Functional Remodeling Following Myofilament Calcium Sensitization in Rats with Volume Overload Heart Failure DISSERTATION Presen

Functional Remodeling Following Myofilament Calcium Sensitization in Rats with

Volume Overload Heart Failure

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Kristin Diane Lewis, D.V.M.

Graduate Program in Comparative and Veterinary Medicine

The Ohio State University

2014

Dissertation Committee:

Pamela Lucchesi, Advisor

Christopher Breuer

Christopher Premanandan

Lynette Rogers

Copyrighted by

Kristin Diane Lewis

2014

Abstract

Hemodynamic volume overload (VO) is characterized by left ventricular (LV) dilation, progressive LV dysfunction, and heart failure (HF). In the aortocaval fistula (ACF) model of VO HF, LV dysfunction is accompanied by a variety of changes at the myocyte level including myofilament proteins including decreased α-to-β-myosin heavy chain

(MHC) expression, decreased myofilament Ca2+ sensitivity and altered regulation of intercalated disc proteins involved in electromechanical coupling. While pharmacologic therapy alone cannot resolve the hemodynamic overload in human patients or in animal models, pre-operative pharmacologic therapy may delay time to surgical intervention, and post-operative pharmacologic therapy may accelerate functional recovery. Current therapies that target neurohormonal pathway activation provide some relief but do not reverse the decreased cardiac contractility, the central feature of systolic HF [60].

Targeting neurohormonal pathway activation can result in increased myocardial oxygen consumption and myocardial Ca2+ overload [140], and therefore it is necessary to investigate drugs that have alternative mechanisms of action. Levosimendan (Levo) stabilizes Ca2+-saturated troponin C (cTnC) prolonging its interaction with cardiac troponin I (cTnI), which promotes contractile force without increasing the intracellular

Ca2+ transient amplitude or myocardial oxygen consumption [126,128,136]. Given these overall factors, my hypothesis is that therapeutic strategies such as Levo that target ii myofilament Ca2+ sensitization will preserve/improve LV function in VO HF. To test this hypothesis, three specific aims were proposed: 1) determine if short-term Levo treatment would preserve/improve LV function in rats with ACF-induced pre-HF or reversed pre-

HF (Chapter 2), 2) determine if Levo could improve the detrimental effects of delayed reversal (Chapter 3), and 3) determine if chronic Levo treatment would preserve LV function when initiated in rats with ACF-induced pre-HF or established HF (Chapter 4).

Sham and ACF surgery were performed at Week 0 following by Levo treatment with and without hemodynamic load reduction surgery (reversal) at various timepoints. Continued

Levo improved systolic and diastolic function regardless of the treatment starting point and hemodynamic load. Improved LV function variably correlated with increased myofilament Ca2+ sensitivity, cMyBP-C/cTnI phosphorylation and normalization of α-to-

β-MHC. Finally, speckle-tracking echocardiographic analysis suggests that Levo improves short-axis, but not long-axis function at end-stage HF. Because of the improved LV function with Levo, with and without hemodynamic load reduction, Levo offers a new therapeutic option in patients with VO HF. More broadly, therapeutic strategies targeting myofilament Ca2+ sensitization may provide a new therapeutic target for patients with volume overload heart failure.

iii

To my husband Travis, my family, and many others for your unwavering patience, love and support through the many highs and lows of this journey.

Acknowledgments

I would like to thank my advisor, Dr. Pamela Lucchesi, the members of my graduate committee, and the members of the Lucchesi lab for their support in the design and execution of the experiments done in this dissertation. I would like to thank the faculty veterinary pathologists at The Ohio State University who have been instrumental in developing and shaping my skills as a veterinary pathologist, and I would like to thank my resident-mates who walked through this journey with me. I would like those who pushed me to become a veterinary pathologist, especially NOD who taught me to ask the question, ―What’s your hypothesis?‖ Finally, I would like to thank my friends and family whose unwavering support and encouragement have helped me through the highs and lows of my educational journey.

Vita

2001...... B.S. Microbiology, California Polytechnic

State University

2001-2006 ...... Research Associate, Genentech, Inc

2010...... D.V.M., University of California Davis,

School of Veterinary Medicine

2010 to present ...... Veterinary Anatomic Pathology Resident,

Graduate Research Associate, Department

of Veterinary Biosciences, The Ohio State

University

Publications

Wilson K, Lucchesi PA. Myofilament dysfunction as an emerging mechanism of volume overload heart failure. Pflugers Arch. 2014 Feb 1. [Epub ahead of print] PMID:

24488008

Fields of Study

Major Field: Comparative and Veterinary Medicine

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... vi

List of Tables ...... xiv

List of Figures ...... xv

Chapter 1: Introduction and Background ...... 1

Abstract ...... 1

Introduction ...... 1

Rodent and non-rodent animal models are used to study VO-HF ...... 4

Neurohormonal activation in volume overload ...... 5

Volume overload physiology and structural alterations differs from pressure

overload ...... 6

LV pump function in VO...... 7

Cardiomyocyte dysfunction in VO ...... 9

Excitation-contraction coupling ...... 9 vii

VO HF, intercalated discs and electromechanical transmission ...... 10

Gap junctions/electrical communication ...... 12

Force transmission through the intercalated disc ...... 13

Myofilament dysfunction in VO ...... 14

Force generation ...... 14

DCM-causing mutation in thin filament proteins ...... 16

Thick filament proteins in VO HF ...... 20

Passive muscle stiffness in VO ...... 24

Extracellular matrix (ECM) alterations in VO...... 24

ECM turnover ...... 24

Mechanical stress-induced ECM remodeling ...... 26

Mast cells and VO-induced ECM remodeling ...... 28

Mast cells and paracrine-induced LV ECM remodeling ...... 30

Pharmacologic therapy in VO ...... 31

Diuretics...... 31

Vasodilators ...... 31

β-adrenergic antagonists ...... 32

Myofilament Ca2+ sensitizers ...... 32

Myosin activators ...... 33

viii

Phosphodiesterase inhibitors ...... 33

Limitations and future directions ...... 34

Chapter 2: Levosimendan preserves cardiac function in rats with ACF-induced pre-

HF and reversed pre-HF ...... 35

Abstract ...... 35

Introduction ...... 36

Materials and Methods ...... 39

ACF Surgical Model in Rats ...... 39

Radiotelemetry...... 41

Levosimendan treatment ...... 41

Echocardiography ...... 42

Measurement of LV hemodynamics ...... 42

LV myocyte isolation ...... 43

Measurement of sarcomere shortening and Ca2 +-transients in isolated LV myocytes

...... 44

Myofilament Force-pCa ...... 44

Immunoblot analysis...... 45

Quantitative real-time PCR analysis...... 45

Statistical analysis...... 46

Results ...... 46

Levo does not significantly alter LV chamber morphology ...... 46

Levo improves LV systolic and diastolic function without altering β-adrenergic

responsiveness ...... 47

Levo does not significantly alter mean arterial pressure ...... 50

Levo improves myofilament Ca2+ sensitivity and maximal force generation without

altering myocyte Ca2+ transient kinetics ...... 50

Levo does not significantly alter LV myocyte sarcomere shortening ...... 52

Levo’s positive lusitropy is associated with increased cMyBP-C phosphorylation,

cTnI phosphorylation, and/or increased α-to-β-MHC ...... 52

Comparison of acute effects of Levo, milrinone and OM on LV function in rats with

established VO-HF ...... 54

Discussion ...... 55

Chapter 3: Impaired function in rats following delayed reversal of volume overload heart failure can be rescued by levosimendan ...... 62

Abstract ...... 62

Introduction ...... 63

Materials and Methods ...... 65

ACF Surgical Model in Rats ...... 65

Levosimendan treatment ...... 67 x

Echocardiography ...... 68

Measurement of LV hemodynamics ...... 68

Immunoblot analysis...... 69

Quantitative real-time PCR analysis...... 69

Statistical analysis...... 70

Results ...... 70

LV chamber morphology rapidly returns to Sham dimensions even after delayed

reversal ...... 70

LV function remains impaired after delayed reversal ...... 72

Levo does not significantly alter LV chamber morphology ...... 74

Continued Levo improves LV systolic and diastolic function ...... 75

Delayed reversal normalizes VO-induced expression of the hypertrophic markers

ANP and BNP ...... 77

Improved LV function with sustained Levo treatment is associated with normalized

α-to-β-MHC ...... 78

Levo’s positive lusitropy is also associated with increased cTnI phosphorylation ... 79

Discussion ...... 80

Chapter 4: Chronic levosimendan administration in rats with volume overload heart failure preserves systolic and diastolic function and circumferential strain ... 89

Abstract ...... 89 xi

Introduction ...... 90

Materials and Methods ...... 92

ACF Surgical Model in Rats ...... 92

Levosimendan treatment ...... 93

Echocardiography ...... 94

Measurement of LV hemodynamics ...... 94

Two-dimensional speckle-tracking analysis ...... 95

Statistical analysis...... 96

Results ...... 96

Levo does not significantly alter LV chamber morphology ...... 96

Levo improves LV systolic and diastolic function without altering survival ...... 97

Levo preserved circumferential strain during end-stage HF ...... 100

Discussion ...... 102

Chapter 5: Summary of findings and future directions ...... 109

Abstract ...... 109

Current clinical problem ...... 109

Summary of findings...... 111

Study Limitations ...... 117

Future directions ...... 120

xii

What is myosin heavy chain (MHC)’s role in functional recovery? ...... 120

What role does cardiac short-axis and long-axis function play in VO HF

progression? ...... 124

Final concluding remarks ...... 127

References ...... 128

xiii

List of Tables

Table 1: LV morphologic, echocardiographic and hemodynamic parameters in rats at

Week 8 ...... 61

Table 2: LV morphologic, echocardiographic and hemodynamic parameters in rats at

Week 19 ...... 87

Table 3: LV morphologic, echocardiographic and hemodynamic parameters in Levo- treated REV@8 rats at Week 19 ...... 88

Table 4: LV morphologic, echocardiographic and hemodynamic parameters in Sham,

ACF-Veh and ACF-Levo rats at Week 19 ...... 108

xiv

List of Figures

Figure 1: Representative pressure-volume loops from mitral (MR) and aortic (AR) regurgitation ...... 3

Figure 2: Effect of VO on intercalated disc proteins ...... 11

Figure 3: Effect of VO on myofilament and myofilament-regulatory proteins ...... 15

Figure 4: Experimental time course ...... 40

Figure 5: Levo improves %FS but not LVEDD ...... 47

Figure 6: Levo improves systolic and diastolic function ...... 48

Figure 7: Beta-adrenergic responsiveness is preserved in ACF and REV ...... 49

Figure 8: Levo improves myofilament calcium sensitivity and force generation without increasing the peak intracellular calcium transient ...... 51

Figure 9: Levo increases cMyBP-C phosphorylation, cTnI phosphorylation and/or alpha to beta MHC ...... 53

Figure 10: Short-term treatment with PDE3a inhibitor improves lusitropy in 8-week ACF rats ...... 55

Figure 11: Experimental time course ...... 67

Figure 12: Effect of delayed reversal on LVEDD and %FS ...... 71

Figure 13: LV systolic and diastolic function are impaired following delayed reversal .. 73

Figure 14: Levo improved %FS without altering chamber morphology ...... 75

Figure 15: Continued Levo improves LV systolic and diastolic function following delayed reversal ...... 76

Figure 16: Continued Levo normalizes alpha-to-beta MHC ...... 79

Figure 17: Pre-reversal LVESD and %FS negatively correlate with post-reversal %FS at

Week 19 ...... 82

Figure 18: Experimental time course ...... 93

Figure 19: Effect of Levo on chamber dilation and % fractional shortening ...... 97

Figure 20: Levo improves LV systolic and diastolic function longterm ...... 98

Figure 21: Overall survival is comparable amongst the ACF groups ...... 99

Figure 22: Representative echocardiograms and strain curves in Sham and ACF rats at

Week 19 ...... 101

Figure 23: Levo preserved circumferential, but not longitudinal or radial, strain ...... 102

Figure 24: Mechanism of improved LV function ...... 112

Figure 25: Experimental time course for proposed Experiment 1 ...... 121

Figure 26: Experimental time course for proposed Experiment 2 ...... 123

Figure 27: Experimental design for proposed Experiment 4 ...... 125

Figure 28: Experimental design for proposed Experiment 5 ...... 126

xvi

Chapter 1: Introduction and Background

Abstract

Two main hemodynamic overload mechanisms (i.e. volume and pressure overload) result in heart failure and these two mechanisms have divergent pathologic alterations and different pathophysiological mechanisms. Extensive evidence from animal models and human studies of pressure overload demonstrate a clear association with alterations in Ca2+ homeostasis. By contrast, emerging evidence from animal models and patients with regurgitant valve disease and dilated cardiomyopathy point toward a more prominent role of myofilament dysfunction. With respect to volume overload heart failure, key features of excitation-contraction coupling defects, myofilament dysfunction and ECM composition will be discussed.

Introduction

Altered hemodynamic load places increased mechanical stress on the heart and initiates a series of compensatory changes within the myocardium that drive physiologic or pathologic structural remodeling depending on the magnitude, duration and type of load. Cardiac volume overload (VO) is associated with increased preload (i.e. increased end-diastolic volume). Causes of physiologic VO range from aerobic exercise training and pregnancy to the post-prandial period in the Burmese python [117]. Physiologic VO 1 is typically associated with an intermittent, non-chronic VO; therefore, chamber geometry is preserved despite biventricular overload. As a result, there is lack of apoptosis, metabolic remodeling, fibrosis and fetal gene re-expression) and hypertrophy is reversible [117]. The mechanisms of physiologic hypertrophy have been reviewed elsewhere [117].

Pathologic volume overload results from persistent abnormal elevation of hemodynamic load and causes include valvular regurgitation, ventricular septal defects, arteriovenous fistulae, and genetic or idiopathic causes in dilated cardiomyopathy. In mitral regurgitation, a fraction of the stroke volume leaks into the left atrium through an incompetent mitral valve during systole. This blood re-enters the ventricle in subsequent cardiac cycles, but because the blood escapes to a low pressure chamber, the ventricle is exposed to pure volume overload [33,117]. Also, because the regurgitant blood shuttles between the left atrium and the left ventricle, the overload is primarily within the left heart. With time, the LV adapts by undergoing eccentric hypertrophy. In aortic regurgitation, a fraction of the stroke volume leaks through an incompetent aortic valve during diastole into the left ventricle which is also receiving blood from the left atrium.

While MR produces pure VO, AR produces both VO and pressure overload because of the increased LV systolic and diastolic pressures [33,117]. As with mitral regurgitation, because the regurgitant blood shuttles between the LV and the aorta, the overload is largely within the left heart. With time, the LV adapts by undergoing a combination of eccentric and concentric hypertrophy. Representative pressure-volume loops for MR and

Figure 1: Representative pressure-volume loops from mitral (MR) and aortic (AR) regurgitation In MR and AR loops, end-diastolic volume (EDV) and stroke volume (SV) are increased. End-systolic volume and pressure (ESV and ESP, respectively) are decreased in MR but increased in AR compared with normal hearts. Pressure volume loops are adapted from RE Klabunde, www.cvphysiology.com, 2013.

AR are shown in Fig 1. In hearts with ventricular septal defects (VSD), a fraction of blood from the LV passes into the RV during systole. A VSD produces pure VO, but because of the direction of blood flow, a VSD results in overload of both the right and left hearts [117]. With arteriovenous fistulae, there is a direct connection between a large artery and vein. Reportedly, a shunt fraction of 20-50% of cardiac output is sufficient to cause heart failure [117]. Similar to a VSD, arteriovenous fistulae result in biventricular overload [117]. In dilated cardiomyopathy, there is unexplained left ventricular dilation with chamber dimensions >115% of expected for a given age and body weight, and this

3 dilation is accompanied by decreased systolic function with fractional shortenings <25%

[119].

Rodent and non-rodent animal models are used to study VO-HF

Aortocaval fistula (ACF) is the most widely used method for inducing VO, especially in rodents [57,74,167]. This model recapitulates most of the pathophysiological features of diseases like MR, including sustained increases in preload,

LV dilation, a supranormal EF, myocyte and extracellular matrix remodeling, and progressive LV dysfunction [40,66,76,160]. It is a valuable tool for identifying molecular and cellular mechanisms that drive disease progression, and testing novel therapeutic approaches to improve LV structure and function [59,84,143,198]. One of the drawbacks to the ACF model is the biventricular VO; however, the fistula can be later closed simulating hemodynamic load reversal surgery (i.e. mitral valve repair surgery)

[76]. More recently, two groups have created MR in the rat by inserting a needle under trans-esophageal echocardiographic guidance to puncture the mitral valve

[87,88,152,153]. As with the ACF model, the size of the puncture dictates the magnitude of the regurgitant flow and ultimately the time course of disease progression. Kim et al

[87,88] show a similar time course in the increase in LV end-diastolic diameter (LVEDD) to the ACF model.

Transgenic models harboring putative dilated cardiomyopathy (DCM)-causing mutations have been used to study the effects of these mutations. In in vivo and ex vivo

4 systems, the mutation is introduced into the myofilaments or myofilament regulatory proteins, which are then exchanged for the native form [53,154,168,179]. The majority of these mutations have been localized primarily to the myofilament machinery (see below).

In the dog, MR is the most widely used method for inducing univentricular VO

[90,104,205]. Similar to the ACF model in rodents, this model recapitulates the pathophysiological features of VO, including sustained increases in preload, LV dilation, a supranormal EF, myocyte and extracellular matrix remodeling, and progressive LV dysfunction. Advantages of the canine model include similar heart rates, excitation- contraction (EC) coupling kinetics, and similar myosin heavy chain isoforms to the human [71,196]. One of the most notable drawbacks to this model is that heart failure develops more slowly in the dog than in the rat [181].

Neurohormonal activation in volume overload

Volume overload stimulates a multifaceted, systemic neurohormonal response that increases cardiac performance, inhibits renal salt and water excretion and results in vasoconstriction, and also stimulates a counter-regulatory response [86]. Neurohormonal alterations have been best studied in the rat ACF model. In this model, there is an immediate decrease in mean arterial pressure along with an increase in venous return to the right ventricle [1]. The neurohormonal systems that become activated include the sympathetic nervous system (SNS), the renin-angiotensin-aldosterone system (RAAS),

5 antidiuretic hormone (ADH) and endothelin-1 (ET-1) [1]. SNS activation results in cardiac stimulation, which increases cardiac output by increasing heart rate and end- diastolic volume and decreases end-systolic volume, and results in vasoconstriction [86].

The RAAS is activated early after ACF creation and stimulates tubular resorption of sodium directly and through aldosterone, stimulates ADH secretion which increases water absorption (and sodium) in the collecting ducts of the kidney, and stimulates the

SNS which results in arteriolar vasoconstriction [1,25]. ET-1 is a potent vasoconstrictor, and in ACF rats, there is a prompt 3-4 fold increase in ET-1 mRNA in the atria and a progressive 5-7 fold increase ET-1 mRNA in the ventricles [29].

The main counter-regulatory system activated in ACF rats includes the natriuretic peptides. Natriuretic peptides of cardiac origin include atrial natriuretic peptide (ANP) and B-type (BNP). Plasma ANP increases 20-fold compared with sham within the first

24 hours following ACF, which mirrors the increase in right atrial pressure, the most potent stimulus for ANP release [1]. The nitric oxide system and vasodilatory prostaglandins (i.e. prostaglandin E2 (PGE2)) have been studied in ACF rats with respect to renal blood flow and are discussed in [1].

Volume overload physiology and structural alterations differs from pressure overload

Hemodynamic overload can broadly be classified into volume and pressure overload, and there are distinct differences between the two with respect to chamber geometry, wall thickness and wall stress. Mechanical wall stress in the LV is

6 approximated by Laplace’s law (σ = (LV pressure x LV radius) / (2 x LV wall thickness).

In pressure overload, there is increased afterload, LV pressure and LV wall thickness increase, while LV radius remains relatively unchanged. Systolic wall stress is increased acutely, which is normalized by the increased wall thickness, while diastolic wall stress remains unchanged to potentially decreased chronically [117]. In volume overload, LV pressure and radius increase while the LV wall thickness remains relatively unchanged.

Systolic wall stress increases acutely but normalizes chronically, while diastolic wall stress increases acutely and remains chronically elevated [117].

LV pump function in VO

LV systolic and diastolic function can be measured through a variety of invasive

(i.e. pressure-volume analysis) and noninvasive techniques (i.e. echocardiography, MRI).

LV function is affected by chamber geometry, mass and loading conditions [117].

Systolic function is associated with excitation-contraction coupling derived Ca2+ delivery to the myofilament and the myofilament’s response to this Ca2+. Additionally, through increased preload (i.e. increased end-diastolic volume (EDV)), there is greater contractile strength due to the Frank-Starling effect, and through decreased afterload, there is increased velocity of shortening. Diastolic function is divided into an active relaxation component and a passive component dictated by structural components of the myocardium (i.e. titin and extracellular matrix (ECM)).

Pump function in VO patients is dependent on the stage of HF progression. In patients with chronic, compensated VO, the chamber dilation and increased compliance allows the LV to accommodate increased blood volume for a given pressure, and this improves passive diastolic filling [33]. By contrast, systolic function is impaired in VO due to changes in intrinsic contractility and alterations in afterload. In early VO, LV afterload is reduced by regurgitation into the LA, but afterload becomes normalized during compensated MR and becomes increased during late VO because of increased myocardial wall stress and increased systemic vascular resistance [33].

Heart failure in rats with ACF-induced VO and in dogs with MR can be divided into three clinically relevant stages: compensated pre-HF (marked LV dilation, increased

LV wall stress), established HF (severe LV dilation) and decompensated, end stage HF

[40,66,76,160,196]. In rats or dogs with compensated pre-HF, there is a mild decrease in systolic and diastolic function, while at established HF, there is moderate systolic and diastolic function [40,76,160,196]. We recently showed that in rats with end-stage HF, systolic function is decreased as measured by both echocardiography (% fractional shortening = 29%) and pressure-volume analysis (end-systolic elastance = 0.36) [40].

These rats have severe pump failure with accompanying pulmonary congestion and edema [40].

Cardiomyocyte dysfunction in VO

Excitation-contraction coupling

Excitation-contraction (E-C) coupling is required for cardiac muscle contraction.

Calcium enters the myocyte through L-type Ca2+ channels during the action potential and induces Ca2+-induced Ca2+ release [25,86]. Ca2+ binds to troponin C and initiates cross- bridge cycling and contraction. Contraction is terminated when Ca2+ is removed from the cytosol primarily by the sarcoplasmic reticulum Ca2+ ATPase (SERCA2a) and to a lesser extent, by extrusion across the plasma membrane via the sodium/calcium exchanger

(NCX) [25,86].

Altered Ca2+ homeostasis is a hallmark of pressure overload (PO) heart failure

[16]. Extensive changes in the expression and post-translational modification of the key

Ca2+ regulatory proteins ryanodine receptor (RyR2) and SERCA2a and phospholamban

(PLB) have been documented in PO [145,170]. For example, in the spontaneously hypertensive rat (SHR) and in rats with transverse aortic constriction (TAC), there is geometric derangement of L-type Ca2+ channels and RyR2 accompanied by dyssynchronous, multiphasic Ca2+ release [105,175,201]. This derangement may lead to defective E-C coupling, in part because the concentration of the initial Ca2+ trigger generated by the L-type Ca2+ channels decays by orders of magnitude with distance

[201]. Recent ultrastructural evidence in TAC rats showed that there was reduction in the size of the L-type Ca2+ channel-RyR2 junction and this may represent a novel mechanism underlying the decreased, desynchronized Ca2+ release in pressure-overload [201]. In

9 contrast, mechanistically distinct alterations in E-C coupling are observed throughout VO

HF progression. Earlier in compensated stages, alterations in E-C coupling are largely due to alterations in myofilament Ca2+ sensitivity, cross-bridge cycling, and electromechanical coupling across the intercalated disc (see below). These myofilament changes persist at end-stage HF, with no significant difference in the amount of Ca2+ available for contraction [66]. We and others have reported alterations in Ca2+ cycling at the late stages of VO HF progression (i.e. end-stage failure) in rats, dogs, and humans

[66,76,114,205]. In end-stage ACF rats, there was a reduction in SR Ca2+ release that was associated with decreased expression of RyR2 and decreased phosphorylation of

PLB without alterations in PLB or Serca2a expression [85]. Similarly McGinley et al

[114] reported that chronic MR in the dog had decreased systolic function (i.e. decreased

2+ dP/dtmax) that was associated with decreased myocyte contractility and peak Ca transient amplitude.

VO HF, intercalated discs and electromechanical transmission

Coordinated LV pump function requires effective electrical and mechanical coupling between individual myocytes. This coupling occurs at the intercalated disc, which is composed of desmosomes, fascia adherens junctions and gap junctions (Fig 2)

[54,58,165]. Desmosomes are intercellular junctions that act as anchors for cytoskeletal intermediate filaments, and they play a role in maintaining structural integrity especially in tissues undergoing constant physical stress [5,176]. Cardiac desmosomes contain tissue-specific cadherins (desmoglein 2, desmocolin 2), plakins

Figure 2: Effect of VO on intercalated disc proteins Mutations and/or alterations in protein level or phosphorylation have been described for these proteins (**). Most of these alterations decrease force transmission. CAR: Coxsackievirus and adenovirus receptor. Figure adapted from Sheikh F. Trends Cardiovasc Med. 2009 August; 19(6): 182–190.

(plakophilin), and armadillo proteins (plakoglobin (i.e. γ-catenin) and plakophilin 2)

[5,176,189]. Desmoplakin, plakoglobin and plakophilin interacts with the intermediate filament desmin, which binds to thin filament proteins [5,176,189]. Fascia adherens junctions link the cell membrane to the actin cytoskeleton, and include cadherins (N- cadherin), catenins (α-, β-, and γ-catenin), and catenin-related proteins (i.e. vinculin, metavinculin and α-actinin) [176].

Gap junctions/electrical communication

Collectively, gap junctions are nonselective, transmembrane channels that connect the cytoplasmic compartments of adjacent cells allowing passage of ions, second messengers, small molecules up to 1 kDa, and electrical current [25,165]. Gap junctions are composed of two hemichannels (connexons), which are composed of six identical

(homomeric) or different (heteromeric) subunits (connexins, Cx) surrounding a central pore [25]. Connexins have two extracellular domains, four helices that span the membrane, and an intracellular carboxy-terminal domain, which through sequence differences may contribute to the different biophysical properties for each of the connexins [62]. Connexin expression in the heart is diverse, with variable expression patterns amongst the atria (Cx40, Cx43), ventricles (Cx43), sinoatrial/atrioventricular

(SA/AV) nodes (Cx40, Cx30.2), and the conduction system (Cx40 and Cx45 in the

Bundle of His, Cx40, Cx43 and Cx45 in the bundle branches/Purkinje fibers) [54,62].

Gap junction remodeling has been identified in ischemic heart disease and heart failure as well as in arrhythmogenic diseases such as ARVC [158,173]. In these diseases,

Cx43 expression is reduced at the gap junction and, depending on the disease, increased on the lateral cell membrane [58,173]. Reduced Cx43 expression results in reduced electrical coupling and reduced conduction velocity, which can increase the likelihood of arrhythmias [58]. One proposed mechanism is increased reentrant activity [58].

Alterations in Cx43 in hypertrophic and ischemic cardiomyopathy have been reviewed in

[58].

In humans and pig, dog, and mouse models of dilated cardiomyopathy, there is a decrease in ventricular Cx43 expression, which in the dog is accompanied by decreased phosphorylation, lateralization, and an increase in the QRS interval [58]. In rats with

ACF-induced end-stage HF, Cx43 expression was reduced at the intercalated disk without evidence of lateralization, and this was accompanied by an increase in the QT interval and QTc without significant differences in QRS or PR interval [66]. These results suggest that decreased Cx43 expression may lead to reduced electrical coupling and contribute to LV pump dysfunction in VO.

Force transmission through the intercalated disc

In cardiac myocytes, force is transmitted between myocytes and between myocytes and the ECM (see below). Force generated at the level of the myofilament is transmitted from cytoskeletal attachments at the Z line to intercalated disc proteins in the fascia adherens and desmosomes. Much of what we have learned about altered force transmission comes from genetic studies of familial DCM-causing mutations in this pathway. For example, numerous single mutations in the desmosomal cadherins desmoglein-2 and desmocollin-2 are associated with altered force transmission [149]. An intact and dynamic cytoskeleton is required for force transmission between the sarcomere and the intercalated disc and plays a key role in cardiomyocyte contraction and relaxation

[116]. This cytoskeleton also connects signaling at the sarcolemma and nucleus with the contractile apparatus, and may in part, dictate the parallel versus serial addition of sarcomeres [116]. The intermediate filament protein desmin links adjacent sarcomeres

13 and myocytes by interacting at the Z-discs and intercalated discs, respectively, and also links the sarcomere with the plasma membrane, ECM and nucleus [116].

Mutations in desmosome proteins are most commonly associated with arrhythmogenic right ventricular cardiomyopathy (ARVC) and myofibrillar myopathies known as desminopathies [5,44,115,176]. In a cohort of 116 families plus 309 additional patients, the prevalence of desmin mutations in DCM patients was 1-2% [44,190]. DCM- linked mutations were present in the 1A helical domain as well as in the 2B rod domain

[190]. While some of these mutations resulted in disruption of desmin filament assembly, other mutations did not and the authors suggest that dysfunction of apparently intact desmin networks is sufficient to cause DCM [190]. Mutations in metavinculin

(R975W, L954del), a splice variant of vinculin, have been linked to DCM [132]. The hypothesized mechanism is disrupted force transmission at the intercalated disc-thin filament interface [132].

Myofilament dysfunction in VO

Force generation

Muscle contraction in the heart is generated through interaction of the thick and thin filaments following activation of the actomyosin Mg2+-ATPase (Fig 3) [110]. Thick filaments are composed of myosin, which is a hexamer composed of two heavy chains, two essential light chains, and two regulatory light chains [25,38,123]. These myosin filaments form the A band within the sarcomere and partially interdigitate with

Figure 3: Effect of VO on myofilament and myofilament-regulatory proteins Mutations and/or alterations in protein level or phosphorylation have been described for these proteins (*). Most of these alterations either impair myofilament Ca2+ sensitivity or alter the effect of β-adrenergic mediated Ca2+ sensitivity. Figure adapted from Morimoto S. Cardiovasc Res. 2008 Mar 1;77(4):659-66.

the thin filaments [25]. Actin, composed of a double helix of actin molecules, forms the thin filament backbone, which is associated with two regulatory proteins—tropomyosin and troponin [25,110]. These thin filaments are tethered together and project from the Z disc or the intercalated disc transmitting force between neighboring sarcomeres or

15 myocytes, respectively [25,133]. Troponin is a heterotrimer composed of troponin T

(cTnT), which binds to tropomyosin, troponin C (cTnC), which binds Ca2+, and troponin

I (cTnI), which binds to actin and inhibits contraction [25].

Force generation occurs through the five-step cross-bridge cycling. Following the previous cycle, ATP binds to the head of the myosin heavy chain (MHC), which causes the myosin head to be released from actin [25]. ATP hydrolysis then occurs, the myosin head pivots into the resting ―cocked‖ state, and the myosin filament advances ~11 nm or

2 actin monomers down the chain [25]. A new cross-bridge forms, followed by release of Pi from myosin, which triggers the power stroke [25]. Finally, ADP is dissociated from the myosin complex, leaving the actomyosin in a rigid state [25].

DCM-causing mutation in thin filament proteins

Two missense mutations (R312H and E361G) causing DCM have been identified in the α-cardiac actin gene [124,133]. These mutations occur in subdomains 1 (E361G) and 3 (R312H) of the actin monomer, which is present in the immobilized end of the actin filament [133]. The E361G mutation is present in a binding domain for α-actinin, which anchors thin filaments to the Z-discs and intercalated discs, and this mutation results in slightly reduced affinity of actin for α-actinin, suggesting a role for reduced force transmission in the pathogenesis of this mutation [124,133,179]. More recently, however, functional studies in mice with the E361G mutation have been performed. In these mice, there is little to no overt cardiac disease, and this mutation does not alter the

16 sliding speed, fraction of motile filaments or the Ca2+ sensitivity [179]. The primary functional change in these mice is that this mutation uncouples Ca2+ sensitivity from cTnI phosphorylation, which ultimately blunts the response to adrenergic stimulation manifesting as contractile dysfunction under stressful conditions [179].

DCM-related mutations have been identified in cTnT, cTnI and cTnC [38]. The largest number of mutations have been identified in cTnT [38]. The deletion ΔK210 in cTnT was identified in 2000 as the first cTnT mutation responsible for familial primary

DCM [38,110,124]. The effect of this deletion on force generation and ATPase activity was studied in rabbit cardiac skinned muscle fibers [125]. This mutation shifted the force-pCa curve to the right but did not decrease maximal force generation, and higher concentrations of free Ca2+ were required for ATPase activation [125]. Additionally in mice, this deletion results in cardiomegaly, heart failure and premature sudden death, and there is a rightward shift in the force-pCa for cardiac myofibers from these mice [53]. An additional 12 mutations in cTnT that are correlated with DCM have been identified, and these mutations affect both the tropomyosin and the cTnC-cTnI binding regions [110].

Nearly all of these DCM-causing mutations result in a Ca2+-desensitizating effect on actomyosin ATPase activity and on skinned fiber force generation [110]. This suggests that the primary mechanism for DCM dysfunction is Ca2+ desensitization of cardiac muscle contraction [110].

Four DCM-causing mutations have been identified in cTnI. The first mutation identified, A2V, caused DCM in an autosomal recessive manner by impairing the interaction of cTnI with cTnT [110]. Two additional mutations (K36Q, N185K) cause

DCM in an autosomal dominant manner by decreasing the binding affinity of Ca2+ to the regulatory site of cTnC [110]. A missense mutation, P16T, has been identified, but the functional effects are not known [110].

A missense mutation in cTnC, G159D, has been linked to an autosomal dominant form of DCM in a single family [121,150]. Unlike the mutations in cTnT and cTnI, this mutation does not decrease the Ca2+ sensitivity of force generation of skinned fibers; however, ATPase activity and filament sliding velocity are decreased in mutant myofilaments [110,150]. Biesiadecki et al [20] confirm that this mutation does not alter myofilament function; however, this mutation blunts the effect of β-adrenergic-like phosphorylation on myofilament function. Typically, phosphorylation of cTnI at

Ser23/24 induces 1) a rightward-shift in the force-pCa curve indicating reduced Ca2+ sensitivity and 2) enhanced cross-bridge cycling [20]. In mutant fibers with Asp substitution at Ser23/24, there was 1) no rightward shift in the force-pCa curve, demonstrating that this mutation nearly completely abolished the characteristic decrease in Ca2+ sensitivity, 2) unaltered cross-bridge cycling. Additional cTnC mutations have been discovered. In one DCM patient, two missense mutations (E59D and D75Y) were identified in the regulatory Ca2+ binding site of cTnC [102]. Expression of these mutations in isolated rat cardiomyocytes resulted in impaired myofilament Ca2+

18 sensitivity and decreased contractility despite normal intracellular Ca2+ homeostasis

[102]. These effects were ultimately attributed to the D75Y mutation, while the E59D mutation was functionally benign [102]. Finally, three additional mutations (Y5H,

M103I, and I148V) have been identified in DCM patients [146]. In skinned porcine fibers with the Y5H and M103I mutations, there is decreased Ca2+-responsiveness [146].

The effect of β-adrenergic-like phosphorylation on these mutations was investigated.

M103I completely abolished the effect of β-adrenergic mediated Ca2+ sensitivity, while the Y5H and I148V mutations only blunted the response [146]. Taken together, these data suggest that cTnC mutations may play a more important role in β-adrenergic responsiveness than in intrinsic basal cross-bridge cycling. This ultimately may lead to heart’s reduced ability to respond to autonomic activation.

In 2001, two missense mutations in α-tropomyosin (E40K and E54K) causing

DCM were identified [37,120,124]. Since then, an additional 10 DCM-causing missense mutations have been identified, and while most of these mutants reduce Ca2+ sensitivity, one α-tropomyosin mutation (D230N) increases Ca2+ sensitivity [155]. One common finding amongst the mutants is the inability to decrease thin filament Ca2+ responsiveness following PKA-mediated phosphorylation of cTnI [155]. In mice with the E54K mutation, there is LV dilation with decreased fractional shortening and progression into heart failure and death within 6 months [154]. These mice have a blunted lusitropic response to β-adrenergic stimulation. In detergent skinned fiber bundles, there is decreased Ca2+ responsiveness, and in LV tissue, there is an increase in the hypertrophic

19 markers β-MHC, BNP and skeletal actin mRNA [154]. In mice and humans with this mutation, the altered cardiac structure and function are thought to be due to impaired actin-tropomyosin binding together with decreased myofilament Ca2+ sensitivity [154].

Thick filament proteins in VO HF

In cardiac muscle, there are two isoforms of myosin heavy chain (MHC)—alpha and beta, and their relative expression is related to species, heart size and heart rate [196].

α-MHC is the predominant isoform in mice, rats and other rodents, while β-MHC is the predominant form in rabbits and larger species [196]. Chemically-skinned myocardium containing primarily α-MHC has a higher ATPase rate, faster sarcomeric shortening velocity, higher power production and greater rate of force development compared with

β-MHC [196]. Compared with β-MHC of the same species, isolated α-MHC has a higher

ATPase rate, faster velocity of actin motility and shorter crossbridge lifetime [196].

The mutations in β-MHC are the most prevalent of the DCM-causing mutations, and the 24 mutations currently described can be grouped into one of three groups—those present in the S1 region, those not present in the S1 region and those that transition from

HCM to DCM [38,202]. Mutations in the S1 region have been implicated in disrupted actin-myosin interactions, while mutations in the tail may affect structure [38]. To study the effects of two of the mutations, mice were created with two DCM-causing missense mutations (S532P and F764L) in α-MHC [168]. Homozygous mice had progressive LV dilation with decreased fractional shortening [168]. The mutant myofilaments from these

20 mice had similar force-generating capacity compared with the wild-type myofilaments, but isolated cardiac myosin had reduced ability to translocate actin, which was attributed to reduced ability of the motor to generate stepwise displacement [168]. More recently, the role of these mutations has been investigated in the context of an intact myofilament lattice in skinned strips of muscle from heterozygous mice [134]. As previously noted, there was a reduction in contractile function for the F764L heterozygote ventricle and skinned myocardium compared with the S532P heterozygote or the wild-type control

[134]. In contrast with the previous findings, differences were noted at the molecular level (i.e. higher MgATP binding rate for the DCM myocardial strips) and this difference may be due to differences in assay geometry and/or post-translational modification [134].

In hemodynamic VO models, the ratio of α-MHC to β-MHC decreases as early as

4 weeks post-ACF during compensated pre-HF and progressively decreases as the rat progresses towards end-stage HF [66,76]. Isoform phenotype was investigated in rats that had undergone surgical reversal of the ACF, which restores chamber geometry with delayed restoration of LV function. The α-MHC to β-MHC ratios were nearly completely restored at 11 weeks post-reversal compared with Sham [76]. Together, these results suggest that hemodynamic load influences MHC isoform expression. This isoform switch has previously been attributed to hemodynamic overload [71,187].

cMyBP-C is a thick filament assembly, accessory and regulatory protein that is a key regulator of cardiac contractility [163]. Unphosphorylated cMyBP-C modulates the

21 speed and force of muscle contraction by regulating the interactions between actin and myosin through constraint of the myosin head [17,151]. In hearts without cMyBP-C, there is biventricular dilation, depressed LV function, and reduced circumferential strain and LV torsion, suggesting the requirement for this protein in maintaining normal structure and mechanical function [67].

cMyBP-C is regulated through proteolytic degradation and phosphorylation. cMyBP-C phosphorylation loosens constraint on the myosin head, allowing strong actin- myosin crossbridges with a net increase in maximal in force generation, crossbridge cycling kinetics and myofilament relaxation [39,46,47,151]. Although multiple phosphorylation sites have been identified, the three sites most commonly studied are present in the M domain (Ser-273, Ser-282 and Ser-302), and in vitro studies have shown that these residues are differentially phosphorylated by the kinases PKA, PKC, PKD,

CamKII, CK2 and RSK [67]. PKA phosphorylates Ser-273, Ser-282 and Ser-302 in response to β-adrenergic stimulation [61,67,122]. By contrast PKC only phosphorylates

Ser-273 and Ser-302 and PKD phosphorylates Ser-302 [61,67,122]. Like PKA, CaMKII phosphorylates Ser-273, Ser-282 and Ser-302, but unlike PKA this phosphorylation occurs in a Ca2+ concentration dependent manner [67,92,164]. Finally, CK2 can phosphorylate Ser-282 [67,92]. Phosphorylation at these sites is coordinated, and phosphorylation alters cardiac function, presumably by altering cMyBP-C’s interactions with thick and thin filaments [67].

To evaluate the effect of the three phosphorylation sites, a series of mutant mice were developed in which phosphorylation sites were replaced by singly or in combination by amino acids that either acted as a phosphomimetic (aspartates) or could not be phosphorylated (alanines). In mice with chronic phosphorylation of Ser-273 and blocked phosphorylation at Ser-282 and Ser-302 (DAA mutant), there was biventricular dilation, myocyte hypertrophy, and interstitial fibrosis accompanied by decreased systolic function

[67]. Interestingly, this phenotype was more severe than cMyBP-C null mice [67]. In mice with chronic phosphorylation at Ser-302 and blocked phosphorylation at Ser-273 and Ser-282 (AAD mutant), there was mild LV dilation and hypertrophy, significant interstitial fibrosis and decreased systolic function [67]. Diastolic function was evaluated by measuring the tissue velocity of the mitral valve annulus during early relaxation (E’) and atrial contraction (A’). The E’/A’ ratio was significantly decreased in all cMyBP-C mutants consistent with diastolic dysfunction [67].

In rats with end-stage ACF-induced HF, total cMyBP-C protein was decreased along with an apparent decrease in phosphorylation at Ser-273, Ser-282 and Ser-302, suggesting that phosphorylation mechanisms remained intact [66]. The role that the loss of cMyBP-C plays in the dysfunction in these end-stage ACF rats is unclear. Congenital loss of cMyBP-C has been linked to LV dilation, cardiac hypertrophy, and systolic and diastolic dysfunction; however, loss of cMyBP-C in adult mice results in diastolic dysfunction and concentric hypertrophy [39].

Passive muscle stiffness in VO

Titin is a giant protein that extends from the Z-disc to the M line, spanning the length of the entire sarcomere [38,64]. Titin plays a role in passive muscle stiffness, assembly of contractile proteins, force transmission at the Z-line as well as protein metabolism including positioning the T-tubules and SR, binding of chaperones, and compartmentalization of metabolic enzymes [38,64]. Several mutations in the titin gene

(TTN) have been associated with DCM. The DCM-causing mutation V54M results in decreased affinity of the titin Z1-Z2 domains for telethonin (or the T-cap) [124]. A different DCM-causing mutation, A743V, results in decreased affinity of the titin Z- repeat region for α-actinin [124]. Finally, the S3799Y mutation linked to DCM increases the affinity of the titin binding region for four and a half LIM protein 2 [124]. This protein is known to bind metabolic enzymes and Morimoto et al [124] suggest that this increase in metabolic enzymes may be important for the pathogenesis of cardiomyopathy.

Extracellular matrix (ECM) alterations in VO

ECM turnover

The ECM is a highly adaptive structure composed of a mixture of collagen, elastin, fibroblasts, macrophages, glycoproteins and glycosaminoglycans, and the ECM surrounds and interconnects myocytes, myofibrils, muscle fibers and the coronary microcirculation [28,83]. Cardiac fibroblasts produce an extensive network composed of fibrillar collagens (primarily types I (85%) and III (15%)), laminin and fibronectin, which maintain the structural framework of the myocardium [197]. ECM composition and

24 turnover is regulated by mechanical stress, neurohormonal activation, inflammation and oxidative stress through contributions from cardiac fibroblasts, myocytes, mast cells and inflammatory cells [77]. These cells secrete matrix metalloproteinases (MMPs) that degrade collagen, secrete tissue inhibitors of MMPs (TIMPs) and secrete other proteolytic enzymes which regulate ECM turnover, like plasminogen activator-1 (PAI-1), which inhibits MMP activation [77].

VO is characterized by chamber dilation and a relative decrease in wall thickness

[117]. ECM composition has been studied in the ACF model in rats. In the first 12 hours to 7 days following ACF surgery, there is rapid ECM degradation (50% decrease in total collagen at 2 days) followed by matrix deposition with changes in collagen isoform from collagen type I to collagen type III with increased elastin and fibronectin and organization by 5 days post-ACF [159,160]. From 3-10 weeks post-ACF (the compensated phase), there is a net loss of ECM turnover through an increase in MMPs and a decrease in TIMPs, as well as a continued isoform shift from collagen I to collagen

III [77]. At end-stage (15-21 weeks post-ACF), increased ECM turnover continues, however, there is net collagen deposition [77,83].

Changes in ECM composition are translated to the myocyte via interaction with integrins at the focal adhesion complex. These complexes mediate mechanical force transmission between the myocyte and ECM [43]. In addition to the ECM and integrins, these focal adhesion complexes are associated with cytoskeletal adaptor proteins such as

25 paxillin, filamin, talin, vinculin and the signaling proteins GTPase RhoA and focal adhesion kinase (FAK) [43]. In adult mouse hearts, myocyte-restricted deletion of β1 integrin results in DCM, while cardiac myocyte-specific inactivation of the non-receptor tyrosine kinase FAK in mice results in eccentric hypertrophy [172]. Additionally, in dogs with MR, there was a decrease in collagen content in LV tissue, which was associated with decreased FAK phosphorylation and decreased FAK interaction with the adapter and cytoskeletal proteins p130Cas and paxillin [162]. Β1-receptor blockade prevented these signaling alterations without restoring ECM accumulation or improving

LV remodeling [162]. Taken together, these data highlight the importance of the interaction between the ECM-myocyte for both matrix composition and myocyte function.

Mechanical stress-induced ECM remodeling

In acute VO, the earliest changes are increases in hemodynamic and mechanical load, which result in increased myocardial stress and diastolic strain [77,117]. The different cell types in the myocardium seemingly respond to these mechanical stimuli in a cell-specific manner, and the combination of their responses alters the balance between pro and anti-fibrotic mechanisms [77]. Mechanical stretch of neonatal rat cardiomyocytes increases MMP-2 and MMP-14 expression, and this increase is blocked by pretreatment with losartan (angiotensin II AT1 receptor antagonist) and AG-490 (Janus kinase 2 tyrosine kinase inhibitor), suggesting this increase occurs through the angiotensin II-

JAK-STAT pathway [195]. Stretch in atrial myocytes induces up-regulation of MMP-2,

MMP-9, and MT1-MMP, which was abolished by treatment with losartan, cyclosporine

A (blocks calcineurin (Cn)) and 11R-VIVIT (blocks NFAT, nuclear factor of activated T cells), suggesting the AT1-Cn-NFAT pathway is a potential mediator of MMP activation

[166]. Cardiac fibroblasts also rapidly respond to mechanical stretch. In neonatal rat cardiac fibroblasts, cyclic uniaxial stretch increases collagen III mRNA and protein at 24 hours, while collagen I mRNA and protein remain unchanged [35]. By contrast, collagen

I and collagen III mRNA increases in fetal rat cardiac fibroblasts stretched cyclically for

48 hours [31]. In adult rat cardiac fibroblasts cultured in serum-free conditions, cyclic stretch caused a net increase in collagens 1 and III, MMP-2, and TIMP-2 mRNA, although there was a significant decrease in collagen I/III mRNA in the presence of serum [75]. Taken together these results seem to indicate that the neonatal cardiac fibroblasts and cardiac myocytes play opposing roles in ECM homeostasis in response to stretch. The cardiac fibroblast response favors net ECM accumulation and expression of pro-fibrotic factors, while the cardiomyocyte response degradation of ECM through increased MMP expression.

Work has been done in our lab with adult rat cardiac fibroblasts from sham and

ACF rats at 4 weeks post-ACF. Our preliminary results show that primary cultures of cardiac fibroblasts from ACF rats display a distinct ―anti-fibrotic‖ phenotype compared to sham cardiac fibroblasts, characterized by decreased protein expression of α-SMA, collagen and CTGF and increased expression of miR-29b and anti-fibrotic peroxisome proliferator-activated receptor-γ (PPAR-γ) and apelin/APJ components (unpublished

27 observations). Interestingly, anti-fibrotic cardiac fibroblasts still appear to have an intact canonical pro-fibrotic TGFβ1 signaling pathway, as measured by Smad2/3 phosphorylation, yet unlike sham cardiac fibroblasts, TGFβ1-treatment of ACF cardiac fibroblasts does not increase expression of pro-fibrotic markers. Our data strongly suggest that the ACF-cardiac fibroblast phenotype does not result from myofibroblast reversion to a quiescent state, but from conversion of quiescent cardiac fibroblasts to an anti-fibrotic state.

Mast cells and VO-induced ECM remodeling

In human hearts with dilated cardiomyopathy and in hearts from animal models of hypertension, myocardial infarction and chronic volume overload, increased numbers of mast cells have been reported [99,100]. Mast cells are present in a variety of tissues, and their phenotype is tissue-specific [156]. Cardiac mast cells produce and store a variety of enzymes, including chymase and tryptase [99]. Tryptase has been shown to activate matrix metalloproteinase 1 (MMP-1; collagenase) and MMP-2 (stromelysin) in peritoneal and cutaneous mast cells in vitro, while chymase has been shown to activate MMP-2 and

MMP-9 in these same mast cells [99]. Finally, TNF-α has also been shown to activate

MMPs [99]. In addition to compounds that activate MMPs, mast cells release factors that can activate the renin-angiotensin system (RAS) locally through the release of renin and subsequent conversion of angiotensin I (Ang I) to angiotenin II (Ang II) by chymase

[156]. Ang II is known to regulate cell communication, impulse propagation and cardiac contractility as well play a role in cardiac remodeling, apoptosis and growth [50].

Much of the work describing the role of mast cells in VO HF has been performed in ACF rats. Levick et al [99] reports that at 3-5 days post fistula there is an increase in mast cell density in the LV in concert with an increase in MMP-2 activity and reduction in collagen volume fraction. In rats given the mast cell stabilizer cromolyn sodium or in mast-cell deficient rats, there was attenuated to absent remodeling, respectively [156] further confirming the role of mast cells in the initial remodeling. Additionally, in dogs with MR, mast cell density was increased 2 and 4 weeks post-MR, which correlated with an increase in MMP-2 activity and increase in collagen degradation [156]. The similarity of the results in dogs with MR and rats with ACF suggest that mast-cell mediated remodeling is not limited to a specific disease model or species.

Although increasing evidence supports the pathologic role of mast cells, little is known about the origins of these cells or the exact mechanisms that trigger mast cell activation. Sources for these cells include proliferation of resident mast cells, maturation of immature mast cells and/or recruitment of hematopoietic precursors [100,156]. Of these three options, maturation is the most likely source of these mast cells. Only a small percentage (1%) of mast cells stain positive for BrdU and following depletion from the peritoneal cavity, it reportedly takes 20 days for full restoration of tissue mast cells [100].

Recently, stem cell factor (SCF)-driven maturation has been shown to account for the increase in mature mast cells [100]. SCF can be released from fibroblasts, endothelial cells and smooth muscle cells. Several mast cell secretagogues have been identified and these include endothelin-1, IL-33, complement factor 5a, reactive oxygen species and

29 other neuropeptides including substance P and neurotensin [99]. Murray et al. [127] implicates endothelin-1 (ET-1) as a causative factor in mast cell degranulation by showing that in blood-perfused isolated rat hearts given ET-1 there is significant cardiac mast cell degranulation, MMP-2 activation and collagen degradation. They have also shown that treatment with bosentan, a non-selective endothelin receptor antagonist, prevented an increase in mast cell density. Preliminary studies by Levick et al [99] have shown that IL-33 activates isolated rat cardiac mast cells to an equal degree as compound

48/80, a known mast cell secretagogue.

Mast cells and paracrine-induced LV ECM remodeling

In rats, acute ACF-induced VO results in a significant increase in interstitial angiotensin II (ANG II) and bradykinin (BK), and treatment with an ACE inhibitor exacerbates LV eccentric remodeling and collagen loss with an increase in mast cell density, without functional improvement [160,200]. The addition of a BK2 receptor antagonist did not reverse the antihypertrophic effects of the ACE inhibitor nor did it improve blood pressure [200]. BK2 receptors have been identified on mast cells and

BK2 receptor blockade prevents mast cell accumulation in the rats 5 days post-ACF induction and, in normal mice, BK2 blockade prevents ACE-inhibitor mediated increases in LV ISF chymase activity [199,200]. Recent evidence suggests that early VO is associated with kallikrein-kinin upregulation and increased interstitial bradykinin [198].

This increased bradykinin mediates mast cell infiltration, followed by ECM loss and LV dysfunction [198]. By contrast in later stages (15 weeks ACF), there is continued LV

30 dysfunction with an increase in LV interstitial catecholamines and perivascular fibrosis

[198].

Pharmacologic therapy in VO

Many different classes of drugs have been used to treat symptoms and dysfunction in patients with VO. These therapeutic options have had variable success and in human patients, the preferred treatment option is surgical repair. In veterinary species (i.e. dogs with myxomatous mitral valve disease (MMVD)) surgical options are limited and pharmacologic therapy is more commonly used.

Diuretics

Clinically used diuretics include the loop diuretics (i.e. furosemide), thiazides (i.e. bendrofluazide), and potassium-sparing diuretics (i.e. amiloride, spironolactone)

[15,48,69]. Diuretics are primarily used for symptomatic relief and effective at treating pulmonary edema; however, evidence of prognostic benefit is scarce [48,141]. In dogs with MR, furosemide reduced left atrial pressure [186]. In dogs with MMVD or DCM, spironolactone was well-tolerated but provided no survival benefit [169].

Vasodilators

Clinically used vasodilators include hydralazine, prazosin, nitroglycerin, nitroprusside, and ACE inhibitors. Vasodilator therapy improves LV function by reducing the regurgitant volume [56]. In patients with aortic regurgitation, vasodilators

31 reduce LV wall stress and have been shown to delay or eliminate the need for valve replacement [7,51,147]. ACE inhibitors reduce systolic blood pressure, systemic vascular resistance as well as left atrial pressure [80]. The efficacy of ACE inhibitors is variable. Several reports indicate that ACE inhibitors have been shown to improve hemodynamics, clinical signs, exercise tolerance and delay the onset of congestive HF in dogs or humans with MR [8,70,97,182], while others indicate that VO hypertrophy and interstitial collagen loss are not improved by ACE inhibitors [52].

β-adrenergic antagonists

Chronic stimulation of the β-adrenergic system, which occurs in chronic HF, is toxic to the heart and the heart compensates with β-AR desensitization [98]. β-adrenergic antagonists (β-blockers) are used to relieve the inhibition and improve LV function and

β-adrenergic responsiveness in MR [2,139]. Although β-blockers improve LV function, they did not attenuate LV remodeling with respect to cell length and ECM loss [139].

Myofilament Ca2+ sensitizers

Myofilament Ca2+ sensitizers include levosimendan, which is approved for the treatment of acute HF in humans, and pimobendan, which is approved for use in dogs with MMVD or DCM and for use in humans (in Japan only) with acute or chronic HF

[26,188]. Levosimendan interacts with the Ca2+-saturated cTnC prolonging interaction between cTnC and cTnI [136]. This promotes contractile force without increasing the amplitude of the Ca2+ transient [136]. Pimobendan is reported to increase the affinity of

32 cTnC regulatory site for Ca2+, but in contrast to levosimendan, pimobendan’s inotropic effects are also derived from phosphodiesterase 3 (PDE3) inhibition [26]. Levosimendan has been shown to improve LV function in acute decompensated HF, pre- and post- operatively and in chronic HF, but no prospective studies have evaluated levosimendan in patients with VO HF [14,45,82,96,180]. In dogs with congestive heart failure from

MMVD or with mitral regurgitation from experimentally-induced chordal rupture, pimobendan improves short-term cardiac function, which is accompanied by decreased left atrial pressure, smaller heart size and less retention of free water [68,69,185].

Myosin activators

Omecamtiv mecarbil, the prototype of this class, increases systolic ejection time, which results in improved stroke volume, fractional shortening and hemodynamics, by increasing the rate of effective myosin cross-bridge formation [191].

Phosphodiesterase inhibitors

Phosphodiesterase inhibitors improve systolic function by inhibiting breakdown of cAMP, which results in an increase in protein kinase A (PKA) activity, which promotes opening of L-type Ca2+ channels. Milrinone inhibits phosphodiesterase 3, while sildenafil inhibits phosphodiesterase 5. Although in patients with DCM milrinone improves myocardial contractility (dP/dtmax) and promotes reduction of systemic and pulmonary vascular resistance, milrinone also increases intracellular Ca2+ and cAMP, which results in increased heart rate, hypotension, arrhythmias and mortality [27,191]. In

33 rats with MR, sildenafil improved ejection fraction and exercise tolerance compared to untreated MR rats [88].

Limitations and future directions

Compared with pressure overload and ischemia, the pathophysiology and pathologic alterations in volume overload heart failure are vastly under studied.

Moreover, much of the work performed is in animal models, primarily in rodents that have very different cardiac physiology from larger animals and humans. Because of this, the results from the studies in transgenic and hemodynamic models in rodents need to be validated against large animal models and human clinical trials. Additionally, evidence for pharmacologic intervention in VO HF is limited. Current treatments for heart failure are largely derived from patients with hypertension and coronary artery disease. For example, traditional pharmacologic therapies include the ACE inhibitors and β-blockers.

In dogs with MR, LV dysfunction is improved by β-adrenergic receptor blockade, but not

ACE inhibition or angiotensin type 1 receptor blockade [162]. These limitations provide many avenues for future research, including developing refined animal models, investigating pharmacologic and/or surgical interventions, and investigating new methods to measure cardiac function (i.e. incorporating strain analysis).

Chapter 2: Levosimendan preserves cardiac function in rats with ACF-induced pre-

HF and reversed pre-HF

Abstract

Aortocaval fistula (ACF)-induced volume overload (VO) heart failure (HF) results in progressive left ventricular (LV) dysfunction. This study investigated whether early intervention with the myofilament Ca2+ sensitizer levosimendan (Levo) could preserve cardiac function during compensated pre-HF and improve function in ACF-VO rats following hemodynamic load reversal. ACF or Sham surgery was performed in male

Sprague-Dawley rats (200-240g) 4 weeks before reversing (REV, closing) the shunt in a subset of rats. From weeks 4-8, ACF and REV rats were given vehicle (Veh; water) or levosimendan (Levo, 1 mg/kg) in drinking water. Levo improved systolic (%FS, Ees,

PRSW) and diastolic (tau, dP/dtmin) function in rats with pre-HF and reversed pre-HF.

Additionally, Levo treatment 1) improved myofilament Ca2+ sensitivity without significantly affecting the amplitude and kinetics of the intracellular Ca2+ transient; and

2) preserved in vivo β-adrenergic responsiveness. In ACF-Levo, increased phosphorylation of cMyBP-C Ser-273 and Ser-302 and cTnI Ser-23/24 correlated with improved diastolic relaxation, while in REV-Levo, increased phosphorylation of cMyBP-

C Ser-273 and increased α-to-β-MHC correlated with improved diastolic relaxation.

Levo improves systolic and diastolic function with and without surgical intervention, and 35 myofilament composition and regulatory protein phosphorylation likely play a key role in improving function.

Introduction

Mitral regurgitation (MR) is the most and second most common valve lesion in the United States and Europe, respectively, affecting >2 million Americans [55,65].

MR’s pathophysiological consequences include chronic left ventricular (LV) hemodynamic volume overload (VO) followed by LV chamber dilation, progressive LV contractile dysfunction and heart failure (HF). Pharmacologic therapy may prolong the time to surgical intervention, or improve function in post-operative patients with continued dysfunction. VO-HF therapeutic options generally target neurohormonal pathways by disrupting receptor-ligand interactions or modulating downstream signaling pathways (e.g. Ca2+-cAMP). Although these therapies successfully manage LV hypertrophy, their inotropic actions do not directly target impaired LV contractility, the central feature of systolic HF [60]. This can result in increased myocardial oxygen consumption and myocardial Ca2+ overload [140].

Therapeutics targeting myofilament activation, including myofilament Ca2+ sensitizers (e.g. levosimendan, Levo) and myosin activators (e.g. omecamtiv mecarbil,

OM), are used to treat acute and chronic HF [14,82,180]. Levo stabilizes Ca2+-saturated troponin C (cTnC) prolonging its interaction with cardiac troponin I (cTnI), which promotes contractile force without increasing the intracellular Ca2+ transient amplitude or

36 myocardial oxygen consumption [126,128,136]. Despite Levo’s common clinical use in

Europe, its effects in specific VO models are not well-characterized.

Diastolic relaxation is largely controlled at the myocyte level through Ca2+ cycling, myofilament composition, and phosphorylation [24], and at the myofilament level by Ca2+ desensitization and increased crossbridge cycling kinetics [192]. Increasing evidence supports a role for phosphorylation of the myofilament regulatory proteins cardiac myosin binding protein-C (cMyBP-C) and cTnI in contraction and relaxation kinetics [10,47,63,95,142,178]. cTnI phosphorylation regulates myofilament crossbridge cycling kinetics and Ca2+ sensitivity, while cMyBP-C phosphorylation appears to accelerate cross-bridge cycling kinetics, increasing force development and subsequent relaxation [10,47,63,95,142,178]. Decreased cMyBP-C expression and altered cTnI phosphorylation are associated with pressure overload-induced and post-infarct-induced

HF in animals and humans [10]. Little is known about the role of altered cMyBP-C and cTnI phosphorylation in VO-induced diastolic dysfunction.

Despite VO’s clinical importance, few models mimic the pathophysiological progression of chronic VO and fewer studies evaluate pharmacologic intervention. The aortocaval fistula (ACF) model of VO-HF in the rodent mimics increased hemodynamic preload observed in human disease irrespective of etiology. In this model, chronically increased LV preload leads to progressive LV pump failure, which is classified into three clinically relevant stages: pre-HF (4 wks ACF; marked LV dilation, increased LV wall

37 stress, mild LV dysfunction), established HF (8 wks ACF; severe LV dilation, significant

LV systolic and diastolic dysfunction) and end-stage HF (15-21 wks ACF; pump failure with pulmonary congestion/edema) [66,76,160]. This model is used to identify molecular and cellular mechanisms driving disease progression and to test novel therapeutic approaches for improving LV structure and function [59,84,198]. Our technique to reverse (REV, close) the ACF allows us to model LV structural and functional changes occurring after hemodynamic load reduction [76]. Reversal during pre-HF (i.e. 4 weeks post-ACF) results in rapid structural remodeling but delayed functional recovery [76].

This mirrors human data where structure, but not function, returns post-surgically in 17% of MR patients [112]. Thus, our model provides an attractive system for studying the physical and biochemical mechanics of VO following surgical and/or pharmacologic intervention.

In this study, we determined if Levo would improve LV function in ACF rats with pre-HF or reversed pre-HF. Our results provide the first evidence that Levo improves systolic function in VO-HF through myofilament Ca2+ sensitization. Levo also improves diastolic function, which correlates with increased cMyBP-C and cTnI phosphorylation in ACF and increased cMyBP-C phosphorylation and increased α-to-β-myosin heavy chain (MHC) in REV.

Materials and Methods

ACF Surgical Model in Rats

Male Sprague-Dawley rats (200-240g; Harlan, Charles River) were housed in a temperature and humidity controlled room using a 12h light/dark cycle with access to standard rat chow and water ad libitum. Studies conformed to the principles of the

National Institutes of Health ―Guide for the Care and Use of Laboratory Animals,‖ (NIH publication No. 85-12, revised 1996). The protocol was approved by the Institutional

Animal Care and Use Committee of The Research Institute at Nationwide Children's

Hospital.

Cardiac VO was induced in rats at Week 0 (Fig 4) under isoflurane anesthesia (2-

2.5%) as previously described [66,76]. Following ventral abdominal midline incision, the abdominal aorta and caudal vena cava were exposed. Cranial and caudal to the fistula site, the adventitia was bluntly separated and 5-0 Ethilon® suture (Ethicon, Cincinnati,

OH) was pre-placed for reversal. An 18g short-bevel needle (Becton Dickinson, Franklin

Lakes, NJ) was inserted through the abdominal aorta into the vena cava creating an aortocaval fistula. The aortic puncture was sealed with cyanoacrylate glue, and the pre- placed suture was loosely tied. Shunt patency was confirmed by visualizing red arterial blood in the vena cava. The abdominal wall and skin were closed with 4-0 chromic gut

Figure 4: Experimental time course ACF was induced at Week 0. In a subset of rats, the ACF was reversed (REV) at Week 4. Subsets of ACF and REV rats were given Levo (1 mg/kg/day in drinking water) from Weeks 4-8. Echo was performed biweekly and hemodynamics, myocyte isolation and tissue collection were performed at Week 8.

fibrovascular tissue bluntly dissected and the previously placed 5-0 Ethilon® suture was ligated. The abdominal wall and skin were closed as above. Shunt closure was defined the lack of arterial blood in the vena cava and/or a >1 mm decrease in LV end-diastolic diameter (LVEDD) as measured by echocardiography. The absence of pedal withdrawal reflex and respiratory rate were monitored during both procedures to ensure adequacy of

40 anesthesia. Buprenorphine (0.03 mg/kg SC) was given immediately post-operatively and every 12 hours for 72 hours post-operatively as needed for pain.

Radiotelemetry

Radiotelemetry catheters (PhysioTel PA-C40; Data Sciences International, St.

Paul, MN) were implanted into a subset of rats immediately following the initial Sham or

ACF surgery. Following ventral neck incision, the right carotid artery was isolated, and the catheter was inserted into the carotid artery to the aortic arch. A subcutaneous pocket was created dorsally over the right shoulder, the radiotransmitter was fed into this pocket, and the neck incision was closed with 5-0 Ethilon®. Blood pressure was recorded in conscious rats using Dataquest A.R.T (Data Sciences International, St. Paul, MN) acquisition software for two days prior to and in select 15 minute intervals during chronic

Levo treatment. Final analysis averaged data from 24 hour periods.

Levosimendan treatment

Four weeks post-ACF induction, rats were given vehicle (Veh; water) or Levo (L- enantiomer of ([4-(1,4,5,6-tetrahydro-4-methyl-6-oxo-3-pyridazinyl)phenyl]-hydrazono)- propanedinitrile; LKT Laboratories, MN) for an additional 4 weeks. Levo (0.0133 mg/mL) was prepared twice weekly in drinking water to deliver a dose of ~1 mg/kg/day

(Fig 4). This dose has improved systolic and diastolic function in previous rodent studies and did not adversely impact water consumption [19,107,108].

Echocardiography

Transthoracic echocardiograms (n=15-24/group) were performed biweekly with an 8.5-mHz transducer (Xario, Toshiba, Tustin, CA) under isoflurane anesthesia (1.5-2%)

[66,76]. M-mode images were obtained in parasternal short axis at the level of the papillary muscles to assess chamber diameters in end-systole (LVESD) and end-diastole

(LVEDD) and LV posterior wall thickness in systole (PWTs) and diastole (PWTd).

These indices were calculated: % Fractional shortening (%FS) = (LVEDD-

LVESD)/LVEDD x 100; Dilation index = (2*PWTd)/LVEDD.

Following baseline echo at Week 8, dobutamine (0.5 mg/kg IP) was given. M- mode images were obtained 2.5-5 min post-injection; %FS at the peak increase was analyzed.

Measurement of LV hemodynamics

LV hemodynamics were assessed invasively using pressure–volume (PV) analysis during Week 8 [66,76]. Rats were anesthetized with 3% isoflurane, intubated by tracheostomy, ventilated with a pressure-controlled rodent ventilator (SAR-830, CWE

Inc, Ardmore, PA), and maintained under 1.75% isoflurane anesthesia. The PV catheter

(1.9F, Transonic SciSense, London, ON) was introduced into the LV via the right carotid artery. Following equilibration, baseline LV hemodynamic parameters were acquired using 5-10 consecutive PV loops. Preload was varied by brief occlusion of the vena cava to obtain preload-recruitable stroke work (PRSW) and end-systolic elastance (Ees), the

42 slope of the end-systolic pressure volume relationship (ESPVR). Data was acquired and analyzed using Labscribe 2 software (iWORX, Dover, NH).

Measures of LV systolic and diastolic function include stroke volume (SV), heart rate (HR), maximum and minimum dP/dt, end-systolic and end-diastolic volume (ESV and EDV), end-systolic and end-diastolic pressure (ESP and EDP), Ees, PRSW, and relaxation constant (tau Weiss). Analysis of covariance (ANCOVA) adjusted marginal means of Ees and PRSW are presented and account for changes in the volume-axis intercept [30].

LV myocyte isolation

Hearts were removed from rats anesthetized with 3% isoflurane for LV myocyte isolation during Week 8 [66]. The heart, mounted on a Langendorff apparatus, was perfused with Tyrode’s solution (131mM NaCl, 4mM KCl, 10mM HEPES, 1mM MgCl2,

1mM CaCl2 and 10mM glucose) supplemented with 10mM 2,3-butanedione monoxime

(BDM), followed by perfusion buffer and then digestion with trypsin/Liberase-TH. LV myocytes were mechanically dispersed in perfusion buffer containing 12.5μM CaCl2 and

10% FBS, filtered, and resuspended in increasing concentrations of CaCl2 to achieve a final concentration of 1mM CaCl2. Isolated myocytes were incubated on laminin-coated chambers in plating media for 1 hour followed by culture media [66] for 1 hour.

Measurement of sarcomere shortening and Ca2 +-transients in isolated LV myocytes

Single myocyte auxotonic sarcomere shortening (field stimulated at 1 Hz) was measured within 2-3 h of initial plating. The following parameters were recorded with an

IonOptix Myocam and analyzed with IonWizard software (IonOptix Corporation, Milton,

MA): sarcomere peak shortening normalized to resting sarcomere length (%PS,), and cell shortening (-dL/dt) and relengthening (+dL/dt) maximal velocities. To evaluate Ca2+, a subset of myocytes was loaded with Fura-2A (1μM, 20-25°C; Molecular Probes, Eugene,

OR) for 20 min, and these myocytes were excited with light in an interleaved pattern at

340/12 nm and 380/12 nm during field stimulation and emission was collected at 510/40

2 + 2 + 2+ nm. Data are expressed as peak Ca amplitude normalized to resting Ca (peak[Cai ]),

2 + 2+ Ca available for contraction (AC/PK), and Ca decay leading to relaxation (AR/PK).

Myofilament Force-pCa

Myocyte fragments were isolated from frozen LV samples by mechanical dissociation and chemically permeabilized (skinned) with 0.3% Triton-X100 as described

[4,13], stored on ice and used within 20 hours of isolation. Opposing ends of cell fragments were attached to a force transducer and a Piezo actuator, and sarcomere length visualized by video micrometry was adjusted to 2.10 µm. The attached cell fragment was exposed to solutions of varying activator [Ca2+]. Active force development was fit to a modified Hill equation that yields maximum force development, Hill coefficient (a

2+ measure of cooperativity), and pCa50 (–log[Ca ] where 50% of maximum force is developed, a measure of myofilament Ca2+ sensitivity).

Immunoblot analysis

LV tissue lysates were prepared as described [66,76]. Proteins (16 μg) were separated by SDS-PAGE and transferred to PVDF membranes. Immunoblotting was performed with antibodies against cTnI (1:2000, Cell Signaling, Beverly, MA), phospho-

TnI Ser23/24 (1:1000, Cell Signaling), cMyBP-C (1:5000), and phospho-cMyBP-C Ser-

273, Ser-282 and Ser-302 (1:5000) [63]. Relative band densities were analyzed using

BioRad Image Lab Software (BioRad, Hercules, CA) and normalized to total ERK1/2 as loading controls [66,76].

Quantitative real-time PCR analysis

RNA was isolated from LV tissue using commercially available kits and protocols

(Qiagen RNeasy microarray tissue mini kit; Qiagen Inc, Valencia, CA). RNA was reverse transcribed using the RevertAid first strand cDNA synthesis kit and protocol (Fermentas;

Thermo Scientific, Hanover, MD). First strand equivalent to 100 ng input RNA was amplified in duplicate for each animal using appropriate Roche universal probe/primer pairs for each target gene: α-myosin heavy chain (MHC), β-MHC ( see [95] for primer sequences) and Maxima Probe quantitative real-time PCR master mix (Fermentas).

Amplifications were carried out for 40 cycles using an Eppendorf MasterCycler-ep

Realplex thermocycler. Parallel amplifications using nonreverse transcribed samples were performed to rule out genomic DNA contamination. Data were analyzed for relative expression using the 2−ΔΔCt method, with ribosomal protein Rpl13a serving as the

45 internal control and the average sham value at each time point serving as a second normalizer.

Statistical analysis

Data are expressed as mean±SEM. Statistical analyses were performed using

GraphPad Prism V6.0 or SPSS V19 (ANCOVA only). One-way or two-way ANOVA or

ANCOVA, followed by Bonferroni's post-hoc test, was used to measure differences between groups. P < 0.05 was considered significant.

Results

Levo does not significantly alter LV chamber morphology

Compared with Sham, LVEDD and LVESD were increased in ACF-Veh and

ACF-Levo; LVEDD and LVESD were comparable to Sham at 6 and 8 weeks in REV-

Veh and REV-Levo (Fig 5; Table 1). Since PWTd remained largely unchanged, the dilation index was significantly lower in ACF but normalized in REV, indicating eccentric hypertrophy and reverse remodeling, respectively (Table 1). EDV and SV were significantly increased in ACF, but reduced in REV (Table 1). In ACF-Veh and ACF-

Levo, these increases were accompanied by increased heart and lung weights (Table 1), consistent with pathologic LV hypertrophy and pulmonary congestion/edema. By contrast, heart and lung weights normalized in REV-Veh and REV-Levo.

Figure 5: Levo improves %FS but not LVEDD A. Representative M-mode images at Week 8. B. LVEDD and %FS. Data are means±SEM (n=15-24/group). *p<0.05, ***p<0.001, ****p<0.0001 vs. Sham-Veh; ^p<0.05, ^^ ^^p<0.0001 vs. ACF-Veh; $$$$ p<0.0001 vs. REV-Veh.

Levo improves LV systolic and diastolic function without altering β-adrenergic responsiveness

%FS was measured by serial echocardiography (Fig 5). At 2-4 weeks post-ACF,

%FS was significantly lower in ACF compared to Sham. Levo significantly improved

%FS (8% increase compared with Week 4, p<0.05), while %FS continued to decline in

ACF-Veh and REV-Veh. During Week 8, load-independent measures of systolic and diastolic function were obtained through pressure-volume analysis (Fig 6, Table 1). As 47

Figure 6: Levo improves systolic and diastolic function A. Representative pressure-volume loops. LV systolic (B) and diastolic (C) functional measurements. Data are means±SEM (n=7-10/group). *p<0.05, ***p<0.001, ****p<0.0001 vs. Sham-Veh; ^p<0.05, ^^^p<0.001 vs. ACF-Veh; $ p<0.05, $$ p<0.01, $$$$ p<0.0001 vs. REV-Veh 48 previously reported [76], Ees and PRSW were decreased in ACF-Veh and REV-Veh vs.

Sham. Levo significantly increased Ees and PRSW over ACF-Veh (67% and 89%, respectively) and REV-Veh (65% and 71%, respectively). In ACF-Veh and REV-Veh, there was significantly impaired diastolic relaxation, (increased tau, a less negative dP/dtmin; Fig 6c, Table 1), which was improved by Levo.

Decreased β-adrenergic responsiveness occurs in end-stage VO [66]. To determine if Levo alters β-adrenergic responsiveness, we measured %FS following dobutamine injection at Week 8. Sham, REV-Veh and REV-Levo had comparable responses to dobutamine (54±3%, 54±6% and 49±5% increase, respectively; Fig 7), which was mildly blunted but not significantly different in ACF-Veh and ACF-Levo

(42±4% and 46±4% increase, respectively).

Figure 7: Beta-adrenergic responsiveness is preserved in ACF and REV Dobutamine (0.5 mg/kg IP) was given during Week 8 and the increase in %FS was measured. Data are means±SEM (n=11-12/group). ****p<0.0001.

Levo does not significantly alter mean arterial pressure

Levo reportedly causes vasodilation and hypotension [136]. To determine if these effects occurred with our Levo dose, we measured blood pressure in conscious rats with radiotelemetry. Mean arterial pressure (MAP) was comparable (p>0.05) amongst all five groups: Sham (110±3 mmHg), ACF-Veh (103±2 mmHg), ACF-Levo (111±4 mmHg),

REV-Veh (120±2 mmHg), and REV-Levo (116±3 mmHg).

Levo improves myofilament Ca2+ sensitivity and maximal force generation without altering myocyte Ca2+ transient kinetics

Myofilament force generation at varying [Ca2+] was measured in skinned myocytes from frozen LV tissue. Compared with Sham, there was a rightward shift in the force-pCa relationship for ACF-Veh and REV-Veh (5.37±0.01 and 5.40±0.02 vs.

5.47±0.02, respectively) and 16% decrease in maximal force generation for ACF-Veh

(Fig 8a). Levo caused the leftward shift in the force-pCa relationship expected for a myofilament Ca2+ sensitizer, and in ACF-Levo, there was a ~50% increase in maximal force generation compared with ACF-Veh.

Intracellular Ca2+ transients in LV myocytes isolated from all groups were measured. Compared to Sham, the Ca2+ transient amplitude was significantly increased in ACF-Veh and REV-Veh but normalized in ACF-Levo and REV-Levo (Fig 8b). To evaluate the Ca2+ available for contraction and relaxation, we integrated the Ca2+ signals during contraction and relaxation, and normalized these values to the peak amplitude

Figure 8: Levo improves myofilament calcium sensitivity and force generation without increasing the peak intracellular calcium transient

2+ A. Force-pCa for LV myofilaments. B. Intracellular calcium amplitude (peak [Cai ]) in response to electrical stimulation. C. Representative Ca2+ transients. D. Ca2+ available for contraction normalized to peak (AC/Pk). Data are means±SEM from 3-8 rats/group. *p<0.05, **p<0.01 vs. Sham-Veh; ^^p<0.01, ^^^p<0.001, ^^^^p<0.0001 vs. ACF-Veh; $$ p<0.01, $$$$ p<0.0001 vs. REV-Veh.

(AC/PK and AR/PK, respectively). AC/PK was similar amongst Sham, ACF-Veh, and

REV-Veh, but decreased in ACF-Levo and REV-Levo (~23%; Fig 8c) indicating lower intracellular Ca2+ availability. Together, these data suggest that Levo increases

51 myofilament Ca2+ sensitivity and force generation without increasing intracellular Ca2+ availability.

Levo does not significantly alter LV myocyte sarcomere shortening

Auxotonic sarcomere peak shortening (%PS) in fresh LV myocytes was reduced in ACF compared to Sham (12.8±0.3%, p<0.05), but was unchanged in ACF-Veh vs.

ACF-Levo (11.8±0.3% vs. 11.3±0.3%, p>0.05). %PS was comparable in Sham, REV-

Veh (13.6±0.3%) and REV-Levo (13.0±0.3%). Sarcomere shortening and re-lengthening velocities were similar in ACF-Veh and ACF-Levo (-2.3±0.1 µm/s vs. -2.4±0.1 µm/s and

1.6±0.1 µm/s vs. 1.7±0.1 µm/s, respectively) and in REV-Veh and REV-Levo (-2.8±0.1

µm/s vs. -3.0±0.1 µm/s and 2.0±0.1 µm/s vs. 2.1±0.1 µm/s, respectively).

Levo’s positive lusitropy is associated with increased cMyBP-C phosphorylation, cTnI phosphorylation, and/or increased α-to-β-MHC

Diastolic relaxation is controlled by changes in Ca2+ cycling, myofilament composition, Ca2+ desensitization, and crossbridge cycling kinetics [3,24]. Coordinated phosphorylation of cMyBP-C and cTnI are central to these processes; therefore, we measured cMyBP-C and cTnI phosphorylation at key serine residues (cMyBP-C Ser-273,

Ser-282 and Ser-302 and cTnI Ser-23/24). Total protein levels, though variable, were not different amongst groups (n=5-6; Fig 9). In ACF-Levo cMyBP-C phosphorylation at Ser-

273 (~2-fold) and Ser-302 (~3-fold), but not Ser-282 (not shown), were significantly increased compared to ACF-Veh (Fig 9a). Additionally in ACF-Levo, cTnI Ser-23/24

Figure 9: Levo increases cMyBP-C phosphorylation, cTnI phosphorylation and/or alpha to beta MHC A. Total cMyBP-C and Ser-273 and Ser-302 phosphorylation. B. Total cTnI and Ser- 23/24 phosphorylation. Representative immunoblots and cumulative data. C. Myosin heavy chain (MHC) mRNA. Data are means±SEM from n=5-8/group. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 vs. Sham-Veh; ^p<0.05, ^^p<0.01 vs. ACF-Veh; $ p<0.05, $$$$ p<0.0001 vs. REV-Veh.

phosphorylation (~2 fold) was significantly increased (Fig 9b). By comparison in REV-

Levo, only cMyBP-C phosphorylation at Ser-273 was significantly increased (~2 fold).

Because of the role of myofilament composition in cross-bridge cycling kinetics, we measured α- and β-MHC mRNA in LV tissue. As previously demonstrated in ACF-

Veh and REV-Veh [76] there is a relative decrease and increase in α- and β-MHC mRNA, respectively (Fig 9c). While the α-to-β-MHC mRNA ratio remains depressed in

ACF-Levo compared with Sham (4.9 vs. 11.2, respectively), the α-to-β-MHC mRNA ratio is nearly normalized in REV-Levo (10.2) and significantly improved relative to

REV-Veh (5.6, p<0.0001). This increase in ratio is driven primarily by decreased β-

MHC mRNA (Fig 9c).

Comparison of acute effects of Levo, milrinone and OM on LV function in rats with established VO-HF

8-week ACF rats were infused with positive inotropes with and without PDE3a inhibitory activity (milrinone (9.9 μg/kg/min) and OM (11.7 μg/kg/min), respectively) for

30 minutes. Control groups included ACF infused with vehicle (1% DMSO) or Levo

(2.8 μg/kg/min) and Sham infused with vehicle. Before and after infusion, %FS and

MAP; tau and dP/dtmin were measured post-infusion.

Following infusion, %FS in ACF (32%) was increased by Levo (40%), milrinone

(38%) and OM (37%); MAP remained unchanged in all groups (Fig 10a). Diastolic relaxation differed amongst groups (Fig 10b). Diastolic relaxation was normal in Sham and impaired in ACF-Veh. Diastolic relaxation was improved in ACF-Levo and ACF-

Milrinone, but impaired in ACF-OM. 54

Figure 10: Short-term treatment with PDE3a inhibitor improves lusitropy in 8-week ACF rats A. Levo (ACF-Levo), milrinone (ACF-Mil) and omecamtiv mecarbil (ACF-OM) improved %FS without altering MAP. B. Levo and milrinone improved, while OM worsened Tau and dP/dtmin. Data are means±SEM. N=5/group. *p<0.05, **p<0.01, ****p<0.0001 vs. Sham-Veh; ^p<0.05 vs. ACF-Veh; $$ p<0.01, $$$ p<0.001, $$$$ p<0.0001 vs. ACF-OM.

Discussion

We treated ACF rats in pre-HF and reversed pre-HF with Levo to determine if treatment with a myofilament Ca2+sensitizer could improve LV function with and without

55 surgical repair of early-stage VO. We show for the first time that despite persistent VO,

Levo improves systolic (%FS, Ees and PRSW; Fig 5-6) and diastolic (tau and dP/dtmin;

Fig 6) function without significantly improving LVEDD (Fig 5). These functional changes are associated with improved myofilament Ca2+ sensitivity with decreased

2+ 2+ intracellular Ca transient amplitude and systolic intracellular Ca (Fig 8); preserved in vivo β-adrenergic responsiveness (Fig 7); and increased cMyBP-C and cTnI phosphorylation in ACF-Levo and increased cMyBP-C phosphorylation and MHC isoform switch in REV-Levo (Fig 9).

ACF is widely used to induce VO in rodents [57,74,167]. This model replicates the hemodynamics and pathophysiology of VO, irrespective of the etiology

[66,74,76,160]. Although Levo is untested in VO-HF models, our findings mirror improved systolic function in Goto-Kakizaki rats post-infarct [107], in dogs with pacing- induced HF [113], and in patients with acute HF [136]. Levo improves systolic function by stabilizing the Ca2+-saturated cTnC, prolonging its interaction with cTnI [136], accelerating actin-myosin crossbridge cycling [3], and increasing contractility without

2+ altering intracellular Ca transient or myocardial O2 demand [136]. Our data confirm increased myofilament Ca2+ sensitivity with decreased intracellular Ca2+ availability, although more Ca2+ could be bound to the sensitized myofilaments leaving less in the cytosol.

Additionally, β-adrenergic responsiveness was preserved in all groups (Fig 7), which differs from the blunted response in ACF at end-stage HF [66], suggesting that β- receptor desensitization is not fully manifest at our timepoint. Interestingly, treating ACF rats with metoprolol 4-15 weeks post-ACF increased mortality and decreased systolic function (unpublished data; Wilson and Lucchesi). This further suggests preserved β- adrenergic responsiveness. Some suggest that phosphodiesterase (PDE) 3 inhibition contributes to Levo’s positive inotropy [113], while others suggest that PDE3 inhibition occurs at supra-therapeutic concentrations and does not contribute to positive inotropy

[140]. With the comparable β-adrenergic responsiveness in Levo and Veh, we suggest that the positive inotropy is independent of PDE3 inhibition.

We also demonstrate positive lusitropy with Levo (Fig 6c), which has also been shown in Dahl/Rapp rats [19], in dogs with pacing-induced HF [113], and in human clinical trials [136]. The mechanism of positive lusitropy is poorly understood; proposed mechanisms include relatively mild Ca2+ sensitization and weaker cTnC-Levo interactions at diastolic Ca2+ levels [136,140]. Our results suggest that cMyBP-C phosphorylation, cTnI phosphorylation and/or MHC isoform contribute to improved relaxation. In ACF-Levo LV tissue collected during Week 8, we show increased phosphorylation of cMyBP-C Ser-273 and Ser-302 and cTnI Ser-23/24 with no alteration in MHC, while in REV-Levo LV tissue, we show increased phosphorylation of cMyBP-C

Ser-273 and a relative increase in α-to-β-MHC. cMyBP-C regulates thick-filament function and phosphorylation at Ser-273, Ser-282 and Ser-302 appears to accelerate

57 cross-bridge cycling, increasing force development as well as relaxation [47,151]. In vitro studies have shown that these cMyBP-C serine residues are differentially phosphorylated by PKA, PKC, PKD, CamKII, CK2 and RSK [67]. Ser-273, Ser-282 and

Ser-302 are phosphorylated by PKA following β-adrenergic stimulation, while PKC phosphorylates Ser-273 and Ser-302 and PKD phosphorylates Ser-302 [61,67,122].

CaMKII phosphorylates Ser-273, Ser-282 and Ser-302 in a [Ca2+]-dependent manner

[67,92,164]. With the overlapping in vitro kinase phosphorylation patterns and the difference between ACF-Levo and REV-Levo, it is difficult to speculate which kinase phosphorylates cMyBP-C. Additionally, hemodynamic load likely alters phosphorylation patterns. cTnI regulates thin-filament function, and Ser-23/24 phosphorylation by PKA decreases Ca2+ sensitivity and myofilament cross-bridge cycling kinetics [142,178] and increases the rate of cTnC-Ca2+ dissociation, which is thought to be rate limiting for diastolic relaxation [103]. Chemically-skinned myocardium containing primarily α-MHC has a higher ATPase rate, faster sarcomeric shortening velocity, higher power production and greater rate of force development compared with β-MHC [196]. In summary our data suggest that cMyBP-C and/or cTnI phosphorylation improves myofilament cross- bridge kinetics in ACF-Levo; whereas cMyBP-C phosphorylation and/or MHC isoform improve cross-bridge cycling kinetics in REV-Levo. The differing roles of cTnI phosphorylation and MHC may be from differences in hemodynamic load.

The mechanisms underlying Levo’s positive lusitropy may involve PDE3a inhibition. To further characterize the role of PDE3a inhibition in lusitropy, we evaluated

58 diastolic relaxation in 8-week ACF rats that were infused with drugs with (milrinone,

Levo) and without (OM, vehicle) known PDE3a inhibitory activity. Lusitropy was improved with Levo and milrinone but not with OM (Fig 10b). This seemingly suggests that Levo’s positive inotropy and lusitropy are mediated by improved myofilament Ca2+ sensitivity and PDE3a inhibition, respectively; while milrinone’s positive inotropy and lusitropy are mediated by PDE3a inhibition. One explanation is that Levo and milrinone differentially target PDE3a. Our data suggest that Levo’s lusitropic actions are associated with increased cTnI and cMyBP-C phosphorylation. By contrast, milrinone’s lusitropic actions appear to be associated with altered Ca2+ transient amplitudes and relaxation times, and increased SR Ca2+ following phospholamban and ryanodine receptor phosphorylation [12]. This may occur through subcellular compartmentalization of PDE and kinase activity [144,183]. To further investigate this, we performed co- immunoprecipitation assays in normal rat LV lysates where we immunoprecipitated with an anti-PDE3a antibody (ab169534 - 1:50 dilution, Abcam, Cambridge, MA) followed by detection with an anti-cTnT antibody (ab8295 – 1:2000; Abcam) and verified our findings by performing the reverse immunoprecipitation. In both cases, cTnT and PDE3a were associated with each other (not shown). Together, these results suggest a role for

PDE3a in lusitropy and that subcellular compartmentalization may explain the disparate findings between PDE3a’s role in Levo’s positive inotropy and lusitropy. Further work to delineate the signaling pathway downstream of Levo leading to cMyBP-C and cTnI phosphorylation is required.

Clinically, arrhythmias and hypotension are associated with high Levo doses

[136]. We noted no significant increase in arrhythmias in Levo-treated rats (not shown); however, heart rhythm was not chronically measured and arrhythmias are difficult to characterize in rodents because of their high heart rates. At our Levo dose, MAP was slightly elevated compared with ACF-Veh and comparable to Sham, suggesting that

Levo’s inotropic and lusitropic effects are not secondary to changes in afterload.

In summary, we demonstrate that 1) Levo improves systolic and diastolic function in rats with pre-HF and reversed pre-HF; 2) targeting the myofilament improved function during early HF even with continued VO; and 3) improved diastolic function correlates with cMyBP-C phosphorylation, cTnI phosphorylation, and/or MHC isoform.

Myofilament-targeting compounds offer alternate pharmacologic treatment options for patients with VO HF.

Table 1: LV morphologic, echocardiographic and hemodynamic parameters in rats at Week 8 Sham-Veh ACF-Veh ACF-Levo REV-Veh REV-Levo N 10-24 7-17 10-19 10-20 10-19 Body weight 392.2 ± 7.7 415.0 ± 7.0 412.4 ± 6.7 352.0 ± 6.6 383.8 ± 5.3 (g) Heart (g) 1.00 ± 0.02 1.91 ± 1.85 ± 1.13 ± 1.18 ± 0.06**** 0.07**** 0.02^^^^ 0.03^^^^ Lung (g) 1.39 ± 0.03 2.24 ± 2.12 ± 1.46 ± 1.59 ± 0.14**** 0.12**** 0.04^^^^ 0.05^^^ LVEDD 7.8 ± 0.06 10.9 ± 10.3 ± 7.9 ± 8.0 ± (mm) 0.22**** 0.11****,^ 0.08^^^^ 0.05^^^^ LVESD 4.3 ± 0.08 7.4 ± 6.1 ± 5.2 ± 4.7 ± (mm) 0.22**** 0.14****,^^^^ 0.25**,^^^^ 0.05^^^^ PWTd (mm) 1.7 ± 0.02 1.8 ± 0.04** 1.9 ± 1.8 ± 0.03 1.8 ± 0.05**** 0.04*** PWTs (mm) 2.8 ± 0.05 2.6 ± 0.07 3.1 ± 2.5 ± 2.8 ± 0.05**,^^^^ 0.04*** 0.05$$$ FS (%) 44.3 ± 0.84 32.2 ± 41.3 ± 32.1 ± 41.7 ± 0.80**** 0.77^^^^ 0.54**** 0.67^^^^ Dilation 0.44 ± 0.01 0.34 ± 0.38 ± 0.45 ± 0.45 ± index 0.01**** 0.01**** 0.01^^^^ 0.01^^^^ HR (bpm) 358 ± 12 315 ± 16 331 ± 7 345 ± 6 362 ± 21 SV (µL) 152 ± 7 220 ± 9**** 209 ± 11*** 140 ± 9^^^^ 146 ± 9^^^^ EDV (µL) 312 ± 15 396 ± 13*** 381 ± 5** 341 ± 15 323 ± 15^^

dP/dtmax 7522 ± 296 7456 ± 414 9788 ± 6907 ± 230 8327 ± 645 (mmHg/sec) 459**,^^ Ees, adj 0.902 ± 0.392 ± 0.654 ± 0.520 ± 0.859 ± (mmHg/µL) 0.049 0.064*** 0.053*,^ 0.041**** 0.039$$$$ PRSW, adj 107.3 ± 9.2 60.2 ± 12.1* 113.6 ± 9.9^ 59.0 ± 7.9* 101.4 ± (mmHg) 7.5$$

dP/dtmin -7887 ± 324 -5739 ± -7626 ± -7693 ± -8327 ± (mmHg/sec) 303**** 197^^^ 469*** 645$ Tau (Weiss) 9.9 ± 0.3 10.8 ± 0.8 8.9 ± 0.3^ 10.6 ± 0.6 9.2 ± 0.4$ LV diameter in end-diastole and end-systole (LVEDD and LVESD, respectively); posterior wall thickness in diastole and systole (PWTd and PWTs, respectively); percent fractional shortening (%FS), Dilation index (2xPWTd/LVEDD). Data are expressed as mean ± SEM. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001 vs Sham-Veh; ^ p<0.05, ^^ p<0.01, ^^^ p<0.001, ^^^^ p<0.0001 vs ACF-Veh; $ p<0.05, $$ p<0.01, $$$ p<0.001, $$$$ p<0.0001 vs REV-Veh.

Chapter 3: Impaired function in rats following delayed reversal of volume overload

heart failure can be rescued by levosimendan

Abstract

Early surgical intervention in aortocaval fistula (ACF)-induced volume overload

(VO) heart failure (HF) results in rapid structural remodeling but delayed functional recovery. This study investigated if delayed surgical intervention in rats with ACF- induced VO HF adversely impacts LV functional recovery and if pre- or post-surgical treatment with levosimendan (Levo) could alter LV functional recovery. ACF or Sham surgery was performed in male Sprague-Dawley rats (200-240 g) for 4 or 8 weeks prior to reversing (REV, closing) the shunt in a subset of animals. REV at 8 weeks resulted in persistent impaired LV systolic and diastolic function, while REV at 4 weeks resulted in delayed, but improved LV systolic and diastolic function. Separate subsets of rats reversed at 8 weeks received Levo (1 mg/kg in drinking water) from weeks 4-8 prior to

REV or from weeks 8-19 following REV. Levo given pre-REV transiently improved LV function, while Levo given post-REV resulted in a sustained increase in LV function to the end of the study. Improved LV function in REV rats given Levo 8-19 wks post-REV correlated with normalized α-to-β-myosin heavy chain (MHC) and increased cTnI Ser-

23/24 phosphorylation. These results demonstrate that delayed REV results in persistent

LV dysfunction that can be rescued with Levo given continuously post-REV. 62

Introduction

Mitral regurgitation (MR) is the most common valvular lesion in the United States affecting >2 million Americans [55,65], and is the second most common valvular disease requiring surgery in Europeans [55]. MR’s pathophysiological consequences include chronic left ventricular (LV) hemodynamic volume overload (VO) followed by LV chamber dilation, eccentric myocardial hypertrophy, progressive LV contractile dysfunction and heart failure (HF). According to American College of Cardiology

(ACC)/American Heart Association (AHA) guidelines, valve surgery is indicated for symptomatic patients and asymptomatic patients with LV systolic dysfunction, and surgery is considered reasonable for asymptomatic patients with pulmonary hypertension at rest and/or atrial fibrillation [23,49,78]. Even with these guidelines, 17% of patients will develop LV dysfunction in the early post-operative period [112,194].

Despite the clinical importance of VO, there are few models described that mimic the pathophysiological progression of chronic VO, and these models are rarely used to assess LV function following hemodynamic load reduction. In the rodent MR model, a needle is inserted into the LV under trans-esophageal echocardiographic guidance to puncture a mitral valve leaflet [87,88,152,153]. While this technique faithfully models

MR in the rat by producing univentricular VO, this technique is not currently reversible and LV remodeling following MR repair cannot be studied. In contrast, the aortocaval fistula (ACF) model of VO HF in the rodent produces biventricular VO but has the advantage that the ACF can be reversed (closed), and LV remodeling can be monitored

63 long term. In ACF rats, the chronic increase in LV preload leads to progressive LV pump failure and decreased myocyte contractility, which can be classified into three clinically relevant stages: pre-HF (4 wks ACF; marked LV dilation, increased LV wall stress, mild

LV dysfunction), established HF (8 wks ACF; severe LV dilation, significant LV systolic and diastolic dysfunction) and end-stage HF (15-21 wks ACF; pump failure with pulmonary congestion/edema) [40,66,76,160]. We have previously shown that reversing

(REV, closing) the ACF during pre-HF (i.e. 4 weeks post-ACF) results in rapid structural remodeling but delayed functional recovery [76]. Reversal during established or end- stage HF has not previously been performed, but could potentially model the stage of HF when many patients would present for treatment.

Although surgery is the only therapeutic option to reduce hemodynamic VO, preoperative pharmacologic therapy may prolong the time to surgical intervention, improve function in poor surgical candidates or improve function in post-operative patients that continue to have depressed function. Therapeutics targeting myofilament activation, including myofilament Ca2+ sensitizers (e.g. levosimendan, Levo) and myosin activators

(e.g. omecamtiv mecarbil). These drugs show promise in treating acute decompensated

HF [14,45], improving pump function pre-operatively [180], and potentially improving outcomes in chronic HF [82]. Levo interacts with and stabilizes Ca2+-saturated troponin

C (cTnC) prolonging its interaction with cardiac troponin I (cTnI) [136]. This interaction promotes contractile force without increasing the amplitude of the intracellular Ca2+ transient or increasing myocardial oxygen consumption [126,128,136]. We have

64 previously treated ACF rats with pre-HF and reversed pre-HF with Levo and have shown improved systolic and diastolic function (Chapter 2).

In the present study, we determined 1) if delayed reversal (8 weeks post-ACF) altered structural and functional recovery compared with early reversal (4 weeks post-

ACF), and 2) if Levo could counteract the detrimental effects of delayed reversal.

Animals reversed during compensated HF (8 weeks post-ACF) were followed for an additional 11 weeks, which is when we observed functional recovery in rats reversed during pre-HF (4 weeks post-ACF). Our results provide the first evidence that delayed surgical intervention results in early structural remodeling but no functional recovery. We also show that although pre-operative treatment with Levo does not preserve function post-REV, continued post-operative Levo improves LV function.

Materials and Methods

ACF Surgical Model in Rats

Male Sprague-Dawley rats (200-240g; Harlan) were housed in a temperature and humidity controlled room using a 12h light/dark cycle with access to standard rat chow and water ad libitum. Studies conformed to the principles of the National Institutes of

Health ―Guide for the Care and Use of Laboratory Animals,‖ (NIH publication No. 85-

12, revised 1996). The protocol was approved by the Institutional Animal Care and Use

Committee of The Research Institute at Nationwide Children's Hospital.

Cardiac VO was induced in rats at Week 0 (Fig 11) under isoflurane anesthesia

(2-2.5%) as previously described [66,76]. Following ventral abdominal midline incision, the abdominal aorta and caudal vena cava were exposed. Cranial and caudal to the fistula site, the adventitia was bluntly separated and 5-0 Ethilon® suture (Ethicon, Cincinnati,

OH) was pre-placed for reversal. An 18g short-bevel needle (Becton Dickinson, Franklin

(Ethicon, Cincinnati, OH) and 4-0 silk (Ethicon, Cincinnati, OH), respectively. Sham animals underwent a similar procedure except for suture pre-placement and aortic puncture. For reversal, a ventral midline incision was made, the fistula site was exposed, fibrovascular tissue was bluntly dissected and the previously placed 5-0 Ethilon® suture was ligated. The abdominal wall and skin were closed as above. Shunt closure was defined as the lack of arterial blood in the vena cava and/or a >1 mm decrease in LV end- diastolic diameter (LVEDD) as measured by echocardiography. The absence of pedal withdrawal reflex and respiratory rate were monitored during both procedures to ensure adequacy of anesthesia. Buprenorphine (0.03 mg/kg SC) was given immediately post- operatively and every 12 hours for 72 hours post-operatively as needed for pain.

Figure 11: Experimental time course ACF was induced at Week 0. In subsets of rats, the ACF was reversed (REV) at Weeks 4 and 8. A subset of rats reversed at Week 8 were given Levo (1 mg/kg/day in drinking water) from Weeks 4-8, while a different subset reversed at Week 8 received Levo from Weeks 8-19. Echo was performed biweekly and hemodynamics and tissue collection were performed at Week 19.

Levosimendan treatment

Rats were given vehicle (Veh; water) or Levo (L-enantiomer of ([4-(1,4,5,6- tetrahydro-4-methyl-6-oxo-3-pyridazinyl)phenyl]-hydrazono)-propanedinitrile; LKT

Laboratories, MN) for the time periods shown in Fig 11. Levo (0.0133 mg/mL) was prepared twice weekly in drinking water to deliver a dose of ~1 mg/kg/day. This dose 67 has improved systolic and diastolic function in previous rodent studies and did not adversely impact water consumption [19,107,108].

Echocardiography

Transthoracic echocardiograms (n=8-10/group) were performed biweekly with an

7.2-8.5-MHz transducer (Xario or Artida, Toshiba Medical Systems, Tustin, CA) under isoflurane anesthesia (1.5-2%) [66,76]. M-mode images were obtained in parasternal short axis at the level of the papillary muscles to assess chamber diameters in end-systole

(LVESD) and end-diastole (LVEDD) and LV posterior wall thickness in systole (PWTs) and diastole (PWTd). These indices were calculated: % Fractional shortening (%FS) =

(LVEDD-LVESD)/LVEDD x 100; Dilation index = (2*PWTd)/LVEDD.

Measurement of LV hemodynamics

LV hemodynamics were assessed invasively using pressure–volume (PV) analysis during Week 19 [66,76]. Rats were anesthetized with 3% isoflurane, intubated by tracheostomy, ventilated with a pressure-controlled rodent ventilator (SAR-830, CWE

Inc, Ardmore, PA), and maintained under 1.75% isoflurane anesthesia. The PV catheter

68 slope of the end-systolic pressure volume relationship (ESPVR). Data was acquired and analyzed using Labscribe 2 software (iWORX, Dover, NH).

Immunoblot analysis

LV tissue lysates were prepared from tissues collected at Week 19 [66,76].

Proteins (16 μg) were separated by SDS-PAGE and transferred to PVDF membranes.

Immunoblotting was performed with antibodies against cTnI (1:2000, Cell Signaling,

Beverly, MA), phospho-TnI Ser23/24 (1:1000, Cell Signaling), serca2a (1:2000, Thermo

Scientific, Waltham, MA), and connexin 43 (1:5000, Santa Cruz Biotechnology, Santa

Cruz, CA). Relative band densities were analyzed using BioRad Image Lab Software

(BioRad, Hercules, CA) and normalized to total ERK1/2 as loading controls [66,76].

Quantitative real-time PCR analysis.

RNA was isolated from LV tissue using commercially available kits and protocols

(Qiagen RNeasy microarray tissue mini kit; Qiagen Inc, Valencia, CA). RNA was reverse

69 transcribed using the RevertAid first strand cDNA synthesis kit and protocol (Fermentas;

Thermo Scientific, Hanover, MD). First strand equivalent to 100 ng input RNA was amplified in duplicate for each animal using appropriate Roche universal probe/primer pairs for each target gene: α-myosin heavy chain (MHC), β-MHC, atrial natriuretic peptide (ANP), and B-type natriuretic peptide (BNP; see [95] for primer sequences) and

Maxima Probe quantitative real-time PCR master mix (Fermentas). Amplifications were carried out for 40 cycles using an Eppendorf MasterCycler-ep Realplex thermocycler.

Parallel amplifications using nonreverse transcribed samples were performed to rule out genomic DNA contamination. Data were analyzed for relative expression using the

2−ΔΔCt method, with ribosomal protein Rpl13a serving as the internal control and the average sham value at each time point serving as a second normalizer.

Statistical analysis

Data are expressed as mean±SEM. Statistical analyses were performed using

GraphPad Prism V6.0 or SPSS V19 (ANCOVA only). One-way or two-way ANOVA or

ANCOVA, followed by Bonferroni's post-hoc test, was used to measure differences between groups. P < 0.05 was considered significant.

Results

LV chamber morphology rapidly returns to Sham dimensions even after delayed reversal

Compared with Sham, LVEDD and LVESD were increased in all experimental groups; these parameters remained elevated in the ACF group at all timepoints compared

Figure 12: Effect of delayed reversal on LVEDD and %FS A. Representative M-mode images at Week 19. B. Echocardiographic measurements: LVEDD and LVESD are LV diameter at end-diastole and end-systole, respectively. Data are means ± SEM (n=8-10/group). ** p<0.01, *** p<0.001, **** p<0.0001 vs. Sham- Veh; ^ p<0.05, ^^ ^^p<0.0001 vs. ACF.

to Sham, while LVEDD and LVESD decreased to Sham values upon reversal at Weeks 4 and 8 for the remainder of the study (Fig 12; Table 2). Compared to Sham at Week 19,

PWTd was increased in ACF but similar in rats reversed at 4 and 8 weeks. As such, the dilation index was significantly lower in ACF indicating eccentric hypertrophy in non- reversed rats. In contrast, the dilation index was normalized in REV rats indicating

71 remodeling to Sham dimensions (Table 2). EDV and SV were measured with a pressure- volume catheter during Week 19. As expected, EDV and SV were significantly increased in ACF, but reduced in REV (Table 2). The increases in EDV and SV were accompanied by increased heart and lung weights in ACF (Table 2), consistent with pathologic LV hypertrophy and pulmonary congestion and edema, which was confirmed by microscopic exam (data not shown). By contrast, heart and lung weights were normalized in the reversal groups.

LV function remains impaired after delayed reversal

%FS, a hemodynamic load-dependent indicator of systolic function, was measured by serial echocardiography. %FS decreased in ACF and REV rats compared to

Sham (Fig 12). Beginning at Week 14, %FS increased in REV@4, indicating recovery of systolic function. By contrast, %FS in ACF and REV@8 remained low. During Week

19, invasive hemodynamic measurements were performed to obtain load-independent measures of systolic function (Fig 13). Similar to previous reports [66,76], Ees and

PRSW were decreased in ACF but increased in REV@4 compared with Sham (Fig 13b,

Table 2). In contrast, Ees and PRSW remained decreased in REV@8. While evaluating systolic function, we examined the effects of delayed reversal on LV diastolic relaxation.

Compared to REV@4, we observed significantly impaired diastolic relaxation, (increased tau, a less negative dP/dtmin; Fig 13c, Table 2), in ACF and REV@8.

Figure 13: LV systolic and diastolic function are impaired following delayed reversal A. Representative pressure-volume loops. B. LV systolic functional measurements: Ees, end systolic elastance; PRSW, preload recruitable stroke work. C. LV diastolic functional measurements: Relaxation constant, tau-Weiss; dP/dtmin, first derivative of LV pressure decline. Data are means ± SEM (n=8-10/group). * p<0.05, ** p<0.01, **** p<0.0001 vs. Sham; ^^^^ p<0.0001 vs. ACF 73

Levo does not significantly alter LV chamber morphology

To determine if Levo could counteract the detrimental effects of delayed reversal, subsets of ACF rats reversed at Week 8 were first treated with Levo for 4 weeks prior to reversal (REV@8-L4-8) or for 11 weeks following reversal (REV@8-L8-19; Fig 11).

Compared with Sham, LVEDD and LVESD were increased in REV@8-Veh, REV@8-

L4-8, and REV@8-L8-19 indicating an increase in chamber dimension until reversal at

Week 8 and comparable to Sham from Weeks 10-19 indicating normalization of chamber dimensions (Fig 14; Table 3). Since PWTd remained largely unchanged and LVEDD normalized, the dilation index was similar to Sham, indicating reverse remodeling (Table

3). These findings indicate that Levo treatment before or after reversal did not significantly alter LV chamber morphology as measured by echocardiography.

In addition to the echocardiographic monitoring of LV size throughout the study,

EDV and SV were measured with a pressure-volume catheter during Week 19. As expected for reversed-VO at Week 8, EDV and SV were comparable to Sham (Table 3).

In these reversed rats, heart and lung weights were comparable to Sham, suggesting resolution of LV hypertrophy and pulmonary congestion/edema (Table 3). These values were comparable between vehicle and Levo-treated groups, suggesting again that Levo treatment does not significantly alter LV morphology.

Figure 14: Levo improved %FS without altering chamber morphology A. Representative M-mode images at Week 19. B. Echocardiographic measurements: LVEDD and LVESD are LV diameter at end-diastole and end-systole, respectively. Data are means ± SEM (n=8-10/group). ** p<0.01, *** p<0.001, **** p<0.0001 vs. Sham- Veh; ^ p<0.05, ^^ ^^p<0.0001 vs. ACF-Veh.

Continued Levo improves LV systolic and diastolic function

While Levo-treatment did not significantly alter LV morphology, previous studies have shown that it can improve LV function during short-term studies (Chapter 2). To this end, we tested LV function in Levo and vehicle-treated rats through 19 weeks post-

ACF. In REV@8 treated with Levo from Weeks 4-8, %FS was significantly improved

Figure 15: Continued Levo improves LV systolic and diastolic function following delayed reversal A. Representative pressure-volume loops. B. LV systolic functional measurements: Ees, end systolic elastance; PRSW, preload recruitable stroke work. C. LV diastolic functional measurements: Relaxation constant, tau-Weiss; dP/dtmin, first derivative of LV pressure decline. Data are means ± SEM (n=7-10/group). ** p<0.01, **** p<0.0001 vs. Sham-Veh; ^^^^ p<0.0001 vs. REV@8wks-Veh 76 during treatment (6% increase compared with Week 4, p<0.05), but %FS dropped to

REV@8-Veh levels by Week 10 and remained low through Week 19 (Fig 14). This decreased systolic function was confirmed by PV analysis (Fig 15), which showed that

Ees and PRSW were decreased compared to Sham, but comparable to REV@8-Veh (Fig

15b, Table 3). In contrast in REV@8 treated from Weeks 8-19, %FS improved to Sham levels by Week 10 and remained elevated through Week 19 (Fig 14). This improved systolic function was confirmed by PV analysis, which showed that Ees and PRSW were comparable to Sham and increased compared to REV@8-Veh (Fig 15b, Table 3). LV diastolic relaxation was evaluated at Week 19. In REV@8-L4-8, there was significantly impaired diastolic relaxation, (increased tau, a less negative dP/dtmin; Fig 15c, Table 3).

In contrast, diastolic relaxation improved to Sham levels in REV@8-L8-19. Together these results demonstrate that continued Levo improves systolic and diastolic function.

Delayed reversal normalizes VO-induced expression of the hypertrophic markers ANP and BNP

To investigate the molecular response to VO and reversed-VO, we measured relative mRNA expression of the hypertrophic markers atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP) in LV tissue at Week 19. ANP and BNP expression was increased in ACF (24-fold and 1.8-fold, respectively; p<0.01 vs. Sham), but normalized in REV@4 (5-fold and 1.2 fold, respectively; p>0.05 vs. Sham) and REV@8

(4-fold and 1.4-fold, respectively; p>0.05 vs. Sham). ANP expression was normalized in

REV@8-L4-8 and REV@8-L8-19 (8-fold and 2-fold, respectively; p>0.05 vs. Sham).

BNP expression was increased in REV@8-L4-8 (1.8-fold, respectively; p<0.01 vs.

Sham), but normalized in REV@8-L8-19 (0.9 fold, respectively; p>0.05 vs. Sham). The increase in BNP mRNA in REV@8-L4-8 was unexpected, though BNP mRNA can increase during pressure overload, β-adrenergic stimulation and hypoxia [18,36,130,203].

Improved LV function with sustained Levo treatment is associated with normalized α-to-

β-MHC

Because of the role of myofilament composition in cross-bridge cycling kinetics, we measured relative α- and β-MHC mRNA expression in LV tissue lysates (Fig 16). In

REV@8-Veh and REV@8-L4-8, α-to-β-MHC ratio is decreased compared with Sham

(2.6 and 2.1 vs. 4.4, respectively; Fig 16), while α-to-β-MHC in REV@8-L8-19 is normalized compared with Sham (5.0 vs. 4.4, respectively). The decreased and increased ratios are driven primarily by increased and decreased β-MHC mRNA, respectively.

Figure 16: Continued Levo normalizes alpha-to-beta MHC Changes in α-to-total, β-to-total and α-to-β-myosin heavy chain (MHC) mRNA ratio. Cumulative data are means ± SEM from n=7-9/group. ** p<0.01, *** p<0.001, **** p<0.0001 vs. Sham-Veh; ^^^^ p<0.0001 vs. REV@8-Veh.

Levo’s positive lusitropy is also associated with increased cTnI phosphorylation

cTnI is a thin filament regulatory protein linking Ca2+ binding to cTnC with actin- myosin interactions and cross-bridge formation. cTnI phosphorylation at Ser-23/24 by

PKA decreases Ca2+ sensitivity and myofilament cross-bridge cycling kinetics[178,142] and increases the rate of Ca2+ dissociation from cTnC, which is thought to be rate limiting for diastolic relaxation[118,103]. Amongst the REV@8 groups, there was no significant

79 difference in total cTnI (n=4/group). In REV@8-Veh and REV@8-L4-8, cTnI Ser 23/24 phosphorylation was comparable to Sham. Compared to Sham and REV@8-Veh, cTnI

Ser-23/24 phosphorylation was increased 2.7-fold (p<0.01) and 1.8-fold (p<0.05), respectively.

Discussion

In this study, we reversed the ACF during Week 8 (i.e. established HF) to characterize the effect of delayed reversal on LV functional recovery. We show for the first time that despite rapid structural remodeling, LV systolic and diastolic function

(%FS, Ees, PRSW, tau and dP/dtmin; Figs 12-13) remains impaired following delayed reversal. We then treated subsets of REV@8 rats with Levo 1) before reversal to determine if Levo pre-treatment could delay the need for surgical reversal and 2) following reversal to determine if Levo could rescue impaired function. Although the improved LV function did not persist when pre-surgical Levo was stopped, post-surgical

Levo improved LV function (Figs 14-15). The improved LV function in REV@8-L8-19 and continued dysfunction in REV@8-Veh and REV@8-L4-8 correlates with normalized and decreased α-to-β-MHC (Fig 16).

Although the ACF model produces biventricular VO rather than the univentricular

VO produced in MR, the ACF model remains attractive because of the relative ease with which the ACF can be reversed. The ability to consistently reverse the ACF enables us to model the hemodynamics and pathophysiology of reversed-VO, irrespective of the

80 etiology [76]. We previously showed that reversal during pre-HF (i.e. 4 weeks post-

ACF) results in rapid structural remodeling but with delayed functional recovery [76]. In that report, a stent-graft was inserted into the aorta to cover the ACF, and that technique was associated with significant morbidity and mortality. As in the studies described in

Chapter 2, we now pre-place suture around the fistula, which is ligated during reversal.

LV structural and functional remodeling follows the same time-course as with the stent graft, but the suture technique is superior because there is decreased morbidity and mortality (0% for both in this study) and is technically easier. In our previous report [76],

LV function was normalized at 15 weeks post-ACF (i.e. 11 weeks post-REV) and no interim data was presented to characterize the time course of recovery. In this study, we show that %FS begins increasing at Week 14 (i.e. 10 weeks post-REV) and the increase continues until Week 19 (Fig 12), demonstrating that the functional recovery we previously described is persistent.

While our early surgical reversal model allows us to study early intervention, typically surgical intervention is delayed until there are clinically apparent symptoms associated with cardiac dysfunction (dyspnea, decreased exercise tolerance). To better model this clinical situation, we reversed the ACF in rats during the established HF phase at 8 weeks and monitored structural and functional recovery for an additional 11 weeks, which is equivalent to the time to functional recovery for REV@4 [76]. Similar to

REV@4, REV@8 results in rapid structural remodeling (Fig 12), which correlates with decreased LV and lung weights suggesting resolution of LV hypertrophy and pulmonary

81 congestion/edema (Table 2). In contrast to REV@4, there is continued dysfunction in

REV@8 (Fig 12-13) which is similar to post-surgical dysfunction, which occurs in 17% of MR patients, even if their surgery is performed within the established ACC/AHA guidelines [112]. Correlation studies in these MR patients suggest that post-surgical dysfunction is predicted by decreased pre-surgical LV ejection fraction, increased pre-

Figure 17: Pre-reversal LVESD and %FS negatively correlate with post-reversal %FS at Week 19 Correlation analysis (Pearson’s correlation analysis) of REV@4 and REV@8-Veh pre- reversal LVEDD, LVESD and %FS with post-reversal %FS at Week 19. Cumulative data are means ± SEM from n=8/group. 82 surgical LVEDD, and increased LVESD [21,171,174]. While pre-reversal LVEDD did not predict %FS in our rats at Week 19 (R2=0.00), increased pre-reversal LVESD and decreased %FS did (R2=0.69 and 0.63, respectively; p<0.001; Fig 17). Together, these results suggest that our model of delayed VO reversal is a relevant model of post-surgical dysfunction in MR patients and that impaired LV function (i.e. %FS and LVESD) prior to reversal is predictive of post-surgical functional recovery in our ACF model.

Levo has previously been shown to improve LV function in Goto-Kakizaki rats post-infarct [107], in Dahl/Rapp rats [19], in dogs with pacing-induced HF [113], and in patients with acute HF [136]. We have also previously shown that Levo improves LV function in ACF rats with pre-HF with and without hemodynamic load reduction

(Chapter 2). Similar to these findings, %FS improved during treatment in REV@8-L4-8 and in REV@8-L8-19, which was confirmed by invasive hemodynamics (Fig 14-15). It is important to note that continued Levo treatment was required to see this effect in our model, as the improved %FS in REV@8-L-4-8 pre-reversal was lost by Week 10 when

%FS became indistinguishable from REV@8Veh. The %FS in REV@8-L4-8 remained low until Week 19 and continued LV dysfunction was confirmed with invasive hemodynamics. These findings suggest that Levo only transiently improved LV function prior to REV. Levo interacts with and stabilizes Ca2+-saturated troponin C (cTnC) prolonging its interaction with cardiac troponin I (cTnI) [136]. The elimination half-lives of Levo and the active metabolite OR-1896 are 0.7 h and 6.5 h, respectively in the rat

[106]. Since Levo has a short elimination half-life and needs to be physically present to

83 be effective, the rapid decrease in function is not unexpected and suggests that the underlying cause of dysfunction persists.

The continued LV dysfunction in REV@8-Veh and REV@8-L4-8 may be due to defective electromechanical coupling, decreased force generation, or alterations in gene and/or protein expression. Because of the lack of differences in Serca2a and Cx43 protein levels amongst the REV@8-Veh, REV@8-L4-8 and REV@8-L8-19 (data not shown), defective electromechanical coupling does not appear to play a role in continued

LV dysfunction. Cardiac myosin heavy chain is intimately involved in cross-bridge cycling which generates the power-stroke required for myocyte contraction and force generation. There are two isoforms of myosin heavy chain (MHC) in cardiac muscle— alpha and beta, and their relative expression is related to heart size, heart rate, and species

[196]. While α-MHC is the predominant isoform in mice, rats and other rodents, β-MHC is the predominant form in rabbits and larger species [196]. Additionally, compared with

β-MHC of the same species, isolated α-MHC has a higher ATPase rate, faster velocity of actin motility and shorter crossbridge lifetime [196]. The difference in MHC percentage has been shown to alter cardiac function, since power output is 52% higher in rat myocyte fragments expressing 12% α-MHC than in in fragments expressing 0% α-MHC

[73]. Our data suggest that MHC isoform dysregulation might contribute to the cardiac dysfunction. This includes the correlation between decreased function in REV@8-Veh and REV@8-L4-8 with decreased α-to-β-MHC and increased function in REV@8-L8-19 with normalized α-to-β-MHC.

One of the potential limitations of our model is the use of rodents, in which α-

MHC is the predominant isoform; however, human and canine data suggests a role for alterations in the α-to-β-MHC isoform ratio in LV dysfunction. Similar to rodents, β-

MHC increases and α-MHC mRNA and protein levels drop from 20-30% to 10% and 7% to <1%, respectively in adult, human failing ventricles [34,157], and α-to-β-MHC decreases in dogs with microsphere emboli-induced HF [161]. Importantly, similar to our results, α-MHC increased in DCM patients that responded to β-blocker therapy [109], and α-to-β-MHC normalized in microsphere emboli HF dogs with improved LV function following the use of a cardiac support device [161]. Overall, these results suggest an important role MHC in LV functional recovery.

Diastolic relaxation is largely controlled through a combination of changes in

Ca2+ cycling, myofilament composition, myofilament Ca2+ desensitization, and increased crossbridge cycling kinetics [3,24]. Cardiac troponin I (cTnI) phosphorylation plays a key role in Ca2+ desensitization and cross-bridge cycling kinetics[178,142] and increases the rate of Ca2+ dissociation from cTnC, which is thought to be rate limiting for diastolic relaxation[118,103]. Accordingly, we measured cTnI phosphorylation at Ser-23/24 and demonstrate an increase in phosphorylation in REV@8-L8-19, which correlates with improved diastolic relaxation (Fig 15C). The increase in cTnI Ser-23/24 phosphorylation is similar to increased phosphorylation and increased diastolic relaxation in ACF-Levo but dissimilar from unchanged phosphorylation but increased diastolic relaxation in

REV@4-Levo (Chapter 2). We previously attributed the difference in cTnI Ser-23/24 phosphorylation at Week 8 in ACF-Levo and REV-Levo to difference in hemodynamic load. However, the increased phosphorylation in REV@8-L8-19 suggests that Levo can phosphorylate cTnI Ser-23/24 in the absence of hemodynamic overload.

In summary, we demonstrate that 1) delayed surgical reversal results in rapid structural remodeling without functional recovery establishing a new model for investigating persistent dysfunction following delayed surgical repair; 2) pre-reversal

Levo only transiently improved LV function, while post-reversal Levo improved LV function until Week 19, and 3) improved LV function correlates with normalization of α- to-β MHC and increased cTnI Ser-23/24 phosphorylation. Levo, or another compound that targets the myofilament, offers an alternate pharmacologic treatment option for post- operative patients with continued LV dysfunction. Further studies are needed to determine if Levo in combination with reversal improves the α-to-β-MHC ratio or if the ratio improves because of improved LV function.

Table 2: LV morphologic, echocardiographic and hemodynamic parameters in rats at Week 19 Sham ACF REV @ 4 wks REV @ 8 wks

N 8-10 7-10 8 8 Body weight 474.2 ± 12.0 476.9 ± 20.8 481.1 ± 22.0 471.4 ± 12.0 (g) Heart (g) 1.15 ± 0.03 2.26 ± 0.12**** 1.23 ± 0.05^^^^ 1.32 ± 0.04^^^^ Lung (g) 1.97 ± 0.40 2.42 ± 0.17**** 1.66 ± 0.04^^^^ 1.70 ± 0.07^^^^ LVEDD 8.2 ± 0.0 11.7 ± 0.2**** 8.5 ± 0.2^^^^ 8.2 ± 0.1^^^^ (mm) LVESD 4.7 ± 0.0 8.2 ± 0.4**** 5.3 ± 0.1**,^^^^ 5.6 ± (mm) 0.1****,^^^^ PWTd (mm) 1.7 ± 0.0 2.0 ± 0.1**** 1.8 ± 0.0^ 1.8 ± 0.0 PWTs (mm) 2.7 ± 0.1 2.8 ± 0.1 2.8 ± 0.1 2.6 ± 0.1 FS (%) 42.0 ± 0.2 30.2 ± 0.4**** 39.7 ± 31.7 ± 0.4**** 0.5****,^^^^ Dilation 0.42 ± 0.01 0.34 ± 0.01**** 0.42 ± 0.01^^^^ 0.45 ± 0.01^^^^ index HR 366 ± 6 323 ± 5 365 ± 8 328 ± 15 (beats/min) SV (µL) 181 ± 7 238 ± 7** 194 ± 10^ 183 ± 7^^^ EDV (µL) 358 ± 8 462 ± 19*** 416 ± 21 400 ± 14^

dP/dtmax 8853 ± 165 8850 ± 372 7553 ± 244 7118 ± 349**,^^ (mmHg/sec) Ees 0.857 ± 0.027 0.341 ± 0.824 ± 0.416 ± (mmHg/µL) 0.028**** 0.014^^^^ 0.027**** Ees, adj 0.848 ± 0.029 0.342 ± 0.813 ± 0.428 ± (mmHg/µL) 0.025**** 0.029^^^^ 0.024**** PRSW 107.7 ± 5.9 53.3 ± 4.3**** 116.4 ± 1.5^^^^ 53.3 ± 3.3**** (mmHg) PRSW, adj 104.5 ± 5.4 53.8 ± 4.7**** 87.4 ± 5.4^^^^ 57.4 ± 4.6**** (mmHg)

dP/dtmin -8587 ± 125 -6160 ± -7128 ± 229* -6294 ± 297**** 243**** (mmHg/sec) Tau (Weiss) 9.6 ± 0.2 12.0 ± 0.5** 10.8 ± 0.2 12.2 ± 0.6** LV diameter in end-diastole and end-systole (LVEDD and LVESD, respectively); posterior wall thickness in diastole and systole (PWTd and PWTs, respectively); percent fractional shortening (%FS), Dilation index (2xPWTd/LVEDD). Data are expressed as mean ± SEM. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001 vs Sham-Veh; ^ p<0.05, ^^ p<0.01, ^^^ p<0.001, ^^^^ p<0.0001 vs ACF-Veh. 87

Table 3: LV morphologic, echocardiographic and hemodynamic parameters in Levo-treated REV@8 rats at Week 19 Sham REV @ 8 wks REV @ 8 wks REV @ 8 wks Levo 4-8 wks Levo 8-19 wks N 8-10 8 8 9 Body weight 474.2 ± 12.0 471.4 ± 12.0 477.8 ± 12.2 466.8 ± 13.0 (g) Heart (g) 1.15 ± 0.03 1.32 ± 0.04 1.24 ± 0.03 1.32 ± 0.05 Lung (g) 1.97 ± 0.40 1.70 ± 0.07 1.62 ± 0.03 1.76 ± 0.07 LVEDD 8.2 ± 0.0 8.2 ± 0.1 8.2 ± 0.1 8.1 ± 0.1 (mm) LVESD 4.7 ± 0.0 5.6 ± 0.1**** 5.6 ± 0.1**** 4.7 ± 0.1^^^^ (mm) PWTd (mm) 1.7 ± 0.0 1.8 ± 0.0 1.8 ± 0.0 1.7 ± 0.0 PWTs (mm) 2.7 ± 0.1 2.6 ± 0.1 2.6 ± 0.1 2.7 ± 0.1 FS (%) 42.0 ± 0.2 31.7 ± 0.4**** 31.8 ± 0.4**** 42.4 ± 0.4^^^^ Dilation 0.42 ± 0.01 0.45 ± 0.01 0.44 ± 0.01 0.41 ± 0.0^ index HR 366 ± 6 328 ± 15 343 ± 13 384 ± 9^^ (beats/min) SV (µL) 181 ± 7 183 ± 7 189 ± 15 179 ± 9 EDV (µL) 358 ± 8 400 ± 14 415 ± 13 386 ± 14

dP/dtmax 8853 ± 165 7118 ± 349** 7222 ± 366* 10322 ± (mmHg/sec) 379^^^^ Ees 0.857 ± 0.027 0.416 ± 0.394 ± 0.874 ± (mmHg/µL) 0.027**** 0.021**** 0.031^^^^ Ees, adj 0.848 ± 0.029 0.428 ± 0.412 ± 0.859 ± (mmHg/µL) 0.024**** 0.031**** 0.029^^^^ PRSW 107.7 ± 5.9 53.3 ± 3.3**** 57.2 ± 4.8**** 104.1 ± 7.8^^^^ (mmHg) PRSW, adj 104.5 ± 5.4 57.4 ± 4.6**** 63.3 ± 5.8**** 98.7 ± 5.4^^^^ (mmHg)

dP/dtmin -8587 ± 125 -6294 ± -6148 ± -9583 ± 420^^^^ 297**** 252**** (mmHg/sec) Tau (Weiss) 9.3 ± 0.2 12.2 ± 0.6** 12.1 ± 0.3** 8.7 ± 0.3^^^^ LV diameter in end-diastole and end-systole (LVEDD and LVESD, respectively); posterior wall thickness in diastole and systole (PWTd and PWTs, respectively); percent fractional shortening (%FS), Dilation index (2xPWTd/LVEDD). Data are expressed as mean ± SEM. ** p<0.01, **** p<0.0001 vs Sham-Veh; ^ p<0.05, ^^ p<0.01, ^^^^ p<0.0001 vs ACF-Veh.

Chapter 4: Chronic levosimendan administration in rats with volume overload

heart failure preserves systolic and diastolic function and circumferential strain

Abstract

Aortocaval fistula (ACF)-induced volume overload (VO) heart failure (HF) results in progressive left ventricular (LV) dysfunction. This study investigated whether chronic intervention with the myofilament Ca2+ sensitizer levosimendan (Levo) could preserve cardiac function in ACF-induced VO longterm. ACF or sham surgery was performed in male Sprague-Dawley rats (200-240 g) 4 or 8 weeks prior to initiating treatment in ACF rats with Levo (1 mg/kg in drinking water) for 15 or 11 weeks, respectively. Levo improved systolic (%FS, Ees, PRSW) and diastolic (tau, dP/dtmin) function. Circumferential, radial and longitudinal strain increased in ACF in early-stage

HF, but decreased back to Sham levels by end-stage HF, respectively. Levo preserved circumferential, but not longitudinal or radial, strain in end-stage HF. These results demonstrate that Levo improves LV systolic and diastolic function. Additionally, speckle-tracking echocardiography indicated that Levo improves short-axis, but not long- axis function.

Introduction

In volume overload (VO) heart failure (HF), there is left ventricular (LV) chamber dilation, eccentric myocardial hypertrophy, progressive LV contractile dysfunction and heart failure (HF). In the aortocaval fistula (ACF) model of VO HF, the chronic increase in LV preload results in progressive LV pump failure, which can be classified into three clinically relevant stages: pre-HF (4 wks ACF; marked LV dilation with mild LV dysfunction), established HF (8 wks ACF; severe LV dilation with significant LV systolic and diastolic dysfunction) and end stage HF (15-21 wks ACF; pump failure with pulmonary congestion/edema) [40,66,76,160]. This model is useful for identifying molecular and cellular mechanisms that drive disease progression and for testing novel therapeutic approaches for improving LV structure and function

[59,84,143,198].

To date, functional studies in ACF rats have used a combination of echocardiography (i.e. % fractional shortening) and pressure-volume analysis (i.e. end- systolic elastance and preload recruitable stroke work) to evaluate systolic function

[66,76]. % fractional shortening (%FS) is non-invasive, can be used to follow the temporal changes from disease progression, and is unaffected by changes in chamber geometry which occur in ACF rats [32]. Alterations in %FS can be verified by pressure- volume analysis, which is considered the gold-standard for evaluating load-independent function [30,32]. While both of these methods are useful, there remain significant limitations when evaluating function in a volume overload model. %FS is affected by

90 preload and afterload, both of which are altered in ACF rats [32]. Additionally, pressure- volume analysis is highly invasive and often performed as a terminal procedure, often preventing serial measurements in the same rat [30,32]. A novel technique for measuring

LV systolic function has recently been identified that may complement our current techniques. Two-dimensional (2D) speckle-tracking echocardiography (STE) follows the ultrasonographic motion of speckles throughout the cardiac cycle to quantify myocardial deformation [93]. STE is a non-invasive echocardiographic technique that can be used to follow temporal changes during disease progression. Circumferential strain is considered an indicator of short-axis function, while longitudinal strain is considered an indicator of long-axis function [89]. Longitudinal and circumferential strain are reduced in rats with hypertensive, ischemic or toxic heart failure [81,93,101,131].

The strain profile in rats with VO has not been characterized; therefore, in the present study, we investigated the temporal changes in longitudinal, circumferential and radial strain during the progression of ACF-induced VO HF. To determine the sensitivity in detecting strain differences with drug-induced improved LV function, we treated subsets of ACF rats with levosimendan (Levo), a myofilament Ca2+ sensitizer, which has improved LV function in ACF rats. Our results provide the first evidence that Levo chronically improves LV function in VO HF, and by STE, Levo improves short-axis, but not long-axis function.

Materials and Methods

ACF Surgical Model in Rats

National Institutes of Health ―Guide for the Care and Use of Laboratory Animals,‖ (NIH publication No. 85-12, revised 1996). The protocol was approved by the Institutional

Animal Care and Use Committee of The Research Institute at Nationwide Children's

Hospital.

Cardiac VO was induced in rats at Week 0 (Fig 18) under isoflurane anesthesia

(2-2.5%) as previously described [66,76]. Following ventral abdominal midline incision, the abdominal aorta and caudal vena cava were exposed. An 18g short-bevel needle

(Becton Dickinson, Franklin Lakes, NJ) was inserted through the abdominal aorta into the vena cava creating an aortocaval fistula. The aortic puncture was sealed with cyanoacrylate glue. Shunt patency was confirmed by directly visualizing red arterial blood in the vena cava. The abdominal wall and skin were closed with 4-0 chromic gut

(Ethicon, Cincinnati, OH) and 4-0 silk (Ethicon, Cincinnati, OH), respectively. Sham animals underwent a similar procedure except for aortic puncture. The absence of pedal withdrawal reflex and respiratory rate were monitored during both procedures to ensure adequacy of anesthesia. Buprenorphine (0.03 mg/kg SC) was given immediately post- operatively and every 12 hours for 72 hours post-operatively as needed for pain.

Levosimendan treatment

Four or eight weeks post-ACF induction, subsets of rats were given Levo (L- enantiomer of ([4-(1,4,5,6-tetrahydro-4-methyl-6-oxo-3-pyridazinyl)phenyl]-hydrazono)- propanedinitrile; LKT Laboratories, MN) until the end of the study at Week 19. Levo

(0.0133 mg/mL) was prepared twice weekly in drinking water to deliver a dose of ~1 mg/kg/day (Fig 18). This dose has improved systolic and diastolic function in previous rodent studies and did not adversely impact water consumption [19,107,108]. A separate subset of ACF rats was given vehicle (Veh; water) for the duration of the study.

Figure 18: Experimental time course ACF was induced at Week 0. Subsets of ACF rats were given Levo (1 mg/kg/day in drinking water) from Weeks 4-19 or from weeks 8-19. Echo was performed biweekly, speckle-tracking echocardiography was performed at Weeks 0, 2, 4, 8, 14 and 19, and hemodynamics were performed at Week 19. 93

Echocardiography

Transthoracic echocardiograms (n=8-12/group) were performed biweekly with a broadband sector transducer utilizing a center frequency of 7.2-MHz (Artida, Toshiba

Medical Systems, Tustin, CA) under isoflurane anesthesia (1.5-2%) [66,76]. The average heart rate of these anesthetized rats ranged from 300-350 beats/min. M-mode images were obtained in parasternal short axis at the level of the papillary muscles to assess chamber diameters in end-systole (LVESD) and end-diastole (LVEDD) and LV posterior wall thickness in systole (PWTs) and diastole (PWTd). These indices were calculated: %

Fractional shortening (%FS) = (LVEDD-LVESD)/LVEDD x 100; Dilation index =

(2*PWTd)/LVEDD.

Measurement of LV hemodynamics

LV hemodynamics were assessed invasively using pressure–volume (PV) analysis during Week 19 [66,76]. Rats were anesthetized with 3% isoflurane to allow for intubation by tracheostomy. These rats were then ventilated with a pressure-controlled rodent ventilator (SAR-830, CWE Inc, Ardmore, PA), and maintained under 1.75% isoflurane anesthesia. The PV catheter (1.9F, Transonic SciSense, London, ON) was introduced into the LV via the right carotid artery. Following equilibration, baseline LV hemodynamic parameters were acquired using 5-10 consecutive PV loops. Preload was varied by brief occlusion of the vena cava to obtain preload-recruitable stroke work

(PRSW) and end-systolic elastance (Ees), the slope of the end-systolic pressure volume

94 relationship (ESPVR). Data was acquired and analyzed using Labscribe 2 software

(iWORX, Dover, NH).

Two-dimensional speckle-tracking analysis

During echocardiography, raw image videos with a frame rate ranging from 380-

400 frames/sec were captured and speckle-tracking analysis was performed at Weeks 0,

2, 4, 8, 14 and 19 (2D Wall Motion Tracking, Advanced Cardiology Package, Toshiba

Medical Systems). The endocardial border was manually traced, the LV was segmented automatically into equidistant segments by the software, and speckle tracking was performed on a frame-to-frame basis. Longitudinal strain values were obtained from the apical 4 chamber view, and radial and circumferential strain measurements from the basal, mid, and apical short-axis planes. Global strain values were calculated from the mean of regional values and stated in per cent (%) change compared to diastole.

Statistical analysis

Data are expressed as mean±SEM. Statistical analyses were performed using

GraphPad Prism V6.0 or SPSS V19 (ANCOVA only). One-way or two-way ANOVA or

ANCOVA, followed by Bonferroni's post-hoc test, was used to measure differences between groups. Survival data were evaluated by the Kaplan–Meier method with pairwise comparison conducted with the log-rank (Mantel-Cox). P < 0.05 was considered significant.

Results

Levo does not significantly alter LV chamber morphology

Compared with Sham, LVEDD and LVESD were increased in ACF-Veh and

ACF-Levo (Fig 19; Table 4). PWTd was increased in ACF-Veh and ACF-Levo and in combination with the larger increase in LVEDD, the dilation index was significantly lower in ACF indicating eccentric hypertrophy (Table 4). EDV and SV were measured with a pressure-volume catheter during Week 19. As expected for VO, EDV and SV were significantly increased in ACF-Veh and ACF-Levo (Table 4). The increases in

EDV and SV were accompanied by increased heart and lung weights in ACF-Veh and

ACF-Levo (Table 4), consistent with pathologic LV hypertrophy and pulmonary congestion and edema.

Figure 19: Effect of Levo on chamber dilation and % fractional shortening Representative M-mode images at week 19. B: Echocardiographic measurements: LVEDD and LVESD are LV diameter at end-diastole and end-systole, respectively. Data are means ± SEM (n=8-12/group). **** p<0.0001 vs Sham-Veh; ^^ ^^p<0.0001 vs ACF-Veh.

Levo improves LV systolic and diastolic function without altering survival

%FS, a hemodynamic load-dependent indicator of systolic function, was measured by serial echocardiography (Fig 19 and Table 4). %FS was decreased in ACF rats (10% and ~25% at 4 and 8-19 weeks, respectively), but improved in ACF rats treated with Levo (~30% increase compared with ACF-Veh at Week 19, p<0.0001). During 97

Figure 20: Levo improves LV systolic and diastolic function longterm A. Representative pressure-volume loops. B. LV systolic functional measurements: Ees, end systolic elastance; PRSW, preload recruitable stroke work. C. LV diastolic functional measurements: Relaxation constant, tau-Weiss; dP/dtmin, first derivative of LV pressure decline. Data are means ± SEM (n=7-8/group). ** p<0.01, *** p<0.001, **** p<0.0001 vs. Sham-Veh; ^^ p<0.01, ^^^ p<0.001, ^^^^ p<0.0001vs. ACF-Veh. 98

Week 19, invasive hemodynamics were performed to obtain load-independent measures of systolic function (Fig 20a). As previously reported [66], Ees and PRSW were decreased in ACF-Veh compared with Sham. Levo from Weeks 4-19 or 8-19 significantly increased Ees and PRSW over ACF-Veh (Ees: 117% and 125%, respectively; PRSW: 89 and 66%, respectively; Fig 20b, Table 4). In addition to these changes in systolic function, Levo also significantly altered LV diastolic relaxation. In

ACF-Veh, there was significantly impaired diastolic relaxation, (increased tau, a less negative dP/dtmin; Fig 20c, Table 4), which was improved in ACF rats treated with Levo from Weeks 4-19 or 8-19. Although Levo improved systolic and diastolic function, survival at 19 weeks was not significantly different amongst the ACF groups (p>0.05;

Fig 21).

Figure 21: Overall survival is comparable amongst the ACF groups Kaplan–Meier survival curve (Log-rank (Mantel-Cox)) shows the comparison between ACF-Veh and ACF-Levo rats (n=12-13/group). Animals were monitored for morbidity and mortality daily.

Levo preserved circumferential strain during end-stage HF

To characterize the effects of ACF-induced VO on strain, speckle-tracking analysis was performed. Representative echo tracings and strain graphs are shown in Fig

22. Circumferential and radial strain were evaluated in parasternal short-axis at the base, mid-wall and apex, and longitudinal strain was evaluated from an apical four chamber view (as diagramed in Fig 23a). Compared with Sham, longitudinal, circumferential and radial strain increased in ACF rats at Weeks 2, 4, and 8 (Fig 23b-d). Near the transition point to end-stage HF and in end-stage HF (Weeks 14 and 19, respectively), mid-wall circumferential strain was significantly higher than in ACF-Veh or Sham (Fig 23b). In contrast, at the end-stage HF timepoint, Levo treatment did not significantly alter longitudinal and radial strain in ACF rats (Figs 23c-d). Additionally, we evaluated rotation at the base and apex and there were no differences between Sham and ACF rats

(data not shown). This is in contrast to enhanced basal and apical rotation in human and canine patients with MR [207,208].

100

Figure 22: Representative echocardiograms and strain curves in Sham and ACF rats at Week 19 Representative mid-wall short-axis echocardiograms (A) with representative circumferential (B) and radial (C) strain. Representative apical-4 chamber echocardiograms (D) with representative longitudinal (E) strain. 101

Figure 23: Levo preserved circumferential, but not longitudinal or radial, strain A. Diagram of LV showing direction for circumferential, longitudinal and radial strain. Adapted from [41]. Mid-wall circumferential (B.), longitudinal (C.) and mid-wall radial strain (D.) increased following ACF. Levo preserved circumferential, but not longitudinal or radial, strain in end-stage HF (Weeks 14 and 19). Data represent means ± SEM from 7-8 rats/group. * p<0.05, **** p<0.0001 vs. Sham-Veh; ^ p<0.05 vs. ACF- Veh.

Discussion

In this study, we treated ACF rats with the myofilament Ca2+ sensitizer Levo from pre-HF to end-stage HF (i.e. 4-19 weeks post-ACF) or from established HF to end-stage

102

HF (i.e. 8-19 weeks post-ACF) to determine whether chronic, long-term treatment would maintain improved LV function. Additionally, we evaluated the effects of ACF-induced

VO on circumferential, longitudinal and radial strain and cardiac rotation. We show that despite persistent hemodynamic VO, Levo improves systolic (%FS, Ees and PRSW; Fig

2-3) and diastolic (tau and dP/dtmin; Fig 20) function without significantly improving

LVEDD (Fig 19). We also show that circumferential, longitudinal and radial strain increase during pre-HF and established HF, but only circumferential strain remains elevated at end-stage HF (Fig 22-23).

The rodent ACF model of VO replicates the hemodynamics and pathophysiology of VO, irrespective of the etiology [59,66,74,76,160,167]. This model enables us to study the effects of chronic pharmacologic intervention on LV function and overall survival. Levo interacts with and stabilizes Ca2+ saturated troponin C (cTnC) prolonging its interaction with cardiac troponin I (cTnI) [136]. This interaction promotes contractile force without increasing myocardial oxygen consumption or increasing the intracellular

Ca2+ transient [126,128,136]. Levo has previously been shown to improve LV systolic and diastolic function in spontaneous diabetic Goto-Kakizaki rats 12 weeks post-infarct,

Dahl/Rapp salt-sensitive hypertensive rats, in dogs with pacing-induced HF, and in human patients [19,107,113,136]. We are the first to evaluate Levo in a model of VO

HF. We previously treated ACF rats with Levo from pre-HF to established HF (i.e. 4-8 weeks post-ACF). In that study, we demonstrated improved LV systolic and diastolic function despite continued VO. In this study, there is continued improvement in LV

103 systolic and diastolic function, which is comparable to human patients in chronic NYHA

III-IV HF given oral Levo [82].

Although Levo improved LV function, there was no difference in survival amongst the ACF groups. This decreased survival differs from the improved survival reported in human patients that were given Levo chronically. This difference in survival between our model and the clinical data is likely due to differences in the pathophysiology of the specific type of HF studied. While the ACF model produces VO- mediated HF, in the human studies, hypertensive or ischemic disease led to HF without

VO [82]. The hemodynamic profiles of hypertensive, ischemic and VO HF differ, and the decreased survival in ACF rats suggests that pharmacologic therapy is insufficient in chronic VO. Additionally, Levo is not known to directly impact the renin-angiotensin- aldosterone (RAAS) or sympathetic nervous systems (SNS), which influence sodium retention [11]. Increased total body sodium (including increased extracellular fluid volume) has been suggested as a cause of the morbidity of decompensated HF [11].

Speckle-tracking echocardiography (STE) is emerging as a non-invasive technique to evaluate LV function. Two-dimensional STE follows the frame-to-frame motion of echo-dense speckles to measure myocardial deformation [111,148]. LV contraction is complex and involves circumferential shortening, radial thickening, longitudinal contraction, and rotation along the long axis of the LV [41,42].

Geometrically, the three dimensions of the heart are represented by longitudinal, radial

104 and circumferential axes (see Fig 23a) and strain (or deformation) is defined as the change in length of a myocardial segment relative to the length at end-diastole [41]. As such, positive strain occurs in longitudinal and circumferential lengthening and radial thickening while negative strain occurs in longitudinal and circumferential shortening and radial thinning [41]. Circumferential strain is considered an indicator of short-axis function, while longitudinal strain is considered an indicator of long-axis function [89].

While STE has been evaluated in ischemic, hypertensive, and toxic models in the rat [81,93,101,131], STE parameters have not been reported for VO models. This paper is the first one to report how STE is altered by VO and additionally demonstrates the effect of Levo therapeutic intervention. In this study during pre-HF and established HF, circumferential and longitudinal strain became more negative while radial strain became more positive in ACF rats compared with Sham (p<0.05). These findings are interpreted as increased longitudinal and circumferential shortening and increased radial thickening.

The increase in strain in early stage HF is similar to increased longitudinal and circumferential strain described in human patients with mild to moderate mitral regurgitation and dogs with mild to marked mitral regurgitation from myxomatous mitral valve disease [177,184,207-209]. By contrast, in the hypertensive Dahl/Rapp salt- sensitive rats, there was no change in strain in early stage disease despite an increase in

LV wall thickness [93]. Together, these results suggest differing strain responses in early

VO and early pressure overload, and the increased strain in VO may be due to activation of the Frank-Starling mechanism to compensate for the increased volume.

105

At the transition to end-stage HF and end-stage HF, radial, longitudinal and circumferential strain decreased to Sham levels in all ACF groups except for circumferential strain for Levo-treated rats (Fig 23). The preservation of circumferential strain in Levo-treated rats but decrease in vehicle-treated rats correlates with the improved LV function measured by %FS, Ees and PRSW (Figs 19-20). By contrast, longitudinal and radial strain decreased compared to earlier stage values, and there was no difference in longitudinal or radial strain amongst the ACF rats at end-stage. The decrease in longitudinal strain is similar to the decrease in Dahl/Rapp salt-sensitive rats where longitudinal strain decreases with chronic pressure overload [74]. The sustained improved circumferential but not longitudinal strain suggests that Levo improves short- axis, but not long-axis function in ACF rats. By contrast in human patients, Levo improved longitudinal strain, but not radial or circumferential strain; however, these patients had acute decompensation of chronic HF primarily caused by ischemic cardiomyopathy (~66%) and dilated cardiomyopathy (~33%) [137,138]. Additionally, the phosphodiesterase 3 inhibitor cilostazol improved circumferential, longitudinal and radial strain [204]. Together, these results suggest the pharmacologic effect on strain varies amongst inotropic agents as well as amongst disease conditions.

In summary, we demonstrate that 1) improved LV function is maintained with chronic administration of the myofilament Ca2+ sensitizer Levo in rats with ACF-induced

VO; 2) circumferential, radial and longitudinal strain increase during early HF, but decreased at end-stage HF; and 3) at end-stage HF, Levo improves circumferential but

106 not longitudinal strain suggesting that Levo improves short-axis, but not long-axis function. Although Levo, or another compound that targets the myofilament, may offer an alternate pharmacologic treatment option for patients with VO HF, excitement is tempered by the lack of difference in long-term survival. Further studies are needed to evaluate strain in reversed-VO HF and to evaluate strain with other positive inotropes.

107

Table 4: LV morphologic, echocardiographic and hemodynamic parameters in Sham, ACF-Veh and ACF-Levo rats at Week 19 Sham ACF ACF ACF Veh Veh Levo 4-19 wks Levo 8-19 wks N 8 10 8 8 Body weight (g) 487.9 ± 16.8 476.8 ± 15.9 507.9 ± 10.4 617.7 ± 20.6 Heart (g) 1.18 ± 0.04 2.33 ± 0.09**** 2.55 ± 0.11**** 3.18 ± 0.14**** Lung (g) 1.49 ± 0.03 2.59 ± 0.29 3.01 ± 0.29* 2.92 ± 0.44* LVEDD (mm) 8.2 ± 0.1 12.6 ± 0.4**** 12.9 ± 0.5**** 12.3 ± 0.5**** LVESD (mm) 4.7 ± 0.1 8.7 ± 0.3**** 7.7 ± 0.3****,^ 7.3 ± 0.3****,^^ PWTd (mm) 1.7 ± 0.0 2.0 ± 0.1**** 2.0 ± 0.0**** 1.9 ± 0.0*** PWTs (mm) 2.9 ± 0.1 3.0 ± 0.1 3.4 ± 0.1***,^^^ 3.4 ± 0.1**,^^ FS (%) 42.6 ± 0.3 30.5 ± 0.4**** 40.2 ± 0.6**,^^^^ 40.5 ± 0.5*,^^^^ Dilation index 0.41 ± 0.01 0.31 ± 0.01**** 0.31 ± 0.01**** 0.32 ± 0.01**** HR (beats/min) 372 ± 4 321 ± 7*** 370 ± 11^^ 365 ± 10^^ SV (µL) 183 ± 9 271 ± 11*** 282 ± 15*** 272 ± 16*** EDV (µL) 363 ± 15 574 ± 23**** 586 ± 27**** 543 ± 27***

dP/dtmax 9203 ± 207 8725 ± 395 12186 ± 11818 ± (mmHg/sec) 558***,^^^^ 326***,^^^^ Ees (mmHg/µL) 0.903 ± 0.032 0.310 ± 0.675 ± 0.707 ± 0.007**** 0.043***,^^^^ 0.050**,^^^^ Ees, adj 0.900 ± 0.037 0.314 ± 0.684 ± 0.708 ± (mmHg/µL) 0.032**** 0.043^^^^ 0.036^^^^ PRSW (mmHg) 106.0 ± 12.8 64.7 ± 7.1* 115.7 ± 10.4^^ 99.9 ± 5.0 PRSW, adj 106.9 ± 8.6 60.0 ± 7.5*** 113.1 ± 10.0^^ 99.6 ± 8.4^^ (mmHg)

dP/dtmin -9124 ± 261 -5828 ± 179*** -7506 ± 490**,^^ -7554 ± 194**,^^ (mmHg/sec) Tau (Weiss) 9.6 ± 0.2 12.1 ± 0.4*** 9.1 ± 0.4^^^^ 9.4 ± 0.3^^^ LV diameter in end-diastole and end-systole (LVEDD and LVESD, respectively); posterior wall thickness in diastole and systole (PWTd and PWTs, respectively); percent fractional shortening (%FS), Dilation index (2xPWTd/LVEDD). Data are expressed as mean ± SEM. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001 vs Sham-Veh; ^ p<0.05, ^^ p<0.01, ^^^ p<0.001, ^^^^ p<0.0001 vs ACF-Veh.

108

Chapter 5: Summary of findings and future directions

Abstract

The work presented in earlier chapters of this dissertation investigated the role of

Levo, a myofilament Ca2+ sensitizer, in improving LV function in rats with ACF-induced

VO HF and reversed VO HF. Differing combinations of pharmacologic and surgical intervention were investigated and include 1) Levo given during pre-HF (Weeks 4-8 post-

ACF) and reversed pre-HF (Weeks 4-8 post reversal at Week 4; Chapter 2), 2) Levo given prior to (Weeks 4-8 prior to reversal at Week 8) and following reversal during established HF (Weeks 8-19 following reversal at Week 8; Chapter 3), and 3) Levo given long term beginning in pre-HF (Weeks 4-19 post-ACF) or beginning in established HF

(Weeks 8-19 post-ACF; Chapter 4). Continuously-administered Levo improved LV function with and without hemodynamic load reduction. Because of this improved function, Levo offers a new therapeutic option in patients with VO HF. More broadly, therapeutic strategies targeting myofilament Ca2+ sensitization may provide a new therapeutic target for patients with VO HF.

Current clinical problem

Significant mitral regurgitation (MR) is estimated to affect >2 million Americans

[65], and the numbers of Americans with asymptomatic mitral regurgitation are 109 uncertain. Surgical intervention to repair or replace the regurgitant valves restores normal hemodynamic load; however, optimal timing for surgical intervention in these patients is controversial. According to American College of Cardiology (ACC)/American

Heart Association (AHA) guidelines, valve surgery is indicated for symptomatic patients and asymptomatic patients with LV systolic dysfunction, and surgery is considered reasonable for asymptomatic patients with pulmonary hypertension at rest and/or atrial fibrillation [22,49,79].

Optimal surgical timing for patients with mild to moderate to severe, asymptomatic, non-dysfunctional mitral regurgitation is controversial. On the one hand, these patients can remain asymptomatic for long periods [79] with one report of patients averaging 16 years from initial diagnosis to onset of symptoms [182] and another report of 30-40% of patients reaching the indication for surgery in 5 years [22]. On the other hand, patients with non-dysfunctional, severe MR had greater left ventricular (LV) contractility, less worsened LV efficiency and better LV mass regression post-surgery than their dysfunctional counterparts [79]. Additionally, in patients with asymptomatic severe MR, there are conflicting data about the increased risk of death prior to developing

Class I and IIa indications for surgery [22]. Complicating this decision is the lack of large-scale, randomized, prospective clinical trials.

Although surgery is the only method to reduce hemodynamic overload, preoperative pharmacologic therapy may delay time to surgical intervention, and post-

110 operative pharmacologic therapy may accelerate functional recovery. Overall, few pharmacologic agents have been studied in patients with MR [2,91,182], and with those studied, only a small clinical benefit has been noted. For example, in patients with chronic MR and preserved LV function, angiotensin-converting-enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs) have shown pooled mean decreases in regurgitant fraction, regurgitant volume and LV end-diastolic volume index of 7.7%, 8.3 mL and 11.5 mL/m2, respectively [182]. Additionally, since these therapies were dependent on blood pressure reduction, there is difficulty in interpreting LV ejection fraction measurements, one of the key determinants of surgical timing, and this may delay surgery [182]. These therapies could increase MR in patients with mitral valve prolapse and may prevent LV adaptive responses [182].

Summary of findings

Results from my studies demonstrate that continued Levo improves LV systolic and diastolic function regardless of treatment starting time point and without respect to hemodynamic load (Figs 5, 6, 14, 15, 19 and 20). These results mirror the results in other rodent and canine models of heart failure (i.e. Goto-Kakizaki rats post-infarct [107], in

Dahl/Rapp rats [19], in dogs with pacing-induced HF [113]) and in human patients with

HF [136]. Because of this improved function, Levo offers a new therapeutic option in patients with VO HF. More broadly, therapeutic strategies targeting myofilament Ca2+ sensitization may provide a new therapeutic target for patients with VO HF. Unlike ACE inhibitors and ARBs which lowered blood pressure and potentially delay surgery, our

111

Levo dose did not alter MAP in ACF or REV rats (Chapter 2). In this regard, Levo offers benefit over current pharmacologic options.

Interestingly, although there was improved function at 8 weeks in ACF rats given

Levo from 4-8 weeks post-ACF, this improved LV function did not persist when reversal was performed at Week 8 and Levo treatment was stopped (Fig 12). This decrease in function may be explained by the combination of the short elimination half-life of Levo in the rat and our proposed mechanism of improved systolic and diastolic function, which includes a role for Levo-induced myofilament Ca2+ sensitization and cMyBP-C and cTnI phosphorylation (Figs 9 & 24). Improved myofilament Ca2+ sensitivity requires the continued presence of Levo, which is rapidly depleted in the rat (t1/2 for Levo is 0.7h and t1/2 for OR-1896, the active metabolite, is 6.5 h) [106]. Similarly, cMyBP-C and cTnI phosphorylation may require the continued presence of Levo.

Figure 24: Mechanism of improved LV function Levo improves LV function through a combination of improved myofilament Ca2+ sensitivity, cMyBP-C and cTnI phosphorylation and normalization of α-to-β-MHC. 112

Additional results from my studies, in combination with canine and human data, suggest an important role for myosin heavy chain (MHC) isoform in LV function.

Cardiac MHC is intimately involved in cross-bridge cycling which generates the power- stroke required for myocyte contraction and force generation. There are two isoforms of

MHC in cardiac muscle—alpha and beta, and their relative expression is related to heart size, heart rate, and species [196]. Compared with β-MHC of the same species, isolated

α-MHC has a higher ATPase rate, faster velocity of actin motility and shorter cross- bridge lifetime [196]. Additionally, the difference in MHC percentage has been show to alter cardiac function, since power output is 52% higher in rat myocyte fragments expressing 12% α-MHC than in in fragments expressing 0% α-MHC [73]

In ACF rats, there was progressive LV dysfunction that correlated with decreased

α-to-β-myosin heavy chain (MHC). Levo given post-reversal at Weeks 4 and 8 improved

LV function at Weeks 8 and 19, respectively, which correlated with nearly normalized to normalized α-to-β-MHC, respectively. In the REV@8-L4-8 rats, Levo improved LV function at Weeks 6 and 8, but by Week 10, there was LV dysfunction which persisted until the end of the study at Week 19. This post-reversal dysfunction suggests that the underlying cause of dysfunction persists. In addition to the possible causes listed above, another possible cause of continued dysfunction is MHC isoform. In these rats at the time of reversal (Chapter 2), α-to-β-MHC remained decreased, compatible with a dysfunctional myofilament phenotype, and by the end of the study, α-to-β-MHC remained decreased. Interestingly, hemodynamic load reversal alone was insufficient to

113 normalize α-to-β-MHC in either rats reversed at 4 or 8 weeks. Instead, the combination of hemodynamic load reversal and Levo correlated with normalized α-to-β-MHC.

The importance of the role of MHC isoform in cardiac function and dysfunction is further highlighted by human and canine data. This is important because β-MHC is the predominant isoform in larger mammals [196]. Similar to rodents, β-MHC increases and

α-MHC mRNA and protein levels drop from 20-30% to 10% and 7% to <1%, respectively in adult, human failing ventricles [34,157], and α-to-β-MHC decreases in dogs with microsphere emboli-induced HF [161]. Importantly, similar to our results, α-

MHC increased in DCM patients that responded to β-blocker therapy [109], and α-to-β-

MHC normalized in microsphere emboli HF dogs with improved LV function following the use of a cardiac support device [161].

Functional studies are required for evaluating the efficacy of new therapeutic candidates. To date, functional studies in ACF rats have used a combination of echocardiography (i.e. % fractional shortening) and pressure-volume analysis (i.e. end- systolic elastance and preload recruitable stroke work) to evaluate systolic function

[66,76]. % fractional shortening (%FS) is non-invasive, can be used to follow the temporal changes of disease progression, and is unaffected by changes in chamber geometry (i.e. elliptical to spherical) which occur in ACF rats [32]. Alterations in %FS can be verified by pressure-volume analysis, which is considered the gold-standard for evaluating load-independent function [30,32]. While both of these methods are useful,

114 there remain significant limitations when evaluating function in a volume overload model. %FS is affected by preload and afterload, both of which are altered in ACF rats

[32]. Additionally, pressure-volume analysis is highly invasive and often performed as a terminal procedure, which prevents serial measurements in the same rat [30,32]. Because of these limitations, additional measures of function are desired.

Geometrically, the three dimensions of the heart are represented by longitudinal, radial and circumferential axes (see Fig 23a) and strain (or deformation) is defined as the change in length of a myocardial segment relative to the length at end-diastole [41]. As such, positive strain occurs in longitudinal and circumferential lengthening and radial thickening while negative strain occurs in longitudinal and circumferential shortening and radial thinning [41]. Circumferential strain is considered an indicator of short-axis function, while longitudinal strain is considered an indicator of long-axis function [89].

While STE has been evaluated in ischemic, hypertensive, and toxic models in the rat [81,93,101,131], STE parameters have not been reported for VO models. Our results from STE demonstrate that ACF increases circumferential, longitudinal and radial strain

115 during pre-HF and established HF. The increase in strain in early stage HF is similar to increased longitudinal and circumferential strain described in human patients with mild to moderate MR and dogs with mild to marked MR from myxomatous mitral valve disease

[177,184,207-209]. By contrast, in the hypertensive Dahl/Rapp salt-sensitive rats, there was no change in strain in early stage disease despite an increase in LV wall thickness

[93]. Together, these results suggest differing strain responses in early VO and early pressure overload, and the increased strain in VO may be due to activation of the Frank-

Starling mechanism to compensate for the increased volume.

In untreated ACF at end-stage HF, circumferential, longitudinal and radial strain decrease to Sham levels. In contrast, Levo preserves circumferential, but not longitudinal or radial, strain in rats with end-stage HF. The preserved circumferential strain in ACF-

Levo at end-stage HF, further suggests that Levo preserves short-axis function but not long-axis function. By contrast in human patients, Levo improved longitudinal strain, but not radial or circumferential strain; however, patients in acute decompensated HF primarily caused by ischemic cardiomyopathy (~66%) and dilated cardiomyopathy

(~33%) [137,138]. Additionally, the phosphodiesterase 3 inhibitor cilostazol improved circumferential, longitudinal and radial strain [204]. Together, these results suggest the pharmacologic effect on strain varies amongst inotropic agents as well as amongst disease conditions.

116

Clinically, use of Levo is controversial. Although early clinical trials demonstrated survival benefit, data from more recent trials was disappointing and there is question as to whether Levo is superior to classic inotropic agents like dobutamine [6].

Although we did not clearly demonstrate a survival benefit with Levo in continued VO, the results in rats given Levo post-reversal suggest that Levo, or another compound that targets the myofilament, offers an alternate pharmacologic treatment option for post- operative patients with continued LV dysfunction.

Study Limitations

The most significant limitations with this work include 1) limited invasive function and tissue sample timepoints, 2) use of MHC mRNA data to suggest changes at the protein level, 3) the field’s limited experience with speckle-tracking echocardiography, and 4) no investigation into durability of improved LV function in the

REV-Levo groups. The impact on our interpretations and conclusions differs with each of these limitations.

In the two 19 week studies (Chapters 3 and 4), invasive hemodynamics and tissue were collected at Week 19 only. While this does not impact the interpretations at Week

19, we lack a complete set of data at intermediate timepoints. For example, in the study where we investigated delayed reversal (Chapter 3), I would suggest adding invasive hemodynamics at Week 12. This timepoint corresponds with 4 weeks post-reversal, which is the early timepoint investigated in [76]. The hemodynamic data would provide

117 the load-independent data to strengthen the claim that there is LV dysfunction in the early post-reversal period. In the two Levo-treated groups, this timepoint corresponds with 4 weeks after Levo treatment ended and 4 weeks after Levo was started. Similar to vehicle-treated rats, invasive hemodynamics at this timepoint would confirm the echo findings. Additionally, tissue collected at this timepoint could be used to evaluate the cellular and subcellular changes associated with delayed reversal with and without treatment. With our current data set, we lack knowledge about cMyBP-C and cTnI phosphorylation as well as the α-to-β-MHC ratio at earlier timepoints.

The second limitation is that we are using MHC mRNA data to suggest changes at the protein level. According to Kurabayashi et al [94], the nucleotide and deduced amino acid sequence homology between α- and β-MHC are 95.1 % and 96.2%, respectively.

Based on Basic Local Alignment Search Tool (BLAST) analysis at the National Center for Biotechnology Information (NCBI), the nucleotide and amino acid sequence homology is 93% and 94%, respectively. Our MHC primers target untranslated regions, where there is decreased homology between α and β MHC. Because of this, we are more confident in our ability to detect differences in mRNA than in protein with immunoblot techniques. The disadvantage of this approach is that our interpretation assumes that the changes in mRNA mirror the changes in protein. For example, microRNAs (miR) can suppress target gene expression by inhibiting target mRNA translation or by promoting degradation [206]. Although miR-27a, miR-208a, miR-208b and miR-499 have been shown to regulate β-MHC gene expression [129,193], these miRs are reported to

118 upregulate β-MHC expression. In our model, β-MHC is increased in dysfunction, but downregulated with improved LV function. Thus, increases in these miRs would not explain the normalized α-to-β-MHC that we observed in REV@8-L8-19. To verify our

MHC mRNA findings, we could use non-immunoblot techniques, including SDS-PAGE gels stained with silver [9], and MALDI-TOF mass spectrometry [72].

The third limitation is that the field’s limited experience with speckle-tracking echocardiography (STE) in VO and following pharmacologic intervention limits our confidence in our data interpretation. STE results in rodent models of VO have not been previously published, and while STE has been used to evaluate function after doxorubicin, one STE study has evaluated LV systolic function following pharmacologic intervention in hypertensive patients given the beta-blocker bisoprolol[135]. While the lack of previous publication allows us to be first in the field, our ability and confidence in interpreting the findings is hindered by our own limited experience with the technique combined with lack of field knowledge.

Finally, with the exception of the REV@8-L4-8 group, we did not stop Levo treatment and continue to monitor LV function. This limits our ability to comment on the persistence of improved LV function. Ultimately, this hinders our ability to correlate normalization of α-to-β-MHC with persistence of improved LV function.

119

Future directions

Based on the findings presented in this dissertation, there are many new research questions. The two most exciting potential avenues include myosin heavy chain and short and long-axis function as measured by speckle-tracking echocardiography.

What is myosin heavy chain (MHC)’s role in functional recovery?

In ACF rats reversed at 4 and 8 weeks given Levo until the end of the study (i.e. 8 and 19 weeks, respectively), the α-to-β MHC ratio normalized (Figs 9 and 16), respectively. This finding raises several questions. First, does the normalized MHC isoform persist if Levo is discontinued? Second, does MHC isoform drive functional recovery or does it respond to functional recovery? Third, what role does MHC isoform play in VO in larger mammals (e.g. dogs and people)? To answer these questions, I would propose the following experiments.

Experiment 1: A) Determine if the improved LV function persists in REV-Levo rats when Levo is discontinued and B) determine if the normalized MHC ratio persists in

REV-Levo rats when Levo is discontinued.

Eight groups of rats will be used in these experiments (Fig. 25). In experiment 1a, rats will undergo Sham or ACF surgery at Week 0. Four weeks later, ACF rats will undergo reversal surgery (REV) and receive either vehicle (water) or Levo (1 mg/kg/day in drinking water) for an additional 4-8 weeks (Fig 25a). In experiment 1b, rats will similarly undergo Sham or ACF surgery at Week 0. Eight weeks later, ACF rats will

120

Figure 25: Experimental time course for proposed Experiment 1 ACF will be induced at Week 0. The ACF will be reversed (REV) at Week 4 (A) or Week 8 (B). Subsets of REV rats will be given Levo (1 mg/kg/day in drinking water) from Weeks 4-8 or 4-12 (A) or Weeks 8-19 or 8-23 (B). Echo will be performed biweekly and hemodynamics and tissue collection will be performed at Weeks 12 or 23.

undergo REV and receive either vehicle or Levo (1 mg/kg/day in drinking water) for an additional 11-15 weeks (Fig 25b). Echocardiography will be performed prior to the ACF surgery (baseline) and biweekly throughout the study duration. Invasive hemodynamic 121 measurements and tissue collection will be performed at 12 or 23weeks. Relative expression of MHC mRNA will be measured in LV tissue collected at 12 or 23 weeks.

Experiment 2: Characterize the pathophysiological relevance of MHC isoform shifts

One of the lingering questions is whether MHC isoform drives functional recovery or if it responds to functional recovery. To answer this question, ten groups of rats will be used (Fig. 26). In experiment 2a, rats will undergo Sham or ACF surgery at

Week 0 (Fig 26a). At the time of surgery, half of the ACF rats will be injected with a viral vector (to be determined) which will replace the native ventricular MHC such that

α-MHC cannot switch to β-MHC. Four weeks later, subsets of the ACF rats with and without native MHC will undergo reversal surgery (REV). Echocardiography will be performed prior to the ACF surgery (baseline) and biweekly throughout the study duration. Invasive hemodynamic measurements and tissue collection will be performed at

8 weeks. Relative expression of MHC mRNA will be measured in LV tissue collected at

8 weeks.

In experiment 2b, rats will undergo Sham or ACF surgery at Week 0 (Fig 26b).

Four weeks later, half of the ACF rats will undergo reversal surgery (REV). At this time, half of the remaining ACF rats and half of the REV rats will be injected with the same viral vector as in experiment 2a. Echocardiography will be performed prior to the ACF surgery (baseline) and biweekly throughout the study duration. Invasive hemodynamic

122 measurements and tissue collection will be performed at 8weeks. Relative expression of

MHC mRNA will be measured in LV tissue collected at 8 weeks.

Figure 26: Experimental time course for proposed Experiment 2 ACF will be induced at Week 0. In subsets of rats, the ACF will be reversed (REV) at Week 4. In Experiment 2A, subsets of ACF rats will be injected with a viral vector that permanently MHC to α-MHC (gray boxes). In Experiment 2B, subsets of ACF and REV rats will be injected with the same viral vector (gray star). Echo will be performed biweekly and hemodynamics and tissue collection will be performed at Week 8.

Experiment 3: Characterize the MHC phenotype in dogs with MR and people with MR and repaired MR

Relative expression of MHC mRNA will be measured in LV tissue collected from dogs with various stages of either experimentally induced MR or with natural MR.

123

Similarly, relative expression of MHC mRNA will be measured in LV tissue collected from people with various stages of natural MR before and after surgical valve repair.

What role does cardiac short-axis and long-axis function play in VO HF progression?

In ACF rats treated with Levo chronically from pre-HF to end-stage HF (i.e. 4-19 weeks post-ACF) and established HF to end-stage HF (i.e. 8-19 weeks post-ACF), there was improved circumferential, but not longitudinal, strain. One interpretation of these findings is that there is improved short axis but not long axis function. Additionally, survival at 19 weeks was no different in ACF-Levo and ACF-Veh rats. Based on these findings, my hypothesis is that impaired long-axis function contributes to VO HF progression. To test this hypothesis, sildenafil, a PDE5 inhibitor, will be used. In the rodent model of MR, sildenafil has previously been shown to improve LV function, exercise tolerance and survival [88]. This will allow us to first determine if improved LV function that leads to improved exercise tolerance and survival correlates with maintained long-axis function (Experiment 4). If sildenafil improves long-axis function, I then propose using sildenafil in the ACF model to determine if there are model-specific differences in long-axis function/survival (Experiment 5).

Experiment 4: Determine whether chronic sildenafil treatment in MR rats alters longitudinal and circumferential strain.

Three groups of rats will be used in these experiments (Fig. 27). Rats will undergo

Sham or mitral regurgitation (MR) surgery at Week 0. Four weeks later, subsets of MR

124 rats will receive either vehicle (water) or sildenafil (Sil; 20 mg/kg/kday in drinking water) for an additional 15 weeks. Echocardiography will be performed prior to the MR surgery

(baseline) and biweekly throughout the study duration. Circumferential, longitudinal and radial strain will be measured at Weeks 0, 2, 4, 8, 14, and 19. Invasive hemodynamic measurements and tissue collection will be performed at Week 19.

Figure 27: Experimental design for proposed Experiment 4 Mitral regurgitation (MR) was induced at Week 0. Subsets of MR rats were given sildenafil (Sil; 20 mg/kg/day in drinking water) from Weeks 4-19. Echo was performed biweekly, speckle-tracking echocardiography was performed at Weeks 0, 2, 4, 8, 14 and 19, and hemodynamics were performed at Week 19.

Experiment 5: Determine whether chronic sildenafil treatment in ACF rats improves LV function and survival as well as improves longitudinal and circumferential strain

Four groups of rats will be used in these experiments (Fig. 28). Rats will undergo

Sham or ACF surgery at Week 0. Four weeks later, subsets of ACF rats will receive 125 either vehicle (water), Levo (1 mg/kg/day in drinking water), or sildenafil (Sil; 20 mg/kg/kday in drinking water) for an additional 15 weeks. Echocardiography will be performed prior to the ACF surgery (baseline) and biweekly throughout the study duration. Circumferential, longitudinal and radial strain will be measured at Weeks 0, 2,

4, 8, 14, and 19. Invasive hemodynamic measurements and tissue collection will be performed at Week 19.

Figure 28: Experimental design for proposed Experiment 5 ACF was induced at Week 0. Subsets of ACF rats were given Levo (1 mg/kg/day in drinking water) or sildenafil (Sil; 20 mg/kg/day in drinking water) from Weeks 4-19. Echo was performed biweekly, speckle-tracking echocardiography was performed at Weeks 0, 2, 4, 8, 14 and 19, and hemodynamics were performed at Week 19.

126

Final concluding remarks

In summary, these data demonstrate that 1) Levo improves LV function in ACF rats with and without hemodynamic load reduction, 2) improved function correlates with increased myofilament Ca2+ sensitivity, normalization of the dysfunctional myosin heavy chain phenotype (i.e. normalization of α-to-β-MHC), and increased phosphorylation of cMyBP-C and cTnI, and 4) Levo improves short-axis but not long-axis function at end- stage HF. Although further work into the mechanism of improved LV function is needed, these results demonstrate that Levo offers a new therapeutic option in patients with VO HF. More broadly, therapeutic strategies targeting myofilament Ca2+ sensitization may provide a new therapeutic target for patients with VO HF.

127

References

1. Abassi Z, Goltsman I, Karram T, Winaver J, Hoffman A (2011) Aortocaval fistula in rat: a unique model of volume-overload congestive heart failure and cardiac hypertrophy. Journal of biomedicine & biotechnology 2011:729497. doi:10.1155/2011/729497

2. Ahmed MI, Aban I, Lloyd SG, Gupta H, Howard G, Inusah S, Peri K, Robinson J, Smith P, McGiffin DC, Schiros CG, Denney T, Jr., Dell'Italia LJ (2012) A randomized controlled phase IIb trial of beta(1)-receptor blockade for chronic degenerative mitral regurgitation. Journal of the American College of Cardiology 60 (9):833-838. doi:10.1016/j.jacc.2012.04.029

3. Ahmed MI, Gladden JD, Litovsky SH, Lloyd SG, Gupta H, Inusah S, Denney T, Jr., Powell P, McGiffin DC, Dell'Italia LJ (2010) Increased oxidative stress and cardiomyocyte myofibrillar degeneration in patients with chronic isolated mitral regurgitation and ejection fraction >60%. Journal of the American College of Cardiology 55 (7):671-679. doi:10.1016/j.jacc.2009.08.074

4. Ait Mou Y, Toth A, Cassan C, Czuriga D, de Tombe PP, Papp Z, Lacampagne A, Cazorla O (2011) Beneficial effects of SR33805 in failing myocardium. Cardiovascular research 91 (3):412-419. doi:10.1093/cvr/cvr096

5. Al-Jassar C, Bikker H, Overduin M, Chidgey M (2013) Mechanistic basis of desmosome-targeted diseases. Journal of molecular biology 425 (21):4006-4022. doi:10.1016/j.jmb.2013.07.035

6. Antoniades C, Tousoulis D, Koumallos N, Marinou K, Stefanadis C (2007) Levosimendan: beyond its simple inotropic effect in heart failure. Pharmacology & therapeutics 114 (2):184-197. doi:10.1016/j.pharmthera.2007.01.008

7. Arsenault M, Zendaoui A, Roussel E, Drolet MC, Dhahri W, Grenier A, Gascon S, Sarrhini O, Rousseau JA, Lecomte R, Couet J (2013) Angiotensin II-converting enzyme inhibition improves survival, ventricular remodeling, and myocardial

128

energetics in experimental aortic regurgitation. Circulation Heart failure 6 (5):1021-1028. doi:10.1161/circheartfailure.112.000045

8. Atkins CE, Keene BW, Brown WA, Coats JR, Crawford MA, DeFrancesco TC, Edwards NJ, Fox PR, Lehmkuhl LB, Luethy MW, Meurs KM, Petrie JP, Pipers FS, Rosenthal SL, Sidley JA, Straus JH (2007) Results of the veterinary enalapril trial to prove reduction in onset of heart failure in dogs chronically treated with enalapril alone for compensated, naturally occurring mitral valve insufficiency. Journal of the American Veterinary Medical Association 231 (7):1061-1069. doi:10.2460/javma.231.7.1061

9. Baptista MJ, Recaman M, Melo-Rocha G, Nogueira-Silva C, Roriz JM, Soares- Fernandes J, Gonzaga S, Santos M, Leite-Moreira A, Areias JC, Correia-Pinto J (2006) Myocardium expression of connexin 43, SERCA2a, and myosin heavy chain isoforms are preserved in nitrofen-induced congenital diaphragmatic hernia rat model. Journal of pediatric surgery 41 (9):1532-1538. doi:10.1016/j.jpedsurg.2006.05.016

10. Barefield D, Sadayappan S (2010) Phosphorylation and function of cardiac myosin binding protein-C in health and disease. Journal of molecular and cellular cardiology 48 (5):866-875. doi:10.1016/j.yjmcc.2009.11.014

11. Bart BA (2009) Treatment of congestion in congestive heart failure: ultrafiltration is the only rational initial treatment of volume overload in decompensated heart failure. Circulation Heart failure 2 (5):499-504. doi:10.1161/circheartfailure.109.863381

12. Beca S, Aschars-Sobbi R, Panama BK, Backx PH (2011) Regulation of murine cardiac function by phosphodiesterases type 3 and 4. Current opinion in pharmacology 11 (6):714-719. doi:10.1016/j.coph.2011.10.017

13. Belin RJ, Sumandea MP, Sievert GA, Harvey LA, Geenen DL, Solaro RJ, de Tombe PP (2011) Interventricular differences in myofilament function in experimental congestive heart failure. Pflugers Archiv : European journal of physiology 462 (6):795-809. doi:10.1007/s00424-011-1024-4

14. Bergh CH, Andersson B, Dahlstrom U, Forfang K, Kivikko M, Sarapohja T, Ullman B, Wikstrom G (2010) Intravenous levosimendan vs. dobutamine in acute

129

decompensated heart failure patients on beta-blockers. European journal of heart failure 12 (4):404-410. doi:10.1093/eurjhf/hfq032

15. Bernay F, Bland JM, Haggstrom J, Baduel L, Combes B, Lopez A, Kaltsatos V (2010) Efficacy of spironolactone on survival in dogs with naturally occurring mitral regurgitation caused by myxomatous mitral valve disease. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 24 (2):331-341. doi:10.1111/j.1939-1676.2009.0467.x

16. Berridge MJ, Bootman MD, Roderick HL (2003) Calcium signalling: dynamics, homeostasis and remodelling. Nature reviews Molecular cell biology 4 (7):517- 529. doi:10.1038/nrm1155

17. Bezold KL, Shaffer JF, Khosa JK, Hoye ER, Harris SP (2013) A gain-of-function mutation in the M-domain of cardiac myosin-binding protein-C increases binding to actin. The Journal of biological chemistry 288 (30):21496-21505. doi:10.1074/jbc.M113.474346

18. Bhavsar PK, Brand NJ, Felkin LE, Luther PK, Cullen ME, Yacoub MH, Barton PJ (2010) Clenbuterol induces cardiac myocyte hypertrophy via paracrine signalling and fibroblast-derived IGF-1. Journal of cardiovascular translational research 3 (6):688-695. doi:10.1007/s12265-010-9199-1

19. Biala A, Finckenberg P, Korpi A, Loytainen M, Martonen E, Levijoki J, Mervaala E (2011) Cardiovascular effects of the combination of levosimendan and valsartan in hypertensive Dahl/Rapp rats. Journal of physiology and pharmacology : an official journal of the Polish Physiological Society 62 (3):275-285

20. Biesiadecki BJ, Kobayashi T, Walker JS, Solaro RJ, de Tombe PP (2007) The troponin C G159D mutation blunts myofilament desensitization induced by troponin I Ser23/24 phosphorylation. Circulation research 100 (10):1486-1493. doi:10.1161/01.RES.0000267744.92677.7f

21. Biscione C, Forleo GB, Costantino MF, Sergnese O, Aluigi S, Sergi D, Mariano E, Dores E, Di Luozzo M, Romeo F (2014) Preoperative echocardiographic measures as predictors of left ventricular dysfunction after mitral valve repair. International journal of cardiology 171 (2):e35-36. doi:10.1016/j.ijcard.2013.11.120

130

22. Bonow RO (2012) Chronic Mitral Regurgitation and Aortic Regurgitation: Have Indications for Surgery Changed? Journal of the American College of Cardiology. doi:10.1016/j.jacc.2012.08.1025

23. Bonow RO (2013) Chronic mitral regurgitation and aortic regurgitation: have indications for surgery changed? Journal of the American College of Cardiology 61 (7):693-701. doi:10.1016/j.jacc.2012.08.1025

24. Borlaug BA, Kass DA (2006) Mechanisms of diastolic dysfunction in heart failure. Trends in cardiovascular medicine 16 (8):273-279. doi:10.1016/j.tcm.2006.05.003

25. Boron WF, Boulpaep EL (2012) Medical Physiology, 2e Updated Edition: with STUDENT CONSULT Online Access. Elsevier Health Sciences,

26. Boyle KL, Leech E (2012) A review of the pharmacology and clinical uses of pimobendan. Journal of veterinary emergency and critical care (San Antonio, Tex : 2001) 22 (4):398-408. doi:10.1111/j.1476-4431.2012.00768.x

27. Bregagnollo EA, Fortes AH, Cicogna AC (1999) Assessment of inotropic and vasodilating effects of milrinone lactate in patients with dilated cardiomyopathy and severe heart failure. Arquivos brasileiros de cardiologia 72 (2):149-160

28. Brown L (2005) Cardiac extracellular matrix: a dynamic entity. American journal of physiology Heart and circulatory physiology 289 (3):H973-974. doi:10.1152/ajpheart.00443.2005

29. Brown LA, Nunez DJ, Brookes CI, Wilkins MR (1995) Selective increase in endothelin-1 and endothelin A receptor subtype in the hypertrophied myocardium of the aorto-venacaval fistula rat. Cardiovascular research 29 (6):768-774

30. Burkhoff D, Mirsky I, Suga H (2005) Assessment of systolic and diastolic ventricular properties via pressure-volume analysis: a guide for clinical, translational, and basic researchers. American journal of physiology Heart and circulatory physiology 289 (2):H501-512. doi:10.1152/ajpheart.00138.2005

131

31. Butt RP, Bishop JE (1997) Mechanical load enhances the stimulatory effect of serum growth factors on cardiac fibroblast procollagen synthesis. Journal of molecular and cellular cardiology 29 (4):1141-1151. doi:10.1006/jmcc.1996.0347

32. Carabello BA (2002) Evolution of the study of left ventricular function: everything old is new again. Circulation 105 (23):2701-2703

33. Carabello BA (2012) Volume overload. Heart failure clinics 8 (1):33-42. doi:10.1016/j.hfc.2011.08.013

34. Carniel E, Taylor MR, Sinagra G, Di Lenarda A, Ku L, Fain PR, Boucek MM, Cavanaugh J, Miocic S, Slavov D, Graw SL, Feiger J, Zhu XZ, Dao D, Ferguson DA, Bristow MR, Mestroni L (2005) Alpha-myosin heavy chain: a sarcomeric gene associated with dilated and hypertrophic phenotypes of cardiomyopathy. Circulation 112 (1):54-59. doi:10.1161/circulationaha.104.507699

35. Carver W, Nagpal ML, Nachtigal M, Borg TK, Terracio L (1991) Collagen expression in mechanically stimulated cardiac fibroblasts. Circulation research 69 (1):116-122

36. Cavallero S, Gonzalez GE, Puyo AM, Roson MI, Perez S, Morales C, Hertig CM, Gelpi RJ, Fernandez BE (2007) Atrial natriuretic peptide behaviour and myocyte hypertrophic profile in combined pressure and volume-induced cardiac hypertrophy. Journal of hypertension 25 (9):1940-1950. doi:10.1097/HJH.0b013e3282435b1e

37. Chang AN, Harada K, Ackerman MJ, Potter JD (2005) Functional consequences of hypertrophic and dilated cardiomyopathy-causing mutations in alpha- tropomyosin. The Journal of biological chemistry 280 (40):34343-34349. doi:10.1074/jbc.M505014200

38. Chang AN, Potter JD (2005) Sarcomeric protein mutations in dilated cardiomyopathy. Heart failure reviews 10 (3):225-235. doi:10.1007/s10741-005- 5252-6

39. Chen PP, Patel JR, Powers PA, Fitzsimons DP, Moss RL (2012) Dissociation of structural and functional phenotypes in cardiac myosin-binding protein C

132

conditional knockout mice. Circulation 126 (10):1194-1205. doi:10.1161/circulationaha.111.089219

40. Chen YW, Pat B, Gladden JD, Zheng J, Powell P, Wei CC, Cui X, Husain A, Dell'italia LJ (2011) Dynamic molecular and histopathological changes in the extracellular matrix and inflammation in the transition to heart failure in isolated volume overload. American journal of physiology Heart and circulatory physiology 300 (6):H2251-2260. doi:10.1152/ajpheart.01104.2010

41. Cheung YF (2012) The role of 3D wall motion tracking in heart failure. Nature reviews Cardiology 9 (11):644-657. doi:10.1038/nrcardio.2012.128

42. Cho GY, Marwick TH, Kim HS, Kim MK, Hong KS, Oh DJ (2009) Global 2- dimensional strain as a new prognosticator in patients with heart failure. Journal of the American College of Cardiology 54 (7):618-624. doi:10.1016/j.jacc.2009.04.061

43. Chopra A, Patel A, Shieh AC, Janmey PA, Kresh JY (2012) alpha-Catenin localization and sarcomere self-organization on N-cadherin adhesive patterns are myocyte contractility driven. PloS one 7 (10):e47592. doi:10.1371/journal.pone.0047592

44. Clemen CS, Herrmann H, Strelkov SV, Schroder R (2013) Desminopathies: pathology and mechanisms. Acta neuropathologica 125 (1):47-75. doi:10.1007/s00401-012-1057-6

45. Cohen-Solal A, Logeart D, Huang B, Cai D, Nieminen MS, Mebazaa A (2009) Lowered B-type natriuretic peptide in response to levosimendan or dobutamine treatment is associated with improved survival in patients with severe acutely decompensated heart failure. Journal of the American College of Cardiology 53 (25):2343-2348. doi:10.1016/j.jacc.2009.02.058

46. Colson BA, Patel JR, Chen PP, Bekyarova T, Abdalla MI, Tong CW, Fitzsimons DP, Irving TC, Moss RL (2012) Myosin binding protein-C phosphorylation is the principal mediator of protein kinase A effects on thick filament structure in myocardium. Journal of molecular and cellular cardiology 53 (5):609-616. doi:10.1016/j.yjmcc.2012.07.012

133

47. Coulton AT, Stelzer JE (2012) Cardiac myosin binding protein C and its phosphorylation regulate multiple steps in the cross-bridge cycle of muscle contraction. Biochemistry 51 (15):3292-3301. doi:10.1021/bi300085x

48. Davies MK, Gibbs CR, Lip GY (2000) ABC of heart failure. Management: diuretics, ACE inhibitors, and nitrates. BMJ (Clinical research ed) 320 (7232):428-431

49. De Bonis M, Bolling SF (2013) Mitral valve surgery: wait and see vs. early operation. European heart journal 34 (1):13-19a. doi:10.1093/eurheartj/ehs248

50. De Mello WC, Danser AH (2000) Angiotensin II and the heart : on the intracrine renin-angiotensin system. Hypertension 35 (6):1183-1188

51. Dell'Italia LJ (2002) The renin-angiotensin system in mitral regurgitation: a typical example of tissue activation. Current cardiology reports 4 (2):97-103

52. Dell'italia LJ, Balcells E, Meng QC, Su X, Schultz D, Bishop SP, Machida N, Straeter-Knowlen IM, Hankes GH, Dillon R, Cartee RE, Oparil S (1997) Volume- overload cardiac hypertrophy is unaffected by ACE inhibitor treatment in dogs. The American journal of physiology 273 (2 Pt 2):H961-970

53. Du CK, Morimoto S, Nishii K, Minakami R, Ohta M, Tadano N, Lu QW, Wang YY, Zhan DY, Mochizuki M, Kita S, Miwa Y, Takahashi-Yanaga F, Iwamoto T, Ohtsuki I, Sasaguri T (2007) Knock-in mouse model of dilated cardiomyopathy caused by troponin mutation. Circulation research 101 (2):185-194. doi:10.1161/circresaha.106.146670

54. Duffy HS (2012) The molecular mechanisms of gap junction remodeling. Heart rhythm : the official journal of the Heart Rhythm Society 9 (8):1331-1334. doi:10.1016/j.hrthm.2011.11.048

55. Enriquez-Sarano M, Akins CW, Vahanian A (2009) Mitral regurgitation. Lancet 373 (9672):1382-1394. doi:10.1016/S0140-6736(09)60692-9

56. Evangelista A, Tornos P, Sambola A, Permayer-Miralda G (2006) Role of vasodilators in regurgitant valve disease. Current treatment options in cardiovascular medicine 8 (6):428-434

134

57. Flaim SF, Minteer WJ, Nellis SH, Clark DP (1979) Chronic arteriovenous shunt: evaluation of a model for heart failure in rat. The American journal of physiology 236 (5):H698-704

58. Fontes MS, van Veen TA, de Bakker JM, van Rijen HV (2012) Functional consequences of abnormal Cx43 expression in the heart. Biochimica et biophysica acta 1818 (8):2020-2029. doi:10.1016/j.bbamem.2011.07.039

59. Gardner JD, Murray DB, Voloshenyuk TG, Brower GL, Bradley JM, Janicki JS (2010) Estrogen attenuates chronic volume overload induced structural and functional remodeling in male rat hearts. American journal of physiology Heart and circulatory physiology 298 (2):H497-504. doi:10.1152/ajpheart.00336.2009

60. Garg V, Frishman WH (2013) A new approach to inotropic therapy in the treatment of heart failure: cardiac myosin activators in treatment of HF. Cardiology in review 21 (3):155-159. doi:10.1097/CRD.0b013e318275889c

61. Gautel M, Zuffardi O, Freiburg A, Labeit S (1995) Phosphorylation switches specific for the cardiac isoform of myosin binding protein-C: a modulator of cardiac contraction? The EMBO journal 14 (9):1952-1960

62. Giovannone S, Remo BF, Fishman GI (2012) Channeling diversity: gap junction expression in the heart. Heart rhythm : the official journal of the Heart Rhythm Society 9 (7):1159-1162. doi:10.1016/j.hrthm.2011.11.040

63. Govindan S, McElligott A, Muthusamy S, Nair N, Barefield D, Martin JL, Gongora E, Greis KD, Luther PK, Winegrad S, Henderson KK, Sadayappan S (2012) Cardiac myosin binding protein-C is a potential diagnostic biomarker for myocardial infarction. Journal of molecular and cellular cardiology 52 (1):154- 164. doi:10.1016/j.yjmcc.2011.09.011

64. Granzier H, Wu Y, Siegfried L, LeWinter M (2005) Titin: physiological function and role in cardiomyopathy and failure. Heart failure reviews 10 (3):211-223. doi:10.1007/s10741-005-5251-7

65. Grayburn PA, Weissman NJ, Zamorano JL (2012) Quantitation of mitral regurgitation. Circulation 126 (16):2005-2017. doi:10.1161/CIRCULATIONAHA.112.121590 135

66. Guggilam A, Hutchinson KR, West TA, Kelly AP, Galantowicz ML, Davidoff AJ, Sadayappan S, Lucchesi PA (2013) In vivo and in vitro cardiac responses to beta- adrenergic stimulation in volume-overload heart failure. Journal of molecular and cellular cardiology 57:47-58. doi:10.1016/j.yjmcc.2012.11.013

67. Gupta MK, Gulick J, James J, Osinska H, Lorenz JN, Robbins J (2013) Functional dissection of myosin binding protein C phosphorylation. Journal of molecular and cellular cardiology 64:39-50. doi:10.1016/j.yjmcc.2013.08.006

68. Haggstrom J, Boswood A, O'Grady M, Jons O, Smith S, Swift S, Borgarelli M, Gavaghan B, Kresken JG, Patteson M, Ablad B, Bussadori CM, Glaus T, Kovacevic A, Rapp M, Santilli RA, Tidholm A, Eriksson A, Belanger MC, Deinert M, Little CJ, Kvart C, French A, Ronn-Landbo M, Wess G, Eggertsdottir A, Lynne O'Sullivan M, Schneider M, Lombard CW, Dukes-McEwan J, Willis R, Louvet A, Difruscia R (2013) Longitudinal Analysis of Quality of Life, Clinical, Radiographic, Echocardiographic, and Laboratory Variables in Dogs with Myxomatous Mitral Valve Disease Receiving Pimobendan or Benazepril: The QUEST Study. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine. doi:10.1111/jvim.12181

69. Haggstrom J, Lord PF, Hoglund K, Ljungvall I, Jons O, Kvart C, Hansson K (2013) Short-Term Hemodynamic and Neuroendocrine Effects of Pimobendan and Benazapril in Dogs with Myxomatous Mitral Valve Disease and Congestive Heart Failure. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine. doi:10.1111/jvim.12217

70. Hamlin RL, Benitz AM, Ericsson GF, Cifelli S, Daurio CP (1996) Effects of enalapril on exercise tolerance and longevity in dogs with heart failure produced by iatrogenic mitral regurgitation. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 10 (2):85-87

71. Hasenfuss G (1998) Animal models of human cardiovascular disease, heart failure and hypertrophy. Cardiovascular research 39 (1):60-76

72. Helmke SM, Yen CY, Cios KJ, Nunley K, Bristow MR, Duncan MW, Perryman MB (2004) Simultaneous quantification of human cardiac alpha- and beta-myosin heavy chain proteins by MALDI-TOF mass spectrometry. Analytical chemistry 76 (6):1683-1689. doi:10.1021/ac035144l

136

73. Herron TJ, McDonald KS (2002) Small amounts of alpha-myosin heavy chain isoform expression significantly increase power output of rat cardiac myocyte fragments. Circulation research 90 (11):1150-1152

74. Houser SR, Margulies KB, Murphy AM, Spinale FG, Francis GS, Prabhu SD, Rockman HA, Kass DA, Molkentin JD, Sussman MA, Koch WJ, American Heart Association Council on Basic Cardiovascular Sciences CoCC, Council on Functional G, Translational B (2012) Animal models of heart failure: a scientific statement from the American Heart Association. Circulation research 111 (1):131- 150. doi:10.1161/RES.0b013e3182582523

75. Husse B, Briest W, Homagk L, Isenberg G, Gekle M (2007) Cyclical mechanical stretch modulates expression of collagen I and collagen III by PKC and tyrosine kinase in cardiac fibroblasts. American journal of physiology Regulatory, integrative and comparative physiology 293 (5):R1898-1907. doi:10.1152/ajpregu.00804.2006

76. Hutchinson KR, Guggilam A, Cismowski MJ, Galantowicz ML, West TA, Stewart JA, Jr., Zhang X, Lord KC, Lucchesi PA (2011) Temporal pattern of left ventricular structural and functional remodeling following reversal of volume overload heart failure. Journal of applied physiology (Bethesda, Md : 1985) 111 (6):1778-1788. doi:10.1152/japplphysiol.00691.2011

77. Hutchinson KR, Stewart JA, Jr., Lucchesi PA (2010) Extracellular matrix remodeling during the progression of volume overload-induced heart failure. Journal of molecular and cellular cardiology 48 (3):564-569. doi:10.1016/j.yjmcc.2009.06.001

78. Imasaka K, Tomita Y, Tanoue Y, Tominaga R, Tayama E, Onitsuka H, Ueda T (2013) Early mitral valve surgery for chronic severe mitral regurgitation optimizes left ventricular performance and left ventricular mass regression. The Journal of thoracic and cardiovascular surgery 146 (1):61-66. doi:10.1016/j.jtcvs.2012.05.056

79. Imasaka KI, Tomita Y, Tanoue Y, Tominaga R, Tayama E, Onitsuka H, Ueda T (2012) Early mitral valve surgery for chronic severe mitral regurgitation optimizes left ventricular performance and left ventricular mass regression. The Journal of thoracic and cardiovascular surgery. doi:10.1016/j.jtcvs.2012.05.056

137

80. Ishikawa T, Tanaka R, Suzuki S, Miyaishi Y, Akagi H, Iino Y, Fukushima R, Yamane Y (2010) The effect of angiotensin-converting enzyme inhibitors of left atrial pressure in dogs with mitral valve regurgitation. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 24 (2):342- 347. doi:10.1111/j.1939-1676.2009.0455.x

81. Ishizu T, Seo Y, Kameda Y, Kawamura R, Kimura T, Shimojo N, Xu D, Murakoshi N, Aonuma K (2014) Left Ventricular Strain and Transmural Distribution of Structural Remodeling in Hypertensive Heart Disease. Hypertension. doi:10.1161/hypertensionaha.113.02149

82. Jalanko M, Kivikko M, Harjola VP, Nieminen MS, Laine M (2011) Oral levosimendan improves filling pressure and systolic function during long-term treatment. Scandinavian cardiovascular journal : SCJ 45 (2):91-97. doi:10.3109/14017431.2011.557779

83. Janicki JS, Brower GL (2002) The role of myocardial fibrillar collagen in ventricular remodeling and function. Journal of cardiac failure 8 (6 Suppl):S319-325. doi:10.1054/jcaf.2002.129260

84. Jobe LJ, Melendez GC, Levick SP, Du Y, Brower GL, Janicki JS (2009) TNF-alpha inhibition attenuates adverse myocardial remodeling in a rat model of volume overload. American journal of physiology Heart and circulatory physiology 297 (4):H1462-1468. doi:10.1152/ajpheart.00442.2009

85. Juric D, Yao X, Thandapilly S, Louis X, Cantor E, Chaze B, Wojciechowski P, Vasanji Z, Yang T, Wigle J, Netticadan T (2010) Defects in ryanodine receptor function are associated with systolic dysfunction in rats subjected to volume overload. Experimental physiology 95 (8):869-879. doi:10.1113/expphysiol.2009.052100

86. Katz AM (2010) Physiology of the Heart. Wolters Kluwer Health/Lippincott Williams & Wilkins Health,

87. Kim KH, Kim YJ, Lee SP, Kim HK, Seo JW, Sohn DW, Oh BH, Park YB (2011) Survival, exercise capacity, and left ventricular remodeling in a rat model of chronic mitral regurgitation: serial echocardiography and pressure-volume analysis. Korean circulation journal 41 (10):603-611. doi:10.4070/kcj.2011.41.10.603 138

88. Kim KH, Kim YJ, Ohn JH, Yang J, Lee SE, Lee SW, Kim HK, Seo JW, Sohn DW (2012) Long-term effects of sildenafil in a rat model of chronic mitral regurgitation: benefits of ventricular remodeling and exercise capacity. Circulation 125 (11):1390-1401. doi:10.1161/circulationaha.111.065300

89. Kim MS, Kim YJ, Kim HK, Han JY, Chun HG, Kim HC, Sohn DW, Oh BH, Park YB (2009) Evaluation of left ventricular short- and long-axis function in severe mitral regurgitation using 2-dimensional strain echocardiography. American heart journal 157 (2):345-351. doi:10.1016/j.ahj.2008.10.004

90. Kleaveland JP, Kussmaul WG, Vinciguerra T, Diters R, Carabello BA (1988) Volume overload hypertrophy in a closed-chest model of mitral regurgitation. The American journal of physiology 254 (6 Pt 2):H1034-1041

91. Knirsch W, Tlach L, Stambach D, Bauersfeld U (2010) Angiotensin-converting enzyme inhibitors in pediatric patients with mitral valve regurgitation-case- control study and review of the literature. Congenital heart disease 5 (3):278-284. doi:10.1111/j.1747-0803.2010.00407.x

92. Kooij V, Holewinski RJ, Murphy AM, Van Eyk JE (2013) Characterization of the cardiac myosin binding protein-C phosphoproteome in healthy and failing human hearts. Journal of molecular and cellular cardiology 60:116-120. doi:10.1016/j.yjmcc.2013.04.012

93. Koshizuka R, Ishizu T, Kameda Y, Kawamura R, Seo Y, Aonuma K (2013) Longitudinal strain impairment as a marker of the progression of heart failure with preserved ejection fraction in a rat model. Journal of the American Society of Echocardiography : official publication of the American Society of Echocardiography 26 (3):316-323. doi:10.1016/j.echo.2012.11.015

94. Kurabayashi M, Tsuchimochi H, Komuro I, Takaku F, Yazaki Y (1988) Molecular cloning and characterization of human cardiac alpha- and beta-form myosin heavy chain complementary DNA clones. Regulation of expression during development and pressure overload in human atrium. The Journal of clinical investigation 82 (2):524-531. doi:10.1172/jci113627

95. Kuster DW, Bawazeer AC, Zaremba R, Goebel M, Boontje NM, van der Velden J (2012) Cardiac myosin binding protein C phosphorylation in cardiac disease.

139

Journal of muscle research and cell motility 33 (1):43-52. doi:10.1007/s10974- 011-9280-7

96. Lahtinen P, Pitkanen O, Polonen P, Turpeinen A, Kiviniemi V, Uusaro A (2011) Levosimendan reduces heart failure after cardiac surgery: a prospective, randomized, placebo-controlled trial. Critical care medicine 39 (10):2263-2270. doi:10.1097/CCM.0b013e3182227b97

97. Lefebvre HP, Brown SA, Chetboul V, King JN, Pouchelon JL, Toutain PL (2007) Angiotensin-converting enzyme inhibitors in veterinary medicine. Current pharmaceutical design 13 (13):1347-1361

98. Lefkowitz RJ, Rockman HA, Koch WJ (2000) Catecholamines, cardiac beta- adrenergic receptors, and heart failure. Circulation 101 (14):1634-1637

99. Levick SP, Melendez GC, Plante E, McLarty JL, Brower GL, Janicki JS (2011) Cardiac mast cells: the centrepiece in adverse myocardial remodelling. Cardiovascular research 89 (1):12-19. doi:10.1093/cvr/cvq272

100. Li J, Lu H, Plante E, Melendez GC, Levick SP, Janicki JS (2012) Stem cell factor is responsible for the rapid response in mature mast cell density in the acutely stressed heart. Journal of molecular and cellular cardiology 53 (4):469-474. doi:10.1016/j.yjmcc.2012.07.011

101. Liao SS, Ruan QY, Lin MY, Yan L (2012) Value of segmental myocardial strain by 2-dimensional strain echocardiography for assessment of scar area induced in a rat model of myocardial infarction. Cardiovascular ultrasound 10:17. doi:10.1186/1476-7120-10-17

102. Lim CC, Yang H, Yang M, Wang CK, Shi J, Berg EA, Pimentel DR, Gwathmey JK, Hajjar RJ, Helmes M, Costello CE, Huo S, Liao R (2008) A novel mutant cardiac troponin C disrupts molecular motions critical for calcium binding affinity and cardiomyocyte contractility. Biophysical journal 94 (9):3577-3589. doi:10.1529/biophysj.107.112896

103. Little SC, Biesiadecki BJ, Kilic A, Higgins RS, Janssen PM, Davis JP (2012) The rates of Ca2+ dissociation and cross-bridge detachment from ventricular

140

myofibrils as reported by a fluorescent cardiac troponin C. The Journal of biological chemistry 287 (33):27930-27940. doi:10.1074/jbc.M111.337295

104. Liu Y, Dillon AR, Tillson M, Makarewich C, Nguyen V, Dell'Italia L, Sabri AK, Rizzo V, Tsai EJ (2013) Volume overload induces differential spatiotemporal regulation of myocardial soluble guanylyl cyclase in eccentric hypertrophy and heart failure. Journal of molecular and cellular cardiology 60:72-83. doi:10.1016/j.yjmcc.2013.03.019

105. Lou Q, Janardhan A, Efimov IR (2012) Remodeling of calcium handling in human heart failure. Advances in experimental medicine and biology 740:1145-1174. doi:10.1007/978-94-007-2888-2_52

106. Louhelainen M, Merasto S, Finckenberg P, Vahtola E, Kaheinen P, Levijoki J, Mervaala E (2010) Effects of the calcium sensitizer OR-1896, a metabolite of levosimendan, on post-infarct heart failure and cardiac remodelling in diabetic Goto-Kakizaki rats. British journal of pharmacology 160 (1):142-152. doi:10.1111/j.1476-5381.2010.00680.x

107. Louhelainen M, Vahtola E, Forsten H, Merasto S, Kyto V, Finckenberg P, Leskinen H, Kaheinen P, Tikkanen I, Levijoki J, Mervaala E (2009) Oral levosimendan prevents postinfarct heart failure and cardiac remodeling in diabetic Goto- Kakizaki rats. Journal of hypertension 27 (10):2094-2107. doi:10.1097/HJH.0b013e32832f0ce4

108. Louhelainen M, Vahtola E, Kaheinen P, Leskinen H, Merasto S, Kyto V, Finckenberg P, Colucci WS, Levijoki J, Pollesello P, Haikala H, Mervaala EM (2007) Effects of levosimendan on cardiac remodeling and cardiomyocyte apoptosis in hypertensive Dahl/Rapp rats. British journal of pharmacology 150 (7):851-861. doi:10.1038/sj.bjp.0707157

109. Lowes BD, Gilbert EM, Abraham WT, Minobe WA, Larrabee P, Ferguson D, Wolfel EE, Lindenfeld J, Tsvetkova T, Robertson AD, Quaife RA, Bristow MR (2002) Myocardial gene expression in dilated cardiomyopathy treated with beta- blocking agents. The New England journal of medicine 346 (18):1357-1365. doi:10.1056/NEJMoa012630

141

110. Lu QW, Wu XY, Morimoto S (2013) Inherited cardiomyopathies caused by troponin mutations. Journal of geriatric cardiology : JGC 10 (1):91-101. doi:10.3969/j.issn.1671-5411.2013.01.014

111. Marwick TH, Leano RL, Brown J, Sun JP, Hoffmann R, Lysyansky P, Becker M, Thomas JD (2009) Myocardial strain measurement with 2-dimensional speckle- tracking echocardiography: definition of normal range. JACC Cardiovascular imaging 2 (1):80-84. doi:10.1016/j.jcmg.2007.12.007

112. Mascle S, Schnell F, Thebault C, Corbineau H, Laurent M, Hamonic S, Veillard D, Mabo P, Leguerrier A, Donal E (2012) Predictive value of global longitudinal strain in a surgical population of organic mitral regurgitation. Journal of the American Society of Echocardiography : official publication of the American Society of Echocardiography 25 (7):766-772. doi:10.1016/j.echo.2012.04.009

113. Masutani S, Cheng HJ, Hyttila-Hopponen M, Levijoki J, Heikkila A, Vuorela A, Little WC, Cheng CP (2008) Orally available levosimendan dose-related positive inotropic and lusitropic effect in conscious chronically instrumented normal and heart failure dogs. The Journal of pharmacology and experimental therapeutics 325 (1):236-247. doi:10.1124/jpet.107.134940

114. McGinley JC, Berretta RM, Chaudhary K, Rossman E, Bratinov GD, Gaughan JP, Houser S, Margulies KB (2007) Impaired contractile reserve in severe mitral valve regurgitation with a preserved ejection fraction. European journal of heart failure 9 (9):857-864. doi:10.1016/j.ejheart.2007.05.013

115. McLaughlin HM, Kelly MA, Hawley PP, Darras BT, Funke B, Picker J (2013) Compound heterozygosity of predicted loss-of-function DES variants in a family with recessive desminopathy. BMC medical genetics 14:68. doi:10.1186/1471- 2350-14-68

116. McLendon PM, Robbins J (2011) Desmin-related cardiomyopathy: an unfolding story. American journal of physiology Heart and circulatory physiology 301 (4):H1220-1228. doi:10.1152/ajpheart.00601.2011

117. Melenovsky V (2013) Cardiac Adaptation to Volume Overload. In: Ostadal B, Dhalla NS (eds) Cardiac Adaptations, vol 4. Advances in Biochemistry in Health and Disease. Springer New York, pp 167-199. doi:10.1007/978-1-4614-5203-4_9

142

118. Memo M, Leung MC, Ward DG, Dos Remedios C, Morimoto S, Zhang L, Ravenscroft G, McNamara E, Nowak KJ, Marston SB, Messer AE (2013) Familial dilated cardiomyopathy mutations uncouple troponin I phosphorylation from changes in myofibrillar Ca2+ sensitivity. Cardiovascular research. doi:10.1093/cvr/cvt071

119. Memo M, Leung MC, Ward DG, dos Remedios C, Morimoto S, Zhang L, Ravenscroft G, McNamara E, Nowak KJ, Marston SB, Messer AE (2013) Familial dilated cardiomyopathy mutations uncouple troponin I phosphorylation from changes in myofibrillar Ca(2)(+) sensitivity. Cardiovascular research 99 (1):65-73. doi:10.1093/cvr/cvt071

120. Mirza M, Marston S, Willott R, Ashley C, Mogensen J, McKenna W, Robinson P, Redwood C, Watkins H (2005) Dilated cardiomyopathy mutations in three thin filament regulatory proteins result in a common functional phenotype. The Journal of biological chemistry 280 (31):28498-28506. doi:10.1074/jbc.M412281200

121. Mogensen J, Murphy RT, Shaw T, Bahl A, Redwood C, Watkins H, Burke M, Elliott PM, McKenna WJ (2004) Severe disease expression of cardiac troponin C and T mutations in patients with idiopathic dilated cardiomyopathy. Journal of the American College of Cardiology 44 (10):2033-2040. doi:10.1016/j.jacc.2004.08.027

122. Mohamed AS, Dignam JD, Schlender KK (1998) Cardiac myosin-binding protein C (MyBP-C): identification of protein kinase A and protein kinase C phosphorylation sites. Archives of biochemistry and biophysics 358 (2):313-319. doi:10.1006/abbi.1998.0857

123. Moore JR, Leinwand L, Warshaw DM (2012) Understanding cardiomyopathy phenotypes based on the functional impact of mutations in the myosin motor. Circulation research 111 (3):375-385. doi:10.1161/circresaha.110.223842

124. Morimoto S (2008) Sarcomeric proteins and inherited cardiomyopathies. Cardiovascular research 77 (4):659-666. doi:10.1093/cvr/cvm084

125. Morimoto S, Lu QW, Harada K, Takahashi-Yanaga F, Minakami R, Ohta M, Sasaguri T, Ohtsuki I (2002) Ca(2+)-desensitizing effect of a deletion mutation Delta K210 in cardiac troponin T that causes familial dilated cardiomyopathy. 143

Proceedings of the National Academy of Sciences of the United States of America 99 (2):913-918. doi:10.1073/pnas.022628899

126. Muller S, How OJ, Jakobsen O, Hermansen SE, Rosner A, Stenberg TA, Myrmel T (2010) Oxygen-wasting effect of inotropy: is there a need for a new evaluation? An experimental large-animal study using dobutamine and levosimendan. Circulation Heart failure 3 (2):277-285. doi:10.1161/CIRCHEARTFAILURE.109.865519

127. Murray DB, Gardner JD, Brower GL, Janicki JS (2008) Effects of nonselective endothelin-1 receptor antagonism on cardiac mast cell-mediated ventricular remodeling in rats. American journal of physiology Heart and circulatory physiology 294 (3):H1251-1257. doi:10.1152/ajpheart.00622.2007

128. Nieminen MS, Pollesello P, Vajda G, Papp Z (2009) Effects of levosimendan on the energy balance: preclinical and clinical evidence. Journal of cardiovascular pharmacology 53 (4):302-310. doi:10.1097/FJC.0b013e31819c9a17

129. Nishi H, Ono K, Horie T, Nagao K, Kinoshita M, Kuwabara Y, Watanabe S, Takaya T, Tamaki Y, Takanabe-Mori R, Wada H, Hasegawa K, Iwanaga Y, Kawamura T, Kita T, Kimura T (2011) MicroRNA-27a regulates beta cardiac myosin heavy chain gene expression by targeting thyroid hormone receptor beta1 in neonatal rat ventricular myocytes. Molecular and cellular biology 31 (4):744-755. doi:10.1128/mcb.00581-10

130. Ohba T, Watanabe H, Murakami M, Takahashi Y, Iino K, Kuromitsu S, Mori Y, Ono K, Iijima T, Ito H (2007) Upregulation of TRPC1 in the development of cardiac hypertrophy. Journal of molecular and cellular cardiology 42 (3):498-507. doi:10.1016/j.yjmcc.2006.10.020

131. Oliveira MS, Melo MB, Carvalho JL, Melo IM, Lavor MS, Gomes DA, de Goes AM, Melo MM (2013) Doxorubicin Cardiotoxicity and Cardiac Function Improvement After Stem Cell Therapy Diagnosed by Strain Echocardiography. Journal of cancer science & therapy 5 (2):52-57. doi:10.4172/1948-5956.1000184

132. Olson TM, Illenberger S, Kishimoto NY, Huttelmaier S, Keating MT, Jockusch BM (2002) Metavinculin mutations alter actin interaction in dilated cardiomyopathy. Circulation 105 (4):431-437

144

133. Olson TM, Michels VV, Thibodeau SN, Tai YS, Keating MT (1998) Actin mutations in dilated cardiomyopathy, a heritable form of heart failure. Science (New York, NY) 280 (5364):750-752

134. Palmer BM, Schmitt JP, Seidman CE, Seidman JG, Wang Y, Bell SP, Lewinter MM, Maughan DW (2013) Elevated rates of force development and MgATP binding in F764L and S532P myosin mutations causing dilated cardiomyopathy. Journal of molecular and cellular cardiology 57:23-31. doi:10.1016/j.yjmcc.2012.12.022

135. Palmieri V, Russo C, Palmieri EA, Pezzullo S, Celentano A (2009) Changes in components of left ventricular mechanics under selective beta-1 blockade: insight from traditional and new technologies in echocardiography. European journal of echocardiography : the journal of the Working Group on Echocardiography of the European Society of Cardiology 10 (6):745-752. doi:10.1093/ejechocard/jep055

136. Papp Z, Edes I, Fruhwald S, De Hert SG, Salmenpera M, Leppikangas H, Mebazaa A, Landoni G, Grossini E, Caimmi P, Morelli A, Guarracino F, Schwinger RH, Meyer S, Algotsson L, Wikstrom BG, Jorgensen K, Filippatos G, Parissis JT, Gonzalez MJ, Parkhomenko A, Yilmaz MB, Kivikko M, Pollesello P, Follath F (2012) Levosimendan: molecular mechanisms and clinical implications: consensus of experts on the mechanisms of action of levosimendan. International journal of cardiology 159 (2):82-87. doi:10.1016/j.ijcard.2011.07.022

137. Paraskevaidis IA, Bistola V, Ikonomidis I, Parissis JT, Papadopoulos C, Filippatos G, Kremastinos DT (2008) Usefulness of dobutamine-induced changes of the two-dimensional longitudinal deformation predict clinical and neurohumoral improvement in men after levosimendan treatment in acutely decompensated chronic heart failure. The American journal of cardiology 102 (9):1225-1229. doi:10.1016/j.amjcard.2008.06.060

138. Paraskevaidis IA, Ikonomidis I, Parissis J, Papadopoulos C, Stassinos V, Bistola V, Anastasiou-Nana M (2012) Dobutamine-induced changes of left atrial two- dimensional deformation predict clinical and neurohumoral improvement after levosimendan treatment in patients with acutely decompensated chronic heart failure. International journal of cardiology 157 (1):31-37. doi:10.1016/j.ijcard.2010.11.007

139. Pat B, Killingsworth C, Denney T, Zheng J, Powell P, Tillson M, Dillon AR, Dell'Italia LJ (2008) Dissociation between cardiomyocyte function and 145

remodeling with beta-adrenergic receptor blockade in isolated canine mitral regurgitation. American journal of physiology Heart and circulatory physiology 295 (6):H2321-2327. doi:10.1152/ajpheart.00746.2008

140. Pathak A, Lebrin M, Vaccaro A, Senard JM, Despas F (2013) Pharmacology of levosimendan: inotropic, vasodilatory and cardioprotective effects. Journal of clinical pharmacy and therapeutics. doi:10.1111/jcpt.12067

141. Peddle GD, Singletary GE, Reynolds CA, Trafny DJ, Machen MC, Oyama MA (2012) Effect of torsemide and furosemide on clinical, laboratory, radiographic and quality of life variables in dogs with heart failure secondary to mitral valve disease. Journal of veterinary cardiology : the official journal of the European Society of Veterinary Cardiology 14 (1):253-259. doi:10.1016/j.jvc.2012.01.003

142. Pena JR, Wolska BM (2004) Troponin I phosphorylation plays an important role in the relaxant effect of beta-adrenergic stimulation in mouse hearts. Cardiovascular research 61 (4):756-763. doi:10.1016/j.cardiores.2003.12.019

143. Perry GJ, Mori T, Wei CC, Xu XY, Chen YF, Oparil S, Lucchesi P, Dell'Italia LJ (2001) Genetic variation in angiotensin-converting enzyme does not prevent development of cardiac hypertrophy or upregulation of angiotensin II in response to aortocaval fistula. Circulation 103 (7):1012-1016

144. Pidoux G, Tasken K (2010) Specificity and spatial dynamics of protein kinase A signaling organized by A-kinase-anchoring proteins. Journal of molecular endocrinology 44 (5):271-284. doi:10.1677/jme-10-0010

145. Pieske B, Maier LS, Bers DM, Hasenfuss G (1999) Ca2+ handling and sarcoplasmic reticulum Ca2+ content in isolated failing and nonfailing human myocardium. Circulation research 85 (1):38-46

146. Pinto JR, Siegfried JD, Parvatiyar MS, Li D, Norton N, Jones MA, Liang J, Potter JD, Hershberger RE (2011) Functional characterization of TNNC1 rare variants identified in dilated cardiomyopathy. The Journal of biological chemistry 286 (39):34404-34412. doi:10.1074/jbc.M111.267211

147. Plante E, Lachance D, Beaudoin J, Champetier S, Roussel E, Arsenault M, Couet J (2009) Comparative study of vasodilators in an animal model of chronic volume 146

overload caused by severe aortic regurgitation. Circulation Heart failure 2 (1):25- 32. doi:10.1161/circheartfailure.108.801548

148. Popovic ZB, Benejam C, Bian J, Mal N, Drinko J, Lee K, Forudi F, Reeg R, Greenberg NL, Thomas JD, Penn MS (2007) Speckle-tracking echocardiography correctly identifies segmental left ventricular dysfunction induced by scarring in a rat model of myocardial infarction. American journal of physiology Heart and circulatory physiology 292 (6):H2809-2816. doi:10.1152/ajpheart.01176.2006

149. Posch MG, Posch MJ, Geier C, Erdmann B, Mueller W, Richter A, Ruppert V, Pankuweit S, Maisch B, Perrot A, Buttgereit J, Dietz R, Haverkamp W, Ozcelik C (2008) A missense variant in desmoglein-2 predisposes to dilated cardiomyopathy. Molecular genetics and metabolism 95 (1-2):74-80. doi:10.1016/j.ymgme.2008.06.005

150. Preston LC, Lipscomb S, Robinson P, Mogensen J, McKenna WJ, Watkins H, Ashley CC, Redwood CS (2007) Functional effects of the DCM mutant Gly159Asp troponin C in skinned muscle fibres. Pflugers Archiv : European journal of physiology 453 (6):771-776. doi:10.1007/s00424-006-0161-7

151. Previs MJ, Beck Previs S, Gulick J, Robbins J, Warshaw DM (2012) Molecular mechanics of cardiac myosin-binding protein C in native thick filaments. Science (New York, NY) 337 (6099):1215-1218. doi:10.1126/science.1223602

152. Pu M, Gao Z, Pu DK, Davidson WR, Jr. (2013) Effects of early, late, and long-term nonselective beta-blockade on left ventricular remodeling, function, and survival in chronic organic mitral regurgitation. Circulation Heart failure 6 (4):756-762. doi:10.1161/circheartfailure.112.000196

153. Pu M, Gao Z, Zhang X, Liao D, Pu DK, Brennan T, Davidson WR, Jr. (2009) Impact of mitral regurgitation on left ventricular anatomic and molecular remodeling and systolic function: implication for outcome. American journal of physiology Heart and circulatory physiology 296 (6):H1727-1732. doi:10.1152/ajpheart.00882.2008

154. Rajan S, Ahmed RP, Jagatheesan G, Petrashevskaya N, Boivin GP, Urboniene D, Arteaga GM, Wolska BM, Solaro RJ, Liggett SB, Wieczorek DF (2007) Dilated cardiomyopathy mutant tropomyosin mice develop cardiac dysfunction with

147

significantly decreased fractional shortening and myofilament calcium sensitivity. Circulation research 101 (2):205-214. doi:10.1161/circresaha.107.148379

155. Redwood C, Robinson P (2013) Alpha-tropomyosin mutations in inherited cardiomyopathies. Journal of muscle research and cell motility 34 (3-4):285-294. doi:10.1007/s10974-013-9358-5

156. Reid AC, Silver RB, Levi R (2007) Renin: at the heart of the mast cell. Immunological reviews 217:123-140. doi:10.1111/j.1600-065X.2007.00514.x

157. Reiser PJ, Portman MA, Ning XH, Schomisch Moravec C (2001) Human cardiac myosin heavy chain isoforms in fetal and failing adult atria and ventricles. American journal of physiology Heart and circulatory physiology 280 (4):H1814- 1820

158. Remo BF, Giovannone S, Fishman GI (2012) Connexin43 cardiac gap junction remodeling: lessons from genetically engineered murine models. The Journal of membrane biology 245 (5-6):275-281. doi:10.1007/s00232-012-9448-0

159. Ruzicka M, Keeley FW, Leenen FH (1994) The renin-angiotensin system and volume overload-induced changes in cardiac collagen and elastin. Circulation 90 (4):1989-1996

160. Ryan TD, Rothstein EC, Aban I, Tallaj JA, Husain A, Lucchesi PA, Dell'Italia LJ (2007) Left ventricular eccentric remodeling and matrix loss are mediated by bradykinin and precede cardiomyocyte elongation in rats with volume overload. Journal of the American College of Cardiology 49 (7):811-821. doi:10.1016/j.jacc.2006.06.083

161. Sabbah HN, Sharov VG, Gupta RC, Mishra S, Rastogi S, Undrovinas AI, Chaudhry PA, Todor A, Mishima T, Tanhehco EJ, Suzuki G (2003) Reversal of chronic molecular and cellular abnormalities due to heart failure by passive mechanical ventricular containment. Circulation research 93 (11):1095-1101. doi:10.1161/01.res.0000101932.70443.fe

162. Sabri A, Rafiq K, Seqqat R, Kolpakov MA, Dillon R, Dell'italia LJ (2008) Sympathetic activation causes focal adhesion signaling alteration in early

148

compensated volume overload attributable to isolated mitral regurgitation in the dog. Circulation research 102 (9):1127-1136. doi:10.1161/circresaha.107.163642

163. Sadayappan S, de Tombe PP (2013) Cardiac myosin binding protein-C as a central target of cardiac sarcomere signaling: a special mini review series. Pflugers Archiv : European journal of physiology. doi:10.1007/s00424-013-1396-8

164. Sadayappan S, Gulick J, Osinska H, Barefield D, Cuello F, Avkiran M, Lasko VM, Lorenz JN, Maillet M, Martin JL, Brown JH, Bers DM, Molkentin JD, James J, Robbins J (2011) A critical function for Ser-282 in cardiac Myosin binding protein-C phosphorylation and cardiac function. Circulation research 109 (2):141- 150. doi:10.1161/circresaha.111.242560

165. Salameh A, Dhein S (2013) Effects of mechanical forces and stretch on intercellular gap junction coupling. Biochimica et biophysica acta 1828 (1):147-156. doi:10.1016/j.bbamem.2011.12.030

166. Saygili E, Rana OR, Meyer C, Gemein C, Andrzejewski MG, Ludwig A, Weber C, Schotten U, Kruttgen A, Weis J, Schwinger RH, Mischke K, Rassaf T, Kelm M, Schauerte P (2009) The angiotensin-calcineurin-NFAT pathway mediates stretch- induced up-regulation of matrix metalloproteinases-2/-9 in atrial myocytes. Basic research in cardiology 104 (4):435-448. doi:10.1007/s00395-008-0772-6

167. Scheuermann-Freestone M, Freestone NS, Langenickel T, Hohnel K, Dietz R, Willenbrock R (2001) A new model of congestive heart failure in the mouse due to chronic volume overload. European journal of heart failure 3 (5):535-543

168. Schmitt JP, Debold EP, Ahmad F, Armstrong A, Frederico A, Conner DA, Mende U, Lohse MJ, Warshaw D, Seidman CE, Seidman JG (2006) Cardiac myosin missense mutations cause dilated cardiomyopathy in mouse models and depress molecular motor function. Proceedings of the National Academy of Sciences of the United States of America 103 (39):14525-14530. doi:10.1073/pnas.0606383103

169. Schuller S, Van Israel N, Vanbelle S, Clercx C, McEntee K (2011) Lack of efficacy of low-dose spironolactone as adjunct treatment to conventional congestive heart failure treatment in dogs. Journal of veterinary pharmacology and therapeutics 34 (4):322-331. doi:10.1111/j.1365-2885.2010.01235.x

149

170. Sedej S, Schmidt A, Denegri M, Walther S, Matovina M, Arnstein G, Gutschi EM, Windhager I, Ljubojevic S, Negri S, Heinzel FR, Bisping E, Vos MA, Napolitano C, Priori SG, Kockskamper J, Pieske B (2013) Subclinical abnormalities in sarcoplasmic reticulum Ca2+ release promote eccentric myocardial remodeling and pump failure death in response to pressure overload. Journal of the American College of Cardiology. doi:10.1016/j.jacc.2013.11.010

171. Senechal M, MacHaalany J, Bertrand OF, O'Connor K, Parenteau J, Dubois- Senechal IN, Costerousse O, Dubois M, Voisine P (2013) Predictors of left ventricular remodeling after surgical repair or replacement for pure severe mitral regurgitation caused by leaflet prolapse. The American journal of cardiology 112 (4):567-573. doi:10.1016/j.amjcard.2013.04.024

172. Seqqat R, Guo X, Rafiq K, Kolpakov MA, Guo J, Koch WJ, Houser SR, Dell'italia LJ, Sabri A (2012) Beta1-adrenergic receptors promote focal adhesion signaling downregulation and myocyte apoptosis in acute volume overload. Journal of molecular and cellular cardiology 53 (2):240-249. doi:10.1016/j.yjmcc.2012.05.004

173. Severs NJ, Bruce AF, Dupont E, Rothery S (2008) Remodelling of gap junctions and connexin expression in diseased myocardium. Cardiovascular research 80 (1):9-19. doi:10.1093/cvr/cvn133

174. Shafii AE, Gillinov AM, Mihaljevic T, Stewart W, Batizy LH, Blackstone EH (2012) Changes in left ventricular morphology and function after mitral valve surgery. The American journal of cardiology 110 (3):403-408 e403. doi:10.1016/j.amjcard.2012.03.041

175. Shah SJ, Aistrup GL, Gupta DK, O'Toole MJ, Nahhas AF, Schuster D, Chirayil N, Bassi N, Ramakrishna S, Beussink L, Misener S, Kane B, Wang D, Randolph B, Ito A, Wu M, Akintilo L, Mongkolrattanothai T, Reddy M, Kumar M, Arora R, Ng J, Wasserstrom JA (2013) Ultrastructural and Cellular Basis for the Development of Abnormal Myocardial Mechanics During the Transition from Hypertension to Heart Failure. American journal of physiology Heart and circulatory physiology. doi:10.1152/ajpheart.00642.2013

176. Sheikh F, Ross RS, Chen J (2009) Cell-cell connection to cardiac disease. Trends in cardiovascular medicine 19 (6):182-190. doi:10.1016/j.tcm.2009.12.001

150

177. Smith DN, Bonagura JD, Culwell NM, Schober KE (2012) Left ventricular function quantified by myocardial strain imaging in small-breed dogs with chronic mitral regurgitation. Journal of veterinary cardiology : the official journal of the European Society of Veterinary Cardiology 14 (1):231-242. doi:10.1016/j.jvc.2011.11.002

178. Solaro RJ, Henze M, Kobayashi T (2013) Integration of troponin I phosphorylation with cardiac regulatory networks. Circulation research 112 (2):355-366. doi:10.1161/CIRCRESAHA.112.268672

179. Song W, Dyer E, Stuckey D, Leung MC, Memo M, Mansfield C, Ferenczi M, Liu K, Redwood C, Nowak K, Harding S, Clarke K, Wells D, Marston S (2010) Investigation of a transgenic mouse model of familial dilated cardiomyopathy. Journal of molecular and cellular cardiology 49 (3):380-389. doi:10.1016/j.yjmcc.2010.05.009

180. Sponga S, Ivanitskaia E, Potapov E, Krabatsch T, Hetzer R, Lehmkuhl H (2012) Preoperative treatment with levosimendan in candidates for mechanical circulatory support. ASAIO journal (American Society for Artificial Internal Organs : 1992) 58 (1):6-11. doi:10.1097/MAT.0b013e318239f401

181. Stewart JA, Jr., Wei CC, Brower GL, Rynders PE, Hankes GH, Dillon AR, Lucchesi PA, Janicki JS, Dell'Italia LJ (2003) Cardiac mast cell- and chymase-mediated matrix metalloproteinase activity and left ventricular remodeling in mitral regurgitation in the dog. Journal of molecular and cellular cardiology 35 (3):311- 319

182. Strauss CE, Duval S, Pastorius D, Harris KM (2012) Pharmacotherapy in the treatment of mitral regurgitation: a systematic review. The Journal of heart valve disease 21 (3):275-285

183. Sumandea CA, Garcia-Cazarin ML, Bozio CH, Sievert GA, Balke CW, Sumandea MP (2011) Cardiac troponin T, a sarcomeric AKAP, tethers protein kinase A at the myofilaments. The Journal of biological chemistry 286 (1):530-541. doi:10.1074/jbc.M110.148684

184. Suzuki R, Matsumoto H, Teshima T, Koyama H (2013) Clinical assessment of systolic myocardial deformations in dogs with chronic mitral valve insufficiency using two-dimensional speckle-tracking echocardiography. Journal of veterinary 151

cardiology : the official journal of the European Society of Veterinary Cardiology 15 (1):41-49. doi:10.1016/j.jvc.2012.09.001

185. Suzuki S, Fukushima R, Ishikawa T, Hamabe L, Aytemiz D, Huai-Che H, Nakao S, Machida N, Tanaka R (2011) The effect of pimobendan on left atrial pressure in dogs with mitral valve regurgitation. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 25 (6):1328-1333. doi:10.1111/j.1939-1676.2011.00800.x

186. Suzuki S, Ishikawa T, Hamabe L, Aytemiz D, Huai-Che H, Fukushima R, Machida N, Tanaka R (2011) The effect of furosemide on left atrial pressure in dogs with mitral valve regurgitation. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 25 (2):244-250. doi:10.1111/j.1939- 1676.2010.0672.x

187. Swynghedauw B (1986) Developmental and functional adaptation of contractile proteins in cardiac and skeletal muscles. Physiological reviews 66 (3):710-771

188. Takeda N, Hayashi Y, Arino T, Takeda A, Noma K (2001) Effect of pimobendan in patients with chronic heart failure. Experimental and clinical cardiology 6 (4):195-199

189. Tavora F, Zhang M, Cresswell N, Li L, Fowler D, Franco M, Burke A (2013) Quantitative Immunohistochemistry of Desmosomal Proteins (Plakoglobin, Desmoplakin and Plakophilin), Connexin-43, and N-cadherin in Arrhythmogenic Cardiomyopathy: An Autopsy Study. The open cardiovascular medicine journal 7:28-35. doi:10.2174/1874192401307010028

190. Taylor MR, Slavov D, Ku L, Di Lenarda A, Sinagra G, Carniel E, Haubold K, Boucek MM, Ferguson D, Graw SL, Zhu X, Cavanaugh J, Sucharov CC, Long CS, Bristow MR, Lavori P, Mestroni L, Familial Cardiomyopathy R, Bank BD (2007) Prevalence of desmin mutations in dilated cardiomyopathy. Circulation 115 (10):1244-1251. doi:10.1161/circulationaha.106.646778

191. Teerlink JR (2009) A novel approach to improve cardiac performance: cardiac myosin activators. Heart failure reviews 14 (4):289-298. doi:10.1007/s10741-009- 9135-0

152

192. van der Velden J (2011) Diastolic myofilament dysfunction in the failing human heart. Pflugers Archiv : European journal of physiology 462 (1):155-163. doi:10.1007/s00424-011-0960-3

193. van Rooij E, Quiat D, Johnson BA, Sutherland LB, Qi X, Richardson JA, Kelm RJ, Jr., Olson EN (2009) A family of microRNAs encoded by myosin genes governs myosin expression and muscle performance. Developmental cell 17 (5):662-673. doi:10.1016/j.devcel.2009.10.013

194. Varghese R, Itagaki S, Anyanwu AC, Milla F, Adams DH (2013) Predicting early left ventricular dysfunction after mitral valve reconstruction: The effect of atrial fibrillation and pulmonary hypertension. The Journal of thoracic and cardiovascular surgery. doi:10.1016/j.jtcvs.2013.08.073

195. Wang TL, Yang YH, Chang H, Hung CR (2004) Angiotensin II signals mechanical stretch-induced cardiac matrix metalloproteinase expression via JAK-STAT pathway. Journal of molecular and cellular cardiology 37 (3):785-794. doi:10.1016/j.yjmcc.2004.06.016

196. Wang Y, Tanner BC, Lombardo AT, Tremble SM, Maughan DW, Vanburen P, Lewinter MM, Robbins J, Palmer BM (2013) Cardiac myosin isoforms exhibit differential rates of MgADP release and MgATP binding detected by myocardial viscoelasticity. Journal of molecular and cellular cardiology 54:1-8. doi:10.1016/j.yjmcc.2012.10.010

197. Weber KT (1989) Cardiac interstitium in health and disease: the fibrillar collagen network. Journal of the American College of Cardiology 13 (7):1637-1652

198. Wei CC, Chen Y, Powell LC, Zheng J, Shi K, Bradley WE, Powell PC, Ahmad S, Ferrario CM, Dell'Italia LJ (2012) Cardiac kallikrein-kinin system is upregulated in chronic volume overload and mediates an inflammatory induced collagen loss. PloS one 7 (6):e40110. doi:10.1371/journal.pone.0040110

199. Wei CC, Hase N, Inoue Y, Bradley EW, Yahiro E, Li M, Naqvi N, Powell PC, Shi K, Takahashi Y, Saku K, Urata H, Dell'italia LJ, Husain A (2010) Mast cell chymase limits the cardiac efficacy of Ang I-converting enzyme inhibitor therapy in rodents. The Journal of clinical investigation 120 (4):1229-1239. doi:10.1172/jci39345

153

200. Wei CC, Lucchesi PA, Tallaj J, Bradley WE, Powell PC, Dell'Italia LJ (2003) Cardiac interstitial bradykinin and mast cells modulate pattern of LV remodeling in volume overload in rats. American journal of physiology Heart and circulatory physiology 285 (2):H784-792. doi:10.1152/ajpheart.00793.2001

201. Wei S, Guo A, Chen B, Kutschke W, Xie YP, Zimmerman K, Weiss RM, Anderson ME, Cheng H, Song LS (2010) T-tubule remodeling during transition from hypertrophy to heart failure. Circulation research 107 (4):520-531. doi:10.1161/circresaha.109.212324

202. Wolny M, Colegrave M, Colman L, White E, Knight PJ, Peckham M (2013) Cardiomyopathy mutations in the tail of beta-cardiac myosin modify the coiled- coil structure and affect integration into thick filaments in muscle sarcomeres in adult cardiomyocytes. The Journal of biological chemistry 288 (51):36260. doi:10.1074/jbc.A113.513291

203. Xia WJ, Huang YY, Chen YL, Chen SL, He JG (2011) Acute myocardial ischemia directly modulates the expression of brain natriuretic peptide at the transcriptional and translational levels via inflammatory cytokines. European journal of pharmacology 670 (1):7-12. doi:10.1016/j.ejphar.2011.09.012

204. Yoshikawa H, Suzuki M, Hashimoto G, Otsuka T, Sugi K (2011) Impact of cilostazol on left ventricular geometry and function: assessment by tissue Doppler imaging and two-dimensional speckle-tracking echocardiography. Echocardiography (Mount Kisco, NY) 28 (4):431-437. doi:10.1111/j.1540- 8175.2010.01357.x

205. Zheng J, Yancey DM, Ahmed MI, Wei CC, Powell PC, Shanmugam M, Gupta H, Lloyd SG, McGiffin DC, Schiros CG, Denney TS, Jr., Babu GJ, Dell'italia LJ (2013) Increased Sarcolipin Expression and Adrenergic Drive in Humans with Preserved Left Ventricular Ejection Fraction and Chronic Isolated Mitral Regurgitation. Circulation Heart failure. doi:10.1161/circheartfailure.113.000519

206. Zhou S, Liu Y, Prater K, Zheng Y, Cai L (2013) Roles of microRNAs in pressure overload- and ischemia-related myocardial remodeling. Life sciences 93 (23):855- 862. doi:10.1016/j.lfs.2013.08.023

207. Zito C, Carerj S, Todaro MC, Cusma-Piccione M, Caprino A, Di Bella G, Oreto L, Oreto G, Khandheria BK (2013) Myocardial deformation and rotational profiles 154

in mitral valve prolapse. The American journal of cardiology 112 (7):984-990. doi:10.1016/j.amjcard.2013.05.031

208. Zois NE, Olsen NT, Moesgaard SG, Rasmussen CE, Falk T, Haggstrom J, Pedersen HD, Moller JE, Olsen LH (2013) Left ventricular twist and circumferential strain in dogs with myxomatous mitral valve disease. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 27 (4):875-883. doi:10.1111/jvim.12114

209. Zois NE, Tidholm A, Nagga KM, Moesgaard SG, Rasmussen CE, Falk T, Haggstrom J, Pedersen HD, Ablad B, Nilsen HY, Olsen LH (2012) Radial and longitudinal strain and strain rate assessed by speckle-tracking echocardiography in dogs with myxomatous mitral valve disease. Journal of veterinary internal medicine / American College of Veterinary Internal Medicine 26 (6):1309-1319. doi:10.1111/j.1939-1676.2012.01017.x

155