MiRNAs in the Development and Progression of Heart Failure
Multiple mechanical and pathological stress triggers and cardiac injury evoke detrimental cardiac remodeling processes in the adult heart leading to chronic heart failure. It is not surprising that miRNAs are also involved in cardiomyocyte pathophysiological mechanisms (including cardiomyocyte hypertrophy, apoptosis, survival, and reactivation of the fetal gene program). However, as the heart comprises other important cell fractions, miRNAs altering functions of cardiac fibroblasts, which trigger extracellular matrix deposition and fibrosis and changes in endothelial-derived miRNAs that regulate angiogenesis, are also important mechanisms for heart failure development and progression. During heart failure development, many fetal genes, which are quiescent in the adult heart, are reactivated (Miyata et al. 2000; Nakao et al. 1997). As activation of gene programs in the fetal heart is regulated by miRNAs, miRNAs inducing fetal genes were also found in experimental and clinical studies. First insights came from a conditional knockout of Dicer, an enzyme needed for processing functional mature miRNAs. Knockout of Dicer led to cardiac remodeling processes and upregulation of fetal gene transcripts (viz., Acta1, Nppb, Myh7, and Nppa) (da Costa Martins et al. 2008). Comparison of miRNAs in experimental hypertrophic cardiomyopathy models showed >12 deregulated miRNAs as compared to sham-operated mice, which showed an overlap when comparing nonfailing versus end-stage heart failure tissues from humans (van Rooij et al. 2006). Seok et al. (2014) observed a downregulation of miR-155 in cardiomyocytes in a pressure overload model. Knockout of miR- 155 in mice repressed cardiac hypertrophy, partly by targeting Jarid2, a key transcriptional regulator of cardiac development and function. Formerly, miR-155 was known as an abundantly expressed miRNA in macrophages and monocytes. The loss of miR-155 in macrophages was found to inhibit leukocyte infiltration and protect murine hearts from hypertrophy, mostly by inhibition of pro-inflammatory macrophage-derived factors and downregulation of adhesion molecules (Heymans et al. 2013). These studies imply that targeted therapy of one miRNA may protect different cell types from pathological mechanisms - either by inhibition of paracrine secretion of macrophages or from hypertrophic response in cardiomyocyte after pressure overload.
Comparison of miRNA expression profiles using left ventricular tissue samples from patients with heart failure, nonfailing hearts, and fetal human heart tissues revealed a profound alteration in miRNA expression of heart failure tissues as compared to healthy tissue samples. In addition, up- and downregulated miRNA showed a > 85% coverage when miRNA expression of heart failure tissues was compared with fetal tissue, indicating a close relationship in molecular miRNA-dependent mechanism with the reactivation of fetal gene programs (Thum et al. 2007). Another elegant study performed a deep-sequencing analysis of RNA of human left ventricular tissue samples derived from nonfailing human LV and failing human LV before and after left ventricular assist device (LVAD) support (Yang et al. 2014). More than 147 miRNAs were differentially regulated when comparing nonfailing with heart failure LV samples; however, only two to five miRNAs returned to normal levels after LVAD support. These observations are consistent with a recent study where only subtle changes in miRNA expression between ischemic and dilated cardiomyopathy were detected, and miRNA profiling did not reveal differences before and after LVAD treatment (Akat et al. 2014). In contrast, >570 lncRNAs were found to be deregulated, mostly of mitochondrial origin, and approximately 10% of these lncRNAs normalized after LVAD support (Yang et al. 2014).
Changes in intracellular calcium handling are critical for heart failure, as they determine cardiac contractility. Gene therapy using an adenoviral vector containing sarcoplasmic reticulum Ca2+-ATPase (SERCA2) led to a decrease in clinical symptoms and reverse remodeling in heart failure patients in a small clinical phase II study (Jessup et al. 2011). SERCA2 is a calcium-transporting ATPase, which enables calcium uptake in the sarcoplasmic reticulum during relaxation of cardio- myocytes. Wahlquist et al. (2014) used the 3'-UTR region of SERCA2 as a sensor construct to identify miRNAs targeting SERCA2. MiR-25 markedly inhibited SERCA2 expression and was upregulated in the myocardium derived from patients with heart failure. Treatment with anti-miR-25 in a transaortic constriction (TAC) model in mice blunted progression of cardiac dysfunction and improved survival. Interestingly, in a study investigating experimental Hand2-induced cardiac hypertrophy, inhibition of miR-25 resulted in impaired cardiac function using a TAC model (Dirkx et al. 2013). Although the two studies used the same experimental hypertrophy model, inhibition of miR-25 was started 3 months (Wahlquist et al.
- 2014) versus 3 days (Dirkx et al. 2013) after TAC operation and cardiac function were assessed at 5.5 months (Wahlquist et al. 2014) as compared to 1 month (Dirkx et al. 2013) after TAC. The findings therefore suggest that miR-25 has different functions in the subacute versus chronic heart failure phase, and miR-25 expression may be dynamic in the course of hypertrophy and heart failure. However, these questions have to be addressed in future studies. Dynamic expression of miRNAs upon a stress trigger has been shown in various studies, such as for miR-212/132 (Ucar et al. 2012), miR-208 (van Rooij et al. 2007), and miR-195 (van Rooij et al.
Cardiac fibroblasts contribute to adverse remodeling processes and progression of heart failure. MiRNAs have been identified as critical regulators in cardiac fibroblasts, thereby contributing to extracellular matrix modulation and cardiac fibrosis. MiR-29 targets multiple collagens and expression of extracellular matrix proteins. In an experimental model of MI, miR-29 expression was markedly reduced in the infarct region and mostly of fibroblast origin when comparing expression between fibroblasts and cardiomyocytes (van Rooij et al. 2008). Interestingly, transforming growth factor p (TGFp), a key enhancer of cardiac fibrosis, decreased miR-29 expression in vitro, suggesting that upregulation of TGFp represses miR-29 expression, resulting in an enhanced deposition of extracellular matrix proteins (van Rooij et al. 2008). MiR-29 downregulation was also identified as a pro-fibrotic mechanism in pulmonary (Cushing et al. 2011) and renal (Qin et al. 2011) fibrosis. However, miR-29 is also involved in the progression of aortic aneurysms. Reduced expression levels of miR-29a and inverse correlation with aortic size have also been observed in patients with aortic aneurysm (Jones et al. 2011). Another miR-29 family member, miR-29b, was reported to be upregulated in patients and in experimental models of aortic aneurysm (no significant difference of miR-29a expression was observed in this study). Downregulation of miR-29 by delivery of LNA-modified antisense oligonucleotides in vivo resulted in enhanced expression of collagen members and reduction in aortic diameter (Boon et al. 2011). Therefore, whereas overexpression of miR-29 after myocardial infarction may reduce cardiac fibrosis, a reduced matrix deposition in miR-29-treated subjects may lead to the progression of aortic dilatation.
Another prominent fibrosis regulating miRNA is miR-21 that is upregulated in rodent models of ischemia-reperfusion (Roy et al. 2009) and hypertrophy (van Rooij et al. 2006). First shown as a pro-fibrotic miRNA in a pressure overload mouse model, which supports cardiac remodeling by an increase of ERK-MAP kinase activity in cardiac fibroblasts, it was later shown that upregulation of miR-21 after experimental ischemic preconditioning was protective in cardiac myocytes by inhibiting the expression of programmed cell death 4 (PDCD4) (Dong et al. 2009). In addition, sodium sulfide (Na2S), a donor for hydrogen sulfide, which is protective in various injury models of the heart, induces miR-21 in cardiomyocytes, indicating that the observed improvement in survival and decreased infarct size in an ischemia-reperfusion injury model are mediated by miR-21 through an inhibition of inflammasome function (Toldo et al. 2014). As miR-21 is expressed in both cardio- myocytes and cardiac fibroblasts, miRNA-mediated cell-to-cell communication was recently investigated. In this study, miR-21-3p (the star strand of miR-21, which is supposedly degraded) showed a high abundance in fibroblast-secreted exosomes and uptake of miR-21-3p in cardiomyocytes that resulted in cardiomyocyte hypertrophy via downregulation of SH3 domain-containing protein 2 (SORBS2) and PDZ and LIM domain 5 (PDLIM5) (Bang et al. 2014). These experimental studies imply that one miRNA may exert different functions in different cell types within the heart. In addition, miR-21-3p, a star strand that was earlier thought to be degraded, may also act as a functional miRNA. Moreover, the observation that miR- 21 is not needed to induce cardiac hypertrophy in a knockout model and an experimental model using a different anti-miR (8-mer LNA miR-21) (Patrick et al. 2010) and different cardiac heart failure models leading to divergent findings of miR-21 action suggests that spatiotemporal expression of miRNAs and different methodological approaches are important determinants of miRNA functional activity.