Clinical significance of circulating microRNAs as diagnostic biomarkers for coronary artery disease

Abstract Coronary artery disease (CAD) is one of the biggest threats to human life. Circulating microRNAs (miRNAs) have been reported to be linked to the pathogenesis of CAD, indicating the possible role in CAD diagnosis. The present study aimed to explore the expression profile of plasma miRNAs and estimate their value in diagnosis for CAD. 67 Non‐CAD control subjects and 88 CAD patients were enrolled. We conducted careful evaluation by RT‐PCR analysis, Spearman rank correlation coefficients analysis, Receiver Operating Characteristic (ROC) curves analysis and so on. The plasma levels of six miRNAs known to be related to CAD were measured and three of them showed obvious expression change. Circulating miR‐29a‐3p, miR‐574‐3p and miR‐574‐5p were all significantly increased. ROC analysis revealed the probability of the three miRNAs as biomarkers with AUCs (areas under the ROC curve) of 0.830, 0.792 and 0.789, respectively. They were significantly correlated with each other in CAD patients, suggesting the possibility of joint diagnosis. The combined AUC was 0.915, much higher than each single miRNA. Therefore, our study revealed three promising biomarkers for early diagnosis of CAD. The combination of these miRNAs may act more effectively than individual ones for CAD diagnosis.

in cardiac tissue protection in stress condition. 9 miR-134 could be biomarkers for acute myocardial infarction. 3 miR-223 secreted by blood cells could enter vascular cells and walls and then play important roles in VSMC function and atherogenesis. 10 miR-574 shows increased expression in infarcted heart disease. 11 miR-574-5p can promote VSMCs proliferation. 12 miR-765 plays a role in CAD diagnosis. 13 In this study, we investigated the plasma levels of six plasma miRNAs (miR-29a-3p, 134-5p, miR-223-3p, miR-574-3p, miR-574-5p and miR-765). The miRNAs with obviously altered expression levels in CAD patients compared with controls patients were selected to explore the diagnostic value and evaluate the combined effect.

| Study participants
A total of 67 control samples and 88 CAD samples were enrolled in this study. The whole protocol of this study was admitted by the ethics committee of the Affiliated Hospital of Qingdao University.

| Plasma collection and RNA isolation
Fast blood samples (~5 mL) were withdrawn in EDTA-anticoagulated tubes, and the plasma (the supernatant) was collected by the centrifugation method.
Total RNA was isolated with the TRIzol extraction method.
Glycogen at a final concentration of 0.1 μg/μL was added to increase the RNA yielding.

| Detection and quantification of miRNAs by qRT-PCR
miRNAs were converted into cDNA and quantified by quantitative real-time polymerase chain reaction (qRT-PCR). SYBR Green miRNA qRT-PCR kits and a Bio-Rad CFX96 system were applied. The primers and specific miRNA sequences are listed in Table S1. We used U6 snRNA as the housekeeping gene.

| Statistical analysis
The distribution of miRNA expression was analysed by GraphPad Prism software. The differences in clinical characteristics between CAD patients and control patients were analysed by Mann-Whitney U test. The correlation analysis was carried out by the Spearman correlation coefficient (GraphPad Prism software). The diagnostic values were evaluated by receiver operating characteristic (ROC) curves analysis (MedCalc software). The area under the ROC curve (AUC) was considered as a critical diagnostic index. All tests were two-sided or two-tailed. A difference with a P < .05 was statistically significant.

| Basic clinical characteristics
The clinical and laboratory characteristics were counted. Significance analysis was performed between the control group and the CAD group ( Figure 1A).

| Correlation analysis
The three selected plasma miRNAs showed different distribution among all patients. We identified the correlation of the three selected plasma miRNAs in both CAD patients and control patients.

| Diagnostic potential of plasma miRNAs by ROC analysis
The ROC curves of miR-29a-3p, miR-574-3p and miR-574-5p revealed the probability of them as valuable biomarkers with AUCs of 0.830, 0.792 and 0.789, respectively ( Figure 2F-H). The related data are briefly summarized in Table S2.
As shown in Figure 2, miR-29a-3p, miR-574-3p and miR-574-5p were significantly correlated with each other, indicating the value of joint diagnosis. The combined AUC value was 0.916 ( Figure 2I; Table   S2), much higher than each single miRNA, suggesting the highest discriminatory power. and miR-574-3p have been found to be co-expressed and regulate physiological processes, such as cancer progression [14][15][16] and cardiovascular diseases. 11 Up till now, most research on miR-574-3p focused on tumorigenesis 15,16 and its role on CAD has not been illustrated.
In this study, we explored the diagnostic potential of circulating miR-29a-3p, miR-574-3p and miR-574-5p. Our data firstly identified the elevated expression of plasma miR-573-3p in CAD patients compared with control patients. We found the correlation of these three circulating miRNAs, indicating their combinative value. Each of them showed considerable diagnostic power from the AUC values of ROC analysis. Furthermore, their recombination exhibited much higher AUC value than each single miRNA. The present study has several limitations. The number of experimental samples was insufficient. Large-scale studies will be performed in the near future. Further experimental studies are needed to explore the mechanisms of up-regulation of miR-29a-3p, miR-574-3p and miR-574-5p. The qRT-PCR method is expensive and time-consuming. Therefore, cheaper and faster techniques are expected to be developed in the near future.
In summary, our study addressed the diagnostic value of miR-29a-3p, miR-574-3p and miR-574-5p and they could be combined into a probe system to provide more efficient, sensitive and non-invasive diagnosis for CAD than individual miRNA.

ACK N OWLED G EM ENTS
The authors would like to thank the funding from National Natural Science Foundation of China (Grant Number 31430041) and Shandong Provincial Natural Science Foundation of China (No. ZR2016CQ23). The authors would also like to thank Liang Xu for the help in revising the manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support this study are within the article and its Supporting Information files.