Cardiac development begins when mesodermal progenitor

Molecular Medicine MicroRNA-22 Regulates Cardiac Hypertrophy and Remodeling in Response to Stress Zhan-Peng Huang,* Jinghai Chen,* Hee Young Seok, Zheng Zhang, Masaharu Kataoka, Xiaoyun Hu, Da-Zhi Wang Rationale: The adult heart is composed primarily of terminally differentiated, mature cardiomyocytes that express signature genes related to contraction. In response to mechanical or pathological stress, the heart undergoes hypertrophic growth, a process defined as an increase in cardiomyocyte cell size without an increase in cell number. However, the molecular mechanism of cardiac hypertrophy is not fully understood. Objective: To identify and characterize micrornas that regulate cardiac hypertrophy and remodeling. Methods and Results: Screening for muscle-expressed micrornas that are dynamically regulated during muscle differentiation and hypertrophy identified microrna-22 (mir-22) as a cardiac- and skeletal muscle enriched microrna that is upregulated during myocyte differentiation and cardiomyocyte hypertrophy. Overexpression of mir-22 was sufficient to induce cardiomyocyte hypertrophy. We generated mouse models with global and cardiacspecific mir-22 deletion, and we found that cardiac mir-22 was essential for hypertrophic cardiac growth in response to stress. mir-22 null hearts blunted cardiac hypertrophy and cardiac remodeling in response to 2 independent stressors: isoproterenol infusion and an activated calcineurin transgene. Loss of mir-22 sensitized mice to the development of dilated cardiomyopathy under stress conditions. We identified Sirt1 and Hdac4 as mir-22 targets in the heart. Conclusions: Our studies uncover mir-22 as a critical regulator of cardiomyocyte hypertrophy and cardiac remodeling. (Circ Res. 2013;112:1234-1243.) Key Words: calcineurin cardiac hypertrophy dilated cardiomyopathy heart isoproterenol microrna-22 posttranscriptional regulation Cardiac development begins when mesodermal progenitor cells adopt cardiac-specific fates and express transcription factors and signaling molecules that control their specification and differentiation. 1,2 During embryonic development, cardiomyocytes continue to proliferate while differentiating. However, cardiomyocyte proliferation dramatically decreases and eventually stops in postnatal hearts. One of the major responses of adult hearts and cardiomyocytes to biomechanical stress and pathological stimuli is to undergo hypertrophic growth, anatomically defined as an increase in the size of cardiomyocytes without an increase in cell number. 3 Cardiac hypertrophy is initially an adaptive response to maintain cardiac output. However, prolonged hypertrophic growth is associated with adverse consequences that may lead to heart failure and sudden death. 4,5 Cardiac hypertrophy is also accompanied by reactivation of a set of cardiac fetal genes, including those that encode atrial natriuretic peptide, brain natriuretic peptide (BNP), β myosin heavy chain (β-mhc), and others, suggesting that molecular events controlling heart development are redeployed to regulate hypertrophic growth. 3,6 MicroRNAs (mirnas) are a class of small noncoding RNAs that modulate gene expression at the posttranscriptional level. The discovery of gene regulation by mirnas added an entirely new layer of complexity to our understanding of how gene expression is regulated. Many mirnas are expressed in developing and adult hearts. Genetic studies demonstrated that mirnas are essential for normal development and function of cardiac and skeletal muscles. 7 13 Further studies uncovered mirna expression profiles in hypertrophic or failing hearts and revealed a collection of mirnas that are dysregulated under those pathological conditions. 10 Furthermore, functional analyses using both gain-of-function and loss-of-function approaches in mice have begun to establish the roles of mirnas in cardiac hypertrophy. 1,10 Original received December 4, 2012; revision received March 15, 2013; accepted March 21, 2013. In February 2013, the average time from submission to first decision for all original research papers submitted to Circulation Research was 11.98 days. From the Department of Cardiology, Boston Children s Hospital, Harvard Medical School, Boston, MA (Z.-P.H., J.C., H.Y.S., Z.Z., M.K., X.H., D.-Z.W.); Department of Cell Biology and Genetics, Chongqing Medical University, Chongqing, China (Z.Z.); and Harvard Stem Cell Institute, Harvard University Cambridge, Cambridge, MA (D.-Z.W.). *These authors contributed equally. The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/circresaha. 112.300682/-/DC1. Correspondence to Da-Zhi Wang, PhD, Department of Cardiology, Boston Children s Hospital, Harvard Medical School, 320 Longwood Ave, Boston, MA 02115. E-mail dwang@enders.tch.harvard.edu 2013 American Heart Association, Inc. Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/CIRCRESAHA.112.300682 1234

Huang et al mir-22 in Cardiac Hypertrophy and Remodeling 1235 Nonstandard Abbreviations and Acronyms ACTN1 BNP CnA ISO KO LV MHC PE TUNEL UTR α-actinin brain natriuretic peptide calcineurin isoproterenol knockout left ventricular myosin heavy chain phenylephrine terminal deoxynucleotidyl transferase dutp nick-end labeling untranslated regions mirna-22 (mir-22) was previously reported as a tumorsuppressive mirna that induces cellular senescence in cancer cell lines. 14,15 mir-22 was shown to repress hypertrophy in cultured cardiomyocytes. 16 Most recently, it was found that mir-22 was required for the heart to adapt to pressure overload induced cardiac hypertrophy. 17 Previously, we investigated the expression and function of mirnas in skeletal muscle differentiation and cardiac hypertrophy. 11,18 We found that the expression of mir-22 was significantly induced during myoblast differentiation. 11 Here, we report that the expression of mir-22 is highly enriched in cardiac and skeletal muscles and is dynamically regulated during cardiac hypertrophy. We show that mir-22 is sufficient to induce cardiomyocyte hypertrophy and that mir-22 is a key regulator of stress-induced cardiac hypertrophy and remodeling. Methods Cell culture, luciferase reporter assays, quantitative reverse transcriptase polymerase chain reaction, Northern blot analyses, Western blot analyses, and immunochemistry were performed according to routine protocols. Generation of mir-22 conventional and cardiac-specific knockout (KO) mice, isoproterenol (ISO) administration, and measurement of cardiac function by echocardiography are described in the Online Materials and Methods. Statistics Values are reported as mean±sd unless indicated otherwise. The 2-tailed Mann-Whitney U test was used for comparing 2 means (Prism, GraphPad). Values of P<0.05 were considered statistically significant. Results mir-22 Is Highly Expressed in Cardiac and Skeletal Muscles and Is Induced in Hypertrophic Cardiomyocytes In a previous study, we used an mirna microarray approach to identify mirnas whose expression increased during skeletal muscle myoblast differentiation. 11 We reported that mir-1 and mir-133 are selectively expressed in cardiac and skeletal muscles and participate in the regulation of myoblast proliferation and differentiation. 11 Among many other mirnas identified from this screen, mir-22 was also found to be significantly induced during myoblast differentiation. 11 We examined the distribution of mir-22 expression in mouse tissue. Northern blot analyses using total RNAs isolated from multiple adult mouse tissues demonstrated that mir-22 is expressed in all tissues tested, with its highest expression detected in cardiac and skeletal muscles (Figure 1A). Next, we determined the expression of mir-22 in the hearts of fetal, postnatal, and adult mice. Northern blot analyses showed that mir-22 expression was increased in the postnatal and adult hearts, with the highest mir-22 expression detected in the hearts of 6-monthold mice (Figure 1B). To specify the cell type distribution of mir-22 in the heart, we examined the expression of mir-22 in isolated neonatal rat cardiomyocytes and nonmyocytes. As shown in Figure 1C, Northern blot demonstrated that mir-22 expression is enriched in neonatal cardiomyocytes. We separated cardiomyocytes and noncardiomyocytes from adult mouse hearts and performed quantitative polymerase chain reaction analyses to detect the expression of mir-22, together with molecular markers for cardiomyoyctes (cardiac troponin T), endothelial cells (platelet endothelial cell adhesion molecules and Flk-1), and fibroblasts (periostin, POSTN). We found that mir-22 is expressed predominantly in cardiomyocytes of adult hearts (Figure 1D). Together, these data suggest that mir-22 may play an important role in the adult heart. Having demonstrated that mir-22 expression increased during heart development and cardiomyocyte differentiation, we next asked whether its expression is also altered in cardiomyocyte hypertrophy. We first examined the expression of mir-22 in neonatal rat cardiomyocytes treated with hypertrophic agonists, phenylephrine (PE), and fetal bovine serum. Both PE and PBS modestly upregulated the expression level of mir-22 in cardiomyocytes 1 or 3 days after treatment (Figure 1E). Consistent with this in vitro result, mir-22 expression level was elevated in vivo in a murine pressure overload model of cardiac hypertrophy (Figure 1F). mir-22 expression was quickly elevated in the hearts 1 week after pressure overload was induced by transverse aortic constriction. This mir-22 upregulation was transient, declining in hearts 2 weeks after transverse aortic constriction and returning to control levels by 4 weeks after transverse aortic constriction (Figure 1F). We confirmed that hypertrophic genes were induced in these hypertrophic hearts (Online Figure IA). Similarly, we found that mir-22 expression was slightly increased in hypertrophic hearts of calcineurin (CnA) transgenic mice (Online Figure IB). Together, these data demonstrate that mir-22 is enriched in cardiac and skeletal muscles and that its expression is induced during cardiac hypertrophy. mir-22 Induces Cardiomyocyte Hypertrophy In Vitro To test the hypothesis that mir-22 is directly involved in cardiomyocyte hypertrophy, we first used both gain-of-function and loss-of-function approaches in cultured cardiomyocytes. We transfected neonatal rat cardiomyocytes with mir-22 mimics or inhibitors as previously reported, 12,18 followed by further culture with or without PE. Indeed, mir-22 was sufficient to induce cardiomyocyte hypertrophy, evidenced by the increase in cell size of cardiomyocytes transfected with mir-22 mimics (Figure 1G and 1H). mir-22 mimic induced hypertrophy was further enhanced by PE (Figure 1G and Online Figure II). However, treating the cells with mir-22 inhibitors alone did not significantly affect the morphology or size of cardiomyocytes (Figure 1G and Online Figure II). However, antagonism of endogenous mir-22 by mir-22

1236 Circulation Research April 26, 2013 Figure 1. MicroRNA-22 (mir-22) induces cardiomyocyte hypertrophy. A, Northern blot analyses of mir-22 expression in different mice tissues. Total RNAs were isolated from indicated adult mice tissues and hybridized with an mir-22 probe. U6 small nuclear RNA (snrna) was used as a loading control. The intensity of the hybridization signal from Northern blot was quantified and shown as the relative expression level after normalized by U6 snrna (bottom). B, Northern blot analyses of mir-22 expression in developing and adult mouse hearts. Total RNAs were isolated from mouse hearts of indicated ages, and the expression level was measured by Northern blot. U6 snrna was used as a loading control. The intensity of the hybridization signal from Northern blot was quantified and shown as relative expression level after normalized by U6 snrna (bottom). C, Northern blot analyses of mir-22 expression in isolated neonatal rat cardiomyocytes and nonmyoyctes. U6 snrna was used as a loading control. D, Quantitative polymerase chain reaction (qpcr) analyses of the expression of mir-22, mir-1, cardiac troponin T (ctnt), fetal liver kinase-1 (Flk-1), platelet endothelial cell adhesion molecule (platelet/endothelial cell adhesion molecule 1, and periostin (POSTN) in isolated cardiomyocytes (CMs) and noncardiomyocytes (Non-CMs) from adult mouse hearts. Relative expression level was normalized by U6 snrna (n=3 in each group; *P<0.05; **P<0.01). E, Northern blot analyses of mir-22 expression in hypertrophic cardiomyocytes. Neonatal rat cardiomyocytes cultured in serum-free medium were treated with hypertrophic agonist phenylephrine (PE) and fetal bovine serum (FBS) for 1 day (1D) or 3 days (3D), and total RNAs were isolated for Northern blot. U6 snrna was used as a loading control. F, Quantitative qpcr analyses of mir-22 expression in hypertrophic mouse hearts induced by transverse aortic constriction (TAC) for 1, 2, and 4 weeks, and control hearts (ctrl; n of each group was indicated; *P<0.05). G, Primary neonatal rat cardiomyocytes were transfected with mir-22 mimics, mir-22 inhibitors, or control. Cells were treated with the hypertrophic agonist PE or without treatment in control. Two days later, cultures were fixed and stained with antibodies for α-actinin, and mean±sd cell areas of α actinin immunostained cardiomyocytes transfected with mir-22 mimics or inhibitors, with or without PE treatment as indicated, were quantified (n>100 cells each condition; **P<0.01). H, Relative expression of cardiac hypertrophic marker genes, measured by quantitative real-time PCR assays, in cardiomyocytes transfected with mir-22 mimics or inhibitors, with or without PE treatment as indicated (n=3 in each group; *P<0.05; **P<0.01). ANP indicates atrial natriuretic peptide; BNP, brain natriuretic peptide; MHC, myosin heavy chain; and SK-α-actin, skeletal muscle α-actin. inhibitors significantly repressed PE-induced hypertrophy (Figure 1G and Online Figure II). Cardiac hypertrophy is accompanied by the increased expression of fetal genes, which are normally expressed in fetal hearts and repressed in adult hearts. 3,5 We examined the expression of hypertrophy-induced fetal genes, including atrial natriuretic peptide, BNP, skeletal muscle α-actin, and β-mhc. We found that atrial natriuretic peptide, BNP, and skeletal muscle α-actin were all substantially upregulated by mir-22 in neonatal cardiomyocytes (Figure 1H). However, mir-22 did not induce β-mhc expression in this setting. More important, mir-22 inhibitors partially suppressed

Huang et al mir-22 in Cardiac Hypertrophy and Remodeling 1237 PE-induced upregulation of hypertrophic marker genes, skeletal muscle α-actin, and β-mhc (Figure 1H), consistent with the observation that inhibition of mir-22 reduced the size of PE-treated cardiomyocytes (Figure 1G; compare columns 4 and 6). Together, these data indicate that mir-22 is sufficient to induce cardiomyocyte hypertrophy. Furthermore, our results also suggest that mir-22 mediates agonist-induced hypertrophic growth. Generation and Characterization of mir-22 Mutant Mice To study the function of mir-22 in vivo, we generated mir-22 KO mice. mir-22 is expressed from a single locus on mouse chromosome 11 that encodes a noncoding transcript with 3 putative exons. We designed a gene targeting strategy in which the mir-22 containing exon (exon 2) is flanked by loxp sites (Online Figure IIIA). After recombination in embryonic stem cells, germline transmission, and removal of the neomycin resistance cassette (Online Figure IIIB), we obtained mir- 22 flox/+ mice. Using the germline deleting EIIaCre transgene, 19 we obtained the mir-22 null allele (mir-22 ) and confirmed global ablation of mir-22 expression (Figure 2A). mir-22 null mice generated from heterozygous intercrosses survived to adulthood at the expected mendelian frequency (Online Figure IIIC). The null mice were viable and fertile. We verified that no mir-22 expression was detected in tissues of mutant mice using sensitive quantitative real-time polymerase chain reaction assays. The gross morphology and the ventricle weight versus body weight ratio of mir-22 mutant mice did not differ significantly from wildtype littermate controls (Figure 2B and 2C). Histological examination and Sirius Red/Fast Green staining did not reveal abnormal cardiac morphology or fibrosis in mir-22 mutant mice (Figure 2D and 2E). Echocardiographic measures of left ventricular (LV) size and function did not differ significantly between mir-22 mutant mice and their littermate controls (Figure 2F). Together, these studies indicate that mir-22 is dispensable for normal mouse development and cardiac function. mir-22 Is Involved in Stress-Dependent Cardiac Hypertrophy and Remodeling Next, we tested whether mir-22 plays a role in stressdependent cardiac response and remodeling. A recent report showed that loss of mir-22 sensitized mice to transverse aortic constriction induced pressure overload. 17 We therefore tested the functional involvement of mir-22 in other types of stress. Chronic infusion of the β-agonist ISO causes ventricular hypertrophy and fibrosis, 20 and ISO-induced cardiac hypertrophy in mice has been used in a number of studies to model human cardiac disease. 20,21 We treated mir 22 null mice and their control littermates with ISO for 2 weeks and then analyzed cardiac function. Consistent with previous reports, 20,22 ISO induced cardiac hypertrophy in wild-type mice, evidenced by an increase in the ratio of ventricle weight to tibial length (Online Figure IV). ISO treatment of wild-type control mice caused dramatic LV wall thickening, as measured by increases in LV posterior wall and interventricular septal thicknesses. However, this hypertrophic response was significantly reduced in mir-22 mutant hearts (Figure 3A and 3B and Online Table I), indicating that mir-22 loss of function reduced ISO-induced cardiac hypertrophy. LV internal dimension and volume were substantially increased in mir-22 null hearts, but not those of controls, after ISO treatment (Figure 3A and 3B and Online Table I), indicating that loss of mir-22 sensitized mice to the development of dilated cardiomyopathy in the face of ISO stress. Functionally, mir-22 was essential for preservation of ventricular systolic function in the face of ISO stress because ventricular contraction, measured by fractional shortening, was dramatically decreased in mir-22 null compared with Figure 2. Generation and characterization of microrna-22 (mir-22) mutant mice. A, Northern blot analyses of mir-22 expression in adult mouse hearts harvested from wildtype (+/+), mir-22 heterozygote (+/ ), and null ( / ) mice. U6 snrna was used as a loading control. B, Gross morphology of adult wild-type (+/+) and mir-22 mutant ( / ) hearts. Bars, 1 mm. C, Ventricle weight (Vw) to body weight (Bw) ratios of 2-monthold wild-type (+/+) and mir-22 null mice ( / ; n of each genotype is indicated). D, Hematoxylin and eosin staining of transverse sections of 2-month-old wildtype (+/+) and mir-22 mutant ( / ) hearts. Bars, 800 μm. E, Sirius Red/Fast Green collagen staining of transverse sections of 2-month-old wild-type (+/+) and mir-22 mutant ( / ) hearts. F, Echocardiography analyses of 2-month-old wild-type (+/+) and mir-22 mutant ( / ) mice. Data represent the mean±sd from at least 3 independent experiments. FS indicates fractional shortening; LV, left ventricular; LVID, left ventricular internal diameter; LVPW, left ventricular posterior wall; and LV Vol;s, LV systolic volume.

1238 Circulation Research April 26, 2013 Figure 3. MicroRNA-22 (mir-22) is required for stress-induced cardiac hypertrophy and function. A, Echocardiography analyses of cardiac function of 3-month-old mir-22 knockout (KO) mice and their control littermates (Ctrl) after isoproterenol (ISO) treatment. Nontreated mice were used as controls (Sham; n of each genotype is indicated). B, Hematoxylin and eosin staining of transverse sections of hearts from 3-month-old mir-22 KO mice (KO) and their control littermates (Ctrl) after ISO treatment. Nontreated mice were used as controls (Sham). Bars, 800 μm. C, Sirius Red/Fast Green collagen staining of transverse sections of hearts from 3-monthold mir-22 KO mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham). Bars, 800 μm. D, Quantification of the stained fibrosis area (n of each genotype is indicated). E, Wheat germ agglutinin staining of transverse sections of hearts from 3-month-old mir-22 KO mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham). Bars, 40 μm. F, Quantification of the size of cardiomyocytes by measuring transverse cell area (n of each genotype is indicated). G, Relative expression of cardiac hypertrophic marker genes of hearts samples from 3-month-old mir-22 KO mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham). n=3 in each group. Data represent the mean±sd from at least 3 independent experiments. *P<0.05; **P<0.01. ANP indicates atrial natriuretic peptide; BNP, brain natriuretic peptide; d, diastolic; FS, fractional shortening; IVS, interventricular septal; LV, left ventricular; LVPW, left ventricular posterior wall; MHC, myosin heavy chain; and s, systolic. wild-type control hearts (Figure 3A and Online Table I). Together, these data indicate that mir-22 loss of function increased cardiac vulnerability to developing decompensated dilated cardiomyopathy in the face of cardiac stress. Histological analysis further confirmed the decrease in ventricular wall thickness and increase in ventricular chamber size in mir-22 null mice in response to ISO treatment, whereas ISO-induced hypertrophy was obvious in control hearts

Huang et al mir-22 in Cardiac Hypertrophy and Remodeling 1239 (Figure 3B). ISO treatment also induced the development of cardiac fibrosis, consistent with a previous report. 20 mir-22 loss of function exacerbated ISO-induced cardiac fibrosis (Figure 3C and 3D), suggesting that mir-22 modulates the development of fibrosis induced by cardiac stress. At the cellular level, ISO treatment significantly increased the size of cardiomyocytes (hypertrophy) in wild-type control hearts. However, loss of mir-22 substantially attenuated ISO-induced cardiomyocyte hypertrophy (Figure 3E and 3F). We examined the expression of hypertrophic markers and found that ISOinduced upregulation of β-mhc and atrial natriuretic peptide, but not BNP, was attenuated in the hearts of mir-22 mutant mice (Figure 3G). We asked whether loss of mir-22 affected the survival of cardiomyocytes after ISO treatment. We performed terminal deoxynucleotidyl transferase dutp nick-end labeling (TUNEL) assay to measure apoptosis and observed an increase of TUNEL signals in the heart of mir-22 KO mice compared with controls (Online Figure V). Quantitative analysis confirmed the increase in apoptosis in cardiomyocytes (Online Figure V). Collectively, these studies indicate that mir-22 participates in the regulation of stress-induced cardiomyocyte survival, cardiac hypertrophy, and remodeling. Cardiac-Specific Deletion of mir-22 Impairs Cardiac Response to Stress The above studies, using an mir-22 loss-of-function genetic model in the mouse, suggest that mir-22 is required for the heart to remodel in response to stress (ISO treatment). However, mir-22 is widely expressed in multiple cell and tissue types (Figure 1), so the global loss-of-function strategy could not elucidate the cellular compartment(s) in which mir- 22 acts to regulate cardiac hypertrophy. To overcome this limitation and more specifically to define mir-22 activity in the heart, we selectively inactivated mir-22 in cardiomyocytes using the conditional mir-22 flox allele and the cardiomyocytespecific α-mhc Cre transgene. We confirmed cardiomyocyte-specific mir-22 deletion in mir-22 flox/flox ;α-mhc Cre mice (hereafter referred to as mir-22 cko) (Figure 4A). Consistent with our observations in conventional KO mice, mir-22 cko mice survived to adulthood at the expected mendelian distribution (Online Figure VI). We did not detect morphological or functional defects in the mir-22 cko mice under baseline conditions. These data indicate that loss of mir-22 in the heart did not result in embryonic lethality. We hypothesized that cardiomyocyte mir-22 expression is required for cardiomyocyte hypertrophy and cardiac remodeling under stress conditions. Indeed, we found that cardiacspecific deletion of mir-22 suppressed ISO-induced cardiac hypertrophy (Figure 4B), evidenced by a reduction in ISOinduced increases in ventricle weight to tibial length ratio and ventricular and interventricular septal thickness (Figure 4C and 4D and Online Table II). Cardiac-specific expression of mir-22 was required for the heart to maintain its function under stress conditions. We found a significant decrease in fractional shortening in mir-22 cko mice after ISO treatment (Figure 4D and Online Table II). The hearts of mir-22 cko mice further progressed to dilated cardiomyopathy on ISO treatment (Online Table II). Histological examination and quantification demonstrated that ISO-mediated cardiomyocyte hypertrophy was inhibited in mir-22 cko hearts (Figure 4E and 4F and Online Figure VII). Furthermore, we observed accelerated cardiac fibrosis in mir-22 cko mice in response to ISO treatment compared with littermate controls (Online Figure VII). These findings suggest that mir-22 expressed within cardiomyocytes participates in the regulation of cardiac fibrosis during cardiac remodeling. Alternatively, mir-22 expressing fibroblasts could alter the fibrosis program in failing hearts of mir-22 cko mice. Finally, we examined molecular markers for cardiac hypertrophy and found that ISO-induced expression of atrial natriuretic factor and β-mhc was reduced in the hearts of mir-22 cko mice (Figure 4G). Together, these results demonstrate that mir-22 regulates cardiomyocyte hypertrophy and heart remodeling in response to stress in vivo, consistent with its role in inducing hypertrophy in cardiomyocytes in vitro. Loss of mir-22 Represses Calcineurin-Induced Cardiac Hypertrophy Cardiac-specific overexpression of CnA, a calcium-regulated phosphatase, in α-mhc CnA transgenic mice induced cardiac hypertrophy. 23 We observed that mir-22 expression was slightly induced in the hearts of α-mhc CnA transgenic mice (Online Figure IB). To test whether mir-22 is required for CnA-induced cardiac hypertrophy, we bred mir-22 null mice with α-mhc CnA transgenic mice. α-mhc CnA transgenic mice underwent dramatic cardiac hypertrophy, evidenced by a significant increase in heart weight to body weight ratio. Heart weight to body weight ratio was increased to a similar extent in 1-month-old CnA/miR-22 KO mice (Online Figure VIIIA). Interestingly, echocardiographic analysis of 2-monthold CnA/miR-22 KO mice indicated that loss of mir-22 reduced CnA-induced thickening of the LV free wall and septum (Online Figure VIIIB and Table III). However, cardiac function was not improved in these mice, probably because of the increase of fibrosis (see below). Histological examination confirmed the decrease in ventricular wall thickness in these CnA/miR-22 KO compound mice (Online Figure VIIIC and VIIID). Loss of mir-22 also aggravated cardiac fibrosis induced by the CnA transgene (Online Figure VIIIE VIIIH). Quantitative measurement confirmed that the increase in fibrosis was statistically significant (Online Figure VIII). The hypertrophic markers BNP and β-mhc were dramatically induced in CnA transgenic hearts, and this upregulation was attenuated by mir-22 loss of function (Online Figure VIIIJ). Our data indicate that mir-22 participates in calcineurin-dependent cardiac hypertrophy and fibrosis. Regulation of mir-22 Target Genes mirnas repress the expression of their target genes primarily by targeting their 3 untranslated regions (UTRs). Several mir-22 target genes have been reported previously, including Sirt1. 15 We screened putative mir-22 targets that are expressed in the heart and known to play a role in cardiac function, including peroxisome proliferator activated receptor-α, peroxisome proliferator activated receptor gamma coactivator 1, phosphatase and tensin homolog, serum response factor, sirtuin 1 (Sirt1), and histone deacetylase 4 (HDAC4). We built luciferase reporters containing the 3 UTRs of these target

1240 Circulation Research April 26, 2013 Figure 4. Cardiac-specific deletion of microrna-22 (mir-22) blunted cardiac hypertrophy in response to stress. A, Quantitative polymerase chain reaction (qpcr) analyses of the expression of mir-22 in heart and skeletal muscle tissues of mir-22flox/flox mice and mir22flox/flox; α-myosin heavy chain (MHC)-Cre mice. N of each genotype is indicated. B, Gross morphology of hearts from 3-month-old mir-22 cardiac-specific knockout (cko; mir-22flox/flox;α-mhc Cre) mice and their control littermates (mir-22flox/+; mir-22flox/flox and mir-22flox/+; α-mhc Cre) after isoproterenol (ISO) treatment. Nontreated mice were used as controls (Sham). Bars, 1.5 mm. C, Ventricle weight (Vw) to tibial length (Tl) ratios of 3-month-old mir-22 cko mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham; n of each genotype is indicated). D, Echocardiography analyses of cardiac function of 3-month-old mir-22 cko mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham; n of each genotype is indicated). E, Wheat germ agglutinin staining of transverse sections of hearts from 3-month-old mir-22 cko mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham). Bars, 40 μm. F, Quantification of the size of cardiomyocytes by measuring transverse cell area (n of each genotype is indicated). G, Relative expression of cardiac hypertrophic marker genes of hearts samples from 3-month-old mir-22 cko mice and their control littermates after ISO treatment. Nontreated mice were used as controls (Sham). n=3 in each group. Data represent the mean±sd from at least 3 independent experiments. *P<0.05; **P<0.01. ANP indicates atrial natriuretic peptide; BNP, brain natriuretic peptide; d, diastolic; FS, fractional shortening; IVS, interventricular septum; LVPW, left ventricular posterior wall; and s, systolic. genes and tested their repression by mir-22. mir-22 repressed the luciferase reporters containing 3 UTRs of peroxisome proliferator activated receptor-α, Sirt1, and HDAC4 (Online Figure IX and Figure 5). The 3 UTRs of Sirt1 and Hdac4 contain highly conserved, computationally predicted mir-22 target sites (Figure 5A). To confirm the requirement of these putative target sites, we mutated residues in the 3 UTRs predicted to bind to the mir-22 seed sequence. The mutant 3 UTRs were no longer repressed by mir-22 (Figure 5B), indicating that these sites mediate mir-22 regulation of Sirt1 and Hdac4 3 UTRs. Next, we investigated whether mir-22 represses the protein expression levels of its targets. Neonatal rat cardiomyocytes were transfected with either mir-22 mimics or inhibitors, together with control mimics or inhibitors, and the protein expression levels of SIRT1 and HDAC4 were determined with Western blot analyses. As shown in Figure 5C, both SIRT1 and HDAC4 proteins were reduced by mir-22 mimics. Conversely, the expression levels of these 2 proteins were increased when endogenous mir-22 was antagonized by mir-22 inhibitors (Figure 5C). These changes in the SIRT1 and HDAC4 protein levels were modest but statistically

Huang et al mir-22 in Cardiac Hypertrophy and Remodeling 1241 Figure 5. Repression of microrna-22 (mir-22) targets in cardiomyocytes. A, Alignment of mir-22 sequences and the 3 untranslated regions (UTRs) of Sirt1 and Hdac4 from different species. Seed sequences are highlighted. B, Luciferase reporters with wild-type or mutant Sirt1-UTR (Luc-Sirt1-Wt-UTR or Luc-Sirt1-Mut-UTR, respectively), as well as wild type and mutant HDAC4-UTR (LUC-HDAC4-WT-UTR or Luc-HDAC4-Mut-UTR, respectively), were cotransfected with mir-22 mimics or control, and luciferase activity was determined. A luciferase construct containing mir-22 antisense sequences (mir-22 sensor) was used as a positive control. Values are presented as relative luciferase activity ±SD relative to the luciferase activity of reporters cotransfected with control mimics (n=4 in each group). C, Neonatal rat cardiomyocytes were transfected with mir-22 mimics, inhibitors, or control. Two days later, the protein levels of mir-22 targets (SIRT1 and HDAC4) were determined by Western blot analyses using indicated antibodies. β-tubulin served as a loading control. The intensity of Western blot signal was quantified and is shown as relative protein expression after normalized by β-tubulin (right). D, Neonatal rat cardiomyocytes were transfected with mir-22 mimics, inhibitors, or control. Twenty-four hours later, the expression levels of Sirt1 and Hdac4 transcripts were documented by quantitative real-time polymerase chain reaction (PCR) assays. E, Total proteins were isolated from the hearts of 2-month-old wild-type (+/+) and mir-22 mutant ( / ) mice. The expression levels of SIRT1 and HDAC4 proteins were measured by Western blot with the indicated antibodies. β-tubulin served as a loading control. The intensity of the Western blot signal was quantified and is shown as relative protein expression after being normalized by β-tubulin (right). F, Total RNAs were isolated from the hearts of 2-month-old wild-type (+/+) and mir-22 mutant ( / ) mice. The expression levels of Sirt1 and Hdac4 transcripts were documented by quantitative real-time PCR assays. Data represent the mean±sd from at least 3 independent experiments. *P<0.05; **P<0.01. significant (Figure 5C). We measured the expression levels of Sirt1 and Hdac4 transcripts after mir-22 overexpression or inhibition, and we found that both Sirt1 and Hdac4 mrnas were repressed by mir-22 (Figure 5D). Next, we examined the expression of endogenous SIRT1 and HDAC4 in the heart of mir-22 KO mice. Indeed, the expression levels of SIRT1 and HDAC4 proteins were elevated in mir-22 mutant mice (Figure 5E). Furthermore, we found that Sirt1 and Hdac4 transcripts were increased in the hearts of mir-22 KO mice (Figure 5F). Similarly, we found that SIRT1 and HDAC4 proteins were increased in the heart of cardiac-specific mir-22 KO mice (Online Figure X). Finally, we examined SIRT1 and HDAC4 protein levels in other tissues of mir-22 global KO mice. Both SIRT1 and HDAC4

1242 Circulation Research April 26, 2013 were elevated in liver and skeletal muscle of mir-22 KO mice (Online Figure XI). Our data suggest that Sirt1 and HDAC4 are canonical targets of mir-22 in the heart and other organs. Discussion In this study, we focused on cardiac- and skeletal muscle enriched mir-22 and found that mir-22 is dispensable for heart development but plays a critical role in cardiomyocyte hypertrophy and cardiac remodeling in response to stress. A recent report on targeted deletion of mir-22 reached similar conclusions on the essential role of mir-22 in the stress response of the heart. 17 However, this previous study reported that about half of mir-22 mutant mice died embryonically, in contrast to our data that show normal survival and cardiac structure and function in mice with global or cardiac-specific mir-22 loss of function under baseline conditions. The reason for the discrepancy is unclear, although the genetic background may be an important factor. We demonstrated in our study that cardiomyocyte hypertrophy induced by ISO treatment or calcineurin transgenesis was attenuated by global or cardiac-specific mir-22 loss of function, strongly suggesting that mir-22 is an essential regulator of cardiac hypertrophy. Furthermore, genetic deletion of mir-22 from the heart resulted in accelerated progression to dilated cardiomyopathy under stress conditions, suggesting that mir-22 plays a key role in the regulation of the transition from hypertrophic to dilated cardiomyopathy in response to pathological stress. Consistent with our study, Gurha et al 17 found that mir-22 participates in the regulation of cardiac contractile function and remodeling in response to stress induced by cardiac pressure overload. In the future, it will be important to further define the role of mir-22 in each of the pathophysiological stress conditions in the heart and, most important, to determine whether mir-22 participates in the regulation of cardiac response to stresses in patients with cardiovascular disease. We found that HDAC4 was significantly repressed by mir- 22. HDAC4 was previously shown as a key cofactor of the myocyte enhancer factor-2 family of transcription factors to modulate the function of myocyte enhancer factor-2c in cardiac and skeletal muscles. 24,25 Interestingly, our previous work also identified HDAC4 as a direct target of mir-1, an mirna specifically expressed in cardiac and skeletal muscle. 11,13 It will be important in the future to functionally test whether HDAC4 contributes to mir-22 mediated cardiac hypertrophy. We noticed that the repression of Hdac4 and Sirt1 by mir-22 was modest, which is not totally surprising. Many mirnas have been shown to inhibit the expression of many of their predicted targets moderately. Given that many mirnas and their targets regulate each other s expression in positive and negative feedback loops, it is intriguing to speculate that the expression of mir-22 is modulated by cardiac transcriptional regulators in the heart under normal or pathological conditions. In their study, Gurha et al 17 found that the expression of many genes encoding contractile proteins was dysregulated in mir-22 mutant hearts. More specifically, they reported that mir-22 directly represses purine-rich element-binding protein B in the heart. 17 Determining how each of the mir- 22 targets, or a combination of some of them, mediates the function of mir-22 in the heart is a challenging task for future investigation. In addition to its role in the regulation of cardiomyocyte hypertrophy, mir-22 was previously indicated to be involved in the control of cell proliferation and tumorigenesis. mir-22 expression was downregulated in several human breast cancer cell lines and clinical samples. One of the mir-22 targets in breast cancer cells is estrogen receptor α. 26,27 mir-22 was shown to inhibit cancer progression, in part, by inducing cell senescence. Several genes have been identified as mir- 22 targets in this setting, including CDK6, Sirt1, and Sp1. 15 Interestingly, HDAC4 was reported as one of the mir-22 targets in cancer cell lines. 28 Those results, together with our studies reported here, indicate that mir-22 plays a critical role in the regulation of cellular proliferation, differentiation, and stress-induced hypertrophy. In the future, it will be important to determine the involvement of this mirna in human cardiovascular disorders and cancer. Acknowledgments We thank members of the Wang laboratory for advice and support. We thank Drs John Mably and William Pu for stimulating discussion and careful reading of the article. Sources of Funding Work in the Wang laboratory was supported by the March of Dimes Foundation and the National Institutes of Health. M. Kataoka is supported by Banyu Life Science Foundation International. Z.-P. Huang is a postdoctoral fellow, and D.Z. Wang is an Established Investigator of the American Heart Association. None. Disclosures References 1. Espinoza-Lewis RA, Wang DZ. MicroRNAs in heart development. Curr Top Dev Biol. 2012;100:279 317. 2. Olson EN. Gene regulatory networks in the evolution and development of the heart. Science. 2006;313:1922 1927. 3. Ahmad F, Seidman JG, Seidman CE. The genetic basis for cardiac remodeling. Annu Rev Genomics Hum Genet. 2005;6:185 216. 4. Berk BC, Fujiwara K, Lehoux S. ECM remodeling in hypertensive heart disease. J Clin Invest. 2007;117:568 575. 5. Hill JA, Olson EN. Cardiac plasticity. N Engl J Med. 2008;358:1370 1380. 6. Seidman JG, Seidman C. The genetic basis for cardiomyopathy: from mutation identification to mechanistic paradigms. Cell. 2001;104:557 567. 7. van Rooij E, Liu N, Olson EN. MicroRNAs flex their muscles. Trends Genet. 2008;24:159 166. 8. van Rooij E, Quiat D, Johnson BA, Sutherland LB, Qi X, Richardson JA, Kelm RJ Jr, Olson EN. A family of micrornas encoded by myosin genes governs myosin expression and muscle performance. Dev Cell. 2009;17:662 673. 9. van Rooij E, Sutherland LB, Qi X, Richardson JA, Hill J, Olson EN. Control of stress-dependent cardiac growth and gene expression by a microrna. Science. 2007;316:575 579. 10. Small EM, Olson EN. Pervasive roles of micrornas in cardiovascular biology. Nature. 2011;469:336 342. 11. Chen JF, Mandel EM, Thomson JM, Wu Q, Callis TE, Hammond SM, Conlon FL, Wang DZ. The role of microrna-1 and micror- NA-133 in skeletal muscle proliferation and differentiation. Nat Genet. 2006;38:228 233. 12. Callis TE, Pandya K, Seok HY, Tang RH, Tatsuguchi M, Huang ZP, Chen JF, Deng Z, Gunn B, Shumate J, Willis MS, Selzman CH, Wang DZ. MicroRNA-208a is a regulator of cardiac hypertrophy and conduction in mice. J Clin Invest. 2009;119:2772 2786. 13. Chen JF, Callis TE, Wang DZ. micrornas and muscle disorders. J Cell Sci. 2009;122:13 20.

Huang et al mir-22 in Cardiac Hypertrophy and Remodeling 1243 14. Xiong J, Du Q, Liang Z. Tumor-suppressive microrna-22 inhibits the transcription of E-box-containing c-myc target genes by silencing c- Myc binding protein. Oncogene. 2010;29:4980 4988. 15. Xu D, Takeshita F, Hino Y, Fukunaga S, Kudo Y, Tamaki A, Matsunaga J, Takahashi RU, Takata T, Shimamoto A, Ochiya T, Tahara H. mir-22 represses cancer progression by inducing cellular senescence. J Cell Biol. 2011;193:409 424. 16. Xu XD, Song XW, Li Q, Wang GK, Jing Q, Qin YW. Attenuation of microrna-22 derepressed PTEN to effectively protect rat cardiomyocytes from hypertrophy. J Cell Physiol. 2012;227:1391 1398. 17. Gurha P, Abreu-Goodger C, Wang T, et al. Targeted deletion of micror- NA-22 promotes stress-induced cardiac dilation and contractile dysfunction. Circulation. 2012;125:2751 2761. 18. Tatsuguchi M, Seok HY, Callis TE, Thomson JM, Chen JF, Newman M, Rojas M, Hammond SM, Wang DZ. Expression of micrornas is dynamically regulated during cardiomyocyte hypertrophy. J Mol Cell Cardiol. 2007;42:1137 1141. 19. Lakso M, Pichel JG, Gorman JR, Sauer B, Okamoto Y, Lee E, Alt FW, Westphal H. Efficient in vivo manipulation of mouse genomic sequences at the zygote stage. Proc Natl Acad Sci USA. 1996;93:5860 5865. 20. Goonasekera SA, Hammer K, Auger-Messier M, et al. Decreased cardiac L-type Ca² + channel activity induces hypertrophy and heart failure in mice. J Clin Invest. 2012;122:280 290. 21. Kook H, Lepore JJ, Gitler AD, Lu MM, Wing-Man Yung W, Mackay J, Zhou R, Ferrari V, Gruber P, Epstein JA. Cardiac hypertrophy and histone deacetylase-dependent transcriptional repression mediated by the atypical homeodomain protein Hop. J Clin Invest. 2003;112:863 871. 22. Kook H, Yung WW, Simpson RJ, Kee HJ, Shin S, Lowry JA, Loughlin FE, Yin Z, Epstein JA, Mackay JP. Analysis of the structure and function of the transcriptional coregulator HOP. Biochemistry. 2006;45:10584 10590. 23. Molkentin JD, Lu JR, Antos CL, Markham B, Richardson J, Robbins J, Grant SR, Olson EN. A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell. 1998;93:215 228. 24. McKinsey TA, Zhang CL, Lu J, Olson EN. Signal-dependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature. 2000;408:106 111. 25. Lu J, McKinsey TA, Zhang CL, Olson EN. Regulation of skeletal myogenesis by association of the MEF2 transcription factor with class II histone deacetylases. Mol Cell. 2000;6:233 244. 26. Pandey DP, Picard D. mir-22 inhibits estrogen signaling by directly targeting the estrogen receptor alpha mrna. Mol Cell Biol. 2009;29:3783 3790. 27. Xiong J, Yu D, Wei N, Fu H, Cai T, Huang Y, Wu C, Zheng X, Du Q, Lin D, Liang Z. An estrogen receptor alpha suppressor, microrna-22, is downregulated in estrogen receptor alpha-positive human breast cancer cell lines and clinical samples. FEBS J. 2010;277:1684 1694. 28. Zhang J, Yang Y, Yang T, Liu Y, Li A, Fu S, Wu M, Pan Z, Zhou W. microrna-22, downregulated in hepatocellular carcinoma and correlated with prognosis, suppresses cell proliferation and tumourigenicity. Br J Cancer. 2010;103:1215 1220. Novelty and Significance What is Known? Hypertrophic growth is one of the most common responses of the heart to pathophysiological stress. Pathological hypertrophy often progresses to dilated cardiomyopathy and heart failure. Although several transcription factors and signaling molecules that regulate cardiac hypertrophy have been well studied, the role of micrornas in cardiac hypertrophy is less clear. What New Information Does This Article Contribute? MicroRNA-22 (mir-22) is a key regulator of cardiac hypertrophy. The expression of mir-22 is induced in hypertrophic cardiomyocytes. Genetic deletion of mir-22 suppresses agonist-induced cardiac hypertrophy in mice. MicroRNAs are recently discovered small noncoding RNAs that have been implicated in a variety of biological processes. We hypothesized that micrornas are previously unrecognized key regulators of cardiac hypertrophy. We found that the expression of mir-22 is induced in hypertrophic cardiomyocytes. Using mouse models of cardiac hypertrophy, we show that deletion of the mir- 22 gene prevented the development of cardiac hypertrophy under stress conditions. Our results also show that mir-22 inhibited Sirt1 and HDAC4, 2 important epigenetic regulators essential for cardiac function. These findings suggest that mir-22 may be a potential therapeutic target in the prevention or treatment of cardiac hypertrophy and heart failure.