FHL2 Inhibits Calcineurin and Represses Pathological Cardiac

FHL2 INHIBITS CALCINEURIN AND REPRESSES PATHOLOGICAL

CARDIAC HYPERTROPHY

APPROVED BY SUPERVISORY COMMITTEE

Joseph A. Hill, M.D., Ph.D. (Mentor)

Joseph A. Garcia, M.D., Ph.D. (Committee Chair)

Michael A. White, Ph.D.

Keith A. Wharton, M.D., Ph.D.

DEDICATION

To Dr. Joseph A. Hill who has for many years been a mentor, a role model, and a friend. What impressed me the most is his passion and enthusiasm for cardiovascular research, his hard work running the division of cardiology and still keep up with everyone’s project in the laboratory. From him I learned not only how to design experiments, but also how to present the findings, and critically address the literature as well as my own data. He was always there to support me when I needed support. Thank you, Joe, I’ve learned a lot from you.

To Dr. Beverly A. Rothermel and her open-door policy. Bev was very influential in my training and education over these years. This work would not have been possible without her contribution, suggestions and troubleshooting.

To Dr. Thomas Gillette for his discussions and expertise inside and outside the laboratory. Tom helped me to wrap up the project by pinpointing the crucial experiments to strengthen my work and put it on paper.

To my high school biology teacher, Rejep Tanish, who recognized my interest in biology and set in motion the trajectory which took me from Turkmenistan, to Turkey, to United States of America.

To my parents, Akmele and Hajymammet, who gave me freedom and support to pursue my interest in biological sciences abroad, even though it meant leaving them for a long time. Their love and belief in me gave me strength throughout my education.

Finally, to my beautiful wife, Jennet, for her love and support. She left her comfortable life and family to come to USA with me. She gave me a beautiful baby boy, Eziz, who was my motivation during writing of my thesis.

FHL2 INHIBITS CALCINEURIN AND REPRESSES PATHOLOGICAL

CARDIAC HYPERTROPHY

BERDYMAMMET HOJAYEV

DISSERTATION

Presented to the Faculty of the Graduate School of Biomedical Sciences

The University of Texas Southwestern Medical Center at Dallas

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

The University of Texas Southwestern Medical Center at Dallas

Dallas, Texas

June, 2010

BERDYMAMMET HOJAYEV, 2010

FHL2 INHIBITS CALCINEURIN AND REPRESSES PATHOLOGICAL

CARDIAC HYPERTROPHY

BERDYMAMMET HOJAYEV, Ph.D.

The University of Texas Southwestern Medical Center at Dallas, 2010

JOSEPH A. HILL, M.D., Ph.D.

Stress-induced cardiac hypertrophy is a hallmark feature of pathological

remodeling which, left unchecked, predisposes hearts to arrhythmia and failure.

FHL2 is a member of the four-and-a-half LIM domain (FHL) family of proteins expressed predominantly in the heart. Targeted disruption of FHL2 leads to an exaggerated response to beta-agonist (isoproterenol)-induced cardiac hypertrophy.

Isoproterenol-induced hypertrophy relies on activation of the calcineurin-NFAT pathway, and inhibition of calcineurin is sufficient to block growth in response to isoproterenol. I also observed that FHL2 is up-regulated in mouse hearts after isoproterenol treatment. Based on this, we hypothesized that FHL2 negatively

regulates the calcineurin-NFAT pathway and consequently, the hypertrophic

growth response. To determine whether calcineurin signaling is enhanced in the

absence of FHL2, wild type (WT) and FHL2 knockout (FHL2-/-) mice were

treated with isoproterenol (32 mg/kg/day). We observed a significant increase in

isoproterenol-induced expression of the NFAT target genes RCAN1.4 and BNP in

FHL2-/- hearts as compared to WT. To determine whether the effect of FHL2 on

the abundance of NFAT target gene transcripts was mediated by calcineurin-

NFAT-dependent transcription, HEK 293 cells were transfected with luciferase

reporter constructs containing the NFAT-driven promoters of either RCAN1 or

IL-2. Consistent with the in vivo data, knockdown of FHL2 message using

siRNA led to increases in both RCAN1 and IL-2 promoter activities elicited by

constitutively active calcineurin or the calcium ionophore, ionomycin.

Importantly, activation of the RCAN1 promoter by ionomycin, in control and

FHL2 knockdown cells, was abolished by the calcineurin inhibitor cyclosporin A, confirming the calcineurin dependence of the response. Over-expression of FHL2 in HEK 293 cells inhibited the activation of both NFAT reporters triggered by either constitutively active calcineurin or ionomycin. Furthermore, neonatal rat ventricular myocytes over-expressing FHL2 exhibited reduced hypertrophic

growth in response to constitutively active calcineurin (measured by cell cross-

sectional area and fetal gene expression). Finally, immunostaining of adult

cardiomyocytes revealed co-localization of FHL2 and calcineurin predominantly

at the sarcomere, and activation of calcineurin by endothelin-1 treatment resulted in interaction between FHL2 and calcineurin as demonstrated by co- immunoprecipitation. These observations demonstrate that FHL2 represses calcineurin-NFAT signaling and thereby suppresses hypertrophic cardiac growth

at least in part by interacting with calcineurin and inhibiting its activation.

vii

TABLE OF CONTENTS

ABSTRACT ...... v TABLE OF CONTENTS ...... viii PRIOR PUBLICATIONS ...... x LIST OF FIGURES ...... xi ABBREVIATIONS ...... xii

CHAPTER 1: INTRODUCTION • Heart failure – epidemiologic burden……………..………………………2 • Progression to heart failure………………………………………………..3 • Hypertrophy as therapeutic target………………………………………....4 • Calcineurin/NFAT pathway…………………………………………….…5 • FHL2……………………………………………………………………....8 o FHL2 expression pattern and regulation…………………………..9 o FHL2 and MEK/ERK pathway………………………………….10 o FHL2 and β-catenin……………………………………………...13 o FHL2 in cancer…………………………………………………..16 o FHL2 as transcription factor……………………………………..18 o FHL2 in the heart………………………………………………...19 • Central thesis……………………………………………………………..22 • References……………………………………………………………….23

CHAPTER 2: FHL2 INHIBITS CALCINEURIN AND REPRESSES PATHOLOGICAL CARDIAC HYPERTROPHY • Introduction…………………………………………………………….30 • Results o Differential roles of FHL2 in various models of hypertrophy….33 β-adrenergic agonist induced hypertrophy………………….33 Pressure overload-induced hypertrophy…………………….34 Constitutively active calcineurin-induced hypertrophy……..35 o Increased FHL2 expression in WT mice treated with isoproterenol……………………………………………………..36 o Activation NFAT target genes in FHL2-/- mouse hearts………..38 o Increased expression of NFAT target genes in FHL2-/- mice…..40 o Knockdown of FHL2 protein in vitro enhances calcineurin- dependent NFAT promoter activation…………………………...41

viii

o FHL2 levels in vitro modulate activation of NFAT target genes by calcineurin……………………………………………………….43 o FHL2 inhibits calcineurin-dependent hypertrophic responses in neonatal rat ventricular myocytes……………………………….44 o FHL2 localizes to the sarcomere in adult and neonatal cardiomyocytes………………………………………………….45 o FHL2 co-localizes and interacts with calcineurin…………….....46 • Interpretation and Conclusion o Inhibition of hypertrophy by FHL2 in vivo……………………..49 o Inhibition of calcineurin/NFAT pathway by FHL2……………..51 o Localization to sarcomere……………………………………….53 o FHL2 interacts with calcineurin………………………………...54 • Figures…………………………………………………………………...55 • Material and Methods…………………………………………………..73 • References………………………………………………………………78

CHAPTER 3: DISCUSSION • Increased FHL2 expression after isoproterenol infusion……………….82 • Inhibition of calcineurin by FHL2………………………………………84 • FHL2 and the MEK1/1-ERK1/2 pathway……………………………..87 • Suppression of MEK1/2-ERK1/2 pathway through calcineurin………..89 • FHL2 and FHL1………………………………………………………..91 • FHL2 in TAC and isoproterenol-induced hypertrophy…………………92 • Regulation of binding of FHL2…………………………………………94 • Perspective………………………………………………………………96 • Figures…………………………………………………………………..97 • References…...... 98

PRIOR PUBLICATIONS

Matta H, Hojayev B, Tomar R, Chugh P, & Chaudhary PM (2003) Use of lentiviral vectors for delivery of small interfering RNA. Cancer Biol Ther 2(2):206-210

LIST OF FIGURES

Figure 2.1: Absence of FHL2 amplifies isoproterenol-induced cardiac

hypertrophy ………………………………………………………55

Figure 2.2: Increased expression of FHL2 in hearts of isoproterenol-treated

mouse hearts…………………………………………..…………..57

Figure 2.3: Increased expression of NFAT target genes in FHL2 knockout mice

following isoproterenol stimulation………………………………59

Figure 2.4: Induction NFAT target gene BNP by constitutively active

calcineurin in FHL2-/- hearts …………………………..………..60

Figure 2.5: Absence of FHL2 enhances NFAT target gene expression in mouse

hearts collected in the morning………………………..………….61

Figure 2.6: Increased heart weight to body weight ratio in FHL2-/- mice at 1

year of age…………………………………………………….…..63

Figure 2.7: Knockdown of FHL2 augments NFAT-dependent expression by

calcineurin…………………………………………………….…..64

Figure 2.8: FHL2 modulates NFAT-dependent promoter activity by

calcineurin………………………………………………………...66

Figure 2.9: FHL2 inhibits hypertrophic growth induced by calcineurin in

neonatal rat ventricular myocytes…………………………………68

Figure 2.10: FHL2 localizes to sarcomere in adult cardiomyocytes…………...69

Figure 2.11: FHL2 colocalizes and interacts with calcineurin…………………71

Figure 3.1: FHL2 is not required to localization of ERK to the sarcomere…...97

ABBREVIATIONS

AKAP A-kinase anchoring protein

ANF atrial natriuretic factor

AngII angiotensin II

AR androgen receptor

BNP brain natriuretic factor

BW body weight

Cain calcineurin inhibitor cAMP cyclic adenosine monophosphate

CBP CREB binding protein

CMV cytomegalovirus

CnA calcineurin A

CnB calcineurin B

CREB cAMP response element-binding

CsA cyclosporine A

ERK extracellular signal-regulated kinase

ET-1 endothelin-1

FAK focal adhesion kinase

FBS fetal bovine serum

FHL2 four and a half LIM domain protein

GFP green fluorescent protein

xii

HEK human embryonic kidney

HW heart weight

IL-2 interleukin-2

LIM Lin11, Mec-3, Isl-3

MEF mouse embryonic fibroblast

MEK MAP kinase kinase

MLP muscle LIM protein

MSC mesenchymal stem cell

NFAT nuclear factor of activated T-cells

NRVM neonatal rat ventricular cardiomyocytes

PE phenylephrine

PKA protein kinase A

PKC protein kinase C

RCAN regulator of calcineurin

RMS rhabdomyosarcoma

SK1 sphingosine kinase-1

SRF serum response factor

TAC thoracic aortic constriction

WT wild type

α-MHC alpha myosin heavy chain

β-MHC beta myosin heavy chain

xiii

CHAPTER 1 INTRODUCTION

Heart failure – epidemiologic burden

Cardiac hypertrophy and heart failure are a growing public health problem and the leading cause of morbidity and mortality in the United States. In 2006, approximately 5.8 million patients had heart failure, with about 670,000 new patients diagnosed with heart failure each year [1]. The likelihood of heart failure increases with age; the prevalence jumps from 3.3% in the 40-59 age group to

15% in the 60-79 age group and to 27.5% in the 80+ group [1]. The estimated cost of health care associated with heart failure in the United States for 2010 is over 39 billion dollars [1]. Despite advances in treatment, the prevalence of heart failure continues to rise due to progressively increasing incidence, thus an increased need for office visits and hospitalizations. Also, rises in the incidence and prevalence of heart failure are partly due to successful treatment of other cardiac conditions, such as acute myocardial infarction.

Heart failure is a condition in which heart has lost the capacity to pump blood to organs of the body [2]. A failing heart is characterized by enlarged left ventricular volume, thinned ventricular walls, and weakened muscle which contributes to systolic dysfunction. The strength of the ventricular contraction becomes inadequate for maintaining normal stroke volume, resulting in reduced cardiac output. Impaired blood flow results in congestion of blood in veins, which causes fluid retention in tissues (edema). Accumulation of fluids in lungs

(pulmonary edema) will cause shortness of breath, especially when lying down.

Other symptoms of heart failure are fatigue, nausea, and increased heart rate.

Progression to heart failure

Heart failure is a progressive disorder which can result from many factors including coronary artery disease, myocardial infarction, hypertension, valve disease, infection of heart valves, and diabetes [2]. Increases in cardiac load often trigger hypertrophic growth to normalize wall stress and oxygen demand. As cardiac myocytes are terminally differentiated cells, heart growth is mainly accomplished by an increase in cell size rather than cell number. Cardiac hypertrophy under these conditions is referred as pathological hypertrophy, and it differs from physiological hypertrophy that occurs in response to exercise and pregnancy. Physiological hypertrophy is characterized by proportional ventricular chamber enlargement and wall thickness. In contrast, pathological hypertrophy results in a disproportionate increase in ventricular chamber size and wall thickness: concentric (thick walls and small ventricular chamber) or eccentric

(large dilated chamber and thin walls). Pathological hypertrophy triggers expression of a “fetal gene program” [3], a gene expression profile which is activated during embryonic development. Although pathological hypertrophy appears initially to be a compensatory mechanism, under conditions of chronic stress it progresses to heart failure. Mechanisms underlying cardiac hypertrophy

4 have been studied extensively and many molecular pathways have been identified to play roles in hypertrophic growth (reviewed in [4]). However, it has been proven difficult to identify when and what triggers the transition to heart failure.

Hypertrophy as therapeutic target

Hypertrophy is initially a compensatory mechanism, often going undetected due to lack of symptoms. Symptoms of heart failure appear after deterioration of heart function, and can be evident during diastole or systole. New guidelines are being developed to identify heart failure or risk factors

(hypertension, structural heart disease, family history and diabetes) at an earlier stage [5]. Detection of heart failure at an earlier stage will likely improve the success of treatment.

Hypertrophy, as mentioned above, is a risk factor for development of heart failure, arrhythmias and sudden death. Recent studies have questioned the necessity of normalization of wall stress. Inhibition of hypertrophic growth induced by pressure overload in mice resulted in sustained heart structure and function with no signs of heart failure or myocyte damage [6]. Thus, pathways critical for pathological hypertrophy have become new therapeutic targets.

Targeting hypertrophy as a proccess may be less complicated than targeting heart failure due to a broader knowledge of pathways activated during hypertrophic growth as compared to pathways activated in the transition to heart failure.

Calcineurin/NFAT pathway

Signal-transduction pathways in hypertrophy are inherently complex.

Studies in animal models have facilitated the identification of a number of important pathways modulating cardiac hypertrophy. One of the major mediators of cardiac hypertrophy is the calcineurin/NFAT pathway. Calcineurin is a Ca2+- dependent serine/threonine protein phosphatase that dephosphorylates the NFAT

(Nuclear Factor of Activated T-cells) family transcription factors, leading to their nuclear translocation and subsequent activation of their target genes. Calcineurin consist of two subunits, CnA and CnB, catalytic subunit calcineurin A and regulatory subunit calcineurin B [7]. In the inactive state calcineurin is inhibited by an autoinhibitory domain located at its C-terminus. Hypertrophic stimuli trigger elevated cytoplasmic calcium levels, which in turn lead to activation of calcineurin by the binding of Ca2+-saturated calmodulin. Binding of

Ca2+/calmodulin to calcineurin triggers a conformational change resulting in the

removal of the autoinhibitory domain from the catalytic domain. Calcineurin is

activated in almost all types of pathological hypertrophy. The role of calcineurin in pathological hypertrophy has been extensively studied (reviewed in [7-8]).

Constitutively active calcineurin, lacking the autoinhibitory domain, is sufficient

to induce hypertrophic growth in vivo and in vitro [9]. Knockout of calcineurin

Aβ impairs hypertrophy induced by pressure overload, and angiotensin II or

6 isoproterenol infusion[10]. A similar effect was observed in NFATc3 knockout mice [11], supporting a model in which both calcineurin and NFAT are required for pathological hypertrophy. Due to its central role in hypertrophy, calcineurin has been viewed as a potential therapeutic target. Pharmacological inhibition of calcineurin by CsA (Cyclosporin A) or FK506, blunts the hypertrophic response in pressure overload and myocardial infarction (MI) models, angiotensin II infusion, and calcineurin transgenic mice (reviewed in [7]). Although these drugs are very effective against cardiac hypertrophy, inhibition of calcineurin/NFAT in other tissues causes side effects such as immunosuppression and nephrotoxicity.

In addition to pharmacological inhibitors of calcineurin, several proteins have been were identified that inhibit calcineurin, including RCAN1, AKAP, cain and calsarcin [7]. RCAN1 is a potent inhibitor of calcineurin, which binds to the catalytic subunit of calcineurin and prevents dephosphorylation of NFAT by calcineurin. RCAN1 preferentially binds to activated calcineurin, suggesting that it competes with the autoinhibitory domain of calcineurin [12]. Moreover, the promoter of the exon4 isoform, RCAN1.4, contains 15 NFAT binding sites, thus activation of the calcineurin/NFAT pathway will specifically induce expression of this isoform. RCAN1.4 can act as a negative feedback regulator of calcineurin, in the sense that its expression is increased with activation of calcineurin; newly synthesized RCAN1.4 binds and inhibits active calcineurin. Once dissociated from calcineurin RCAN1.4 is targeted for degradation. Overexpression of

RCAN1.4 inhibits hypertrophic growth in vivo and in vitro, suggesting that inhibition of calcineurin by RCAN1.4 inhibits the calcineurin/NFAT pathway in hypertrophy induced by various stimuli.

AKAP79 (A-kinase associated protein 79) is a scaffolding protein that has been shown to mediate interaction of calcineurin with PKA and PKC [13].

AKAP79 anchors protein kinases and phosphatases to the membrane where it also interacts with β1 and β2 adrenergic receptors and the L-type calcium channel.

Cain (calcineurin inhibitor) is a large protein (240 kDa) that is believed to be a

scaffolding protein. It has been demonstrated that overexpression of cain leads to

inhibition of calcineurin [14]. Overexpression of the calcineurin binding regions

of AKAP79 and cain is sufficient to inhibit hypertrophic growth in vitro and in

vivo [15-16]. However, endogenous expression of AKAP79 and cain is low in the

heart. Recently a novel calcineurin inhibitor has been identified, calsarcin [17].

Calsarcin has been shown to localize to the sarcomere through its interaction with

α-actinin. Calsarcin knockout mouse hearts were sensitized to pressure overload

and constitutive activation of calcineurin [17]. Consistent with hypertrophic

growth, expression of hypertrophic markers and NFAT target genes were elevated

in calsarcin-deficient hearts. In vitro, calsarcin overexpression inhibits

calcineurin dependent dephosphorylation of NFAT. Transgenic mice that

overexpress calsarcin manifest a decrease in angiotensin II-induced hypertrophy

[18].

FHL2

The LIM domain was named after the first initials of the first genes shown to possess this domain Lin11 (C. elegans), Isl-3 (rat), and Mec-3 (C. elegans).

The LIM domain consists of cysteine-histidine rich, zinc binding, double zinc- finger like structures. The classical sequence of the LIM domain is CX2CX16-

23HX2CX2CX2CX16-21CX2(C, H, D) (X any amino acid) [19].

LIM domains are thought to be protein interaction domains, suggesting that LIM proteins might be involved in anchoring partners to a specific subcellular location, modulating their activity, or promoting interaction between proteins [20]. LIM domains are present in a variety of eukaryotic proteins with diverse biological functions. Depending on the presence of other functional domains, LIM proteins are divided into three groups: LIM-only, LIM- homeodomain and LIM-kinase.

FHL2, a LIM-only protein, was first identified from a subtractive cDNA hybridization screen of normal myoblasts and rhabdomyosarcoma (RMS, skeletal muscle cancer) cells [21]. FHL2 was expressed in primary myoblasts but was down-regulated in a RMS cell line, RD, which has been shown to express mutated p53 and an activated N-ras oncogene. Further analysis of other RMS cell lines demonstrated consistently low expression of FHL2 in these lines [21], which was the rationale for its initial nomenclature DRAL, down-regulated in

9 rhabdomyosarcoma LIM protein. FHL2 was also identified by another group in

skeletal muscle and was named SLIM3, skeletal LIM protein 3 [22]. Upon

discovery of additional homologous proteins, these proteins were classified into

the Four and a Half LIM domain family, which consist of four LIM domains and

an N-terminal half LIM domain. Currently there are five members in this family:

FHL1, FHL2, FHL3, FHL4, and ACT. Comparison of FHL2 protein to other

members for FHL in humans reveal 48% identity with FHL1, 52% with FHL3,

47% with FHL4 and 59% with ACT. Analysis of FHL gene family expression

shows a tissue specific patterns. FHL2 is mainly expressed in the heart, while the

highest expression of FHL3 is in skeletal muscle. FHL4 and ACT are only

detected in testis. However, FHL1 is more ubiquitous, detected at high levels in

many tissues including heart, skeletal muscle, ovary, kidney, lung and brain.

FHL2 expression pattern and regulation

Although FHL2 was first shown to be down-regulated in rhabdomyosarcoma, extensive expression analysis of human and mouse tissues demonstrated that FHL2 is primarily expressed in the heart [21]. Lower expression was detected in ovary, testis, adrenal gland, skeletal muscle and small intestine [21, 23-25]. Very low mRNA levels were observed in lung, liver and kidney by northern blot analysis [26], however no expression was observed using the dot blot hybridization method [24]. Some studies have demonstrated

10 expression in brain and skeletal muscle. No expression was detected in skin, spleen and thymus. Western blot analysis could only detect FHL2 protein expression in the heart, which was localized to ventricles and was undetectable in atria. Disagreement regarding expression of FHL2 in tissues could be explained by different detection methods utilized by different groups.

FHL2 expression has been further analyzed during embryonic development in mice. FHL2 expression has been tracked by a β-galactosidase activity assay, which was introduced into the FHL2 locus by homologous recombination. At E7.5-E8.0, β-galactosidase activity was detected in the cardiac crescent. Heart-specific expression of FHL2 is persistent in later embryonic stages from E9.0 to E14 [27]. These results were further confirmed by in situ hybridization, which showed expression in the heart, both in atria and ventricles.

This suggests that atrial expression of FHL2 decreases later during development, as no FHL2 was detected in atria of adult mouse hearts.

The chromosomal location of FHL2 is 2q12-q14 and 1B in human and mouse, respectively. The human FHL2 gene consists of 7 exons, the first three of which are non-coding. Mouse FHL2 message consist of 6 exons, and expression starts from the second exon. Promoter analysis suggests the presence of many putative transcription binding sites including NKX2.5, p53, SRF, MEF-2, E2F and AP-1. However, only two of them have been confirmed so far.

Low expression of FHL2 in rhabdomyosarcoma RD cells with mutant p53 led to a study to test whether FHL2 was a target gene of p53 [24]. An experiment was carried out in RD cells stably transfected with temperature sensitive-p53 or empty vector. FHL2 mRNA levels were increased when p53 was expressed at

32° C, however no changes were observed when p53 was expressed at 38° C or in empty vector transfected cells at both temperatures. These data suggest that endogenous p53 was sufficient to induce FHL2 expression in primary human myoblasts (containing wild type p53) when treated with ionizing radiation (IR).

RD cells and p53-/- mouse embryo fibroblasts failed to induce FHL2 expression after IR. Re-expression of p53 in RD cells restored FHL2 induction by IR.

A screen for target genes for the transcription factor SRF (serum response factor) identified FHL2. Promoter analysis confirmed the presence of a conserved

CArG consensus sequence that binds SRF. FHL2 transcript was upregulated when Srf-/- cells were transfected with constitutively active SRF fusion protein

SRF-VP16 [28]. Moreover, SRF was able to bind the CArG box of FHL2, as demonstrated by ChIP assay. Consistent with mRNA findings, FHL2 protein was induced by SRF-VP16, but not mutant SRFΔM-VP16, in Srf-/- cells. FHL2 promoter activity was increased by constitutively active RhoAV14. RhoA- dependent activation was completely abolished when the CArG box was mutated, which suggests that SRF is essential for regulation of FHL2 expression by RhoA.

FHL2 is predominantly expressed in the heart, which suggests regulation by cardiac specific transcription factors. NKX2.5 is a cardiac transcription factor that is essential for cardiac development of the embryo and abundantly expressed in adult heart, and FHL2 has multiple putative NKX2.5 binding sites in its promoter. Furthermore, whole-mount in situ hybridization in early mouse embryos revealed that both FHL2 and NKX2.5 mRNA are detected in the cardiac crescent at E7.5, and continue to be expressed in similar pattern. The presence of both SRF and NKX2.5 binding sites in the FHL2 promoter, along with the fact that they interact with each other, consistent with the hypothesis that SRF and

NKX2.5 act synergistically on the FHL2 promoter.

FHL2 and MEK/ERK pathway

MEK1/2-ERK1/2 pathway have been implicated in the hypertrophic growth response of cardiomyocytes both in vivo and in vitro. Phosphorylation of

ERK1/2 has been observed in cultured myocytes treated with hypertrophic agonists such as isoproterenol, endothelin-1 (ET-1), and angiotensin II (AngII), and this hypertrophic response could be blocked when cells were infected with virus expressing dominant negative MEK1. Similarly, ERK1/2 activation is augmented in a pressure overload mouse model [29]. Transgenic mice expressing constitutively active MEK1 develop a hypertrophic phenotype [30]. A search for

ERK2 binding partners identified FHL2 as possible interaction partner [31].

FHL2 inhibited activation of promoters downstream of the MEK1/2-ERK1/2 pathway through reduced phosphorylation of the GATA-4 transcription factor.

Whereas phosphorylation of ERK2 was not modified by FHL2, the nuclear localization of ERK2 was diminished in FHL2-expressing cells [31].

Immunoprecipitation analysis demonstrated that FHL2 has a higher affinity for phosphorylated ERK2 compared to unphosphorylated ERK2, suggesting that

FHL2 binds to active ERK2 and prevents its nuclear translocation, thus blocking downstream signaling. Furthermore, FHL2 inhibited hypertrophic growth triggered by PE, constitutively active MEK1, and GATA-4 in cultured neonatal ventricular myocytes [31]. Although FHL2 had no effect on ERK2 phosphorylation in cardiomyocytes, others have shown that FHL2 modulates

ERK2 phosphorylation in HEK 293 and MCF-7 cells [32-33]. Increased ERK2 phosphorylation and nuclear translocation were associated with increased expression of FHL2 in HEK 293 cells [32]. In another study, low FHL2 levels were associated with increased ERK2 phosphorylation in MCF-7 breast cancer cells and in bone marrow-derived stem cells [33-34].

FHL2 and β-catenin

Given that FHL2 is downregulated in rhabdomyosarcoma, a skeletal muscle cancer, its role in muscle tissue has been investigated further. A yeast- two-hybrid screen was performed on cDNA from differentiated mouse myoblasts using FHL2 as bait. β-catenin was identified as one of the binding proteins, which was confirmed in vitro and in vivo [35-36]. β-catenin is an essential molecule of the Wnt signaling pathway. When Wnt signaling is off, β-catenin is phosphorylated by GSK-3β, targeted for ubiquitination by β-TrCP, and subsequently degraded by the proteasome. Activation of the Wnt receptor stabilizes β-catenin, which in turn translocates into the nucleus and binds

TCF/LEF transcription factors. TCF/LEF transcription factors occupy target promoters in the absence of β-catenin and act as repressors. However, association of β-catenin activates TCF/LEF dependent expression. Overexpression of FHL2 inhibited activation of a TCF/LEF-dependent promoter in cultured cells in a dose- dependent manner [35]. Furthermore, cells transfected with FHL2 formed myotubes more efficiently compared to control cells. After 3 days in differentiation medium, FHL2-transfected cells formed well-developed myotubes, whereas in control cells this stage is reached by day 5. Consistent with this visual evidence for differentiation, increased expression of myosin heavy chain was also detected in FHL2-expressing cells. [35]. However, FHL2 activated a TCF/LEF- dependent promoter in HEK 293 cells, suggesting the inhibitory effect of FHL2 is

15 specific to myoblasts [35-37]. Activation of TCF/LEF promoters by FHL2 was further enhanced by the coactivator CBP/p300 [38]. Interaction of FHL2 with

CBP or p300 was demonstrated in vitro and in vivo. Furthermore, FHL2, p300 and β-catenin formed a ternary complex in which FHL2 increased acetylation of

β-catenin by p300, which increases association of β-catenin to TCF.

Transactivation of β-catenin by FHL2 was reported to increase mesenchymal stem cell (MSC) differentiation into osteoblasts [39].

Dexamethasone, an inducer of MSC differentiation into osteoblasts, increased the expression of FHL2 in MSCs. Overexpression of FHL2 led to induction of osteoblast cell markers, whereas knockdown of FHL2 attenuated their expression.

FHL2 interacted with β-catenin and increased its nuclear localization and promoter activation. Wnt increased expression of osteoblast markers by FHL2, whereas the Wnt inhibitor DKK1 abolished the effect of FHL2, confirming that

FHL2 can act on the Wnt/β-catenin pathway.

The effect of FHL2 in β-catenin signaling was also demonstrated through study of the cyclin D1 promoter. ChIP (chromatin immunoprecipitation) assay showed that FHL2 binds to the cyclin D1 promoter at the TCF/CRE site [40].

Expression of cyclin D1 was lower in FHL2-null MEF cells, which also display slow cell cycle progression and lower expression of genes involved in the cell cycle, such as cyclin A, cyclin E and p16INK4. Re-expression of FHL2 or cyclin

D1 in FHL2-null MEF cells restored their proliferative capacity. Microarray

16 analysis showed deregulation of a broad range of proteins involved in cell cycle regulation in FHL2-null MEF cells.

FHL2 in cancer

Identification of FHL2 in cancer cell lines allowed investigation into

FHL2’s role in malignancy. A number of studies have demonstrated that FHL2 expression is often deregulated in various cancer tissues, upregulated in some and down-regulated in others [41]. Surprisingly, FHL2 can function as either an oncogene or a tumor suppressor depending on the cell type. To date, FHL2 has been shown to interact with several proteins involved in cancer.

Expression analysis of FHL2 in tumors revealed that it is down-regulated in rhabdomyosarcoma and prostate cancer. However, FHL2 expression was highly elevated in breast[42] and ovarian cancer[43], human melanoma[44] and colon carcinoma[45].

Of the breast tumors tested, these tissues expressed either low or high levels of FHL2 compared to normal tissues. High levels of expression correlated with worse patient survival compared to tumors with low FHL2 expression [42].

Another study demonstrated interaction of FHL2 with BRCA1 tumor suppressor

[46], however, the physiological relevance of this interaction has not been elucidated. FHL2 was also overexpressed in epithelial ovarian cancer [43], where it localizes to the cytoplasm and membrane. It interacts with and colocalizes with

FAK (focal adhesion kinase). However, the contribution of this interaction and localization to cancer prognosis remains unknown.

FHL2 is expressed in normal prostate tissue, and FHL2 levels are increased in prostate cancer cells [47]. Immunohistochemical analysis of FHL2 in prostate cancer tissues revealed that nuclear localization of FHL2 correlated with poorer prognosis measured by Gleason score. FHL2 interacts with the androgen receptor (AR) and potentiates expression from AR-dependent promoters [47].

This suggests that localization of FHL2 to the nucleus and interaction with AR, and consequent increased expression of its target genes, contribute to prostate cancer progression.

FHL2 expression is also up-regulated in gastrointestinal cancer cells [45].

To understand the role of FHL2 in these cells, FHL2 expression was targeted by either antisense or siRNA. Targeting FHL2 led to improved differentiation capacity in these cells [45]. Moreover, it led to decreased expression of oncogenes, and serum-dependent and anchorage-dependent cell growth.

Moreover, suppression of FHL2 reduced the capability of cancer cells to induce formation of tumors in vivo when injected in nude mice [45].

FHL2 interacts with many proteins that are involved in cancer pathogenesis, including β-catenin, AP-1, androgen receptor, BRCA1, FoxO1, SKI and ERK2. The role of FHL2 in carcinogenesis appears to depend on the origin

18 of the tumor, its subcellular localization, and the role of binding partners in the particular tumor type.

FHL2 as transcription factor

Although LIM domains consist solely of two zinc finger-like structures, they do not bind DNA. Indeed, there is no evidence that FHL2 binds DNA. That said, numerous studies suggest that FHL2 plays a role in transcription as either activator or repressor. For example, a GAL4-DNA binding domain fused to

FHL2 stimulated transcription from a GAL-4 minimal promoter [48].

As noted above, FHL2 has been demonstrated to interact with the androgen receptor (AR) and act as coactivator of AR-dependent gene expression

[47]. Coactivator functions of FHL2 are ligand-dependent, and they are abolished by mutation of the activation domain of AR. AR activity is critical for prostate cancer progression. FHL2 levels are upregulated in prostate cancer tissues, with increased nuclear localization detected [47]. FHL2 was also identified as an activator of peroxisome proliferator activated receptor α/ retinoid receptor X receptor α (PPARα/RXRα) in keratinocytes [49], however the mechanism remains unknown due to limited interaction of FHL2 with PPARα.

FHL2 binds CREB and stimulates CREB-mediated transcription [23]. An interaction site was mapped to the kinase-inducible domain of CREB, the same location required for CBP association. CBP interaction requires phosphorylation

19 of serine-133, whereas FHL2 interaction is independent of phosphorylation status.

Both FHL2 and CBP stimulate CREB-mediated transcription, but it is unknown whether they are mutually exclusive, because they bind to the same region of

CREB. However, FHL2 and CBP/p300 synergistically stimulated β-catenin- dependent transcription and formed a ternary complex [38]. This synergistic effect was also observed on androgen-dependent promoters [38].

FHL2 has also been identified as a coactivator of AP-1, which consists of

Jun and Fos transcription factors. FHL2 interacts with Jun and Fos and strongly stimulates their target promoters [50]. Coexpression of FHL2 with GAL4-Jun or

GAL4-Fos increased activation of a GAL4-dependent minimal promoter.

Recently it was shown that FHL2 interacts with the transcription factor

Runx2 [51], a protein essential for normal function of osteoblasts. Mice null for

Runx2 completely lack mineralized tissue [52]. FHL2 stimulates Runx2- dependent transcription. FHL2 knockout mice develop osteopenia [51, 53], while overexpression in osteoblasts results in increased bone mass. A separate study suggested that FHL2 might a play role in differentiation of osteoblasts via stimulation of CREB activity [54].

FHL2 in the heart

FHL2 has been studied extensively since its initial identification in 1996

[22]. Over 100 studies have analyzed the role of FHL2 in various cellular

20 functions and tissues, with over 50 interaction partners identified. However, even though the highest expression levels of FHL2 are in the heart, a very limited number of studies have addressed the role of FHL2 in this tissue.

Two groups independently developed FHL2 knockout mice [27] . In both groups, FHL2 knockout mice displayed no obvious phenotype, with normal heart function and structure. Interestingly, treatment of FHL2 knockout mice with isoproterenol resulted in an exaggerated hypertrophic response accompanied with increased expression of the hypertrophic marker ANF compared to wild-type mice [27]. Isoproterenol induced a 21% increase in heart weight to body weight ratio in wild-type mice as compared to a 58% increase in FHL2 knockout mice.

However, hypertrophic growth in response to pressure overload was indistinguishable from wild type mice [55].

FHL2 was identified as a binding partner for ERK2, and they both localize to the sarcomere in isolated adult cardiomyocytes [31]. The MAP (mitogen activated protein) kinase MEK1/2-ERK1/2 is an important pathway in hypertrophy, which is activated in many hypertrophic conditions. FHL2 interacts with ERK2, and the affinity of FHL2 for ERK2 was increased when ERK2 was phosphorylated [31]. Interaction of FHL2 with phospho-ERK2 blocked its translocation into the nucleus, and as a result blunted its downstream signaling.

FHL2 inhibited MEK1-activated ANF and GATA-responsive promoters in a dose-dependent manner. In addition, overexpression of FHL2 in neonatal

21 cardiomyocytes inhibited PE and MEK-1-induced hypertrophy as determined by cell cross-sectional area and hypertrophic marker expression [31].

Several studies have implicated FHL2 in apoptosis; overexpression of

FHL2 induced apoptosis in RD, COS-1 and NIH 3T3 cells [24]. A recent study suggests that FHL2 can also regulate apoptosis in cardiomyocytes [56]. FHL2 was identified as a SK-1 (sphingosine kinase-1) interacting protein from a yeast- two-hybrid assay, and interaction was confirmed in vitro and in vivo. SK-1 has been demonstrated to play an important role in diverse biological processes including cell survival. Overexpression of SK1 promotes cell survival and protects cells from apoptotic insults [57]. FHL2 associates with SK1 and colocalizes to the cytoplasm. Overexpression of FHL2 markedly decreased SK1 activity and its protective effects in cardiomyocytes [56]. Knockdown of FHL2 in neonatal cardiomyocytes increased SK1 activity and protected cells from apoptosis induced by H2O2 [56]. Endothelin-1, which is known to induce SK1

activity, triggered dissociation of FHL2 from SK1 and increased activity of SK1

[56]. This suggests that FHL2 may act as a switch between apoptosis and survival, promoting apoptosis when it binds and inhibits SK-1, and promoting survival when it dissociates from SK-1.

Central thesis

Hypertrophy is initially an adaptive mechanism to cope with increased hemodynamic or neurohumoral stress, but it is also a risk factor for heart failure and arrhythmia. A recent study has demonstrated that the hypertrophic response to β-adrenergic stimulation was amplified in FHL2 knockout mice suggesting an inhibitory role of FHL2 in hypertrophy. In addition, calcineurin/NFAT signaling has been demonstrated to be a critical regulator of hypertrophy; inhibition of this pathway protects hearts from hypertrophic growth.

Based on these facts, we hypothesized that FHL2 suppresses isoproterenol-induced hypertrophy through inhibition of the calcineurin/NFAT pathway. I report that calcineurin/NFAT-dependent expression is increased in

FHL2 knockout mice and in isoproterenol-treated mouse hearts. Overexpression of FHL2 suppressed activation of NFAT-dependent promoters by endogenous or constitutively active calcineurin. Conversely, FHL2 siRNA increased expression from these promoters following calcineurin activation. Additionally, I demonstrate that FHL2 and calcineurin colocalize at the sarcomere and physically associate upon activation of calcineurin. Finally, overexpression of FHL2 blunted the hypertrophic response to constitutively active calcineurin in neonatal cardiomyocytes at least in part by interacting with and inhibiting calcineurin signaling.

References

1. Lloyd-Jones, D., et al., Heart disease and stroke statistics--2010 update: a report from the American Heart Association. Circulation, 2010. 121(7): p. e46-e215.

2. Hill, J.A. and E.N. Olson, Cardiac plasticity. N Engl J Med, 2008. 358(13): p. 1370-80.

3. Colucci, W.S., Molecular and cellular mechanisms of myocardial failure. Am J Cardiol, 1997. 80(11A): p. 15L-25L.

4. Heineke, J. and J.D. Molkentin, Regulation of cardiac hypertrophy by intracellular signalling pathways. Nat Rev Mol Cell Biol, 2006. 7(8): p. 589-600.

5. Jessup, M., et al., 2009 focused update: ACCF/AHA Guidelines for the Diagnosis and Management of Heart Failure in Adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation, 2009. 119(14): p. 1977-2016.

6. Hill, J.A., et al., Cardiac hypertrophy is not a required compensatory response to short-term pressure overload. Circulation, 2000. 101(24): p. 2863-9.

7. Wilkins, B.J. and J.D. Molkentin, Calcium-calcineurin signaling in the regulation of cardiac hypertrophy. Biochem Biophys Res Commun, 2004. 322(4): p. 1178-91.

8. Wilkins, B.J. and J.D. Molkentin, Calcineurin and cardiac hypertrophy: where have we been? Where are we going? J Physiol, 2002. 541(Pt 1): p. 1-8.

9. Molkentin, J.D., et al., A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell, 1998. 93(2): p. 215-28.

10. Bueno, O.F., et al., Impaired cardiac hypertrophic response in Calcineurin Abeta -deficient mice. Proc Natl Acad Sci U S A, 2002. 99(7): p. 4586-91.

11. Wilkins, B.J., et al., Targeted disruption of NFATc3, but not NFATc4, reveals an intrinsic defect in calcineurin-mediated cardiac hypertrophic growth. Mol Cell Biol, 2002. 22(21): p. 7603-13.

12. Rothermel, B.A., et al., Myocyte-enriched calcineurin-interacting protein, MCIP1, inhibits cardiac hypertrophy in vivo. Proc Natl Acad Sci U S A, 2001. 98(6): p. 3328-33.

13. Faux, M.C., et al., Mechanism of A-kinase-anchoring protein 79 (AKAP79) and protein kinase C interaction. Biochem J, 1999. 343 Pt 2: p. 443-52.

14. Lai, M.M., et al., Cain, a novel physiologic protein inhibitor of calcineurin. J Biol Chem, 1998. 273(29): p. 18325-31.

15. De Windt, L.J., et al., Targeted inhibition of calcineurin attenuates cardiac hypertrophy in vivo. Proc Natl Acad Sci U S A, 2001. 98(6): p. 3322-7.

16. Taigen, T., et al., Targeted inhibition of calcineurin prevents agonist- induced cardiomyocyte hypertrophy. Proc Natl Acad Sci U S A, 2000. 97(3): p. 1196-201.

17. Frey, N., et al., Mice lacking calsarcin-1 are sensitized to calcineurin signaling and show accelerated cardiomyopathy in response to pathological biomechanical stress. Nat Med, 2004. 10(12): p. 1336-43.

18. Frank, D., et al., Calsarcin-1 protects against angiotensin-II induced cardiac hypertrophy. Circulation, 2007. 116(22): p. 2587-96.

19. Sanchez-Garcia, I. and T.H. Rabbitts, The LIM domain: a new structural motif found in zinc-finger-like proteins. Trends Genet, 1994. 10(9): p. 315- 20.

20. Dawid, I.B., J.J. Breen, and R. Toyama, LIM domains: multiple roles as adapters and functional modifiers in protein interactions. Trends Genet, 1998. 14(4): p. 156-62.

21. Genini, M., et al., Subtractive cloning and characterization of DRAL, a novel LIM-domain protein down-regulated in rhabdomyosarcoma. DNA Cell Biol, 1997. 16(4): p. 433-42.

22. Morgan, M.J. and A.J. Madgwick, Slim defines a novel family of LIM- proteins expressed in skeletal muscle. Biochem Biophys Res Commun, 1996. 225(2): p. 632-8.

23. Fimia, G.M., D. De Cesare, and P. Sassone-Corsi, A family of LIM-only transcriptional coactivators: tissue-specific expression and selective activation of CREB and CREM. Mol Cell Biol, 2000. 20(22): p. 8613-22.

24. Scholl, F.A., et al., DRAL is a p53-responsive gene whose four and a half LIM domain protein product induces apoptosis. J Cell Biol, 2000. 151(3): p. 495-506.

25. Chan, K.K., et al., Protein-protein interaction of FHL2, a LIM domain protein preferentially expressed in human heart, with hCDC47. J Cell Biochem, 2000. 76(3): p. 499-508.

26. Tanahashi, H. and T. Tabira, Alzheimer's disease-associated presenilin 2 interacts with DRAL, an LIM-domain protein. Hum Mol Genet, 2000. 9(15): p. 2281-9.

27. Kong, Y., et al., Cardiac-specific LIM protein FHL2 modifies the hypertrophic response to beta-adrenergic stimulation. Circulation, 2001. 103(22): p. 2731-8.

28. Philippar, U., et al., The SRF target gene Fhl2 antagonizes RhoA/MAL- dependent activation of SRF. Mol Cell, 2004. 16(6): p. 867-80.

29. Ocaranza, M.P. and J.E. Jalil, Mitogen-activated protein kinases as biomarkers of hypertension or cardiac pressure overload. Hypertension, 2010. 55(1): p. 23-5.

30. Bueno, O.F., et al., The MEK1-ERK1/2 signaling pathway promotes compensated cardiac hypertrophy in transgenic mice. EMBO J, 2000. 19(23): p. 6341-50.

31. Purcell, N.H., et al., Extracellular signal-regulated kinase 2 interacts with and is negatively regulated by the LIM-only protein FHL2 in cardiomyocytes. Mol Cell Biol, 2004. 24(3): p. 1081-95.

32. Lauffart, B., et al., Interaction of TACC proteins with the FHL family: implications for ERK signaling. J Cell Commun Signal, 2007. 1(1): p. 5- 15.

33. Saeki, Y., et al., Ligand-specific sequential regulation of transcription factors for differentiation of MCF-7 cells. BMC Genomics, 2009. 10: p. 545.

34. Park, J., et al., Deficiency in the LIM-only protein FHL2 impairs assembly of extracellular matrix proteins. FASEB J, 2008. 22(7): p. 2508-20.

35. Martin, B., et al., The LIM-only protein FHL2 interacts with beta-catenin and promotes differentiation of mouse myoblasts. J Cell Biol, 2002. 159(1): p. 113-22.

36. Wei, Y., et al., Identification of the LIM protein FHL2 as a coactivator of beta-catenin. J Biol Chem, 2003. 278(7): p. 5188-94.

37. Chen, D., et al., SKI activates Wnt/beta-catenin signaling in human melanoma. Cancer Res, 2003. 63(20): p. 6626-34.

38. Labalette, C., et al., Interaction and functional cooperation between the LIM protein FHL2, CBP/p300, and beta-catenin. Mol Cell Biol, 2004. 24(24): p. 10689-702.

39. Hamidouche, Z., et al., FHL2 mediates dexamethasone-induced mesenchymal cell differentiation into osteoblasts by activating Wnt/beta- catenin signaling-dependent Runx2 expression. FASEB J, 2008. 22(11): p. 3813-22.

40. Labalette, C., et al., The LIM-only protein FHL2 regulates cyclin D1 expression and cell proliferation. J Biol Chem, 2008. 283(22): p. 15201-8.

41. Kleiber, K., K. Strebhardt, and B.T. Martin, The biological relevance of FHL2 in tumour cells and its role as a putative cancer target. Anticancer Res, 2007. 27(1A): p. 55-61.

42. Gabriel, B., et al., Expression of the transcriptional coregulator FHL2 in human breast cancer: a clinicopathologic study. J Soc Gynecol Investig, 2006. 13(1): p. 69-75.

43. Gabriel, B., et al., Focal adhesion kinase interacts with the transcriptional coactivator FHL2 and both are overexpressed in epithelial ovarian cancer. Anticancer Res, 2004. 24(2B): p. 921-7.

44. Morgan, M.J. and S.A. Whawell, The structure of the human LIM protein ACT gene and its expression in tumor cell lines. Biochem Biophys Res Commun, 2000. 273(2): p. 776-83.

45. Wang, J., et al., Suppression of FHL2 expression induces cell differentiation and inhibits gastric and colon carcinogenesis. Gastroenterology, 2007. 132(3): p. 1066-76.

46. Yan, J., et al., BRCA1 interacts with FHL2 and enhances FHL2 transactivation function. FEBS Lett, 2003. 553(1-2): p. 183-9.

47. Muller, J.M., et al., FHL2, a novel tissue-specific coactivator of the androgen receptor. EMBO J, 2000. 19(3): p. 359-69.

48. Yan, J.H., et al., [Mapping of FHL2 transcription activation domain]. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai), 2003. 35(7): p. 643-8.

49. Ciarlo, J.D., et al., FHL2 expression in keratinocytes and transcriptional effect on PPARalpha/RXRalpha. J Dermatol Sci, 2004. 35(1): p. 61-3.

50. Morlon, A. and P. Sassone-Corsi, The LIM-only protein FHL2 is a serum- inducible transcriptional coactivator of AP-1. Proc Natl Acad Sci U S A, 2003. 100(7): p. 3977-82.

51. Gunther, T., et al., Fhl2 deficiency results in osteopenia due to decreased activity of osteoblasts. EMBO J, 2005. 24(17): p. 3049-56.

52. Xiao, Z., et al., Selective Runx2-II deficiency leads to low-turnover osteopenia in adult mice. Dev Biol, 2005. 283(2): p. 345-56.

53. Govoni, K.E., et al., Disruption of four-and-a-half LIM 2 decreases bone mineral content and bone mineral density in femur and tibia bones of female mice. Calcif Tissue Int, 2006. 79(2): p. 112-7.

54. Lai, C.F., et al., Four and half lim protein 2 (FHL2) stimulates osteoblast differentiation. J Bone Miner Res, 2006. 21(1): p. 17-28.

55. Chu, P.H., et al., FHL2 (SLIM3) is not essential for cardiac development and function. Mol Cell Biol, 2000. 20(20): p. 7460-2.

56. Sun, J., et al., FHL2/SLIM3 decreases cardiomyocyte survival by inhibitory interaction with sphingosine kinase-1. Circ Res, 2006. 99(5): p. 468-76.

57. Olivera, A., et al., Sphingosine kinase expression increases intracellular sphingosine-1-phosphate and promotes cell growth and survival. J Cell Biol, 1999. 147(3): p. 545-58.

CHAPTER 2

FHL2 INHIBITS CALCINEURIN AND REPRESSES

PATHOLOGICAL CARDIAC HYPERTROPHY

Introduction

A large number of physiologic and pathologic stimuli are able to trigger

hypertrophy of the heart, though the molecular profiles and structural changes can

be completely different. Physiological hypertrophy is stimulated by pregnancy

and exercise, whereas pathological hypertrophy is elicited by increased cardiac

load, myocardial injury or neurohormonal stimulation [1]. As cardiac myocytes

are terminally differentiated, heart growth is accomplished by an increase in

myocyte size rather than cell number. Pathological hypertrophy is believed to be

an initially compensatory mechanism to lessen wall stress and normalize

myocardial oxygen demand. However, under conditions of chronic stress,

hypertrophy may progress to heart failure and arrhythmia, the number one cause

of morbidity and mortality in Western society [2].

Calcineurin, a Ca2+-calmodulin-dependent serine/threonine protein

phosphatase, is activated in a variety of hypertrophic insults including

isoproterenol-induced hypertrophy [3]. Upon activation, calcineurin

dephosphorylates NFAT (Nuclear Factor of Activated T-cells), which in turn translocates into the nucleus and activates expression from its target promoters.

The calcineurin-NFAT cascade is considered a critical pathway in cardiac hypertrophy. Transgenic mice overexpressing calcineurin [4-5] or NFAT [5] develop remarkable cardiac hypertrophy followed by rapid progression to heart failure. Inhibition of calcineurin genetically or pharmacologically is sufficient to

31 block hypertrophic growth in response to pressure overload or neurohormonal stimulation, as well as in transgenic models of hypertrophy (reviewed [3]).

FHL2, four and a half LIM domain protein, is expressed predominantly in the heart. LIM domains have been implicated in protein-protein interactions, and recent studies have identified over 50 FHL2 binding partners (reviewed [6]).

FHL2 is involved in many processes including cell cycle regulation [7-8], apoptosis [9-10], differentiation [11-14], extracellular matrix assembly [15], bone formation, [16] and wound healing [17-18]. Although the highest expression of

FHL2 is in the heart, knockout mice are viable and display no obvious cardiac phenotype under normal conditions [19-20]. However, when treated with the β- adrenergic agonist isoproterenol, FHL2 knockout mice develop an exaggerated hypertrophic phenotype [19].

The fact that FHL2 knockout mice exhibit an exaggerated hypertrophic response to isoproterenol suggests that FHL2 plays an inhibitory role in hypertrophy. Isoproterenol infusion activates calcineurin [21], and inhibition of calcineurin blunts isoproterenol-induced cardiac hypertrophy [22-23].

Independent analyses examining FHL2 or calcineurin localization reveal that they are localized to the Z-disk and M-band of the sarcomere [24-25]. Based on this, I hypothesized that FHL2 inhibits isoproterenol-induced hypertrophy by inhibiting calcineurin-NFAT signaling. Here, I demonstrate that isoproterenol stimulates

FHL2 expression in the heart. Inhibition of FHL2 results in increased activation

32 of calcineurin/NFAT target genes in vivo and in vitro. Overexpression of FHL2 in vitro reduces NFAT-dependent gene expression and hypertrophic growth of myocytes induced by calcineurin. FHL2 and calcineurin colocalize to the sarcomere in adult cardiomyocytes. Interaction between endogenous FHL2 and calcineurin was observed upon activation of calcineurin with ET-1. Thus, I propose that the inhibitory role of FHL2 in isoproterenol-induced hypertrophy is accomplished, at least partly, through an interaction of FHL2 with calcineurin leading to inhibition of calcineurin/NFAT signaling.

Results

Differential roles of FHL2 in various models of hypertrophy

Although FHL2 is highly expressed in both the neonatal and adult heart,

knockout mice develop normally and do not manifest an overt phenotype.

However, when hypertrophy was induced with isoproterenol treatment, FHL2-/-

mice developed exaggerated hypertrophic growth compared to wild-type mice

[19]. To explore further the effect of FHL2 in hypertrophic growth, wild type and

FHL2-/- mice were subjected to three different hypertrophic models: isoproterenol, pressure overload, and constitutively active calcineurin.

β-adrenergic agonist induced hypertrophy

Isoproterenol-induced hypertrophy is a neurohumoral response and does not require invasive surgery. Isoproterenol is introduced chronically via use of a minipump inserted subcutaneously, which continuously releases isoproterenol.

Isoproterenol works through activation of β-adrenergic G-coupled receptors, which activate adenylate cyclase. Increased levels of cyclic-AMP trigger activation of PKA. Stimulation of β-adrenergic receptors leads to PKA- dependent phosphorylation of multiple intracellular targets including the L-type

Ca2+ channel, the ryanodine receptor (RyR2), and phospholamban.

First, I tested whether exaggerated hypertrophic growth in FHL2-/- mice

in response to isoproterenol treatment could be replicated. 9 week-old wild-type

(WT) mice were treated with 32mg/Kg/day body weight of isoproterenol or saline

(diluent control) by minipump infusion over 7 days. After the 7th day of treatment,

mice were sacrificed and the extent of the hypertrophic response was measured as heart weight to body weight ratios. Heart masses in saline-treated control WT and

FHL2-/- mice showed a similar proportion, confirming that FHL2-/- mice have no hypertrophic phenotype under basal conditions. Isoproterenol infusion induced

robust cardiac hypertrophy in both WT and FHL2-/- mice. In agreement with

earlier studies, FHL2-/- mice showed an increase in hypertrophic growth relative to WT, with an increase in the heart weight to body weight ratio of 39% (p<0.01)

and 27% (p<0.01), respectively (Figure 2.1A). These results confirm that

FHL2-/- hearts exhibit higher hypertrophic growth in response to isoproterenol

treatment.

Pressure overload-induced hypertrophy

Thoracic Aortic Constriction (TAC) is a model for pressure overload-

induced hypertrophy. Specifically, TAC is a model of left ventricular

hypertrophy in response to hemodynamic stress, where the thoracic aorta is

constricted resulting in increased afterload and wall stress in the left ventricle. 3 weeks of TAC is sufficient to induce cardiac hypertrophy with a 50% increase in

heart weight/body weight ratio compared to sham-treated animals.

Next, pressure overload-induced hypertrophy was analyzed in FHL2-/- mice. WT and FHL2-/- mice underwent TAC surgery, and a control group underwent open chest surgery but without aortic constriction. Mice were sacrificed at the end of 3 weeks, and hypertrophic growth was analyzed gravimetrically. Sham-treated WT and FHL2-/- mice were similar. TAC surgery induced robust hypertrophy in both groups with 38% (p<0.01) and 35% (p<0.01) increases in WT and FHL2-/- mice, respectively (Figure 2.1B). Thus, TAC- induced hypertrophy is preserved and not altered by inactivation of FHL2.

Constitutively active calcineurin-induced hypertrophy

Next, I tested whether loss of FHL2 could alter the hypertrophic response in transgenic mice expressing constitutively active calcineurin (CnA, lacking the autoinhibitory domain) under the heart-specific α-MHC promoter. CnA transgenic mice develop cardiac hypertrophy within 2 months, which then rapidly progresses to heart failure.

CnA transgenic mice were crossed with FHL2-/- mice to analyze hypertrophic growth in WT and FHL2-/- mice induced by constitutively active calcineurin. 9 week old control and calcineurin transgenic mice were sacrificed, and hypertrophy was measured by heart weight to body weight ratio. Consistent with previous results, WT and FHL2-/- were indistinguishable. Constitutively active calcineurin induced robust hypertrophic growth in both WT and FHL2-/-

36 backgrounds. Increases in HW/BW ratio were 139% and 144% in WT and

FHL2-/- respectively (Figure 2.1C). Thus FHL2 deficiency did not result in increased hypertrophy triggered by constitutively active calcineurin in vivo.

These data collected so far demonstrate that FHL2 can suppress hypertrophy induced by isoproterenol, but not by pressure overload or constitutively active calcineurin. This could be due to nature and amplitude of activation of signaling pathways in these models. Isoproterenol activates downstream signaling through

β-adrenergic receptors, whereas signaling in the TAC model is triggered by increased wall stress which activates stretch sensors at the cell membrane and sarcomeres. In addition, TAC is associated with a systemic response involving circulating neurohumoral elements and more. In the calcineurin the transgenic model, constitutively active calcineurin is expressed under a powerful cardiac promoter, which results in tremendous hypertrophy followed rapidly by heart failure. It is possible, even likely, that inhibition of calcineurin by FHL2 in this model is not sufficient to impact the overly activated hypertrophic growth.

Increased FHL2 expression in WT mice treated with isoproterenol

Inactivation of the FHL2 gene leads to an exaggerated hypertrophic response to beta-adrenergic stimulation, indicates that FHL2 is involved in modifying this response. We analyzed whether the expression of FHL2 is regulated during isoproterenol-induced hypertrophy. 9 week-old wild type mice

37 were treated with 32mg/Kg/day of isoproterenol or saline (diluent control) by minipump infusion over 7 days. After the 7th day of treatment, ventricular tissue

was collected for both protein and RNA isolation. Western blot analysis using an

antibody specific to FHL2 showed a nearly two-fold increase in the abundance of

FHL2 protein after isoproterenol treatment over the saline control (Figure

2.2A,B). Real time PCR analysis of FHL2 mRNA levels showed a similar result

with the isoproterenol treated group displaying consistently higher levels of FHL2

expression (Figure 2.2C). These data show that FHL2 is up-regulated during

isoproterenol-induced hypertrophy and suggest that this up-regulation may play a

role in modifying the hypertrophic response.

I also measured the expression of FHL2 in other hypertrophy models. 8

week old mice underwent TAC or sham surgery, and were monitored daily. After

one week of surgery, mice were sacrificed and hearts were collected. TAC-

induced hypertrophy was confirmed by an increase in heart weight to body weight

ratio. Lysates were prepared from the same part of the hearts and analyzed by

western blot. No change in FHL2 levels was observed between sham and TAC

animals (Figure 2.2D). Thus, not all hypertrophic stimuli induce FHL2

expression, and FHL2 expression is not dependent on hypertrophic growth.

Next, I tested the expression of FHL2 in a calcineurin-dependent model of

hypertrophy. Mice expressing constitutively active calcineurin and wild type

siblings were sacrificed at 9 weeks of age. Calcineurin transgenic mice developed

38 massive hypertrophic growth compared to wild-type mice. Analysis of FHL2 mRNA levels by real time PCR revealed that FHL2 expression was unchanged in calcineurin transgenic mice (Figure 2.2E). As a control, I tested RCAN expression, which was dramatically increased in the transgenic background

(Figure 2.2F). Thus, constitutive activation of calcineurin was not sufficient to induce increases in FHL2 expression. This shows that calcineurin activation does not induce FHL2 expression. This further demonstrates that hypertrophic growth per se is not a trigger for FHL2 expression.

In addition, expression of FHL2 in response to hypertrophic agonists was also evaluated in vitro using cultured neonatal rat ventricular myocytes. After 48 hours in 1% FBS media, cells were treated with either AngII (angiotensin II) or

PE (phenylephrine). 24 hours post-treatment, cells were lysed and analyzed by western blot. FHL2 levels were upregulated in samples treated with AngII or PE

(Figure 2.2G). Elevated ANF expression confirmed the hypertrophic growth response in AngII and PE treated cells. Thus, FHL2 expression can be induced by adrenergic agonists but not by pressure overload-induced hypertrophy or calcineurin/NFAT signaling.

Activation NFAT target genes in FHL2-/- mouse hearts.

To determine whether the amplified growth response in the absence of

FHL2 is modulated through the calcineurin/NFAT pathway, the transcription

39 levels of two NFAT target genes RCAN1.4 and BNP were examined. Real time

PCR showed that without isoproterenol treatment, FHL2-/- mice have no change in the mRNA levels of RCAN1.4 or BNP, as compared to the WT mice. In contrast, isoproterenol treatment resulted in a significant increase in both

RCAN1.4 (Figure 2.3A) and BNP (Figure 2.3B) expression in FHL2-/- mice as compared to the wild-type control. Importantly, the expression of the ubiquitous isoform RCAN1.1 was not significantly altered in the isoproterenol treatment group compared to control in either genotype (Figure 2.3C). These data suggest that amplified cardiac hypertrophy in FHL2-/- mice could be due to increased activation of calcineurin/NFAT signaling.

As increased expression of NFAT target genes was observed, I hypothesized that in the calcineurin transgenic model, expression of these genes will be elevated in an FHL2-/- background. Analysis of expression of RCAN1.4 promoter showed that the expression is similar in FHL2-/- calcineurin transgenic hearts (Figure 2.4A). However, expression of BNP was somewhat higher in

FHL2-/- mice (Figure 2.4B). In conclusion, loss of FHL2 in the background of overexpressed constitutively active calcineurin has little or no effect on NFAT target gene expression. This, in turn, suggests that although FHL2 can repress endogenous activation of calcineurin in isoproterenol-induced hypertrophy, its effect is limited in constitutively active calcineurin transgene-induced hypertrophy.

Increased expression of NFAT target genes in FHL2-/- mice.

While FHL2 has little effect on CnA transgene-induced hypertrophy, I hypothesized that it may affect NFAT activity at more physiological levels. Our collaborator, Dr Beverly Rothermel, has recently demonstrated that activation of the calcineurin/NFAT pathway follows a circadian rhythm in the heart, so that highest activation of the NFAT target promoter, RCAN1.4, is in the morning and the lowest in the evening. This led me to hypothesize that in FHL2-/- mice activation of RCAN1.4 and BNP will be higher than in WT in the morning, when

RCAN1.4 activation is at its highest. 12 week old WT and FHL2-/- mice were sacrificed between 7 and 8 am in the morning, and left ventricles were collected for gene expression analysis. Heart weight to body weight measurements were similar in WT and FHL2-/- mice (Figure 2.5A). Western blot analysis showed that expression of RCAN1.4 was higher in FHL2 -/- mice compared to WT

(Figure 2.5B,C). However, no change was observed in RCAN1.1 isoform

(Figure 2.5B,D), which is not regulated by NFAT. To further confirm increased activation of these genes, we analyzed mRNA expression of RCAN1.4 and BNP.

Compared to WT, expression of RCAN1.4 (Figure 2.5E) and BNP (Figure 2.5F) was 2.1 and 2.6 fold higher in FHL2-/- mice, respectively. Expression of the

RCAN1.1 isoform was similar in both genotypes (Figure 2.5G). Although expression of calcineurin/NFAT target genes was increased in FHL2-/- mice, no

41 hypertrophy was detected by HW/BW ratio. β-MHC mRNA levels were similar in both genotypes (Figure 2.5H), whereas ANF was slightly increased in FHL2-/- mice (Figure 2.5I), although it was not statistically significant.

Although FHL2-/- mice have higher activation of calcineurin in the morning, it is not sufficient to induce cardiac hypertrophy in these younger mice.

Next, I wanted to test whether older mice would be more susceptible to hypertrophy by higher basal activation of calcineurin in FHL2-/- mice. I have collected 1-year old WT and FHL2-/- mice and analyzed HW/BW ratio. 1-year old FHL2-/- mice have a modest 17% increase in HW/BW ratio as compared to

WT (Figure 2.6). Thus, while loss of FHL2 does not have an effect in young mice, long term effects, perhaps due to chronically increased calcineurin activation or stresses that accumulate with older age, lead to increased heart weight in aged FHL2-/- mice. Taken together, these data support a model where

FHL2 exerts tonic suppression on pro-growth pathways in the myocardium.

Knockdown of FHL2 protein in vitro enhances calcineurin-dependent NFAT promoter activation.

The in vivo data suggest that loss of FHL2 protein results in increased calcineurin/NFAT signaling in response to hypertrophic stimuli. To test this directly, we employed a luciferase reporter construct driven by the RCAN1 promoter, which contains an NFAT binding site and is activated by calcineurin.

Promoter activity was measured as a ratio of luciferase activity to an internal control (β-galactosidase), with the control extract set to 1. Neonatal rat ventricular cardiomyocytes (NRVM) were infected with an adenoviral vector expressing constitutively active calcineurin (lacking the C-terminal auto-inhibitory domain) to activate the NFAT pathway. FHL2 steady-state protein levels were decreased by siRNA knockdown with oligos targeted to the FHL2 mRNA, as shown in

Figure 2.7C (two independent targeting sequences were used to confirm specificity). While control cells showed around a 3-fold increase in calcineurin- stimulated RCAN1 reporter activity, simultaneous knockdown of FHL2 nearly doubled that response (Figure 2.7A).

I next set out to determine whether knockdown of FHL2 levels had a similar ability to enhance the effect of endogenous calcineurin on NFAT promoter activation. To do this, I activated endogenous calcineurin by treating cells with the calcium ionophore, ionomycin. FHL2 levels were decreased by siRNA treatment for 48 hours followed by treatment of the cells with ionomycin for 4 hours. Promoter activity was measured as above. Ionomycin treatment activated the RCAN1 promoter approximately 4 fold in control cells, and decreasing FHL2 levels in these cells resulted in a significant increase in the responsiveness of the

RCAN1 promoter to stimulation by ionomycin (Figure 2.7B). We have also analyzed the effect of FHL2 on endogenous expression of RCAN1.4. Western blot analysis demonstrated that cells transfected with FHL2 siRNA have higher

43 expression of RCAN1.4 when cells are stimulated with ionomycin compared to cells transfected with control siRNA (Figure 2.7D). These data show that decreasing FHL2 levels can increase the responsiveness of a cell to calcineurin/NFAT signaling and point to a mechanism for the effect of FHL2 in

ß-adrenergic receptor agonist induced hypertrophy.

FHL2 levels in vitro modulate activation of NFAT target genes by calcineurin

Our in vivo and in vitro data suggest the steady-state levels of FHL2 protein may play a role in modulating calcineurin/NFAT signaling and blunting the response to hypertrophic stimuli. To test this possibility further, we used

HEK (human embryonic kidney) cells in which FHL2 levels were modulated, at first by siRNA knockdown of FHL2 mRNA, and then by over-expression using a

CMV-driven FHL2 expression construct. Promoter activity was measured as a ratio of luciferase activity to an internal control (Renilla luciferase) and the control extract set to 1. FHL2 levels were decreased by siRNA treatment for 48 hours, or increased by transfection of the FHL2 construct for 24 hours, followed by treatment of the cells with ionomycin for 4 hours. Ionomycin treatment activated the RCAN1 promoter by 40% in control cells (Figure 2.8A). As observed in the NRVMs, decreasing FHL2 levels in HEK cells resulted in a substantial increase in the responsiveness of the RCAN1 promoter to stimulation by ionomycin (Figure 2.8A). In contrast, titration of the FHL2 expression

44 plasmid resulted in increasing levels of FHL2 protein and decreasing levels of ionomycin induced RCAN1 promoter activity in a dose-responsive manner

(Figure 2.8B). Additionally, over-expression of FHL2 also decreased the ability of the constitutively active calcineurin to activate expression from the RCAN promoter, confirming that the inhibitory effect of FHL2 on NFAT promoter activation is acting downstream of calcineurin activation (Figure 2.8C). The siRNA and overexpression results were confirmed using an IL-2 promoter construct as an alternative in vitro NFAT target (Figure 2.8D). These data show that FHL2 levels can modulate the responsiveness of a cell to calcineurin/NFAT signaling and point to a physiologic role for the observed increase in FHL2 expression in response to ß-agonist-induced hypertrophy.

FHL2 inhibits calcineurin-dependent hypertrophic responses in neonatal rat ventricular myocytes

It has been demonstrated that constitutively active calcineurin is sufficient to induce a hypertrophic response in neonatal rat ventricular myocytes. Our data demonstrate that overexpression of FHL2 leads to reduced activation of NFAT- dependent promoters, suggesting that increased levels of FHL2 might inhibit calcineurin/NFAT-dependent gene expression and subsequent hypertrophic growth. To test this, I used adenovirus infection with vectors expressing constitutively active calcineurin and either FHL2 or a control protein (ß-gal).

Cells were infected and hypertrophic growth assessed 48 hours post infection through the measurement of cross-sectional area and fetal gene expression. As anticipated, FHL2 significantly inhibited hypertrophic growth induced by calcineurin as determined by cell cross-sectional area measurements (Figure

2.9A,B). In addition, expression of hypertrophic genes ANF and BNP, elevated in calcineurin infected cells, was blunted by FHL2 over-expression (Figure

2.9C,D), consistent with the blunting of cell growth. These results support a model in which FHL2 levels blunt hypertrophy through the suppression of the calcineurin/NFAT signaling pathway.

FHL2 localizes to the sarcomere in adult and neonatal cardiomyocytes

I analyzed the localization of FHL2 in primary cardiomyocytes. GFP- fused FHL2 was transfected into cardiomyocytes and analyzed by fluorescence microscope. Control cells, transfected with GFP alone, showed diffuse GFP signal throughout the cell. However, the GFP signal in GFP-FHL2 transfected cells was organized in a clear sarcomeric pattern (Figure 2.10A), and in some cells a nuclear signal was present (Figure 2.10A, left). Control GFP transfected cells showed diffuse nuclear and cytoplasmic localization. This suggests that, in cardiac cells, FHL2 localizes mainly to the sarcomere, but in some cells it is also targeted to the nucleus.

Localization of endogenous FHL2 was investigated in adult heart tissue.

Four chamber heart sections were prepared from wild type and FHL2 knockout mice. Immunostaining with FHL2 antibody confirmed sarcomeric localization of

FHL2 as was observed in neonatal cardiomyocytes (Figure 2.10B, center). No signal was detected in FHL2 knockout mouse hearts (Figure 2.10B, right), validating the specific staining of FHL2 in wild-type hearts. No nuclear localization was observed in the adult heart (Figure 2.10B, left). Next, I hypothesized that FHL2 might translocate into the nucleus under conditions of stress. Wild-type mice were subjected to sham or TAC operation, and hearts were fixed and collected for analysis at the end of day 7 post-operation. Staining for

FHL2 revealed no difference between sham and TAC-subjected hearts, and no nuclear localization was detected in either group (Figure 2.10C). This suggests that FHL2 does not localize to the nucleus in TAC-induced hypertrophy. As I have observed augmented hypertrophy in FHL2 knockout mice in response to isoproterenol, it would be important of analyze whether isoproterenol induces nuclear localization of FHL2 in adult cardiomyocytes.

FHL2 co-localizes and interacts with calcineurin

Our results suggest a model in which FHL2 acts downstream of calcineurin activation to inhibit NFAT target gene expression. To examine whether this inhibition could be through a direct interaction with calcineurin, we

47 first looked to see if their localization would display similar patterns. I demonstrated that FHL2 localizes at or in close proximity to the Z-disks. In recent years it has been revealed that Z-disks harbor a wide array of signaling molecules, including calcineurin. To confirm that FHL2 and calcineurin co-localize to the same region of the sarcomere, adult heart tissues were co-immunostained with

FHL2 and calcineurin antibody. Figure 2.11A shows that FHL2 and calcineurin co-localize to the same region in adult cardiomyocytes.

LIM domains have been implicated in protein-protein interactions. Muscle

LIM protein MLP has been shown to interact with calcineurin. To test whether

FHL2 can also interact with calcineurin, we first expressed either full length or constitutively active HA-tagged calcineurin along with GFP-FHL2 in HEK cells.

Figure 2.11B shows that FHL2 co-precipitates more efficiently with active calcineurin than inactive full length calcineurin.

These data suggest that FHL2 might inhibit NFAT gene activation through a direct interaction with activated calcineurin. To test this model, co- immunoprecipitations of endogenous proteins were performed from neonatal rat cardiomyocytes in which calcineurin was activated. Cells were either treated or untreated with ET-1 for 1 hour, which has been shown to increase intracellular calcium levels and activate calcineurin. In untreated cells, the calcineurin antibody failed to pull down detectable amounts of FHL2 (Figure 2.11C,D); however, under conditions where calcineurin is activated, calcineurin and FHL2

48 can be readily co-immunoprecipitated. As a control we pulled down FHL2 with an antibody to sphingosine kinase 1 (SK1). FHL2 has been previously shown to interact with SK1 and this interaction is disrupted by treatment with endothelin-1

(Figure 2.11D). These data suggest that the ability of FHL2 to modulate NFAT signaling acts through the formation of a protein-protein complex between FHL2 and active calcineurin.

Interpretation and Conclusion

FHL2 null mice are viable with no initial overt phenotype; however they display an exaggerated hypertrophic response to isoproterenol and aging. We have assessed isoproterenol treatment of FHL2-/- mice and observed similar results as reported previously [19]. Although, we used younger mice, 2 months old and treated for 7 days, versus 7-8 months and 10 days of treatment, I observed a significant increase in heart mass in FHL2-/- mice compared to WT mice in the isoproterenol treatment group; hypertrophic growth, as measured by heart weight to body weight ratio, was increased by 44% in FHL2-/- mice (Figure 2.1A). No changes were observed between FHL2-/- and WT mice in the control group.

Surprisingly, the hypertrophic response to TAC was indistinguishable in

WT and FHL2-/- mice. TAC is a surgical procedure where the thoracic aorta of mice is constricted at a certain level, any variation of which may result either weaker or tighter constriction. Thus, the TAC model may not be sensitive enough to detect small changes in hypertrophic growth. Although FHL2-/- mice had slightly higher HW/BW ratios, the difference was not significant due to high variations in both groups (Figure 2.1B). Similar results were reported by Chu et al [20] in similar FHL2 knockout mice; FHL2 deficiency had no effect on the

TAC procedure. Since a higher difference of hypertrophy was observed in older mice treated with isoproterenol compared to younger groups used in my experiments, we have attempted to band 6 months old mice to test whether older

FHL2-/- mice will demonstrate higher hypertrophic growth in response to TAC.

However, since the TAC procedure in our lab was optimized to 2 months old mice, banding 6 months old mice led to inconsistent results and a more severe phenotype. The possible reasons for this discrepancy between isoproterenol and

TAC-induced hypertrophy in FHL2 knockout mice will be discussed in chapter 3.

Perhaps not surprisingly, no change was observed when FHL2-/- mice were crossed to calcineurin transgenic mice. These mice express constitutively active calcineurin under a potent cardiac-specific promoter. These mice manifest tremendous hypertrophy which rapidly progresses to heart failure by 2 months of age. I have analyzed mice at 2 months; at this stage they were already in heart failure stage with HW/BW ratio of 11 (Figure 2.1C). At that point there were no differences between the genotypes, but perhaps an earlier time point analysis would uncover a role for FHL2 in hypertrophic growth in the calcineurin transgenic model. Analysis of the expression of NFAT target genes showed variability in some FHL2-/- mice compared to WT (data not shown), but changes were minor and did not result in an increased HW/BW ratio. This could be explained by the already high activation of calcineurin resulting in a maximum hypertrophic response, and increasing it further does not have any additional effect. This also supports our model in which FHL2 functions through calcineurin and not via a parallel pathway, which may have resulted in an increased effect.

Expression of NFAT target genes in the isoproterenol model was consistent with the hypertrophic phenotype observed in WT and FHL2-/- mice.

The more severe hypertrophic phenotype was consistent with a higher activation of the calcineurin/NFAT target genes RCAN1.4 and BNP in FHL2-/- mouse hearts. As these data are only correlative, where increased calcineurin might cause increased hypertrophy or vice-versa, I looked at endogenous activation of calcineurin in the absence of hypertrophic stimuli. Activation of calcineurin manifests circadian rhythmicity, with highest activation in the morning and lowest in the evening. WT and FHL2-/- hearts collected at 7 AM had the same HW/BW ratio. Expression of RCAN1.4 was increased in FHL2-/- mice both at the mRNA and protein levels (Figure 2.5B,E). BNP expression also showed an increased expression in FHL2-/- mice in the AM compared to WT (Figure 2.5F). This suggests that FHL2 deficiency results in higher activation of calcineurin/NFAT pathway even in the absence of hypertrophic stimuli, which in the long run induces modest cardiac hypertrophy in FHL2-/- mice (Figure 2.6).

In vivo results were further tested in vitro in neonatal rat ventricular myocytes. Two independent FHL2 siRNAs increased activation of the RCAN promoter by calcineurin. The same results were observed when endogenous calcineurin was activated with ionomycin, an effect that was completely blocked by calcineurin inhibitor CsA. Ionomycin treatment was carried out for only four hours, sufficient to trigger robust expression and avoid possible secondary effects

52 by calcineurin activation or ionomycin toxicity. NRVM cells are capable of developing significant hypertrophy in response to hypertrophic stimuli, of which overexpression of constitutively active calcineurin is one example. Hence, to avoid possible effects of hypertrophy itseft on analyzed promoters, I repeated these experiments in HEK 293 cells. Knockdown of FHL2 resulted in increased activation of calcineurin, whereas overexpression of FHL2 attenuated activation

RCAN1 and IL-2 promoters by calcineurin (Figure 2.8 C,D).

Transduction of neonatal myocytes with active calcineurin induced significant hypertrophic growth. Consistent with inhibition of NFAT-dependent promoters, overexpression of FHL2 also suppressed calcineurin-induced hypertrophy, as determined by cell cross-sectional area and expression of hypertrophic markers. Previously, it has been demonstrated that FHL2 also inhibits hypertrophy induced by PE and active MEK1. Active MEK-1 induced robust cell growth, of which only 20% was inhibited by FHL2. PE treatment induced moderate hypertrophy which was completely suppressed by FHL2.

Interestingly, inhibition of MEK/ERK signaling with dominant negative MEK1 suppressed only 50% of PE-induced hypertrophy. This implies that FHL2 inhibits pathways other than MEK/ERK during PE-induced hypertrophy. Thus, it is possible that complete inhibition of hypertrophic growth by FHL2 is accomplished by suppression of both calcineurin/NFAT and MEK/ERK pathways, rather than MEK/ERK alone.

Inhibition of calcineurin and MEK/ERK pathways will be further discussed in chapter 3.

Localization to sarcomere

I have shown that FHL2 localizes to the sarcomere in adult and neonatal cardiomyocytes. It also localizes to the nucleus in neonatal cells, however no nuclear localization was observed in adult cells. It was reported that FHL2 localizes to the Z-disk and M-band, which is accomplished by association with

N2B and is2 regions of titin [24]. However, my data show that most of the FHL2 signal localized to the Z-disks, and very low or undetectable levels to the M-band.

Z-disks of cardiomyocytes are hubs for signaling pathways, and many regulatory proteins have been shown to localize to Z-disks, including proteins involved in hypertrophy such as calcineurin, ERK, MLP, PKC and PKA. In addition, the evasive cardiac stretch sensing protein/complex is believed to localize at the Z- disk, which activates hypertrophic pathways in pressure overload models.

Consistent with the function of LIM domain, FHL2 may act as a scaffolding protein localized to Z-disks. However, FHL2 does not appear to contribute to mechanosensing, as FHL2 knockout mice develop comparable hypertrophy in response to pressure overload.

FHL2 interacts with calcineurin

I also demonstrate that FHL2 interacts with calcineurin. Calcineurin localizes in a similar pattern to FHL2 in myocytes at the Z-disk and M-band.

Calcineurin has been shown to interact with MLP (muscle LIM protein) [26], which consist of 2 LIM domains. Overexpression of FHL2 and calcineurin showed that both active mutant and wild-type calcineurin interact with FHL2.

However, under native conditions this interaction requires activation of calcineurin, in this case with ET-1.

In conclusion, I have demonstrated that FHL2 inhibits calcineurin in vivo and in vitro. FHL2 deficiency leads to increased calcineurin/NFAT gene expression and increased hypertrophic growth, whereas overexpression of FHL2 suppresses NFAT target gene expression and hypertrophic growth induced by calcineurin.

Figure 2.1 Absence of FHL2 amplifies isoproterenol-induced cardiac hhypertrophy. A) Cardiac hypertrophy induced by isoproterenol. FHL2 wild type (WT) and knockout (FHL2-/-) mice were subjected to chronic isoproterenol (28 mg/mL per 25g body weight) infusion using minipumps. Mice were sacrificed after 7 days and hypertrophic growth was analyzed by measuring heart weight to body weight ratio (HW/BW). B) Cardiac hypertrophy induced by pressure overload. HW/BW ratio was measured in WT and FHL2-/- mice subjected to sham or thoracic aortic constriction (TAC) surgery for 3 weeks. C) Cardiac hypertrophy induced by calcineurin. Hypertrophy induced by constitutive active calcineurin driven by α-MHC promoter was analyzed in WT and FHL2-/- mice were sacrificed and analyzed at 2 months of age. Note that HW/BW ratio in WT

56 mice was significantly increased in FHL2-/- mice following isoproterenol infusion but not pressure overload or calcineurin transgenic mice. * p<0.05 vs WT, # p<0.05 vs WT isoproterenol.

Figure 2.2 Increased expression of FHL2 in isoproterenol-treated mouse hearts. A,B) Western blot(A) and quantification (B) for FHL2 and GAPDH ( loading control) in samples prepared from control and isoproterenol-treated hearts. C) FHL2 transcript levels in control and isoproterenol treated hearts were measured using qRT-PCR and normalized to cyclophylin mRNA. n=3, * p<0.05 vs control. Note that FHL2 expression was increased ~2 fold is isoproterenol- treated hearts. D) Western blot of FHL2 and GAPDH in samples prepared from hearts of mice that underwent sham or TAC surgery. Note that no significant change was detected in FHL2 levels following TAC.

Figure 2.2 Increased expression of FHL2 in hearts of isoproterenol-treated mice. Quantitative RT-PCR analyses of FHL2 (E) and RCAN1.4 (F), a target of calcineurin/NFAT pathway, expression in WT and calcineurin transgenic mice (CnA Tg) at 2 months of age. Note the induction of RCAN1.4 but not FHL2 in calcineurin transgenic mice. * p<0.01 vs WT. G) Western blot analysis shows increased expression of FHL2 in neonatal cardiomyocytes treated with hypertrophic agonist, angiotensin II (AngII) and phenylephrine (PE) for 24 hours.

Figure 2.3 Increased expression of NFAT target genes in FHL2 knockout mice followingn isoproterenol stimulation. WT and FHL2-/- mice were treated with control or isoproterenol for 7 days. Total RNA, isolated from WT and FHL2- /- mouse hearts, was used to prepare cDNA and to quantify the expression of RCAN1.4 (A), BNP (B), and RCAN1.1 (C), * p<0.05 vs WT, # p<0.05 vs WT isoproterenol. Note the increased induction of calcineurin/NFAT target genes in isoproterenol treated FHL2-/- hearts.

Figure 2.4 Induction NFAT tarrget gene BNP by constitutivelly active calcineurin in FHL2-/- hearts. A) Similar RCAN1.4 expression was observed in calcineurin transgenic WT and FHL2-/- mice. B) Increased BNP mRNA in some calcineurin transgenic FHL2-/- mice. * p<0.05 vs WT.

Figure 2.5 Absence of FHL2 enhances NFAT target gene expression in mouse hearts collected in the morning. A) HW/BW ratio was measured in WT and FHL2-/- mice collected at 7 am. Relatively similar HW/BW ratios were observed. B-D) Protein extracts were subjected to western blot analyses (B) of the indicated proteins and quantification of RCAN1.4 (C) and RCAN1.1 (D) are shown. Note that RCAN1.4 but not RCAN1.1 protein levels are increased. E-J) qRT-PCR analyses of the indicated gene expression revealed that only NFAT target genes RCAN1.4 (E) and BNP (F) but not RCAN1.1 (G) or hypertrophic markers β- MHC (H) and ANF (I) were upregulated in FHL2-/- hearts. * p<0.05 vs WT. Also note the absence of FHL2 message (J) in FHL2-/- hearts.

* *

Figure 2.6 Increased heart weight to body weight ratio in FHL2-/- mice at 1 year of age. WT and FHL2-/- mice were sacrifice at 12-14 months of age, and analyzed gravimetrically. * p<0.05 vs WT.

Figure 2.7 Knockdown of FHL2 augments NFAT-responsive promoter activity by calcineurin. A) Cotransfection of neonatal cardiomyocytes with FHL2 siRNA and constitutively active calcineurin shows increased activation of

RCAN promoter compared to control siRNA transfected cells, * p<0.05 vs GFP, # p<0.05 vs siControl+calcineurin. B) Increased activation of RCAN promoter by ionomycin treatment (1µM, 4 hours) in FHL2 siRNA transfected cells, which was completely blocked by Cyclosporin A (CsA), * p<0.05 vs siControl no treatment, # p<0.05 vs siControl+ionomycin. C) Western blot representing FHL2 knockdown in neonatal rat ventricular myocytes transfected with two FHL2 specific and a control siRNA. D) Increased expression of RCAN1.4 following ionomycin treatment in FHL2 siRNA transfected cells in neonatal cardiomyocytes.

* * # #

Figure 2.8 FHL2 modulates NFAT-dependent promoter activity by calcineurin. A) Reporter gene expression from RCAN1.4 promoter was analyzed in HEK 293 cells following transfection of control or FHL2 siRNA and ionomycin treatment. Ionomycin treatment (1µM, 4 hours) significantly amplified promoter activity in FHL2 siRNA transfected cells, * p<0.05 vs siControl no treatment, # p<0.05 vs siControl+ionomycin. B) FHL2 overexpression inhibits activation of RCAN promoter by ionomycin in a dose-dependent manner, p<0.01 vs Control, # p<0.05 vs Ionomycin. C,D) Overexpression of FHL2 inhibits CnA- dependent induction of RCAN1.4 (C) and IL-2 (D) promoter activities in HEK 293 cells, * p<0.01 vs Control, # p<0.05 vs Calcineurin.

Figure 2.9 FHL2 inhibits hypertrophic growth induced by calcineurin in neonatal rat ventricular myocytes (NRVMs). A) NRVMs infected with βgal (control) or FHL2, were co-infected with GFP or Calcineurin. Cells were fixed and immunostained with α-actinin antibody 48 hours post infection. B) Cell cross- sectional area of cells immunostained with α-actinin was measured, n~180 cells in each group. * p<0.01 vs βgal GFP, # p<0.01 vs βgal Calcineurin, Note that calcineurin-dependent increase in cell size was blunted by FHL2. C,D) Expression of hypertrophic markers ANF (C) and BNP (D) was quantified under indicated conditions using qRT-PCR, . ** p<0.05 vs βgal GFP, ## p<0.05 vs βgal+Calcineurin. Note the suppressed hypertrophic marker expression in FHL2 infected cells, consistent with hypertrophic growth. (Scale bar, 20 µm.)

Figure 2.10 FHL2 localizes to sarcomere in adult cardiomyocytes. A) Immunostaining of WT and FHL2-/- hearts with FHL2 antibody. WT transverse (left) and longitudinal (center) view of cardiomyocytes. No staining was detected

70 in control FHL2-/- mice (right). B) FHL2 does not localize to the nucleus under normal and pressure overload conditions. C) GFP-tagged FHL2 localized to sarcomere in neonatal cardiomyocytes. FHL2 is also targeted to the nucleus in some cells (left) or excluded from nucleus (right) in others. (Scale bar, 20 µm.)

Figure 2.11 FHL2 colocalizes and interacts with calcineurin. A) Co-staining of WT heart tissues with calcineurin and FHL2 antibodies. Overlay of each section indicates colocalization of FHL2 and CnA. B) Physical interaction of FHL2 and calcineurin. GFP-tagged FHL2 was transfected into HEK 293 cells alone or along with HA-tagged calcineurin. FHL2 immunoprecipitates with constitutively active and wild type calcineurin in HEK 293 cells. C,D) Interaction of endogenous FHL2 and calcineurin in ET-1 treated NRVMs. NRVMs were maintained in media containing 1% FBS serum for 48 hours. Cell lysates, prepared from ET-1- treated (100nm for 60 min) or untreated cells, were incubated with calcineurin (C,D) or SK1 (D) antibodies. Proteins were separated on SDS-PAGE gel and FHL2 antibody was used for immoprecipitation analyses. Note that calcineurin interacts with FHL2 only in ET-1-treated lysates. (Scale bar, 5 µm.)

Material and Methods

Chronic isoproterenol infusion

Avertin (anesthetic) was prepared in PBS with final concentration of 20mg/ml in

50 ml conical tube, rocked until completely dissolved, filtered and stored at 4C covered with aluminum foil. Male C57BL6 mice (9 week old) were injected with

0.5cc avertin for a final concentration of 0.4mg/g body weight. Mice were returned to cage until anesthetized, which was confirmed by inability to react to toe pinching. 1cm incision was made on the back of a mouse and Alzet minipumps containing 200ul of Isoproterenol (32mg/Kg/day) or saline were inserted subcutaneously. Skin was reattached with special staple. Mice were returned to animal facility and monitored daily. Mice were sacrificed at the end

of day 7. Body weight and heart weight were measured to analyze hypertrophic

phenotype. Heart were frozen in liquid nitrogen and stored at -80 C.

Pressure overload hypertrophy model

Male C57BL6 mice (8 week old) were subjected to pressure overload by thoracic

aortic banding (TAB). In this procedure, the constriction is placed in the

transverse aorta between the innominate and left common carotid arteries.

Constriction to a 27-gauge stenosis induces moderate hypertrophy (~40% increases in heart mass). Mice were returned to animal facility and monitored

74 daily. Body weight and heart weight were measured to analyze hypertrophic phenotype. Heart were frozen in liquid nitrogen and stored at -80 C.

Cell culture and transfections.

C2C12 skeletal myoblasts were maintained in DMEM media supplemented with

20% FBS. Cells were transfected at 30-50% confluency with Lipofectamine

(Invitrogen) and PLUS (Invitrogen) according to manufacturer’s protocol. HEK

293 cells were maintained in DMEM supplemented with 10% FBS. Cells were transfected at 60-70% confluency with Lipofectamine 2000 reagent (Invitrogen).

Cells were harvested after 24 hours and lysed and luciferase assay was performed according to manufacturers protocol (Promega). Cells were co-transfected with pCMV-LacZ of pRL-CVM (Promega), which was used to normalize for transfection efficiency. siRNA were transfected at 30% confluency and reporter plasmid were transfected at 24 hours and cells were harvested and analyzed at 48 hours. Negative control and FHL2 siRNA was obtained from Ambion.

Negative control #1, cat#: 4390844

FHL2-1, sense 5’-CAACGACUGCUUUAACUGUtt-3’ ,

antisense 5’-ACAGUUAAAGCAGUCGUUGtg -3’

FHL2-2, sense 5’-GCAAGGACUUGUCCUACAAtt -3’ ,

antisense 5’- UUGUAGGACAAGUCCUUGCtt-3’

Primary culture of neonatal rat ventricular myocytes.

Cardiomyocytes were isolated from ventricles of 1-2 day old rat pups (Sprague

Dawley) and plated in medium containing 10% fetal calf serum at density of 1250 cells/mm2 as described(Yan Ni paper). Myocyte cultures obtained from this

differential plating method contained less than 5% non-cardiomyocytes as

determined microscopically after immunostaining with myocyte-specific α-

actinin antibody. After 48 hours plating media was replaced by 1% FBS media at

which point cells were infected with adenoviral constructs expressing protein of

interest at 30 MOI (multiplicity of infection) concentration. 3 hours post infection,

media containing viruses were replaced with fresh 1% FBS media. Analyses were

performed at 24 or 48 hours after infection.

Immunocytochemistry

Cultured cardiomyocytes were washed with PBS 3 times and fixed with 4% PFA

for 10 minutes. Cells were permeabilized in 0.05% Triton-X for 15 minutes. Cells

were blocked in 3% normal goat serum and 1% bovine serum albumin prepared in

PBS for 30 minutes. α-actinin antibody(Sigma) was prepared 1/100 in blocking

solution and incubated for 1 hour at room temperature, followed by 1 hour of Cy-

5 conjugated secondary antibody. Pictures were acquired on Leica DM2000

microscope. Cross-sectional area was measured using ImageJ software.

Immunoprecipitation

Following 48 hours of 1% FBS media neonatal rat ventricular myocytes were treated with 1µM ionomycin or 100nm ET-1 for 1 hour. Cells were washed 3 times with PBS and harvested in NP-40 lysis buffer (50mM Tris-HCl,150mM

NaCl, 1% NP-40), supplemented with complete protease inhibitor (Roche). Cell debri was pelleted down and supernatant used for immunoprecipitation.

Calcineurin antibody (Stressgen, 1/100) antibody was pulled down with TrueBlot anti-rabbit beads (eBioscience). Western blot was performed using calcineurin

(1/1000) and FHL2 (MBL, 1/1000) antibodies and TrueBlot secondary antibodies

(1/2000).

RNA Purification and real-time PCR

RNA isolation from tissues and cells was performed using Trizol according to manufacturer’s protocol (Invitrogen) and concentration was measured on

Nanodrop-1000 (Thermo Scientific). cDNA was synthesized using high capacity cDNA reverse transcription kit (Applied Biosystems). Real-time PCR was performed using SYBR Green mix (Applied Biosystems) on ABI-PE Prism 7000 sequence detection system. The relative quantities of mRNA were determined and normalized to cyclophilin A or 18S.

Primers:

Table 1. Real time primers. Forward Reverse FHL2 TCACAGCACGGGATGAGTTTC GTGCCACCCAGACCACTAATG RCAN1.4 CCCGTGAAAAAGCAGAATGC TCCTTGTCATATGTTCTGAAGAGGG RCAN1.1 GACCCGCGCGTGTTC TGTCATATGTTCTGAAGAGGGATTC BNP CACCGCTGGGAGGTCACT GTGAGGCCTTGGTCCTTCAA ANP CTTCTTCCTCTTCCTGGCCT TTCATCGGTCTGCTCGCTCA β-MHC CTGACTGAACAGCTGGGCTC AACTCTGGAGGCTCTTCACT Cycloph. CAGACGCCACTGTCGCTTT TGTCTTTGGAACTTTGTCTGCAA A 18S AAACGGCTACCACATCCAAG CCTCCAATGGATCCTCGTTA

Immunoblot Analysis

Protein lysates were fractionated by SDS polyacrylamide gel electrophoresis,

transferred to supported nitro-cellulose membrane (Bio-Rad, 0.2µm) and

subjected to immunoblot analysis. The ratio of antigen:antibody, detergent

concentration, and duration and temperature of the reactions were optimized for

each antibody.

Statistical significance.

Averaged data are reported as means ± SEM. Statistical significance was analyzed

using a Student’s unpaired t test.

References

1. McMullen, J.R. and G.L. Jennings, Differences between pathological and physiological cardiac hypertrophy: novel therapeutic strategies to treat heart failure. Clin Exp Pharmacol Physiol, 2007. 34(4): p. 255-62.

2. Lloyd-Jones, D., et al., Heart disease and stroke statistics--2010 update: a report from the American Heart Association. Circulation, 2010. 121(7): p. e46-e215.

3. Wilkins, B.J. and J.D. Molkentin, Calcium-calcineurin signaling in the regulation of cardiac hypertrophy. Biochem Biophys Res Commun, 2004. 322(4): p. 1178-91.

4. Molkentin, J.D., et al., A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell, 1998. 93(2): p. 215-28.

5. Lim, H.W., et al., Reversal of cardiac hypertrophy in transgenic disease models by calcineurin inhibition. J Mol Cell Cardiol, 2000. 32(4): p. 697- 709.

6. Johannessen, M., et al., The multifunctional roles of the four-and-a-half- LIM only protein FHL2. Cell Mol Life Sci, 2006. 63(3): p. 268-84.

7. Labalette, C., et al., The LIM-only protein FHL2 regulates cyclin D1 expression and cell proliferation. J Biol Chem, 2008. 283(22): p. 15201-8.

8. Martin, B.T., et al., FHL2 regulates cell cycle-dependent and doxorubicin- induced p21Cip1/Waf1 expression in breast cancer cells. Cell Cycle, 2007. 6(14): p. 1779-88.

9. Scholl, F.A., et al., DRAL is a p53-responsive gene whose four and a half LIM domain protein product induces apoptosis. J Cell Biol, 2000. 151(3): p. 495-506.

10. Sun, J., et al., FHL2/SLIM3 decreases cardiomyocyte survival by inhibitory interaction with sphingosine kinase-1. Circ Res, 2006. 99(5): p. 468-76.

11. Han, W., et al., FHL2 interacts with and acts as a functional repressor of Id2 in human neuroblastoma cells. Nucleic Acids Res, 2009. 37(12): p. 3996-4009.

12. Lai, C.F., et al., Four and half lim protein 2 (FHL2) stimulates osteoblast differentiation. J Bone Miner Res, 2006. 21(1): p. 17-28.

13. Martin, B., et al., The LIM-only protein FHL2 interacts with beta-catenin and promotes differentiation of mouse myoblasts. J Cell Biol, 2002. 159(1): p. 113-22.

14. Wang, J., et al., Suppression of FHL2 expression induces cell differentiation and inhibits gastric and colon carcinogenesis. Gastroenterology, 2007. 132(3): p. 1066-76.

15. Park, J., et al., Deficiency in the LIM-only protein FHL2 impairs assembly of extracellular matrix proteins. FASEB J, 2008. 22(7): p. 2508-20.

16. Gunther, T., et al., Fhl2 deficiency results in osteopenia due to decreased activity of osteoblasts. EMBO J, 2005. 24(17): p. 3049-56.

17. Kirfel, J., et al., Impaired intestinal wound healing in Fhl2-deficient mice is due to disturbed collagen metabolism. Exp Cell Res, 2008. 314(20): p. 3684-91.

18. Wixler, V., et al., Deficiency in the LIM-only protein Fhl2 impairs skin wound healing. J Cell Biol, 2007. 177(1): p. 163-72.

19. Kong, Y., et al., Cardiac-specific LIM protein FHL2 modifies the hypertrophic response to beta-adrenergic stimulation. Circulation, 2001. 103(22): p. 2731-8.

20. Chu, P.H., et al., FHL2 (SLIM3) is not essential for cardiac development and function. Mol Cell Biol, 2000. 20(20): p. 7460-2.

21. Zou, Y., et al., Isoproterenol activates extracellular signal-regulated protein kinases in cardiomyocytes through calcineurin. Circulation, 2001. 104(1): p. 102-8.

22. De Windt, L.J., et al., Targeted inhibition of calcineurin attenuates cardiac hypertrophy in vivo. Proc Natl Acad Sci U S A, 2001. 98(6): p. 3322-7.

23. Rothermel, B.A., et al., Myocyte-enriched calcineurin-interacting protein, MCIP1, inhibits cardiac hypertrophy in vivo. Proc Natl Acad Sci U S A, 2001. 98(6): p. 3328-33.

24. Lange, S., et al., Subcellular targeting of metabolic enzymes to titin in heart muscle may be mediated by DRAL/FHL-2. J Cell Sci, 2002. 115(Pt 24): p. 4925-36.

25. Torgan, C.E. and M.P. Daniels, Calcineurin localization in skeletal muscle offers insights into potential new targets. J Histochem Cytochem, 2006. 54(1): p. 119-28.

26. Heineke, J., et al., Attenuation of cardiac remodeling after myocardial infarction by muscle LIM protein-calcineurin signaling at the sarcomeric Z-disc. Proc Natl Acad Sci U S A, 2005. 102(5): p. 1655-60.

CHAPTER 3

DISCUSSION

Increased FHL2 expression after isoproterenol infusion

FHL2 protein and mRNA are most abundant in the heart, with low but yet detectable levels in other tissues (reviewed [1]). FHL2 expression initiates during embryogenesis and remains high through out adulthood [2]. Analyses of FHL2 expression in the adult heart reveal that FHL2 levels are unchanged in pressure overload-induced hypertrophy [3]. FHL2 levels are elevated in hearts after cardiopulmonary bypass [4], whereas its protein abundance is decreased in failing human hearts [5]. Here, I demonstrate that FHL2 expression is increased in isoproterenol-induced hypertrophy at both the protein and mRNA levels (Figure

2.2A,C). There are several putative factor transcription binding sites identified within the FHL2 promoter, including NKX2.5, SRF, p53, MEF2 and AP-1.

However, only two of them have been confirmed thus far: SRF [6] and p53 [7].

The contribution of MEF2 and AP-1 to the expression of FHL2 remains unclear.

SRF, a member of the MADS family of transcriptional activators, has been shown to regulate the expression of cardiac-specific genes, including ANF, cardiac α-actin, α-MHC and β-MHC [8]. Mice overexpressing wild-type SRF develop hypertrophic cardiomyopathy with activation of fetal gene expression [9], whereas deletion of SRF in adult mouse hearts results in dilated cardiomyopathy with suppressed expression of contractile and structural proteins [10]. Recent studies have demonstrated that FHL2 expression is upregulated by SRF [6, 11].

Given this, I considered a connection between isoproterenol and SRF. However,

in one study, SRF levels were increased by isoproterenol stimulation [12],

whereas others have shown that isoproterenol inhibits expression from SRF-

dependent promoters [13]. Thus, it is not clear whether SRF is involved in

isoproterenol-induced expression of FHL2. It would be interesting to analyze the

expression of FHL2 in hearts of SRF transgenic mice, to test the role of SRF on

the FHL2 promoter in vivo. FHL2 expression could be analyzed in neonatal

cardiomyocytes transfected with SRF siRNA treated with adrenergic agonist. If

FHL2 expression is SRF dependent in cardiac cells, knockdown of SRF would blunt the increase in FHL2 following treatment with adrenergic agonists.

The tumor suppressor p53 was recently identified as a regulator of FHL2 expression [14]. Low expression of FHL2 was observed in a rhabdomyosarcoma cell line which expresses mutant p53. Overexpression of wild-type p53 in this cell line restored FHL2 expression. Similarly, γ-irradiation increased expression of

FHL2 in a p53-dependent manner [7]. It is unclear whether isoproterenol can induce p53 in cardiomyocytes. Thus, it is unclear whether p53 might play a role in upregulation of FHL2 expression after isoproterenol treatment.

FHL2 expression is turned on early in embryogenesis and expressed through adulthood [2], which coincides with the expression of NKX2.5, a homeodomain transcription factor [15] and the earliest known marker of vertebrate heart development. The FHL2 promoter contains multiple putative binding sites for the transcription factor NKX2.5 [1], which may be responsible

for the expression of FHL2 in early cardiac progenitor cells. Expression of

NKX2.5 is upregulated in adrenergic agonist-induced cardiac hypertrophy [16].

Moreover, transcriptional activity of NKX2.5 is augmented by SRF [17]. Thus

NKX2.5 and SRF have the potential to synergistically regulate FHL2 expression

in isoproterenol-treated hearts.

I have tested whether calcineurin/NFAT signaling might regulate FHL2

expression. FHL2 mRNA levels in calcineurin transgenic mice were compared to wild-type mice, and no significant difference was observed in FHL2 expression between the two genotypes (Figure 2.2E). I have analyzed RCAN1.4 expression as a control, which was over 17-fold higher in calcineurin transgenic mice compared to wild type (Figure 2.2F). These data strongly suggests that FHL2 is not directly regulated by the calcineurin/NFAT pathway.

Inhibition of calcineurin by FHL2

Calcineurin is a Ca2+-dependent serine/threonine protein phosphatase that

dephosphorylates NFAT family transcription factors, which in turn leads to their

nuclear translocation and activation of their target genes. Calcineurin is activated

in the majority of models of pathological hypertrophy. Constitutively active

calcineurin, lacking the autoinhibitory domain, is sufficient to induce

hypertrophic growth in vivo and in vitro [18]. Also, targeted inhibition of

calcineurin Aβ impairs hypertrophy induced by pressure overload, AngII, or isoproterenol infusion[19].

Here, I tested whether activation of calcineurin/NFAT signaling was higher in FHL2-/- mice. RCAN1.4 and BNP mRNA were analyzed as readouts of calcineurin/NFAT pathway activation, because they are direct targets of NFAT

[18, 20-21], and both are activated in hypertrophy [22-23]. Indeed, activation of

RCAN1.4 and BNP was increased in isoproterenol-treated FHL2-/- mice compared to WT in the same treatment group (Figure 2.3A,B). Elevated expression of these genes was also observed with endogenous calcineurin in vivo

(Figure 2.5C,E,F). Conversely, the expression of the calcineurin-independent

RCAN1.1 isoform was not changed (Figures 2.3C and 2.5D,G). Knockdown of

FHL2 resulted in augmented activation of the RCAN1.4 promoter by calcineurin and ionomycin in vitro (Figures 2.7A,B,D and 2.8A). Increased ionomycin- dependent activation of the RCAN promoter in FHL2 knockdown cells was completely inhibited by CsA (Figure 2.7B), confirming calcineurin dependence.

On the other hand, overexpression of FHL2 inhibited activation of the RCAN1.4 promoter (Figure 2.8B,C,D). Moreover, FHL2 blunted the hypertrophic growth induced by constitutively active calcineurin in neonatal cardiomyocytes (Figure

2.9). These data support a model in which FHL2 inhibits activation of NFAT target genes.

FHL2 has been demonstrated to play a role in cardiomyocyte apoptosis in vitro [7]. Knockdown of FHL2 protected cells from H2O2-induced apoptosis, and re-expression of FHL2 reversed this effect [24]. ET-1 is a potent survival factor against apoptosis in various cell lines including cardiomyocytes, which involves activation of calcineurin/NFAT pathway. Inhibition of calcineurin, with CsA or

FK506, abolishes the protective effect of ET-1 in cardiomyocytes [25]. Inhibition of calcineurin/NFAT signaling by FHL2 would explain the pro-apoptotic effect of

FHL2 in cardiomyocytes, where the knockdown of FHL2 would result in improved survival through increased activation of calcineurin/NFAT signaling.

The impact of the apoptotic effect of FHL2 will need to be studied in vivo in

FHL2 knockout mice under conditions of severe stress, which trigger apoptotic cardiomyocyte death, such as ischemia.

I have shown that FHL2 and calcineurin colocalize and form a complex with each other. In order to identify a possible mechanism of inhibition of the calcineurin/NFAT pathway by FHL2, we have examined to the regulation of this pathway by endogenous inhibitors. Some of the proteins that have been identified as calcineurin inhibitors are RCAN1.4, cyclophilin (complexed with CsA),

FKBP12 (complexed with FK506), AKAP79, cain, Bcl-2 and calsarcin.

Interaction of NFAT with calcineurin depends on 2 conserved sites of NFAT,

PxIxIT and LxVP [26]. AKAP79 and cain inhibit calcineurin through their

PxIxIT motif [27]. RCAN1.4 has a similar motif in addition to the calcineurin-

inhibitor calcipresin 1 (CIC) motif [27]. CsA and FK506 bind to cyclophilin and

FKBP12, respectively, which in turn bind and inhibit calcineurin through their

LxVP motifs [28]. However, neither calsarcin nor Bcl-2 contains these motifs.

Analysis of the FHL2 sequence reveals that FHL2 does not contain those motifs either. It has been proposed that Bcl-2 inhibits calcineurin/NFAT by sequestering calcineurin within the ER or the mitochondrial membrane [29]. Thus, it is possible that inhibition by FHL2 is accomplished by sequestration of calcineurin to the sarcomere, consistent with increased affinity between FHL2 and calcineurin following adrenergic stimulation (Figure 2.11C,D). The exact mechanism of inhibition of calcineurin by FHL2 will require further study. This includes mapping the sites on FHL2 and calcineurin required for interaction and possible complexes with NFAT or inhibitors of calcineurin.

FHL2 and the MEK1/1-ERK1/2 pathway

Purcell et al. reported that FHL2 inhibits MAPK-ERK signaling and hypertrophic growth in neonatal cardiomyocytes by binding to activated phosphorylated ERK2 and inhibiting its nuclear accumulation [30]. It has been demonstrated that the MEK1/2-ERK1/2 pathway is activated by isoproterenol,

AngII and in pressure overload induced hypertrophy [31-32]. Overexpression of an active mutant of MEK-1 is sufficient to induce hypertrophy in neonatal cardiomyocytes [33]. MEK1 inhibitors, PD98059 and U0126, have been

demonstrated to inhibit hypertrophy by agonists such as PE, AngII and ET-1 [34].

To understand the role of the Ras/MEK/ERK pathway in vivo, hypertrophy was

evaluated in transgenic mice overexpressing constitutively active Ras or MEK1.

Ras transgenic animals developed a pathological hypertrophy phenotype

characterized by fetal gene induction, interstitial fibrosis, diastolic dysfunction

and arrhythmic sudden death [35]. Furthermore, overexpression of active MEK-1 by a heart-specific promoter supports the in vitro data. These mice develop compensated hypertrophy by two months which does not progress to heart failure

[33]. This is a mild phenotype compared to the one observed in Ras transgenics suggesting that other pathways downstream of Ras are responsible for the transition to heart failure.

Although activation of Ras/MEK1 leads to a hypertrophic phenotype, recent studies of ERK knockout mice questioned the role of ERK1/2 in hypertrophy. Mice lacking ERK1 and one allele of ERK2 (ERK1-/- ERK2-/+) had a normal response to pressure overload-induced hypertrophy [36]. The contribution of ERK1/2 was also studied in inducible transgenic mice that express dual specificity phosphatase 6 (DUSP6), an ERK1/2-specific phosphatase that dephosphorylates and inactivates ERK1/2 [37]. Induction of DUSP6 completely suppressed phosphorylation of ERK1/2 in a pressure-overload model, whereas

MEK1 phosphorylation was unaffected. Also, Purcell et al. found that DUSP6

transgenic mice retained the ability to mount a hypertrophic response in the

pressure overload model [36], confirming limited or no function of ERK1/2 in hypertrophy. This appears contradictory to their conclusion that FHL2 prevents hypertrophy through inhibition of nuclear localization of ERK2, as ERK2 appears to have a limited role in hypertrophy. Whether constitutively active ERK2 could induce hypertrophy in vivo or in vitro requires further study. If constitutively active ERK2 fails to induce hypertrophic growth, this will suggest that FHL2 acts on a different pathway downstream of MEK1.

FHL2 and ERK2 have been shown to localize to the sarcomere in adult cardiomyocytes [30]. It was proposed that FHL2 inhibits MAPK-ERK signaling by binding to ERK and inhibiting its translocation into the nucleus. I have confirmed interaction of FHL2 with ERK2 (immunoprecipitation) and their colocalization in adult heart. I next tested whether FHL2 was required for localization of ERK2 to the sarcomere (Figure 3.1A). Staining with ERK antibody showed that ERK localization was unchanged in FHL2 knockout mice compare to wild type (Figure 3.1B), suggesting that FHL2 is not responsible for sarcomeric localization of ERK.

Suppression of MEK1/2-ERK1/2 pathway through calcineurin

Both calcineurin/NFAT and MEK1/2-ERK1/2 pathways are activated through α- and β-adrenergic receptors [38]. Calcineurin is activated by increased calcium levels following isoproterenol, AngII or PE stimulation. Interestingly,

recent studies have shown that elevated calcium levels are involved in activation of ERKs. Calcium chelation by EGTA suppressed phosphorylation of ERK1/2 in response to adrenergic stimulation [32]. Similar results were observed when cells were pretreated with the L-type calcium antagonist, nifedipine [39]. Further analysis of the MEK1/2-ERK1/2 pathway in hypertrophic settings revealed that activation of this pathway depends on activation of calcineurin. Inhibition of calcineurin with FK506 or CsA blunts phosphorylation of ERK1/2 by isoproterenol or AngII treatment in vitro [39-40]. In vivo analysis in a model of renovascular hypertension induced-hypertrophy (two kidney-one clip (2K1C)) likewise show that CsA blunts phosphorylation of ERK [40]. Furthermore, hypertrophy in response to constitutively active MEK1 was abolished when

MEK1 transgenic mice were crossed to calcineurin Aβ knockout mice [38]. In summary, in vitro and in vivo data suggest that calcineurin is required for activation and hypertrophy induced by MEK1/2-ERK1/2 pathway. Thus, inhibition of calcineurin/NFAT signaling by FHL2 suppresses activation of

MEK1/ERK1/2 signaling.

Purcell et al. have reported in their study that FHL2 suppresses hypertrophy through inhibition of nuclear accumulation of ERK2 [30]. In their study, luciferase assays were performed 48 hours after transfection with MEK1, which triggers hypertrophic growth as demonstrated by cell cross-sectional area and expression of hypertrophic markers and which is inhibited by FHL2. As

calcineurin is required for the hypertrophic effect of MEK1, I hypothesize that

calcineurin signaling will also be activated during hypertrophic growth. Thus, it

is possible that FHL2 inhibits hypertrophic growth in this model by means of its

suppression of calcineurin/NFAT signaling. Inhibition of transcription from

ANF and GATA-4 promoters by FHL2 could be explained by the secondary suppression of hypertrophy by FHL2, hence lower expression of these promoters.

In my study, to avoid confounding effects of hypertrophy and inhibition of hypertrophy, experiments were repeated in HEK cells, which do not undergo hypertrophy. Cells were treated with ionomycin for only 4 hours to get optimal data in a short time and avoid off-target effects. All transfection experiments which involve overexpression of protein were carried out for no longer than 24 hours, and every promoter of interest was cotransfected with control promoter to normalize for non-specific effect of FHL2 on transcription.

FHL2 and FHL1

FHL2 is detected in a number of tissues, with the highest level of expression in the heart, both at protein and mRNA levels. FHL1 is another member of the four and a half LIM protein family, which is highly expressed in skeletal muscle, lung, kidney and to a lower extent in the heart [3]. Although

FHL2 is more abundant in heart, its expression is unchanged in pressure overload

induced cardiac hypertrophy, whereas FHL1 is highly upregulated [3]. FHL2

levels are elevated in hearts after cardiopulmonary bypass [4], whereas FHL2 protein abundance is decreased in failing hearts [5]. Here, I demonstrate that

FHL2 expression is increased in isoproterenol-induced hypertrophy. Although both FHL1 and FHL2 have similar structure and 50% identity in amino acid sequence, they play opposite roles when it comes to hypertrophy. Whereas FHL2 knockout mice display exaggerated hypertrophy in response to isoproterenol,

FHL1-null mice exhibit blunted hypertrophic growth induced by either transverse aortic constriction or by constitutively active Gq [41]. These authors proposed that the blunted phenotype in FHL1 knockout mice was due to reduced activation of the ERK pathway, suggesting that FHL1 promotes activation of ERK. In another study, FHL1 also activated NFAT-dependent gene expression in C2C12 cells and FHL1 transgenic mice [42]. However, FHL2, consistent with the in vivo phenotype, inhibits both ERK and calcineurin-NFAT dependent gene expression.

Collectively, these data show FHL1 and FHL2 have opposite effects on MAPK-

ERK and calcineurin-NFAT signaling, thus resulting in opposite hypertrophic phenotypes.

FHL2 in TAC and isoproterenol-induced hypertrophy

FHL2 knockout mice have similar pressure overload-induced hypertrophic growth as reported by Chu et al [43] and data presented here, whereas when treated with isoproterenol, FHL2-/- mice exhibit exaggerated hypertrophy. This

could be explained by the differences in the signaling cascades activated in these

two different models of hypertrophy. Pressure overload induces a more global

response which involves diverse hypertrophic signaling pathways, whereas

isoproterenol infusion may induce a focused response that is dependent on calcineurin/NFAT signaling. Mechanical stress is considered to be a trigger to

activation of a hypertrophic response in an overloaded heart. However, in addition

to activating signaling pathways, mechanical stress will also induce activation of

α-adrenergic receptors (AngII and ET-1) and β-adrenergic receptors (epinephrine

and norepinephrine), suggesting a role of these receptors in pressure overload hypertrophy. However, null mice for either α- or β-adrenergic [44] receptors still develop hypertrophy in response to TAC, supporting a global response, rather than dependence on a single pathway.

Another explanation for the growth response observed in response to TAC and isoproterenol in FHL2-/- mice could be the increased expression of FHL2 in isoproterenol-treated hearts (Figure 2.2A,C), whereas it remains unchanged in

pressure overload model (Figure 2.2G). The effect of FHL2 on transcriptional

regulation of gene expression is well established in vitro, although there are no

reports of regulation of transcription by FHL2 in cardiomyocytes. It is possible to

attribute this role to FHL2 in primary cardiac cells, where nuclear localization is

observed in some cells (Figure 2.10A). However, it is harder to accept it in adult

cardiomyocytes, where FHL2 solely localizes to the sarcomere. In fact, nuclear

localization of FHL2 was not detected in cardiomyocytes even under hypertrophic stimulation induced by the TAC procedure (Figure 2.10C). However, since we

have observed increased expression of FHL2 in isoproterenol induced

hypertrophy, it is possible that newly synthesized FHL2 translocates into the

nucleus to regulate gene expression, whereas sarcomeric FHL2 remains in the

cytoplasm.

This is similar to the case of β-catenin which is mainly found in two pools: cytoplasmic, and e-cadherin bound β-catenin at adherens junctions [45]. The latter

is required for proper formation of adherens junctions and is not involved in Wnt

signaling. However, levels of cytoplasmic β-catenin are tightly regulated by

ubiquitination and proteasome-dependent degradation in the absence of Wnt

ligand. Activation of Wnt signaling stabilizes β-catenin, which translocates into

the nucleus and activates expression with TCF/LEF transcription factors. In our

case it is possible that FHL2 localized to Z-disks does not dissociate from the

sarcomere, so it does not localize to the cytoplasm or nucleus. However, newly synthesized FHL2 might translocate to the nucleus, as it is not associated yet with the sarcomere. Thus, localization of newly synthesized FHL2 will need to be studied in hearts treated with isoproterenol; nuclear localization and its function in the nucleus will have to be evaluated.

Regulation of binding of FHL2

Here, I have demonstrated that FHL2 binds to calcineurin. In

overexpression experiments, FHL2 binds both wild-type calcineurin and

constitutively active calcineurin (Figure 2.11B). However, at endogenous levels

of calcineurin and FHL2 in neonatal rat ventricular myocytes, interaction between

them was almost undetectable in conditions where calcineurin was inactive.

However, stimulation of cells with ET-1 increased the amount of FHL2

immunoprecipitated with calcineurin (Figure 2.11C,D), suggesting that FHL2 has

higher affinity for active calcineurin compared to inactive calcineurin.

Immunoprecipitation of control SK-1 from the same lysates used for calcineurin

showed that FHL2 binds SK-1 under basal conditions and dissociates when

treated with ET-1 (Figure 2.11D), as reported previously [24]. This type of

regulation has also been shown for other proteins. For instance, FHL2 has low

affinity for unphosphorylated ERK2 (inactive) and high affinity for

phosphorylated ERK2 (active) [30]. Another case is regulation of FHL2 and c-

Fos interaction. Heregulin (EGF-like growth factor) stimulation triggers

differentiation in the MCF-7 breast cancer cell line. Treatment of MCF-7 cells

with heregulin increased association of FHL2 and c-Fos [46].

Immunoprecipitation from neonatal cardiomyocytes showed that FHL2

and calcineurin interaction is enhanced following activation of adrenergic receptors by ET-1. One question is whether this interaction explains the

hypertrophic response of FHL2-/- mice to isoproterenol and TAC. Do FHL2 and calcineurin interact in response to mechanical stress? Thus, it would be important to observe whether the interaction between FHL2 and calcineurin can be triggered by biomechanical stress by stretching neonatal cardiomyocytes on special culture plates. If stretching does not stimulate FHL2-calcineurin complex formation, this could explain the lack of effect of FHL2 in TAC-induced hypertrophy.

Perspective

Here I present data that shows an interaction between FHL2 and calcineurin at endogenous expression levels, which is triggered by the hypertrophic agonist ET-1. Knockdown of FHL2 augments activation of NFAT target genes by calcineurin. Overexpression of FHL2 suppresses NFAT- dependent promoter activation and hypertrophic growth stimulated by calcineurin.

FHL2 knockout mice display higher activation of calcineurin/NFAT signaling and therefore are more susceptible to hypertrophic insults. Thus, FHL2 is involved in hypertrophy through calcineurin/NFAT signaling and may represent a new therapeutic target.

Figure 3.1 FHL2 is not required to localization of ERK to the sarcomere. A) FHL2 and ERK colocalize to the sarcomere in adult cardiomyocyte. B) Immuno- staining with ERK specific antibody of WT and FHL2-/- hearts. (Scale bar, 5µm.)

References

1. Johannessen, M., et al., The multifunctional roles of the four-and-a-half- LIM only protein FHL2. Cell Mol Life Sci, 2006. 63(3): p. 268-84.

2. Kong, Y., et al., Cardiac-specific LIM protein FHL2 modifies the hypertrophic response to beta-adrenergic stimulation. Circulation, 2001. 103(22): p. 2731-8.

3. Chu, P.H., et al., Expression patterns of FHL/SLIM family members suggest important functional roles in skeletal muscle and cardiovascular system. Mech Dev, 2000. 95(1-2): p. 259-65.

4. Wan, S., et al., Expression of FHL2 and cytokine messenger RNAs in human myocardium after cardiopulmonary bypass. Int J Cardiol, 2002. 86(2-3): p. 265-72.

5. Bovill, E., et al., Reduction of four-and-a-half LIM-protein 2 expression occurs in human left ventricular failure and leads to altered localization and reduced activity of metabolic enzymes. J Thorac Cardiovasc Surg, 2009. 137(4): p. 853-61.

6. Philippar, U., et al., The SRF target gene Fhl2 antagonizes RhoA/MAL- dependent activation of SRF. Mol Cell, 2004. 16(6): p. 867-80.

7. Scholl, F.A., et al., DRAL is a p53-responsive gene whose four and a half LIM domain protein product induces apoptosis. J Cell Biol, 2000. 151(3): p. 495-506.

8. Nelson, T.J., et al., SRF-dependent gene expression in isolated cardiomyocytes: regulation of genes involved in cardiac hypertrophy. J Mol Cell Cardiol, 2005. 39(3): p. 479-89.

9. Zhang, X., et al., Cardiomyopathy in transgenic mice with cardiac-specific overexpression of serum response factor. Am J Physiol Heart Circ Physiol, 2001. 280(4): p. H1782-92.

10. Parlakian, A., et al., Temporally controlled onset of dilated cardiomyopathy through disruption of the SRF gene in adult heart. Circulation, 2005. 112(19): p. 2930-9.

11. Heemers, H.V., et al., Androgen induction of the androgen receptor coactivator four and a half LIM domain protein-2: evidence for a role for serum response factor in prostate cancer. Cancer Res, 2007. 67(21): p. 10592-9.

12. Lu, Y., et al., MicroRNA-1 downregulation by propranolol in a rat model of myocardial infarction: a new mechanism for ischaemic cardioprotection. Cardiovasc Res, 2009. 84(3): p. 434-41.

13. Davis, A., et al., Functional significance of protein kinase A activation by endothelin-1 and ATP: negative regulation of SRF-dependent gene expression by PKA. Cell Signal, 2003. 15(6): p. 597-604.

14. Genini, M., et al., Subtractive cloning and characterization of DRAL, a novel LIM-domain protein down-regulated in rhabdomyosarcoma. DNA Cell Biol, 1997. 16(4): p. 433-42.

15. Komuro, I. and S. Izumo, Csx: a murine homeobox-containing gene specifically expressed in the developing heart. Proc Natl Acad Sci U S A, 1993. 90(17): p. 8145-9.

16. Saadane, N., L. Alpert, and L.E. Chalifour, Expression of immediate early genes, GATA-4, and Nkx-2.5 in adrenergic-induced cardiac hypertrophy and during regression in adult mice. Br J Pharmacol, 1999. 127(5): p. 1165-76.

17. Chen, C.Y. and R.J. Schwartz, Recruitment of the tinman homolog Nkx- 2.5 by serum response factor activates cardiac alpha-actin gene transcription. Mol Cell Biol, 1996. 16(11): p. 6372-84.

18. Molkentin, J.D., et al., A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell, 1998. 93(2): p. 215-28.

19. Bueno, O.F., et al., Impaired cardiac hypertrophic response in Calcineurin Abeta -deficient mice. Proc Natl Acad Sci U S A, 2002. 99(7): p. 4586-91.

20. Rothermel, B.A., R.B. Vega, and R.S. Williams, The role of modulatory calcineurin-interacting proteins in calcineurin signaling. Trends Cardiovasc Med, 2003. 13(1): p. 15-21.

21. Cohen, T.V. and W.R. Randall, NFATc1 activates the acetylcholinesterase promoter in rat muscle. J Neurochem, 2004. 90(5): p. 1059-67.

22. Hill, J.A., et al., Targeted inhibition of calcineurin in pressure-overload cardiac hypertrophy. Preservation of systolic function. J Biol Chem, 2002. 277(12): p. 10251-5.

23. Hasegawa, K., et al., Ventricular expression of atrial and brain natriuretic peptides in dilated cardiomyopathy. An immunohistocytochemical study of the endomyocardial biopsy specimens using specific monoclonal antibodies. Am J Pathol, 1993. 142(1): p. 107-16.

24. Sun, J., et al., FHL2/SLIM3 decreases cardiomyocyte survival by inhibitory interaction with sphingosine kinase-1. Circ Res, 2006. 99(5): p. 468-76.

25. Kakita, T., et al., Calcineurin pathway is required for endothelin-1- mediated protection against oxidant stress-induced apoptosis in cardiac myocytes. Circ Res, 2001. 88(12): p. 1239-46.

26. Martinez-Martinez, S., et al., Blockade of NFAT activation by the second calcineurin binding site. J Biol Chem, 2006. 281(10): p. 6227-35.

27. Sieber, M. and R. Baumgrass, Novel inhibitors of the calcineurin/NFATc hub - alternatives to CsA and FK506? Cell Commun Signal, 2009. 7: p. 25.

28. Liu, J.O., et al., It takes two binding sites for calcineurin and NFAT to tango. Mol Cell, 2009. 33(6): p. 676-8.

29. Shibasaki, F., et al., Suppression of signalling through transcription factor NF-AT by interactions between calcineurin and Bcl-2. Nature, 1997. 386(6626): p. 728-31.

30. Purcell, N.H., et al., Extracellular signal-regulated kinase 2 interacts with and is negatively regulated by the LIM-only protein FHL2 in cardiomyocytes. Mol Cell Biol, 2004. 24(3): p. 1081-95.

31. Bueno, O.F. and J.D. Molkentin, Involvement of extracellular signal- regulated kinases 1/2 in cardiac hypertrophy and cell death. Circ Res, 2002. 91(9): p. 776-81.

32. Yamazaki, T., et al., Norepinephrine induces the raf-1 kinase/mitogen- activated protein kinase cascade through both alpha 1- and beta- adrenoceptors. Circulation, 1997. 95(5): p. 1260-8.

33. Bueno, O.F., et al., The MEK1-ERK1/2 signaling pathway promotes compensated cardiac hypertrophy in transgenic mice. EMBO J, 2000. 19(23): p. 6341-50.

34. Wang, L. and C.G. Proud, Ras/Erk signaling is essential for activation of protein synthesis by Gq protein-coupled receptor agonists in adult cardiomyocytes. Circ Res, 2002. 91(9): p. 821-9.

35. Zheng, M., et al., Sarcoplasmic reticulum calcium defect in Ras-induced hypertrophic cardiomyopathy heart. Am J Physiol Heart Circ Physiol, 2004. 286(1): p. H424-33.

36. Purcell, N.H., et al., Genetic inhibition of cardiac ERK1/2 promotes stress-induced apoptosis and heart failure but has no effect on hypertrophy in vivo. Proc Natl Acad Sci U S A, 2007. 104(35): p. 14074- 9.

37. Li, C., et al., Dusp6 (Mkp3) is a negative feedback regulator of FGF- stimulated ERK signaling during mouse development. Development, 2007. 134(1): p. 167-76.

38. Sanna, B., et al., Direct and indirect interactions between calcineurin- NFAT and MEK1-extracellular signal-regulated kinase 1/2 signaling pathways regulate cardiac gene expression and cellular growth. Mol Cell Biol, 2005. 25(3): p. 865-78.

39. Zou, Y., et al., Isoproterenol activates extracellular signal-regulated protein kinases in cardiomyocytes through calcineurin. Circulation, 2001. 104(1): p. 102-8.

40. Murat, A., et al., Calcineurin blockade prevents cardiac mitogen-activated protein kinase activation and hypertrophy in renovascular hypertension. J Biol Chem, 2000. 275(52): p. 40867-73.

41. Sheikh, F., et al., An FHL1-containing complex within the cardiomyocyte sarcomere mediates hypertrophic biomechanical stress responses in mice. J Clin Invest, 2008. 118(12): p. 3870-80.

42. Cowling, B.S., et al., Identification of FHL1 as a regulator of skeletal muscle mass: implications for human myopathy. J Cell Biol, 2008. 183(6): p. 1033-48.

43. Chu, P.H., et al., FHL2 (SLIM3) is not essential for cardiac development and function. Mol Cell Biol, 2000. 20(20): p. 7460-2.

44. Palazzesi, S., et al., Pressure overload causes cardiac hypertrophy in beta1-adrenergic and beta2-adrenergic receptor double knockout mice. J Hypertens, 2006. 24(3): p. 563-71.

45. Papkoff, J., et al., Wnt-1 regulates free pools of catenins and stabilizes APC-catenin complexes. Mol Cell Biol, 1996. 16(5): p. 2128-34.

46. Saeki, Y., et al., Ligand-specific sequential regulation of transcription factors for differentiation of MCF-7 cells. BMC Genomics, 2009. 10: p. 545.