Immunogenic cell death-related biomarkers in heart failure probed by transcriptome and single-cell sequencing

Abstract

Background: Heart failure (HF) represents the terminal stage of various cardiovascular disorders, with immunogenic cell death (ICD) potentially influencing HF progression through modulation of immune cell activity. This study aimed to identify ICD-associated biomarkers in patients with HF and explore their underlying mechanisms.

Methods: Data from GSE57338, GSE3586 and GSE5406 were retrieved from the Gene Expression Omnibus (GEO) database. Differential expression analysis and weighted gene co-expression network analysis (WGCNA) were employed to identify candidate genes, followed by enrichment analysis and Protein-Protein Interaction (PPI) network construction. Candidate biomarkers were selected using two machine learning approaches and validated for expression levels, with receiver operating characteristic (ROC) curve analysis determining the final biomarkers. A nomogram model was built based on the biomarkers, followed by molecular regulatory network analysis, gene set enrichment analysis (GSEA), immune infiltration assessment, and drug prediction. Additionally, key cells were selected for pseudo-time and cell communication analysis using the GSE183852 dataset. Next, pseudotemporal analysis was also performed on key cell subpopulations. Real-time quantitative PCR (RT-qPCR) was employed to validate the biomarkers.

Results: Three biomarkers, CD163, FPR1, and VSIG4, were identified as having significant diagnostic value for HF. GSEA revealed their enrichment in ribosomal and immune cell-related pathways. These biomarkers were notably correlated with CD8 T cells and M2 macrophages. Carbachol and etynodiol were predicted to interact with all three biomarkers. Single-cell RNA sequencing identified nine cell types, with expression of the biomarkers confined to monocytes and macrophages. Strong cell communication was observed between these cell types and fibroblasts. Expression of CD163 and VSIG4 decreased over time in monocytes and macrophages, whereas FPR1 showed an upward trend. In addition, the expression levels of CD163 and VSIG4 increased in subpopulations of monocytes and macrophages, whereas FPR1 showed a decreasing trend. RT-qPCR results confirmed significant down-regulation of CD163, FPR1, and VSIG4 in patients with HF and animal models.

Conclusions: This study identified and validated three ICD-related biomarkers in HF-CD163, FPR1, and VSIG4-offering a novel theoretical foundation for the clinical diagnosis and treatment of HF.

Keywords: biomarker; heart failure; immunogenic cell death; monocytes and macrophages; single-cell RNA sequencing analysis.

Figure 1

Acquisition of key module genes. (a) Co-expression module identification. (b) Heatmap showing the correlation between modules and phenotypes.

Figure 2

Identification and enrichment analysis of candidate genes and PPI. (a) Venn diagram depicting the overlap between differentially expressed genes (DEGs) and key module genes. (b) Gene Ontology (GO) enrichment analysis results. (c) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis results. (d) Protein-Protein Interaction (PPI) network. (e) Upset plot representing the PPI network.

Figure 3

Machine learning for candidate biomarker screening. (a) Results of the Support Vector Machine-Recursive Feature Elimination (SVM-RFE) model. (b) Bar chart depicting the MeanDecreaseGini scores for candidate genes. (c) Correlation analysis of candidate biomarkers. *P < 0.05, ***P < 0.001.

Figure 4

Diagnosis and evaluation of biomarkers. (a) Expression levels of candidate genes in the training set, with the horizontal axis representing genes and the vertical axis indicating gene expression levels (Wlicoxon rank sum test, ****P < 0.0001). (b) Expression levels of candidate genes in the validation set, with similar axis labels and significance markers (Wlicoxon rank sum test, *P < 0.05, **P < 0.01, ns: P > 0.05). (c) ROC curve analysis of the VSIG4 biomarker in the validation set. (d) ROC curve analysis of the CD163 biomarker in the training set. (e) ROC curve analysis of the FPR1 biomarker in the training set. (f) ROC curve analysis of the VSIG4 biomarker in the training set. (g) ROC curve analysis of the FPR1 biomarker in the validation set. (h) ROC curve analysis of the CD163 biomarker in the validation set.

Figure 5

Functional analysis of biomarkers. (a) GSEA enrichment analysis of the CD163 gene. (b) GSEA enrichment analysis of the FPR1 gene. (c) GSEA enrichment analysis of the VSIG4 gene.

Figure 6

Analysis of immune cell infiltration. (a) Box plot illustrating immune cell infiltration differences (Wlicoxon rank sum test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: P > 0.05). (b) Correlation of differential immune cell types (*P < 0.05, **P < 0.01, ***P < 0.001). (c) Correlation between biomarkers and differential immune cells (*P < 0.05, **P < 0.01, ***P < 0.001).

Figure 7

Molecular regulatory network and drug prediction. (a) Regulatory network between the CD163 gene and miRNAs, where pink nodes represent biomarkers and blue nodes represent interacting miRNAs. (b) Regulatory network between the CD163 gene and transcription factors (TFs). (c) Diagram of drug prediction for biomarkers.

Figure 8

Single-cell RNA sequencing analysis. (a) UMAP plot for different cell types. (b) Expression profile plot of biomarkers. (c) Pathway enrichment analysis of cell subtypes.

Figure 9

Cell subtype communication analysis. (a-b) Diagrams of cell communication results. (c) Results of pseudotime and state analysis in single-cell pseudotemporal analysis. Darker colors indicate the most advanced state of development, while lighter colors indicate more mature development. The cells were divided into 9 different periods according to their developmental state. (d) Cell pseudotime analysis of Seurat clusters. Different cell clusters presented different positions at various nodes of the developmental trajectory. (e) Dynamic atlas of biomarkers.

Figure 10

Clinical and animal validation of hub genes. (a) Expression of CD163, FPR1, and VSIG4 in peripheral blood mononuclear cells of patients with HF and NHF individuals (Unpaired t test, **P < 0.01, ***P < 0.001). (b-c) Echocardiograms of the HF rat model and sham group (Unpaired t test, ***P < 0.001). (d) The ratios of heart weight to body weight and lung weight to tibia length in rat model (Unpaired t test, *P < 0.05). (e-f) HE and Masson staining of rat hearts. (g) Collagen volume fraction(%) calculated by Masson staining (Unpaired t test, ***P < 0.001). (h) Expression of CD163, FPR1, and VSIG4 in the hearts of HF and sham groups (Unpaired t test, *P < 0.05, **P < 0.01).