Systemic approach to identify serum microRNAs as potential biomarkers for acute myocardial infarction

Abstract

Background: Recent studies have revealed the role of microRNAs (miRNAs) in a variety of biological and pathological processes, including acute myocardial infarction (AMI). We hypothesized that ST-segment elevation myocardial infarction (STEMI) may be associated with an alteration of miRNAs and that circulating miRNAs may be used as diagnostic markers for STEMI.

Methods: Expression levels of 270 serum miRNAs were analyzed in 8 STEMI patients and 8 matched healthy controls to identify miRNAs differentially expressed in the sera of patients with AMI. The differentially expressed miRNAs were evaluated in a separate cohort of 62 subjects, including 31 STEMI patients and 31 normal controls.

Results: The initial profiling study identified 12 upregulated and 13 downregulated serum miRNAs in the AMI samples. A subsequent validation study confirmed that serum miR-486-3p and miR-150-3p were upregulated while miR-126-3p, miR-26a-5p, and miR-191-5p were significantly downregulated in the sera of patients with AMI. Ratios between the level of upregulated and downregulated miRNAs were also significantly different in those with AMI. Receiver operator characteristics curve analysis using the expression ratio of miR-486-3p and miR-191-5p showed an area under the curve of 0.863.

Conclusion: Our results suggest that serum miRNAs may be used as potential diagnostic biomarkers for STEMI.

Figure 1

Detection of differentially expressed circulating miRNAs in serum samples. (a) Detection of miRNAs in serum from healthy subjects. Expression levels of 270 miRNAs in 300 μL of serum were quantified using a multiplexed RT-qPCR assay. The relative expression levels of the top 50 miRNAs detected in serum from three healthy subjects (mean ± SD) are shown. (b) Principle component analysis using expression levels of the 270 human miRNAs in serum samples from 8 healthy (blue) and 8 AMI (red) subjects. (c) Volcano plot indicated that 25 miRNAs were significantly altered in sera from the AMI patients. Red lines indicate the Student's t-test P = 0.05 and ±2-fold change. (d) Unsupervised hierarchical clustering of healthy and AMI samples using the 25 differentially expressed serum miRNAs. The hierarchical clustering was generated using Pearson's dissimilarity as the distance measure and Ward's method for linkage analysis. A: AMI patients; N: healthy subjects.

Figure 2

Expression levels and predictive power of five candidate miRNAs in the healthy controls and acute myocardial infarction (AMI) patients. (a) Expression levels of 5 candidate miRNAs in serum samples from 31 healthy controls and 31 AMI patients. Ct values generated from RT-qPCR were normalized to the spiked-in synthetic miRNA and then converted to 39-Ct. Data are presented as mean ± SD. P values were calculated using the t-test. (b) ROC analysis using expression levels of individual miRNAs in healthy controls and AMI samples.

Figure 3

Predictive power and expression ratios between two candidate serum miRNAs in the healthy controls and acute myocardial infarction (AMI) patients. (a) ROC analysis using expression ratios between two miRNAs (control N = 31, AMI N = 31 patients). (b) Expression ratios of miRNA combinations in AMI and control samples. Data are presented as mean ± SD. P values were calculated using the t-test.