Author + information
- Received July 17, 2015
- Revision received February 8, 2016
- Accepted February 15, 2016
- Published online May 3, 2016.
- Kathleen Meyer, MSca,b,
- Bettina Hodwin, MSca,
- Deepak Ramanujam, PhDa,b,
- Stefan Engelhardt, MD, PhDa,b and
- Antonio Sarikas, MDa,b,∗ ()
- aInstitute of Pharmacology and Toxicology, Technische Universität München, Munich, Germany
- bDeutsches Zentrum für Herz-Kreislauf-Forschung (DZHK), partner site Munich Heart Alliance, Munich, Germany
- ↵∗Reprint requests and correspondence:
Dr. Antonio Sarikas, Institute of Pharmacology and Toxicology, Technische Universität München, Biedersteiner Strasse 29, 80802 Munich, Germany.
Background Fibrosis is a hallmark of many myocardial pathologies and contributes to distorted organ architecture and function. Recent studies have identified premature senescence as a regulatory mechanism of tissue fibrosis, but its relevance in the heart remains to be established.
Objectives This study investigated the role of premature senescence in myocardial fibrosis.
Methods Murine models of cardiac diseases and human heart biopsies were analyzed for characteristics of premature senescence and fibrosis. Loss-of-function and gain-of-function models of premature senescence were used to determine its pathophysiological role in myocardial fibrosis.
Results Senescence markers p21CIP1/WAF1, senescence-associated ß-galactosidase (SA-ß-gal), and p16INK4a were increased 2-, 8-, and 20-fold (n = 5 to 7; p < 0.01), respectively, in perivascular fibrotic areas after transverse aortic constriction compared with sham-treated control subjects. Similar results were observed with cardiomyocyte-specific β1-adrenoceptor transgenic mice and human heart biopsies. Senescent cells were positive for platelet-derived growth factor receptor-α, vimentin, and α-smooth muscle actin, specifying myofibroblasts as the predominant cell population undergoing premature senescence in the heart. Inactivation of the premature senescence program by genetic ablation of p53 and p16INK4a (Trp53-/- Cdkn2a-/- mice) resulted in aggravated fibrosis after transverse aortic constriction, when compared with wild-type control subjects (49 ± 4.9% vs. 33 ± 2.7%; p < 0.01), and was associated with impaired cardiac function. Conversely, cardiac-specific expression of CCN1 (CYR61), a potent inducer of premature senescence, by adeno-associated virus serotype 9 gene transfer, resulted in ∼50% reduction of perivascular fibrosis after transverse aortic constriction, when compared with mock- or dominant-negative CCN1-infected control subjects, and improved cardiac function.
Conclusions Our data establish premature senescence of myofibroblasts as an essential antifibrotic mechanism and potential therapeutic target in myocardial fibrosis.
- antifibrotic therapy
- cardiac fibroblasts
- extracellular matrix
- gene therapy
- transverse aortic constriction
Fibrosis is a hallmark of most myocardial pathologies with limited treatment options (1). Whereas extracellular matrix (ECM) production during wound healing provides structural integrity to damaged tissue due to the negligible self-regenerative capacity of the mammalian heart, excessive fibrosis imposes detrimental effects and may result in scarring and loss of heart function (1). The molecular mechanisms that regulate ECM homeostasis and fibrogenesis are complex and incompletely understood. A signaling network of growth factors and cytokines is thought to cooperate to induce activation of cardiac fibroblasts (CF) and regulate the expression of ECM components and matricellular proteins (e.g., matrix metalloproteinases) to modulate the fibrotic response (2). Moreover, it is increasingly appreciated that the cellular microenvironment and cross-talk between different cell populations contribute to regulation of fibrosis (3).
Recent studies have uncovered a critical role for premature cellular senescence in tissue remodeling (4). In contrast to replicative senescence, originally characterized in human fibroblasts undergoing telomere erosion in culture (5), premature senescence is an irreversible form of cell-cycle arrest that can be triggered by various cellular stresses, including deoxyribonucleic acid damage, oncogene activation, and oxidative stress (6). Seminal studies have established premature senescence as an essential tumor-suppressive mechanism by stalling proliferation of oncogene-harboring cells at risk of neoplastic transformation (7–9). Additionally, senescent cells are found in aged tissues (4,10), although their role in this context has not been fully investigated. Senescent cells remain viable and metabolically active, but they are unable to proliferate despite the presence of mitogens. In addition to cell-cycle arrest, senescent cells exhibit the up-regulation of p16INK4a, p21CIP1/WAF1, and senescence-associated ß-galactosidase (SA-ß-gal) that distinguishes them from most quiescent cells (11,12). Furthermore, senescent cells are characterized by the up-regulation of secreted proteins and microenvironment modulators that comprise the senescent-associated secretory phenotype or senescence messaging secretome (13,14).
More recently, senescent cells were shown to accumulate in liver and skin damage, dampening the fibrotic response through the expression of antifibrotic proteins (15,16). This suggests that senescent cells may arise to limit fibrosis during tissue repair, although this has not been examined in the heart.
In this study, we analyzed different murine models of cardiac diseases and human heart biopsies for characteristics of premature senescence and fibrosis. In addition, loss-of-function and gain-of-function models were used to determine an essential role of premature senescence in restraining cardiac fibrosis.
Transverse aortic constriction (TAC) was performed on 8-week-old male C57BL/6N mice as described previously (17). In sham surgery, only the chest was opened, but no ligation of the aorta was performed. Cardiac dimensions and function were analyzed by pulse-wave Doppler echocardiography before TAC/sham surgery and before the animals were euthanized. ß1-adrenoceptor transgenic mice (18) on Friend Virus B NIH (FVB/N) background were analyzed at the age of 2.5, 5, and 10 to 12 months. We also obtained p53 knockout mice (B6.129-Trp53tm1Brd N12) and CDKN2 knockout mice (B6.129-Cdkn2atm1Rdp/Nci); the latter carry a targeted deletion of exons 2 and 3 of the INK4a/ARF locus that eliminates both p16 (Ink4a) and p19 (Arf). Trp53-/- Cdkn2a-/- compound mutant mice were generated by crossbreeding Trp53+/- and Cdkn2a+/- mouse strains. For cardiotropic expression of exogenous CCN1, 3-week-old male wild-type (WT) mice received adeno-associated virus serotype 9 (AAV9)-CCN1 or AAV9 dominant-negative CCN1 mutant (AAV9-CCN1-DN) (1 × 1012 genome copies per mouse) or iodixanol by tail vein injection before TAC. All animal studies were performed in accordance with the relevant guidelines and regulations of the responsible authorities.
Data are shown as mean ± SEM. Statistical analysis was performed with Prism (version 6, GraphPad Software, Inc., La Jolla, California). Data distribution was assessed by a Shapiro-Wilk test for normality. Differences between 2 means were assessed by a 2-tailed paired or unpaired Student t test. Differences among multiple means were assessed by 1-way or 2-way analysis of variance followed by the Bonferroni correction as indicated. A p value of <0.05 was considered significant.
Accumulation of senescent cells in fibrotic heart tissue
To investigate the relationship between fibrosis and premature senescence in the heart, we analyzed 2 established animal models of cardiac fibrosis and human heart biopsies for markers of cellular senescence. In TAC, an experimental model for pressure overload-induced cardiac hypertrophy and fibrosis (17), fibrosis typically originates from areas surrounding the coronary arteries and spreads throughout the myocardium. Perivascular fibrosis was significantly increased in TAC hearts when compared with hearts of sham control subjects 2 and 6 weeks after surgery (∼1.7- and ∼1.3-fold, respectively) (Figure 1A, Online Figure 1). To detect senescent cells, heart sections were stained for proliferation marker KI-67 and a panel of senescence markers were used that included SA-ß-gal, p16INK4a, and p21CIP1/WAF1. SA-ß-gal is the most established assay for senescence detection, in vitro and in vivo (12,21), based on increased lysosomal content of senescent cells (22). Additionally, p16INK4a and p21CIP1/WAF1 expression is a hallmark of senescent cells (23), which prevent cell cycle progression by inhibiting cyclin-dependent kinases (24).
The percentage of perivascular KI-67–positive cells was increased 5 days after TAC when compared with sham (6.4 ± 1.1% vs. 1.5 ± 0.4%) and declined thereafter (Online Figure 2), indicating a proliferative impulse immediately after TAC-induced pressure overload. In contrast, TAC animals exhibited a constant increase in the expression of senescence marker p21CIP1/WAF1, SA-ß-gal, and p16INK4a (2-, 8-, and 20-fold, respectively), when compared with sham control subjects 6 weeks after TAC (n = 5 to 7; p < 0.01) (Figures 1B to 1D).
Similar results were obtained with the ß1-adrenoceptor (ADRB1) transgenic (TG) model. TGAdrb1 mice develop cardiac hypertrophy that progresses to dilated cardiomyopathy with loss of ventricular function, cardiomyocyte (CM) degeneration, and replacement fibrosis (18). Compared with WT control subjects, TGAdrb1 mice displayed marked fibrosis at 5 and 10 months of age (∼2.0- and ∼4.1-fold, respectively) (Online Figure 3). Consistent with the TAC model, we observed a significant accumulation of SA-ß-gal, p16INK4a, and p21CIP1/WAF1-positive cells in TGAdrb1 mice in the course of fibrogenesis compared with WT control subjects (Online Figure 3).
KI-67–positive cells were increased ∼1.7-fold in TGAdrb1 mice versus WT from 2.5 to 10 months of age (Online Figure 4), which is consistent with a continuous tissue remodeling process due to cardiomyocyte degeneration and replacement fibrosis.
Similar to the data obtained in animal models, expression of senescence markers SA-ß-gal (R2 = 0.70; p < 0.001; n = 25) and p16INK4a (R2 = 0.62; p < 0.0001; n = 22) correlated positively with fibrosis in heart biopsies of patients experiencing idiopathic cardiomyopathy (Online Figure 5).
Cardiac myofibroblasts and premature senescence
Transformation of fibroblasts to myofibroblasts, characterized by expression of α-smooth muscle actin and production of ECM components, is a key event in cardiac remodeling and fibrosis (1,25). Our observation that senescent cells accumulate within fibrotic areas (Figure 1, Online Figures 3 and 5) suggested that (myo)fibroblasts may constitute the main cell population undergoing senescence upon fibrosis induction. Indeed, the majority of senescent (p21CIP1/WAF1-positive) cells expressed the fibroblast markers platelet-derived growth factor receptor α (92.5 ± 1.8%) (Figures 2A and 2B) or vimentin (92.1 ± 1.5%) (Online Figure 6). In contrast, only 10.3 ± 1.9% and 5.9 ± 2.3% of senescent cells expressed the endothelial cell marker CD31 or cardiac myocyte marker troponin T, respectively (Figures 2A and 2B). Additionally, 65.2 ± 10.5% of p21CIP1/WAF1-positive cells expressed α-smooth muscle actin (Figures 2A and 2B), indicating that the majority of CF undergoing senescence are myofibroblasts. Similar results were obtained with tissue of TGAdrb1 mice (data not shown). To further corroborate these findings, SA-ß-gal (Glb1) transcript level was measured by quantitative real-time polymerase chain reaction in CM and CF isolated from TAC- or sham-treated mice (Figure 2C and Online Table 1). Basal Glb1 gene expression was 4.4-fold higher in CF compared with CM in sham-treated hearts. Furthermore, Glb1 expression significantly increased in CF (+125%; p < 0.05), but not in CM, after TAC (Figure 2C and Online Table 1), indicating that SA-ß-gal is mainly expressed in CF.
Enhanced fibrosis with dysfunctional senescence machinery
To evaluate the biological relevance of premature senescence in cardiac fibrogenesis, we first studied the cardiac phenotype of Trp53-/- and Cdkn2a-/- mice, respectively. In many cell types, p53 and p16INK4a pathways contribute to cellular senescence to a varying degree (26). Surprisingly, we did not detect a significant reduction of SA-ß-gal expression in hearts of Trp53-/- or Cdkn2a-/- mice after TAC compared with WT littermates (Online Figures 7A and 8A). We also observed no difference in the severity or progression of perivascular fibrosis (Online Figures 7B and 8B).
To determine the impact of disrupting both gene loci on the development of premature senescence and fibrosis in the heart, we generated Trp53-/- Cdkn2a-/- compound mutant mice and subjected them to TAC (Online Figure 9). Consistent with a critical role of p53 and p16INK4a for cellular senescence, heart sections of Trp53-/- Cdkn2a-/- mice contained a markedly reduced number of SA-ß-gal-positive CF after TAC compared with WT littermates (Figures 3A and 3B). Remarkably, Trp53-/- Cdkn2a-/- mice developed a more severe fibrotic phenotype after TAC than did Trp53+/+ Cdkn2a+/+ mice (49 ± 4.9% vs. 33 ± 2.7%; p < 0.01) (Figures 3A and 3B, Online Figure 10A) that was associated with impaired cardiac function as judged by ejection fraction, fractional shortening, and left ventricular volume (Figure 3C, Online Figure 10B). Of note, angiogenesis (based on the number of CD31-positive endothelial cells) was not altered in Trp53-/- Cdkn2a-/- mice after TAC when compared with WT control subjects (Online Figure 11). In summary, these data demonstrate that CF lacking both the Trp53 and Cdkn2a genes (and thus functional ARF/p53 and p16/retinoblastoma [Rb] pathways) fail to senesce.
Induction of premature senescence improves heart function
The earlier observations suggested that induction of the senescence program in CF may limit fibrosis progression, thus implicating premature senescence as a regulatory mechanism to restrain excess ECM production and tissue remodeling in the heart. To test this hypothesis, we expressed the matricellular protein CCN1 in the heart of sham or TAC-treated mice to trigger premature senescence of CF in vivo.
CCN1 (also known as CYR61) is a secretory protein that is expressed at sites of wound healing to induce fibroblast senescence through binding to integrin α6β1 and cell surface heparin sulfate proteoglycans and activation of p53 and p16INK4a pathways (27,28).
To establish the ability of CCN1 to trigger senescence in the heart, we first evaluated the effect of wild-type CCN1 (CCN1-WT) and CCN1-DN devoid of integrin α6β1 and heparin sulfate proteoglycans binding sites critical for senescence induction (16) on CF in vitro (Online Figures 12A and 13). Incubation of neonatal rat cardiofibroblasts with conditioned medium of neonatal rat cardiomyocytes infected with adenoviruses expressing CCN1 resulted in a robust induction of cellular senescence as evidenced by SA-ß-gal staining (+83%; p < 0.01) and Cdkn1a (p21CIP1/WAF1) gene expression (+175%; p < 0.05). In contrast, no effect was observed with CCN1-DN-containing medium (Online Figures 12B to 12D).
Next we assessed whether heart-specific overexpression of CCN1 can induce CF senescence in vivo using AAV9 gene transfer. AAV9 was chosen for its strong heart tropism (29,30). Mice were injected with AAV9-CCN1, AAV9-CCN1-DN, or mock treated, and subjected to TAC or sham surgery. After 2 weeks, echocardiography was performed and tissue samples were collected for further analysis (Figure 4A). Immunoblotting of heart lysates showed an ∼1.5-fold increase each of CCN1 and CCN1-DN compared with endogenous CCN1 of uninfected control hearts (Figure 4B).
In confirmation of our in vitro experiments with neonatal rat cardiofibroblasts, premature senescence was increased (as evidenced by SA-ß-gal staining) in CCN1-infected animals (+225%; p < 0.05; and +240%) compared with mock-treated or CCN1-DN-infected control subjects after TAC, respectively (Figure 4C), and predominantly affected (myo)fibroblasts (Online Figure 14). Importantly, mice with CCN1-triggered senescence displayed markedly reduced perivascular fibrosis compared with mock or CCN1-DN infected control subjects (11.9 ± 1.4% vs. 22.4± 4.0% and 22.1 ± 1.8%, respectively; both p < 0.01) (Figure 4C). Of note, a similar antifibrotic effect was observed for interstitial ventricular fibrosis (Online Figure 15A). Additionally, the attenuated fibrotic remodeling after TAC in the gain-of-function senescence animal model was accompanied by improved cardiac function (Figure 4D, Online Figure 15B).
In this study, we presented comprehensive evidence for an essential role of premature cellular senescence as a mechanism to restrain cardiac fibrosis (Central Illustration). We showed that senescent cells accumulate in fibrotic myocardial tissue and identified myofibroblasts as the predominant cardiac cell population undergoing senescence. Mechanistically, we demonstrated that the cellular senescence program of CF is dependent on both p53/p21 and p16/Rb pathways, and that genetic ablation of Trp53/Cdkn2a resulted in diminished senescence that is associated with aggravated fibrosis and functional impairment of the heart after TAC. Conversely, induction of the cellular senescence program in CF by cardiotropic expression of the matricellular protein CCN1 had cardioprotective effects and led to reduced fibrosis and improved cardiac function after TAC. Collectively, our results revealed senescent fibroblasts as critical regulators of cardiac fibrogenesis and established the cellular senescence program as a potential target for antifibrotic therapies.
Senescence was initially described in the context of replicative exhaustion and telomere attrition (5). There is now substantial evidence that other stressors (e.g., oncogenic signaling or deoxyribonucleic acid damage) can elicit premature types of senescence independent of telomeres (11,31). Our observation, that CF undergo an initial phase of hyperproliferation (as evidenced by KI-67 expression) that is followed by the accumulation of senescence markers after TAC (e.g., SA-ß-gal) suggested an oncogene-induced senescence–like mechanism. In oncogene-induced senescence, a promitogenic stimulus triggers enhanced cell proliferation before inducing irreversible cell-cycle arrest (31). Moreover, murine cells have long telomeres (32) that make it unlikely to shorten sufficiently to initiate senescence during the 6-week observation period after TAC in this study. Thus, although correlative, our results were consistent with the hypothesis that fibroblast senescence during cardiac fibrosis results from hyperproliferative signals that triggered their initial expansion.
In the heart, senescent cells have been reported in the context of hypoxia and ischemia and mechanistically shown to be dependent on p53 (33). In contrast to these findings, but in line with those of others (34,35), we found that genetic ablation of p53 was insufficient to reduce the number of SA-ß-gal–positive CF (Online Figure 7). Moreover, TAC-induced fibrosis developed to a similar extent in the hearts of Trp53+/+ and Trp53-/- mice (Online Figure 7). Similar results were obtained with a Cdkn2a knockout animal model (Online Figure 8). Instead, we found that genetic disruption of both Trp53 and Cdkn2a gene loci was required to sufficiently reduce CF senescence (Figure 3), and that heart lysates of Trp53-/- Cdkn2a+/- mice displayed a 2-fold compensatory up-regulation of p16INK4a messenger ribonucleic acid level when compared with WT littermates (Online Figure 9). These results underscore that in the heart, p53/p21 and p16/Rb pathways act in parallel to execute the cellular senescence program and that a high degree of functional redundancy exists to perform partially redundant roles. Such a mechanism could provide an extra layer of protection against the bypass of senescence in CF and thus unrestrained myocardial fibrosis.
Which upstream signaling factors and mechanisms trigger senescence in the heart upon cardiac damage await further investigation. We showed that ectopic expression of matricellular protein CCN1 elicits senescence of CF in vitro and in vivo. Mechanistically, CCN1 exerts its prosenescence activity largely through integrin α6β1 and cell surface heparin sulfate proteoglycans, thereby activating RAC-1–dependent nicotinamide adenine dinucleotide phosphate-oxidase 1 to trigger a robust and sustained accumulation of reactive oxygen species. Consequently, CCN1 induces deoxyribonucleic acid damage response and p53 activation, and triggers the reactive oxygen species -dependent activation of p38 mitogen-activated protein kinases and extracellular signal-related kinases, which in turn activate the p16INK4a/Rb pathways to induce senescence (28). Of note, in addition to senescence induction, CCN1 conveys other pathophysiological functions (e.g., angiogenesis) that are mediated by distinct integrin binding sites (e.g., αvβ3 and α6β1) (36). We observed that endogenous CCN1 is strongly expressed in fibrotic areas after TAC, coinciding with a high percentage of senescent cells (data not shown). This concurred with reports of high CCN1 expression levels at sites of tissue repair in the skin (27). In the heart, ischemia/reperfusion was shown to induce a robust CCN1 expression in isolated perfused hearts that was associated with cardioprotection (37). Whether CCN1-mediated cellular senescence contributes to the beneficial effects observed by Zhao et al. (37) remains to be investigated. In addition to CCN1, other factors of the myocardial microenvironment or senescence mechanisms (e.g., Cdkn2a locus de-repression) (38) and altered micro ribonucleic acid expression (39) may contribute to fibroblast senescence. Furthermore, non-CF cell populations of the heart may modulate CF senescence by direct (cell-cell contacts) or indirect (paracrine) signaling mechanisms, for example.
What is the biological function of cellular senescence in the heart? Senescent cells accumulate with age and contribute to age-related diseases, such as atherosclerosis (40). However, there is mounting evidence for a beneficial and protective role of senescence, as evidenced in tumor suppression and the control of tissue remodeling on tissue damage in the liver and skin (6). Our findings expanded the beneficial roles of premature senescence to the heart, further underscoring its importance for tissue remodeling. In accordance with previous observations (15,16), the extent of senescence in the heart correlated positively with the degree of fibrosis, suggesting that the same cell population that produces ECM in wound healing and tissue repair is converted to senescent cells to self-limit fibrosis.
The molecular mechanisms involved in the regulation of ECM homeostasis and fibrosis are complex and include a signaling network of growth factors, cytokines, and cell-cell communications within the tissue microenvironment (2,3). Several mechanisms may contribute to restrain fibrosis via induction of premature senescence in the heart: 1) cell cycle arrest of CF, reducing the number of ECM-producing cells; 2) removal of ECM-producing cells; and 3) curtailing the production of ECM promoting the degradation of matrix components. It has been reported that senescent cells are cleared by the immune system, for example, by natural killer cells (15) or macrophages (34). In addition to cell clearance, senescent cells can also display dramatic changes in their secretory properties. For instance, senescent cells down-regulate genes encoding ECM components and up-regulate ECM-degrading enzymes (e.g., matrix metalloproteinases) (11,13,14). In liver fibrosis, activated stellate cells are the primary source of myofibroblasts, which initially proliferate and produce matrix proteins but eventually become senescent and cease to promote fibrosis (15). Similarly, in excisional skin wounds, recruited fibroblasts and differentiated myofibroblasts initially proliferate and deposit ECM to form granulation tissue. At later stages, myofibroblasts are driven into senescence, where they cease to proliferate and up-regulate matrix-degrading enzymes (16). Thus, the same cell population that produces ECM in wound healing appears to be converted in to matrix-degrading senescent cells to self-limit fibrogenesis.
It is tempting to speculate that harnessing senescence for therapy may be a promising strategy for the prevention or resolution of excess ECM deposition and fibrosis. Of note, clinical trials are currently underway to evaluate the clinical benefit of senescence-inducing drugs for the treatment of cancer (41).
We cannot exclude a potential contribution of other (non-CF) cell types to senescence induction and fibrosis modulation in the gain-of-function and loss-of-function models. Therefore, the role of cell-cell communication within the microenvironment of the heart for senescence-modulated fibrogenesis warrants further investigation.
Our results supported the notion that premature cellular senescence occurs as a general, programmed tissue repair response that functions in disparate organ systems to control fibrosis. An intriguing result of our study was the finding that cardiotropic induction of myofibroblast senescence significantly restrained both perivascular and interstitial fibrosis and improved heart function in vivo. Whether “prosenescence” therapies are a useful strategy to impede excess fibrosis in the heart awaits future investigation.
COMPETENCY IN MEDICAL KNOWLEDGE: Premature senescence of (myo)fibroblasts restrains myocardial fibrosis, contributes to tissue remodeling, and improves cardiac function in vivo.
TRANSLATIONAL IMPLICATIONS: Additional studies should be undertaken to evaluate the clinical benefit of senescence-inducing therapies for patients with diseases mediated by myocardial fibrosis.
The authors thank Sabine Brummer for performing cardiac histology, Kornelija Sakac for echocardiographic analysis and TAC surgeries, Urszula Kremser for primary cell isolations, Stanislas Werfel for help with automated microscopy. We thank M. Klingenspor (TUM), Y. Sassi (TUM), and S. Muehlich (Ludwig Maximilian University, Munich, Germany) for critical discussions. The authors are grateful to R. Hajjar (Icahn School of Medicine at Mount Sinai, New York, New York) for providing human heart biopsies. The authors also thank L. Lau (University of Illinois at Chicago) for kindly providing the CCN1 plasmids.
For a supplemental Materials and Methods section as well as figures and a table, please see the online version of this article.
This study was supported by a research grant from the German Research Foundation (SA 1706/3-1 to Dr. Sarikas), a Marie Curie International Reintegration grant (256584 to Dr. Sarikas), and a Laura-Bassi stipend of Technische Universität München (to K. Meyer). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- adeno-associated virus serotype 9
- cardiac fibroblasts
- extracellular matrix
- senescence-associated ß-galactosidase
- transverse aortic constriction
- wild type
- Received July 17, 2015.
- Revision received February 8, 2016.
- Accepted February 15, 2016.
- American College of Cardiology Foundation
- Leask A.
- Frangogiannis N.G.
- Herbig U.,
- Ferreira M.,
- Condel L.,
- Carey D.,
- Sedivy J.M.
- Dimri G.P.,
- Lee X.,
- Basile G.,
- et al.
- Young A.R.J.,
- Narita M.
- Rockman H.A.,
- Ross R.S.,
- Harris A.N.,
- et al.
- Engelhardt S.,
- Hein L.,
- Wiesmann F.,
- Lohse M.J.
- Ganesan J.,
- Ramanujam D.,
- Sassi Y.,
- et al.
- Kurz D.J.,
- Decary S.,
- Hong Y.,
- Erusalimsky J.D.
- Sherr C.J.,
- Roberts J.M.
- Suckau L.,
- Fechner H.,
- Chemaly E.,
- et al.
- Hulot J.S.,
- Fauconnier J.,
- Ramanujam D.,
- et al.
- Leu S.J.,
- Lam S.C.,
- Lau L.F.
- Minamino T.