Introduction

MI is one of the leading causes of death worldwide1,2, inflicting irreversible damage on cardiac muscle tissue. The aging of cardiomyocytes is considered to play a significant role in age-related cardiovascular diseases3. Researchers have utilized single-cell resolution to elucidate the transcriptomic changes of various cardiac cell types, including cardiomyocytes, fibroblasts, endothelial cells, macrophages, T cells, B cells, and more than a dozen others, discovering that cardiomyocytes are particularly sensitive to aging4. Clinical evidence and experimental studies have linked cellular senescence, the accumulation of senescent cells, and the production and release of SASP components to age-related cardiac pathologies, such as heart failure, myocardial ischemia, and infarction. However, the precise role of senescent cells under these conditions remains unclear, and for some cardiac diseases, such as cardiac fibrosis and MI, cellular senescence has been reported to have both detrimental and beneficial effects5,6.

Feng et al. 7 found that after eliminating senescent cells using the drug ABT263 or reducing the formation of senescent cells in Trp53 gene-knockout mice, the regenerative capacity of the neonatal heart post-injury was compromised, with fibrotic scars persisting and a decreased rate of cardiomyocyte proliferation. Concurrently, the research team also discovered that intravenous injection of CCN1 in adult mice post-MI promoted cellular senescence, reduced cardiac fibrosis, and improved cardiac pumping function. This indicates that cellular senescence is not only necessary for the of the neonatal heart but also holds significant therapeutic implications for the treatment of MI in adults. Similarly, studies have induced a P53-mediated transient senescence phenotype in animal models in the early stages post-cardiac injury, which coincides with cardiac phenomena8, whereas treatment with navitoclax to clear senescent cells or Tp53 knockout to inactivate them has been shown to impair cardiac. Thus, acute cellular senescence plays a crucial physiological role in heart development, and the transient presence of senescent cells in the heart under temporary stress may be beneficial, but the long-term accumulation of senescent cells can impair cardiac function and promote the progression of cardiac diseases9.

Both cellular repair and senescence involve changes in the cell cycle physiologically. Although the proliferation rate of adult cardiomyocytes is very low, extensive use of advanced lineage tracing methods has shown that the renewal of cardiomyocytes comes solely from existing ones10,11,12. Therefore, cell cycle regulators have emerged as potential therapeutic interventions to promote cardiomyocyte proliferation post-cardiac injury. For instance, a combination of cell cycle activators (including CDK1, CDK4, cyclin B1, and cyclin D1) has effectively improved cell division and cardiac function in mice, rats, and human cardiomyocytes post-MI13. In summary, cell cycle-related genes are closely associated with the fate of cardiomyocytes post-injury.

However, the mechanisms by which cell cycle-related genes are involved in the repair and senescence of cardiomyocytes post-MI are not yet fully understood. Investigating the fate and regulatory mechanisms of the heart post-MI, deeply understanding the interrelationship between cardiomyocyte repair and senescence, comprehensively monitoring and assessing cardiomyocytes post-infarction, and searching for effective treatment methods and drugs are important topics in the current field of cardiovascular medicine.

Methods

Data acquisition and preprocessing

We obtained chip data of MI post-event based on the Agilent platform from the GEO database, specifically the GSE161427 and GSE59867 datasets, serving as our test sets. The dataset GSE16142714 consists of myocardial tissue samples from 5 mice that underwent myocardial infarction for 3 days and 4 sham-operated mice. The GSE5986715 dataset consists of 101 peripheral blood mononuclear cell (PBMCs) samples from patients with myocardial infarction and 46 samples from healthy controls. Additionally, 494 SRGs were retrieved from the Cellage database. To compensate for the limited sample size of our test sets, we also utilized two new datasets of post-infarction myocardial samples from mice (GSE775:includes myocardial tissue samples from 18 mice with normal left ventricular function and 18 mice with left ventricular myocardial infarction)16 and blood samples from patients (GSE60993:includes peripheral blood samples obtained from patients with acute coronary syndrome who presented to the emergency department within 4 h of chest pain onset: 7 cases of ST-segment elevation myocardial infarction and 7 normal controls)17 as external validation sets. Figure 1 illustrates the flowchart of this study.

Fig. 1
figure 1

The flowchart of the study.

Full size image

Identification of differentially expressed genes and analysis of differential expression of post-infarction senescence-associated genes

Using R software (version 4.2.1), we standardized the upregulated genes in the GSE161427 dataset and performed differential expression analysis with the “limma” package. Volcano plots and heatmaps were created using “ggplot2” and "pheatmap," respectively. DEGsup were selected based on the criteria of |Log2(FC)|> 2 and P < 0.05. Gene Set Enrichment Analysis (GSEA) was conducted on the sorted genes using the “clusterprofiler” package with the KEGG database, selecting pathways with an absolute normalized enrichment score (NES) > 2, a false discovery rate (FDR) < 0.25, and a p < 0.05. The intersection of DEGsup and SRGs, termed DEupSGs, was visualized using an online Venn diagram tool (http://bioinformatics.psb.ugent.be/webtools/Venn/).

WGCNA of GSE59867

Weighted Gene Co-expression Network Analysis (WGCNA)18 is an effective method for exploring the connections between phenotypic traits and genes. We performed this analysis using the “WGCNA” package in R. Considering the requirement of WGCNA for a large sample size and to avoid the bias introduced by manually setting thresholds in differential analysis, we used the GSE59867 dataset for WGCNA. Initially, missing values and outliers were identified and removed from GSE59867. Subsequently, a scale-free R-squared of 0.85 and a soft threshold β of 14 were selected to meet the criteria for constructing a scale-free network, resulting in the formation of a topological overlap matrix (TOM). A hierarchical clustering dendrogram for module identification was generated, and modules with similarity > 0.3 were merged using dynamic pruning. Finally, the correlation between modules and MI was calculated, with a p < 0.05 considered a reliable association. The “clusterProfiler” package in R was used for Gene Ontology (GO) functional analysis and KEGG pathway enrichment analysis19,20,21 of modules with a positive correlation coefficient to MI, to assess the function, location, or pathways of key modules and genes. The results were visualized using bubble charts. The intersection of DEupSGs and core genes of the WGCNA purple module yielded 12 genes.

Machine learning algorithm (LASSO)

LASSO regression, a machine learning algorithm, was employed for variable selection to enhance predictive accuracy and for regularization to improve the predictive precision and interpretability of statistical models22. The analysis was conducted using the “glmnet” package in R.

Construction and analysis of protein–protein interaction (PPI) networks

The online String database (https://string-db.org) predicts interactions between proteins. To investigate the interactions among the 12 genes, these genes were uploaded to the String database with the minimum interaction score set to 0.40, using default settings for other parameters, and the results were exported. The exported files were imported into Cytoscape 3.10.0 for visualization. The PPI diagram was drawn with color depth representing the magnitude of differences, using the CytoHubba plugin to calculate and rank genes based on the Degree value.

Identification of the core target gene CDC6

The top 3 genes from the PPI analysis based on Degree value were intersected with genes selected by LASSO to identify potential core target genes. The 2 target genes was further validated using ROC curve analysis with external validation sets GSE775 and GSE60993, employing the pROC package in R to calculate the area under the curve (AUC). Genes with an AUC > 0.700 were selected as final core target genes from both validation sets. The dynamic changes of the core target gene CDC6 post-MI were observed, comparing the expression levels of CDC6 in MI mice and sham-operated mice at different time points.

Construction of in vivo MI model and ex vivo OGD model

In Vivo Model: Male C57BL/6 J mice (6–8 weeks old, 22 ± 2 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. Mice were housed under conditions of 23 ± 3 °C, 30–70% humidity, and a 12-h light/dark cycle. The experimental group underwent ligation of the coronary artery under anesthesia to induce acute MI; the control group was treated similarly without coronary artery ligation. Samples were collected at 24 h, 48 h, 1 week, 4 weeks, and 8 weeks post-MI. Mice were anesthetized with Avertin and their intraperitoneal venous blood was drawn. Subsequently, the mice were euthanized by quickly cutting the medulla oblongata with scissors and their hearts were removed.

Ex Vivo Model: 10 neonatal C57BL/6J mice, aged three days or less, were sourced from the Experimental Animal Center at Southern Medical University. Euthanasia by decapitation was performed under the approval of the Ethics Committee of Jinan University, and the hearts were aseptically extracted under sterile conditions on a super-clean bench. After trimming away connective tissue, the myocardial tissue was cut into 1 mm pieces and digested with pre-warmed Type II collagenase. The cell suspension was centrifuged, and primary fibroblasts and cardiomyocytes were separated after 90 min of culture. Primary cardiomyocytes were transferred to new 12-well plates, and their morphology and adhesion were observed after 48 h before proceeding with hypoxia intervention experiments. Samples were collected at 6 h, 8 h, 12 h, and 16 h post-hypoxia.

Transfection of overexpressed RNA plasmid vector and intervention with senescence inhibitor navitoclax (ABT263)

The overexpressed RNA plasmid vector was extracted and purified by Shandong Weizhen Biotechnology Co., Ltd. Transfection was performed when the density of primary neonatal cardiomyocytes reached 60–80%. After preparing the transfection solution, pure DMEM culture medium was used for transfection. The culture medium containing the transfection solution was removed after 8 h and replaced with complete culture medium. Hypoxia treatment was performed after 48 h of incubation. The hypoxia duration was 6 h, followed by mRNA level detection. Navitoclax (ABT-263), as a Bcl-2 family protein inhibitor, exerts anti-aging effects by inducing apoptosis to eliminate senescent cells7,23. In the ABT263 intervention group, 1umol/L ABT263 was added when the density of primary neonatal cardiomyocytes reached 40–50%. After 72 h of intervention, transfection was performed followed by subsequent hypoxia steps.

Validation of CDC6 gene and genes mediating cardiac repair using quantitative real-time PCR

Total RNA was extracted using Trizol reagent (Beyotime Biotechnology, Shanghai, China). The concentration of total RNA was measured, and RNA was reverse-transcribed into cDNA using a transcription kit. The diluted cDNA was processed with PerfectStart® Green qPCR SuperMix for qPCR. Primers were synthesized by Beijing Qingke Biotechnology Co., Ltd. and Shenggong Bioengineering (Shanghai) Co., Ltd.

Primers sequence:

  • CDC6: forward 5′-3′ cctcaaaccactctccgaatgta

    •    reverse 5′-3′ tgttcccatcaacttccgaaatg;

  • CNPY2: forward 5′-3′ caatccagatggcagccagtcag

    •    reverse 5′-3′ gtcacacacctcctcaagcaactc;

  • MFGE8: forward 5′-3′ ggagcaaggaagcagcaaggtc

    •    reverse 5′-3′ ggtgatgcggttatgccaggac

  • MRPS5: forward 5′-3′ gaaaggcaaagaaccgagcgattc

    •    reverse 5′-3′ gaggcggcagatggtgataatgg;

  • CPT1B: forward 5′-3′ gaggcacttctcagcatggtcatc

    •    reverse 5′-3′ tggacaggagacggacacagatag;

All experiments were repeated at least three times.

Statistical analysis

Statistical analysis and plotting were performed using GraphPad Prism 8.0. We employed paired t-tests and analysis of variance (ANOVA) as statistical methods. A p < 0.05 was considered statistically significant. Data are presented as the mean ± SD.

Results

Screening of differentially expressed genes

We downloaded the gene expression dataset GSE161427 from GEO and analyzed it using the “limma” package in R, identifying 1173 differentially expressed genes. Among these, 894 genes were significantly upregulated and 279 genes were significantly downregulated in the myocardial ischemia group. To further understand the signaling pathways involved in MI within the GSE161427 dataset, we performed GSEA enrichment analysis on all genes after differential analysis, sorted by Log2 (FC) value (Fig. 2). Then, the upregulated differentially expressed genes were displayed as a volcano plot and a heatmap (the heatmap displays the top 200 DEGs) (Fig. 3A and B). The results indicated that pathways such as cell cycle, oxidative phosphorylation, viral protein interaction with cytokine and cytokine receptor, propanoate metabolism, cytokine-cytokine receptor interaction, and citrate cycle were enriched in mice with MI. Concurrently, a total of 494 SRGs were obtained from the Cellage database and intersected with DEGsup, resulting in 50 common DEupSGs, which were displayed in a Venn diagram (Fig. 3C).

Fig. 2
figure 2

Enrichment analyses using gene set enrichment analysis (GSEA)of GSE161427. (A) Cell cycle. (B) Viral protein interaction with cytokine and cytokine receptor. (C) Oxidative phosphorylation. (D) Propanoate metabolism. (E) Cytokine-cytokine receptor interaction. (F) Citrate cycle. (G) Chemokine signaling pathway. (H) Phagosome. (I) NF-kappa B signaling pathway.

Full size image
Fig. 3
figure 3

Identification of DEGs and DEupSGs in MI cardiomyocyte cells in the GSE161427. (A) Volcano plot depicting the differential gene expression in myocardial tissue between coronary artery ligation and sham-operated mice (GSE161427). (B) Heatmap depicting the differential gene expression in myocardial tissue between coronary artery ligation and sham-operated mice (GSE161427). (C) Venn diagram of overlapping genes between the GSE161427 and Cellage database. DEGs up, upregulated genes in differentially expressed genes; Cellage, senescence-related genes.

Full size image

Weighted gene co-expression network analysis and trait module genes

To further identify target genes for MI, we conducted weighted gene co-expression network analysis (WGCNA) using another dataset, GSE59867. A clustering tree was constructed based on the PBMCs samples from 101 patients with acute MI within 4–6 days and 46 patients with stable coronary heart disease (Fig. 4A). One outlier was removed based on Pearson correlation coefficients. We found that the optimal soft threshold β was 14 when the scale-free R2 = 0.85. With β = 14, a scale-free model co-expression network was constructed (Fig. 4B), and the MergeCutHeight was used as the tree cutting height for module merging. Modules were merged based on similar expression patterns among patients. By setting the mergeCutHeight to 0.3 (Fig. 4C and D), dynamic tree cutting resulted in the construction of 11 gene co-expression modules (Fig. 4E). We selected modules that were positively correlated with AMI for further analysis, consistent with the DEGsup genes studied in the previous differential analysis. Among the positively correlated modules with MI, the red and purple modules were the most strongly associated with MI (with the red module having a Correlation Coefficient of 0.36, P = 8e−6; and the purple module having a Correlation Coefficient of 0.25, P = 0.003) (Fig. 4E). Subsequently, we applied the “clusterProfile” package in R to perform functional enrichment analysis on the core genes of the two modules, including GO and KEGG analyses. We found that the KEGG analysis of the purple module indicated that differentially expressed genes post-MI were mainly involved in the cell cycle, P53 signaling pathway, and cellular senescence, etc. According to the GO analysis results, they were involved in biological processes (BP), cellular components (CC), and molecular functions (MF), with the most significant biological processes being related to the mitotic and meiotic cell cycles (Fig. 4F and G). Therefore, the subsequent identification of target genes focused on the purple module. Through the intersection of DEupSGs and core genes of the purple module, 12 target genes were identified: CDC6, PLK1, KIF11, BUB1, UHRF1, RRM2, KIF20A, CCNA2, HELLS, ECT2, MYBL2 and BRIP1 (Fig. 5A).

Fig. 4
figure 4

The results of weighted gene co-expression network analysis (WGCNA) of GSE59867. (A) Sample clustering and trait heatmap of GSE59867(101 patients with acute MI within 4–6 days and 46 patients with stable coronary heart disease). (B) The relationship between the scale-free fit index, the mean connectivity and various soft-thresholding powers. (C) Cluster dendrogram. (D) Clustering dendrogram of genes, various colors represent different modules. (E) The 11 gene co-eigengene adjacency heatmap of the correlation between modules of MI(The purple module having a Correlation Coefficient of 0.25, P = 0.003). (F) Pathways of DESRGs in Kyoto encyclopedia of genes and genomes (KEGG) analysis (Figure adapted from KEGG database). (G) Pathways of DESRGs in Gene Ontology (GO) analysis.

Full size image
Fig. 5
figure 5

Integration of Machine Learning Algorithm (Lasso Regression) and Protein–Protein Interaction (PPI) Analysis. (A) Venn diagram illustrating the intersection between differentially expressed upregulated genes (DEupSGs) and the core genes of the purple module. (B) Coefficient plot of Lasso regression across various penalty parameter values. (C) Cross-validation process of Lasso regression, highlighting the point with the minimum error, which corresponds to a core gene count of five. (D) Protein–protein interaction network constructed from the top five hub genes identified through the analysis. (E) Venn diagram representing the overlap between Lasso regression core genes and the top five hub genes from the PPI network.

Full size image

LASSO

We then utilized LASSO with tenfold cross-validation to select the number of core genes corresponding to the Lambda (lambda.1se) within one standard error range, which resulted in 7 genes: PLK1, CDC6, ECT2, HELLS, MYBL2,UHRF1 and KIF20A (Fig. 5B and C).

PPI protein interaction analysis

Using the STRING database, we identified the protein interactions among the 12 genes. With the minimum interaction score set to 0.40, the PPI network analysis yielded a total of 12 hub nodes (56 edges, an average node degree of 9.33, and an average clustering coefficient of 0.936). Subsequently, the PPI network was imported into Cytoscape and visualized using the MCODE and cytohubba plugins. The nodes were ranked by degree, and the top 3 genes were selected, including CDC6, PLK1 and KIF11 (Fig. 5D).

Evaluation of the diagnostic efficacy of post-MI senescence-related core target genes

We obtained the intersection of the two methods, resulting in 2 target genes common to MI and cellular senescence: CDC6 and PLK1 (Fig. 5E).

Additionally, we selected the serum sample dataset GSE60993 from patients after MI and the myocardial tissue sample dataset GSE775 from mice to validate the diagnostic efficiency of the target genes. Among the 2 target genes, only the gene CDC6 showed an area under the curve (AUC) > 0.7 in both datasets (Fig. 6A–D). We then examined the expression levels of CDC6 at different time points after MI in the dataset GSE775 and displayed them using violin plots (Fig. 6E), finding that CDC6 expression significantly increased at 24 h post-MI compared to the sham-operated group and persisted for a period, showing no statistical difference with the sham-operated group at 8 weeks (p < 0.05).

Fig. 6
figure 6

Diagnostic value of target genes in the GSE60993 and GSE775 set and the expression levels of Cdc6 at different time points post-MI in GSE775. (A) CDC6 in the GSE60993(AUC = 0.776). (B) PLK1 in the GSE60993(AUC = 0.510). (C) Cdc6 in the GSE775(AUC = 0.725). (D) Plk1 in the GSE775(AUC = 0.510). (E) The expression level of the CDC6 gene in the GSE775 dataset across different time periods following MI.

Full size image

Construction of in vivo and ex vivo hypoxia models and core target gene validation

We searched the NCBI website and found that the CDC6 gene is highly expressed in fetal heart tissue (Fig. 7A), but almost not expressed in normal adult heart tissue after maturation (Fig. 7B). We also constructed a mouse MI model (Fig. 7C), collected myocardial tissue from mice at different time points, and validated the core target gene using qPCR. The results showed that CDC6 expression began to significantly increase at 24 h post-MI (p < 0.05), peaked between 48 h and 1 week, and then showed no statistical difference with the sham-operated group at 8 weeks (Fig. 7D). We also constructed an ex vivo OGD model of primary neonatal cardiomyocytes under hypoxia (Fig. 7C), and collected primary cardiomyocytes at different hypoxia times to validate the expression of the CDC6 gene. The results indicated that CDC6 gene expression was highest at 6 h post-OGD (Fig. 7E).

Fig. 7
figure 7

Analysis and validation of core target genes involved in mediating cardiac repair. (A) Expression levels of the CDC6 gene in normal human fetal tissues, as obtained from NCBI. (B) Expression levels of the CDC6 gene in normal adult human tissues, as sourced from NCBI. (C) Establishment of in vivo mouse myocardial infarction animal models and ex vivo primary neonatal cardiomyocyte OGD models. (D) Animal models were utilized to verify the expression levels of the CDC6 gene at various time points following myocardial infarction. (E) Cellular models confirmed the expression levels of the CDC6 gene under different durations of hypoxia. (F) Validation of genes mediating cardiac repair (Cnpy2, Mfge8, Gdfl5, Atf4, Mrps5, and Cpt1b).

Full size image

Validation of genes mediating cardiac repair

We detected the influence of CDC6 overexpression on the expression of genes mediating cardiac repair in the in vitro model under hypoxic conditions. After transfection with the Lipo2000 overexpression plasmid, primary neonatal cardiomyocytes were placed in a hypoxic incubator for 6 h. Compared with the NC group, the mRNA levels of genes related to cardiac repair were significantly increased (Fig. 7F).

To verify whether the increased expression of genes mediating early cardiac repair after MI is related to CDC6-induced cellular senescence, we intervened with the senescence inhibitor ABTA263 in primary neonatal cardiomyocytes overexpressing CDC6 exposed to the OGD environment. Compared with the group not treated with ABT263, the mRNA expression levels of genes related to cardiac repair decreased, indicating that the target gene CDC6 promotes the expression increase of genes mediating cardiac repair by promoting acute cellular senescence of cardiomyocytes in the early stage of hypoxia.

Discussion

This study utilized transcriptomic analysis of myocardial samples post-MI and peripheral blood mononuclear cells (PBMCs) to identify differentially expressed upregulated genes (DEupSGs) and 494 senescence-related genes (SRGs). GSEA and WGCNA analyses, along with functional enrichment results, indicated that both MI and cellular senescence processes are accompanied by changes in cell cycle-related genes. Employing machine learning and protein–protein interaction (PPI) network analysis, two common target genes were identified from the aforementioned methods. ROC curve analysis using two new external validation datasets post-MI suggested that the core target gene CDC6 holds significant diagnostic value for both MI and cellular senescence. Observations of CDC6 expression in cardiomyocytes at various times post-infarction revealed temporal differences, with an increase 24 h post-acute MI peaking at 48 h to 1 week, and then no difference from the sham surgery group by the 8 weeks. In vitro primary cardiomyocyte models under oxygen–glucose deprivation showed an upregulation of CDC6 gene expression after 6 h. Both in vivo and in vitro model results were consistent with bioinformatics analysis, indicating an upregulation of the cell senescence-associated CDC6 gene in the early stages post-MI.

CDC6, a key gene associated with the early regulation of the cell cycle, produces the CDC6 protein that forms the replication initiation complex, promoting the separation of DNA strands and the initiation of DNA replication, a protein essential for the start of DNA replication. Studies have shown that overexpression of CDC6 in primary cells may promote excessive DNA replication and induce senescence responses, akin to senescence induced by oncogene activation24. The Cellage database also indicates that the CDC6 gene has an effect on inducing cellular senescence. Although the mechanism is not yet clear, CDC6 plays a pivotal role in biological processes such as DNA replication, cell cycle regulation, and cellular senescence.

Over the past 2 decades, numerous studies have explored the regulatory mechanisms of the cell cycle post-MI25,26. The limited regenerative capacity of the adult mammalian heart is mainly due to the cessation of the myocardial cell cycle after birth. Evidence has confirmed that post-MI, some viable myocardial cells spontaneously reactivate their proliferative capacity and produce new myocardial cells in an attempt to repair damaged cardiac tissue27,28. Wang WE et al. published a study on the regenerative capacity of adult myocardial cells. The study found that the proliferation of mature myocardial cells may be a potential source of new myocardial cells. Researchers observed the proliferation of mature myocardial cells in in vivo experiments, proving that mature myocardial cells can re-enter the cell cycle through dedifferentiation, proliferation, and redifferentiation to form new myocardial cells29. Additionally, studies have confirmed that during the maturation process of myocardial cells after birth, their energy metabolism also undergoes changes, shifting from glycolysis to fatty acid metabolism reprogramming, accompanied by chromatin reconfiguration and exit from the cell cycle, setting an obstacle for adult heart regeneration30. In the early stages post-MI, some myocardial cells can spontaneously reverse this process, attempting to stimulate myocardial cells to re-enter the cell cycle and proliferate, thus allowing the heart to regenerate after ischemic injury31.

The spontaneous re-entry of mature myocardial cells into the cell cycle is a short-term adaptive response to cardiac injury or stress. Some data in MI research indicate that this repair phenomenon of myocardial cells is related to early acute cellular senescence7,8. However, this natural phenomenon of low-level myocardial cell proliferation post-MI in adult mammals rarely has a meaningful impact on cardiac repair and this ability cannot be spontaneously maintained long-term32. Researchers have found that after myocardial cell stress injury, myocardial cells with the ability to replicate DNA do not fully divide into two new cells but remain in a polyploid state33. This polyploid phenomenon is related to the injurious regenerative capacity of the heart, aiming to increase the volume and function of myocardial cells. Unfortunately, polyploid myocardial cells can change the epigenetic and transcriptional characteristics of myocardial cells, making them immature and unstable, thereby accelerating the accumulation of senescent myocardial cells. Moreover, the abnormal increase in the number and volume of myocardial cells may also lead to the disorder of myocardial cell arrangement and connections, affecting the structure and function of myocardial tissue. At the same time, the spontaneous induction of myocardial cell glycolysis increase and the decrease in fatty acid oxidation rate due to acute ischemia and hypoxia of the myocardium reduce the metabolic flexibility of myocardial cells. This energy deficit is commonly seen in elderly patients with heart failure, leading to impaired cardiac contractile function and the basis for the deterioration of heart function34.

This may explain the results of this study, where CDC6 expression is transiently elevated post-MI, inducing acute myocardial cell senescence, while also promoting some viable myocardial cells to re-enter the cell cycle to enhance myocardial repair post-infarction. Subsequently, the expression level of CDC6 decreases, the cell cycle is halted, energy metabolism is reshaped, and polyploid myocardial cells typically lose normal cell cycle control, leading to the accumulation of senescent cells and impaired heart function35. To verify this hypothesis, we constructed an in vitro OGD cardiomyocyte model and intervened with the senescence inhibitor ABT263 in primary cardiomyocytes overexpressing CDC6. We were pleasantly surprised to find that compared to the NC group, overexpression of CDC6 significantly increased the mRNA levels of genes related to heart repair and the intervention with the drug ABT263 reduced the expression of these genes. This result proves that the early increase in CDC6 post-OGD in myocardial cells may accelerate the acute senescence of myocardial cells, promoting an increase in the expression of genes related to heart repair.

It is worth noting that no study to date has investigated the interaction of CDC6 in the early stages of MI in cell repair and senescence. Our research explores the relationship between MI and cellular senescence, discovering and validating the significant early upregulation of the myocardial cell senescence-related gene CDC6 in in vivo and in vitro models of MI. This overexpression-induced acute myocardial cell senescence may have a beneficial impact on myocardial cell. In future research, we will focus on exploring the mechanism by which the cell cycle gene CDC6 is involved in the early stages of MI in myocardial cell senescence and repair and further investigate what determines the balance when the effects of cell senescence coexist.

Our study has some notable limitations. Firstly, our research only extracted datasets from the Cellage database; in fact, there should be more SRGs that have not yet been identified. Secondly, our sample size is relatively small, especially for myocardial tissue samples, although we have constructed animal models to validate and compensate for this limitation. Lastly, the gene CDC6 we selected has not yet had sufficient research to confirm its role in regulating cell repair and senescence in the early stages of MI, therefore, more in vitro and in vivo studies are needed to confirm these findings.