AUTHOR=Wei Jingjing , Yu Peng , Hu Yucai , Qiao Lijie , Li Bin , Li Haitao , Xie Shiyang , Dong Zhengwei , Wang Aolong , Zhang Yilin , Wang Xinlu , Wang Yongxia , Zhu Mingjun 

TITLE=CLU, FOS, and CXCL8 as diagnostic biomarkers for heart failure progression post-acute myocardial infarction: an integrated RNA-Seq and multi-machine learning study

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2025.1611668

DOI=10.3389/fcvm.2025.1611668

ISSN=2297-055X

ABSTRACT=BackgroundHeart failure (HF) secondary to acute myocardial infarction (AMI) remains a public health concern. Peripheral blood mononuclear cells (PBMCs) are the essential initiators of heart failure after myocardial infarction (HFpAMI). We aimed to identify PBMCs-related critical genes as diagnostic biomarkers for HFpAMI and analyze the immune infiltration patterns.MethodsDifferential expression genes (DEGs) from PBMCs microarray data of AMI with or without HF were identified. Functional enrichment analysis was used to explore the biological roles of DEGs. Subsequently, candidate biomarkers were identified using machine learning and the MCODE plugin, with ROC used to describe the accuracy. CIBERSORT was utilized to investigate immune infiltration. Multi-level validation of our findings was conducted, including RNA-seq profiling of the external cohort, RT-qPCR, and flow cytometry analyses on PBMCs samples.ResultsIn the comparison between 30 HFpAMI and 34 non-HF samples, 27 DEGs were identified. Functional enrichment analysis suggested that DEGs may be involved in the pathological process of HFpAMI by participating in immune-inflammatory response. Employing machine learning and MCODE assessment, we identified three robust potential biomarkers (CLU, FOS, and CXCL8). Immunological analysis revealed a marked increase in neutrophils and decrease in CD4T cells. In the external validation cohort, RNA-seq analysis demonstrated consistent upregulation of CLU, FOS, and CXCL8 in HFpAMI compared to non-HF controls. RT-qPCR and flow cytometry further corroborated these expression trends and their correlations with neutrophil infiltration, CD4T cells and M2 macrophage concentration reductio, aligning with bioinformatics predictions. ROC analysis validated the diagnostic efficacy of these biomarkers, with CLU exhibiting the highest AUC (0.833, 95% CI: 0.679–0.988), followed by FOS (0.809, 95% CI: 0.64–0.977) and CXCL8 (0.802, 95% CI: 0.635–0.970).ConclusionsSignificantly upregulated DEGs, including CLU, FOS, and CXCL8, might be served as novel diagnostic biomarkers for HFpAMI, and dysregulated immune infiltration hinted possible the immune system intervention point in the setting of HFpAMI.