Journal of Molecular Medicine

, Volume 90, Issue 8, pp 865–875

Circulating microRNAs: novel biomarkers for cardiovascular diseases

Authors

  • Jiahong Xu
    • Department of Cardiology, Tongji HospitalTongji University School of Medicine
    • Key Laboratory of Arrhythmias, Ministry of EducationChina (Shanghai East Hospital, Tongji University School of Medicine)
  • Jiangmin Zhao
    • Department of RadiologyShanghai Third People’s Hospital, Shanghai Jiaotong University School of Medicine
  • Graham Evan
    • Cardiovascular InstituteBeth Israel Deaconess Medical Center and Harvard Medical School
  • Chunyang Xiao
    • Cardiovascular InstituteBeth Israel Deaconess Medical Center and Harvard Medical School
    • Key Laboratory of Arrhythmias, Ministry of EducationChina (Shanghai East Hospital, Tongji University School of Medicine)
    • Department of Psychiatry, Tongji HospitalTongji University School of Medicine
    • Key Laboratory of Arrhythmias, Ministry of EducationChina (Shanghai East Hospital, Tongji University School of Medicine)
    • Cardiovascular InstituteBeth Israel Deaconess Medical Center and Harvard Medical School
Review

DOI: 10.1007/s00109-011-0840-5

Cite this article as:
Xu, J., Zhao, J., Evan, G. et al. J Mol Med (2012) 90: 865. doi:10.1007/s00109-011-0840-5
  • 1.5k Views

Abstract

MicroRNAs (miRNAs) are a novel class of small, non-coding, single-stranded RNAs that negatively regulate gene expression via translational inhibition or mRNA degradation followed by protein synthesis repression. Many miRNAs are expressed in a tissue- and/or cell-specific manner and their expression patterns are reflective of underlying patho-physiologic processes. miRNAs can be detected in serum or in plasma in a remarkably stable form, making them attractive biomarkers for human diseases. This review describes the progress of identifying circulating miRNAs as novel biomarkers for diverse cardiovascular diseases, including acute myocardial infarction, heart failure, coronary artery disease, diabetes, stroke, essential hypertension, and acute pulmonary embolism. In addition, the origin and function and the different strategies to identify circulating miRNAs as novel biomarkers for cardiovascular diseases are also discussed. Rarely has an opportunity arisen to advance such new biology for the diagnosis of cardiac diseases.

Keywords

MicroRNAsPlasmaSerumBiomarkerCardiovascular diseases

Introduction

MicroRNAs (miRNAs, miRs) are endogenous, non-coding, single-stranded RNAs of 19–25 nucleotides in length [1]. They post-transcriptionally regulate gene expression via translational inhibition or mRNA degradation followed by protein synthesis repression [2, 3]. It is estimated that more than 1,000 different miRNAs exist in humans, and many are expressed in a tissue- and/or cell-specific manner [1, 4]. Their expression patterns are reflective of underlying patho-physiologic processes and, interestingly, miRNAs can be detected in serum or in plasma in a remarkably stable form. Circulating miRNAs can withstand repetitive freezing and thawing cycles, making them attractive as potential biomarkers for diverse human diseases [5, 6].

Claiming 17.1 million lives per year, cardiovascular diseases are the largest killers worldwide, making them major global health problems and extreme financial burdens [7]. An early and correct diagnosis might afford great benefits in treatment [8]. To that end, the exploration of new biomarkers with high sensitivity and specificity in diagnosis of cardiovascular diseases continues [9]. An ideal biomarker should be reproducibly measurable with high sensitivity and specificity for the clinical outcome of interest and should reflect an important pathogenetic process [10]. miRNAs are exciting as potential biomarkers because they fulfill many of these criteria [6]. Emerging evidences have indicated circulating miRNAs as novel biomarkers for cardiovascular diseases including acute myocardial infarction (AMI), heart failure (HF), coronary artery disease (CAD), diabetes mellitus (DM), stroke, essential hypertension, and acute pulmonary embolism (APE) [1125]. Here we have highlighted and summarized recent advances in the identification of circulating miRNAs as novel biomarkers for cardiovascular diseases (Table 1) and briefly discuss the origin and function and the different strategies to identify circulating miRNAs as novel biomarkers for cardiovascular diseases.
Table 1

Potential circulating microRNA biomarkers for cardiovascular diseases

Diseases

MicroRNAs

Sources

Regulation

Time point

No. samples

Potential use

Methods

Species

Reference

Acute myocardial infarction

miR-1

Plasma

++

 

93 AMI, 66 healthy

Diagnostic

qRT-PCR

Human

[13]

 

miR-499

Plasma

++

48 h

9AMI, 5 UAP, 9CHF, 10 healthy

Diagnostic

qRT-PCR

Human

[12]

 

miR-208a

Plasma

++

4.8 ± 3.5 h

33 AMI and non-AMI, 30 healthy;

Diagnostic

qRT-PCR

Human

[8]

    

3 h

8

Diagnostic

qRT-PCR

Rat

[19]

 

miR-1; miR-133; miR-499

Plasma

+∼+++

8.6 ± 5.1 h

33 STEMI,17 healthy

Diagnostic

qRT-PCR

Human

[20]

    

18 h, 6–18 h, 24 h

4–5

Diagnostic

qRT-PCR

Mice

[20]

 

miR-208b; miR-499

Plasma

+++∼++++

12 h

36 STEMI, 36 healthy

Diagnostic

qRT-PCR

Human

[38]

 

miR-1

Serum

+++

6 h

12

Diagnostic

qRT-PCR

Rat

[15]

 

miR-1291; miR-663b

Whole blood

3 ± 2.3 h

20 STEMI, 20 healthy

Diagnostic

Array

Human

[41]

 

miR-328

Plasma

++

24 h

51 AMI, 28 healthy

Diagnostic

qRT-PCR

Human

[43]

  

Whole blood

++

24 h

51 AMI, 28 healthy

Diagnostic

qRT-PCR

Human

[43]

 

miR-21, miR-29a, miR-208a

Plasma

−∼+

2–90 days

12 post-MI 12 healthy

Diagnostic, prognostic

qRT-PCR

Human

[44]

Heart failure

miR423-5p

Plasma

+

 

30HF, 20 non-HF, 39 healthy

Diagnostic

Array

Human

[46]

 

miR-126

Plasma

 

33 ischemic HF, 17 asymoptomatic controls

Diagnostic

qRT-PCR

Human

[47]

 

miR-361-5p

Plasma

+

 

10 HF, 10 healthy

Diagnostic

Array

Human

[14]

 

miR-107;miR-139; miR-142-5p

PBMCs

 

26 NIDCM, 23 ICM, 27 healthy

Diagnostic

Array

Human

[48]

Coronary artery disease

miR-126; miR-17; miR-92a miR-155; miR-145

Plasma

 

67 CAD, 31 healthy

Disease monitoring

Array

Human

[16]

 

miR-135a/miR-147 ratio

PBMCs

++

 

25 UAP, 25 SAP, 20 healthy

Disease monitoring

Array

Human

[49]

Diabetes mellitus

miR-15a;miR-126;miR-320; miR-223

Plasma

 

80 DM, 80 healthy

Disease monitoring

Array

Human

[18]

 

miR-28-3p

Plasma

+

 

80 DM, 80 healthy

Disease monitoring

Array

Human

[18]

 

miRNA-9, miRNA-29a, miRNA-30d, miRNA-34a, miRNA-124a, miRNA-146a, and miRNA-375

Serum

+

 

18 n-T2D, 19 pre-T2D, 19s-NGT

Disease monitoring

qRT-PCR

Human

[52]

 

miRNA-144, miRNA-150, miRNA-182, miRNA-192, miRNA-29a, and miRNA-320a

Whole blood

+

 

6 IGF, 8 T2D, 7 healthy

Disease monitoring

qRT-PCR

Human

[53]

 

miRNA-146a, miRNA-30 d

Whole blood

 

6 IGF, 8 T2D, 7 healthy

Disease monitoring

qRT-PCR

Human

[53]

Stroke

miR-124

Plasma

+++

9–24 h

3–5

Diagnostic

qRT-PCR

Rat

[23]

Essential hypertension

hcmv-miR-UL112;miR-605; miR-623;let-7e;miR-516b; miR-600;kshv-miR-K12-6-3p; miR-602; miR-1252

Plasma

+

 

13 Essential hypertension, 5 healthy

Disease monitoring

Array

Human

[24]

 

miR-296-5p;miR-133b; miR-30 d; miR-625*; miR-1236; miR-518b; miR-1227;miR-664; miR-615-5p; miR-18b*; miR-1249; miR-324-3p; ebv-miR-BART17-3p; miR-634; ebv-miR-miR-BART19-5p;miR-486-5p;kshv-miR-K12-10a;kshv-miR-K12-10b

Plasma

 

13 Essential hypertension, 5 healthy

Disease monitoring

Array

Human

[24]

Acute pulmonary embolism

miR-134

Plasma

++

 

32 APE, 32 healthy, 22 non-APE

Diagnostic

Array

Human

[25]

Up-regulation, <10 “+”, 10–100 “++”, 100–1,000 “+++”, >1,000 “++++”

Down-regulation: < 10 “−”, 10–100 “−−”, 100–10,00 “−−−”, >1,000 “−−−−”

miR microRNA; AMI aute myocardial infarction; qRT-PCR quantitative real-time polymerase chain reaction; UAP unstable angina pectoris; CHF congestive heart failure; STEMI ST-segment elevated myocardial infarction; HF heart failure; PBMCs peripheral blood mononuclear cells; NIDCM nonischemic dilated cardiomypathy; ICM ischemic cardiomypathy; CAD coronary artery disease; UAP unstable angina pectoris; SAP stable angina pectoris; DM diabetes mellitus; n-T2D newly diagnosed type 2 diabetes; pre-T2D pre-diabetes (impaired glucose tolerance and/or impaied fasting glucose); s-NGT T2D-susceptible individuals with normal glucose tolerance; APE acute pulmonary embolism

Origin and function of circulating miRNAs

The origins or sources of circulating miRNAs are largely unknown. Some hypotheses have proposed that they are secreted in membrane-bounded-vesicles (apoptotic bodies, microvesicles, exosomes, etc.) [26], while others state that they are secreted in vesicle-free but protein-protected protein–miRNA complexes (Ago2, NPM1, and other RNA-binding proteins) [2729]. It has also been hypothesized that they are produced as by-products of dead cells [26, 30] (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00109-011-0840-5/MediaObjects/109_2011_840_Fig1_HTML.gif
Fig. 1

Origin of circulating miRNAs

To explore whether or not the heart is a source for circulating miRNAs, concentration gradients across the coronary circulation for muscle-enriched (miR-133a, miR-499, and miR-208a), vascular (miR-126, miR-92a), leukocyte-related (miR-155), and platelet-enriched (miR-223) miRNAs were measured using plasma from the aorta and coronary venous sinus in patients without coronary artery disease, with stable coronary artery disease, and with troponin-positive acute coronary syndromes [31]. miR-499, miR-133a, and miR-208a were found to be up-regulated in the aortas of patients with troponin-positive acute coronary syndromes as compared to those with stable coronary artery disease [31]. Interestingly, a significant increase in miR-499 and miR-133a across the coronary circulation in troponin-positive acute coronary syndromes was observed, suggesting that these muscle-derived miRNAs were released into the coronary circulation during myocardial injury [31]. In addition, in patients with troponin-positive acute coronary syndromes, miR-126 showed an inverse transcoronary concentration gradient, suggesting consumption during transcoronary passage [31]. A similar pattern like miR-126 was found for miR-92a though it failed to reach statistical significance [31]. As for miR-155 and miR-223, no significant difference was observed in patients with troponin-positive acute coronary syndromes [31]. This study was the first to give direct evidence showing that the heart is a source for circulating miRNAs in myocardial injury [31].

It is still debated whether circulating miRNAs are simply waste products or whether they have potential function. There are some reports pointing to their potential functions, including their ability to modulate immune cells and their potential transfer to other cells or organs to induce more organ-specific functions [3033] (Fig. 2). miR-146a in peripheral blood mononuclear cells (PBMCs) was first reported to modulate T helper (Th) cells Type 1 (Th1) function in patients with acute coronary syndrome [32]. As Th1 is essential in the process of plaque instability and plaque rupture—which is a common pathogenetic feature in acute manifestations of atherosclerosis, such as acute coronary syndrome—it was speculated that miR-146a might contribute to the onset of acute coronary syndrome [32]. The altered expression of inflammation-related miRNA (miR-155) was subsequently found to be correlated with T helper 17 cell differentiation in patients with acute coronary syndrome, indicating that it might contribute to atherosclerosis, as atherosclerosis is a chronic inflammatory disease that is upregulated by immune cells, especially Th cells [33]. In addition, a calcium ionophore, A23187, was used to stimulate the release of miR-133a followed by exosome harvest. After adding these exosomes into human embryonic kidney 293 cells expressing the miR-133a sensor vector, the luciferase activity was significantly reduced, indicating that circulating miRNAs may be transferred into other cells or organs to induce their specific functions [30].
https://static-content.springer.com/image/art%3A10.1007%2Fs00109-011-0840-5/MediaObjects/109_2011_840_Fig2_HTML.gif
Fig. 2

Function of circulating miRNAs

Acute myocardial infarction

Acute myocardial infarction (AMI) is a leading cause of morbidity and mortality worldwide [34]. Creatine kinase-MB, cardiac myoglobin, and troponins have been widely used in the clinic [34, 35]. However, continuous efforts are paid to explore novel approaches that can complement and improve current strategies for AMI diagnosis [34].

miR-208a

In an isoproterenol-induced myocardial injury rat model, the plasma concentration of miR-208a, a cardiac-specific miRNA, was found to be significantly increased with good correlation to the concentration of cardiac troponin I (cTnI), a classic biomarker of myocardial injury. This indicated that miR-208a was leaked into the circulating blood from injured cells [19]. To exclude the possibility that the increased plasma concentration of miR-208a was caused by non-specific insult, its level was also measured in a renal-infarction model. There it was found that the plasma level of miR-208a remained undetectable, suggesting that non-specific insult does not appreciably increase its circulating concentration [19]. In end-stage renal disease, even in the absence of an acute coronary syndrome, the concentration of cardiac troponins will occasionally increase as they are excreted from the kidney. Therefore, it seems that plasma miR-208a is superior to cTnI and/or cardiac troponin I (cTnT) for detecting acute myocardial injury in patients with renal dysfunction [19].

Subsequently, the potential of muscle/cardiac specific miRNAs including miR-1, miR-133a, miR-133b, and miR-208a as AMI biomarkers was explored [8]. Their levels in plasma were significantly increased in AMI rats. However, further studies showed that elevated plasma levels of miR-1, miR-133a, and miR-499 might also be caused by skeletal muscle injury, suggesting the lack of cardio-specificity. Thus, plasma miR-208a seems to be an ideal biomarker of AMI. Moreover, plasma miR-208a could be easily detected in 90.9% of AMI patients and in 100% of AMI patients within 4 h of the onset of symptoms. By the receiver operating characteristic (ROC) curve analysis, miRNA-208a revealed high sensitivity and specificity for diagnosing AMI, pointing to its potential as novel biomarker in human AMI [8].

However, contrary to the previous report [8], it was found that miR-208a was expressed at low levels in three of nine ST elevation MI patients (STEMI) patients and was undetectable in other AMI patients [20]. Moreover, a most recent study also failed to detect miR-208a in the plasma of acute STEMI patients [21]. They speculated that the inconsistence of the miR-208a result reported in theirs and in that one showing it to be an ideal biomarker for AMI might be due to the extremely short half life of miR-208a. The earliest patients' samples in that study were taken within 4 h of the onset of symptoms while the samples in their study were taken after 12 h [21]. In light of these results, due to its short half life, miR-208a does not appear to be an ideal biomarker of human AMI even if it is increased in AMI patients.

Interestingly, the elevation of circulating miR-208a in AMI could be consistently found in mice but not in human. We speculated that it might be due to the baseline expression differences of this miRNA between these two species. miR-208a is encoded within an intron of α-cardiac muscle myosin heavy chain gene (α-MHC) while miR-208b is encoded within an intron of β-cardiac muscle myosin heavy chain gene (β-MHC) [36]. β-MHC is the predominant isoform in adult human hearts, while in rodent, it is a-MHC [37]. As the expression level of these two miRNAs often parallel with their host genes [36], thus their expression level in human and rodent could be different, which means it is possible that in rodent, miR-208a is more abundant while in human is miR-208b. Thus, in rodent, miR-208a might be much easier to be leaked into the circulation in AMI.

miR-1, miR-133a, miR-133b, miR-499, and miR-208b

Other muscle-specific miRNAs include miR-1, miR-133a, miR-133b, miR-499, and miR-208b. Within 12 h of the onset of symptoms in acute STEMI, the levels of plasma miR-1, miR-133a, miR-499-5p, and miR-208b increased 300-fold, 70-fold, 250-fold, and 3,000-fold, respectively [21]. Although muscle injury might also lead to their release into circulation, it has already been found that in an acute hind-limb ischemia, unlike in AMI, plasma levels of miR-1, miR-133a, miR-133b, and miR-499-5p did not increase, indicating that they are ideal biomarkers for AMI [20]. However, in that study showing miR-208 to be an ideal biomarker for AMI, they indicated that plasma levels of miR-1, miR-133a, and miR-499 were also increased in skeletal muscle injury [8]. Therefore, more studies are needed to determine whether muscle injury might also contribute to the release of muscle specific miRNAs into circulation.

miR-499 was found to be highly enriched in the heart and that its plasma concentrations were increased in all AMI individuals but were below the limit of detection for acute coronary syndrome and congestive HF patients and for all healthy controls [12]. This result indicated that plasma miR-499 was a potential biomarker for the diagnosis of AMI. Recently, plasma levels of heart-associated miRNAs (miR-1, miR-133a, miR-208b, and miR-499), fibrosis-associated miRNAs (miR-21 and miR-29b), and leukocyte-associated miRNAs (miR-146, miR-155, and miR-223) in patients with various cardiac damage including AMI, viral myocarditis, diastolic dysfunction, and acute HF were assessed [38]. miR-208b and miR-499 were found to be elevated 1,600-fold and 100-fold in AMI patients, respectively, compared to healthy controls. The ROC curve analysis revealed an area under the ROC curve (AUC) of 0.94 for miR-208b and 0.92 for miR-499, indicating that they were ideal biomarkers for AMI. The plasma level of these two miRNAs correlated well with plasma cTnT, indicating that they were released from injured cardiomyocytes [38].

The plasma miR-1 level was also reported to be significantly increased in AMI patients and subsequently dropped to the normal level upon discharge following medication [13]. Using the plasma level of miR-1, the AUC was 0.7740 for separation between non-AMI and AMI patients and 0.8522 for separation between AMI patients under hospitalization and those discharged, indicating plasma miR-1 to be a potential biomarker for AMI [13]. Interestingly, no correlation between plasma miRNA-1 and the other well-known biomarkers of AMI including cTnI and CK-MB were found, indicating that miR-1 might be a unique, independent biomarker for AMI [13].

Interestingly, in a cardiac cell necrosis model induced by Triton X-100, miR-1 was also found to be released into the culture medium and was stable at least for 24 h [15]. In addition, in a rat AMI model, serum miR-1 was increased over 200-fold with a peak at 6 h [15]. Moreover, the serum miR-1 level in AMI rats had a strong positive correlation with MI size [15]. This study opens the possibility of using serum miRNA as biomarkers of AMI. In addition, miR-1 and miR-133a were also reported to be elevated in urine of acute STEMI patients and that the miRNA stability in urine would not be affected by freezing and thawing [21]. The elevated levels of miR-1 and miR-133a in urine were due plasma excretion as they both correlated strongly with the glomerular filtration rate (GFR) [21]. In support of this hypothesis, miR-208b or miR-499-5p were unable to be detected in the urine of acute STEMI patients as these two miRNAs did not correlate with the GFR, indicating they are not due to renal elimination [21]. This study indicates that urine miRNA might also be used as an AMI biomarker. In line with these two reports, it seems that circulating miRNA in plasma, serum, and other body fluids may represent a gold mine of non-invasive biomarkers for AMI [15, 21].

As indicted above, we can find that great bifurcations in regarding cardiac/muscle-specific miRNAs as biomarkers of AMI exist, possibly due to the great heterogeneity of human plasma samples. Thus, the route from proof of principle to the creation of reliable and reproducible miRNA clinical test is a long one. Practically, quantification of circulating miRNAs is challenging due to the low amounts of RNA recovered from plasma or serum and to the lack of proper endogenous controls for normalization. Several endogenous circulating miRNAs such as miR-16, U6, and 5S rRNA cannot be ruled out with regards to the changes with cardiovascular disease or even with cardiovascular risk factors. Therefore, the use of spike-in, control miRNAs such as synthetic Caenorhabditis elegans miRNA (cel-miR-39, cel-miR-54, and cel-miR-238 etc.) for normalizing is necessary [11, 39]. Also, a standard quantitative real-time polymerase chain reaction (qRT-PCR) outline that includes blood acquisition, handling, storage, and plasma internal controls for plasma or serum should be set up worldwide in the near future. Details about a suggested protocol for the analysis of circulating microRNA biomarkers in plasma and serum using qRT-PCR have been recently described elsewhere [40].

miR-1291, miR-663b, miR-328, miR-21, and miR-29a

All the studies mentioned above are based on the hypothesis that circulating miRNAs are released from the injured heart. For this reason, heart or muscle specific miRNAs are paid special attention. However, another possibility is that circulating miRNAs do not originate in the injured heart. To test this possibility, the whole-genome miRNA expressions in peripheral total blood samples of AMI patients were detected and 121 significantly dysregulated miRNAs were identified [41]. Of these, miR-1291 and miR-663b showed the highest sensitivity and specificity for the discrimination of cases from controls [41]. They also indicated a unique signature of 20 miRNAs that predicted AMI with a specificity of 96%, a sensitivity of 90%, and an accuracy of 93% [41]. These results indicate that single miRNA and especially miRNA signatures derived from peripheral blood could function as novel, potentially valuable biomarkers for AMI.

It was also found that miR-328, a miRNA which has been reported to contribute to adverse electrical remodeling in atrial fibrillation [42], has also been reported to increase by 10.9-fold and 16.1-fold in plasma and whole blood of AMI patients, respectively [43]. Although miR-328 is not a cardiac/muscle-specific miRNA, it correlated well with cTroI [43]. Moreover, circulating miR-328 showed no significant changes in AMI patients with arrhythmia compared to those without [43]. ROC analysis revealed that AUCs of miR-328 in plasma and whole blood were 0.810 and 0.872, respectively, suggesting that miR-328 might also represent a novel biomarker of AMI [43].

In contrast to all of the studies conducted in AMI patients measuring selected miRNAs in a single time point following AMI, a recent study reported a time-dependent change in plasma miRNA levels after AMI [44]. They found that miR-21 initially fell 2 days post-AMI, increased 5 days post-AMI, and returned to baseline levels at later post-AMI time points. In addition, miRNA-29a increased 5 days post-AMI and fell to baseline levels at later post-AMI time points while miR-208a increased 5 days post-AMI and remained up-regulated up to 90 days post-AMI. Interestingly, a significant association between miR-29a early after AMI (5 days post-AMI) and left ventricle end-diastolic volume (LVEDV) later after an AMI (90 dayays post-AMI) was observed [44]. These results indicated that serially profiling miRNAs in the plasma of post-AMI patients may have both mechanistic and prognostic significance. However, as that study only examined 12 patients following an AMI for a 90-day period of time, the findings still need to be confirmed in larger studies [44].

Heart failure

Heart failure (HF) is the leading cause of hospitalization in people older than 65, affecting nearly 5 million Americans [45]. A simple and reliable measurement of circulating biomarker as objective measures of HF is highly desirable [45].

After performing a miRNA array on plasma of healthy controls and on HF patients, Tijsen and others selected 16 miRNAs for validation in independent samples, including 39 healthy controls and 50 cases with reports of dyspnea, of whom 30 were diagnosed with HF and 20 with dyspnea attributable to non-HF-related causes [46]. It was found that the plasma miR-423-5p level was specifically elevated in HF and with an AUC of 0.91, the ROC curve analysis further confirmed it to be a diagnostic predictor of HF [46]. However, other researchers found that the miRNA array was not proper for plasma miRNA detection due to its low detection rate [47]. Instead, they assessed the plasma concentrations of miR-126, miR-122, and miR-499 by qRT-PCR in congestive HF patients and revealed that the plasma miR-126 level was negatively correlated with age and the log (B-type natriuretic peptide). When the New York Heart Association (NYHA) Functional Classification was improved from IV to III, plasma concentrations of miRNA-126 were up-regulated, further indicating that it could be a useful biomarker for HF [47]. Later, the plasma miRNA expression levels of chronic HF patients were compared to healthy controls using a microRNA array, and it was found that plasma miR-361-5p concentrations were decreased significantly in all individuals with chronic HF [14]. This result implicated plasma miR-361-5p as a potential, novel biomarker for chronic HF.

In order to characterize the miRNA activity in an environment other than human plasma, the miRNA signatures in peripheral blood mononuclear cells (PBMCs) from chronic HF patients were also revealed. In that study, it was found that miR-107, miR-139, and miR-142-5p were down-regulated in both non-ischemic dilated (NIDCM) and ischemic cardiomyopathy (ICM) patients compared to healthy controls. In addition, miR-142-3p and miR-29b were increased in NIDCM while miR-125b and miR-497 were decreased in ICM patients [48]. This study indicated the potential of using these miRNAs in peripheral blood mononuclear cells as biomarkers of chronic HF.

Coronary artery disease

Coronary artery disease (CAD) is a major cause of death and disability in developed countries and some developing countries [16, 49]. For example, in the United States, it remains responsible for about one third of all deaths in individuals over age 35 [11, 49].

Plasma levels of endothelial cell enriched miRNAs (miR-126, miR-17, and miR-92a), inflammation associated miR-155, and smooth muscle-enriched miR-145 were reported to be significantly reduced in CAD patients compared to healthy controls. These results offered potential biomarker candidates for CAD [16]. After that, a commercial miRNA platform was used to detect 157 different miRNAs in PBMCs of CAD patients and a fivefold increase in the concentration of miR-135a as well as a fourfold decrease in miR-147 concentration were found, indicating that miR-135a/miR-147 ratio could be a potential biomarker for CAD. Interestingly, it was also indicated that unstable angina pectoris patients could be discriminated from stable patients based on their relatively high expression level of miR-134, miR-198, and miR-370, suggesting that this miRNA signature could be used to identify patients at risk for acute coronary syndromes [49].

Diabetes mellitus

The complications of cardiovascular diseases are the major sources of expenses for patients with DM [50]. Two thirds of diabetes mellitus (DM) patients will die from cardiovascular diseases [50, 51]. The prevalence of DM in the United States was about 3.1% in the general population. Of these cases, 90% are type 2 DM [50]. Moreover, the number of DM patients is predicted to double within 20 years due to people in developing countries adopting a less healthy lifestyle [51].

In type 2 DM patients, decreased plasma levels of miR-20b, miR-21, miR-24, miR-15a, miR-126, miR-191, miR-197, miR-223, miR-320 and miR-486 were reported, while a modest increase of miR-28-3p was observed [18]. Using the expression profiles of the five most significant miRNAs (miR-15a, miR-126, miR-320, miR-223, miR-28-3p), DM patients (70%) and healthy controls (92%) could be correctly classified. Interestingly, using this profile, normoglycemic subjects that developed DM over the 10-year follow-up period were already classified as diabetics. Thus, these 5 miRNAs might be potential biomarkers for the diagnosis or even prediction of DM [18].

Besides plasma miRNAs, the potential of using serum miRNAs as DM biomarkers is also explored [52]. In newly diagnosed type 2 DM patients, all seven DM-related miRNAs detected (miR-9, miR-29a, miR-30d, miR-34a, miR-124a, miR-146a, and miR-375) in serum were significantly increased compared with type 2 DM-susceptible individuals with normal glucose tolerance, pointing towards their potential as biomarkers for type 2 DM. Interestingly, a similar pattern of these seven miRNAs was observed in pre-DM (impaired glucose tolerance and/or impaired fasting glucose) and type 2 DM-susceptible individuals with normal glucose tolerance, indicating that during the pathogenesis of type 2 DM, the DM-related miRNAs will not be significantly changed in serum [52].

The miRNA expression profiles for type 2 DM in whole blood was also recently highlighted [53]. Using eight miRNAs (miR-144, miR-146a, miR-150, miR-182, miR-192, miR-29a, miR-30d, and miR-320a), impaired fasting glucose and type 2 DM could be well distinguished [53]. Increased circulating level of miR-144 was also found to be correlated with its decreased target gene (insulin receptor substrate 1), a key molecule in insulin signaling [53]. However, because blood samples also contain circulating cells which would also affect the miRNA signature, the data from this study cannot be directly compared with other studies using plasma or serum.

Stroke

Globally, there are over 50 million stroke patients, producing an immense burden on the economic and healthcare infrastructure [54]. Ischemic stroke accounts for approximately 85% of the total number of strokes [54]. A miRNA profile described under ischemic conditions in both the brain and blood of rats subjected to middle cerebral artery occlusion was performed and indicated that studies are highly needed to evaluate the possibility of using circulating miRNAs as biomarkers in stroke and related pathologies [22]. A miRNA array was therefore used to identify the miRNA signature in peripheral blood of patients with ischemic stroke [54]. A miRNA expression profile was found to be implicated in the endothelial/vascular function, erythropoiesis, angiogenesis, and neural function, demonstrating the potential of using peripheral blood miRNA profiles as biomarkers of cerebral ischemic stroke [52]. Similar to the use of cardiac-specific miRNA as biomarkers of AMI, the concentration of the brain-specific miR-124 was also reported to be increased in the plasma of a rat acute ischemic stroke model, garnering interest in exploring brain-specific miRNAs including miR-124, miR-134, miR-153, miR-9, and miR-219 as potential plasma biomarkers of human acute ischemic stroke [23].

Essential hypertension

Essential hypertension accounts for approximately 90–95% of hypertension cases, affecting over 1 billion people all over the world. It is a major risk factor for AMI, HF, stroke, and chronic renal failure [55].

Using microarray-based miRNA expression profiling, plasma miRNA expression of 13 essential hypertension patients and five healthy controls was detected and 46 dysregulated miRNAs were identified, 11 of which were up-regulated and 35 down-regulated [24]. Among all 46 miRNAs, 27 were deposited into the miRBase (Table 1). Interestingly, interferon regulatory factor 1 was found to be a direct target of hcmv-miR-UL112. Increased HCMV seropositivity and quantitative titers were identified in the hypertension group compared with the control group. In addition, seropositivity, log-transformed copies of HCMV, and hcmv-miR-UL112 were independently associated with an increased risk of hypertension. This study was the first to report a plasma miRNA profile for hypertensive patients and demonstrated a novel link between HCMV infection and essential hypertension [24]. However, high degrees of inter-patient differences exist in their study, and it is still unclear whether or not these miRNAs can be used as biomarkers for diagnosis or for hypertension screening.

Acute pulmonary embolism

Acute pulmonary embolism (APE) is a dreaded yet frequent cardiovascular emergency with an estimated annual incidence of approximately 100 cases per 100,000 individuals in the United States and Europe [56, 57].

Our group identified 30 differentially expressed plasma miRNAs in APE patients as compared to healthy controls using the TaqMan miRNA microarray [25]. From these miRNAs, miR-134 and miR-410 were selected for further validation based on the changes (>10-fold) and probability values (P < 0.05). Using the TaqMan-based miRNA qrtPCR reactions, the plasma miRNA-134 level was found to be increased in the APE group compared to both healthy controls and the non-APE group (patients who had only reported dyspnea, chest pain, or cough). However, plasma miRNA-410 was also increased in the non-APE group. ROC curve analysis showed that plasma miR-134 was a specific diagnostic predictor of APE with an AUC of 0.833 [25].

However, because the number of APE patients in our study is relatively small, the results obtained in this small group will require replication in large, independent studies of APE. Additional work is also required to determine the additive benefit of miR-134 in algorithms for APE detection that incorporate other diagnostic modalities in a prospective fashion. Our study nevertheless provides a basis for future efforts to develop plasma miRNA-based assays to diagnose APE [25].

Strategies for identifying miRNAs as novel cardiovascular disease biomarkers

Based on different understandings of the origins of circulating miRNAs, two major strategies are widely used to identify circulating miRNAs as novel biomarkers for cardiovascular diseases (Fig. 3). First, with the understanding that circulating miRNAs were thought to be released from injured cells, cardiac/muscle specific miRNAs including miR-208, miR-499, miR-1, and miR-133 have been identified as potential biomarkers of AMI (Table 1) [11]. This hypothesis was supported by the fact that the changes of plasma miRNA levels were associated with reciprocal changes in their cardiac expression; miR-1, miR-133a, miR-133b, and miR-499-5p exhibited a decrease either in the infarct zone, in the border zone, or in both zones [26]. However, other highly enriched cardiac miRNAs, including miR-24, miR-26a, miR-126, and miR-30c, were not found to be increased in the plasma of AMI patients, indicating that release from injured cells is not the only mechanism [26]. Therefore, the miRNA array and next-generation sequence analyses have become more common ways to identify potential circulating miRNAs as biomarkers of cardiovascular diseases (Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs00109-011-0840-5/MediaObjects/109_2011_840_Fig3_HTML.gif
Fig. 3

Two major strategies used to identify circulating miRNAs as novel biomarkers for cardiovascular diseases, qRT-PCR, quantitative real-time polymerase chain reaction; ROC curve, receiver operating characteristic curve

Future directions

Until now, all studies aiming to identify circulating miRNAs as biomarkers of cardiovascular diseases use relatively small samples, which results in many divergences between different reports. All of the miRNAs identified as potential biomarkers in these small, pilot studies should be validated in independent, large cohort studies designed specifically for biomarker validation in order to expedite the transition from bench to bedside.

Moreover, as ideal biomarkers, it would be interesting to explore the role of circulating miRNAs in prognosis [58]. Other than one report which showed that plasma miR-208b and miR-133a were associated with all-cause mortality at 6 months in acute coronary syndrome patients [59], the information regarding the prognostic impact of circulating miRNAs in cardiovascular diseases is still largely lacking.

It would also be interesting to identify circulating miRNAs that could predict response to a particular therapy in cardiovascular diseases. With this tool, caregivers would be able to optimize the risk-to-benefit ratio of a treatment in individual patients. Furthermore, by identifying specific circulating miRNA signatures related to a specific treatment, it would be possible to better determine the efficacy of treatments.

Lastly, RNA isolation from blood and subsequent quantification by real-time PCR is time consuming, thereby making the measurement of circulating miRNAs for use as a rapid beside test challenging [60]. As technologies continue to improve, miRNA profiling will get even easier, faster, and cheaper, and eventually clinically available. In fact, miRNA-based diagnostic assays have already been developed and approved for certain neoplastic diseases. Rarely has an opportunity arisen to advance such new biology for the diagnosis of cardiac diseases.

Conclusions

Circulating miRNAs have emerged as potential novel biomarkers for cardiovascular diseases. Their role as predictors of prognosis and therapy response remains to be determined.

Disclosures

None declared.

Copyright information

© Springer-Verlag 2011