Identification MYD88 as a candidate molecule of sodium overload-induced cell death in myocardial ischemia-reperfusion injury Authors: Xiuqi Wang, Chenchao Zou, Xiangyuan He, Sheng Liu, Jichun Liu Published: 2026 ABSTRACT AbstractBackgroundMyocardial ischemia-reperfusion (I/R) injury disrupts intracellular ion homeostasis, triggering intracellular sodium overload and ultimately leading to cell death. However, the specific molecular mechanisms driving this pathological process remain incompletely understood.MethodsWeighted gene co-expression network analysis (WGCNA) was applied to myocardial I/R injury transcriptomic datasets from the GEO database to identify genes associated with sodium accumulation. Subsequently, three machine learning algorithms (LASSO, Random Forest, and Support Vector Machine) were integrated to further screen for core hub genes. Single-cell RNA sequencing (scRNA-seq) data were utilized to determine the cellular localization of these key genes, followed by in vivo validation using a murine myocardial I/R injury model.ResultsWGCNA identified 11 candidate genes highly correlated with sodium accumulation. Further screening via machine learning algorithms pinpointed MYD88 and GPR35 as core genes. Single-cell analysis revealed that MYD88 was specifically upregulated in macrophages, suggesting that sodium overload might exacerbate myocardial I/R injury by mediating inflammatory responses. In vivo experiments confirmed a significant elevation of sodium ion concentration in myocardial tissues following I/R, accompanied by a marked upregulation of MYD88 expression.ConclusionThis study confirms that myocardial I/R injury triggers sodium overload and identifies MYD88 as a candidate molecule closely associated with sodium overload-induced cell death (SOICD). These findings provide a novel perspective for further exploring the pathological mechanisms underlying myocardial I/R injury.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13019-026-04333-0.Keywords: Myocardial ischemia-reperfusion injury, Sodium overload-induced cell death, MYD88, machine learning, Macrophage Introduction Myocardial ischemia-reperfusion (I/R) injury represents a pivotal challenge in managing acute myocardial infarction [1–3]. Although timely reperfusion restores blood flow, it paradoxically precipitates secondary damage characterized by oxidative stress, inflammation, and ion homeostasis imbalance [4, 5]. During ischemia, ATP depletion disrupts transmembrane ion gradients, causing intracellular sodium (Na⁺) accumulation [6, 7]. Upon reperfusion, this “sodium overload” exacerbates calcium influx via the Na⁺/Ca²⁺ exchanger, triggering cell swelling, membrane rupture, and necrosis [8–10]. This specific pathological cascade, termed Sodium Overload-Induced Cell Death (SOICD), links ion dysregulation to necrotic cell death [11, 12]. While transcriptomics offers tools to dissect myocardial I/R injury molecular networks, the specific molecular bridges connecting inflammation, gene regulation, and sodium overload mechanisms remain elusive. Therefore, the aim of this study is to identify the candidate genes related to SOICD after myocardial I/R injury, providing ideas for further exploration of the mechanism of myocardial I/R injury (Fig. 1). Flowchart for the study Materials and methods Acquisition and preprocessing of myocardial I/R injury datasets Datasets were retrieved from the GEO database using the keywords “myocardial ischemia–reperfusion injury " with the following inclusion criteria: (i) whole-genome mRNA microarray data; (ii) heart tissue source; (iii) inclusion of I/R and control groups; and (iv) minimal sample size of n = 3 per group with complete annotation. Datasets GSE61592 and GSE160516 were selected. Raw matrices underwent background correction and quantile normalization. Probes lacking gene symbols or exhibiting redundancy were collapsed to the maximum expression value. After merging, batch effects were corrected using the ComBat method in the sva package, with quality assessing via Principal Component Analysis (PCA). The Sodium Overload-Induced Cell Death related genes (SOICDGs) set comprising 646 genes was compiled from GeneCards and literature (Supplementary Table 1). Differentially expressed genes (DEGs) were identified using the limma package in R, with significance defined as an adjusted P-value (FDR) < 0.05 and an absolute log2 fold change (|log2FC|) > 1. Functional enrichment analysis Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using the clusterProfiler package. Using all detectable genes as the background, enrichment for Biological Processes (BP), Cellular Components (CC), Molecular Functions (MF), and KEGG pathways was assessed, applying Benjamini–Hochberg correction for multiple testing [13]. Weighted Gene Co-Expression Network Analysis (WGCNA) A signed co-expression network was constructed using the WGCNA package on the batch-corrected matrix after filtering low-variance genes and outliers. The adjacency matrix was transformed into a Topological Overlap Matrix (TOM). Modules were identified using dynamic tree cutting (minModuleSize = 30) and merged if similarity coefficients exceeded 0.25. Significant modules were identified based on an FDR < 0.05. Candidate gene set for machine learning The genes obtained by WGCNA were intersected with the DEGs and SOICDGs from the three datasets for further machine learning analysis and screening. Cross-analysis was conducted and visualized using a Venn diagram. Machine learning-based feature selection This study employed three machine learning algorithms - LASSO, Random Forest (RF), and Support Vector Machine (SVM). For LASSO (glmnet), the penalty coefficient (lambda.1se) was determined for the Z-score standardized data through 10-fold cross-validation. For RF (random forest), 1000 trees were generated, with mtry set to the square root of the number of features; importance was ranked by averaging the reduction in the Gini index. For SVM (e1071), recursive feature elimination (SVM-RFE) was applied, and its parameters were optimized through grid search. The genes obtained from the three machine learning methods were intersected to determine the key SOICDGs. Independent validation with external datasets This study used another dataset of myocardial I/R injury (GSE108940) to validate the key genes. The data in this dataset was organized in the same way as before. Differential expression was assessed via limma (FDR < 0.05) to confirm consistent effect direction. Processing and clustering of scRNA-seq data The mouse Myocardial I/R injury scRNA-seq dataset (GSE227238) was acquired from GEO and analyzed using the Seurat package. Quality control retained cells with 200–6,000 detected genes (nFeature_RNA) and < 15% mitochondrial content; genes expressed in fewer than three cells were excluded. Data were normalized via the LogNormalize method, and the top 2,000 variable genes were identified. The Harmony algorithm was applied to correct for batch effects. Following PCA, cells were clustered (resolution 0.6–1.2) and visualized using t-SNE or UMAP. Cell types were identified by combining SingleR automated annotation with canonical markers. The expression of key SOICDGs in cell populations To determine the cellular distribution of GSOICD genes, FeaturePlots were generated in t-SNE space. Expression differences between Sham and MIRI groups were visualized using violin plots and statistically assessed via the Wilcoxon rank-sum test. Gene set enrichment analysis (GSEA) Analysis and immune cell infiltration GSEA was performed using the clusterProfiler R package to identify significantly enriched functional pathways, with statistical significance defined as an adjusted P-value < 0.05. Additionally, the immune microenvironment was characterized by applying the CIBERSORT algorithm (utilizing the LM22 reference signature) to the batch-corrected gene expression profiles to estimate the relative proportions of infiltrating immune cells. Establish the myocardial I/R model in mice All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Jiangsu Haohan Biotechnology Co., Ltd. (Approval No. HJSW-25093001). Eight-week-old male C57BL/6 mice (weighing 20–25 g) were purchased from Gemei Pharmaceutical Co., Ltd. Animals were housed in individually ventilated cages under standard specific-pathogen-free (SPF) conditions (22 ± 2 °C, 50 ± 10% humidity, 12-hour light/dark cycle) with ad libitum access to food and water. After a one-week acclimatization period, the mice were randomly assigned to either the Sham group or the ischemia/reperfusion (I/R) group (n = 15 per group). For the establishment of the I/R model, mice were anesthetized with isoflurane (3%–4% for induction, 1.5%–2% for maintenance) and mechanically ventilated. Core body temperature was maintained at 37 °C using a heating pad. Preoperative analgesia (0.1 mg/kg, subcutaneously) was administered. Following a left thoracotomy at the fourth intercostal space, the left anterior descending (LAD) coronary artery was reversibly ligated using a slipknot with a 6 − 0 silk suture. Ischemia was maintained for 30 min, followed by knot release for 24 h of reperfusion. The Control group underwent identical surgical procedures without LAD ligation. Continuous surface electrocardiogram (ECG, Lead II) monitoring was performed throughout the procedure. Successful ischemia was confirmed by immediate ST-segment elevation, and reperfusion was verified by ST-segment resolution. Mice failing to meet these ECG criteria were excluded from subsequent analyses. Following 24 h of reperfusion, the mice were deeply anesthetized with 4%–5% isoflurane and euthanized via exsanguination, in strict compliance with AVMA guidelines. 2,3,5-Triphenyltetrazolium Chloride (TTC), Hematoxylin and Eosin staining (HE), and Masson Staining To assess myocardial damage and fibrosis after reperfusion, hearts were collected, fixed, and paraffin-embedded. Sections were stained with HE and Masson’s Trichrome. Masson staining highlighted myocardial fibres in red and collagen fibres in blue, from which the fibrotic area ratio was calculated. TTC (Solarbio, Cat# G3005) staining was used to assess infarction: hearts were quickly washed, frozen at -4 °C for 30 min, and cut into 1–2 mm sections. The slices were incubated in 2% TTC solution at 37 °C for 15 min in the dark, washed with PBS, and fixed in 4% paraformaldehyde. Viable myocardium was stained red, while infarct areas remained white. Measurement of Heart Tissue Sodium (Na+) Content Heart tissue Na⁺ content was measured using an enzymatic assay kit (Solarbio, BC2785). Briefly, approximately 0.1 g of heart tissue was homogenized on ice in the provided sodium-free extraction buffer at a 1:10 (w/v) ratio. The homogenate was centrifuged at 10,000 g for 10 min at 4℃. The supernatant was collected, and the reaction system was prepared in a microplate according to the manufacturer’s instructions. Absorbance was measured at 405 nm using a microplate reader, and Na+ concentrations were calculated based on a standard curve. Western blot Total protein was extracted from the cells and quantified. Equal amounts of protein were resolved via SDS-PAGE and transferred onto PVDF membranes. Following a 1-hour blocking step with 5% non-fat dry milk in TBST at room temperature, the membranes were incubated overnight at 4 °C with primary antibodies targeting MYD88 (Proteintech, 23230-1-AP, 1:2000), ANP (Proteintech, 27426-1-AP, 1:800), and BNP (Invitrogen, PA5-98294, 1:500). β-Tubulin and GAPDH (Proteintech, 10094-1-AP, 1:20,000) served as the internal loading control. After washing with TBST, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 h at room temperature. Immunoreactive bands were visualized using an enhanced chemiluminescence (ECL) system, and densitometric analysis was performed using ImageJ software. Target protein signals were normalized to the corresponding β-Tubulin signals for statistical comparisons. Result Identification of DEGs Following batch correction, PCA demonstrated a marked reduction in batch effects and distinct separation between myocardial I/R injury and control groups (Fig. 2A–B). Differential analysis identified 1,678 DEGs, including 937 upregulated and 741 downregulated genes (Fig. 2C). Identification of DEGs from the GSE61592 and GSE160516 datasets. (A) Volcano plot illustrating the differentially expressed genes (DEGs) in the integrated GSE61592 and GSE160516 datasets. (B) Principal component analysis (PCA) plot showing the clustering of the integrated samples. (C) Integration of the two datasets identified 1,678 DEGs (937 upregulated and 741 downregulated), as illustrated by the volcano plot and heatmap. (D) KEGG enrichment analysis and GO analysis of DEGs Function enrichment analysis KEGG enrichment analysis revealed that the DEGs were primarily enriched in key pathways such as Regulation of actin cytoskeleton, MAPK signaling pathway, and Propanoate metabolism. GO analysis further indicated that in the BP, genes were mainly involved in phagocytosis, heart contraction, and heart process; in CC, enrichment was observed in focal adhesion and cell-substrate junction; in MF, genes were primarily associated with actin filament binding and actin binding (Fig. 2D). These results highlight the critical roles of cytoskeletal remodeling, MAPK signaling, and cardiac contraction in myocardial I/R injury. 11 genes related to myocardial I/R injury were identified by WGCNA Following WGCNA of the batch-corrected integrated matrix (Fig. 3A), the soft threshold was selected based on scale-free network criteria, and module identification was performed. Three co-expression modules were identified: blue, brown, and yellow (Fig. 3B). Module-trait correlation analysis revealed that the brown and blue modules were significantly correlated with I/R treatment (|r| ≥ 0.72, P ≤ 2 × 10⁻⁴), while the yellow module showed no significant correlation. The brown module exhibited the highest gene significance (Fig. 3C). The genes from the brown modules were intersected with the previously identified DEGs and SOICDGs, resulting in 11 genes associated with myocardial I/R injury: MYD88, MVP, LGALS4, ATP8B1, GPR35, EFHD2, FBLIM1, PLEKHA6, SH3RF2, MYO5C, and RIN1 (Fig. 3D). Identification of 11 SOICD-related candidate genes via WGCNA. (A) WGCNA of the integrated expression matrix identified three distinct co-expression modules. (B) Module-trait correlation analysis revealed significant associations of the brown and blue modules with I/R injury. (C) The brown module exhibited the highest gene significance. (D) Final identification of the 11 candidate genes associated with SOICD MYD88 and GPR35 were further screened using machine learning Manhattan and circos plots revealed the distribution of 11 genes related to myocardial I/R injury across chromosomes (Fig. 4A-B). The significance of these genes across various chromosomal locations further supports their potential roles in myocardial injury. LASSO regression, using 10-fold cross-validation, selected five stable feature genes: MYD88, MVP, GPR35, PLEKHA6, and RIN1 (Fig. 4C). RF analysis calculated gene importance and further confirmed the critical roles of FBLIM1, GPR35, MYD88, and LGALS4 (Fig. 4D). SVM-RFE recursive feature elimination ultimately identified 11 genes (Fig. 4E). The Venn diagram confirmed that MYD88 and GPR35 were key SOICDGs identified by all three methods (Fig. 4F). MYD88 and GPR35 as the key SOICDGs in myocardial I/R injury (A-B) Manhattan and Circos plots showing the distributions of 11 candidate genes across chromosomes. (C) LASSO regression identified MYD88, MVP, GPR35, PLEKHA6, and RIN1. (D) Random forest analysis confirmed the importance of FBLIM1, GPR35, MYD88, and LGALS4 in myocardial I/R injury. (E) SVM-RFE revealed MYD88, GPR35, and 9 other genes associated with I/R injury. (F) The Venn diagram find that MYD88 and GPR35 are critical genes identified by all three methods. (G) Validation using the GSE108940 dataset revealed a significant upregulation of MYD88 (P = 0.04) but no significant change in GPR35 (P = 0.95) High expression of MYD88 demonstrates greater robustness in the independent validation dataset In an independent validation dataset of myocardial I/R injury (GSE108940), MYD88 was consistently and significantly upregulated (P = 0.04), whereas GPR35 showed no significant alteration (P = 0.95) (Fig. 4G). This indicates that the upregulation of MYD88 exhibits greater robustness across cohorts, indicating superior robustness for MYD88. Identification of transcriptional heterogeneity and hyper-variable features in the mouse heart To deconvolute the cellular heterogeneity induced by myocardial I/R injury, we performed a genome-wide variability analysis on 16,200 background genes. We identified 1,500 hyper-variable genes (HVGs) as core drivers of cell clustering. As shown in Fig. 5A, Spp1 (fibrosis), Nppa (cardiac stress), and hemoglobin genes (Hba-a1, Hbb-bs) exhibited the highest standardized variance. This suggests that inflammation, hemodynamic stress, and hypoxia are primary drivers of transcriptional heterogeneity. Additionally, cytoskeletal genes (Acta2, Tagln) and chemokines (Ccl21a) were top HVGs, indicating drastic transcriptional reprogramming of cardiac stromal cells. The heatmap revealed distinct modular patterns. Stress response genes and extracellular matrix remodeling genes were highly enriched in specific clusters, highlighting the transcriptional signatures driving tissue injury (Fig. 5B). Based on these features, t-SNE projection identified six major cell populations (Fig. 5C). These included Cardiomyocytes, Fibroblasts, Vascular Smooth Muscle Cells, Neutrophils, Macrophages, and Resident Macrophages. Comparison between groups showed a marked increase in Neutrophils and Infiltrating Macrophages in the myocardial I/R injury group. In contrast, the Sham group was dominated by homeostatic Cardiomyocytes and Resident Macrophages. Preliminary identification of MYD88 as a key candidate molecule. (A) Analysis of transcriptional heterogeneity identifying HVGs in the single-cell dataset. (B) Heatmap visualization of the top feature genes driving the first four principal components (PC_1 to PC_4). (C) t-SNE projection of major cell subpopulations in the mouse heart, illustrating cellular heterogeneity and microenvironmental remodeling between the Control and myocardial I/R injury groups. (D) Violin plots detailing the expression distribution of MYD88 across distinct cell types. MYD88 is significantly upregulated in macrophages, fibroblasts, and vascular smooth muscle cells in the myocardial I/R injury group, highlighting its robust activation across multiple cell compartments in response to injury. (E) GSEA revealing pathways associated with MYD88, including B-cell receptor signaling and the TCA cycle. (F) Immune infiltration analysis demonstrating significant correlations between MYD88 expression and diverse immune cell populations, notably B cells, T cells, and macrophages MYD88 is significantly upregulated in macrophages To further elucidate the underlying cellular mechanisms, this study analyzed the expression profiles of MYD88 and GPR35 across various cell populations. The results revealed that both MYD88 and GPR35 were predominantly enriched in macrophages (Fig. 5D). Further comparative analysis demonstrated that, within the macrophage subpopulation, the expression level of MYD88 was significantly upregulated in the myocardial I/R group compared to the control group, whereas GPR35 showed no significant alteration (Fig. 5E) (Table 1). Taking into account the consistent findings from the independent validation, we ultimately selected MYD88 as the pivotal SOICDG for subsequent experimental validation. Table 1Multiple-testing for MYD88 in MacrophagesGeneAverage log2FCP-valueP-value (Padj)clusterMYD882.3981.48 × 10− 752.41 × 10− 71Macrophages Multiple-testing for MYD88 in Macrophages GSEA and Immune cell correlation of MYD88 GSEA revealed that MYD88 was significantly enriched in the B cell receptor signaling pathway and the tricarboxylic acid (TCA) cycle and ECM receptor (Fig. 5F). Furthermore, immune cell infiltration analysis demonstrated significant correlations between MYD88 expression levels and the relative abundances of multiple immune cell subpopulations. Specifically, MYD88 exhibited strong associations with B cells, T cells, and macrophages, suggesting a critical regulatory role in the immune response following myocardial I/R injury (Fig. 5G). Establishment of Mouse Myocardial I/R injury Model After ligation of the LAD in mice, ECG studies revealed significant ST segment elevations, confirming successful induction of ischemia. Following reperfusion, the ST segment significantly decreased, and the mice survived for 24 h, confirming the successful establishment of the mouse myocardial I/R injury model (Fig. 6A). TTC staining revealed clear white infarct areas in the myocardial tissue of the reperfusion group, indicating ischemic injury (Fig. 6B). HE staining revealed structural damage to the myocardial cells, with some areas showing cell swelling and necrosis. No significant fibrosis was observed by Masson staining, suggesting that significant fibrosis had not yet occurred in this model (Fig. 6C). In vivo validation of elevated MYD88 expression and sodium overload following myocardial I/R injury. (A) ECG tracings showing ST-segment elevation and recovery, confirming successful I/R injury. (B) TTC staining showing myocardial infarct areas. (C) HE staining revealing structural tissue damage, cell swelling, and necrosis. (D) Measurement of myocardial sodium ion concentrations, showing a significant increase in the I/R group relative to controls. (E) Western blot analysis and quantification confirming elevated MYD88 protein levels in the I/R group, consistent with bioinformatic predictions Myocardial I/R injury induces sodium overload in mouse heart tissues To evaluate the impact of myocardial I/R injury on cardiac ion homeostasis in mice, we extracted heart tissue homogenates from each group and measured the sodium ion (Na+) concentrations. The results revealed that, compared to the control group, the sodium content in the myocardial tissue of the I/R model group was significantly elevated (I/R vs. Control: 56.8 ± 6.65 vs. 29.4 ± 5.27 µmol/g, P < 0.05) (Fig. 6D). This abnormal ion accumulation at the tissue level indicates that I/R stress led to a pronounced intracellular “sodium overload” in cardiomyocytes, thereby providing crucial evidence for the subsequent induction of cell damage. MYD88 is upregulated in the myocardial I/R injury group Total protein was extracted from the heart tissue of both groups and subjected to Western blot analysis. The results revealed that the protein levels of ANP and BNP were significantly greater in the reperfusion injury group than in the control group, further confirming the successful establishment of the mouse myocardial I/R injury model (Fig. 6E). Additionally, the expressions of MYD88 was significantly greater in the reperfusion injury group, which is consistent with the findings from previous bioinformatics analysis and supports the role of these proteins in myocardial I/R injury. Discussion In the present study, by integrating multiple transcriptomic datasets related to myocardial I/R injury and employing a combination of WGCNA and three machine learning algorithms (LASSO, Random Forest, and SVM), we identified MYD88 as a pivotal candidate gene associated with SOICD. Subsequent in vivo validation using a murine myocardial I/R model confirmed a significant accumulation of sodium ions alongside a marked upregulation of MYD88 expression, reinforcing the robustness of our bioinformatic predictions. Crucially, scRNA-seq analysis revealed that MYD88 is specifically upregulated within the macrophage subpopulation. Taken together with the results of GSEA (highlighting B cell receptor signaling and the TCA cycle) and immune infiltration analysis, these findings suggest that MYD88 may act as a central immunometabolic hub, participating in the pathological progression of I/R injury by modulating macrophage-mediated inflammatory responses and metabolic dysfunction. Specifically, machine learning algorithms were utilized to identify MYD88 and GPR35 as core hub genes. Subsequent validation using an independent external dataset revealed that the expression level of MYD88 remained significantly upregulated, whereas GPR35 showed no significant alteration. This indicates that the upregulation of MYD88 exhibits stronger robustness across cohorts. Furthermore, the single-cell transcriptomic analysis demonstrated that MYD88 is predominantly enriched and specifically highly expressed in the macrophage subpopulation. This suggests MYD88 may promote sodium overload through specific cellular mechanisms. While our data highlights the robust expression of MYD88, particularly in infiltrating macrophages, the precise mechanism linking MYD88 to sodium overload remains to be functionally validated. As a central adaptor in TLR/IL-1R signaling, MYD88 is commonly elevated during inflammatory injury. We hypothesize that its activation may promote a pro-inflammatory microenvironment. Previous literature indicates that this pathway can trigger the release of cytokines and Reactive Oxygen Species (ROS) [14, 15]. Given that ROS are known to disrupt ion homeostasis—potentially by activating the Na⁺/H⁺ Exchanger 1 (NHE-1) and inhibiting the Na⁺/K⁺-ATPase pump-it is plausible that macrophage-derived MYD88 signaling indirectly contributes to sodium accumulation in cardiomyocytes via oxidative stress [16, 17]. Furthermore, the elevated MYD88 levels observed in fibroblasts and smooth muscle cells suggest that I/R injury may induce a broader pro-inflammatory transition in stromal compartments, potentially exacerbating tissue damage and cell death [18–20]. We acknowledge that this proposed MYD88-ROS-NHE1/Na⁺ axis is currently speculative. It serves as a theoretical framework derived from our bioinformatic findings and existing literature, which warrants rigorous investigation in future functional studies. Immune Microenvironment and SOICD Our immune infiltration analysis revealed a strong correlation between MYD88 and the recruitment of macrophages and neutrophils. These immune cells release pore-forming toxins and proteolytic enzymes that can compromise membrane integrity, further aggravating osmotic swelling and sodium overload. Thus, MYD88 likely orchestrates a pathological microenvironment that accelerates SOICD [15]. Several limitations of this study should be acknowledged. First, while we preliminarily identified MYD88 as a potential mediator of myocardial I/R injury, this study lacks direct functional experiments to robustly validate the downstream molecular mechanisms. Additionally, we were unable to directly measure dynamic changes in intracellular sodium concentrations to experimentally confirm the sodium overload phenotype. Second, although we observed a significant upregulation of MYD88 in macrophages, the specific mechanisms by which it orchestrates cardiomyocyte injury remain to be fully elucidated. Future research should prioritize investigating the intercellular communication (crosstalk) between macrophages and cardiomyocytes, which may be crucial for thoroughly decoding the pathogenesis of I/R-induced myocardial injury. Conclusion In summary, this study confirms the occurrence of tissue sodium overload during myocardial I/R injury and identifies MYD88 as a key candidate molecule associated with SOICD. Our integrated analyses reveal that MYD88 is significantly upregulated, particularly within infiltrating macrophages. While the precise causal mechanisms remain to be functionally validated, we hypothesize that MYD88-associated inflammatory responses may indirectly contribute to the disruption of cardiac ion homeostasis, potentially exacerbating sodium overload and subsequent cell death. Therefore, further investigation into MYD88 and its associated pathways may provide a novel therapeutic perspective for mitigating myocardial reperfusion injury. Supplementary Information Supplementary Material 1 Supplementary Material 2 Acknowledgements None. Author contributions This study was designed by X.W and J.L. X.W and C.Z contributed to the data collection and analysis. X.W and X.H wrote the first draft of the paper. S.L and J.L revised the article. All authors agreed to the published version of the manuscript. Funding None. Data availability GSE61592, GSE160516 and GSE108940 can be downloaded for free from the official website of the GEO database. Declarations Ethics approval and consent to participate All animal protocols were approved by the Institutional Animal Care and Use Committee of Jiangsu Haohan Biotechnology Co., Ltd. (Approval No: HJSW-25093001). Competing interests The authors declare no competing interests. Clinical trial number Not applicable. Patient consent for publication Not applicable. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Xiuqi Wang, Chenchao Zou, and Xiangyuan He, are the co-first authors. Contributor Information Sheng Liu, Email: Tantan138@163.com. Jichun Liu, Email: ljcncdxefy0729@163.com. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement GSE61592, GSE160516 and GSE108940 can be downloaded for free from the official website of the GEO database.