Journal of Cardiovascular Translational Research

, Volume 6, Issue 6, pp 884–898

Circulating MicroRNAs as Novel Biomarkers for the Early Diagnosis of Acute Coronary Syndrome

  • J. C. Deddens
  • J. M. Colijn
  • M. I. F. J. Oerlemans
  • G. Pasterkamp
  • S. A. Chamuleau
  • P. A. Doevendans
  • J. P. G. Sluijter
Article

DOI: 10.1007/s12265-013-9493-9

Cite this article as:
Deddens, J.C., Colijn, J.M., Oerlemans, M.I.F.J. et al. J. of Cardiovasc. Trans. Res. (2013) 6: 884. doi:10.1007/s12265-013-9493-9

Abstract

Small non-coding microRNAs (miRNAs) are important physiological regulators of post-transcriptional gene expression. miRNAs not only reside in the cytoplasm but are also stably present in several extracellular compartments, including the circulation. For that reason, miRNAs are proposed as diagnostic biomarkers for various diseases. Early diagnosis of acute coronary syndrome (ACS), especially non-ST elevated myocardial infarction and unstable angina pectoris, is essential for optimal treatment outcome, and due to the ongoing need for additional identifiers, miRNAs are of special interest as biomarkers for ACS. This review highlights the nature and cellular release mechanisms of circulating miRNAs and therefore their potential role in the diagnosis of myocardial infarction. We will give an update of clinical studies addressing the role of circulating miRNA expression after myocardial infarction and explore the diagnostic value of this potential biomarker.

Keywords

ACSCirculating microRNAsMyocardial infarctionBiomarkers

Introduction

Cardiovascular disease is still one of the leading causes of death in the Western world [1]. Acute coronary syndrome (ACS), a clinical diagnosis used to identify patients in need for cardiac revascularization, includes ST-elevated myocardial infarction (STEMI), non-ST-elevated myocardial infarction (NSTEMI), and unstable angina pectoris (UA). Here, myocardial infarction (MI) is defined as myocardial cell death due to prolonged periods of ischemia characterized by an imbalance between myocardial oxygen supply and demand [2, 3], which is mediated by necrotic and apoptotic cell death [2, 4, 5]. The main mechanisms of MI include coronary artery occlusion caused by atherosclerotic plaque rupture followed by thrombus formation and acute occlusion of a coronary artery by emboli or spasm.

The clinical diagnosis of ACS in the hospital is based on (1) clinical symptoms, like chest pain and a vasovagal response; (2) changes on the electrocardiogram, consisting of ST-segment elevation or depression or new pathological Q-waves; and (3) circulating biomarkers level changes, preferably troponin as described in the guidelines of the European Society of Cardiology and the American College of Cardiology [2]. At the emergency department, these parameters are used to classify patients with ACS. As explained above, different causes can lead to ACS, thereby including three subtypes: STEMI, NSTEMI, and UA. In the clinical setting, circulating biomarkers are very important to confirm the diagnosis of ACS since patients with NSTEMI or UA cannot be diagnosed based on clinical symptoms and ECG only. After final diagnosis of ACS, including STEMI, NSTEMI, and UA, clinical treatment for the different subgroups is similar.

Up to date, the most commonly used biomarker for MI is cardiac troponin (cTn) [2]. Troponin I and T subunits are part of the contractile myofibrils of cardiomyocytes and are released into the circulation after cardiomyocyte necrosis. The disadvantage of cTn is its time restrain since it is only detectable 3 to 6 h after clinical symptoms of cardiac ischemia have started [68], in contrast to, e.g., chemical cardiac injury where cTnT can already be detected 15 min after injury [9]. Therefore, 10 % to 15 % of patients with MI presenting to the emergency department do not yet have a positive cTn test [10, 11]. Since early diagnosis of MI maximizes the effects of cardiac revascularization therapy, there is an ongoing search for early diagnostic biomarkers.

Newly developed high-sensitive cardiac troponin T (hs-cTnT) assays have led to an improved diagnosis of MI as early as 3 h after the onset of symptoms [12, 13]. Nevertheless, guidelines still advise measuring troponin at presentation to the hospital and 3-6 h later [2] in order to see temporal changes in cTn concentrations, suggestive for MI. The main disadvantage of hs-cTnT assays is that in a substantial part of the population, especially elderly, continuously elevated levels of cTn are measured due to heart failure or chronic kidney disease [14, 15], resulting in false positive hs-cTnT assays [14].

An alternative to cTn is creatine kinase MB (CK-MB). Similar to cTn, CK-MB is released into the circulation upon cardiomyocyte necrosis and is related to infarct size and inversely to long-term survival [16]. The disadvantage of CK-MB as a biomarker of MI is its lower tissue specificity compared to cTn [17] and its delayed elevation (7 h) after the onset of symptoms [18]. Besides cTn and CK-MB, there are more candidate biomarkers for MI investigated that are released in the circulation upon cardiac damage, such as Heart-type fatty acid binding protein (H-FABP) [19, 20], C-reactive protein (CRP) [21], B-type natriuretic peptide (BNP) [22], C-terminal portion of provasopressin (copeptin) [23], and ischemia modified albumin (IMA) [24]. However, none of these biomarkers are elevated as early or are as specific as hs-cTnT [12, 25].

Despite the fact that hs-cTnT reveals an excellent diagnostic potential for the diagnosis of MI, having a large area under the curve (AUC) in receiver operator characteristics (ROC) analysis, there is still a need for new biomarkers that are able to reduce time between the onset of MI and the actual therapeutic interventions to further prevent cardiac ischemia and subsequent cell death. An ideal biomarker has seven different properties, including (1) high sensitivity and specificity for the disease of interest, (2) high reproducibility, (3) high stability inside and outside the human body, (4) measurable in linear proportion to myocardial damage, (5) directly measurable after disease onset, (6) a long biological half-life, and (7) advantages above the existing biomarkers. Moreover, preferably it should be measured relatively inexpensive [2527]. In the continuous search for new biomarkers, post-transcriptional processing have recently received an increased interest via the release and presence of small regulatory RNAs, called microRNAs (miRNAs), into several body fluids [2831]. Emerging evidence indicate that miRNAs are involved in various diseases and appeared to be circulating in the blood [30, 32], providing opportunities for new biomarker discoveries. Here, we review the advantages and disadvantages of circulating miRNAs as biomarkers for the diagnosis of myocardial infarction and briefly discuss the potential origin of identified miRNAs and their expression in cardiovascular disease.

MicroRNAs

MircroRNAs (miRNAs) are ~22 nucleotides-long non-coding RNAs that regulate gene expression at post-transcriptional level [33]. miRNAs are transcribed in the nucleus as precursor molecules, called primary miRNAs (pri-miRNA), and subsequently processed by RNase III enzyme, Drosha and DCGR8, into a precursor miRNA (pre-miRNAs). This molecule is transported by exportin-5 into the cytoplasm, where they are further processed by Dicer, another RNase III enzyme, into the mature single-stranded miRNA. Mature miRNAs regulate post-transcriptional gene expression by their uptake into the RNA-induced silencing complex (RISC) and subsequently binding to the 3′-, and sometimes the 5′-, untranslated region (UTR) of mRNA through imperfect base pairing [34], thereby repressing its translation via deadenylation or promotion of mRNA degradation [35, 36]. More detailed information on the biogenesis and function of miRNAs is described in reviews written by Bartel et al. [37], Ghildiyal et al. [38], Winter et al. [39], and Guo et al. [40]. Translational repression by miRNAs is thought to regulate various cellular processes in the human body including cell differentiation [41], proliferation [41, 42], and apoptosis [43]. Since mRNAs targets, regulated by miRNAs, are still largely unknown, possible target sites can be predicted genome-wide by evaluation of complementary nucleotide sequences.

Deregulation of miRNAs is associated with abnormal expression and translation of mRNAs and consequently with development of various pathological conditions. The first discovered miRNAs that were found to be involved in human disease were miRNA-15 and -16 that were both downregulated due to a deletion of a small region on the coding 13q14 gene, thereby causing lymphatic leukemia [44]. Emerging evidence has indicated that miRNAs also play a role in various other diseases such as progressive liver disease [45], neurodegenerative disease [28], and cardiovascular diseases [46]. Specific up- and downregulation of miRNAs give cardiovascular disease-specific miRNA expression signatures for, e.g. atherosclerosis [47, 48], angiogenesis [49], heart failure [50], and coronary artery disease (CAD) [51]. These differences in expression patterns are found through genome-wide miRNA profiling techniques, mainly via miRNA microarray analysis or deep sequencing techniques. Other traditional techniques that can be used for detection of these small RNA molecules are Northern blotting, the “golden standard”, and quantitative real-time PCR [52, 53].

Circulating miRNAs

Originally, miRNAs appeared to be located in the cytoplasm, but later studies revealed the presence of miRNAs in extracellular fluid, e.g. blood [30] and urine [54]. At first, the function of circulating miRNAs was unclear, but later, it was proposed that these extracellular circulating miRNAs are transported to different cells and are able to regulate post-transcriptional gene expression in recipient cells [55, 56].

Moreover, despite the presence of ribonucleases, miRNAs appeared to be stable in the extracellular fluids. Interestingly, it was observed that circulating miRNAs could resist multiple freezing and thawing cycles, high and low pH, and boiling [32], which makes them extremely suitable for biomarker research from large existing biobanks, and raised the question on how miRNAs are kept from degradation outside the protective cellular environment. There has been an increasing focus on the biological role of circulating miRNAs and their different transports methods in fluids [57]. Up to now, there are three established modes of miRNA transport and stabilization outside the cell (see Fig. 1):
Fig. 1

miRNA production, processing, and different modes of transport. Pri-miRNA is transcribed from DNA, processed by DCGR8 and Drosha to pre-miRNA, and subsequently transported out of the nucleus by exportin-5. Dicer processes the pre-miRNA into a single-stranded mature miRNA which can be loaded into the RISC complex and thereby will silence mRNA translation. It is thought that normal miRNA function is cytosolic, but precursor and mature miRNAs can also be transported outside the cell via (1) protein–miRNA complexes, (2) HDL–miRNA complexes, (3) exosomes which are released when a multivesicular body fuses with the cell membrane, (4) microvesicles or apoptotic bodies, and (5) free miRNAs by natural spill or necrosis

  1. 1.

    Protein-miRNA complexes: The majority of circulating miRNAs are stabilized by binding proteins [58]. The protein and miRNA form a protein-miRNA complex that protects the miRNA from RNAses in the plasma. Argonaute2 proteins [58] and nucleophosmin [59] are identified binding proteins found to pack and protect miRNAs in plasma and cell culture medium, respectively, but there may be other proteins that carry miRNAs which remain to be discovered.

     
  2. 2.

    High-density lipoproteins: High-density lipoproteins (HDL) bind and transport endogenous miRNA to recipient cells [60]. The trigger for miRNA export out of the cell is not known, although it is postulated that the ceramide pathway plays a role in inhibiting the export of miRNA to HDL. Ceramide is a sphingolipid which is regulated by sphingomyelinases, and inhibition of neutral sphingomyelinase 2 (nSMase 2) in macrophages results in increased miRNA export via HDL [60]. In addition, it is demonstrated that HDL-bound miRNA delivery to cells in vitro is mediated by scavenger receptor class B type I (SR-BI) [60, 61]. Recent studies suggest that miRNA-92a, -126, and -223 are miRNAs most abundantly bound to HDL [61] and might be taken up by endothelial cells [60]. Moreover, HDL isolated from patients with stable CAD alters the miRNA expression pattern of smooth muscle cells and peripheral blood mononuclear cells in vitro [61].

     
  3. 3.

    Encapsulation in vesicles: Microvesicles, formed by outward budding of the cell membrane [62], contain miRNAs, which are protected from RNAse-dependent degradation outside the cell [63]. Some of the miRNAs found in microvesicles are hardly located in the cytoplasm [62], suggesting active transport of miRNAs into microvesicles. Since it is demonstrated that microvesicles incorporate into target cells and regulate gene expression [62], it is thought that intercellular communication is, among other communication mechanisms, mediated by these microvesicles. Another type of membranous microvesicles is apoptotic bodies. These apoptotic bodies can also contain miRNAs and release them to neighboring cells [55]. Interestingly, there are only a few miRNAs enriched in apoptotic bodies compared to the abundant level of other miRNAs in the cell, e.g. miRNA-21 and -126. This would suggest, along with miRNA enrichment in microvesicles, that there is an active form of transport of certain miRNAs to other cells [55, 56, 62, 64].

    An additional mode of miRNA transport is by exosomes, which are smaller nano-sized vesicles that can also export miRNAs to other cell types [56, 65]. Exosomes are released from the cell when a multivesicular body fuses with the plasma membrane [66, 67] (see Fig. 1). It was observed that ceramide triggers the formation and release of exosomes [68], which is in contrast to the repression of miRNA export to HDL. Another stimulus that can trigger the extracellular amount of miRNA transported by exosomes, in a dose-dependent manner, is an increased intracellular Ca2+ concentration [69, 70], mimicked by calcium ionophore A23 [71].

     

Additionally, there is possibly also a passive release of free miRNAs into the circulation after, e.g., cardiomyocyte necrosis. However, which of these modes of transport plays a role under physiological conditions and in states of disease remains to be elucidated. Each mechanism alone or together could play a role in the elevation or decrease of the amount of circulating miRNAs upon stress responses.

Circulating miRNAs in Cardiovascular Disease

As stated before, several recent studies have shown that miRNAs play a physiological role in cardiovascular homeostasis and in the pathogenesis of cardiovascular diseases (CVD). For comprehensive overviews of the biological role of miRNAs, we would like to refer to excellent reviews on target recognition [72], regulatory functions [7276], and clinical implications [73, 75] in the cardiovascular field. Many miRNAs are known to be up- or downregulated as a cause or consequence of these pathologic processes; moreover, upon these stress signals, also the extracellular presence of miRNAs is altered and even detectable in body fluids. By using differences in miRNA expression levels, new cardiovascular biomarkers can be discovered and used for diagnostic, prognostic, or therapeutic purposes [77, 78]. Several interesting new biomarker candidates have promising potential, including miRNA 423-5p, which is expressed in cardiomyocytes and increased in blood samples from patients having heart failure [50], and miRNA-145, -155, -92a, -17, and -126 which are significantly altered in patients with CAD, compared to healthy controls [79]. Here, we will focus on miRNAs that are identified and could be used for the diagnosis of ACS and thereby monitor myocardial damage.

Cardiac miRNA Expression after MI

Expression pattern studies of myocardial tissue reveal that several miRNAs are up- or downregulated upon MI [49, 80]. Studies in this field are predominantly performed in murine models, showing deregulation of multiple miRNAs in the border zone of the infarcted area. Of these, miRNA-1, -29b, -126, and -499-5p are downregulated 1 week after MI. Interestingly, these miRNAs all recovered towards baseline expression levels 14 days after MI [81]. In contrast, another study revealed that miRNA-1 and -206 are increasingly elevated in the infarcted ventricle over a period of weeks after MI [82]. A study in mice showed that levels of miRNA-1, -133a, -208a, and -499 were reduced in the infarcted tissue 24 h after MI compared to healthy sham-operated mice hearts [71]. The reduction of miRNAs in the border zone of the infarction might suggest a release of miRNAs out of the cells upon MI, but could also reflect the remaining viable myocardial tissue.

In confirmation of the results, already shown in mice and rats, studies are performed on miRNA expression after MI in human patient heart samples. It was found that miRNA-1 and -133a/b were downregulated in infarcted tissue, and miRNA-1 was upregulated in the remote tissue of the heart. In addition, miRNA-150, -186, -208, -210, and -451 levels were increased in infarcted heart tissue and border zone samples, compared to healthy hearts [8385]. Furthermore, an upregulation of miRNA-21, -214, and -233 and downregulation of miRNA-29b and -149 in the border zone of human infarcted myocardium was shown [86].

The shift in miRNA expression is caused by (1) miRNA release from the cells, (2) changes in miRNA production, and (3) post-transcriptional processing. Furthermore, it is demonstrated that MI can induce a specific expression pattern of miRNAs, which are directly related and expressed by the myocardium or specifically induced by other cell types upon myocardial damage. These miRNAs can be potential therapeutic targets to reduce myocardial damage [73] and to influence adverse remodeling, but also used as biomarkers in a clinical setting as circulating miRNAs for the diagnosis of MI.

Circulating miRNAs after MI

Up to date, ~20 clinical studies have been performed to investigate whether circulating miRNAs can serve as potential new biomarkers for myocardial infarction (see summary in Table 1). The major goal of these studies was to assess whether miRNAs are suited for diagnosis of MI. Besides the diagnostic value of miRNAs, biological characteristics of the studied miRNAs are sometimes addressed as well. Both spatial changes [87, 88] and temporal changes are determined, which are of importance for the evaluation of potential new biomarkers and demonstrate the direct causal link to the disease. The different miRNAs studied, which are expressed in several tissues or cell types that are involved upon myocardial damage [89], can be divided into four groups:
Table 1

Overview of clinical studies describing the role of circulating miRNAs as diagnostic biomarkers for ACS

(1) Myocardium and Muscle-Expressed miRNAs

Probably the most useful biomarker for myocardial damage will be coming from the heart itself, and since some muscle-specific miRNAs are known, including miRNA-1, -133a/b, -208, and -499 [9092], these are the most abundantly studied.

miRNA-1

miRNA-1 is expressed both in the myocardium as in skeletal muscles, but most abundantly expressed in cardiomyocytes. The miRNA-1 family consists of miRNA-1-1 and miRNA-1-2, which have the same sequence but are located on different chromosomes (Chr), Chr 20 and 18, respectively [93]. miRNA-1 is well known to play a role in cardiogenesis [94], muscle differentiation [90, 94], and cardiac conduction [90, 95, 96]. It is demonstrated that miRNA-1 inhibits both KCNJ2, which codes for a K+ channel sub-unit Kir2.1, and GJA1, coding for connexin 43 gap junction channels [95]. In addition, a relation between cardiac conduction, widening of the QRS complex, and the expression of miRNA-1 is shown [97].

All studies that measured miRNA-1 expression in the circulation have shown that miRNA-1 is increased in the circulation after STEMI [71, 78, 88, 97105], NSTEMI [11, 88, 101, 104, 106], and UA [11, 104]. The level of upregulation differs between study groups. It is reported that miRNA-1 expression levels increased 100-fold to 300-fold within 12 h after the onset of symptoms [98, 100]. In contrast, another study demonstrated only a mild increase of three times, which did not reach significance [99]. Since it is demonstrated that these cardiac miRNAs are correlated with the glomerular filtration rate (GFR) [100], fast renal elimination of miRNA-1 might explain the discrepancy in levels between these studies. In addition, for the muscle-related miRNAs, skeletal muscle turnover can very well influence the baseline values of the circulating miRNAs. An increase in miRNA-1 expression in the circulation after the onset of MI can already be measured within 3 to 4 h after the onset of symptoms [71, 102], even before the peak time and potential detection of cTnT [71, 78] and normalizes to baseline only after 3 days [98]. Interestingly, another study could not detect an elevation of miRNA-1 already after 15 h [71]. Additionally, miRNA-1 is correlated with GFR [100], CK-MB [98], and hs-TnT [104], but not with risk factors such as age, gender, blood pressure, and diabetes mellitus [11]. Overall, the ROC analysis showed AUC values between 0.78 [71] and 0.98 [100], which ranges for a biomarker from reasonable to excellent.

miRNA-133a/b

miRNA-133a and miRNA-133b differ from each other by a single nucleotide [93, 107]. Although these two miRNAs are located on different chromosomes, they are both expressed in smooth muscle cells, skeletal muscle cells, and cardiomyocytes [90]. miRNA-133 is a key regulator of vascular smooth muscle cell proliferation, influencing the progression of atherosclerosis and playing a role in vascular remodeling [77, 90, 96, 108]. Furthermore, miRNA-133 is associated with enhanced myoblast proliferation [90] in skeletal muscles and with cardiogenesis [96] and modulation of cardiac hypertrophy [77] in pathophysiological conditions.

Circulating miRNA-133b expression has been rarely investigated, but limited studies that included miRNA-133b [78, 104] found an increased expression after MI for all subgroups of ACS. Circulating miRNA-133a is upregulated after MI and correlates with the concentration of cTn, suggesting that the release of miRNA-133a is associated with the extent of myocardial damage [87]. This was confirmed by the study of De Rosa et al., in which they explored the trans-coronary gradient of miRNA-133a in the aorta and in the coronary sinus, observing a positive gradient and thereby indicating that miRNA-133a is released from cardiomyocytes in the coronary circulation [87]. Like miRNA-1, miRNA-133 is significantly increased in UA patients as well, although lower as compared to patients presenting with a (N)STEMI [71, 104], possibly explained by less myocardial damage in UA patients. miRNA-133a and miRNA-133b reached their peak levels even before cTnI, which peaked at 3 h [71, 78, 109]. The increased serum presence in time of miRNA-133b remains debatable since it varies from 33 h [71] up to 90 days [105]. According to Wang et al., miRNA-133a is superior to hs-cTnT since it is demonstrated that the concentration of miRNA-133a in suspected patients is not subject to chronic kidney disease [109], and thereby very promising since the weakness of the currently used biomarker cTn, especially the cTnT isoform [110], is the late elevation after MI combined with the low specificity in older patients with chronic kidney disease.

miRNA-208a/b

miRNA-208a is a cardiomyocyte-specific miRNA and therefore has gained high interest in diagnosing acute MI [111]. The miRNA-208 family consists of three members: miRNA-208a, miRNA-208b, and miRNA-499. miRNA-208a and miRNA-208b are both located on Chr 14 and have identical nucleotide sequences of the seed region, the main target recognition site, but have a difference of three nucleotides in the remaining sequence. miRNA-208a is transcribed from an intron in the Myh6 gene (also known as α-MHC) and is expressed in the heart and miRNA-208b from an intron in the Myh7 gene (also known as β-MHC) in both heart and skeletal muscle, where they control myosin characteristics and myofibril specifications [87]. Functionally, miRNA-208 plays a role in cardiac fibrosis and cardiomyocyte hypertrophy during stress and hypothyroidism [111]. Moreover, miRNA-208a plays a role in cardiac conduction by regulating, among other proteins, gap junction protein connexin 40 [112].

The findings on miRNA-208a are diverse, and several groups were not able to conclusively detect circulating miRNA-208a [11, 78, 100, 106, 113]. It is therefore suggested that miRNA-208a is not a good biomarker; however, we were able to find a significant increase in miRNA-208a levels in NSTEMI patients [11] and not in patients with UA or non-ACS patients. Wang et al. found detectable miRNA-208a in 90.9 % of patients with MI and not in healthy people or with coronary heart disease without MI [103]. Consistent with these results, higher levels of circulating miRNA-208a in (N)STEMI patients compared to patients with UA are shown [104]. In contrast to the disappointing results mentioned before [71, 78], these results demonstrated the discriminative power of miRNA-208a for MI. The ROC analysis showed an AUC of 0.965, which is very close to the AUC of cTn (0.987) [103]. In addition, it has been demonstrated that within 4 h after the onset of symptoms, 100 % of the MI patients have elevated levels of miRNA-208a, while only 85 % of these patients have a cTnT elevation [103]. Subgroup analysis revealed a better specificity and sensitivity for miRNA-208a in patients presenting with chest pain and an initially negative cTn, suggesting an important role for miRNA-208a as an early diagnostic biomarker for ACS [103]. A possible explanation for variation in detection of miRNA-208 between different study groups is the variety in the time of sampling (8.5-48 h) after onset of symptoms. It is thought that miRNA-208a is rapidly cleared from circulation and therefore is not detectable at later time points, e.g., 48 h.

miRNA-208b is also upregulated after MI [113], ranging from 1,600-fold to [99] 5 × 105-fold [114] and remained up to 3 days [100]. A difference was also found between STEMI vs. NSTEMI patients with an eight-fold increase of miRNA-208b levels in STEMI patients [114]. In addition, there is a good correlation with cTn 24 h post-MI [100], suggesting a correlation with myocardial damage. ROC analysis demonstrated an AUC varying from 0.89 [101] to 1 [100], suggesting good to excellent accuracy and specificity to diagnose MI.

miRNA-499

As stated before, miRNA-499 is a family member of miRNA-208 and is located in an intron of the Myh7b gene and almost exclusively produced in the heart [113], although maybe limited in slow contracting skeletal muscles [92]. miRNA-499 is co-expressed with MYH7Bβ [41] and regulates the myosin gene and mitochondrial dynamics [41, 115]. Also, miRNA-499 was demonstrated to be involved in cell proliferation and cardiomyocyte differentiation [41] and is therefore proposed as a target for regenerative therapy [116].

Similar as the other muscle-specific miRNAs, miRNA-499 levels were found to be increased between 100-fold [99] and 3,000-fold [114] after MI. miRNA-499-5p circulating levels increase in the circulation within a few hours [11] and have a peak time between 6 and 12 h [113] after MI, which is slower compared to cTnI [78]. A positive transcoronary gradient was also demonstrated for miRNA-499, which suggests its release from myocardial cells during MI [87], further supported by the correlation of miRNA-499-5p with cTnT [106], hs-TnT [87, 114], and CK-MB [113].

The ROC of miRNA-499 showed AUC values between 0.822 [103] and 0.97 [114], suggesting that miRNA-499 would be a good diagnostic biomarker. In a subgroup of samples, obtained from patients with initially low levels of cTn, the AUC of miRNA-499-5p was superior to both cTnT and hs-cTnT [11], suggesting that miRNA-499-5p could improve fast diagnosis of MI, although we do not know yet how fast this is detectable post-MI.

Based on these results, miRNA-1, -208, and -499 seem to be good biomarkers for early detection of MI. It is demonstrated that miRNA-1 and -133a, which are co-transcribed and correlated after MI [100], are not upregulated in other heart diseases [103] and after non-myocardial ischemia such as hindlimb ischemia. However, healthy people might show low levels in their blood of all three miRNAs due to normal muscle turnover [103], and it is demonstrated that in situations of skeletal muscle damage, an upregulation of all three miRNAs could be observed [103, 117], which might give false positive outcomes. miRNA-1, -208, and -499 demonstrate a better diagnostic value compared to miRNA-133 [99, 100], as shown by the AUC of the ROC analysis curve. For early diagnosis of ACS, a single miRNA as biomarker might not be as powerful as the combination of several miRNAs in a biomarker panel. It is postulated that [118] a multi-marker approach is cost attractive, might increase diagnostic power, and provides more insight in the process of myocardial damage after MI. In addition to these muscle-specific miRNAs, e.g. vascular wall- and leukocyte-associated miRNAs might add more profound information of underlying mechanisms and an earlier detection of myocardial stress.

(2) Vascular Wall-Expressed miRNAs

Since vascular changes and stress signals are also following MI, miRNAs associated with the vasculature might be of interest to add to the cardiac-specific miRNAs, e.g. miRNA-126 [119], miRNA-92a [49], and miRNA-145 [120].

miRNA-126

miRNA-126 is located on Chr 9 and is highly expressed in endothelial cells where it is found to modulate both vasculogenesis and angiogenesis [121, 122]. Moreover, miR-126 regulates the turnover of vascular smooth muscle cells and plays a role in the protection against atherosclerotic lesion development in ApoE knockout mice [55, 123].

Circulating levels of miRNA-126 are downregulated post-MI, as compared to healthy subjects [102], and these levels correlate with the time trend seen in cTnI expression [102]. Downregulation is already seen 4 h after the onset of symptoms and reaches its lowest point after 24 h, after which, miRNA-126 normalizes and reaches a second low point again after 1 week. This suggests that it is also involved in later phases post-MI, like inflammation or angiogenesis that usually follow MI.

Interestingly, a five-fold increase in the expression of miRNA-126 is found in blood sample taken from the aortic bulb [87]. In trans-coronary experiments, they observed a further decrease in levels of circulating miRNA-126 within the coronary system as compared to the aortic bulb, and this decrease was related to the amount of cardiac damage. This would suggest a form of “consumption” of miRNA-126 in the coronary arteries after MI [87] and possibly in the arterial and venous system as well, which is possibly explained by an increased uptake of miRNA-126 in blood vessels after MI, thereby referring to an active signaling cascade.

miRNA-92a and miRNA-145

Circulating miRNA-92a is located on Chr 13 and controls angiogenesis as described in mouse models [49]. Like miRNA-126, circulating miRNA-92a seems to be upregulated two-fold after MI in the aortic bulb blood sample, compared to people with stable coronary artery disease. A similar decrease in gradient of this miRNA was seen in the trans-coronary experiment, although it did not reach statistical significance [87].

Circulating miRNA-145 plays a role in the differentiation of vascular smooth muscle cells and represses their proliferation [120]. After MI, an increase in miRNA-145 levels is observed, which might be related to plaque rupture and vessel injury [10], which is correlated with cTnT (0.71).

Although differently expressed, both miRNA-92a and miRNA-126 would not serve as good biomarkers since a decrease in abundance is more difficult to detect as an increase in a general patient population. In addition, the AUC of miRNA-126 is not as good as studies have shown for cTn. In order to determine the clinical diagnostic value for miRNA-145 as a biomarker, more research is necessary.

(3) Leukocyte and Platelet-Expressed miRNAs

miRNAs associated with circulating leukocytes, such as miRNA-146 and -155, can be valuable for the diagnosis of MI since, e.g. peripheral blood monocytes are involved in plaque stability, plaque rupture, and have a role in early inflammation after MI [124], and therefore may be monitors of early stress and for myocardial damage [51].

To our knowledge, several miRNAs are associated with platelets and are functional released upon MI, thereby taken up by endothelial cells and affect their cellular behavior [125]; these include miRNA-21, -22, -24, -34, -126, -185, -191, -197, -223, -320b, and -423-5p.

miRNA-146, miRNA-155, and miRNA-223

miRNA-155 is located on Chr 21 and present in activated macrophages [126]. Additionally, miRNA-155 is expressed in several other cell types [127], playing a role in the Ig class switch of B-lymphocytes [128], or is involved in cell death, by improving cell survival and modulating protease activity [129, 130].

miRNA-233 is located on Chr X [93, 107] and is involved in several processes, e.g. progenitor cell proliferation, granulocyte function [131], and attenuation of inflammation [132]. Furthermore, a micro-array on myocardial tissue after ischemia reperfusion injury revealed that miRNA-223 is upregulated after 1 and 3 days and correlates with the inflammatory response. Besides its expression in leukocytes, miRNA-223 is also highly enriched in platelets [132]. In the study of De Rosa et al., miRNA-155 and -223 levels were modestly, not statistically significant, elevated after MI [84]. Corsten et al. found a statistically significant decrease of miRNA-223 in AMI patients [99]. Both studies did not find an association of miRNA-223 with cTn [87, 99]; also, miRNA-155 was not associated with circulating cTn [87].

miRNA-146 is located on Chr 5 and involved in the inflammatory response of leukocytes; its expression can be induced by LPS, and it regulates toll-like receptor (TLR) and cytokine signaling [133]. We have observed that circulating miRNA-146a levels were increased 11.7-fold 3.2 h after MI [11] and apparently more elevated in NSTEMI patients compared to UA. In a subgroup of 152 patients, who presented in the hospital within 3 h after the start of complains, the miRNA-146a concentration was already elevated [11]. In addition, in the hs-cTnT-negative subgroup of 194 patients, miRNA-146a was significantly elevated, suggesting that miRNA-146a is an earlier biomarker than hs-cTnT for myocardial damage; however, when three miRNAs, miRNA-1, -21, and -499, are taken together to diagnose MI, miRNA146a does not add any additional diagnostic value for hs-cTnT-negative patients [11].

Up to now, although miRNAs associated with leukocytes are increased upon MI, their predictive value seems to add little to muscle cell-related miRNAs for the diagnosis of MI.

miRNA-21

miRNA-21 is located on Chr 17 and plays a role in the regulation of cardiac hypertrophy, where miRNA-21 is downregulated compared to healthy hearts [134]; moreover, it plays a role in the formation of cardiac fibrosis by influencing the ERK-MAP kinase signaling pathway [135].

Circulating miRNA-21 was elevated 18.9-fold after MI and also detectable in the hs-cTnT-negative subgroup; moreover, miRNA-21 was elevated even within 3 h after the onset of symptoms [11]. At day 2 post-MI, the levels of miRNA-21 decreased but surprisingly increased again at day 5. At day 90, the concentration had returned to baselines [105]. The individual ROC of miRNA-21 had an AUC of 0.76 and is thereby lower than cTn or hs-cTnT. However, combined with miRNA-499 and -1, miRNA-21 adds diagnostic value (AUC 0.89) compared to hs-cTnT (0.86) [11]. Therefore, miRNA-21 does have an advantage at early time points after MI when cTn is not detectable yet.

(4) Other miRNAs

Besides the most abundantly studied miRNAs described above, relatively new miRNAs are discovered due to high throughput micro-array screening, e.g. miRNA-122, -145, -375, -663b, and -1291 [10, 78, 113]. After MI, miRNA-122 and -375 are downregulated and reach their minimum at approximately 1-2 days after onset of symptoms [78]. Furthermore, a selection of miRNAs (both up- and downregulated) can be combined for a higher discriminative and diagnostic power, but still need validation in large patient cohorts to elucidate their role as diagnostic biomarkers.

Discussion

miRNAs play an important role in almost all cellular processes and are involved in the initiation and progression of different cardiovascular diseases where they are deregulated in both tissue as well as extracellular fluids. Although miRNAs are well known for their complex working mechanism, targeting multiple genes in different types of tissue, the specific expression patterns of circulating miRNAs in CVD, their stability and early release from, e.g., cardiomyocytes into the circulation make them extremely suitable as biomarkers for ACS. The release of cardiac tissue-specific miRNAs have a possible functional advantage compared to non-cardiac tissue-specific miRNAs; however, both can be used as biomarkers for ACS when proved specific.

Compared to cTn, the biochemical golden standard for diagnosis of ACS, miRNAs are detectable in the circulation at an earlier time point after MI. cTn is part of the myofibrillar apparatus, being large protein complexes, whereas the majority of miRNAs are also protein bound, but relatively smaller and are released in a controlled way, which can explain the time difference in release after MI (Fig. 2). miRNAs can be released more quickly, which benefits early MI diagnosis and revascularization therapy.
Fig. 2

Detection of miRNAs after MI. The moment of first detection (blue), peak range (red), and concentration above baseline level (green) in hours after MI [11, 71, 78, 102, 113, 114], compared to cTn and CK-MB [18, 137]

Although miRNAs as early biomarkers for ACS appear to be very promising, the translation of these markers from bench to bedside is delayed by some limitations, e.g. the currently used study populations and the time for RNA detections. From these, the major drawback for using miRNAs as biomarkers for clinical diagnosis of ACS is their laborious isolation and detection procedures. Compared to the detection of cTn, which is an ELISA-based method, circulating miRNAs are detected based on PCR techniques which are very time consuming, and for that reason, cTn is still the best biomarker in clinical practice. In order to decrease the time of uncertainty until the diagnosis, there is an urgent need for further technical development of miRNA detection in non-classical RNA detection approaches. Another limitation is that most of the clinical studies included small numbers of patients, usually using healthy controls as references and displayed a large variation in the time of blood sampling after MI and in isolation and detection methods to analyze miRNAs. The use of a healthy control group is debatable since a biomarker for MI should discriminate between patients presenting to daily clinical practice, e.g. at the emergency department with chest pain caused by ACS or non-ACS. It is demonstrated that in the majority of the clinical studies, the discriminative power of a single miRNA is still lower compared to cTn, possibly caused by the type of study group, study group size, and unstandardized methods.

Fortunately, three studies did investigate miRNAs of patients with chest pain suspicious for MI presenting at the emergency department [11, 99, 103]. In a cumulative comparison of multiple miRNAs, an increased specificity and sensitivity for the diagnosis of MI is shown [11, 136]. Even in combination with hs-cTnT and a clinical model, including patient history and cardiovascular risk factors, miRNAs add discriminative power, which support the notion that circulating miRNAs can be used as independent early diagnostic markers for ACS [86]. Even more promising is the presence of miRNAs in ACS patients that presented with chest pain suspicious for ACS with an initial negative troponin [11]. The combined assessment of multiple circulating miRNAs together with cTn could be used to increase specificity and sensitivity and accelerate MI diagnosis. A deliberately chosen selection of miRNAs, e.g. muscle-, vascular wall-, and leukocyte-associated miRNAs, can increase diagnostic power. A panel like this should contain miRNAs, both up- and downregulated, that embody the full mechanism of myocardial damage and subsequent inflammatory and reparative process, thereby optimizing diagnostic power of miRNAs for early diagnosis of ACS.

In conclusion, circulating miRNAs as biomarkers for ACS holds great potential. Up to date, ~20 studies are performed which show an early upregulation of muscle-specific miRNAs after MI, related to the extent of myocardial damage with clear discriminating and diagnostic power. Currently investigated miRNAs for early diagnosis of MI should be validated in larger patient cohorts, and the additional value of new biomarker candidates from micro-array discoveries need to be elucidated.

Acknowledgements

We acknowledge the support from “Stichting Swaeneborgh”, a ZonMW Translational Adult Stem Cell (TAS) grant 1161002016, and the Netherlands CardioVascular Research Initiative: the Dutch Heart Foundation, Dutch Federation of University Medical Centers, the Netherlands Organization for Health Research and Development, and the Royal Netherlands Academy of Sciences.

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • J. C. Deddens
    • 1
  • J. M. Colijn
    • 1
  • M. I. F. J. Oerlemans
    • 1
  • G. Pasterkamp
    • 1
  • S. A. Chamuleau
    • 1
  • P. A. Doevendans
    • 1
    • 2
  • J. P. G. Sluijter
    • 1
    • 2
  1. 1.Department of Cardiology, Division Heart and LungsUniversity Medical Center UtrechtUtrechtThe Netherlands
  2. 2.Netherlands Heart Institute (ICIN)UtrechtThe Netherlands