Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice J Physiol Biochem . 2026 Apr 10;82(1):36. doi: 10.1007/s13105-026-01176-0 Search in PMC Search in PubMed View in NLM Catalog Add to search Identification of mitochondria-related genes in calcific aortic valve disease by integrated analysis of single-cell and bulk transcriptomic atlases Ziling Mai Ziling Mai 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Ziling Mai 1, 2, 3, # , Huijun Du Huijun Du 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Huijun Du 1, 2, 3, # , Yuhan Chen Yuhan Chen 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Yuhan Chen 1, 2, 3 , Yaner Lu Yaner Lu 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Yaner Lu 1, 2, 3 , Bingchen Liu Bingchen Liu 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Bingchen Liu 1, 2, 3, ✉ , Deqiang Sun Deqiang Sun 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China Find articles by Deqiang Sun 1, 2, 3, ✉ Author information Article notes Copyright and License information 1 Department of Cardiology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, 310009 China 2 State Key Laboratory of Transvascular Implantation Devices, Hangzhou, 310009 China 3 Heart Regeneration and Repair Key Laboratory of Zhejiang Province, Hangzhou, 310009 China ✉ Corresponding author. # Contributed equally. Received 2025 Nov 25; Accepted 2026 Mar 19; Issue date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13068690 PMID: 41961374 Abstract Calcific aortic valve disease (CAVD) is a highly prevalent heart valve disorder in which mitochondria act as critical regulators of calcification, yet their precise pathogenic mechanisms remain unclear. To elucidate these mechanisms, we integrated single-cell transcriptomic datasets comparing normal and calcified human aortic valves to identify 200 differentially expressed mitochondria-related genes (DE-MRGs), each exhibiting distinct expression patterns across diverse cellular subpopulations. Pseudotime trajectory analysis revealed 18 DE-MRGs with dynamic changes during the endothelial-to-mesenchymal transition, and intercellular communication analysis highlighted enhanced signaling between valve interstitial cells (VICs) and macrophages. Specifically, we hypothesized that distinct mitochondrial hubs may modulate these interactions. Using machine learning and bulk transcriptomic data, we identified microsomal glutathione S-transferase 1 (MGST1), an enzyme located on the outer mitochondrial and endoplasmic reticulum membranes, as a hub gene with high predictive performance. Subsequent validation confirmed that MGST1 is functionally involved in calcification, as its expression was markedly upregulated under calcifying conditions. Molecular docking further predicted that ritlecitinib exhibits the highest binding affinity for MGST1, and this molecule was shown to ameliorate calcification. In conclusion, this study delineates a comprehensive molecular network of MRGs in CAVD pathogenesis and identifies MGST1 as a mitochondria‑related hub gene that is upregulated in CAVD and functionally promotes calcification in vitro. Supplementary Information The online version contains supplementary material available at 10.1007/s13105-026-01176-0. Keyword: Calcific aortic valve disease (CAVD), Mitochondria-related genes, Diagnostic biomarker Introduction Calcific aortic valve disease (CAVD) is a prevalent heart valve disorder among aging populations [ 1 ]. Approximately one-third of older adults exhibit echocardiographic or radiological evidence of calcific aortic valve sclerosis (CAVS), which represents an early or subclinical stage of CAVD [ 2 , 3 ]. CAVD is characterized by progressive leaflet thickening and calcification, which ultimately lead to chronic heart failure. The pathogenesis of valve calcification involves a complex interplay of factors, including inflammatory processes [ 4 ], dysregulated lipid metabolism [ 5 , 6 ], heightened oxidative stress [ 7 ], mitochondrial dysfunction [ 8 ], and activation of osteogenic pathways [ 9 ]. Despite advances in our understanding of these mechanisms, there are currently no pharmacological agents available to prevent or reverse CAVD progression [ 10 ]. Mitochondria are not only the primary source of energy production in CAVD but also serve as central regulators of valve calcification. Previous studies have reported impaired mitochondrial respiration and reduced ATP generation in calcified aortic valve tissues [ 11 ]. In addition, significant dysregulation of mitophagy and mitochondrial biogenesis pathways has been observed [ 12 ]. Mitochondrial fission, mediated by dynamin-related protein 1 (DRP1), also plays a pivotal role in cardiovascular calcification [ 13 , 14 ]. Interestingly, agents such as rapamycin have been shown to reverse calcific phenotypes by enhancing autophagy [ 12 ]. Based on these studies, it is clear that mitochondrial dysfunction contributes directly to aortic valve calcification through disrupted energy metabolism, increased oxidative stress, upregulation of osteogenic genes, and impaired autophagy. Therefore, identifying key genes involved in mitochondrial injury during CAVD progression is crucial for early patient detection and therapeutic intervention. Current research lacks a systematic understanding of how mitochondrial gene networks coordinate complex tissue-level responses in CAVD. Beyond bioenergetics, it is recognized that mitochondrial dysfunction can trigger inflammatory signaling and alter matrix secretion [ 15 , 16 ]. Therefore, we hypothesized that specific mitochondria-related hub genes act as critical checkpoints modulating endothelial-to-mesenchymal transition (EndMT), extracellular matrix (ECM) remodeling, and immune–stromal crosstalk during valve calcification. In this context, MGST1 warrants particular attention. It is an integral membrane protein localized predominantly to the outer mitochondrial membrane and endoplasmic reticulum, where it catalyzes glutathione conjugation and protects cellular membranes from oxidative stress and lipid peroxidation [ 17 , 18 ]. MGST1 has been implicated in redox homeostasis, ferroptosis regulation, and tissue remodeling in several cardiovascular and metabolic conditions [ 19 , 20 ]. However, whether MGST1 is dysregulated in CAVD and how it might influence valvular calcification remain unknown. Here, we leveraged single-cell transcriptomic sequencing to systematically elucidate the multicellular cooperative mechanisms of CAVD from a mitochondrial perspective for the first time. Our findings provide novel insights into the complex pathology of CAVD and establish a theoretical foundation for the development of mitochondria-targeted therapeutic strategies. Material and methods Overall study design The study workflow is illustrated in Fig. 1 . Briefly, the study was conducted in four phases: (1) Single-cell Atlas Construction: Single-cell transcriptomic datasets from normal and calcified human aortic valves were integrated and batch-corrected to identify key cell populations, including valve-derived stromal cells (VDSCs). (2) MRG Screening & Functional Analysis: First, 200 differentially expressed mitochondria-related genes (DE-MRGs) were identified. Subsequently, pseudotime trajectory analysis and cell–cell communication analysis were performed to uncover their roles in EndMT and immune-stromal crosstalk. (3) Key Target Identification: Machine learning algorithms (LASSO and SVM-RFE) were applied to screen for robust diagnostic markers, identifying MGST1 as a hub gene correlated with macrophage infiltration. (4) Experimental Validation & Drug Discovery: The function of MGST1 was validated via in vitro experiments (western blotting). Furthermore, ritlecitinib was identified as a potential therapeutic candidate through molecular docking and preliminarily evaluated via in vitro pharmacological assays. Fig. 1. Open in a new tab Overall study design and analytical workflow Acquisition of gene expression data Transcriptomic datasets ( GSE12644 , GSE51472 , GSE77287 , and GSE83453 ) comprised of 26 normal and 27 calcified aortic valve samples were retrieved from the GEO repository (Supplementary Table 1 ) [ 21 ]. Single-cell RNA-sequencing data (PRJNA562645; four CAVD vs. two controls) were similarly sourced from the GEO database. A curated set of 1,136 mitochondria-related genes (MRGs) was obtained from MitoCarta3.0 ( http://www.broadinstitute.org/mitocarta ), representing the most comprehensive inventory of mitochondrial-localized proteins to date (Supplementary Table 2 ) [ 22 ]. Data preprocessing, normalization, and batch effect correction To ensure robust integration across multiple datasets, all expression data were first normalized using established platform-appropriate strategies. For the four bulk transcriptomic datasets, raw counts were converted to transcripts per million (TPM) and subsequently log₂-transformed (log₂[TPM + 1]) to approximate a normal distribution and reduce variance heterogeneity across samples. For the single-cell RNA-seq data, expression values were normalized using the standard LogNormalizemethod in Seurat (scale factor = 10,000). Following normalization, rigorous batch effect correction was applied. For the bulk datasets, known batch effects attributable to dataset source (GEO accession) were corrected using the ComBatfunction from the sva package, followed by refinement with the remove batch effect function from the limma package [ 23 , 24 ]. For the single-cell data, cells from different samples were integrated using the harmony package, which effectively removes inter-sample technical variation while preserving biological heterogeneity [ 25 ]. Single-cell data quality control and identification of cell subpopulations Single-cell data processing and cell type annotation were performed using Seurat (v5.0.1). Cells were filtered with standard quality control thresholds: 200 < nFeature_RNA < 3,000 and mitochondrial gene fraction < 5%. After log-normalization, the top 50 principal components were used for clustering via the “FindNeighbors” and “FindClusters” functions with the resolution at 0.3 and visualization via UMAP. Cell types were annotated based on established marker genes. Differentially expressed genes (DEGs) in single-cell analysis for each cell cluster were identified using |log₂FC|≥ 0.25, detection rate ≥ 25% per cluster, and FDR‑adjusted P < 0.05. Scoring of MRGs DE-MRGs were identified by intersecting marker genes from each cell subpopulation within the set of MRGs. DE-MRGs were ranked in descending order based on their average log 2 fold change (avg_log 2 FC), and the top 20 were selected for co-expression analysis. Enrichment of the DE-MRG set in each individual cell was quantified using the AUCell algorithm, which calculated an area under the curve (AUC) score for each cell [ 26 ]. The “AUCell_exploreThresholds” function was employed to determine the activity threshold, and cells with AUC values exceeding this threshold were considered to have activated DE-MRG signatures. These MRG-activated cell clusters were subsequently visualized on the UMAP plot with distinct color labeling. Cell differentiation trajectory analysis Pseudotime analysis was primarily conducted using the Monocle package [ 27 ]. All valve interstitial and endothelial cell subpopulations were extracted, and the “orderCells” function was used to infer their differentiation states. The Branched Expression Analysis Modeling (BEAM) function was then applied to systematically examine the dynamic expression patterns of DE-MRGs across various stages of differentiation. Intercellular signaling network analysis Patterns of cell–cell communication were investigated at the molecular level using the CellChat package [ 28 ]. Ligand-receptor interactions were annotated based on the CellChatDB human database, and both the quantity and intensity of interactions between cell types were assessed and visualized. Identification of differentially expressed genes (DEGs) The limma algorithm [ 29 ] was used to identify DEGs between CAVD and normal valve tissues, with significant genes defined by a P-value of < 0.05 and |logFC|> 0.585. The results were visualized using volcano plots and heat maps. Functional enrichment analysis Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed to identify common driver genes using the clusterProfiler package [ 30 ]. Statistical significance for all enrichment results was set at P < 0.05. Machine learning-based identification of biomarkers Potential key targets for CAVD were identified using a combined approach of Least Absolute Shrinkage and Selection Operator (LASSO) regression and Support Vector Machine Recursive Feature Elimination (SVM-RFE). LASSO regression selected optimal feature variables and determined their best parameter by minimizing classification errors [ 31 ], while SVM-RFE applied a support vector machine with recursive feature elimination to refine the selection of critical variables [ 32 ]. The diagnostic performance of candidate genes in both the training sets ( GSE12644 , GSE51472 , and GSE77287 ) and the independent validation set ( GSE83453 ) was assessed using Receiver Operating Characteristic (ROC) curves generated via the pROC package, with AUC values used to evaluate their potential as diagnostic biomarkers. Immune cell infiltration analysis The relative abundances of 22 distinct immune cell types in tissue samples were estimated using the CIBERSORT algorithm [ 33 ], which is based on the specific gene expression profiles of immune cells, generating an immune cell infiltration matrix. Correlations between the expression levels of core biomarkers and degrees of immune infiltration were further examined (P < 0.05 was considered significant). In addition, single-gene Gene Set Enrichment Analysis (GSEA) was performed to explore key pathways regulated by the identified genes. Collection of human aortic valve specimens Calcified aortic valve tissues were obtained from patients with severe aortic stenosis undergoing valve replacement surgery at the Second Affiliated Hospital of Zhejiang University School of Medicine. Control (non-calcified) valve samples were collected from individuals undergoing Bentall procedures or surgical intervention for aortic regurgitation at the same hospital. Exclusion criteria included significant aortic regurgitation, rheumatic valve disease, infective endocarditis, congenital valvular malformation, and diabetes-related valvulopathy. Following explantation, a portion of each valve was processed for primary cell isolation, while the remainder was dehydrated in 30% sucrose and embedded in OCT compound for subsequent immunofluorescence analysis. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Second Affiliated Hospital, School of Medicine, Zhejiang University (No. 2024–0099). Written informed consent was obtained from all participants. Primary cell culture and osteogenic induction Primary human valve interstitial cells (VICs) were isolated from aortic valve leaflets following previously established protocols, with the purity of cell preparations verified as described earlier [ 34 ]. VICs obtained from normal valves between passages three and seven were selected for subsequent experiments. For osteogenic differentiation, cells were cultured in basal medium supplemented with 10 mmol/L β-glycerophosphate (β-GP), 10 nmol/L dexamethasone, and 0.25 mmol/L ascorbic acid (β-GA protocol), or induced with 2 mmol/L Na₂HPO₄, and maintained for 5 days. MGST1 knockdown in VICs VICs were transfected with MGST1-specific siRNA or non-targeting control siRNA (si-NC) using Lipofectamine RNAiMAX (Invitrogen) following the manufacturer’s protocol. After 12 h, cells were cultured in calcification medium for 5 days. Knockdown efficiency was confirmed by western blotting. Alkaline phosphatase (ALP) activity and runt-related transcription factor 2 (RUNX2) expression levels were measured as described above. Functional validation of MGST1 in VICs To investigate the functional role of MGST1 in calcification, both loss-of-function and gain-of-function studies were performed. For loss-of-function studies, VICs were transfected with MGST1-specific siRNA or non-targeting control siRNA (si-NC) using Lipofectamine RNAiMAX (Invitrogen) following the manufacturer’s protocol. After 12 h, cells were cultured in calcification medium for 5 days. For gain-of-function studies, VICs were transfected with a mammalian expression plasmid encoding full-length human MGST1 or an empty vector control using Lipofectamine 3000 (Invitrogen). After a 6-h recovery period, cells were similarly subjected to calcification induction. In both experimental setups, transfection efficiency was confirmed by western blotting, and calcification phenotypes were assessed by measuring ALP and RUNX2 expression as described above. Quantitative real-time PCR analysis Total RNA was extracted from VICs using TRIzol reagent (Sigma) according to the manufacturer’s protocol. Complementary DNA was synthesized with PrimeScript™ RT Master Mix (TaKaRa) following the supplier’s instructions. Quantitative real-time PCR was performed on a Roche LightCycler® 480II system using TB Green® Premix Ex Taq™ (TaKaRa). The following primer sequences were used: MGST1: forward, 5′-TAGAACGTGTACGCAGAGCC-3′; reverse, 5′-ATGGTGTGGTAGATCCGTGC-3′, and GAPDH: forward, 5′-ATCATCAGCAATGCCTCCTG-3′; reverse, 5′-GGCCATCCACAGTCTTCTG-3′. Western blot analysis Cells were collected and lysed on ice for 30 min using RIPA lysis buffer. Lysates were centrifuged at 16,000 × g for 10 min at 4 °C, and the resulting supernatant was collected. Protein concentration was determined using the BCA assay, and samples were diluted to a final concentration of 3 mg/mL. Five-fold sample loading buffer was added, and the mixture was boiled at 100 °C for 10 min. Proteins were separated by SDS-PAGE and subsequently transferred onto PVDF membranes, which were then blocked with 5% non-fat milk in TBST. Membranes were incubated with primary antibodies overnight at 4 °C, followed by washing and incubation with secondary antibodies at room temperature for 1 h. Protein bands were visualized using an enhanced chemiluminescence detection system. The following primary antibodies were used: MGST1 (Abcam, #ab131059, 1:1,000), ALP (R&D Systems, #MAB29092, 1:1,000), RUNX2 (Cell Signaling Technology, #8486S, 1:1,000), and GAPDH (Huabio, #ET1702-66, 1:10,000). Immunofluorescence staining Frozen sections of human aortic valve tissue were prepared for staining. Residual OCT compound was removed by rinsing with PBS, and sections were fixed in 4% paraformaldehyde at room temperature for 30 min. Membrane permeabilization was achieved with 0.2% Triton X-100 for 20 min, followed by blocking with 5% bovine serum albumin (BSA) for 30 min. Sections were incubated overnight at 4℃ with a primary antibody against MGST1 (Abcam, #ab131059, 1:50). After washing with PBS, samples were incubated with appropriate fluorescent secondary antibodies at room temperature for 60 min in the dark. Nuclei were counterstained with DAPI. Fluorescence images were acquired using a ZEISS LSM980 with Airyscan2, and quantitative analysis was performed with ImageJ software (NIH). Drug prediction and molecular docking analyses of MGST1 Drugs associated with MGST1 were retrieved from the DrugBank database ( https://go.drugbank.com ). Molecular docking experiments were performed to examine the binding modes and affinities between candidate drugs and key target proteins. The crystal structure of MGST1 was predicted using AlphaFold. The three-dimensional structures of the screened drug compounds were obtained from the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ) . The structures of MGST1 and drug compounds were preprocessed using AutoDockTools software, including the addition of hydrogen atoms and removal of water molecules [ 35 ]. Subsequently, AutoDock Vina software was employed for molecular docking to calculate the binding energy between the target protein and drug compounds (kcal/mol). Finally, PyMOL software was used to visualize the docking results and generate detailed binding mode diagrams. Ritlecitinib treatment in VICs To experimentally validate the top candidate compound ritlecitinib predicted by molecular docking to target MGST1, an in vitro intervention study was conducted. VICs were seeded in culture plates and pre-cultured until 70–80% confluence. Cells were treated with ritlecitinib or vehicle control (DMSO), and the treatment was maintained throughout the subsequent calcification induction. Following treatment, total protein was extracted using RIPA lysis buffer. Protein concentration was determined by BCA assay, and the expression levels of MGST1 and calcification markers (RUNX2, ALP) were analyzed by western blotting as described above. Statistical analysis Statistical analyses were primarily conducted using R Studio (v4.3.2). For comparisons between groups of continuous variables with normal distribution, independent samples t-tests were employed. The Wilcoxon rank-sum test was used to analyze non-normally distributed data. A P-value of < 0.05 was considered statistically significant. Results Overall study design The overall workflow of this study is summarized in Fig. 1 . A detailed description of each analytical and experimental step is provided in the Methods section. Single-cell atlas of aortic valve tissues in control and CAVD groups Single-cell RNA-sequencing data was obtained from the GEO database. After stringent quality control, 12,390 high-quality cells were retained for further analysis, including 3,094 (25%) cells from the control group and 9,296 (75%) cells from the CAVD group (Fig. 2 b). Using the UMAP technique to visualize high-dimensional single-cell RNA sequencing data, cells were classified into five clusters (Fig. 2 a), which were annotated as VDSCs, VICs, valve endothelial cells (VECs), lymphocytes, and macrophages. Among them, VDSCs had the highest proportion among the six samples (Fig. 2 c ) . Next, we validated the cell subgroups using classical cellular markers (Fig. 2 d). The top five significant DEGs in each cell cluster are shown as a heat map (Fig. 2 e). Fig. 2. Open in a new tab Single-cell dimensionality reduction clustering and characterization. a) UMAP projection of cell clusters. b) Proportion of cells per sample relative to the total cell count. c) Sample-wise distribution of cell clusters. d) Cluster identity validation based on canonical marker gene expression. e) Heat map of DEGs across cell subpopulations Identification and functional profiling of DE-MRGs across cell subpopulations To investigate the potential associations between MRGs and specific cellular subpopulations, we intersected 2,191 differentially expressed cell marker genes (log 2 FC > 0.25, adjusted P < 0.05) with a curated list of 1,136 MRGs, and identified 200 DE-MRGs ( Fig. 3 a , Supplementary Table 3 ). The top 20 DE-MRGs included C15orf48, ACSL1, OCIAD2, STOM, LACTB, IFI27, MRPS6, SLC25A37, PPIF, ACOT9, BCL2A1, PMAIP1, FTH1, HIGD2A, MPST, COX17, PISD, BAX, MRPL3, and HTRA2, with notable correlations such as the positive correlation between FTH1 and C15orf48 (R = 0.37, P < 0.001) and SLC25A37 (R = 0.35, P < 0.001) ( Fig. 3 c , Supplementary Table 4 ). Among these, COX17 was broadly distributed across all cell types, OCIAD2 and STOM were enriched in VECs, C15orf48 and ACSL1 were predominantly expressed in macrophages, and MRPL3 was highly expressed in VDSCs ( Fig. 3 b , Supplementary Fig. 1 ). AUCell analysis was used to assess DE-MRG functional activity, and cells with enrichment scores > 0.203 were defined as “MRGs high-scoring cells.” A total of 3,109 high-scoring cells were identified. These cells were mainly distributed in the VIC, VDSC, and VEC subgroups ( Fig. 3 d–f ). DEGs within each high-scoring subpopulation were visualized via volcano plots ( Fig. 3 g ). KEGG pathway enrichment analysis (log 2 FC > 0.25, adjusted P < 0.05) indicated that these DE-MRGs were mainly involved in the PI3K-AKT, TNF, and AGE-RAGE signaling pathways ( Fig. 3 h ). Fig. 3. Open in a new tab Identification and characterization of DE-MRGs. a) Identification of DE-MRGs. b) Cell-type specific distribution of COX17, OCIAD2, STOM, C15orf48, ACSL1, and MRPL3. c) Correlation matrix among the top 20 DE-MRGs. d) Histogram showing AUC scores for DE-MRGs. e–f) Distribution of MRGs high-scoring cells across different subpopulations. g) Volcano plots of DE-MRGs for each subpopulation. h) KEGG pathway enrichment analysis of DE-MRGs within the high-scoring cell subpopulation Differentiation trajectory and intercellular communication of MRGs high-scoring cells Therefore, to determine whether MRGs are involved in key cellular programs underlying CAVD as hypothesized, we first characterized DE-MRGs across major cell types and along the EndMT trajectory. First, we isolated MRGs high-scoring VECs and VICs (Fig. 4 a) and subjected them to pseudotime trajectory analysis to simulate their differentiation states (Fig. 4 b ) . The trajectory map revealed five distinct states, with darker blue denoting earlier differentiation stages. VECs were predominantly located at the early stages of differentiation, whereas VICs appeared at later stages, indicating a progressive transition from VECs to VICs consistent with the involvement of EndMT in CAVD. BEAM analysis identified 18 DE-MRGs between two key nodes along the trajectory, with the pre-branch representing the root state, fate 1 corresponding to VECs, and fate 2 to VICs. Based on their expression dynamics across cell populations, these DE-MRGs were classified into six distinct clusters, reflecting coordinated transcriptional programs during pseudotime differentiation (Fig. 4 c ) . Fig. 4. Open in a new tab Cell-to-cell communication analysis of MRGs high-scoring cells. a) UMAP plot showing high-scoring VECs and VICs based on MRG expression. b) Pseudotime trajectory analysis of cellular differentiation. c) Dynamic heat map showing DE-MRG expression on both sides of the branch nodes. d) Network analysis of cell–cell interactions among high-scoring populations. e) Ligand-receptor interaction networks between VICs and other cell types. f) Collagen signaling pathway network analysis Consistent with the premise that mitochondrial dysfunction may orchestrate immune–stromal crosstalk beyond cell differentiation, we then investigated cell-to-cell communication in MRGs high-scoring cells (Fig. 4 d), and found that VIC–macrophage interactions were the most frequent and intensive, suggesting an active immune–stromal interface in CAVD. Ligand–receptor network analysis highlighted significant MIF-(CD74 + CXCR4) and MIF-(CD74 + CD44) signaling specifically between VICs and macrophages (Fig. 4 e). In addition, multiple collagen family members were repeated identified in VIC-related networks, implying a pivotal role of collagen-mediated signaling. Collagen pathway analysis demonstrated that VICs occupied a central hub position, with significantly higher roles as “Mediator,” “Influencer,” and “Sender” compared to other cell types ( Fig. 4 f ), highlighting their importance in matrix remodeling and tissue homeostasis within CAVD. Integrated transcriptomic analysis, key target identification, and immune-related characterization of MGST1 To validate the significance of MRGs in CAVD, we performed differential expression analysis on three transcriptome datasets, and identified 477 DEGs, including 305 upregulated and 172 downregulated genes ( Fig. 5 a–b ) . Among the 200 DE-MRGs previously identified, four genes—MGST1, GPX1, C15orf48, and BCL2A1—were significantly differentially expressed ( Fig. 5 c ) . Functional enrichment of the 477 DEGs revealed that GO biological processes were predominantly associated with inflammation (e.g. leukocyte migration and chemokine response), while KEGG pathways highlighted NF-κB signaling, cytokine–cytokine receptor interaction, PI3K–AKT signaling, and ECM–receptor interaction, indicating the involvement of inflammatory and ECM remodeling mechanisms (Supplementary Fig. 2a–b ). Fig. 5. Open in a new tab Identification of key genes. a) Heat map of DEGs. b) Volcano plot of DEGs. c) Venn diagram showing the overlap between DE-MRGs and DEGs identified by transcriptomic analysis. d) Selection of key genes using LASSO regression and SVM-RFE. e) Overlap of DEGs identified by LASSO regression and SVM-RFE. f) ROC curves for the three most common key genes in the test datasets. g) ROC curve for MGST1 in the validation dataset. h) Single-gene GSEA enrichment of MGST1-associated GO terms. i) Single-gene GSEA enrichment of MGST1-associated KEGG pathways To pinpoint core molecular targets, LASSO regression and SVM-RFE analyses were conducted, with both consistently identifying MGST1, GPX1, and C15orf48 as key genes ( Fig. 5 d–e ) . ROC curve analysis across the GSE51472 , GSE12644 , and GSE77287 datasets revealed that MGST1 achieved the highest average diagnostic performance (AUC = 0.818) ( Fig. 5 f ) . In the independent validation dataset GSE83453 , MGST1 showed excellent diagnostic accuracy (AUC = 0.986) ( Fig. 5 g ) , highlighting its potential as a biomarker for CAVD. Given the pronounced inflammatory signatures, we further investigated immune infiltration patterns (Supplementary Fig. 2 c–d), and found significantly elevated proportions of M0 and M1 macrophages in CAVD. Single-gene GSEA indicated that MGST1 was enriched in immune-related processes (e.g. immune activation, inflammatory response, and leukocyte-mediated immunity) ( Fig. 5 h ) , while KEGG pathway enrichment analysis linked MGST1 to adipocytokine signaling, cardiac muscle contraction, and chemokine signaling pathways ( Fig. 5 i ) . Moreover, MGST1 expression correlated positively with T follicular helper (Tfh) cells and M0 macrophages(Supplementary Fig. 2 e), suggesting that MGST1 may be involved in macrophage-driven immune modulation in CAVD. Experimental validation and drug interaction analysis of MGST1 Finally, to systematically validate MGST1 as a functional driver and therapeutic target in CAVD, we characterized its expression dynamics, biological function, and druggability. After 5 days of induction, the expression of calcification markers ALP and RUNX2 was significantly upregulated, and was accompanied by marked increases in both MGST1 mRNA and protein levels under calcifying conditions ( Fig. 6 a–b ) . Immunofluorescence staining of human aortic valves further confirmed a pronounced elevation of MGST1 expression in calcified tissues compared to normal controls (Fig. 6 c–d ) . Fig. 6. Open in a new tab Validation of MGST1. a) MGST1, ALP, and RUNX2 protein expression levels were assessed in primary VICs cultured in control medium (CON) and osteogenic medium (OM). ALP and RUNX2 served as markers of calcification. b) MGST1 mRNA expression levels were detected in VICs cultured in CON and OM. c–d) Immunofluorescence staining of human calcified and normal aortic valves. Nuclei are stained with DAPI (blue) and MGST1 expression is shown in red, n = 6. e) Effects of MGST1 knockdown on calcification markers in VICs under calcifying conditions. siRNA-mediated suppression of MGST1 significantly decreased ALP activity and reduced RUNX2 mRNA and protein expression levels compared with controls. f) The binding interactions of MGST1 with ritlecitinib and glutathione disulfide. *P < 0.05, **P < 0.01, ***P < 0.001 To determine whether MGST1 actively drives calcification, we performed gain- and loss-of-function experiments. Notably, overexpression of MGST1 robustly elevated, while its knockdown significantly reduced, the protein levels of key calcification markers ALP and RUNX2 in VICs under calcifying conditions (Fig. 6 e , Supplementary Fig. 4 a). These opposing phenotypes establish MGST1 not merely as an associated marker but as a functional regulator capable of modulating the calcification process. Based on this confirmed function, DrugBank-based prediction identified ritlecitinib, glutathione, and glutathione disulfide as potential compounds targeting MGST1. Molecular docking analysis revealed that ritlecitinib showed the strongest binding affinity (− 6.4 kcal/mol) through hydrogen bonds with ARG-147 and SER-96 (Fig. 6 f). Glutathione disulfide formed hydrogen bonds with TYR-121, ASN-78, ARG-74, and ARG-71 (Fig. 6 f ) , exhibiting a binding energy of − 5.3 kcal/mol, while glutathione interacted with TYR-121, HIS-117, and ASN-78 with a binding energy of − 4.6 kcal/mol (Supplementary Fig. 3 ). The lower binding energy of the ritlecitinib–MGST1 complex suggested a stable interaction, supporting its potential as a therapeutic candidate. To validate this prediction functionally, we treated calcifying VICs with ritlecitinib. Consistent with the docking results, ritlecitinib treatment not only downregulated MGST1 protein expression but also reduced the levels of the calcification markers ALP and RUNX2 (Supplementary Fig. 4 b). These results confirm the docking prediction and demonstrate the efficacy of ritlecitinib in attenuating the calcification pathway. Discussion CAVD exhibits a marked age-dependent progression in incidence. Due to its complex and poorly understood pathogenesis, limited interventions currently available to halt or reverse the progression of CAVD. Mitochondria, as the principal organelles responsible for aerobic respiration, are essential for maintaining normal cellular physiological functions. Under physiological conditions, mitochondrial quality control mechanisms including mitochondrial dynamics, function, and metabolism are maintained in a dynamic equilibrium. Increasing evidence suggests that mitochondria play a pivotal role in the development of CAVD. However, the precise molecular pathways underlying mitochondrial alterations in CAVD remain poorly defined. Therefore, comprehensive investigations into the pathogenic mechanisms of mitochondrial dysfunction in CAVD, as well as identification of potential therapeutic targets, are crucial for slowing disease progression. Furthermore, understanding how genetic predispositions (e.g. in NOTCH1 or ECM-related genes) intersect with mitochondrial dysfunction could offer new insights into the heterogeneity of CAVD progression. Several previous studies have applied bulk or single‑cell omics approaches to characterize the molecular landscape of CAVD and have identified inflammatory, ECM, and osteogenic signaling pathways as central contributors to disease progression [ 36 – 38 ]. Mitochondrial dysfunction has also been implicated in CAVD, but in most omics‑based studies it has appeared as one element within broader gene sets rather than as an explicit primary focus, and mitochondria‑related transcriptional programs have not been systematically delineated across distinct valve cell populations [ 39 , 40 ]. In parallel, prior mitochondrial research in CAVD has largely relied on targeted assays in VICs or animal models, providing important mechanistic insights but offering less resolution of cell type‑specific and trajectory‑dependent changes at the transcriptomic level. In this context, our study was designed to specifically interrogate MRGs across the single‑cell landscape of human CAVD valves, integrating pseudotime analysis, cell–cell communication, and machine‑learning‑based prioritization to relate mitochondrial gene dysregulation to EndMT, immune–stromal crosstalk, and calcific remodeling. This mitochondria‑focused, multi‑level framework is intended to complement and extend existing omics and experimental studies by highlighting mitochondrial signaling as a potential unifying axis connecting diverse pathogenic processes in CAVD. Integrated analysis of single-cell transcriptomic data revealed heterogeneous expression patterns of MRGs across distinct cell types in CAVD. For example, OCIAD2 and STOM were found to be specifically enriched in VECs, while C15orf48 and ACSL1 were predominantly found in macrophages, suggesting that these MRGs may exert distinct functions in different cell populations to influence calcification. Previous studies have also demonstrated that mitochondrial dysfunction in VICs can drive oxidative stress and promote calcification [ 12 ], while other studies have reported that macrophages may contribute to a pro-calcific environment through heightened mitophagy and the release of inflammatory mediators [ 41 , 42 ]. MRGs have previously been shown to play significant roles in osteogenesis and calcification [ 43 ]. Notably, DEGs in cells with high MRG expression have been found to be significantly enriched in signaling pathways such as the PI3K-AKT [ 44 ], TNF [ 45 ] and AGE-RAGE [ 44 ] signaling pathways, all of which have been implicated in promoting a pro-calcific microenvironment. Similarly, the dynamic regulation of MRG expression is evident during EndMT, particularly genes involved in apoptosis and mitochondrial fission, which may drive the transition of cells towards profibrotic and pro-calcific phenotypes [ 46 ]. Interestingly, recent studies have highlighted the critical role of ECM remodeling in valve pathology. For instance, genetic defects in ECM components have been linked to impaired mechanosensing and mitochondrial dysfunction in connective tissue disorders [ 47 ]. Our finding that VICs serve as the primary hub for collagen signaling aligns with recent evidence suggesting that aberrant ECM secretion acts as a driver for fibro-calcific remodeling [ 48 ]. This suggests a potential feedback loop where mitochondrial dysfunction in VICs not only alters cellular metabolism but also disrupts the ECM niche, further propagating calcification signals. MGST1 is a microsomal glutathione S‑transferase localized primarily to the outer mitochondrial membrane and endoplasmic reticulum membrane, where it catalyzes the conjugation of glutathione to electrophilic substrates and helps detoxify reactive oxygen species by stabilizing the glutathione thiolate anion [ 17 , 49 , 50 ]. Emerging evidence suggests that the function of MGST varies across different disease states. For example, MGST1 has been shown to mitigate ischemia–reperfusion injury in cardiomyocytes [ 19 ]; while MGST1 expression levels in cardiac fibroblasts were found to decrease progressively with advancing heart failure [ 20 ]. In addition, MGST1 has been implicated in promoting chemoresistance in certain tumor types, including melanoma and pancreatic ductal adenocarcinoma [ 49 , 51 ]. Here, we showed that MGST1 was significantly upregulated in VICs, suggesting a possible role in driving pathological remodeling in CAVD. First, MGST1 may exert a pro-fibrotic effect, because activated VICs have been shown to contribute to a pro-calcific microenvironment through excessive collagen secretion [ 52 ]. Our findings suggest that VICs are the primary effector cells for collagen signaling in CAVD valves. Moreover, MGST1 may participate in regulation of the immune microenvironment. Immune infiltration analysis revealed a positive correlation between MGST1 expression and levels of Tfh cells and M0 macrophages, suggesting a potential role for metabolic-immune crosstalk in modulating VIC calcification. Thus, the high expression of MGST1 in VICs under CAVD pathological conditions, together with its impact on calcification markers in vitro, supports MGST1 as a promising candidate for therapeutic intervention, warranting future confirmation in in vivo models. Our study provided a mitochondrial perspective on the pathogenesis of CAVD. By integrating MRG expression profiling, we identified DE-MRGs and their cell-type specific distribution patterns within valve cell subpopulations. Through a combination of trajectory and cell–cell communication analyses, we demonstrated that dynamic regulation of MRGs orchestrated phenotypic transitions during EndMT and elucidated the molecular interplay between VICs and immune cells in driving calcification. An independent, orthogonal validation using a murine CAVD model yielded concordant results (Supplementary Fig. 5 ). Utilizing machine learning algorithms and ROC curve analysis, we further pinpointed MGST1 as a key regulatory gene associated with mitochondrial dysfunction in CAVD. For the first time, our study systematically characterized the dynamic expression of MRGs in CAVD and their involvement in cell phenotype modulation, as well as identified MGST1 as a potential biomarker of mitochondrial dysfunction. Notably, through DrugBank screening and molecular docking, we identified ritlecitinib as a top candidate targeting MGST1. Experimental validation confirmed that ritlecitinib not only downregulates MGST1 expression but also reduces key calcification markers in VICs, demonstrating its functional efficacy in mitigating the pro-calcific phenotype. These findings thus move beyond computational prediction to provide preliminary experimental evidence supporting MGST1 inhibition as a viable strategy. Overall, our findings provide a solid theoretical foundation for understanding disease heterogeneity and for developing precision therapies targeting mitochondrial pathways in CAVD. However, several limitations of this study should be noted. First, the overall sample size was relatively small—particularly in the control group—which may increase individual heterogeneity-related variability. Second, the lack of integration between our sequencing data and spatial transcriptomics or subcellular imaging limits precision in mapping the spatial distribution of DE-MRGs within valve architecture. Third, the subcellular localization of MGST1 in CAVD requires further experimental validation. Fourth, functional studies investigating MGST1 localization and its involvement in mitochondrial dynamics under calcifying conditions are necessary to elucidate its mechanism of action. Finally, the clinical relevance of MGST1 remains to be explored. Specifically, its potential as a circulating biomarker was not assessed in this study. Future research should validate MGST1 protein expression in larger clinical cohorts using ELISA or immunohistochemistry, evaluate its utility as a serum biomarker, and examine its correlation with echocardiographic parameters such as valve calcification scores. In summary, this study is the first to systematically elucidate the role of MRGs in CAVD and identify MGST1 as a regulatory gene with biomarker potential. These findings provide a novel theoretical framework for understanding disease heterogeneity and for developing precision therapies targeting mitochondrial dysfunction in CAVD. Conclusion This study is the first to systematically describe the regulatory roles of mitochondria-related molecules in the valve cells of CAVD patients. Notably, we highlight the pivotal role of the mitochondria-related gene MGST1 in the initiation and progression of valve calcification, as well as provide preliminary evidence supporting its feasibility as a potential therapeutic target. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 49733 KB) (48.6MB, pdf) Supplementary file2 (PDF 4522 KB) (4.4MB, pdf) Supplementary file3 (PDF 5492 KB) (5.4MB, pdf) Supplementary file4 (PDF 2700 KB) (2.6MB, pdf) Supplementary file5 (PDF 2979 KB) (2.9MB, pdf) Supplementary file6 (XLSX 10 KB) (9.8KB, xlsx) Supplementary file7 (XLSX 23 KB) (23.1KB, xlsx) Supplementary file8 (XLSX 12 KB) (11.8KB, xlsx) Supplementary file9 (XLSX 18 KB) (17.6KB, xlsx) Acknowledgements We thank the GEO database for providing valuable public datasets used in this study, and International Science Editing ( http://www.internationalscienceediting.com ) for editing this manuscript. Authors’ contributions All authors made a significant contribution to the work reported, whether that was in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Funding This research was funded by financial support from the National Key Research and Development Program of China (Project 2023YFA1800700 and 2022YFA1105200), the National Natural Science Foundation of China (NSFC, Project 81773012), and the Key Research and Development Program of Zhejiang Province (No.2025C02144). Data availability No datasets were generated or analysed during the current study. Declarations Ethical approval Cardiac valve tissue collection was approved by the Ethics Committee of the Second Affiliated Hospital, School of Medicine, Zhejiang University (No. 2024–0099) and written informed consent was obtained from all donors. Clinical trial number Not applicable. Data sharing statement The public datasets analyzed in this study are available in the GEO repository ( https://www.ncbi.nlm.nih.gov/ ). Disclosure There are no conflicts of interest to be disclosed by all authors. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Ziling Mai and Huijun Du contributed equally to this work. Contributor Information Bingchen Liu, Email: [email protected]. Deqiang Sun, Email: [email protected]. References 1. Coffey S, Cox B, Williams MJ (2014) The prevalence, incidence, progression, and risks of aortic valve sclerosis: a systematic review and meta-analysis. J Am Coll Cardiol 63:2852–2861 [ DOI ] [ PubMed ] [ Google Scholar ] 2. Towler DA (2013) Molecular and cellular aspects of calcific aortic valve disease. Circ Res 113:198–208 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, Enriquez-Sarano M (2006) Burden of valvular heart diseases: a population-based study. Lancet 368:1005–1011 [ DOI ] [ PubMed ] [ Google Scholar ] 4. García-Rodríguez C, Parra-Izquierdo I, Castaños-Mollor I, López J, San Román JA, Sánchez Crespo M (2018) Toll-like receptors, inflammation, and calcific aortic valve disease. Front Physiol 9:201 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Cho KI, Sakuma I, Sohn IS, Jo SH, Koh KK (2018) Inflammatory and metabolic mechanisms underlying the calcific aortic valve disease. Atherosclerosis 277:60–65 [ DOI ] [ PubMed ] [ Google Scholar ] 6. Tsimikas S, Fazio S, Ferdinand KC, Ginsberg HN, Koschinsky ML, Marcovina SM, Moriarty PM, Rader DJ, Remaley AT, Reyes-Soffer G et al (2018) NHLBI working group recommendations to reduce lipoprotein(a)-mediated risk of cardiovascular disease and aortic stenosis. J Am Coll Cardiol 71:177–192 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Farrar EJ, Huntley GD, Butcher J (2015) Endothelial-derived oxidative stress drives myofibroblastic activation and calcification of the aortic valve. PLoS One 10:e0123257 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Pedriali G, Morciano G, Patergnani S, Cimaglia P, Morelli C, Mikus E, Ferrari R, Gasbarro V, Giorgi C, Wieckowski MR et al (2020) Aortic valve stenosis and mitochondrial dysfunctions: clinical and molecular perspectives. Int J Mol Sci 21:4899 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Rajamannan NM, Subramaniam M, Rickard D, Stock SR, Donovan J, Springett M, Orszulak T, Fullerton DA, Tajik AJ, Bonow RO et al (2003) Human aortic valve calcification is associated with an osteoblast phenotype. Circulation 107:2181–2184 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Qian X, Xu L, Geng B, Li F, Dong N (2025) Navigating the landscape of translational medicine of calcific aortic valve disease: bridging bench to bedside. JACC Asia 5:503–515 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Liu C, Liu R, Cao Z, Guo Q, Huang H, Liu L, Xiao Y, Duan C, Ma R (2022) Identification of MMP9 as a novel biomarker to mitochondrial metabolism disorder and oxidative stress in calcific aortic valve stenosis. Oxid Med Cell Longev 2022:3858871 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [ Retracted ] 12. Morciano G, Patergnani S, Pedriali G, Cimaglia P, Mikus E, Calvi S, Albertini A, Giorgi C, Campo G, Ferrari R et al (2022) Impairment of mitophagy and autophagy accompanies calcific aortic valve stenosis favouring cell death and the severity of disease. Cardiovasc Res 118:2548–2559 [ DOI ] [ PubMed ] [ Google Scholar ] 13. Zhu Y, Zhang JL, Yan XJ, Ji Y, Wang FF (2023) Exploring a new mechanism between lactate and VSMC calcification: PARP1/POLG/UCP2 signaling pathway and imbalance of mitochondrial homeostasis. Cell Death Dis 14:598 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Su ZD, Li CQ, Wang HW, Zheng MM, Chen QW (2023) Inhibition of DRP1-dependent mitochondrial fission by Mdivi-1 alleviates atherosclerosis through the modulation of M1 polarization. J Transl Med 21:427 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Yu L, Su Z, Tian D, Liu S, Zhang L, Wang Z, Guo S, Zhu W, Wang P, Zhang N (2025) Piezo1 induces mitochondrial autophagy dysfunction leading to cartilage injury in knee osteoarthritis. Mol Med 31:272 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Schaefer PM, Scherer Alves L, Lvova M, Huang J, Rathi K, Janssen K, Butic A, Yardeni T, Morrow R, Lott M et al (2022) Combination of common mtDNA variants results in mitochondrial dysfunction and a connective tissue dysregulation. Proc Natl Acad Sci U S A 119:e2212417119 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Cebula M, Morgenstern R (2023) Enzymology of reactive intermediate protection: kinetic analysis and temperature dependence of the mesophilic membrane protein catalyst MGST1. FEBS J 290:3448–3460 [ DOI ] [ PubMed ] [ Google Scholar ] 18. Kuang F, Liu J, Xie Y, Tang D, Kang R (2021) MGST1 is a redox-sensitive repressor of ferroptosis in pancreatic cancer cells. Cell Chem Biol 28:765–75.e765 [ DOI ] [ PubMed ] [ Google Scholar ] 19. Yang Y, Zhao C, Li C, Lu Z, Cao X, Wu Q (2025) MGST1 overexpression ameliorates mitochondrial dysfunction and ferroptosis during myocardial ischemia/reperfusion injury after heart transplantation. Int J Biol Macromol 299:140135 [ DOI ] [ PubMed ] [ Google Scholar ] 20. Wang X, Mu X, Li X, Yang C, Cai Y, Liu C, Liu Z, He Z (2025) Construction of a deep learning model and identification of the pivotal characteristics of FGF7- and MGST1- positive fibroblasts in heart failure post-myocardial infarction. Int J Biol Macromol 310:143171 [ DOI ] [ PubMed ] [ Google Scholar ] 21. Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, Marshall KA, Phillippy KH, Sherman PM, Holko M et al (2013) NCBI GEO: archive for functional genomics data sets—update. Nucleic Acids Res 41:D991–D995 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Rath S, Sharma R, Gupta R, Ast T, Chan C, Durham TJ, Goodman RP, Grabarek Z, Haas ME, Hung WHW et al (2021) Mitocarta3.0: an updated mitochondrial proteome now with sub-organelle localization and pathway annotations. Nucleic Acids Res 49:D1541–D1547 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Johnson WE, Li C, Rabinovic A (2007) Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 8:118–127 [ DOI ] [ PubMed ] [ Google Scholar ] 24. Chen C, Grennan K, Badner J, Zhang D, Gershon E, Jin L, Liu C (2011) Removing batch effects in analysis of expression microarray data: an evaluation of six batch adjustment methods. PLoS One 6:e17238 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Korsunsky I, Millard N, Fan J, Slowikowski K, Zhang F, Wei K, Baglaenko Y, Brenner M, Loh PR, Raychaudhuri S (2019) Fast, sensitive and accurate integration of single-cell data with harmony. Nat Methods 16:1289–1296 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Aibar S, González-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, Hulselmans G, Rambow F, Marine JC, Geurts P, Aerts J et al (2017) SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14:1083–1086 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Qiu X, Mao Q, Tang Y, Wang L, Chawla R, Pliner HA, Trapnell C (2017) Reversed graph embedding resolves complex single-cell trajectories. Nat Methods 14:979–982 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q (2021) Inference and analysis of cell-cell communication using CellChat. Nat Commun 12:1088 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Diboun I, Wernisch L, Orengo CA, Koltzenburg M (2006) Microarray analysis after RNA amplification can detect pronounced differences in gene expression using limma. BMC Genomics 7:252 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Yu G, Wang LG, Han Y, He QY (2012) ClusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16:284–287 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Yang C, Delcher C, Shenkman E, Ranka S (2018) Machine learning approaches for predicting high cost high need patient expenditures in health care. Biomed Eng Online 17:131 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Sanz H, Valim C, Vegas E, Oller JM, Reverter F (2018) SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics 19:432 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 33. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, Hoang CD, Diehn M, Alizadeh AA (2015) Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 12:453–457 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Chen J, Ren T, Xie L, Hu H, Li X, Maitusong M, Zhou X, Hu W, Xu D, Qian Y et al (2024) Enhancing aortic valve drug delivery with PAR2-targeting magnetic nano-cargoes for calcification alleviation. Nat Commun 15:557 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Eberhardt J, Santos-Martins D, Tillack AF, Forli S (2021) Autodock Vina 1.2.0: new docking methods, expanded force field, and python bindings. J Chem Inf Model 61:3891–3898 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Li Q, Song C, Wei Z, Zhou H, Wang S, Li H, Yang H, Luo Q, Li J, Chen M (2025) Age-associated methionine sulfoxide reductase A protects against valvular interstitial cell senescence and valvular calcification. Geroscience. 10.1007/s11357-025-01675-w [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Yang R, Han C, Xie Y, Qiu S, Zhang S, He J, Wang Z, Zhang Z, Liu H, Chen L et al (2025) The pathogenic roles of local vitamin D metabolism defect in valve inflammation and calcification. Adv Sci 12:e01250 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Zhang H, Yan W, Ling J, Shi X, Lai P, Wang B, Liao Y, Zhang H (2026) Silencing CCL5 suppresses ferroptosis to alleviate calcific aortic valve disease through chemokine pathway inhibition. Atherosclerosis 414:120640 [ DOI ] [ PubMed ] [ Google Scholar ] 39. Song N, Ji E, Yu JE, Choi KH, Kim DH, Song JM, Kang DH, Song JK, Yu J, Kim K et al (2025) Spermidine enhances mitochondrial function and mitigates aortic valve calcification: implications for DNA methyltransferase-1 activity. JACC Basic Transl Sci 10:345–366 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Liu F, Chen J, Hu W, Gao C, Zeng Z, Cheng S, Yu K, Qian Y, Xu D, Zhu G et al (2022) PTP1B Inhibition improves mitochondrial dynamics to alleviate calcific aortic valve disease via regulating OPA1 homeostasis. JACC Basic Transl Sci 7:697–712 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Wang K, Zheng Q, Liu X, Geng B, Dong N, Shi J (2022) Identifying hub genes of calcific aortic valve disease and revealing the immune infiltration landscape based on multiple WGCNA and single-cell sequence analysis. Front Immunol 13:1035285 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Tanase DM, Valasciuc E, Gosav EM, Floria M, Costea CF, Dima N, Tudorancea I, Maranduca MA, Serban IL (2022) Contribution of oxidative stress (OS) in calcific aortic valve disease (CAVD): from pathophysiology to therapeutic targets. Cells 11:2663 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Voicu G, Mocanu CA, Safciuc F, Rebleanu D, Anghelache M, Cecoltan S, Droc I, Simionescu M, Manduteanu I, Calin M (2024) VCAM-1 targeted nanocarriers of shRNA-Smad3 mitigate endothelial-to-mesenchymal transition triggered by high glucose concentrations and osteogenic factors in valvular endothelial cells. Int J Biol Macromol 281:136355 [ DOI ] [ PubMed ] [ Google Scholar ] 44. Adhikari R, Jung J, Shiwakoti S, Park EY, Kim HJ, Ko JY, You J, Lee M, Oak MH (2023) Capsaicin inhibits aortic valvular interstitial cell calcification via the redox-sensitive NFκB/AKT/ERK1/2 pathway. Biochem Pharmacol 212:115530 [ DOI ] [ PubMed ] [ Google Scholar ] 45. Sánchez-Bayuela T, Peral-Rodrigo M, Parra-Izquierdo I, López J, Gómez C, Montero O, Pérez-Riesgo E, San Román JA, Butcher JT, Sánchez Crespo M et al (2025) Inflammation via JAK-STAT/HIF-1α drives metabolic changes in pentose phosphate pathway and glycolysis that support aortic valve cell calcification. Arterioscler Thromb Vasc Biol 45:e232–e249 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Tu P, Pan Y, Wang L, Li B, Sun X, Liang Z, Liu M, Zhao Z, Wu C, Wang J et al (2024) CD62E- and ROS-responsive ETS improves cartilage repair by inhibiting endothelial cell activation through OPA1-mediated mitochondrial homeostasis. Biomater Res 28:0006 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Marcos-Ríos D, Rochano-Ortiz A, Méndez-Barbero N, Oller J (2025) Defective mitochondrial respiration in hereditary thoracic aneurysms. Cells 14:768 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Zhang H, Tsui CK, Garcia G, Joe LK, Wu H, Maruichi A, Fan W, Pandovski S, Yoon PH, Webster BM et al (2024) The extracellular matrix integrates mitochondrial homeostasis. Cell 187:4289–304.e4226 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 49. Zhang J, Ye ZW, Morgenstern R, Townsend DM, Tew KD (2023) Microsomal glutathione transferase 1 in cancer and the regulation of ferroptosis. Adv Cancer Res 160:107–132 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Dodson M, Anandhan A, Zhang DD (2021) MGST1, a new soldier of NRF2 in the battle against ferroptotic death. Cell Chem Biol 28:741–742 [ DOI ] [ PubMed ] [ Google Scholar ] 51. Zhang J, Ye ZW, Chakraborty P, Luo Z, Culpepper J, Aslam M, Zhang L, Johansson K, Haeggström JZ, Xu J et al (2023) Microsomal glutathione transferase 1 controls metastasis and therapeutic response in melanoma. Pharmacol Res 196:106899 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Ballester-Servera C, Alonso J, Cañes L, Vázquez-Sufuentes P, Puertas-Umbert L, Fernández-Celis A, Taurón M, Rodríguez-Sinovas A, López-Andrés N, Rodríguez C, et al. (2023) Lysyl oxidase-dependent extracellular matrix crosslinking modulates calcification in atherosclerosis and aortic valve disease. Biomed Pharmacother 167:115469. [ DOI ] [ PubMed ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary file1 (PDF 49733 KB) (48.6MB, pdf) Supplementary file2 (PDF 4522 KB) (4.4MB, pdf) Supplementary file3 (PDF 5492 KB) (5.4MB, pdf) Supplementary file4 (PDF 2700 KB) (2.6MB, pdf) Supplementary file5 (PDF 2979 KB) (2.9MB, pdf) Supplementary file6 (XLSX 10 KB) (9.8KB, xlsx) Supplementary file7 (XLSX 23 KB) (23.1KB, xlsx) Supplementary file8 (XLSX 12 KB) (11.8KB, xlsx) Supplementary file9 (XLSX 18 KB) (17.6KB, xlsx) Data Availability Statement No datasets were generated or analysed during the current study. Articles from Journal of Physiology and Biochemistry are provided here courtesy of Springer ACTIONS View on publisher site PDF (7.2 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top