Using the R package GEOquery (14), we downloaded two datasets each for AMI [GSE48060 (15) and GSE29532 (16)] and ICM [GSE116250 (17) and GSE46224 (18)] from the GEO (19) database (https://www.ncbi.nlm.nih.gov/geo/). Comprehensive details are available in Supplementary Tables S1 , S2 . The R package sva (20) was utilized for batch correction and integration, producing the consolidated GEO datasets for AMI and ICM. The R package limma (21) facilitated normalization and standardization, followed by principal component analysis (22). Mitophagy-related genes (MRGs) were sourced from the GeneCards database (23) (https://www.genecards.org/) and the Molecular Signatures Database (MSigDB) (24) (https://www.gsea-msigdb.org/gsea/msigdb), yielding a total of 1633 unique MRGs (mitophagy-related genes) after merging and deduplication, as detailed in Supplementary Table S3 .

The analysis of differential gene expression was carried out for both AMI and ICM using the limma package in R. After reviewing literature (25, 26) and testing various thresholds, we chose |logFC| > 0 and P < 0.05 to ensure robust results while maximizing the inclusion of as many biologically significant differentially expressed genes as possible. Genes with logFC above 0 and a p-value below 0.05 were categorized as up-regulated, whereas those with logFC below 0 and the same p-value threshold were categorized as down-regulated. Venn diagrams were employed to depict the overlap between up-regulated and down-regulated genes, and further intersections with MRGs (mitophagy-related genes) were analyzed to pinpoint MRDEGs (mitophagy-related differentially expressed genes).

The STRING database (27) (https://string-db.org/) facilitated the construction of a PPI network based on MRDEGs(mitophagy-related differentially expressed genes), employing a minimum interaction confidence score of 0.400(medium confidence). Interactions with a confidence score above this threshold are considered to be sufficiently supported by evidence, thereby filtering out potential false-positive results. The CytoHubba (28) plugin within Cytoscape (29) software applied five algorithms—Maximum Neighborhood Component (MNC), Maximal Clique Centrality (MCC), Edge Percolated Component (EPC), Degree, Closeness (30)—to compute scores for MRDEGs, selecting the top 20 MRDEGs. The intersection of results from these algorithms identified hub genes related to AMI and ICM. By performing multi-analysis screening using the STRING database and five algorithms in Cytoscape, the reliability of the results was enhanced, and errors that might arise from relying on a single algorithm were minimized.

AlphaFoldDB (31) (https://alphafold.com) predicted and visually displayed the protein structures of hub genes, assessed by a Predicted Local Distance Difference Test (pLDDT) score ranging from 0 to 100. The regulatory network between the mRNA of 9 hub genes and 48 transcription factors (TFs) was predicted using the ChIPBase (32) database (http://rna.sysu.edu.cn/chipbase/). Potential interactions between mRNA and miRNAs, as well as mRNA and RNA-binding proteins (RBPs) (33), were screened using the StarBase v3.0 database (34) (https://starbase.sysu.edu.cn/), and the networks were visualized using Cytoscape software. This analysis included 4 hub genes and 27 miRNAs, as well as 10 hub genes and 43 RBPs. Furthermore, the Comparative Toxicogenomics Database(CTD) (35) (https://ctdbase.org/) was employed to identify potential drugs or molecular compounds associated with the hub genes. The mRNA-Drug regulatory network was constructed and subsequently visualized using Cytoscape software, comprising 8 hub genes and 15 drugs or molecular proteins.

Expression levels of MRDEGs(mitophagy-related differentially expressed genes) in the Combined Datasets were compared using group comparison graphs. The Spearman algorithm analyzed the correlation of hub gene expressions, with the R packages igraph (36) and ggraph illustrating correlations and chord diagrams. Scatter plots by the ggplot2 R package displayed the strongest correlated hub genes.

Gene Ontology(GO) (37) and Kyoto Encyclopedia of Genes and Genomes(KEGG) (38) enrichment analysis of hub genes was performed using the R package clusterProfiler (39), electing results based on an adjusted p-value < 0.05. The Pathview R package (40) visualized the pathway enrichment analysis results.

GSEA (41) (Gene Set Enrichment Analysis)analysis was executed on the combined datasets for AMI and ICM using the clusterProfiler package in R, with the following settings: seed value at 2023, a gene set size range from 10 to 500, and the gene set c2.cp.all.v2022.1.Hs.symbols.gmt [All Canonical Pathways]. The threshold for significance was set at a p-value below 0.05. Additionally, GSVA (42) (Gene Set Variation Analysis)was applied to all genes within the combined datasets of AMI and ICM, utilizing gene sets from MSigDB (24), adhering to the same p-value criterion for selection.

We utilized multiple machine learning algorithms, including Logistic Regression, Support Vector Machine (SVM) (43), and Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis, to identify key genes for constructing diagnostic models for AMI and ICM. This approach is grounded in several studies of significant scientific value in the field of bioinformatics (44, 45). The models were implemented using the R package glmnet, with parameters set.seed (500) and family=‘binomial’ (46).The key genes chosen from AMI and ICM to determine the RiskScore, employing coefficients obtained from LASSO regression analysis.

ROC(Receiver Operating Characteristic) curves were plotted for the diagnostic models of AMI Key Genes and ICM Key Genes using the pROC package in R. Additionally, nomograms (47) illustrating the relationships between Key Genes were generated with the rms package in R. Calibration analysis was conducted to evaluate the precision and discriminatory capacity of the diagnostic models for AMI and ICM. Decision Curve Analysis (DCA) (48) for predicting clinical outcomes using AMI Key Genes and ICM Key Genes was performed using the ggDCA package in R. Moreover, Functional Similarity (Friends) analysis was carried out with the GOSemSim R package (49).

To enhance the reliability of our methodology, we drew upon the approach proposed by Zhang L et al. (50), and utilized mitophagy-related RiskScore to subgroup the AMI group for further in-depth analysis. Based on the formula outlined in section 2.8, we calculated the RiskScore for acute myocardial infarction (AMI) samples within the AMI Combined Datasets, utilizing the regression coefficients derived from the LASSO model specifically for AMI. The median RiskScore was instrumental in categorizing the samples into HighRisk and LowRisk groups. Samples with a risk score above the median were classified into the HighRisk group, while those with a risk score equal to or below the median were classified into the LowRisk group. A similar methodology was applied to determine the RiskScore for ischemic cardiomyopathy (ICM) samples in the ICM Combined Datasets, again using the LASSO regression coefficients pertinent to ICM. The samples were classified into HighRisk and LowRisk categories based on their median RiskScores. These two sets of high- and low-risk classifications will be utilized for subsequent subgroup analyses independently. Differential analysis was carried out with the limma package in R, with visualization of the results achieved through the ggplot2 and pheatmap packages in R.

GSEA (41) was conducted on AMI samples in the AMI Combined Datasets and ICM samples in the ICM Combined Datasets with clusterProfiler package in R. GSVA (42) was applied to the HighRisk and LowRisk groups of AMI and ICM samples, respectively. The same gene sets, parameters, and screening criteria were used as in previous analyses.

Immune cell infiltration matrices were determined through single sample gene set enrichment analysis (ssGSEA) (51) for samples of AMI and ICM. Comparison graphs for the groups were created using ggplot2 to illustrate the variance in immune cell expression between the LowRisk and HighRisk groups in AMI and ICM. The detailed subgroup classification method can be found in section 2.10.

The Ethics Committee of Shanghai East Hospital, affiliated with Tongji University, approved this study(Approval number 2024-175), which follows the Declaration of Helsinki guidelines. Once written informed consent was secured from all participants, peripheral blood samples were collected from six individuals each with diagnoses of AMI and ICM, and from six normal subjects.

Venous blood samples were collected into whole blood RNA preservation tubes (model ZXQX-10). After centrifugation and sedimentation, total RNA was extracted from peripheral blood mononuclear cells (PBMCs) employing the TransZol Up Plus RNA kit (TransGen, China). The integrity and purity of the extracted RNA were evaluated using the GEN5 microplate reader (biotek, USA). Quantitative real-time PCR (RT-qPCR) experiments were performed on the Q5 Real-Time PCR Detection System (Thermo, USA). Glyceraldehyde-3-phosphate dehydrogenase (GADPH) was used as the internal control to normalize the data. Relative expression levels of the target genes were determined using the 2-ΔΔCt method.

This article’s data processing and analyses were performed with R software (Version 4.3.0). We assessed the statistical significance of continuous variables across two groups using the independent Student’s T-Test. For variables that did not follow a normal distribution, the Mann-Whitney U test (also referred to as the Wilcoxon Rank Sum Test) was utilized. The Kruskal-Wallis test was applied to analyze data involving three or more groups. Spearman’s correlation analysis determined the relationships among various molecules. All statistical tests were two-sided, and a p-value threshold of less than 0.05 was considered significant.

Batch effects were meticulously removed from AMI datasets GSE48060 and GSE29532, culminating in the creation of the AMI Combined Datasets ( Figures 2A, B, E, F ). In a similar manner, batch effects were eliminated from RPKM(Reads Per Kilobase of transcript per Million mapped reads)data for ICM datasets GSE116250 and GSE46224, producing the ICM Combined Datasets ( Figures 2C, D, G, H ).

Within the AMI Combined Datasets, a total of 1472 genes were identified as differentially expressed, encompassing 677 upregulated and 795 downregulated genes ( Figures 2I, K ). Concurrently, the ICM Combined Datasets revealed 7572 DEGs(differentially expressed genes), including 3557 upregulated and 4015 downregulated genes ( Figures 2J, L ).

A comparison of genes from both datasets showed that 429 genes were differentially expressed in both AMI and ICM, with 237 downregulated ( Figure 2M ) and 192 of these genes upregulated ( Figure 2N ). Additionally, 61 MRDEGs(mitophagy-related differentially expressed genes) were isolated through the intersection of the DEGs with genes related to mitophagy ( Figure 2O ), detailed further in Supplementary Table S4 .

Utilizing the STRING database, an interrelationship was established among 52 MRDEGs (mitophagy-related differentially expressed genes), forming a robust PPI network ( Figure 3A ). These genes were evaluated and ranked using five distinct algorithms within Cytoscape, leading to the identification of the top 20 MRDEGs ( Figures 3B–F ). A cross-analytical approach among these algorithms revealed 11 critical Hub Genes associated with both AMI and ICM: POLR2B, RPS11, MRPS5, METAP1, HNRNPA2B1, XRN1, GART, GFM1, TNPO1, LIG3, and AGK ( Figure 3G ).

Protein structures for the 11 Hub Genes were predicted and visualized using AlphaFoldDB ( Figures 4A–K ). Nine of these genes demonstrated high structural confidence (pLDDT > 90) across their main domains: POLR2B, RPS11, MRPS5, METAP1, HNRNPA2B1, XRN1, GART, TNPO1, and AGK; the remaining two, LIG3 and GFM1, showed substantial confidence (70 < pLDDT < 90).

The regulatory network was expanded to include 48 transcription factors linked to 9 Hub Genes, constructing an mRNA-TF network ( Figure 4L , detailed in Supplementary Table S5 ). Moreover, regulatory networks involving 27 miRNAs binding to 4 hub genes and 43 RBPs associated with 10 Hub Genes were elucidated ( Supplementary Table S5 , Figures 4M, N , respectively). An mRNA-drug interaction network involving 8 Hub Genes and 15 drugs or molecular compounds was also constructed ( Figure 4O , detailed in Supplementary Table S5 ).

Substantial differences were detected in the expression levels of 9 Hub Genes between the AMI group and the control group, with genes AGK, GART, HNRNPA2B1, LIG3, METAP1, POLR2B, RPS11, TNPO1, and XRN1 showing statistically significant differences (p-value < 0.05) ( Figures 5A, D ). The most pronounced positive correlation was between GFM1 and AGK, showing a p-value less than 0.001 and a correlation coefficient (r-value) of 0.716 ( Figure 5B ), while the strongest negative correlation was noted between RPS11 and POLR2B, with an r value of -0.371 and a p-value less <0.001 ( Figure 5C ).

When contrasting the ICM group with the control group, 10 Hub Genes revealed noteworthy distinctions in their expression levels, with statistical significance (p-value < 0.05): AGK, GART, HNRNPA2B1, LIG3, METAP1, MRPS5, POLR2B, RPS11, TNPO1, and XRN1 ( Figures 5E, H ). In this group, the strongest positive correlation was found between MRPS5 and HNRNPA2B1 (r-value = 0.645, p-value < 0.001, Figure 5F ), while the most significant negative correlation was noted between RPS11 and POLR2B (r- value = -0.698, p-value < 0.001, Figure 5G ).

After converting the hub genes into gene IDs, comprehensive GO and KEGG analyses were performed. Key biological processes identified included the DNA biosynthetic process, the regulation of telomere maintenance via telomere lengthening, telomere maintenance via telomere lengthening, regulation of DNA biosynthetic process ( Figures 6A–D , Supplementary Table S6 ). The KEGG pathway analysis further elucidated the relationship between Hub Genes and critical signaling pathways. These genes exhibited notable enrichment primarily in pathways such as Ribosome Pathway, the Antifolate Resistance, One Carbon Pool by Folate, RNA polymerase, and Base excision repair ( Figures 6E, F , Supplementary Figure 1 ).

The GSEA highlighted that all genes in the AMI Combined Datasets were significantly associated with biological functions including regulation of cell death, inflammatory mediators, and their signaling pathways ( Figures 7A, C–F , Supplementary Table S7 ). Similarly, genes in the ICM Combined Datasets were predominantly linked to inflammatory responses and pathways like Tgf Beta, Nfkb, among others ( Figures 7B, G–J , Supplementary Table S8 ).

In the GSVA, significant distinctions were noted in the enrichment of specific functions and pathways between the datasets. For the AMI Combined Datasets, notable pathways included Reactive Oxygen Species Pathway, Cholesterol Homeostasis, and Tgf Beta Signaling ( Figures 7K, L , Supplementary Table S9 ). In the ICM Combined Datasets, numerous pathways demonstrated significant enrichment. By applying stringent criteria, including a p-value < 0.05 and ranking by logFC, we identified the top 10 pathways demonstrating positive enrichment as well as the top 10 pathways displaying negative enrichment. This selection includes pathways such as Unfolded Protein Response, E2f Targets, Spermatogenesis, Heme Metabolism, Peroxisome, and Myogenesis, among others ( Figures 7M, N , Supplementary Table S10 ).

For AMI, a logistic regression model utilizing 11 hub genes identified 10 significant contributors ( Figure 8A ). Additionally, an SVM(support vector machine) model highlighted 7 pivotal genes with minimal error rates and maximum accuracy ( Figures 8B, C ). Critical genes such as RPS11 and AGK were further affirmed through LASSO(Least Absolute Shrinkage and Selection Operator) regression, which refined the AMI diagnostic model to include 4 key genes: RPS11, METAP1, HNRNPA2B1, and AGK ( Figures 8D, E ). In ICM diagnosis, logistic regression, SVM, and LASSO models emphasized the diagnostic relevance of genes like MRPS5, METAP1, and HNRNPA2B1 ( Figures 8N–R ).

The ROC(Receiver Operating Characteristic) curve generated from the RiskScore of the diagnostic model for AMI demonstrated an AUC (Area Under the Curve) of 0.833 ( Figure 8F ), indicating high diagnostic accuracy. The ROC curve for the key gene RPS11 had an AUC value of 0.794, with other genes having AUC values between 0.5 and 0.7 ( Figures 8G–J ). The nomogram highlighted the significant contribution of RPS11 expression to improving the diagnostic utility of the AMI model over other factors ( Figure 8M ).

Similarly, the ICM diagnostic model showed high diagnostic accuracy with a risk score AUC of 0.996, and key genes MRPS5 (AUC 0.929), HNRNPA2B1 (AUC 0.877), and METAP1 (AUC 0.855) also demonstrating high diagnostic accuracy ( Figures 8S–V ). The nomogram indicated that the expression of MRPS5 notably enhances the diagnostic utility of the ICM model over other variables ( Figure 8W ). Calibration curve analysis and DCA(Decision Curve Analysis) showed that both AMI and ICM diagnostic models perform well with significant net benefits ( Figures 8K, L, X, Y ).

Analysis conducted via the Friends algorithm indicated that RPS11 is the gene proximate to the critical threshold (cut-off value = 0.60) in the context of AMI as depicted in Figure 9S . Similarly, HNRNPA2B1 emerges as the gene nearest to the critical threshold (cut-off value = 0.60) within ICM, as illustrated in Figure 9T . These findings suggest that each of these genes holds a significant role in the pathogenesis of AMI and ICM, respectively.

An analysis within the AMI samples identified 2504 DEGs(differentially expressed genes), meeting the criteria of having an |logFC |> 0 and a p-value < 0.05 between high and low-risk groups. The detailed subgroup classification method can be found in section 2.10. Among these DEGs, 1193 genes were upregulated, while 1311 were downregulated ( Figures 9A, B ). Subsequent GSEA revealed significant enrichment across various biological functions and signaling pathways including WNT5A-dependent internalization of FZD4, the Hedgehog signaling pathway, Wnt ligand biogenesis and trafficking, and ADORA2B-mediated production of anti-inflammatory cytokines ( Figures 9C–G ). The specific results of this analysis are documented in Supplementary Table S11 .

Similarly, for ICM samples, 1875 DEGs (differentially expressed genes) met the established criteria, with 803 genes upregulated and 1072 genes downregulated ( Figures 9H, I ). The GSEA for these samples indicated significant enrichment in pathways associated with Oxidative Stress Response, IL1 and Megakaryocytes In Obesity, Photodynamic Therapy-induced NFkb Survival Signaling, and TGF-beta Receptor Signaling In Skeletal Dysplasias ( Figures 9J–N ), with comprehensive details provided in Supplementary Table S12 .

For AMI samples, GSVA differentiated the top 10 positively and negatively enriched pathways between HighRisk and LowRisk groups based on p-values less than 0.05 and logFC rankings ( Figures 9O, P , Supplementary Table S13 ). The detailed subgroup classification method can be found in section 2.10. Validation via the Mann-Whitney U test reaffirmed the statistical significance (p-value < 0.05) of 10 pathways between the HighRisk and LowRisk groups, including pathways such as Myogenesis, KRAS Signaling Down, Epithelial-Mesenchymal Transition, Pancreas Beta Cells, Heme Metabolism, UV Response Down, Estrogen Response (Late and Early), TGF-beta Signaling, and Protein Secretion.

The analysis of ICM samples highlighted 12 pathways showing statistically significant differences between HighRisk and LowRisk groups ( Figures 9Q, R , Supplementary Table S14 ), including pathways involved in P53 Pathway, IL6_JAK_STAT3 Signaling, TNFA Signaling Via NFKB, Hypoxia, and Spermatogenesis, KRAS Signaling Up, IL2_STAT5 Signaling, Coagulation.

In AMI samples, ssGSEA analysis revealed distinct variations in the presence of six types of immune cells between the high and low-risk groups. Please refer to Section 2.10 for the detailed grouping method of high and low-risk subgroups. These included activated CD4+ T cells, CD56 bright natural killer cells, central memory CD4+ T cells, natural killer cells, Type 17 T helper cells, and Type 2 T helper cells, as illustrated in Figure 10A . Notably, the LowRisk group exhibited a substantial negative correlation between Type 17 T helper cells and activated CD4+ T cells(r-value = -0.579, p-value < 0.05) ( Figure 10B ), and similar trends were observed in the HighRisk group ( Figure 10C ). The gene AGK exhibited a significant positive association with activated CD4+ T cells in both risk groups, as shown in Figures 10D, E . Furthermore, the gene RPS11 shows a positive correlation with natural killer cells in the HighRisk group ( Figure 10E ).

For ICM samples, ssGSEA indicated significant variations in 10 types of immune cells, including activated CD4+ T cells, central memory CD8+ T cells, effector memory CD8+ T cells, activated dendritic cells, eosinophils, macrophages, memory B cells, mast cells, myeloid-derived suppressor cells (MDSCs), and regulatory T cells ( Figure 10F ). In the LowRisk group, regulatory T cells and macrophages demonstrated a marked positive association, with an r-value of 0.952 and a p-value less than 0.05 ( Figure 10G ); while in the HighRisk group, mast cells and MDSCs displayed a significant positive correlation (r-value = 0.936, p-value < 0.05) ( Figure 10H ). Moreover, the gene MRPS5 showed a notable negative association with activated CD4+ T cells in the LowRisk group ( Figure 10I ) and with memory B cells in the HighRisk group ( Figure 10J ).

Key genes were validated in peripheral blood from three groups using RT-PCR. The results showed a significant statistical difference in RPS11 in the AMI group (p<0.05) and a significant statistical difference in MRPS5 in the ICM group (p<0.05) ( Figures 11A–E ).This suggests that different mitophagy genes are involved and play distinct roles in the AMI phase and the chronic phase of ICM. During the acute phase, the increased expression of RPS11 indicates its potential as a valuable diagnostic marker.

To further investigate their effect on MI (Myocardial Infarction), we induced MI mice and performed Electrocardiographic (ECG) monitoring ( Figures 11K, L ) and cardiac ultrasonography ( Figures 11H–J ) to assess cardiac function 48 hours after MI. Electrocardiogram results confirmed the successful establishment of the acute myocardial infarction model. Ultrasonography results showed a significant decline in heart function following acute myocardial infarction, consistent with the heart failure criteria of ischemic cardiomyopathy. Mice were raised until 28 days of post-ligation, at which point cardiac tissue was collected. At 28 days (4 weeks), the myocardial tissue had already undergone the acute inflammatory phase, with increased fibrosis and the initiation of myocardial remodeling, marking a critical stage that impacts subsequent heart function. We performed hematoxylin and eosin (HE) staining ( Figures 11F, G ) on the heart tissue of the mice to visualize the area of infarction as well as the surrounding area. Additionally, further RT-PCR analyses ( Figures 11M, N ) were conducted, which revealed that the expression of RPS11 was elevated in the infarcted mice group. The outcome of Masson staining ( Figures 12A, B ) showed that the myocardial tissue of the sham group was structurally intact, the cardiomyocytes were aligned and the collagen fibers in the interstitium were less stained and had a sporadic distribution, whereas the myocardial tissue of the infarct group was significantly damaged, with some areas of cardiomyocytes disorganized or broken, and the staining of collagen fibers in the fibrotic region was significantly increased (blue color).

Immunohistochemical staining of the animal model ( Figures 12C–I ) showed that the RPS11 protein was mainly distributed in the cytoplasm of the cells, and the positive signals were brownish-yellow in color. The positive expression of RPS11 in the infarct area was significantly increased in the MI group compared with the Sham group (p < 0.05). These findings suggest that RPS11 exhibits significant differential expression not only during the acute phase of myocardial infarction (AMI) but also in the transition from the acute to the chronic phase.

Observation under the light microscope showed that the morphology of H9c2 cardiomyocytes in the normoxic control group was intact and evenly arranged, and the nuclei were clear, while the cells in the hypoxia-reoxygenation group (the H/R group) showed the damage characteristics of rounding, crumbling, widening of gaps, and partial rupture and detachment (as shown in Figures 12K, L ), indicating that hypoxia-reoxygenation treatment had significantly affected the morphology of cardiomyocytes.

Cell flow assay results ( Figures 12M, N, J ) showed that the apoptosis rate of cardiomyocytes in the H/R group was significantly higher than that in the Control group (P < 0.05).

Western blot results ( Figures 12O–R ) showed that the expression of RPS11, BNIP3(BCL2 protein-interacting protein 3), and the LC3II/I(Microtubule-associated protein 1 light chain 3II/I) ratio was significantly lower in the H/R+siRPS11 group than in the H/R group (p < 0.05). The expression of the internal control protein as an internal reference protein was not significantly different in the two groups (p > 0.05). The expression levels of the individual proteins in the H/R+siCON group did not differ significantly from those of the H/R group (p > 0.05).The myocardial hypoxia-reoxygenation model simulates the cardiac status of acute myocardial infarction patients following emergency revascularization in real-world settings, providing strong evidence for a comprehensive understanding of the mechanisms underlying the transition from AMI to ICM.