phosphate\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ecAMP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ecyclic adenosine monophosphate\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ePKA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ecAMP-dependent protein kinase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ebHLH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ebasic helix-loop-helix\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eOXPHOS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eOxidative Phosphorylation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eTEM CD4 cells\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eEffector Memory CD4 T cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eROS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eReactive Oxygen Species\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eTFAM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003emitochondrial transcription factor A\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eJNK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003ec-Jun N-terminal Kinase\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eHSCs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eHematopoietic Stem Cells\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGEO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGene Expression Omnibus\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eCTD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eComparative Toxicogenomics Database\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eSTRING\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eSearch Tool for the Retrieval of Interacting Genes/Proteins\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eCytoscape\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eCytoscape Software\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eMSigDB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eMolecular Signatures Database\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003ePDB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eProtein Data Bank\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eCB-Dock2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eCB-Dock2 Web Server\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGROMACS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGROningen MAchine for Chemical Simulation\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eAMBER\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eAssisted Model Building with Energy Refinement\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eGAFF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eGeneral Amber Force Field\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eNVT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eCanonical Ensemble (Constant Number of particles, Volume, Temperature)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 152px;\"\u003e\n \u003cp\u003eNPT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 405px;\"\u003e\n \u003cp\u003eIsothermal-Isobaric Ensemble (Constant Number of particles, Pressure, Temperature)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets analysed during the current study are available in the GEO database (GSE1428 and GSE38718, http://www.ncbi.nlm.nih.gov/geo/) and the CNGB Nucleotide Sequence Archive database (CNP0004394, CNP0004395, CNP0004494, and CNP0004495, https://db.cngb.org/cdcp/hlma/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWang Fang: Conceptualization, Data curation, Validation, Visualization, Writing\u0026ndash;original draft, Writing\u0026ndash;review \u0026amp; editing. Zhao Fei: Data curation, Validation, Visualization, Writing\u0026ndash;review \u0026amp; editing. Zhao Yue: Data curation, Validation, Visualization, Writing\u0026ndash;review \u0026amp; editing. Cai Liangling: Visualization, Writing\u0026ndash;review \u0026amp; editing. Wei Yuanhuan: Supervision, Writing\u0026ndash;review \u0026amp; editing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our sincere gratitude to all individuals and organizations who supported and assisted us throughout this research. Special thanks to the following authors: Wang Fang, Zhao Fei, Zhao Yue, Cai Liangling, Wei Yuanhuan.In conclusion, we extend our thanks to everyone who has supported and assisted us along the way. Without your support, this research would not have been possible.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLi X, Wu C, Lu X, Wang L. Predictive models of sarcopenia based on inflammation and pyroptosis-related genes. Front Genet. 2024;15:1491577.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerez K, Ciotlos S, McGirr J, Limbad C, Doi R, Nederveen JP, Nilsson MI, Winer DA, Evans W, Tarnopolsky M, et al. Single nuclei profiling identifies cell specific markers of skeletal muscle aging, frailty, and senescence. Aging. 2022;14(23):9393\u0026ndash;422.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang S, Xiang C, Song Y. Identification of the shared gene signatures and pathways between sarcopenia and type 2 diabetes mellitus. PLoS ONE. 2022;17(3):e0265221.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Zhu G, Zhang F, Yu D, Jia X, Ma B, Chen W, Cai X, Mao L, Zhuang C, et al. Identification of a novel immune-related transcriptional regulatory network in sarcopenia. BMC Geriatr. 2023;23(1):463.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDamluji AA, Alfaraidhy M, AlHajri N, Rohant NN, Kumar M, Al Malouf C, Bahrainy S, Ji Kwak M, Batchelor WB, Forman DE, et al. Sarcopenia and Cardiovascular Diseases. Circulation. 2023;147(20):1534\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuzuya M. Drug-related sarcopenia as a secondary sarcopenia. Geriatr Gerontol Int. 2024;24(2):195\u0026ndash;203.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCho MR, Lee S, Song SK. A Review of Sarcopenia Pathophysiology, Diagnosis, Treatment and Future Direction. J Korean Med Sci. 2022;37(18):e146.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReig-L\u0026oacute;pez J, Garc\u0026iacute;a-Arieta A, Mangas-Sanju\u0026aacute;n V, Merino-Sanju\u0026aacute;n M. Current Evidence, Challenges, and Opportunities of Physiologically Based Pharmacokinetic Models of Atorvastatin for Decision Making. Pharmaceutics 2021, 13(5).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReig-L\u0026oacute;pez J, Merino-Sanjuan M, Garc\u0026iacute;a-Arieta A, Mangas-Sanju\u0026aacute;n V. A physiologically based pharmacokinetic model for open acid and lactone forms of atorvastatin and metabolites to assess the drug-gene interaction with SLCO1B1 polymorphisms. Biomed pharmacotherapy = Biomedecine pharmacotherapie. 2022;156:113914.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGanga HV, Slim HB, Thompson PD. A systematic review of statin-induced muscle problems in clinical trials. Am Heart J. 2014;168(1):6\u0026ndash;15.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng Z, Hu J, Xiao G, Liu Y, Jia D, Wu G, Xie C, Li S, Bi X. Integrating network toxicology and molecular docking to explore the toxicity of the environmental pollutant butyl hydroxyanisole: An example of induction of chronic urticaria. Heliyon. 2024;10(15):e35409.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou GL, Su SL, Yu L, Shang EX, Hua YQ, Yu H, Duan JA. Exploring the liver toxicity mechanism of Tripterygium wilfordii extract based on metabolomics, network pharmacological analysis and experimental validation. J Ethnopharmacol. 2025;337(Pt 2):118888.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z, Ding H, Zou Q. Identifying cell types to interpret scRNA-seq data: how, why and more possibilities. Brief Funct Genomics. 2020;19(4):286\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng C, Chen W, Jin H, Chen X. A Review of Single-Cell RNA-Seq Annotation, Integration, and Cell-Cell Communication. Cells 2023, 12(15).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLai Y, Ram\u0026iacute;rez-Pardo I, Isern J, An J, Perdiguero E, Serrano AL, Li J, Garc\u0026iacute;a-Dom\u0026iacute;nguez E, Segal\u0026eacute;s J, Guo P, et al. Multimodal cell atlas of the ageing human skeletal muscle. Nature. 2024;629(8010):154\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLove MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGustavsson EK, Zhang D, Reynolds RH, Garcia-Ruiz S, Ryten M. ggtranscript: an R package for the visualization and interpretation of transcript isoforms using ggplot2. Bioinf (Oxford England). 2022;38(15):3844\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinf (Oxford England). 2016;32(18):2847\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen H, Boutros PC. VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R. BMC Bioinformatics. 2011;12:35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innov (Cambridge (Mass). 2021;2(3):100141.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498\u0026ndash;504.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw. 2010;33(1):1\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi H, Yuan X, Liu G, Fan W. Identifying and Validating GSTM5 as an Immunogenic Gene in Diabetic Foot Ulcer Using Bioinformatics and Machine Learning. J Inflamm Res. 2023;16:6241\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobles-Jimenez LE, Aranda-Aguirre E, Castelan-Ortega OA, Shettino-Bermudez BS, Ortiz-Salinas R, Miranda M, Li X, Angeles-Hernandez JC, Vargas-Bello-P\u0026eacute;rez E, Gonzalez-Ronquillo M. Worldwide Traceability of Antibiotic Residues from Livestock in Wastewater and Soil: A Systematic Review. Animals: open access J MDPI 2021, 12(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann S, Castelo R, Guinney J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics. 2013;14:7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu ZY, Huang RH. Integrating single-cell RNA-sequencing and bulk RNA-sequencing data to explore the role of mitophagy-related genes in prostate cancer. Heliyon. 2024;10(9):e30766.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSatija R, Farrell JA, Gennert D, Schier AF, Regev A. Spatial reconstruction of single-cell gene expression data. Nat Biotechnol. 2015;33(5):495\u0026ndash;502.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGriss J, Viteri G, Sidiropoulos K, Nguyen V, Fabregat A, Hermjakob H. ReactomeGSA - Efficient Multi-Omics Comparative Pathway Analysis. Mol Cell Proteom. 2020;19(12):2115\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCornwell M, Vangala M, Taing L, Herbert Z, K\u0026ouml;ster J, Li B, Sun H, Li T, Zhang J, Qiu X et al. VIPER: Visualization Pipeline for RNA-seq, a Snakemake workflow for efficient and complete RNA-seq analysis. \u003cem\u003eBMC Bioinformatics\u003c/em\u003e 2018, 19(1):135.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12(1):1088.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmalley I, Chen Z, Phadke M, Li J, Yu X, Wyatt C, Evernden B, Messina JL, Sarnaik A, Sondak VK, et al. Single-Cell Characterization of the Immune Microenvironment of Melanoma Brain and Leptomeningeal Metastases. Clin cancer research: official J Am Association Cancer Res. 2021;27(14):4109\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong B, Li Y, Yang R, Dai S, Zhan Y, Zhang WB, Dong R. Single-cell transcriptional profiling reveals heterogeneity and developmental trajectories of Ewing sarcoma. J Cancer Res Clin Oncol. 2022;148(12):3267\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe T, d'Uscio LV, Sun R, Santhanam AVR, Katusic ZS. Inactivation of BACE1 increases expression of endothelial nitric oxide synthase in cerebrovascular endothelium. J Cereb blood flow metabolism: official J Int Soc Cereb Blood Flow Metabolism. 2022;42(10):1920\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKatusic ZS. Vascular endothelial dysfunction: does tetrahydrobiopterin play a role? Am J Physiol Heart Circ Physiol. 2001;281(3):H981\u0026ndash;986.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward C, Ferrucci L, Sun K, Fried LP, Walston J, Varadhan R, Guralnik JM, Semba RD. Oxidative protein damage is associated with poor grip strength among older women living in the community. \u003cem\u003eJournal of applied physiology (Bethesda, Md\u003c/em\u003e: 1985) 2007, 103(1):17\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiu PM, Pistilli EE, Alway SE. Age-dependent increase in oxidative stress in gastrocnemius muscle with unloading. J Appl Physiol (Bethesda Md: 1985). 2008;105(6):1695\u0026ndash;705.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoopman R, van Loon LJ. Aging, exercise, and muscle protein metabolism. J Appl Physiol (Bethesda Md: 1985). 2009;106(6):2040\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeng SJ, Yu LJ. Oxidative stress, molecular inflammation and sarcopenia. Int J Mol Sci. 2010;11(4):1509\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrancis SH, Blount MA, Corbin JD. Mammalian cyclic nucleotide phosphodiesterases: molecular mechanisms and physiological functions. Physiol Rev. 2011;91(2):651\u0026ndash;90.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eObrietan K, van den Pol AN. GABA activity mediating cytosolic Ca2\u0026thinsp;+\u0026thinsp;rises in developing neurons is modulated by cAMP-dependent signal transduction. J neuroscience: official J Soc Neurosci. 1997;17(12):4785\u0026ndash;99.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArnold WD, Clark BC. Neuromuscular junction transmission failure in aging and sarcopenia: The nexus of the neurological and muscular systems. Ageing Res Rev. 2023;89:101966.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBao Z, Cui C, Chow SK, Qin L, Wong RMY, Cheung WH. AChRs Degeneration at NMJ in Aging-Associated Sarcopenia-A Systematic Review. Front Aging Neurosci. 2020;12:597811.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWinter B, Braun T, Arnold HH. cAMP-dependent protein kinase represses myogenic differentiation and the activity of the muscle-specific helix-loop-helix transcription factors Myf-5 and MyoD. J Biol Chem. 1993;268(13):9869\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang YF, Yang W, Liao ZY, Wu YX, Fan Z, Guo A, Yu J, Chen QN, Wu JH, Zhou J, et al. MICU3 regulates mitochondrial Ca(2+)-dependent antioxidant response in skeletal muscle aging. Cell Death Dis. 2021;12(12):1115.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu H, Ranjit R, Richardson A, Van Remmen H. Muscle mitochondrial catalase expression prevents neuromuscular junction disruption, atrophy, and weakness in a mouse model of accelerated sarcopenia. J cachexia sarcopenia muscle. 2021;12(6):1582\u0026ndash;96.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez-Hedstr\u0026ouml;m D, Priego T, Amor S, de la Fuente-Fern\u0026aacute;ndez M, Mart\u0026iacute;n AI, L\u0026oacute;pez-Calder\u0026oacute;n A, Inarejos-Garc\u0026iacute;a AM, Garc\u0026iacute;a-Villal\u0026oacute;n \u0026Aacute;L, Granado M. Olive Leaf Extract Supplementation to Old Wistar Rats Attenuates Aging-Induced Sarcopenia and Increases Insulin Sensitivity in Adipose Tissue and Skeletal Muscle. Antioxid (Basel Switzerland) 2021, 10(5).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaram C, Yi J, Xiao Y, Dhakal K, Zhang L, Li X, Manno C, Xu J, Li K, Cheng H, et al. Absence of physiological Ca(2+) transients is an initial trigger for mitochondrial dysfunction in skeletal muscle following denervation. Skelet Muscle. 2017;7(1):6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCovarrubias AJ, Perrone R, Grozio A, Verdin E. NAD(+) metabolism and its roles in cellular processes during ageing. Nat Rev Mol Cell Biol. 2021;22(2):119\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMigliavacca E, Tay SKH, Patel HP, Sonntag T, Civiletto G, McFarlane C, Forrester T, Barton SJ, Leow MK, Antoun E, et al. Mitochondrial oxidative capacity and NAD(+) biosynthesis are reduced in human sarcopenia across ethnicities. Nat Commun. 2019;10(1):5808.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrederick DW, Loro E, Liu L, Davila A Jr., Chellappa K, Silverman IM, Quinn WJ 3rd, Gosai SJ, Tichy ED, Davis JG, et al. Loss of NAD Homeostasis Leads to Progressive and Reversible Degeneration of Skeletal Muscle. Cell Metabol. 2016;24(2):269\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCiurea AV, Mohan AG, Covache-Busuioc RA, Costin HP, Glavan LA, Corlatescu AD, Saceleanu VM. Unraveling Molecular and Genetic Insights into Neurodegenerative Diseases: Advances in Understanding Alzheimer's, Parkinson's, and Huntington's Diseases and Amyotrophic Lateral Sclerosis. Int J Mol Sci 2023, 24(13).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIn: \u003cem\u003eParkinson\u0026rsquo;s Disease: Pathogenesis and Clinical Aspects.\u003c/em\u003e Edited by Stoker TB, Greenland JC. Brisbane (AU): Codon Publications Copyright \u0026copy; 2018 Codon Publications.; 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKedlian VR, Wang Y, Liu T, Chen X, Bolt L, Tudor C, Shen Z, Fasouli ES, Prigmore E, Kleshchevnikov V, et al. Human skeletal muscle aging atlas. Nat aging. 2024;4(5):727\u0026ndash;44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan H, Yang K, Li Y, Shaw TI, Wang Y, Blanco DB, Wang X, Cho JH, Wang H, Rankin S, et al. Integrative Proteomics and Phosphoproteomics Profiling Reveals Dynamic Signaling Networks and Bioenergetics Pathways Underlying T Cell Activation. Immunity. 2017;46(3):488\u0026ndash;503.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeder RA, Ahmed R. Similarities and differences in CD4\u0026thinsp;+\u0026thinsp;and CD8\u0026thinsp;+\u0026thinsp;effector and memory T cell generation. Nat Immunol. 2003;4(9):835\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTan S, Li S, Min Y, Gister\u0026aring; A, Moruzzi N, Zhang J, Sun Y, Andersson J, Malmstr\u0026ouml;m RE, Wang M, et al. Platelet factor 4 enhances CD4(+) T effector memory cell responses via Akt-PGC1α-TFAM signaling-mediated mitochondrial biogenesis. J Thromb haemostasis: JTH. 2020;18(10):2685\u0026ndash;700.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDimeloe S, Mehling M, Frick C, Loeliger J, Bantug GR, Sauder U, Fischer M, Belle R, Develioglu L, Tay S, et al. The Immune-Metabolic Basis of Effector Memory CD4\u0026thinsp;+\u0026thinsp;T Cell Function under Hypoxic Conditions. J Immunol (Baltimore Md: 1950). 2016;196(1):106\u0026ndash;14.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Y, Ye Y, Krau\u0026szlig; PL, L\u0026ouml;we P, Pfeiffenberger M, Damerau A, Ehlers L, Buttgereit T, Hoff P, Buttgereit F, et al. Age-related increase of mitochondrial content in human memory CD4\u0026thinsp;+\u0026thinsp;T cells contributes to ROS-mediated increased expression of proinflammatory cytokines. Front Immunol. 2022;13:911050.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHayashi T, Kato N, Furudoi K, Hayashi I, Kyoizumi S, Yoshida K, Kusunoki Y, Furukawa K, Imaizumi M, Hida A, et al. Early-life atomic-bomb irradiation accelerates immunological aging and elevates immune-related intracellular reactive oxygen species. Aging Cell. 2023;22(10):e13940.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDorshkind K, H\u0026ouml;fer T, Montecino-Rodriguez E, Pioli PD, Rodewald HR. Do haematopoietic stem cells age? Nat Rev Immunol. 2020;20(3):196\u0026ndash;202.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDella Bella S, Bierti L, Presicce P, Arienti R, Valenti M, Saresella M, Vergani C, Villa ML. Peripheral blood dendritic cells and monocytes are differently regulated in the elderly. Clin Immunol (Orlando Fla). 2007;122(2):220\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHernandez-Torres F, Matias-Valiente L, Alzas-Gomez V, Aranega AE. Macrophages in the Context of Muscle Regeneration and Duchenne Muscular Dystrophy. \u003cem\u003eInternational journal of molecular sciences\u003c/em\u003e 2024, 25(19).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinari ALA, Thomatieli-Santos RV. From skeletal muscle damage and regeneration to the hypertrophy induced by exercise: what is the role of different macrophage subsets? Am J Physiol Regul Integr Comp Physiol. 2022;322(1):R41\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaclier M, Cuvellier S, Magnan M, Mounier R, Chazaud B. Monocyte/macrophage interactions with myogenic precursor cells during skeletal muscle regeneration. FEBS J. 2013;280(17):4118\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTidball JG. Regulation of muscle growth and regeneration by the immune system. Nat Rev Immunol. 2017;17(3):165\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen J, Zhu X, Liu H. MiR-483 induces senescence of human adipose-derived mesenchymal stem cells through IGF1 inhibition. Aging. 2020;12(15):15756\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKamata T, Giblett S, Pritchard C. KRAS(G12D) expression in lung-resident myeloid cells promotes pulmonary LCH-like neoplasm sensitive to statin treatment. Blood. 2017;130(4):514\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Table 1","content":"\u003cp\u003e\u003cstrong\u003eTable 1 The binding energy of molecular docking.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eKey genes\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003edrug\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTotal Score\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBACE1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eATS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-9.3 Kcal/mol\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePDE4A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eATS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-8.9 Kcal/mol\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Sarcopenia, Atorvastatin, Network toxicology, Single-cell RNA sequencing","lastPublishedDoi":"10.21203/rs.3.rs-9058922/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9058922/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eSarcopenia (SA) significantly affects the quality of life in the elderly. Atorvastatin (ATS), a lipid-lowering drug, may cause myopathy as a side effect. This study aimed to predict key genes involved in ATS-induced SA using network.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003ePublic databases were used to obtain single-cell RNA sequencing (scRNA-seq) data, transcriptome data, and targets for SA and ATS toxicity. Toxicity prediction was conducted, and candidate genes were identified through differentially expressed genes (DEGs) associated with SA, as well as ATS toxicity targets. Machine learning and gene expression analyses were used to identify key genes, followed by functional enrichment and immune infiltration analysis. Molecular docking and dynamics simulations assessed the binding affinity between ATS and key genes. Key cell types were identified through scRNA-seq analysis.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThe toxicity prediction indicated that ATS exhibited relatively low toxicity in experimental animals. A total of 16 candidate genes were identified from the intersection of 2,402 DEGs, 101 toxicity targets, and 7,859 targets. BACE1 and PDE4A were selected as key genes. Functional enrichment analysis revealed their association with Parkinson's disease, oxidative phosphorylation, Alzheimer's disease, and Huntington's disease. Immune infiltration analysis showed that BACE1 and PDE4A were positively correlated with effector memory CD4 T cells. Molecular docking confirmed strong binding affinity between ATS and key genes, forming stable complexes. Myeloid cells were identified as key cells involved in SA.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThis study provides insights into the molecular mechanisms of ATS-induced SA and supports strategies for preventing ATS-induced myopathy.\u003c/p\u003e","manuscriptTitle":"The mechanism of atorvastatin-induced sarcopenia elucidated based on network toxicology, single-cell RNA sequencing, and bulk transcriptomics data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-10 06:03:40","doi":"10.21203/rs.3.rs-9058922/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7afe345d-c92f-4c5a-abad-330a3210574f","owner":[],"postedDate":"March 10th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-10T12:10:50+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-10 06:03:40","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9058922","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9058922","identity":"rs-9058922","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}