PRKAR1A and SDCBP Serve as Potential Predictors of Heart Failure Following Acute Myocardial Infarction

Abstract

Background and objectives: Early diagnosis of patients with acute myocardial infarction (AMI) who are at a high risk of heart failure (HF) progression remains controversial. This study aimed at identifying new predictive biomarkers of post-AMI HF and at revealing the pathogenesis of HF involving these marker genes.

Methods and results: A transcriptomic dataset of whole blood cells from AMI patients with HF progression (post-AMI HF, n = 16) and without progression (post-AMI non-HF, n = 16) was analyzed using the weighted gene co-expression network analysis (WGCNA). The results indicated that one module consisting of 720 hub genes was significantly correlated with post-AMI HF. The hub genes were validated in another transcriptomic dataset of peripheral blood mononuclear cells (post-AMI HF, n = 9; post-AMI non-HF, n = 8). PRKAR1A, SDCBP, SPRED2, and VAMP3 were upregulated in the two datasets. Based on a single-cell RNA sequencing dataset of leukocytes from heart tissues of normal and infarcted mice, PRKAR1A was further verified to be upregulated in monocytes/macrophages on day 2, while SDCBP was highly expressed in neutrophils on day 2 and in monocytes/macrophages on day 3 after AMI. Cell-cell communication analysis via the "CellChat" package showed that, based on the interaction of ligand-receptor (L-R) pairs, there were increased autocrine/paracrine cross-talk networks of monocytes/macrophages and neutrophils in the acute stage of MI. Functional enrichment analysis of the abovementioned L-R genes together with PRKAR1A and SDCBP performed through the Metascape platform suggested that PRKAR1A and SDCBP were mainly involved in inflammation, apoptosis, and angiogenesis. The receiver operating characteristic (ROC) curve analysis demonstrated that PRKAR1A and SDCBP, as well as their combination, had a promising prognostic value in the identification of AMI patients who were at a high risk of HF progression.

Conclusion: This study identified that PRKAR1A and SDCBP may serve as novel biomarkers for the early diagnosis of post-AMI HF and also revealed their potentially regulatory mechanism during HF progression.

Keywords: acute myocardial infarction; biomarkers; cell–cell communication; heart failure; leukocytes.

Figure 1

Study flowchart. GEO, Gene Expression Omnibus; AMI, acute myocardial infarction; HF, heart failure; PBMCs, peripheral blood mononuclear cells; WGCNA, weighted gene co-expression network analysis; cor, correlation coefficient; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; Mon/Mac, monocytes/macrophages; Neu, neutrophils; L–R pairs, ligand–receptor pairs; ROC, receiver operating characteristic; AUC, area under the ROC curve.

Figure 2

Identification of highly correlated modules of post-AMI HF. Heatmap of the association between module eigengenes and HF progression after AMI. The correlation coefficient (cor) and Q-value are presented in each cell. The greenyellow module was found to be significantly correlated with post-AMI HF according to |cor|>0.5 and Q-value < 0.05. Red indicates a positive correlation between two modules, and blue indicates a negative correlation. AMI, acute myocardial infarction; HF, heart failure.

Figure 3

Validation of the expression levels of key bio-correlated genes in post-AMI HF and post-AMI non-HF patients based on GSE11947 and GSE59867 datasets. GSE11947: HF n = 16, non-HF n = 16; GSE59867: HF n = 9, non-HF n = 8. P-value was shown above the asterisk (*). *P < 0.05, **P < 0.01. Data are presented as mean ± SEM. AMI, acute myocardial infarction; HF, heart failure; SEM, standard error of mean.

Figure 4

The changes in expression levels of key bio-correlated genes over time, following AMI in the GSE59867 dataset. The changes in gene expression levels were observed in the HF and non-HF groups at different time points after AMI (1d: n = 9 vs. 8, 4–6 days: n = 9 vs. 6, 30 days: n = 8 vs. 8, 180 days: n = 9 vs. 8). The P-value was shown above the asterisk (*). *P < 0.05; n.s., no significance vs. non-HF group. Data are presented as mean ± SEM. AMI, acute myocardial infarction; HF, heart failure; SEM, standard error of mean.

Figure 5

Identification of cell type of CD45+ lymphocytes in infarcted cardiac tissues of mice. (A) tSNE visualization of cell type in heart tissues of control and infarcted mice. (B) tSNE visualization of lymphocytes in heart tissues of control mice at 1 day (1 d), 2 days (2 d) and 3 days (3 d) after AMI. (C) Dot plot of canonical markers of different cell types. Dot size reflects the percentage of cells expressing the markers in each cell type. The scale color represents the gene expression levels from low to high.

Figure 6

Dot plot of expression profiles of PRKAR1A (A) and SDCBP (B) in monocytes/macrophages and neutrophils at different time points after AMI. Dot size reflects the percentage of cells expressing the markers in each cell type. The scale color represents the gene expression levels from low to high. AMI, acute myocardial infarction.

Figure 7

Upregulated ligand–receptor signaling received by monocytes/macrophages (Mon/Mac) on days 2 (A) and 3 (C) after AMI and neutrophils (Neu) on days 2 after AMI (B). Cell–cell communication of recipient Mon/Mac and Neu was analyzed by the “CellChat” package. Dot size reflects the P-value of the signaling. The scale color represents the level of communication probability from minimum to maximum. AMI, acute myocardial infarction; Neu, neutrophils.

Figure 8

Receiver operating characteristic (ROC) curve analysis of potential biomarkers. ROC curves of PRKAR1A, SDCBP, and the combination of SDCBP/PRKAR1A, and the areas under the ROC curves were presented based on the expression levels of GSE59867 (A) and GSE11947 (B) datasets, respectively.