Integrated analysis of bulk transcriptomics and single-cell transcriptomics revealed the association characteristics between heart failure and β-hydroxybutyrylation-related genes

BACKGROUND: As a major cause of global mortality, heart failure (HF) has mechanisms that remain elusive. Recent studies suggest β-hydroxybutyrylation (Kbhb) may contribute to its pathogenesis, driving this investigation into Kbhb-associated HF biomarkers. METHODS: We collected HF-associated data and information on Kbhb-related genes (KRGs) from publicly available databases and published studies. Potential biomarker candidates were identified by determining the overlap between differentially expressed genes (DEGs) and KRGs. Subsequently, machine learning algorithms, ROC curve evaluation, and expression verification were applied to identify robust biomarkers. To investigate the biological roles of these biomarkers in HF, we performed functional enrichment analysis, assessed the immune microenvironment, predicted regulatory networks and therapeutic compounds, and conducted molecular docking simulations. Furthermore, we utilized scRNA-seq data to determine critical cell populations and examine the expression patterns of biomarkers across different cellular contexts. Additionally, we used serum samples from four heart failure patients and four healthy controls to validate the protein expression levels of the identified biomarkers using enzyme-linked immunosorbent assay (ELISA). RESULTS: Three genes-HMOX2, H2AFZ, and KPNA2-were determined as diagnostic biomarkers for HF. Functional enrichment analyses demonstrated that these biomarkers exhibited significant associations with essential biological pathways, such as proteasome-mediated processes, oxidative phosphorylation mechanisms, and amino acid metabolic pathways. Furthermore, neutrophil infiltration showed significant positive correlations with both H2AFZ (cor = 0.42, p < 0.05) and KPNA2 (cor = 0.39, p < 0.05). Moreover, regulatory factor SP1 was predicted to target both KPNA2 and HMOX2. Thereafter, molecular docking confirmed potent binding of Bisphenol A to all 3 biomarkers, especially to HMOX2 (binding energy = -7.3 kcal/mol). Interestingly, T cells were identified as the key cell type in HF, and biomarkers exhibited dynamic changes during their differentiation. ELISA verification further confirmed that protein levels of HMOX2(P< 0.05) and KPNA2(P< 0.05) were significantly downregulated in HF patients, consistent with bioinformatics predictions. By contrast, H2AFZ exhibited a decreasing trend but without statistical significance (p> 0.05). CONCLUSION: HMOX2 and KPNA2 were identified as Kbhb-related biomarkers in HF, while T cells served as a key cell type in the disease. Meanwhile, this study provided novel therapeutic targets for HF patients.