Despite a recent decline of in-hospital mortality attributable to acute myocardial infarction (AMI), the incidence of ischemic heart failure (HF) in post-AMI patients is increasing. Although various microRNAs have been proposed as diagnostic indicators for AMI, no microRNAs have been established as predictors of ischemic HF that develops after AMI.
We attempted to identify circulating microRNAs that can serve as reliable predictors of ischemic HF in post-AMI patients.
Using sera collected a median of 18 days after AMI onset, we screened microRNAs in 21 patients who experienced development of HF within 1 year after AMI and in 65 matched controls without subsequent cardiovascular events after discharge. Among the 377 examined microRNAs, the serum level of only miR-192 was significantly upregulated in AMI patients with development of ischemic HF. Because miR-192 is reported to be p53-responsive, the serum levels of 2 other p53-responsive microRNAs, miR-194 and miR-34a, also were investigated. Interestingly, both microRNAs were coordinately increased with miR-192, particularly in exosomes, suggesting that these microRNAs function as circulating regulators of HF development via the p53 pathway. Furthermore, miR-194 and miR-34a expression levels were significantly correlated with left ventricular end-diastolic dimension 1 year after AMI.
In the sera of post-AMI patients who experienced development of de-novo HF within 1 year of AMI onset, the levels of 3 p53-responsive microRNAs had been elevated by the early convalescent stage of AMI. Further investigations are warranted to confirm the usefulness of these circulating microRNAs for predicting the risk of development of ischemic HF after AMI.
Although recent advances in the management of acute myocardial infarction (AMI), including primary percutaneous coronary intervention strategies and evidence-based therapies, have resulted in a substantial decline in mortality, the number of post-AMI patients who survive AMI but experience development of ischemic heart failure (HF) is increasing worldwide.1 Therefore, the identification of biomarkers that can predict risk of HF development in post-AMI patients is needed for optimizing management and treatment strategies. To date, several types of biomarkers, such as N-terminal probrain natriuretic peptide and cardiac troponin T, have been shown to predict cardiovascular events after AMI; however, it remains inconclusive whether these biomarkers can predict future HF in post-AMI patients.2
In This Issue, see p 239
Editorial see p 242
MicroRNAs are small endogenous noncoding RNAs that regulate gene expression posttranscriptionally. Recently, various cell types were found to release microRNAs in membrane-bound vesicles termed exosomes, which circulate stably in the bloodstream.3–5 Although the physiological significance of circulating microRNAs is not fully understood, they have attracted attention as potential diagnostic and prognostic biomarkers for various diseases, particularly cancer.3,4
With respect to cardiovascular disease, several cardiac microRNAs, including miR-1, miR-133a, and miR-208a, have been detected in the serum in the acute phase of AMI and thus represent potentially useful diagnostic markers for AMI.5,6 However, these microRNAs are most likely released from necrotic heart tissue into the blood directly and are not encapsulated within exosomes and, therefore, have short half-lives. For this reason, these cardiac microRNAs are unlikely to be predictive of future HF development in post-AMI patients.6 The aim of the present study was to identify circulating microRNAs that can serve as predictors of HF development in patients who survive the acute stage of AMI.
We retrospectively analyzed the records of patients registered in the Osaka Acute Coronary Insufficiency Study, which has been described elsewhere.7 The study protocol was approved by the ethics committee of each participating hospital, and written informed consent was provided by each patient at the time of registration. On the basis of the results of an initial screening (Online Table I), we performed a second screening to examine the microRNA profiles of an increased number of matched patients (HF group, n=21; control group, n=65; Online Table II). Detailed methods are described in the Online Data Supplement. Results were analyzed by the Mann–Whitney U test or Student t test. Correlations were tested using Pearson correlation coefficient. Statistical significance was set as P<0.05, P<0.01, or P<0.001.
To identify microRNAs with altered expression in accordance with HF development after the convalescent stage of AMI, we initially compared the expression level of individual microRNAs in sera collected a median of 18 days after AMI onset using a high-throughput array between HF and control groups (n=7, respectively). Among 377 microRNAs, the expression levels of 14 microRNAs were found to significantly differ between the 2 groups when U6 small nuclear RNA, miR-766, or let-7d was used as an internal control (Online Table I). The serum levels of the 14 microRNAs were further examined in a second screening. Two of the microRNAs, miR-485-3p and miR-518d-3p, were below the limit of detection of the assay. However, the expression level of miR-192 was significantly upregulated in the HF group, whereas those of the remaining 11 microRNAs were comparable between the 2 groups (Figure 1A; Online Figure I).
Figure 1. Upregulation of 3 specific microRNAs in sera of the heart failure (HF) group. The expression levels of miR-192 (A) and of miR-194 and miR-34a (B) in sera of the HF (black bars, n=21) and control groups (white bars, n=65) are displayed after normalization to U6 snRNA. Similar results were obtained when miR-766 or let-7d was used as an internal control. *P<0.05, **P<0.01.
Because miR-192 is reported to be p53-responsive, the levels of 2 p53-responsive microRNAs, miR-194 and miR-34a,8,9 also were measured. As anticipated, the serum levels of miR-194 and miR-34a also were upregulated in the AMI patients who experienced development of ischemic HF (Figure 1B). Furthermore, the miR-192, miR-194, and miR-34a expression levels were significantly correlated with one another, suggesting that the 3 microRNAs were coordinately upregulated in a single cascade, such as the p53 signaling pathway. In particular, a high correlation (r=0.86) was detected between miR-194 and miR-34a (Online Figure II). We also examined the serum levels of miR-208 and miR-499, which are candidate markers for severe myocardial damage and necrosis,10 but neither microRNA was detected in the serum of either the HF group or the control group.
To investigate whether miR-192, miR-194, and miR-34a were released into the serum within exosomes or had directly leaked from necrotic heart tissue, we separated collected serum into exosome and supernatant fractions. Western blot analysis against CD63, an exosome marker protein, confirmed the purity of the 2 fractions (Figure 2A). The levels of the 3 microRNAs were then quantified in each fraction, demonstrating that all 3 microRNAs were highly enriched in the exosome fraction (Figure 2B). We further compared the expression levels of these microRNAs in each fraction between the HF and control groups. Although no significant differences were detected in the supernatant fraction, the expression levels of miR-192 and miR-194 in the exosome fraction of the HF group were significantly higher than those of the control group (Figure 2C). A similar tendency also was observed for miR-34a, suggesting that the 3 microRNAs were mainly released into the serum within exosomes. Using cultured myoblasts derived from embryonic rat heart, we confirmed that both intracellular and extracellular levels of miR-192, miR-194, and miR-34a were increased after p53 activation, and that knockdown of these 3 microRNA increased cell viability, whereas transfer of extracellular microRNAs via the culture media decreased cell viability of myoblasts (Figure 3; Online Figure III).
Figure 2. Circulating miR-34a, miR-192, and miR-194 are highly enriched in serum exosomes. A, Western blot analysis against CD63 in the exosome (Exo) and supernatant fractions (Sup) of the serum. B, Expression levels of miR-192, miR-194, and miR-34a in the serum exosome fraction (black bars, n=4) and supernatant fraction (white bars, n=4) are displayed after normalization to U6 snRNA. C, Expression levels of miR-192, miR-194, and miR-34a in the serum Exo fraction and Sup fraction obtained from the heart failure group (gray bars, n=4) and control group (white bar, n=4) are shown. *P<0.05.
Figure 3. p53-responsive microRNAs were upregulated after p53 activation and influenced cell viability. Treatment of rat H9c2 cells with doxorubicin (Dox) increased p53 protein level (A), followed by an increase of caspase 3/7 activity (B) and a decrease of cell viability (C). Along with p53 activation, intracellular (D) and subsequently extracellular (E) levels of miRNA-192, miRNA-194, and miRNA-34a were upregulated. Dox treatment after knockdown of all 3 microRNAs, but not individual microRNA, in H9c2 cells increased cell viability not in 24 hours, but in 36 hours (F), whereas coculture with media containing 3 p53-responsive microRNAs decreased cell viability in the presence of Dox (G). n=3-5 for each experiment (B-G). *P<0.05, **P<0.01, ***P<0.001.
Finally, we analyzed whether the serum levels of the 3 p53-responsive microRNAs were correlated with left ventricular diastolic dimension, a clinical parameter of cardiac remodeling. The expression levels of miR-194 and miR-34a, but not miR-192, were positively correlated with left ventricular diastolic dimension, which was measured for 58 patients by echocardiography ≈1 year (median, 402 days) after the onset of AMI (miR-194: r=0.33, P=0.01; miR-34a: r=0.38, P=0.003; and miR-192: r=0.09, P=0.52; n=58; Figure 4). Furthermore, the serum levels of miR-194 (r=−0.28; P=0.03) and miR-34a (r=−0.27; P=0.04), but not miR-192 (r=−0.21; P=0.10), were correlated with left ventricular ejection fraction measured ≈1 year after AMI onset. These correlations were not altered after consideration of the history of HF-related medications as a potential confounding factor (Online Table III). This finding suggested that the serum levels of these p53-responsive microRNAs could predict left ventricular remodeling after the convalescent stage of AMI.
Figure 4. Circulating levels of miR-194 and miR-34a are correlated with left ventricular diastolic dimension (LVDd) measured ≈1 year after acute myocardial infarction (AMI) onset. Each dot represents the serum expression levels of miR-192 (left), miR-194 (middle), and miR-34a (right) relative to U6 snRNA (−ΔCt) and LVDd that was measured ≈1 year after the onset of AMI. Values for the heart failure group are indicated by black dots (n=12) and those of the control group are indicated by open dots (n=46). CI indicates confidence interval.
In this study, we demonstrated that serum levels of the microRNAs, miR-192, miR-194, and miR-34a, are upregulated as early as a median of 18 days after AMI onset in patients who survived AMI but experienced development of HF within 1 year. Furthermore, circulating miR-194 and miR-34a levels in the convalescent stage of AMI were associated with left ventricular diastolic dimension. Thus, these p53-responsive microRNAs may be useful for stratifying the risk of future ischemic HF events and cardiac remodeling in post-AMI patients.
Because the expression levels of the 3 identified microRNAs were well-correlated with each other, it is conceivable that their expression was coordinately upregulated in a single cascade. One of the most likely candidates for an inducer of this coordinated expression is the tumor suppressor p53, because all 3 microRNAs are reported to be induced by the direct binding of p53 to the promoter regions of the corresponding genes.8,9 In particular, it was recently demonstrated that miR-34a is induced in a p53-dependent manner, and that cardiomyocyte apoptosis can be suppressed through the blockade of miR-34a processing in a mouse model of AMI.11 We confirmed that these 3 microRNAs are induced by p53 activation in cultured myoblasts and are subsequently released from cells. Furthermore, transfer of these extracellular microRNAs accelerated cell death (Figure 3). Taken together with our recent finding that p53 accumulation in the myocardium in response to pressure overload or after MI plays an essential role in HF progression in mice,12,13 our present results suggest that activation of p53 and the increased expression of the p53-responsive microRNAs, miR-192, miR-194, and miR-34a, are likely involved in the pathogenesis of HF after AMI. However, it remains undetermined whether circulating p53-responsive microRNAs were derived from infarct or ischemic heart tissues.
Interestingly, circulating miR-208b and miR-499, which are both reported to reflect severe myocardial damage or necrosis,10 were not detected in the serum of either HF patients or control patients in the present study. Thus, detection of p53-responsive microRNAs may have reflected ongoing myocardial damage associated with the development of future HF, which was qualitatively different from necrosis.
It was recently reported that miR-34 family members are upregulated in the heart in response to stress, including AMI, and that systemic injection of miR-34 anti-miR significantly attenuated cardiac remodeling and dysfunction in a mouse AMI model.14 Accordingly, modification of the p53 pathway using microRNA-based technologies may be a novel therapeutic approach to prevent ischemic HF in post-AMI patients, particularly for those with increased levels of circulating p53-responsive microRNAs.
In conclusion, we identified 3 p53-responsive microRNAs that were upregulated in the serum of post-AMI patients who experienced development of HF within 1 year of AMI onset. Further investigations with the increased number of samples and in other cohorts are required to confirm the present findings and future clinical applications of miR-192, miR-194, and miR-34a as predictive indicators of HF.
The authors thank all of the collaborators, research nurses, and coordinators who participated in the Osaka Acute Coronary Insufficiency Study (OACIS).
This research was supported by grants-in-aid for
None.
| AMI | acute myocardial infarction |
| HF | heart failure |
In May 2013, the average time from submission to first decision for all original research papers submitted to Circulation Research was 15 days.
*These authors contributed equally to this work.
This article was sent to Mark Sussman, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.113.301209/-/DC1.
A number of patients survive acute myocardial infarction (AMI) but experience development of chronic ischemic heart failure (HF).
MicroRNAs are noncoding small RNAs that are released from various cell types and circulate in the bloodstream.
Accumulation of p53, a multifunctional protein, after AMI plays an essential role in HF development in mice.
Serum levels of 3 p53-responsive microRNAs, miR-192, miR-194, and miR-34a, are higher in the early convalescent stage of AMI (median 18 days after onset) in patients with development of HF within 1 year.
Serum levels of miR-194 and miR-34a correlate with left ventricular diastolic dimension, a clinical parameter of cardiac remodeling, measured 1 year after AMI.
Cultured myoblasts exposed to miR-192, miR-194, and miR-34a released from other cells have decreased viability.
Although measurements of circulating microRNAs have been reported to be useful for the diagnosis of AMI and HF, it is not clear whether microRNA levels are reliable predictors of ischemic HF that develops after AMI. We found that 3 p53-responsive microRNAs are upregulated during the early convalescent stage of AMI in the sera of patients who experienced development of HF within 1 year after discharge. Moreover, serum levels of these microRNAs were positively correlated with left ventricular diastolic dimension 1 year after AMI. These findings suggest that circulating p53-responsive microRNAs may be clinically useful as predictive indicators of HF in post-AMI patients.
In May 2013, the average time from submission to first decision for all original research papers submitted to Circulation Research was 15 days.
*These authors contributed equally to this work.
This article was sent to Mark Sussman, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.113.301209/-/DC1.
联系客服
