Acute myocardial infarction (AMI) remains a major cause of cardiovascular death despite advances in reperfusion and secondary prevention (1). AMI typically occurs on the background of coronary atherosclerosis and is frequently triggered by plaque disruption, thrombosis, and abrupt coronary occlusion. Although cardiac troponins are central to diagnosis, there is a continuing need to better understand the molecular mechanisms that drive the transition from stable coronary artery disease (sCAD) (2) to acute coronary events and to identify biomarkers that reflect immuno-inflammatory activity.

Chest pain is the earliest and most common clinical manifestation of AMI. Diagnosis is typically based on characteristic clinical symptoms, elevated levels of cardiac troponin I (cTnI), and electrocardiographic (ECG) changes, such as ST-T abnormalities. However, cTnI elevation often takes 3–4 h to become detectable. While myoglobin levels rise more rapidly, they are influenced by various non-cardiac factors, making myoglobin a less specific biomarker and more suitable as an auxiliary diagnostic indicator. Moreover, in NSTEMI cases, ECG often fails to show significant ST-segment elevation. Since AMI itself triggers a local and systemic inflammatory response (3–5), identifying novel biomarkers capable of detecting this immune-inflammatory reaction at an earlier stage could potentially improve the early prediction and diagnosis of AMI, particularly in patients with sCAD, before it progresses to a full-blown infarction. Such markers could also enable more precise and timely therapeutic interventions in the early stages of AMI. Sterile inflammation is initiated immediately after myocardial ischemia through the release of damage-associated molecular patterns (DAMPs) that activate pattern-recognition receptors and downstream pathways such as TLR/NF-kappaB signaling, inflammasome activation, and cytokine cascades (6, 7). Neutrophils and monocytes are rapidly mobilized and recruited to the infarcted myocardium, where they contribute to tissue injury in the early phase but also participate in debris clearance and subsequent repair. Clinically and experimentally, the balance and timing of these immune responses are strongly linked to infarct size, healing quality, ventricular remodeling, and heart failure risk (8).

Importantly, AMI is not immunologically homogeneous. Transcriptomic and single-cell studies have suggested that circulating immune cell subsets and inflammatory gene programs vary across patients and may correlate with plaque phenotypes and clinical outcomes (9). Therefore, integrating immune infiltration estimation with gene co-expression and clustering analyses may help identify immune-related genes and molecular subtypes relevant to AMI progression and patient stratification. In this study, we integrated two microarray datasets to identify immune-related differentially expressed genes and construct an AMI-associated co-expression network. We combined differential expression analysis, WGCNA, immune cell deconvolution, protein-protein interaction (PPI) analysis, and consensus clustering to determine hub immune genes and molecular subtypes. Finally, we performed preliminary validation by RT-qPCR in a clinical cohort and by re-analysis of an independent dataset.

In this study, we integrated PBMC transcriptomic datasets to identify immune-related genes and molecular subtypes associated with AMI. By combining differential expression analysis, WGCNA, immune cell deconvolution, and PPI network analysis, we identified six hub immune genes (CSF3R, CD14, AQP9, S100A9, SLC11A1, and IL1RN) that were strongly linked to neutrophil and monocyte fractions. Our enrichment analyses highlighted leukocyte chemotaxis/migration and cytokine signaling, consistent with the established concept that AMI triggers rapid systemic and local immune activation. Neutrophils and monocytes are among the earliest responders and can amplify injury through proteases, reactive oxygen species, and inflammatory cytokines, while later reparative programs facilitate resolution and scar formation. Therefore, circulating immune signatures may capture both disease activity and heterogeneity in inflammatory states among AMI patients.

The molecular subtypes identified by consensus clustering provide an additional perspective on AMI heterogeneity. Cluster 3 exhibited consistently higher hub gene expression, suggesting a subgroup with more pronounced innate immune activation, whereas Cluster 2 showed the lowest expression. Future studies integrating clinical outcomes, imaging features, and longitudinal sampling are needed to determine whether these transcriptomic subtypes correspond to differences in plaque phenotype, reperfusion injury, infarct healing, or prognosis.

We then conducted a clinical data analysis of the six selected genes. First, we enrolled clinically confirmed AMI patients and matched them with sCAD controls. The results showed that three genes—AQP9, S100A9 and SLC11A1—were significantly upregulated in AMI patients, suggesting that they may be involved in the pathophysiological processes of myocardial infarction. However, the expression levels of CSF3R, IL1RN and CD14 showed no significant difference between the AMI and sCAD groups, which was not entirely consistent with our previous bioinformatics predictions. We speculate that there might be the following reasons: ① Differences in data acquisition methods and methodological distinctions in detection approaches have led to inconsistencies between bioinformatics analysis data and validation cohort. This discrepancy may be attributed to the standardized processing of high-throughput sequencing data, whereas RT-qPCR relies on amplification signals from specific primers, potentially leading to deviations in the expression trends of certain genes between the two methods.② It is undeniable that our sample size is relatively small, which may cause errors in the research results. ③ Furthermore, since RT-qPCR typically analyzes specific samples while high-throughput data may encompass multiple patients, physiological variations among individuals could also contribute to observed differences. In future studies, we aim to refine the standardization of data collection procedures to enhance the robustness and representativeness of the experimental design.

Among the identified three hub genes (AQP9, S100A9 and SLC11A1), AQP9 and S100A9 have been repeatedly implicated in inflammatory responses relevant to coronary artery disease and AMI. AQP9 is expressed in myeloid cells and has been associated with neutrophil and macrophage function as well as lipid metabolism, suggesting potential links between metabolic dysregulation and inflammation (22). Another study confirmed that AQP9 is a direct target gene of miR-212. Through inhibition of the PI3K/Akt pathway, miR-212 reduces AQP9 gene expression, thereby modulating the inflammatory response and myocardial damage (23). Previous studies have shown that AQP9 plays a critical role in lipid metabolism. In the liver, AQP9-mediated glycerol uptake is a key step in glycerol synthesis, while in adipose tissue, AQP9 is involved in lipid degradation. Mice with AQP9 gene deficiency exhibit elevated plasma triglyceride and cholesterol levels, highlighting its role in lipid homeostasis (24, 25). Given the direct link between lipid metabolism and atherosclerosis, AQP9 may contribute to atherosclerotic progression by regulating both lipid metabolism and inflammatory responses (26).

S100A9, often forming the S100A8/A9 complex, is a prototypic alarmin abundantly expressed in neutrophils and monocytes (27, 28) and can promote cytokine production through receptors such as TLR4 and RAGE (29). Studies have shown that in patients with bacterial infections, S100A9 exerts dual effects: on the one hand, it binds to calmodulin in cardiomyocytes, thereby reducing myocardial contractility; on the other hand, it induces neutrophil chemotaxis through the RAGE receptor and triggers an inflammatory cascade via the TLR/NF-κB and MYD88 signaling pathways (30, 31). Building on this, another study demonstrated that S100A9 is associated with myocardial injury following AMI. The use of S100A9 inhibitors was found to regulate the expression of proteins involved in five biological processes: leukocyte adhesion, cardiomyocyte apoptosis, intrinsic apoptotic signaling, sarcomere organization, and myocardial hypertrophy. By downregulating the expression of inflammatory markers such as MOESN (32), S100A9 inhibition reduces the inflammatory response in AMI. S100A9 also plays a pivotal role in atherosclerosis development, which is recognized as a chronic inflammatory process. S100A9 promotes neutrophil adhesion to endothelial cells and enhances macrophage uptake of LDL-C, thereby accelerating the inflammatory cascade (33). A study using S100A9-targeting antibodies demonstrated a reduction in atherosclerosis-associated inflammation, suggesting that this approach could potentially serve as an alternative to lipid-lowering therapies in the future, offering a novel strategy for the clinical treatment of coronary artery disease (34).

SLC11A1 (also known as NRAMP1) emerged as a hub gene in our integrative analysis. SLC11A1 encodes a proton-coupled divalent metal ion transporter that shapes myeloid-cell iron/manganese handling and inflammatory activation (35–37). Although SLC11A1 has been implicated in macrophage/foam-cell biology in atherosclerosis and has appeared in transcriptomic biomarker signatures of coronary artery disease (38, 39), its association with AMI has not been systematically characterized. In our study, SLC11A1 was consistently upregulated in AMI and correlated with myeloid cell fractions, supporting its potential relevance to acute coronary inflammation. Mechanistically, SLC11A1 encodes a proton-coupled divalent metal ion transporter predominantly expressed in macrophages and localized to late endosomal/lysosomal membranes. By regulating iron and manganese handling within myeloid cells, SLC11A1 may influence oxidative stress, inflammatory signaling, and macrophage activation states (37). Given the importance of monocyte-to-macrophage differentiation in plaque inflammation and post-infarction repair, elevated SLC11A1 expression in AMI may reflect heightened innate immune activation and altered metal metabolism during acute coronary events. This hypothesis warrants direct experimental testing in future work, including functional assays in macrophages and in vivo models.

Several limitations should be noted. First, the discovery datasets were derived from PBMCs rather than myocardial tissue, and CIBERSORT estimates may not fully capture the complexity of circulating immune states. Second, our RT-qPCR validation cohort was small and lacked healthy controls; although we used an external dataset for complementary validation, larger multi-center cohorts are needed. Third, we validated mRNA expression only; protein-level validation and functional experiments were not performed. Finally, although our analyses suggest associations with immune infiltration, causality cannot be inferred from transcriptomic correlations alone.

In summary, we identified immune-related hub genes and molecular subtypes in AMI using an integrated bioinformatics framework and preliminary validation. AQP9, S100A9, and SLC11A1 were consistently upregulated in AMI and may serve as candidate biomarkers reflecting innate immune activation. These results provide a foundation for mechanistic studies and for developing immune-informed strategies to stratify and potentially intervene in AMI.