Identification of iron metabolism-related genes in the circulation and myocardium of patients with sepsis via applied bioinformatics analysis

Abstract

Background: Early diagnosis of septic cardiomyopathy is essential to reduce the mortality rate of sepsis. Previous studies indicated that iron metabolism plays a vital role in sepsis-induced cardiomyopathy. Here, we aimed to identify shared iron metabolism-related genes (IMRGs) in the myocardium and blood monocytes of patients with sepsis and to determine their prognostic signature.

Methods: First, an applied bioinformatics-based analysis was conducted to identify shared IMRGs differentially expressed in the myocardium and peripheral blood monocytes of patients with sepsis. Second, Cytoscape was used to construct a protein-protein interaction network, and immune infiltration of the septic myocardium was assessed using single-sample gene set enrichment analysis. In addition, a prognostic prediction model for IMRGs was established by Cox regression analysis. Finally, the expression of key mRNAs in the myocardium of mice with sepsis was verified using quantitative polymerase chain reaction analysis.

Results: We screened common differentially expressed genes in septic myocardium and blood monocytes and identified 14 that were related to iron metabolism. We found that HBB, SLC25A37, SLC11A1, and HMOX1 strongly correlated with monocytes and neutrophils, whereas HMOX1 and SLC11A1 strongly correlated with macrophages. We then established a prognostic model (HIF1A and SLC25A37) using the common differentially expressed IMRGs. The prognostic model we established was expected to better aid in diagnosing septic cardiomyopathy. Moreover, we verified these genes using datasets and experiments and found a significant difference between the sepsis and control groups.

Conclusion: Common differential expression of IMRGs was identified in blood monocytes and myocardium between sepsis and control groups, among which HIF1A and SLC25A37 might predict prognosis in septic cardiomyopathy. The study may help us deeply understand the molecular mechanisms of iron metabolism and aid in the diagnosis and treatment of septic cardiomyopathy.

Keywords: diagnostic biomarkers; iron metabolism; peripheral blood monocytes; sepsis; septic cardiomyopathy.

Figure 1

Screening of the differentially expressed iron metabolism-related genes (IMRGs). (A) Principal components analysis for GSE79962. (B) Volcano plot of GSE79962. (C) Heatmap of the common differentially expressed IMRGs in septic myocardiopathy and controls. (D) Venn diagram showing the overlap of genes among differentially expressed genes (DEGs) of GSE79962, GSE46955, and IMRGs. (E) Principal components analysis of GSE46955. (F) Volcano plot of GSE46955. (G) Heatmap of the common differentially expressed IMRGs in blood monocytes of sepsis and control groups.

Figure 2

Enrichment analysis for the differentially expressed genes (DEGs) in circulation and the myocardium. (A) Gene Ontology (GO) enrichment analysis, (B) Bubble plot of enriched GO terms in GSE79962. (C) GO enrichment analysis, (D) Bubble plot of enriched GO terms in GSE46955. (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis in GSE79962. (F) KEGG pathway analysis in GSE46955. Abbreviations: BP, biological process; CC, cellular component; MF, molecular function.

Figure 3

Protein–protein interaction (PPI) networks in common DEGs of iron metabolism. (A) PPI network formed by common differentially expressed IMRGs. (B–D) The top three MCODE complexes. (E) The top 10 hub genes identified by Cytohubba. (F–H) Enrichment analysis performed by Metascape.

Figure 4

Landscape of immune infiltration in the myocardium of patients with sepsis and controls. (A) Heatmap of the proportions of the 28 immune cell types. (B) Box plots of immune cell proportions. (C) Correlation analysis of immune cells. (D) Correlation between common differentially expressed IMRGs and immune cells.

Figure 5

Prognostic prediction model based on common IMRGs. (A) Kaplan–Meier survival curves of 28-day mortality between high- and low-risk groups (p = 2.052e-03). (B) The risk score analysis between high-risk and low-risk groups and survival status analysis. (C) The receiver operating characteristic (ROC) curve of the two-gene model in patients with sepsis and controls (AUC:0.708). (D) The expression of the two genes between sepsis and healthy samples.

Figure 6

Validation of SLC25A37 and HIF1A in a mouse model of LPS-induced sepsis using quantitative PCR analysis. (A) Transthoracic echocardiography between sepsis and control groups. (B) Hematoxylin and eosin staining. (C) Relative expression of SLC25A37 and HIF1A between sepsis and control groups in mice, along with the expression of β-actin as an internal standard for normalization. Asterisks indicate a significant statistical p-value calculated using Student’s t-test (*p < 0.05; **p < 0.01; ***p < 0.001).

Figure 7

Relationship between the genes of the prognostic model and various diseases based on the comparative toxicogenomics database (CTD). Interactions between HIF1A (A) and SLC25A37 (B) and cardiovascular and inflammatory diseases, respectively.