Identification of ferroptosis-related genes in syncytiotrophoblast-derived extracellular vesicles of preeclampsia

Abstract

Preeclampsia (PE), defined as new-onset hypertension and multi-organ systemic complication during pregnancy, is the leading cause of maternal and neonatal mortality and morbidity. With extracellular vesicles research progresses, current data refers to the possibility that ferroptosis may play a role in exosomal effects. Evidence has suggested that ferroptosis may contribute to the pathogenesis of preeclampsia by bioinformatics analyses. The purpose of the current study is to identify the potential ferroptosis-related genes in syncytiotrophoblast-derived extracellular vesicles (STB-EVs) of preeclampsia using bioinformatics analyses. Clinical characteristics and gene expression data of all samples were obtained from the NCBI GEO database. The differentially expressed mRNAs (DE-mRNAs) in STB-EVs of preeclampsia were screened and then were intersected with ferroptosis genes. Functional and pathway enrichment analyses of ferroptosis-related DE-mRNAs in STB-EVs were performed. Ferroptosis-related hub genes in STB-EVs were identified by Cytoscape plugin CytoHubba with a Degree algorithm using a protein-protein interaction network built constructed from the STRING database. The predictive performance of ferroptosis-related hub genes was determined by a univariate analysis of receiver operating characteristic (ROC). The miRNA-hub gene regulatory network was constructed using the miRwalk database. A total of 1976 DE-mRNAs in STB-EVs were identified and the most enriched item identified by gene set enrichment analysis was signaling by G Protein-Coupled Receptors (normalized enrichment score = 1.238). These DE-mRNAs obtained 26 ferroptosis-related DE-mRNAs. Ferroptosis-related DE-mRNAs of gene ontology terms and Encyclopedia of Genes and Genomes pathway enrichment analysis were enriched significantly in response to oxidative stress and ferroptosis. Five hub genes (ALB, NOX4, CDKN2A, TXNRD1, and CAV1) were found in the constructed protein-protein interaction network with ferroptosis-related DE-mRNAs and the areas under the ROC curves for ALB, NOX4, CDKN2A, TXNRD1, and CAV1 were 0.938 (CI: 0.815-1.000), 0.833 (CI: 0.612-1.000), 0.875 (CI: 0.704-1.000), 0.958 (CI: 0.862-1.000), and 0.854 (CI: 0.652-1.000) in univariate analysis of ROC. We constructed a regulatory network of miRNA-hub gene and the findings demonstrate that hsa-miR-26b-5p, hsa-miR-192-5p, hsa-miR-124-3p, hsa-miR-492, hsa-miR-34a-5p and hsa-miR-155-5p could regulate most hub genes. In this study, we identified several central genes closely related to ferroptosis in STB-EVs (ALB, NOX4, CDKN2A, TXNRD1, and CAV1) that are potential biomarkers related to ferroptosis in preeclampsia. Our findings will provide evidence for the involvement of ferroptosis in preeclampsia and improve the understanding of ferroptosis-related molecular pathways in the pathogenesis of preeclampsia.

Figure 1.

Schematic presentation of the analysis process.

Figure 2.

The PCA analysis, the volcano plot of the identified DE-mRNAs and representative results of GSEA analysis in the gene expression data in GSE190971. (A) The visualization of principal component analysis for GSE190971. (B) Volcano plot of the identification of DE-mRNAs, including 1880 upregulated and 96 downregulated genes, Red, upregulation; Blue, downregulation. (C) The most significant enriched gene set correlated with the PE group was signaling by Gpcr (NES = 1.238, P.adj < .001, FDR < 0.001). (D) The second significant enriched gene set correlated with the PE group was matrisome associate (NES = 1.229, P.adj < .001, FDR < 0.001). DE-mRNAs = differentially expressed mRNAs, FDR = false discovery rate, Gpcr = G-protein coupled receptor, GSEA = gene set enrichment analysis, NES = normalized enrichment score, PCA = principal component analysis.

Figure 3.

Ferroptosis-related DE-mRNAs between PE STB-EVs and normal STB-EVs samples. (A) Volcano plot of 216 ferroptosis-related genes including 23 upregulated and 3 downregulated genes, Red, upregulation; blue, downregulation. (B) Venn diagram to screen ferroptosis-related DE-mRNAs. (C) The heat map of 26 ferroptosis-related DE-mRNAs between PE STB-EVs and Normal STB-EVs samples. DE-mRNAs = differentially expressed mRNAs, PE = preeclampsia, STB-EVs = syncytiotrophoblast-derived extracellular vesicles.

Figure 4.

GO functional analysis of 26 ferroptosis-related DE-mRNAs, Chord (A) and Bubble plot (B) of enriched GO terms of 26 ferroptosis-related DE-mRNAs, Y-axis: name of GO items; X-axis: percentage of the number of genes assigned to a term mong the total number of genes annotated in the network; Bubble size, number of genes assigned to a pathway; Color: enriched −log10(P-value). DE-mRNAs = differentially expressed mRNAs, GO = gene ontology.

Figure 5.

KEGG of 26 ferroptosis-related DE-mRNAs, Chord (A) and Bubble plot (B) of enriched KEGG terms of 26 ferroptosis-related DE-mRNAs, Y-axis: name of the KEGG signaling pathway; X-axis: percentage of the number of genes assigned to a term among the total number of genes annotated in the network; Bubble size, number of genes assigned to a pathway; Color: enriched −log10(P-value). DE-mRNAs = differentially expressed mRNAs, KEGG = Kyoto Encyclopedia of Genes and Genomes analysis.

Figure 6.

Spearman correlation analysis of the 26 ferroptosis-related DE-mRNAs. DE-mRNAs = differentially expressed mRNAs.

Figure 7.

Identification of ferroptosis-related hub genes in PE STB-EVs. (A) PPI network constructed with ferroptosis-related DE-mRNAs were performed using the STRING, the nodes represent proteins, and the edges represent the interaction of the proteins. (B) Cytohubba in Cytoscape was used to find the top five hub genes in the PPI network by Degree, PPI network of the top 5 hub genes was visualized by Cytoscape, and the top 5 hub genes are displayed from red (high Degree value) to yellow (low Degree value). DE-mRNAs = differentially expressed mRNAs, PE = preeclampsia, PPI = protein-protein interaction, STB-EVs = syncytiotrophoblast-derived extracellular vesicles.

Figure 8.

Receiver operating characteristic analysis revealed the predictive performance of ferroptosis-related hub genes for PE. (A) The AUC of ALB (A), NOX4 (B), CDKN2A (C), TXNRD1 (D), CAV1 (E) in ROC monofactor analysis. ALB = albumin, AUC = area under the curve, CAV1 = caveolin 1, CDKN2A = cyclin dependent kinase inhibitor 2A, NOX4 = nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4, ROC = receiver operating characteristic, TXNRD1 = thioredoxin reductase.

Figure 9.

A miRNA–hub gene regulatory network construction in PE STB-EVs, Interaction network between hub genes and its targeted miRNAs, Genes were colored in red, miRNAs were colored in blue, the higher amounts of cross-linked genes including hsa-miR-26b-5p, hsa-miR-192-5p, hsa-miR-124-3p, hsa-miR-492, hsa-miR-34a-5p and hsa-miR-155-5p. PE = preeclampsia, STB-EVs = syncytiotrophoblast-derived extracellular vesicles.

Figure 10.

Biological pathways, (A) the molecular function, (B) and prediction of potential transcription factors (C) of miRNAs.