Inhibition of miR-154 Protects Against Cardiac Dysfunction and Fibrosis in a Mouse Model of Pressure Overload

Expression of miR-154 is upregulated in the diseased heart and was previously shown to be upregulated in the lungs of patients with pulmonary fibrosis. However, the role of miR-154 in a model of sustained pressure overload-induced cardiac hypertrophy and fibrosis had not been assessed. To examine the role of miR-154 in the diseased heart, adult male mice were subjected to transverse aortic constriction for four weeks, and echocardiography was performed to confirm left ventricular hypertrophy and cardiac dysfunction. Mice were then subcutaneously administered a locked nucleic acid antimiR-154 or control over three consecutive days (25 mg/kg/day) and cardiac function was assessed 8 weeks later. Here, we demonstrate that therapeutic inhibition of miR-154 in mice with pathological hypertrophy was able to protect against cardiac dysfunction and attenuate adverse cardiac remodelling. The improved cardiac phenotype was associated with attenuation of heart and cardiomyocyte size, less cardiac fibrosis, lower expression of atrial and B-type natriuretic peptide genes, attenuation of profibrotic markers, and increased expression of p15 (a miR-154 target and cell cycle inhibitor). In summary, this study suggests that miR-154 may represent a novel target for the treatment of cardiac pathologies associated with cardiac fibrosis, hypertrophy and dysfunction.

treatment of hepatitis C virus 19 . Thus, there is promise for the successful development of miRNA-based therapies for the treatment of HF.
In a previous genome-wide transcriptome study we screened for miRNAs in the heart which were regulated by PI3K(p110α ), elevated in a diseased setting, and decreased in a protected setting 4 . We subsequently inhibited miRNA candidates (miR-34, miR-34a and miR-652) with LNA-antimiR-based drugs and demonstrated that this approach was associated with reduced pathology and improved cardiac function in mouse models of cardiac disease [15][16][17][18] . The goal of the current study was to target miR-154, a PI3K-regulated miRNA, which had been reported to have a profibrotic role in the human lung 20 . The role of miR-154 in the heart, and more specifically cardiac fibrosis, had not previously been examined.
Cardiac fibrosis leads to stiffness of the heart, and negatively impacts cardiac function leading to HF 21 . With limited effective therapies, treating fibrotic disorders represents a major unmet need 22,23 . In the current study, we assessed the therapeutic potential of inhibiting miR-154 in a mouse model with pre-existing pressure overload-induced pathological hypertrophy and cardiac dysfunction due to transverse aortic constriction (TAC). This model is associated with significant cardiomyocyte hypertrophy, cardiac fibrosis and elevated expression of HF molecular markers including atrial and B-type natriuretic peptides (ANP and BNP) 8,15,16 . We report here, that treatment with LNA-antimiR-154 in a setting of pressure overload was able to i) maintain cardiac function, ii) attenuate hypertrophy (increases in heart and cardiomyocyte size), iii) improve expression of cardiac pathology markers, and iv) prevent a significant increase in cardiac fibrosis.
Results miR-154 expression was increased in the diseased heart and depressed in the protected heart. Further analysis from our previous microarray screen 4 identified miR-154 expression to be i) upregulated in ventricular tissues from a mouse model of myocardial infarction (MI) and pathological hypertrophy (Fig. 1A); ii) downregulated in a mouse model of cardioprotection (due to cardiac specific over-expression of PI3K[p110α ], Fig. 1A); and iii) inversely correlated with cardiac function (Fig. 1B) 4 . In addition, analysis of miR-154 expression by qPCR in hearts of mice subjected to pressure overload (induced by TAC) for 12 weeks from a previous study 15 demonstrated that miR-154 is increased in a setting of pathological hypertrophy (Fig. 1C). Furthermore, miR-154 expression in hearts of mice tended to increase following 1 week of pressure overload and was significantly elevated at 4 weeks of pressure overload (Fig. 1D). Finally, by mining and analyzing publically available profiling datasets we found increased expression of miR-154 in hearts from HF patients with hypertrophic cardiomyopathy (compared to a non-failing cohort) (Fig. 1E).
LNA chemistry had no effect on organ weights in adult mice. First, we compared LNA-control treated mice to saline treated mice for any effects of the LNA chemistry. In adult male sham and unoperated, aged matched C57BL/6 mice, the LNA-control compound had no effect on body weight, tibial length, and organ weights (including the heart, atria, lung, kidney and liver) compared with the same dosing regimen of saline ( Supplementary Fig. S1). Thus, sham and unoperated mice administered saline or LNA-control have been combined.
Inhibition of miR-154 attenuated pathological cardiac remodelling and lung congestion. Adult mice were subjected to a sham or TAC operation. TAC induced a chronic pressure load on the heart and is associated with progressive pathological hypertrophy and cardiac dysfunction. Following 4 weeks of TAC, left ventricular (LV) remodelling and cardiac dysfunction was confirmed by echocardiography and mice were then subcutaneously administered LNA-control or LNA-antimiR-154 ( Supplementary Fig. S2). miR-154 expression increased in TAC LNA-control hearts compared to sham LNA-control hearts, and was inhibited in the hearts of sham and TAC LNA-antimiR-154 treated mice ( Fig. 2A). This supports our previous observation of increased expression of miR-154 in the heart following a pathological insult (Fig. 1A,C).
Pathological cardiac hypertrophy together with depressed cardiac function is typically associated with an increase in heart and atrial size, and lung congestion. TAC LNA-control mice displayed an increase in heart weight to tibial length ratio of approximately 47% versus sham mice (Fig. 2B,C; Supplementary Table S1) compared to only a 29% increase in TAC LNA-antimiR-154 treated mice (Fig. 2B,C; Supplementary Table S1). TAC LNA-control mice also displayed an increase in atrial weight to tibial length ratio (AW/TL, ~80% increase, Fig. 2C) and significant lung congestion (lung weight to tibial length ratio [LW/TL] increased ~55%, Fig. 2C, Supplementary Table S1) compared to sham control mice. In contrast, these parameters were not elevated in TAC LNA-antimiR-154 treated mice (Fig. 2C, Supplementary Table S1). Differences in heart size were accompanied by similar differences in cardiomyocyte size (Fig. 2D). Compared with sham mice, myocyte size increased by approximately 93% in TAC LNA-control hearts but only 60% in TAC LNA-antimiR-154 treated hearts (Fig. 2D). TAC-induced hypertrophy is typically associated with an increase in ANP, BNP, and β -myosin heavy chain (β -MHC). ANP and BNP expression was significantly increased in TAC LNA-control hearts but not in TAC LNA antimiR-154 treated hearts (Fig. 2E). β -MHC expression was also elevated in TAC LNA-control hearts and tended to be lower in TAC LNA-antimiR-154 treated hearts (Fig. 2E).
LNA-antimiR-154 treatment was associated with preserved cardiac function. Following 4 weeks of pressure overload, TAC mice displayed increased LV wall thickness, increased LV mass and depressed cardiac function (fractional shortening, FS, reduced by 20-25% compared with pre-surgery values and sham mice, Fig. 3A Fibrotic and autophagic pathways were enriched for predicted miR-154 targets. To examine additional mechanisms via which LNA-antimiR-154 might protect the stressed heart we interrogated predicted gene targets of miR-154 using TargetScan Mouse 6.2, and performed pathway enrichment analysis (Partek ® Genomics Suite v6.5). Consistent with miR-154 playing a key role in regulating fibrosis, the transforming growth factor beta (TGFβ ) and Wnt signaling pathways were significantly enriched ( Fig. 4, P = 0.02 and P = 0.04, respectively) 20,24,25 . In addition, the "Regulation of autophagy" pathway was enriched (Fig. 4, P < 0.02).  miR-154 does not regulate autophagy in the heart. Since the "Regulation of autophagy" pathway was enriched and autophagy-related protein 7 (Atg7) is a predicted target of miR-154, we assessed the potential contribution of autophgy. Autophagy is an essential cellular process that degrades and recycles senescent or damaged proteins and organelles, and is required for normal homeostasis of cardiomyocytes 26 . In the present study, whilst mRNA expression of Atg7 was increased in the hearts of LNA-antimiR-154 treated TAC mice compared with LNA-control-treated TAC mice (Fig. 5A), this was not accompanied by an increase in ATG7 protein expression (Fig. 5B). In addition, protein expression of another autophagic marker, LC3-II was not significantly different between TAC LNA-control and TAC LNA-antimiR-154 treated hearts (Fig. 5C). Thus, it is unlikely that miR-154 regulates autophagy and contributes to the cardiac phenotype, at least at the time point examined. 20 and pathway enrichment analysis ( Fig. 4) implicated miR-154 in regulating fibrosis. Fibrosis is a common feature of the failing myocardium and this is typically associated with increased levels of profibrotic factors and extracellular matrix proteins 27 . On histological analysis, hearts of TAC LNA-control mice displayed increased deposition of fibrosis ( Fig. 6A) and this was accompanied by increased expression of collagen 3 (Col3α1), collagen 1 (Col1α1) (Fig. 6B), and latent MMP2 by gelatin zymography (Fig. 6C). These features were all attenuated by treatment with LNA-antimiR-154 ( Fig. 6A-C).

LNA-antimiR-154 treatment was associated with less fibrosis. Previous studies in the lung
We next investigated miR-154 targets which had been implicated in being involved with liver fibrosis including Dickkopf-related protein 2 (Dkk2) 28 , Friend leukemia integration-1 (Fli1) that has been implicated in extracellular matrix (ECM) gene regulation in skin fibroblasts 29 , a novel predicted target involved in calcium transporting, ATPase, Ca+ + Transporting, Plasma Membrane 1 (Atp2b1), and cyclin-dependent kinase inhibitor 2B (Cdkn2b), also known as p15, a cell cycle inhibitor shown to be decreased with lung fibrosis 20 . The gene expression of Dkk2, Fli1 and Atp2b1 did not change with LNA-antimiR-154 treatment (Fig. 7A). Protein expression of p15 increased with LNA-antimiR-154 treatment in TAC hearts compared to TAC LNA-control hearts (Fig. 7B). This suggests a potential mechanism by which miR-154 may regulate fibrosis in the heart.
Chronic inhibition of miR-154 was not associated with morphological disarray. Consistent with the long half-life of LNA-antimiRs, inhibition of miR-154 was achieved in adult mouse organs including the heart, kidney, liver and lung, 8 weeks after LNA-oligonucleotide administration ( Supplementary Fig. S3A). Chronic inhibition of miR-154 had no effect on liver organ weight in sham and TAC mice administered LNA-antimiR-154 compared to LNA-control mice ( Supplementary Fig. S3B), and the heart and liver displayed no evidence of morphological disarray by histological assessment (Supplementary Fig. S3C). Due to the reported role of miR-154 in a setting of lung disease 20 , and the known crosstalk between the lungs and heart in disease settings 30 , we measured miR-154 in lungs of our disease model. Expression of miR-154 was not elevated in the lungs of TAC LNA-control mice compared with sham mice (Supplementary Fig. S4).

Discussion
Numerous studies have highlighted the key role of miRNAs contributing to processes leading to HF, including cardiac hypertrophy, apoptosis and fibrosis 12,14,23,31 . Inhibition of pathogenic miRNAs with LNA-antimiRs has had a beneficial effect on cardiac function and pathology in animal models of HF [15][16][17][18] . We previously identified miRNAs that were differentially regulated in settings of cardiac stress and protection 4 , and demonstrated this as a valuable approach in identifying and targeting candidate miRNAs as a therapy for HF [15][16][17] . In the current study, we examined the effect of silencing miR-154 in a TAC mouse model which is characterized by hypertrophy, fibrosis and cardiac dysfunction. miR-154 was an appealing candidate as we previously reported that miR-154 expression was decreased in hearts of mice with increased PI3K activity, and increased in the diseased heart due to MI 4 . Furthermore, other investigators reported that the expression of miR-154 was elevated in two mouse models with established pathological hypertrophy due to TAC or cardiac-specific transgenic over-expression of activated calcineurin A 32 . Finally, it had previously been shown that miR-154 expression was increased in lungs of patients with pulmonary fibrosis, and a miR-154 inhibitor was able to attenuate TGFβ induced proliferation of human lung fibroblasts 20 . However, the role of miR-154 had not previously been examined in an in vivo model. This study represents the first to directly assess the role of miR-154 in a setting of pathological hypertrophy associated with fibrosis. Fibrosis is a common feature of heart disease of various aetiologies. Thus, the identification and development of therapies with antifibrotic properties is of great interest 23,33,34 .
Pressure overload is associated with numerous cardiac adaptive and maladaptive changes and responses including cardiac myocyte hypertrophy, fibrosis, autophagy and inflammation 35 . Over time, these features lead to pathological heart enlargement and cardiac dysfunction which progresses to heart failure. In the present study, treatment with LNA-antimiR-154 was associated with favourable outcomes in TAC mice including attenuation (WGA) from sham and TAC LNA-control and LNA-antimiR-154 treated mice and quantification of cell area. Scale bar = 100 μM. Data are expressed as mean ± SEM. Upper graph: N = 3 (sham groups) 7-8 (TAC groups). *P < 0.05 vs. sham of the same treatment group; ***P < 0.001 vs. sham LNA-control. Lower graph: Cardiomyocyte size expressed as a percent increase compared to combined sham groups (as sham LNA-control and sham LNA-antimiR-154 were not different). ***P < 0.001 vs. sham; **P < 0.01 vs. sham; † P < 0.05. (E) qPCR analysis of ANP (Nppa), BNP(Nppb) and β-MHC (Myh7) standardized to Hprt1 in sham and TAC LNA-control and LNA-antimiR-154 treated mice. Data are expressed as mean ± SEM. N = 3-8 per group. For ANP and BNP: *P≤ 0.05 vs. sham of the same treatment group (One way ANOVA followed by Fisher's post-hoc test). For β -MHC: *P < 0.05 vs. sham of same treatment group (Mann Whitney nonparametric test for LNA-control sham vs. TAC. Unpaired t-test for LNA-antimiR-154 sham vs. TAC).
Scientific RepoRts | 6:22442 | DOI: 10.1038/srep22442 of heart size, lower atrial and lung weights, and preserved cardiac function. To explore potential mechanisms by which miR-154 may mediate protection in a setting of pathological hypertrophy, we investigated the effect of LNA-antimiR-154 on cardiomyocyte size, cardiac fibrosis, and autophagy. Here, we show that LNA-antimiR-154 mediated protection was associated with anti-hypertrophic actions and anti-fibrotic properties. Cardiac hypertrophy was examined by calculating LV mass during the time course of the study by echocardiography, as well as measuring heart size, cardiomyocyte size and hypertrophic genes (ANP and BNP) at study end. Collectively,    the attenuation in cardiac enlargement, smaller myocyte size, and the absence of significant increases in ANP and BNP demonstrate that LNA-antimiR-154 treatment attenuated TAC-induced hypertrophy. Based on histological assessment of collagen deposition, MMP2 abundance and gene expression of extracellular markers, LNA-antimiR-154 treatment also attenuated TAC-induced fibrosis. To determine how silencing miR-154 could mediate antifibrotic effects in the heart, we investigated the expression of a number of miRNA gene targets previously linked with pulmonary or liver fibrosis 20,28 . Here we show that the protein expression of the cell cycle inhibitor, p15 (which is decreased in a setting of lung fibrosis) was increased with LNA-antimiR-154 treatment in TAC hearts compared to TAC LNA-control hearts. Thus, the regulation of p15 by miR-154 represents one mechanism by which LNA-antimiR-154 is likely to mediate anti-fibrotic effects in the heart.
We also investigated the possibility that miR-154 may mediate protection by regulating autophagy because Atg7, a predicted target of miR-154, is a key autophagy protein that had been shown to protect the stressed heart 36 . Hearts from transgenic mice with increased Atg7 expression had enhanced autophagy without heart pathology and cardiac dysfunction 36 . Furthermore, when Atg7 transgenic mice were bred with a mouse model with reduced autophagy and pathological remodelling, sustained Atg7-induced autophagy in the heart decreased fibrosis, attenuated pathological cardiac remodelling and improved survival 36 . In the current study, despite increased Atg7 gene expression with LNA-antimiR-154 treatment, this was not accompanied by increased Atg7 protein expression. In addition, LC3 protein expression was not different between LNA-control and LNA-antimiR-154 treated hearts. Thus, it would appear autophagy has not contributed to the improved cardiac outcome in LNA-antimiR-154 TAC mice.
miRNAs are often associated in clusters or families, and miR-154 is highly conserved between species (Supplementary Fig. S5A). Several studies have shown a role for clustered miRNAs in disease 18,37 , and the therapeutic potential of targeting miRNA clusters and families in cancer and cardiovascular disease 16,18,38 . The therapeutic capacity of LNA-antimiR-154 may be increased by targeting other miRNAs that are mapped to the human chromosome 14q32 cluster, where miR-154 is located 20,39 . Expression of miR-495 and miR-299-5p, both localized to the chromosome 14q32 cluster, were also upregulated in a setting of MI and inversely correlated to cardiac function ( Supplementary Fig. S5B,C) 4 . Thus, further improvements in cardiac function and pathology may be observed if multiple members of the 14q32 cluster are inhibited. This could be explored in future studies.
Finally, since i) miR-154 plays a role in lung disease 20 , ii) there was a moderate increase in lung weight in TAC LNA-control mice (Fig. 2C), and iii) organ crosstalk between the heart and lungs exists in disease settings 30 , it was considered possible that LNA-antimiR-154 therapy may have indirectly affected cardiac function via an effect in the lungs. To examine this possibility, we assessed miR-154 expression in the lungs of sham and TAC LNA-control mice. As miR-154 was not elevated in the lungs of our TAC model, we consider it unlikely that a decrease in miR-154 in the lungs of TAC LNA-antimiR-154 treated mice contributed to the cardiac phenotype. However, examining a potential lung-heart interaction would be of interest in more severe cardiac disease models in which miR-154 is elevated in the lungs.
In conclusion, we show that in vivo subcutaneous delivery of LNA-antimiR-154 offers a promising therapeutic approach in a model of pathological hypertrophy associated with cardiac fibrosis. This study is the first to show that inhibition of miR-154 is able to attenuate cardiac hypertrophy, maintain cardiac function and protect against fibrosis. Mechanisms via which silencing miR-154 has antifibrotic properties and protects against cardiac dysfunction are highlighted in Fig. 8. Fibrosis is a pathological hallmark of chronic organ failure, including disorders of the heart, kidney, liver and lung, and chronic autoimmune diseases (e.g. scleroderma) 33 . Fibrotic disorders are associated with significant morbidity and mortality worldwide and there are limited effective therapies 22 . Since fibrosis is a common feature of numerous cardiac conditions which progress to HF including MI, hypertrophic cardiomyopathy and dilated cardiomyopathy, as well as being a predominant feature in a number of diseases in other organs 22,33,40 , inhibition of miR-154 using LNA therapeutics may aid in the treatment of numerous diseases associated with organ fibrosis. Pressure overload. Adult (~12 week old) male C57BL/6 mice were subjected to TAC (n = 15) or a sham (n = 6) operation as previously described 8 . The TAC model induces a chronic pressure load on the heart and is associated with progressive pathological hypertrophy and cardiac dysfunction within four weeks of surgery 8 . Briefly, prior to surgery, mice were anaesthetized with a combination of ketamine, xylazine and atropine (100:20:1.2 mg/kg, i.p.), administered an analgesic (carprofen, 5 mg/kg, s.c.) and intubated for ventilation. Mice were then administered a local anaesthetic (lignocaine, 7 mg/kg, s.c.) at the site of incision. A sternectomy was performed to access the aorta and a non-absorbable 5-0 braided silk suture was tied around the aorta between the right innominate and left carotid arteries, causing a constriction of approximately 65% using a 0.46 mm probe as a guide. For sham operations to serve as controls, mice (sham, n = 6) received the same surgical procedure except that the suture around the aorta was removed prior to closing the chest cavity.
In vivo delivery of LNA-antimiR oligonucleotides. Following 4 weeks of pressure overload, all mice underwent echocardiography to confirm the degree of LV remodelling and cardiac dysfunction. Sham and TAC mice were then randomized into control or treated groups. The TAC groups had a similar degree of LV hypertrophy based on LV wall thickness and LV mass after 4 weeks of pressure overload, prior to commencement of treatment. Mice were subcutaneously administered LNA-control or LNA-antimiR-154 (25 mg/kg/day) over three consecutive days and left for a period of 8 weeks (Supplementary Fig. S2). Our previous studies demonstrated that 3 consecutive daily subcutaneous injections of an LNA-antimiR was sufficient to inhibit microRNA gene expression for at least 2 months [15][16][17] .
An additional subset of unoperated, aged matched male mice (n = 13) were included within the study. Since the sham operation had no effect on morphological, functional or biochemical parameters, unoperated and sham mice were combined and randomized into saline/LNA-control or LNA-antimiR-154 treated groups. For clarification, the "sham LNA-control group" includes 3 sham LNA-control mice and 4 unoperated saline mice, and the "sham LNA-antimiR-154 group" includes 3 sham LNA-antimiR-154 treated mice and 9 unoperated LNA-antimiR-154 treated mice.

RNA isolation.
Total RNA was isolated from frozen mouse tissues using TRI Reagent (Sigma-Aldrich, St Louis, MO, USA) and quantitated on a Nanodrop ™ Spectrometer (Thermo Scientific, Waltham, MA, USA).

Quantitative RT-PCR (qPCR).
For mRNA expression analysis, 2 μg of total RNA was reverse transcribed using the High Capacity RNA-to-cDNA kit (Life Technologies, Carlsbad, CA, USA) according to manufacturer's recommendations. qPCR was performed using TaqMan ® probes (Life Technologies) and amplified on an Applied Biosystems 7500 real-time PCR instrument according to manufacturer's instructions. Hypoxanthine phosphoribosyltransferase 1 (Hprt1) was used to standardize for cDNA concentration and data was analysed using the 2 −∆∆Ct method of quantification. For miRNA expression analysis, 50-100 ng of total RNA was reversed transcribed using TaqMan ® MicroRNA Reverse Transcription Kit (Life Technologies) according to manufacturer's recommendations. qPCR was performed using TaqMan ® MicroRNA Assays (Life Technologies), expression was normalized against snoU6 and data analysed using the 2 −∆∆Ct algorithm.
Histological analyses. Tissue samples were fixed in 4% paraformaldehyde overnight and paraffin embedded for histological analysis at 6 μm cross-sections. To assess tissue morphology, sections were stained with haematoxylin and eosin (H&E). Cardiac collagen deposition/interstitial fibrosis (Masson's trichrome stain) and cell area (wheat germ agglutinin stain, WGA) of ventricles were assessed as previously described 17 . Statistical Analyses. Statistical analyses were performed using StatView (Version 5.0.1, SAS Institute Inc., Cary, NC, USA). Results are presented as means ± SEM. For normally distributed data, differences between groups were identified using one-way analysis of variance (ANOVA) followed by Fisher's post-hoc test. Unpaired t-tests were used when comparing two groups for a single measure. If data were not normally distributed, differences between groups were identified using a Nonparametric test (Mann-Whitney test). For echocardiography parameters, differences between groups were identified using a two-way repeated measures ANOVA followed by Fisher's post-hoc test. A value of P < 0.05 was considered significant. All relative units are expressed as a fold change with the relevant control group normalised to 1.