Metabolic Variation Dictates Cardiac Pathogenesis in Patients With Tetralogy of Fallot

Abstract

Background: Herein, we aimed to analyze cardiac metabolic reprogramming in patients with tetralogy of Fallot (ToF).

Methods: Cardiac metabolic reprogramming was analyzed through comprehensive bioinformatics analysis, which included gene set enrichment, gene set variation, and consensus clustering analyses, so as to assess changes in metabolic pathways. In addition, full-spectrum metabolomics analysis was performed using right atrial biopsy samples obtained from patients with ToF and atrial septal defect (ASD) before cardiopulmonary bypass; ultrahigh performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) was used to construct a metabolic map of cardiac metabolic reprogramming in cyanotic congenital heart disease.

Results: The metabolic maps of carbohydrate metabolic process and heme metabolism were significantly activated, while bile acid metabolism, lipid droplet, and lipid binding were primarily restrained in ToF samples as compared with that in ASD samples. The reprogramming of butanoate metabolism was identified basing on the UPLC-MS/MS detection and analysis in myocardial hypoxia damage in cyanotic heart disease. Finally, the butanoate metabolism-related hub regulators ALDH5A1 and EHHADH were identified and they were significantly downregulated in ToF samples.

Conclusions: The metabolic network of butanoate metabolism involved ALDH5A1 and EHHADH, which could contribute to myocardial tissue damage in cyanotic congenital heart of ToF. Our results provide further insights into the mechanisms underlying metabolic reprogramming in cyanotic congenital heart disease and could lead to the identification of potential therapeutic targets.

Keywords: UPLC–MS/MS; butanoate metabolism; metabolic reprogramming; metabonomic analysis; tetralogy of Fallot.

Figure 1

Workflow of metabolic reprogramming analysis.

Figure 2

Differentially expressed gene (DEG) identification and functional enrichment analysis. (A) Volcano plot showing DEG distribution. (B) Clustering heatmap and biological analysis presenting the biological function of up- and downregulated DEGs. (C) Biological function terms based on gene ontology were calculated by the GSEA algorithm in response to the comparison between ToF and ASD RA specimens.

Figure 3

Metabolic pathway detection and interaction network construction. (A) Clustering heatmap showing GSVA-quantized pathway score of metabolic maps. (B) Metabolic map interaction analysis was identified via weighted partial correlation analysis. (C) Differential analysis of metabolic pathways of GSVA score in Figure 2C. (D) Three-dimensional principal component analysis, showing right atrial biopsy sample distribution of ToF and ASD patients before cardiopulmonary bypass. (E) Clustering heatmap depicting the standard quantification of metabolites in ToF and ASD RA specimens. ^#p < 0.05.

Figure 4

Detection of differentially expressed metabolites in ToF and ASD RA specimens via the UPLC–MS/MS spectra. (A) T-score of OPLS-DA showing the distribution of ToF and ASD specimens. (B) Volcano plot showing the distribution of differentially expressed metabolites. (C) Differential expression of metabolites in ToF and ASD specimens. (D) Scatter diagram of Figure 3D showing the standardized Z-score for each sample.

Figure 5

Pathway enrichment analysis with differentially expressed metabolites. (A) The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis involved in classification of environmental information processing, metabolism, and cellular processes were identified. (B) Of the biologic procession, the significantly enriched terms were identified in Figure 4B. (C) Metabolic pathway enrichment was detected based on metabolite set enrichment analysis. (D) Overlap between the gene list involved in butanoate metabolism and DEGs of GSE132176.

Figure 6

Hub regulator expression and functional enrichment analysis of ToF and ASD RA specimens. (A,B) Hub genes showed differences in ToF and ASD RA specimens in the GSE132176 and GSE169214 datasets, respectively. (C) Hub terms were enriched in response to hub regulators. ^#p < 0.05.