To Phosphorylate Or Not to Phosphorylate: the Role of Tropomyosin Phosphorylation in Cardiac Function and Disease

TO PHOSPHORYLATE OR NOT TO PHOSPHORYLATE: THE ROLE OF TROPOMYOSIN PHOSPHORYLATION IN CARDIAC FUNCTION AND DISEASE

A dissertation submitted to the

Graduate School

of the University of Cincinnati

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

(Ph.D.)

In the Department of Molecular Genetics, Biochemistry and Microbiology

of the College of Medicine

2012

Emily M. Schulz

B.S. Ohio University, 2004

Committee Chair: David F. Wieczorek, Ph.D.

ABSTRACT

Tropomyosin (Tm) is an α-helical coiled-coil protein crucial in the calcium dependent regulation of the thin filament of the sarcomere. α-Tm is phosphorylated solely at serine 283 and phosphorylation levels are tightly regulated. Approximately 70% of total cardiac Tm phosphorylated during fetal development, decreasing to 30% during adulthood. Total Tm phosphorylation is altered in multiple mouse models of cardiac disease, including hypertrophic cardiomyopathy, dilated cardiomyopathy and myocardial infarction, indicating that Tm phosphorylation may play a role in the initiation, progression or modulation of cardiac disease.

To determine the effect of loss of α-Tm phosphorylation, a transgenic (TG) mouse model was generated in which cardiac specific α-Tm expresses an alanine at amino acid 283 rather than a serine (α-Tm

S283A). Counter to previous studies on Tm phosphorylation performed in vitro, at basal levels, significantly decreased α-Tm phosphorylation has no effect. TG animals exhibit normal cardiac function, efficient contractility and relaxation under basal conditions and under β-adrenergic stimulation.

However, when α-Tm S283A TG animals are subjected to transaortic constriction (TAC), the TG TAC operated animals fail more quickly than the non-transgenic (NTG) TAC operated littermates.

Interestingly, in TG hearts, there is an increase in SERCA2a expression and an increase in PLN phosphorylation at Ser16. This increase in energetic demand placed on the heart both by the increase in

SERCA2a activity coupled with the increased energetic demand that occurs during the onset of cardiac disease may be the basis of the more rapid cardiac failure in the TG TAC mice.

After examining the effect of decreased α-Tm phosphorylation in the context of an acute, extrinsic cardiac stressor, a model of decreased α-Tm phosphorylation was made in the context of a chronic, intrinsic stressor. These α-Tm 180-S283A double mutant TG (DMTG) animals were surprising in that the α-Tm

180 familial hypertrophic cardiomyopathy phenotype was completely rescued in a TG line that exhibited

Tm phosphorylation levels similar those seen in the α-Tm S283A TG hearts. This rescue occurred at

iii morphological and physiological levels, including echocardiographic analysis. In addition, examination

2+ of detergent extracted skinned fiber bundles indicate that Ca sensitivity and cooperativity (nH) is rescued to NTG levels, indicating that the rescue occurs primarily at the level of the sarcomere.

The DMTG mutant hearts show a significant decrease in PLN phosphorylation at both Ser16 and Thr17 compared to NTG and α-Tm S283A TG hearts. However, DMTG levels of PLN phosphorylation at both sites are significantly increased compared to α-Tm 180 levels, which may possibly account for the hypercontractile phenotype seen in the DMTG mice when evaluated via echocardiography.

These data seem to indicate that dephosphorylating Tm in the context of hypertrophy might be of benefit to the overall function of the heart. However, it is important to examine the effect of increased Tm phosphorylation in the context of other cardiac disease, such as dilated cardiomyopathy or myocardial infarction. The studies presented here, in addition to the proposed studies, would possibly increase understanding of the functional consequences of Tm phosphorylation which may lead to possible therapeutic interventions.

DEDICATION

This work is dedicated to Brant Schulz. Husband, cheerleader, voice of reason, support system, chef and cat-wrangler. I couldn’t have done this without you.

NANOS GIGANTUM HUMERIS INSIDENTES My deep thanks and gratitude go to Dr. David F. Wieczorek. The continuous support and encouragement he offered has been crucial to any small contributions I have made to our field. I especially thank him for teaching me to think independently, creatively and critically. I also extend my thanks to the members of my committee. Dr. Andrew Herr, Dr. James Lessard, Dr. Jeffery Molkentin and Dr. Gary Shull have supported me and significantly contributed to my understanding of how to do great science.

I thank members of my lab: Dr. Ganapathy Jagatheesan and Dr. Sudarsan Rajan for helping me to understand what type of scientist I wanted to become. Hannah Yaejee Hong generated the construct for the α-Tm 180-S283A studied in this dissertation. Shelby Moore generated the companion construct, α- Tm 180-S283D.

My deep thanks go to Dr. Vikram Prasad and Dr. Tracy Pritchard for their technical support, vast store of knowledge they never hesitated to share and their interest in the research of a graduate student from a neighboring lab.

Maureen Bender has been crucial to the work I have done in this lab.

I am lucky to have many good friends who have supported me throughout this process. We are spread out through the country and the world but their support and love has been very important to me. Although equally all are dear to me, Dr. Susan Vidovichenko, Dr. Emily Bradford, Dr. Naomi Oshiro, Dr. Palanikumar Manoharan, Kanimozhi Vairamani, Robyn Pilcher-Roberts, Cat Tucker and Lana Goodrich deserve special mention here. Thank you.

This work would not have been possible without the financial support of TG T36-HL07382 to Dr. A. Schwartz.

Finally, I must express my deepest thanks and gratitude to my family. My parents, Russell and Robin Thompson, my sisters, my extended family and my family by marriage have done so much to support me in my endeavors. Stress relief and puffy support was provided by Nimbus, Olliver, Gilbert and Jayne. Without the love and support of my husband, Brant Schulz, this work would not have been possible.

vii

TABLE OF CONTENTS

ABSTRACT ...... iii DEDICATION...... v NANOS GIGANTUM HUMERIS INSIDENTES ...... vii TABLE OF CONTENTS ...... viii LIST OF FIGURES ...... xi LIST OF TABLES ...... xiii LIST OF ABBREVIATIONS ...... xiv

CHAPTER 1: INTRODUCTION ...... 1 Tropomyosin ...... 1 Tropomyosin Genes ...... 1 Tropomyosin Protein ...... 3 Striated Muscle Tropomyosins ...... 6 Tropomyosin in Muscle Contraction ...... 9 Ca2+ Flux Proteins ...... 12 Exercise Training and Physiological Hypertrophy ...... 16 Phosphorylation of Myofibrillar Proteins ...... 17 Phosphorylation of Tropomyosin ...... 21 Cardiac Hypertrophy and Tropomyosin ...... 22 The Goal of this Dissertation ...... 27 CHAPTER 2: MATERIALS AND METHODS ...... 29 Generation of α-Tm S283A and α-Tm 180-S283A TG Mice ...... 29 Genotyping of TG Mice ...... 30 Northern Blot Analyses ...... 30 Real Time RT-PCR Analysis ...... 32 Copy Number by Real Time RT-PCR ...... 33 Southern Blot Analysis ...... 33 Generation of an α-Tm S283 Phospho-Specific Antibody ...... 35 Heart Homogenate, Myofibrillar Preparation and Cytoplasmic Preparation ...... 35 Quantitation of TG Protein and Investigation of Myofibrillar Proteins ...... 36

viii

Matrix Assisted Laser Desorption/Ionization – Time of Flight (MALDI-TOF) ...... 37 Two Dimensional Isoelectric Focusing ...... 37 Western Blot Analysis of Cardiac and Myofibrillar Proteins ...... 38 Investigation of the Effect of Sacrifice on Tm Phosphorylation ...... 39 Gravimetrics ...... 40 Histology and Pathology ...... 40 Cardiomyocyte Cross-Sectional Areas ...... 41 Echocardiographic Analysis ...... 41 Transaortic Constriction and Echocardiography ...... 42 Measurements of Ca2+ Dependent Activation of Tension...... 42 Isolated Work-Performing Heart Model ...... 43 Blood Pressure Studies ...... 43 Calcineurin/ Protein Phosphatase Activity Assay ...... 44 Statistics ...... 44 Generation of Recombinant α-Tm, β-Tm and γ-Tm NTG and Mutant Proteins ...... 44 Actin Binding Assay ...... 47 Circular Dichroism Measurements ...... 48 CHAPTER 3: TROPOMYOSIN DEPHOSPHORYLATION RESULTS IN COMPENSATED HYPERTROPHY ...... 49 Summary ...... 49 Introduction ...... 50 Results ...... 52 Discussion ...... 66 CHAPTER 4: INVESTIGATION OF Tm PHOSPHORYLATION AND α-Tm S283A MICE...... 71 Striated Muscle Tm Ser283A Specific Antibody ...... 71 Phosphorylation of Human Cardiac Tm ...... 72 Tm Phosphorylation is Sensitive to Method of Sacrifice...... 74 Early Studies on the α-Tm S283A Mice ...... 75 Additional Studies of Line 2 α-Tm S283A TG Animals ...... 78 The Mystery of α-Tm S283A Line 25 ...... 82 Conclusions ...... 87 CHAPTER 5: TROPOMYOSIN DEPHOSPHORYLATION RESCUES TROPOMYOSIN INDUCED FAMILIAL HYPERTROPHIC CARDIOMYOPATHY ...... 89 Summary ...... 89

Introduction ...... 91 Results ...... 93 Discussion ...... 104 CHAPTER 6: α-Tm 180-S283A LINE 335 TG ANIMAL STUDIES AND α-Tm 180 CROSSED WITH α-Tm S283A/D TG ANIMAL STUDIES ...... 108 α-Tm 180-S283A Line 335 Double Mutant Transgenic Mouse Studies ...... 108 Conclusions from Studies Performed on α-Tm 180-S283A DMTG Line 335 Hearts ...... 116 α-Tm 180 TG Animals Bred With Tm Phosphorylation Mutants ...... 118 Conclusions on α-Tm 180 TG Animals Bred with Tm Phosphorylation Mutants ...... 122 CHAPTER 7: INVESTIGATING POSSIBLE BIOCHEMICAL CONTRIBUTORS TO DIFFERENTIAL TM ISOFORM FUNCTION ...... 123 Rationale ...... 123 Generation and Purification of Recombinant Proteins ...... 126 Actin Binding Assays ...... 127 Circular Dichroism ...... 129 Conclusions ...... 131 CHAPTER 8: CONCLUSIONS AND FUTURE DIRECTIONS ...... 133 REFERENCES ...... 143

LIST OF FIGURES

Figure 1. Tropomyosin (Tm) isoform diversity...... 2 Figure 2. The heptad repeat structure of each α-helical Tm protein...... 3 Figure 3. Importance of alanine clusters...... 5 Figure 4. Periods one and five of striated muscle α, β and γ-Tm...... 6 Figure 6. Figure showing the Tm coiled-coil in each of the three states of contraction...... 11 Figure 7. Activation pathways of Ca2+ flux proteins involved in the regulation of cardiac contraction. ... 13 Figure 8. Sarcomeric and myofilament protein schematics...... 18 Figure 9. Examples of hypertrophic growth caused by Tm mutations...... 23 Figure 10. Plasmids involved in generating Tm recombinant proteins...... 46 Figure 11. α-Tm S283A TG animal characterization ...... 53 Figure 12. Phosphorylation status of NTG and α-Tm S283A TG animals...... 55 Figure 13. Histology and Gravimetrics...... 57 Figure 14. Ca2+ sensitivity and work performing heart parameters...... 59 Figure 16. Echocardiographic analyses, gravimetrics and histology of NTG sham, TG sham, NTG transaortic constriction operated (TAC) and TG TAC hearts from 12-16 week old mice...... 65 Figure 17. Validating the Tm phosphorylation specific antibody ...... 72 Figure 18. Phosphorylation of Tm in human heart samples...... 73 Figure 20. Northern blot analyses of 8 TG lines generated from the cardiac specific α-Tm S283A construct...... 75 Figure 21. Northern blot analyses of α-Tm S283A Line 2, Line 25 and Line 97 ...... 76 Figure 22. Southern blot analysis of α-Tm S283A TG lines...... 77 Figure 23. Gravimetric analysis of α-Tm S283A Line 2 mice ...... 79 Figure 24. Phosphate release in NTG and α-Tm S283A TG heart homogenate by both PP1 and PP2a. ... 80 Figure 25. Systolic blood pressure in 12-16 week old male NTG and TG animals ...... 80 Figure 26. Masson’s Trichrome staining of whole heart sections from control and TAC operated hearts...... 81 Figure 27. Kaplan Meier Survival Curve for NTG, typical Line 25 and extreme Line 25 TG animals. ... 82 Figure 28. Whole hearts and sections from Line 25 typical and extreme and control mice...... 84 Figure 29. Gravimetrics of NTG, Line 25 typical and Line 25 extreme mice...... 85 Figure 30. Real Time RT-PCR analysis of typical and extreme Line 25...... 86 Figure 31. Characterizing the α-Tm 180-S283A DMTG animals ...... 93 Figure 32. Phosphorylation status of Tm in NTG, α-Tm 180, α-TM S283A and α-Tm 180-S283A DMTG hearts...... 95 Figure 33. Histology and gravimetrics...... 97 Figure 34. Parameters involved in Ca2+ tension relations...... 99 Figure 35. Real Time RT-PCR analysis of cardiomyopathy markers...... 101 Figure 36. Alterations in Ca2+ flux proteins ...... 103 Figure 37. Characterization of α-Tm 180-S283A DMTG Line 325 and Line 335...... 109 Figure 38. Histology and gravimetrics...... 111 Figure 39. Real Time RT-PCR analysis of cardiomyopathy markers...... 113 Figure 40. Alterations in Ca2+ flux proteins...... 115 Figure 41. Copy numbers of 180 mice bred with phosphorylation mutants...... 120

Figure 42. Western blots of 180 mice bred with phosphorylation mutants...... 121 Figure 43. Heptad motif of WT Tm isoforms...... 124 Figure 44. Standard curve of 0.0 µM α-Tm to 20.0 µM α-Tm...... 127 Figure 45. Images of supernatant and pellet fractions from actin binding assay...... 128 Figure 47. Tm recombinant proteins...... 131 Figure 48. Model of the Ser283 residues at the C-terminus of tropomyosin...... 139

xii

LIST OF TABLES

Table 1. Primers utilized to generate TG constructs ...... 29 Table 2. Genotyping primers ...... 30 Table 3. Primers used to generate probe fragments for Northern blot analysis ...... 31 Table 4. Primers used in Real Time RT-PCR ...... 32 Table 5. Primers generated for copy number by Real Time RT-PCR ...... 33 Table 6. Primers utilized for Southern blot analysis ...... 35 Table 7. Conditions used in Western blot analyses...... 39 Table 8. Primers used to clone striated muscle Tm isoforms into pGEM-T easy and pET-Sumo ...... 45 Table 9. Primers used to generate α-Tm, β-Tm and γ-Tm mutants ...... 47 Table 10. Parameters involved in Ca2+-tension relations in skinned fiber bundles...... 58 Table 11. Cardiac function of NTG and α-TM S283A TG mice at 3 months of age as assessed by M- Mode echocardiography ...... 60 Table 12. Cardiac function of NTG, α-Tm 180 and α-Tm E180G-S283A DMTG at 3 months of age as assessed by echocardiography ...... 98 Table 13. Parameters involved in Ca2+ tension relations ...... 100

xiii

LIST OF ABBREVIATIONS

3’UTR- 3 prime untranslated region 5’UTR- 5 prime untranslated region α-MHC: α-myosin heavy chain α-Tm 175: Tropomyosin mutation at codon 175 (Asp-Asn) and the corresponding TG mice α-Tm 180: Tropomyosin mutation at codon 180 (Glu-Gly) and the correspondent TG mice α-Tm 180-S283A: Tropomyosin mutation at codon 180 (Asp-Asn) and at codon 283 (Ser-Ala) and the corresponding TG mice α-Tm E54K: Tropomyosin mutation at codon 54 (Asp-Lys) and the corresponding TG mice β-MHC: β-myosin heavy chain %FS: Percent Fractional shortening AAV: Adeno-associated virus Amp: Ampicillin ANP: Atrial natriuretic peptide BNP: Brain natriuretic peptide Bp: Base pair BP: Blood pressure BSA: Bovine serum albumin CaM kinase: Ca2+/calmodulin-dependent protein kinase CD: Cervical dislocation OR Circular Dichroism CH1: Tm specific 1’ antibody CIP: Calf intestinal phosphatase CnA: Calcineurin CO2:Carbon dioxide c-Raf1:proto-oncogene serine/threonine- protein kinase DCM: Dilated cardiomyopathy DMTG: Double mutant transgenic DTG: Double transgenic DT: Deceleration time DTT: Dithiothreitol E/A ratio: ratio of early-to-late ventricular filling velocities E/Em ratio: (ratio of early pulse-Doppler filling velocity to early tissue Doppler velocity) EF: Ejection fraction EC coupling: Excitation contraction-coupling Epoch: Epoch Spectrometer ERK1/2: Extracellular signal-regulated kinases 1 and 2 F1: FHC: Familial hypertrophic cardiomyopathy FS: Fractional shortening Glu180Gly: H&E: Hemotoxylin and eosin stain HCM: Hypertrophic cardiomyopathy HF: Heart failure HGH: Human growth hormone HMW: High molecular weight HRP: Horse radish peroxidase HW:BW: Heart weight to body weight IPG: Immobilized pH gradient IVRT: Iso-volumetric relaxation time IVS: Intraventricular septum

xiv

Kapp: Apparent binding constant LMW: Low molecular weight LTCC: L-type Ca2+ channel, Dihydropyridine Receptor LVCW:BW: Left ventricular chamber weight to body weight ratio LV: Left ventricular LVIDd,s: Left ventricular internal dimension in diastole, systole LVOT: Left ventricular outflow tract LVOT CSA (cm2)= (D/2)2: LVOT cross-sectional area MALDI-TOF: Matrix assisted laser desorption/ionization time of flight MEK1: Dual specificity mitogen-activated protein kinase 1 MCIP1: Myocyte enriched calcineurin-interacting protein 1 MI: Myocardial infarction MLC: Myosin light chain MLC-2: Myosin light chain 2 MyBPC: Myosin binding protein C nH: Hill Coefficient NO: Nitric oxide NTG: Non-transgenic OD: Optical density p38 MAPK: p38 mitogen-activated protein kinase pBS +: pBlueScript II SK (+) pCa50: Calcium sensitivity Pfu: High fidelity polymerase PKA: Protein kinase A PKC: Protein kinase C PKCδ: Protein Kinase C delta PKCε: Protein Kinase C epsilon PKCζ: protein kinase C zeta PKG: Protein kinase G PLN: Phospholamban PLN Ser16: Phosphorylated phospholamban serine 16 PLN Thr 17: Phosphorylated phospholamban threonine 17 PLNKO: PLN knockout PP1: Protein phosphatase 1 PP2a: Protein phosphatase 2a PP2b: Protein phosphatase 2b pTM: Phosphorylated Tm RLC: Regulatory light chain RT-PCR: Reverse transcriptase polymerase chain reaction RV: Right Ventricle RWT: Relative wall thickness RyR: Ryanodine receptor S16E: PLN mutation Serine16 to Aspartic Acid S283A:Tm mutation Serine283 to Alanine SDM: Site directed mutagenesis SDS-PAGE: Sodium dodecyl sulfate polyacrylamide gel electrophoresis SERCA2a: Sarcoplasmic ATPase isoform 2a SERCA2a-/-: Knockout (or) null SERCA2a-/+: Heterozygous SLN: Sarcolipin SR: Sarcoplasmic reticulum

TAC: Transaortic constriction TBP: Tributylphosphate TCEP: tris(2-carboxyethyl)phosphine TG: Transgenic TG Tm: Transgenic Tm Tm: Tropomyosin TnC: Troponin C TnI: Troponin I TnT: Troponin T TPP: Time to peak pressure TAC: Transaortic constriction Vcf: Velocity of circumferential shortening VO2 max: Maximal oxygen consumption VTI: Velocity-time integral VW:BW: Ventricle weight to body weight ratio WGA: Wheat germ agglutinin WT: Wild type

xvi

CHAPTER 1: INTRODUCTION

Tropomyosin

Tropomyosin (Tm) is a canonical alpha helical coiled-coil dimer that is highly conserved from yeast to humans. In striated muscle, Tm is a sarcomeric thin filament protein crucial for the control of Ca2+ regulated contraction. This archetypal coiled-coil protein serves to modulate the actin myosin interaction in order to regulate thin filament function. Mutations in sarcomeric Tm have been identified in cardiac diseases such as hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM).

Tropomyosin Genes

Four Tm genes have been characterized in vertebrates to date and members of this family are widely distributed. In vertebrates, the Tm family is separated into two groups, a high molecular weight (HMW) group containing 284 amino acid (aa) residues and low molecular weight (LMW) group containing 248 aa. The genes have been named TPM1, TPM2, TPM3 and TPM4, though the genes are also referred to as

α-Tm, β-Tm, γ-Tm and δ-Tm respectively [1]. The Tm genes and gene products will be referred to utilizing the later naming convention.

Each of the four genes generates multiple isoforms with varying tissue and developmental specificity.

These isoforms are generated by mutually exclusive alternative splicing of the multiple exons encoded by each gene as well as use of alternative promoters and differential 3’ end processing [2-5]. The genes are highly homologous, although exons involved in alternative splicing are more widely divergent. Exon 1a in α- and β-Tm are highly homologous while exon 1a and exon 1b in α-Tm exhibit a higher level of dissimilarity. Some of the Tm isoforms encode part of the structure of the cytoskeleton microfilament and are ubiquitously expressed. Non-muscle isoforms play specific roles in cytokinesis and neurite outgrowth, among many other functions (Figure 1).

Figure 1. Tropomyosin (Tm) isoform diversity. Four genes (α, β, γ, δ) generate multiple isoforms through the use of alternative promoters, alternative splicing and differential 3’ end processing [3].

Tropomyosin Protein

Tm, as stated previously, is an archetypal alpha helical coiled-coil dimeric protein that lies in the groove of the actin helix. Tm polymerizes head-to-tail with a short overlap of 9-11 amino acids [6, 7]. The coiled-coil motif is a consequence of the chemical nature of the 40 individual 7 amino acid (a-b-c-d-e-f-g) repeat termed the heptad repeat (Figure 2). The 40 heptad repeats are unbroken, making Tm unique among fibrous proteins as most coiled-coils have breaks in the coil. Amino acids at the a and d position are hydrophobic in nature and protrude from the helix into holes between residues on the other Tm helix, facilitating coiling. Amino acids at the e and g positions often have opposite charges and interact electrostatically, stabilizing the coiled-coil [8-10].

Figure 2. The heptad repeat structure of each α-helical Tm protein. The hydrophobic interactions between the d and a position amino acids facilitate coiling while the interactions between the e and g position amino acids stabilize the coiled coil [11].

Tm has an unusually high content of alanine residues at the d position of the heptad repeat when compared to similar coiled coil proteins [12]. Clusters of alanines occur at the d position separated by segments that are leucine rich in the d position of the heptad. It is suggested that the alanine clusters allow for a narrowing of the coiled-coil given the small size of the residue, introducing areas of local flexibility in the molecule and facilitating the winding of Tm around the actin helix (Figure 3). The alanine clusters are not evenly spaced throughout the molecule, however, as there are seven in α-Tm, the occurrence of the clusters can be related to the putative actin binding domains of Tm, commonly referred to as periods [13, 14].

Tm has been roughly divided into seven periods, each containing 35-42 amino acids and these periods correspond to quasi-equivalent actin binding regions. These periods have been shown to play different functions within the Tm molecule. Periods one and five are specialized for maximal actin binding affinity as they include alanine clusters in the d position which allow the Tm dimer to assume the optimal helical confirmation necessary for strong Tm-actin binding. These periods also include consensus residues that appear to be necessary for Tm-actin interaction [13, 15, 16].

Figure 3. Importance of alanine clusters. A. Sequence of chicken striated muscle α-Tm. Alanine residues at the d position are highlighted by blue triangles. Negatively charged amino acids implicated in actin binding are highlighted in black boxes. Amino acids common in coiled-coils are highlighted with red diamonds. B. Simplified secondary structure of Tm showing the flexibility conferred by the alanine clusters. C. A simplified actin/Tm structure. Figure adapted from [12].

Striated Muscle Tropomyosins

There are three striated muscle tropomyosins, α-Tm, β-Tm and γ-Tm. Both α- and β-Tm are expressed in the heart while all three isoforms are expressed in skeletal muscle. Interestingly, the three striated muscle isoforms have 7 alanine clusters with different numbers of alanines present in the d position of the heptad repeat in periods one and five (Figure 4). Brown et al [12] showed that the alanine clusters are important for actin binding. This difference in the number of alanines in the clusters between the α, β and γ-Tm isoforms likely contribute to differential isoform function in the murine heart, which will be discussed shortly.

Figure 4. Periods one and five of striated muscle α, β and γ-Tm. Differences in d position alanine clusters highlighted in red [Ahmed and Wieczorek, unpublished].

Striated muscle α-, β- and γ-Tm are highly homologous, sharing greater than 85% amino acid identity

(Figure 5). α-Tm is predominantly expressed in fast-twitch skeletal and cardiac muscle. β-Tm is expressed at increased levels in skeletal muscle with exact amounts depending on which muscle is being examined while γ-Tm makes up approximately 30% of total Tm in slow-twitch muscle [17, 18].

Recently, a novel isoform of α-Tm containing exon 2a, normally restricted to smooth muscle Tm, was discovered in human cardiac tissue [19]. This isoform has been termed TPM1κ as it is an isoform generated from the TPM1 gene, also known as κ-Tm and seems to be differentially regulated in cardiomyopathy patients [20]. While κ-Tm is not normally expressed in the murine heart, biochemical studies indicate that κ-Tm binds actin less effectively than the major striated muscle α-Tm isoform.

Additionally, κ-Tm overexpression in transgenic mice studied by the work-performing heart model indicated that transgenic (TG) hearts exhibited a moderate but consistent decrease in the rates of contraction and relaxation with an increase in time to peak pressure (TPP). Isolated, detergent extracted

TG fibers exhibit decreased myofilament Ca2+ sensitivity compared to non-transgenic (NTG) animals that express α-Tm in the heart, indicating that even among the same gene family, Tm proteins behave very differently [20].

Figure 5. Aligned amino acid sequences of striated muscle α, β and γ-Tm.

During cardiac development, the ratio of α-Tm mRNA to β-Tm mRNA expression is 5:1, decreasing to a ratio of 60:1 after birth [21], suggesting not only tissue specific roles for Tm isoforms, but also temporally significant roles. β-Tm protein is expressed at an approximately 20% of total Tm during fetal development and at approximately 1-2% of total Tm expressed in the adult heart [22, 23]. Investigation into the in vivo function of β-Tm protein expression was performed by over-expressing the β-Tm cDNA under the cardiac specific α-myosin heavy chain (α-MHC) promoter. Interestingly, there is a feedback mechanism regulating the precise amounts of sarcomeric proteins, including Tm, maintaining total expression level at 100%, meaning that overexpressing any TG Tm protein decreases the amount of endogenous Tm protein expressed [24]. Curiously, this feedback mechanism does not exist for cytoskeletal Tm [25]. If β-Tm over-expression results in approximately 50% of the total Tm protein, the level of the endogenous α-Tm protein is reduced to approximately 50%, in keeping with the feedback mechanism responsible for maintaining proper amounts of total Tm protein in the myofilament. These β-

Tm TG hearts exhibit impairments in diastolic parameters, specifically decreases in the maximum rate of relaxation as well as time to half-relaxation. These animals also exhibit increased Ca2+ sensitivity of skinned fiber bundles [22, 26, 27]. Interestingly, when β-Tm is expressed at higher levels in the murine heart, the animals die 10-14 days post-natally and exhibit thrombus formation in both atria and the lumen of the left ventricle as well as excessive deposition of fibrotic tissue and myocyte disarray [28]. These data indicate essential differences between the functions of α-Tm and β-Tm (which share 86% amino acid identity) as well as emphasizing the importance of maintaining an appropriate ratio of α-Tm to β-Tm in the murine heart.

In order to determine the in vivo role of γ-Tm, γ-Tm was overexpressed in the heart under the α-MHC promoter. Although γ-Tm is not expressed in the murine heart, there have been reports of γ-Tm expression in the human heart [29]. α and γ-Tm share approximately 93% amino acid identity. TG mouse lines were generated, expressing 40-60% γ-Tm protein with a concomitant decrease in α-Tm expression. Interestingly, all γ-Tm animals have a normal life span, are hypercontractile and demonstrate

8 a decrease in Ca2+ sensitivity, opposite of what is seen in the β-Tm over-expression mice [30], further highlighting the essential differences in isoform function.

Expression of α-, β- and γ-Tm at approximately equal levels in a mouse heart sheds light on whether one isoform is dominant in the conferral of physical and biochemical properties. These βγ-DTG hearts show no pathological alterations in cardiac morphology or function. Interestingly, the βγ-Tm DTG hearts seem to owe most of the functional alterations in contractile parameters to the expression of the γ-Tm isoform.

Work performing heart studies indicate that the βγ-Tm DTG hearts exhibit increases in the rate of contraction and relaxation, very similar to γ-Tm TG hearts and opposite of the β-Tm TG hearts.

Additionally, NTG, γ-Tm and βγ-Tm DTG hearts respond appropriately to β-adrenergic agonists while the β-Tm TG hearts exhibit a blunted response that is not enhanced at high doses of isoproterenol.

Finally, the βγ-Tm DTG myofibers exhibit decreased Ca2+ sensitivity, as does the γ-Tm TG myofiber indicating that the presence of γ-Tm seems to dominate the functional phenotype when all isoforms are equally expressed. These results seem to indicate that all isoforms contribute to muscle function and holds interesting implications for skeletal muscle, which has higher levels of β- and γ-Tm [31].

Tropomyosin in Muscle Contraction

Tropomyosin plays a major role in contraction through the regulation of the thin filament of the sarcomere via modulation of actin-myosin interactions. Several models have been proposed elucidating the position of Tm on the actin filament and the effect the position of Tm has on myosin binding and muscle contraction. The generally accepted model suggests that Tm has occupational states on the thin filament. The tethered/blocked position is described as the position that Tm occupies when no Ca2+ is bound to troponin C (TnC) (Figure 6). When no Ca2+ is bound to TnC, TnC binds TnI and TnT loosely while TnI tightly binds actin. Tm lies along the actin filament and sterically blocks the interaction between actin and myosin. At this position, the muscle is relaxed. The closed/cocked/Ca2+ induced position of Tm occurs after binding of Ca2+ to TnC, strengthening the interaction between TnI-TnC, which allows the initiation of a conformational change in the Tn complex, releasing TnI from actin. It is

9 suggested, at this point, that crossbridges are weakly associated but incapable of forming strong, force- generating crossbridges, serving to hold the thick and thin filaments in parallel. Interactions between

TnT-Tm, TnI-actin and TnI-Tm weaken upon binding of Ca2+ to TnC, allowing Tm to be pushed from the position it occupies on actin during the blocked position toward the position it will occupy in the open position. The final position of the Tm is referred to as the open position. After intracellular Ca2+ increases, the weakly associated cross-bridges are strengthened and Tm is shifted into the open position, well away from the myosin binding sites on actin, allowing myosin full access to actin and activating the myofilament to initiate contraction [11, 32-34].

Figure 6. Figure showing the Tm coiled-coil in each of the three states of contraction. A. Blocked state. B. Closed/Cocked state. C. Open state [11].

Ca2+ Flux Proteins

During excitation contraction-coupling (EC coupling), extracellular Ca2+ entry through the L-type Ca2+ channel (LTCC) prompts release of sarcoplasmic reticulum (SR) Ca2+ through the ryanodine receptor

(RyR). This SR release of Ca2+ raises cytosolic Ca2+ concentrations and initiates muscle contraction. As the sarcolemma repolarizes, Ca2+ is resequestered into the SR through the activity of the sarcoplasmic reticulum Ca2+ - ATPase isoform 2a (SERCA2a) and sarcolemmal Ca2+ transporters, resulting in muscle relaxation. SERCA2a activity controls both the rate of Ca2+ removal from the cytosol and the amount of

SR Ca2+ load, largely determining the rate of relaxation and the pump is regulated by phospholamban

(PLN) and sarcolipin (SLN), a PLN homologue (Figure 7).

As SERCA2a is a major regulator of intracellular Ca2+ homeostasis, SERCA2a has been extensively studied in experimental animal models as well as tissues isolated from failing human hearts. SERCA2a expression and/or activity are found to be decreased in failing human hearts [35-38]. Homozygous knockout of SERCA2a (SERCA2a-/-) in the mouse heart is an embryonic lethal phenotype while heterozygous mice (SERCA2a+/-) expressing approximately 60% of WT SERCA2a levels exhibit no overt cardiac pathology and respond appropriately to β-adrenergic stimulation [39-41]. However, when

SERCA2a +/- animals are stressed by pressure overload, the heterozygotes develop heart failure (HF) more quickly than WT littermates [42]. Interestingly, there was no compensatory upregulation of any of the other isoforms of SERCA, including SERCA1a or SERCA2b, indicating that SERCA2a is critical for maintaining proper Ca2+ fluxes.

Figure 7. Activation pathways of Ca2+ flux proteins involved in the regulation of cardiac contraction. The complex interactions and interplays between crucial cardiac Ca2+ regulatory proteins [43].

Over-expression of SERCA2a in the murine heart results in increased Ca2+ transport from the cytosol into the SR, increases in relaxation and contraction and does not cause cardiac pathology [44-46]. SERCA2a over-expression or gene transfer in animal models of cardiac disease has been shown to be beneficial, often partially restoring or normalizing cardiac function and physiology. For example, conditional overexpression of SERCA2a can restore normal cardiac function, Ca2+ handling and contractility in animals that have pressure overload-induced cardiac hypertrophy [47]. α-Tm 180 (Glu180Gly) TG animals exhibit a severe form of familial hypertrophic cardiomyopathy (FHC), characterized by defects in relaxation, a grossly enlarged left atria and death occurring between 6-8 months of age [48-50].

Administration of a single dose of SERCA2a via adenovirus to 1 day old α-Tm 180 mice delayed the onset of the FHC phenotype and improved cardiac function up to 3-4 months of age [51]. Phase I human clinical trials in humans were conducted in HF patients utilizing gene transfer of SERCA2a cDNA via an adeno-associated virus (AAV). At 6 months, the study demonstrated sufficient safety and improvements in symptoms, biomarkers and LV function and remodeling. In double blind phase II trials, HF patients show improvement in maximal oxygen consumption (VO2 max), LV end-systolic volumes with no increases in adverse events, disease related events or arrhythmias [52].

In contrast to these data, when SERCA2a is over-expressed prior to pressure overload induced hypertrophy, the TG animals develop HF more quickly than their NTG littermates likely due to the increased energetic demand placed on the diseased heart by increased SERCA2a activity [53]. SERCA2a is the second most energetically demanding pump in the heart, after the myosin ATPase, making

SERCA2a vulnerable to changes in energetics and ATP supply. In the failing heart, energetic demand typically increases due to poor pump function [54].

The major regulator of SERCA2a activity is PLN. Alterations in the PLN to SERCA2a ratio have been observed to be altered in cardiac disease. PLN is a transmembrane phosphoprotein that binds to and inhibits SERCA2a depending up its phosphorylation state. When PLN is dephosphorylated, it binds to

SERCA2a and inhibits the activity of the pump [55]. However, upon phosphorylation by protein kinase

A (PKA) or Ca2+/calmodulin-dependent protein kinase (CaM kinase), the inhibitory activity of PLN is blocked and the activity of SERCA2a can increase four fold or more [56-58].

In the cardiac SR, PLN exists in both a pentameric and monomeric form. When cytoplasmic Ca2+ concentrations are low, PLN interacts strongly with SERCA2a, inhibiting SERCA2a activity. When PLN is phosphorylated, PLN dissociates from SERCA2a due to the aggregation of PLN monomers into pentamers, indicating that monomeric PLN is the species that actively interacts with and inhibits

SERCA2a [59-61].

Interestingly, knocking out PLN in mice (PLNKO) results in significantly increased rates of contraction and relaxation in both isolated heart preparations and intact mice [62, 63], indicating that knocking out

PLN has a similar effect to increasing SERCA2a. PLNKO hearts are only minimally stimulated by β- agonists as PLNKO results in nearly maximum rates of contraction and relaxation. Additionally, models of cardiac disease can also be rescued by PLNKO, including the FHC model α-Tm 180 [64-66]. When

PLNKO mice were crossed with α-Tm 180 mice, permanent increases in SR Ca2+ uptake occur, improving contraction and relaxation and preventing the development of cardiac hypertrophy.

Phospholamban is phosphorylated at Serine 16 and Threonine 17 and that phosphorylation has been shown to be altered in disease conditions [67]. Transgenic studies have shown that phosphorylating

Ser16 is sufficient to mediate the maximal mechanical response to β-adrenergic stimulation [68].

Additionally, adenoviral delivery of a constitutively phosphorylated PLN (S16E) inhibited the progression of heart failure upon myocardial infarction [69]. Thr17 phosphorylation also relieves PLN inhibition of SERCA2a and increases SERCA2a activity, resulting in a similar phenotype to SERCA2a overexpression [43, 57].

Exercise Training and Physiological Hypertrophy

SERCA2a and PLN play an important role during exercise training and physiological cardiac hypertrophy. Physiological cardiac hypertrophy is heart growth in response to chronic exercise training and is reversible and characterized by normal or enhanced cardiac function. Interestingly, both types of hypertrophic growth that occur in pathological hypertrophy also occur in physiological hypertrophy.

Eccentric hypertrophy often occurs in response to aerobic exercise or pregnancy and results in volume overload and is characterized by chamber enlargement and a proportional change in wall thickness.

Eccentric growth in response to disease, however, is associated with a thinning of the ventricular walls.

Strength training results in a pressure load on the heart, rather than the volume load that drives eccentric growth, and results in concentric hypertrophy [70, 71].

There are biochemical and molecular alterations that occur in exercise trained hearts. Contrary to what is seen in pathological hypertrophy, there are increases in SERCA2a expression and activity in exercised trained animals. This increase occurs both in the healthy exercise trained heart and results in increased rates of Ca2+ cycling and faster relaxation rates, likely due to the decrease in the PLN to SERCA2a ratio

[72, 73]. Additionally, studies have indicated exercise training results in increased CAM kinase dependent phosphorylation of PLN Thr17, which, in addition to the decrease in PLN to SERCA2a ratio, increases the activity of the pump and Ca2+ cycling even further [74].

Even after the onset of cardiac disease, such as pacing-induced heart failure or myocardial infarction, exercise training results in an increase in SERCA2a (compared to disease state) and a partial rescue of the disease phenotype [75, 76]. The positive effects of exercise training on animal models of disease have led to human studies on HF patients. Individuals who underwent exercise training exhibit increases in VO2 max, similar to human patients in the SERCA2a clinical study, as well as increased exercise duration, independent of age or sex [77] although other studies indicated variability in the response of diseased patients to exercise [78].

Phosphorylation of Myofibrillar Proteins

Isoform switching is a method used to alter the function of the contractile apparatus. However, most isoform changes occur over days to weeks. Another, more acute method of altering the contractile properties and Ca2+ sensitivity of the sarcomere involves altering the phosphorylation status of proteins in the sarcomere. Phosphorylation of sarcomeric proteins allows alterations in the force of contraction independent of sarcoplasmic Ca2+ fluxes (Figure 8 for a diagram of the myofilament and sarcomere).

Titin, the giant filamentous sarcomeric backbone, mediates myocyte compliance. Titin can be phosphorylated by PKA in response to β-adrenergic stimulation by catecholamines and protein kinase G

(PKG) activated by nitric oxide (NO) or natriuretic peptides [79-81]. Both PKA and PKG phosphorylate the same site, serine 469, found in the elastic segment of the protein [82]. Interestingly, both PKA and

PKG phosphorylation at the same site reduces passive stiffness although PKA reduces passive stiffness to a greater degree. This is likely due to the fact that PKG phosphorylates additional sites, possibly affecting protein-protein interactions rather than effecting titin-based stiffness.

Pathologically increased passive myocardial stiffness often occurs during HF, likely due to deranged PKA signally that typically occurs during disease. This loss of PKA mediated titin phosphorylation is likely responsible for the increases in passive stiffness and, in fact, PKA administration to myocytes isolated from patients with diastolic heart failure decreased passive stiffness from significantly elevated levels indicated that PKA-based phosphorylation of titin maybe be of clinical importance [83, 84].

Figure 8. Sarcomeric and myofilament protein schematics. A. Schematic of thick and thin filament protein members. Adapted from [85]. B. Relationship of titin to the thick and thin filament [84].

Myosin binding protein C (MyBPC) is a thick filament associated protein important in the regulation of contraction. The N-terminus of MyBPC interacts at the C-terminus with the light meromyosin section of the myosin rod and with titin. MyBPC also interacts with the subfragment-2 portion of myosin, reducing actomyosin ATPase activity. Interestingly, phosphorylation of MyBPC by PKA prevents the interaction between MyBPC and subfragment-2, releasing the hold on cross-bridge cycling. Mutations in MyBPC are responsible for greater than 40% of all FHC cases [86]. In human cardiac disease, there are decreases in MyBPC phosphorylation [87]. When a transgenic animal expressing MyBPC with the three well knows MyBPC phosphorylation sites (Ser273, 283 and 302) mutated to alanines is generated on a

MyBPC null background, cardiomyocytes exhibit reduced stretch activation and do not respond appropriately to dobutmine treatment [88, 89]. However, generation of a MyBPC mutant with the three phosphorylation sites mutated to aspartic acids, meant to mimic constitutive phosphorylation, increases dobutamine induced cardiac contractility and also seems be protective against ischemia-reperfusion injury, with a decreased ischemic area and increased fractional shortening compared to WT littermates

[90], indicating that phosphorylation of MyBPC is crucial for proper contraction [91].

Myosin light chain 2 (MLC-2), also known as the myosin regulatory light chain (RLC), has been shown to be crucial in the assembly of myofibers in zebrafish [92]. Interestingly, when a non-phosphorylatable

MLC-2 TG mouse is generated (TG-RLC(P-)) with serine 15 mutated to an alanine in the myocardium,

TG animals display a lack of a functional response to dobutamine and skinned fiber bundles show a decrease in the maximum tension generated. MLC-2 is phosphorylated by myosin light chain kinase, which increases force and accelerates the rate of force development while decreasing the rate of relaxation [93-96]. Interestingly, the alterations in contraction seem to be mechanical as phosphorylation of MLC-2 results in movement of cross-bridges away from the thick filament [97].

Troponin T is considered the ‘lever’ of the thin filament responsible for transmitting the conformational changes in TnC-TnI structure upon Ca2+ binding or release to Tm. PKC has been shown to phosphorylate four main sites, Thr197, Ser201, Thr206 and Thr287 in the mouse TnT protein [98, 99]. These sites are

19 located in the C-terminal half of the molecule that interacts with both TnI, TnC and possibly Tm as well.

Additionally, this Tm binding region is likely essential in transmitting the Ca2+ induced conformational changes to Tm and actin [100]. Research involving phosphorylation of both TnI and TnC showed a 30% decrease in maximum force generated by cardiac mouse myofilaments that is rescued when partially replaced with a non-phosphorylatable fast skeletal muscle TnT isoform, indicating that TnT phosphorylation might be pivotal in the PKC dependent decrease in myofilament tension [99]. Thorough characterization of a number of transgenic animals expressing TnT phosphorylation mimetics singly or in concert as well as transgenic animals expressing non-phosphorylatable TnT mutations singly or in concert emphasized the importance of the PKC-α site Thr206. Phosphorylation or expression of a phosphomimetic (Thr206Glu) results in significant reduction of tension and actomyosin Mg-ATPase activity, Ca2+ sensitivity and cooperative activation of the thin filaments. Interestingly, none of the other sites (Thr197, Ser201 and Thr287) had any effect on the studied parameters [101, 102].

Troponin I (TnI) phosphorylation is considered to be a major or master regulator of sarcomeric function.

TnI is phosphorylated at Ser23/24 by PKA and phosphorylation of these sites results in decreased myofilament Ca2+ sensitivity associated with increased heart rate [103]. Multiple studies provide compelling evidence that phosphorylation of Ser23/24 plays a crucial role in the maintenance of the force-frequency response in contracting ventricles [104, 105]. Interestingly, phosphorylation of TnI by

PKC seems to have an opposing effect in comparision to TnI phosphorylation by PKA. Phosphorylation of TnI at Ser42/44 induces a depression in maximum tension and phosphorylation of Thr143 increases sensitivity to Ca2+ [106-108]. Of particular interest is the fact that alterations or mutations in other sarcomeric proteins, such as MyBPC, result in altered TnI phosphorylation, further pointing toward the importance of TnI phosphorylation in the proper maintenance of contraction.

Phosphorylation of Tropomyosin

Striated muscle Tm is phosphorylated at a sole reported site, the penultimate amino acid, serine 283 [109-

111]. During fetal development, approximately 70% of α-Tm is phosphorylated, decreasing to approximately 30% after birth [110], indicating that Tm phosphorylation may play a crucial role in filament formation during myofibrillogenesis. A model of DCM generated by this lab shows significant decreases in Tm phosphorylation while studies of FHC models show significant differences in Tm phosphorylation depending on the age of the animal studied [48, 49, 112-115]. The significant alterations in Tm phosphorylation in disease models suggest a possible role of Tm phosphorylation in the development or modulation of hypertrophy. Although studies in the past have implicated a currently unknown ‘tropomyosin kinase’ in the phosphorylation of α-Tm, as yet, the kinase involved has not been conclusively identified [116, 117].

In vitro studies on isolated phosphorylated and nonphosphorylated α-Tm have indicated that phosphorylation may play a role in enhancing nearest neighbor interactions in the sarcomere.

Phosphorylated α-Tm increases the strength of Tm head-to-tail interactions, thought to be important in the cooperative activation of the thin filament, and also increases the activity of the actin-activated myosin S1

ATPase compared to α-Tm that was not phosphorylated. While it has been shown that binding to actin is unchanged regarding altered Tm phosphorylation status, when troponin is added to phosphorylated α-Tm, viscosity is reduced compared to non-phosphorylated protein, indicating a significant weakening of the

Tm-Tn interaction. Interestingly, studies showed that there was no alteration in Tm-Tn binding strength, indicating that the reduced viscosity in the previous experiment may indicate that altering the phosphorylation status of α-Tm may, in some way, affect the ability of troponin T (TnT) to bridge the Tm overlap [118-120].

As stated previously, in a model of DCM generated by this lab, a significant decrease is seen in α-Tm phosphorylation, indicating that alteration of phosphorylation status of Tm in vivo may be play a role during the development of cardiac disease [113]. A transgenic mouse was generated expressing a

21 phosphorylation mimetic rather than a serine at aa 283. Preliminary data indicates that α-Tm S283D TG animals with high expression of the transgene die early from DCM. However, these animals do not exhibit any change in cooperativity as measured by the Hill Coefficient (nH), contrary to expectations from previously published in vitro work. Animals expressing more moderate levels of the transgene do not seem to exhibit gross alterations in cardiac function with the exception of a decrease in the relaxation parameter measured using work performing heart preparations and development of a mild to moderate

DCM phenotype beginning at 6 months of age [Jagatheesan and Wieczorek, unpublished]. More work is necessary to elucidate the role of α-Tm phosphorylation during both basal cardiac function and during the development of cardiac disease.

Cardiac Hypertrophy and Tropomyosin

Hypertrophic cardiomyopathy (HCM) is a common genetic cardiac disease occurring in approximately

1:500 individuals and is the leading cause of sudden cardiac death among young athletes. There are many different HCM-causing mutations and there are a wide variety of clinical manifestations, leading to many prognoses and treatments. HCM is characterized by a hypertrophied but non-dilated left ventricle with myocyte disarray [121-123] (Figure 9). It was noted that some cases of HCM seemed to be familial in origin, passed on in a Mendelian autosomal dominant manner and this subset of the disease was termed familial hypertrophic cardiomyopathy (FHC).

FHC has been classified as a disease of the sarcomere due to discovery of a large number of mutations in sarcomeric proteins such as α-Tm, MYBPC, α-MHC and titin. In the United States, FHC caused by mutations in α-Tm are relatively rare, occurring in approximately 5% or fewer cases. Patients with mutations in α-Tm typically experience mild symptoms and late onset. Prevalence of FHC is similar in

Japan, although symptoms tend to be more severe and often result in sudden death [124, 125].

Interestingly, in Finland, mutations in α-Tm are responsible for approximately 30% of all FHC patients, likely due to a founder effect [126].

Figure 9. Examples of hypertrophic growth caused by Tm mutations. A. Whole hearts from a mouse model of hypertrophy (α-Tm 180), a normal animal (WT) and a model of dilated cardiomyopathy (α-Tm E54K). B. Cross sections of the same from A [31].

Eleven mutations in α-Tm have been identified that are responsible for causing FHC. Six of these mutations occur in the TnT binding region, possibly altering the interaction between Tm and TnT. Three mutations can be found at the N-terminal end of the molecule, one mutation is found in the center of the molecule and one mutation at the C-terminal end [123, 127]. Five of the mutations result in charge changes. Given the heptad repeat of the coiled-coil, alterations in charge changes at any of the positions

(a, b, c, d, e, f, g) can perturb coiling, coiled-coil interactions and interactions with sarcomeric binding partners. Evidence for this has been seen in several studies investigating the biochemical effects of the

Asp175Asn mutation and the Glu180Gly mutations in α-Tm. Both mutations cause familial hypertrophic cardiomyopathy and both mutations result in partial unwinding of the coiled coil and alterations in surface charge in the region of the Tm molecule implicated in TnT binding [128-130]. Interestingly, the study by

Golitsina [129] and the study by Bing [130] disagree on the effect of these mutations on actin binding, with the Golitsina study indicating that α-Tm 175 bound much more weakly to actin than either WT α-Tm or α-Tm 180. In contrast, the Bing study indicated that WT α-Tm and α-Tm 175 bind actin with equal affinity and that α-Tm 180 binds actin with less affinity. Additionally, it was shown that, in activating conditions (pCa5), the velocity of the actin filaments containing the mutant α-Tms increases significantly, indicating that the increased velocity is likely due to an altered interaction between Tm and TnT.

Recent biochemical studies investigating the flexibility of the D175N mutation and the E180G mutation of the overall structure of the Tm molecule have provided interesting insights in the possible mechanism of sarcomeric dysfunction caused by these FHC mutations. Both the D175N mutation and the E180G mutation confer significantly increased flexibility to the α-Tm molecule compared to WT protein.

Interestingly, in the presence of actin, the flexibility of both mutants is blunted but still significantly more flexible than WT [131-133]. It is likely that this increase in flexibility results in less required Ca2+ to induce conformational changes through the Tn complex, allowing Tm to shift off the myosin head binding sites on actin and initiate contraction, leading to the increased Ca2+ sensitivity found in mouse models expressing these mutations, discussed below.

This laboratory was the first to develop in vivo models of FHC. The first model generated was α-Tm in which aa 175, an aspartate, was mutated to an asparagine (α-Tm 175). Physiological studies assessing functional changes in cardiac performance showed that there are significant functional differences between mice expressing greater than 40% TG protein. TG hearts exhibit decreases in maximum rate of contraction, increased time to peak pressure (TPP), decreases in the maximum rate of relaxation; in effect, these hearts are hypodynamic when compared to NTG littermates. Histopathology revealed myocyte hypertrophy, disarray and fibrosis. Interestingly, these mice have a normal life expectancy, similar to human patients with the same mutation [48, 49, 134].

Another FHC model was generated by this lab. These animals expressed α-Tm in which aa 180, a glutamic acid, was mutated to a glycine (α-Tm 180). At TG protein expression levels above 35%, TG skinned fibers are more sensitive to Ca2+, depressed relaxation at both 2.5 and 4.5 months of age as well as morphological anomalies. The left atria of the α-Tm 180 animals is greatly enlarged and exhibits calcification and thrombus formation. Fibrotic tissue is excessively deposited in both the left ventricular

(LV) free wall and the left atria. These animals also die between 6-8 months of age, due to the severe phenotype [48-50]. Interestingly, if the α-Tm 180 animals are crossed with a TG mouse expressing a chimeric α/β protein, the disease phenotype of the α-Tm 180 mice are rescued. The chimeric protein encodes α-Tm from aa 1-257 and β-Tm from 258-284 and includes the β-Tm 3’ UTR, altering one of the

TnT binding regions. This chimeric protein causes no pathology, but decreases Ca2+ sensitivity and depresses the rates of contraction and relaxation. The α-Tm 180 crossed with α-Tm expressing the chimeric protein double transgenic (DTG) hearts exhibit morphology similar to NTG littermates, have rates of contraction and relaxation similar to NTG littermates, and respond appropriately to β-adrenergic stimulation, indicating that Ca2+ may be a crucial modulator of FHC disease progression [135, 136].

Interestingly, a TG mouse model expressing α-Tm 180 on a different background (C57 Blk6, rather than

FVB/N) exhibited no abnormal morphology or hypertrophy. However, these animals did exhibit diastolic dysfunction that was rescued after administration of β-blockers [137]. These seemingly conflicting

25 results emphasize the importance of possible modifier genes on FHC severity and provide an explanation as to why individuals with the same mutation exhibit symptoms of variable severity.

Similar to FHC, DCM is associated with mutations in sarcomeric proteins including α-Tm, MyBPC and actin. DCM is a disease that is often associated with congestive heart failure and is characterized by depressed systolic function, cardiomegaly and ventricular dilation (Figure 9). There are three reported mutations in α-Tm that are known to cause DCM: Glu40Lys, Glu54Lys and Glu180Val [138]. This laboratory generated an in vivo mouse model in which α-Tm has been mutated to express a lysine at aa 54 rather than a glutamic acid (α-Tm E54K) [112, 113]. TG animals exhibit increases in heart weight to body weight ratios in conjunction with echocardiographic evidence indicating an increase in LV diastolic dimensions. Both contraction and relaxation are impaired in these mice, as well as a decrease in Ca2+ sensitivity of skinned fiber bundles. This is the first TG animal generated expressing a known human mutation that causes DCM.

The study of TG animals expressing human mutations that lead to FHC and DCM enhance the understanding of the role of Tm in sarcomeric function and the development of cardiac disease. It is to be hoped that greater study of these animals may elucidate the mechanism behind the development of pathological hypertrophy, allowing the development of more effective prevention and treatments of cardiac disease.

The Goal of this Dissertation

The main objective of this dissertation is to explore the importance of striated mucle α-Tm phosphorylation in cardiac function in an in vivo setting. Extensive research has been done on the significance of phosphorylation of other myofibrillar proteins both in vitro and in vivo. Although Tm phosphorylation was discovered several decades ago, investigation of the functional role of Tm phosphorylation has been limited to in vitro studies. Those in vitro studies suggested that altering Tm phosphorylation might result in altered nearest neighbor interactions, leading to possible contractile changes when studied in vivo. Additionally, the tight regulation the phosphorylation status of Tm both during fetal development and through the life time of animals seems to indicate that Tm phosphorylation has specific roles in cardiac contraction. α-Tm phosphorylation also changes in disease models generated by this laboratory, suggesting that α-Tm phosphorylation status may play some role in either the development or modulation of cardiac disease.

In order to address this issue, a construct was generated substituting serine 283 with an alanine, rendering the TG α-Tm unphosphorylatable. This construct was expressed under the cardiac specific α-MHC promoter. This α-Tm S283A TG animal was investigated under basal conditions as well as under β- adrenergic stimulation and during acute disease caused by an extrinsic stressor (transaortic constriction –

TAC). The data indicate that male α-Tm S283A animals exhibit a very mild compensated hypertrophy with no attendant cardiac dysfunction. Interestingly, these animals have a slight but significant increase in SERCA2a expression and PLN Ser16 phosphorylation although these increases do not confer increased contractility. When these animals are stressed with TAC, the TG α-Tm S283A animals exhibit impaired cardiac function compared to NTG littermates, likely due to the increased energetic demand placed on the heart by increased SERCA2a activity.

In order to study the role of α-Tm dephosphorylation in the context of a cardiac disease caused by an intrinsic stressor, a double mutant transgenic (DMTG) construct was generated expressing both the FHC mutation α-Tm 180 (Glu180Gly) and the dephosphorylating S283A mutation (α-Tm 180-S283A) and

27 expressed under the α-MHC promoter. Interestingly, these DMTG animals show a nearly complete rescue of the cardiac FHC phenotype with Ca2+ sensitivity returned to NTG levels in comparison to the greatly increased Ca2+ sensitivity found in the α-Tm 180 mice. Echocardiographic analyses indicates that these hearts are actually hypercontractile compared to NTG littermates. Additionally, there is no significant increase in SERCA2a expression, though there are alterations in PLN phosphorylation, indicating that dephosphorylation of α-Tm may initiate a signaling cascade that is cardioprotective in the case of an intrinsic or sarcomeric stressor.

A secondary objective of this dissertation is to document the attempt to study the biochemical contributors to differential isoform function. Although striated muscle α, β and γ-Tm share more than

85% amino acid identity, they confer very different function to the heart when overexpressed. Multiple mutant proteins were generated, altering both β and γ-Tm to contain similar numbers of alanine clusters in order to determine if the alanine clusters were responsible for some of the biochemical differences seen.

A full discussion of work done and the development of experimental protocols will be included.

CHAPTER 2: MATERIALS AND METHODS

Generation of α-Tm S283A and α-Tm 180-S283A TG Mice

Mouse α-Tm cDNA in the pBlueScript II SK (+) (pBS +) plasmid was subject to site directed mutagenesis (SDM) (Agilent Technologies) utilizing α-Tm S283A primers listed in Table 1.

Table 1. Primers utilized to generate TG constructs Primer Name Primer Sequence α-Tm S283A Forward 5’-CAC GCT CTC AAC GAT ATG ACT GCC ATA TAA GTT TCT TTG CTT CAC-3’ α-Tm S283A Reverse 5’-GTG AAG CAA AGA AAC TTA TAT GGC AGT CAT ATC GTT GAG AGC GTG-3’ E180G SDM Forward 5’- ACG TGC AGA GGG GCG GGC TGA-3’ E180G SDM Reverse 5’-TCA GCC CGC CCC TCT GCA CGT-3’

Clones expressing the mutations listed in Table 1 were grown up on LB agar plates containing the selective antibiotic 50mg/ml Ampicillin (Amp) (Sigma Aldrich). Clones were sequenced to determine whether the mutation was present (GeneWiz) and positive clones were digested with SalI and HindIII and ligated into the linear Robbins Vector (pBS+ including the α-MHC promoter and the human growth hormone (HGH) 3’UTR and poly A tail) that had also been digested with SalI and HindIII. This subcloning from pBS+ into the Robbins Vector was necessary as the high fidelity polymerase (Pfu) utilized in the SDM process becomes error prone after 5000 base pairs (bp). The TG constructs were then transformed into DH5α E. coli competent cells and plated on LB agar plates containing Amp. Clones were chosen and sequenced for the presence of the mutation. Positive clones were digested with NotI to release the α-MHC promoter, α-Tm S283A cDNA and 3’UTR and the HGH poly-A tail. To generate TG mice, this purified fragment was then provided to the University of Cincinnati Transgenic Core for microinjection into fertilized FVB/N mouse blastocyts.

The α-Tm 180-S283A TG animals were generated in the same way as the α-Tm S283A TG animals with one change. α-Tm S283A cDNA in the pBS+ plasmid was subjected to site directed mutagenesis

(Agilent Technologies) to introduce the E180G mutation utilizing primers listed in Table 1. Founders for each transgenic mouse colony were identified utilizing PCR performed on tail clip DNA and were bred to

FVB/N animals to generate offspring.

Genotyping of TG Mice

Tail clips from 5-7 day old mice were digested overnight in 1X tail lysis buffer (10X tail lysis buffer -

0.5M Tris pH8.5, 1.0M KCl, 2mL HCL in 100 mL), 0.45% Tween 20, 0.45% Igepal-CA603/NP-40 and

0.5ug/ul proteinase K at 55°C. To inactivate the proteinase K, the tails are boiled for 10 minutes.

Digested DNA is then subject to PCR analysis. Genotyping primer sequences are found in Table 2. After

PCR, the amplicons are resolved on a 0.8% agarose gel. The α-MHC Forward and α-Tm reverse primers produce a PCR product of ~200 bp and the GAPDH control primers produce a PCR product of ~500bp.

Table 2. Genotyping primers Primer Name Primer Sequence α-MHC Forward 5’- GCC CAC ACC ACA AAT GAC AGA-3’ α-Tm Reverse 5’- TCC AGT TCA TCT TCA GTG CCC -3’ GAPDH Control Forward 5’- AGC GAG CTC AGG ACA TTC TGG -3’ GAPDH Control Reverse 5’ – CTC CTA ACC ACG CTC CTA GCA -3’

Northern Blot Analyses

Animals were sacrificed using cervical dislocation. The hearts were rapidly dissected out and snap frozen. The hearts were then added to 1 mL of Trizol Reagent (Life Technologies) and were homogenized using a Polytron PT 10-35 homogenizer. Samples were prepared as per Trizol Reagent instruction in a cold room. Samples were quantitated on an Epoch Spectrometer (Epoch). 10 µg of RNA was electrophoresed on a 1% formaldehyde agarose gel with ethidium bromide at 90 V in 1X MOPS running buffer (Sigma Aldrich) for 3-4 hours. After running the gel, the 18S and 8S ribosomal RNA bands were visualized using a UV light gel box.

A passive blot was set up overnight utilizing capillary action to transfer the RNA to a Hybond XL nitrocellulose membrane optimized for nucleotide transfer (GE Healthcare). The RNA was UV crosslinked to the membrane using a UV Stratalinker (Stratagene). Membranes were briefly rinsed in DEPC-

H2O and were then placed in Super Hyb (Molecular Research Center, Inc) prewarmed to 42°C and placed in a 42°C incubator for 4 hours. Probes were made to the α-Tm 5’UTR, α-MHC promoter or GAPDH utilizing primers listed in Table 3 and were gel purified prior to use. The specific probe being utilized would be radio-labeled with 32P labeled dCTP using the Amersham Rediprime II Random Prime

Labelling System (GE Healthcare). Radiolabeled probe was purified using the illustra ProbeQuant G-50

Micro Columns (GE Healthcare) to remove unincorporated nucleotides and was then added to the membrane with new, pre-warmed Super Hyb buffer and incubated overnight at 42°C. The membrane was then washed twice in 2X SSC with 0.1% SDS and washed twice in 0.2X SSC with 0.1% SDS for 30 minutes each wash at 65°C. The blot was then placed on a phosphor screen overnight and was visualized using a StormScanner 840 phospho-imager (Amersham) and quantitated using ImageQuant 5.1

(Molecular Probes). α-MHC endogenous (5.5 kb) and TG (~2.2 kb) transcripts of two distinc sizes, α-Tm

5’ UTR (endogenous RNA) levels were both normalized to GAPDH expression.

Table 3. Primers used to generate probe fragments for Northern blot analysis Primer Name Primer Sequence GAPDH Forward 5’- TGA CCA CAG TCC ATG CCA TG -3’ GAPDH Reverse 5’- GAC GGA CAC ATT GGG GGT AG -3’ MHC Exon 1 Forward 5’- TCA GAG ATT TCT CAA ACC CAG -3’ MHC Exon 1 Reverse 5’- GGT TAA CTT TTC AGA GAA TCC TG -3’ αTM 5’UTR Forward 5’- AAG TAT TGG CTG TCC TAA GGA ATC-3’ αTM 5’UTR Reverse 5’- GCG TCC ATG GTG GCG GTG GC -3’

Real Time RT-PCR Analysis

RNA was isolated from 3 month old NTG and TG mouse ventricular tissue was using TRIZOL Reagent

(Life Technologies). 5µg of each RNA sample was used in reactions performed to generate cDNA in triplicate following the SuperScript III RT (Life Technologies) protocol utilizing random hexamers. Two of the cDNA reactions for each RNA sample was pooled to generate a standard and the remaining cDNA reaction was reserved for testing. Real time RT-PCR samples were prepared in 20 µl reactions using the

Absolute Blue QPCR Master Mix (Thermo Fisher) and were performed using an Opticon 2 real time RT-

PCR machine (MJ Research). Each sample was run in triplicate and each experiment was repeated twice.

Target mRNA was normalized to GAPDH expression as described by Pfaffl’s method [139]. Primers can be found listed in Table 4.

Table 4. Primers used in Real Time RT-PCR Primer Name Primer Sequence ANP Forward 5’-GCTTCCAGGCCATATTGGAG-3’ ANP Reverse 5’-GGGGGCATGACCTCATCTT-3’ BMHC Forward 5’-TTCATCCGAATCCATTTTGGGG-3’ BMHC Reverse 5’-GCATAATCGTAGGGGTTGTTGG-3’ BNP Forward 5’-GAGGTCACTCCTATCCTCTGG-3’ BNP Reverse 5’-GCCATTTCCTCCGACTTTTCTC-3’ GAPDH Forward 5’-TGACCACAGTCCATGCCATC-3’ GAPDH Reverse 5’-GACGGACACATTGGGGGTAG-3’ L-TYPE (CaCNa1c) Forward 5’-CCTGCTGGTGGTTAGCGTG-3’ L-TYPE (CaCNa1c) Reverse 5’-TCTGCCTCCGTCTGTTTAGAA-3’ MICP1 (RCAN1) Forward 5’- TTGTGTGGCAAACGATGATGT- 3’ MICP1 (RCAN1) Reverse 5’- CCCAGGAACTCGGTCTTGT-3’ NCX Forward 5’-CACTATTTTTGCTTTCATCAAC-3’ NCX Reverse 5’-ATCTGATTGTTCCTTTAGAAGC-3’ PLN Forward 5’-AAAGTGCAATACCTCACTCGC-3’ PLN Reverse 5’-GGCATTTCAATAGTGGAGGCTC-3’ RyR2 Forward 5’ATGGCTTTAAGGCACAGCG-3’ RyR2 Reverse 5’-CAGAGCCCGAATCATCCAGC-3’ Serca2a Forward Proprietary V. Prasad, PhD. Shull lab. Serca2a Reverse Proprietary V. Prasad, PhD. Shull lab.

Copy Number by Real Time RT-PCR

Development of a fast, high through-put method of determining copy number by utilizing Real Time RT-

PCR was developed. Originally, the Solaris qPCR (Thermo Scientific) system was utilized as the primers were purported to correspond to ‘all’ α-Tm isoforms. Upon closer examination, the reverse primer was specific to smooth muscle α-Tm only. Instead, BioRad Sybr Green was utilized with specially made primers to determine copy number (Table 6). All TG lines studied were normalized to NTG copy number

(2).

Table 5. Primers generated for copy number by Real Time RT-PCR Primer Name Sequence α-Tm 3’ UTR RTPCR Forward 5’- TTT GCT TCA CTC CTC CCA AGA CTC -3’ α-Tm 3’ UTR RTPCR Reverse 5’- AAA GAA ACC TGG GTC AGC TGG AGA -3’

Southern Blot Analysis

Animals were sacrificed using cervical dislocation. Hearts, tails or spleens were removed and placed in a digestion buffer (100mM Tris pH 8.0, 10mM EDTA, 100mM NaCl, 0.5% SDS, 10% v/v 10mg/mL

Proteinase K) overnight at 55°C. The digest solution was added to 5 Prime Phase Lock Gel Heavy tubes

(Fisher Scientific), with 50% v/v phenol and 50% v/v chloroform and mixed to homogenous suspension.

The tubes were centrifuged between 12,000-16,000 x g for 5 minutes. The phase lock gel forms a barrier, keeping the organic phase below the gel and the aqueous (DNA containing) phase above the gel. Another

50% v/v chloroform is added to additionally purify the DNA and the tubes are spun again at 12,000-

16,000 x g for 5 minutes. The aqueous phase is then decanted into a separate 1.5mL eppendorf tube.

10% v/v of 3.3M sodium acetate is added to the aqueous phase, in addition to 50% v/v ice cold ethanol.

The tubes are gently inverted until the DNA becomes visible as a globular structure. The DNA is then carefully spooled onto a pipette tip and is transferred to a tube containing 75% ethanol. The samples are incubated at -80°C for 30 minutes in order to slow the Brownian motion of the molecules. The samples are then spun at 16,000 x g for 5 minutes. The supernatant is decanted and the samples are allowed to dry for 10 minutes at room temperature. The pellets are then resuspended in TE buffer and quantitated. 20

µg of genomic DNA is digested with EcoR I overnight at 37°C in a total volume of 100 µL. The samples are subjected to speed vacuum to decrease the overall volume to 20 µL, allowing easy loading of the samples onto a 1.0% SDS page gel and run slowly, at 35V overnight. Interestingly, genomic DNA digest with EcoR I, after visualization on a UV gel box, show a smear of digested DNA with a single discrete band, likely representing the digest of highly repetitive ribosomal DNA. The 1kb size ladder run along with the DNA is marked with a needle full of India ink for easy visualization different sized DNA bands after transfer to nitrocellulose.

The gel containing the genomic DNA must be prepared in order for the DNA to easily transfer to the nitrocellulose membrane. First, the gel must be washed in depurination buffer (1 L H2O, 21mL HCl).

Secondly, the gel must be washed in denaturation buffer (348g NaCl, 80g NaOH into 4L H2O) to separate the double stranded genomic DNA into single stranded DNA and finally, the gel must be neutralized

(348g NaCl, 484 Tris HCL, 4L H2O, pH to 7.0 using HCl). Each of these steps must be repeated twice for 20 minutes at room temperature. The DNA will then be transferred overnight at room temperature using capillary action to to Hybond XL, a nitrocellulose membrane optimized for nucleotide transfer (GE

Healthcare). The next day, the blot will be crosslinked using a UV Stratalinker (Stratagene) and will be placed in pre-warmed Super Hyb (Molecular Research Center) for 4-6 hours in an incubator at 42°C. A probe fragment (primer sequences in Table 5) generated to the α-Tm 3’UTR, allowing visualization of both endogenous α-Tm and TG α-Tm will be radiolabeled with 32P dCTP using Amersham Rediprime II

Random Prime Labelling System (GE Healthcare), purified using illustra ProbeQuant G-50 Micro

Columns (GE Healthcare) and incubated with new, pre-warmed Super Hyb at 42°C overnight. The blot is then washed with 2X SSC with 0.1% SDS page and 0.2X SSC with 0.1% SDS twice for 30 minutes for each solution at 65°C. The blot will then be place on a phosphor screen overnight and will be imaged using a StormScanner 840 phosphor-imager (Amersham) and copy number will be quantified using

ImageQuant 5.1 (Molecular Probes). Each sample is normalized to that sample’s endogenous Tm copy number, allowing an internal control (2).

Table 6. Primers utilized for Southern blot analysis Primer Name Primer Sequence ATM 3’UTR Southern Forward 5’- CCA AGA CTC CTT CGT CAA GC -3’ ATM 3’UTR Southern Reverse 5’- TGA GCG TTG AGA CGA AAA TG -3’

Generation of an α-Tm S283 Phospho-Specific Antibody

The α-Tm protein sequence was provided to YenZym Antibodies, LLC. YenZym generated two peptides: the phospho-peptide CEELDHALNDMT-pS-I-COOH and the non-phosphorylated peptide

CEELDHALNDMTSI-COOH. The antigen peptide was injected into rabbits and subjected to ELISA to determine antigen-specific affinity. The phospho-antibody was purified by adsorption with the non- phosphorylated peptide. Both Tm 283 phospho-specific antibodies generated from the two rabbits injected with the antigen were extensively tested against recombinant and myofibrillar Tm. The YZ520 antibody was found to bind exclusively to phosphorylated Tm. There was some non-specific binding to the non-phosphorylated peptide in the YZ519 antibody that could likely be reduced with additional purification using the non-phosphorylated peptide conjugated to an affinity matrix.

Heart Homogenate, Myofibrillar Preparation and Cytoplasmic Preparation

Animals were sacrificed using cervical dislocation and hearts were briefly washed in ice cold PBS pH 7.4.

Hearts were then snap frozen in liquid N2. Hearts were homogenized in 1X Rigor buffer (100mM KCl,

20mM NaPO4, 5mM MgCl2, 5mM EGTA pH 8.0, 1mM DTT) containing proteinase inhibitor (Roche) and phosphatase inhibitor (Thermo Fisher) as previously described [22]. For experiments only requiring whole heart homogenate, the homogenate is sonicated and quantified using Bradford reagent (BioRad).

In order to extract myofibers, 1% Triton X 100 is added to the 1X Rigor buffer. The pellet is then resuspended and stored in a buffer containing 5mM ATP and 5mM CaCl2, sonicated and quantified using

Bradford reagent (BioRad).

In order to isolate the cytoplasmic fraction, the supernatant from early stages of myofibrillar preparation is collected, spun at 100,000 x g in a micro-ultracentrifuge (Sorvall) at 4°C for 1 hour. The supernatant is then carefully decanted and any remaining pellet is discarded.

Quantitation of TG Protein and Investigation of Myofibrillar Proteins

The α-Tm E180G mutation confers differential mobility on an SDS-PAGE gel while the S283A mutation does not [48]. Multiple methods are useful for quantifying TG vs. endogenous protein. The simplest method is loading 25-30 µg of myofibrillar preparation on a mini 10% SDS-PAGE gel with a crosslinking ratio of 29:1. Another similar method is to load 25-30µg of myofibrillar preparation on a 15cm 10%

SDS-PAGE gel with a crosslinking ratio of 200:1. After the gels are run, the gels are stained with

Coomassie Blue overnight and destained with 50% methanol and 10 % acetic acid. The gels are then scanned and the total Tm present in a sample is measured using ImageQuant 5.1 (Molecular Probes) followed by measurement of only the TG Tm, which runs below endogenous Tm. Percent TG Tm is calculated by dividing the value of TG Tm by total Tm and multiplying by 100. These methods also allow careful examination of the other proteins present in myofibrillar preparations in order to determine whether expression of a TG Tm results in altered expression of other myofibrillar proteins. Additionally, cytoplasmic preparations are also examined in this way to determine whether there are any detectable changes in TG samples versus NTG samples.

The final method useful in separating TG and endogenous protein levels is loading 4µg of myofibrillar protein on a 15cm 3.4M urea gel, transferring the gel to nitrocellulose overnight at 4°C and probing for

Tm (CH1) [17, 30, 31]. ImageQuant 5.1 is used to evaluate the total Tm present in the sample and then,

ImageQuant 5.1 is used to measure TG protein, if the endogenous protein and TG protein separate.

Matrix Assisted Laser Desorption/Ionization – Time of Flight (MALDI-TOF)

α-Tm S283A protein did not separate from NTG Tm by SDS-PAGE under any conditions tested, therefore, additional methods of analysis were employed. 30 µg of myofibrillar proteins from α-Tm

S283A Line 2, Line 25 and Line 97 TG hearts were resolved on a 10% SDS-PAGE gel along with 200 ng of α-Tm and α-Tm S283A recombinant proteins. The recombinant proteins were used as positive controls. The gels were stained with Coomassie Blue and then destained with 50% methanol and 10% acetic acid. The gels were provided to the University of Cincinnati Proteomics Core. The Tm band was excised from the gel, reduced, alkylated and subjected to a tryptic digest. Recovered peptides were desalted with a µ-C18 ZipTip and spotted onto a MALDI-TOF target plate. All spectra were acquired in

Reflector Positive Ion Mode on an ABSciex 4800 MALDI-TOF/TOF instrument. The percentage TG protein was calculated after normalization and subtraction of background contributions.

Two Dimensional Isoelectric Focusing

The 2-D PAGE was performed utilizing the procedure described by Warren [113] with slight modifications. 3 µg of myofibrillar protein preparation was mixed with 450 µl of IEF solubilization buffer (8 M urea, 2 M thiourea, 4% w/v CHAPS, 10 mM EDTA, 252 mM dithiothreitol (DTT), 2% v/v

3.5-5 pH immobilized pH gradient (IPG) buffer and 2mM tributylphosphate (TBP) and 0.001% bromophenol blue). Each sample was placed in the well of the IPG strip tray, and the IPG strip was placed on the sample. Active rehydration of the strip was done for 12 hours at 50 V. The focusing was done in 3 steps: step I: 250 V for 15 minutes with rapid ramp, step II: 10,000 V for 2 hours with slow ramp, step III: 10,000 V until 90,000 Vh with rapid ramp. During all steps the maximum current per gel was 50 μA.

After the isoelectric focusing, the strip was equilibrated in denaturation buffer (6 M urea, 2% SDS, 0.375

M Tris HCl pH 8.8, 40 % v/v glycerol, and 2% DTT ) for 10 minutes and then in neutralization buffer (6

M urea, 2% SDS, 0.375 M Tris HCl pH 8.8, 40 % v/v glycerol, and 2.5% iodoacetamide) for 10 minutes.

The gel is then run in the second dimension, which consists of a 10% SDS-PAGE resolving gel with a 4%

SDS-PAGE stacking gel. The gels are run at 8mA per gel overnight.

The SDS-PAGE gel was processed for Western blotting. Nitrocellulose membrane was used for transfer and the run conditions were 200mA for overnight at 4°C. Tm was visualized with the CH1 antibody

(Sigma Aldrich) at a dilution of 1:5000. Anti-mouse horse radish peroxidase (HRP) was used at dilution of 1:10000. Thermo Scientific Pico chemiluminescent substrate was incubated with the blot for 5 minutes and the signal was detected using autoradiography.

Intensity of the Tm protein signal was quantified using ImageQuant v5.1 software (Molecular Probes).

Quantification was done in such a way that the total tropomyosin on each blot was considered 100 percent. The percent phosphorylation of Tm was estimated as phosphorylated Tm / (phosphorylated Tm + non-phosphorylated Tm) x 100.

Western Blot Analysis of Cardiac and Myofibrillar Proteins

Western blot analyses on myofibrillar protein preparations (4 µg) from 3 month male α-Tm S283A TG,

180-S283A double mutant transgenic (DMTG) and appropriate NTG littermate hearts were conducted using the Tm specific antibody CH1 (Sigma Aldrich), pTm S283 phosphorylation specific antibody

(YenZyme) and sarcomeric α-actin antibody 5C5 (Sigma Aldrich) as a loading control.

Whole ventricular homogenates from 3 month old male α-Tm S283A TG, 180-S283A DMTG and appropriate NTG littermate controls were utilized to determine the expression levels of Ca2+ handling proteins. Western blots were used to visualize SERCA2a (Thermo Scientific), TnI (Cell Signaling), pTnI23/24 (Cell Signaling), phospholamban (PLN) (Thermo Scientific), phosphorylated serine 16 PLN

(PLN Ser16) (Badrilla), phosphorylated PLN threonine 17 (PLN Thr17) (Badrilla) and calcineurin (CnA)

(Sigma). Sarcomeric actin (Sigma) was used as a loading control. Table 7 details conditions utilized in

Western blot analyses in these studies.

Table 7. Conditions used in Western blot analyses Target [Protein] Source 1’Ab 1’Ab 2’Ab 2’Ab Blocking dilution dilution Total Tm 3-5 µg Myo Prep CH1 1:5000 Anti Mouse 1:5000 5% milk HRP pTm 3-5 µg Myo Prep pTM 1:10000 Anti Rabbit 1:10000 10% milk YZ520 HRP Actin1 3-5 µg Myo 5C5 1:10000 Anti-Rabbit 1:10000 5% milk Prep/Whole HRP Heart SERCA2a2 10-20 µg Whole MA3-910 1:5000 KPL anti 1:10000 5% milk heart mouse HRP pPLN3 10 µg Whole A010-12 1:1000 KPL anti- 1:10000 5% milk Ser16 heart rabbit HRP pPLN3 5 µg Whole A010-13 1:1000 KPL anti 1:10000 5% milk Thr17 heart rabbit HRP PLN3 5-10 µg Whole MA3-922 1:2000 KPL anti 1:10000 5% milk heart mouse HRP TnI 4-10 µg Myo Prep/ 4002 1:1000 KPL anti 1:10000 5% BSA Whole rabbit HRP heart pTnI 4-10 µg Myo Prep/ 4004 1:1000 KPL anti 1:10000 5% BSA Ser23/24 Whole rabbit HRP heart 1 Actin is often probed on the same membrane as other proteins (SERCA2a, PLN, etc). Decrease 1’Ab dilution to 1:20000. 2Boiling samples probed with SERCA2a alters the mobility of the protein. 3 Boiling PLN samples results in all protein running as the monomeric form. Unboiled samples run as both pentameric and monomeric PLN.

Investigation of the Effect of Euthanasia on Tm Phosphorylation

In order to determine whether the method of sacrifice of mice can influence the phosphorylation status of

Tm, mice were sacrificed separately using Avertin, isoflurane, CO2 and cervical dislocation in conformance with protocols established in “Guiding Principles for the Care and Use of Vertebrate

Animals in Research and Training” by the American Physiological Society. The Institution of Animal

Care and Use Committee approved the handling and maintenance of the animals. 6 week old male

FVB/N mice were given an intraperitoneal injection of Avertin at a dosage of 375ng/g body weight. As the anesthetic took effect, the animals were placed on a heating pad for 10 minutes to keep the animal warm and allow normalization of catecholamines and stress hormones. After toe pinch indicated the depth of anesthetic, the heart was dissected out and snap frozen. Similar to Avertin, animals were

39 anesthetized using 2.5-3% isoflurane at a 1L/min O2 flow rate, were placed on a heating pad with a nose- cone flowing by cervical dislocation, and hearts were dissected out and snap frozen. Finally, animals were sacrificed using CO2 followed by cervical dislocation and hearts were dissected out and snap frozen.

Myofibers were extracted as described earlier and 4µg of protein from each sample were run on a 10%

SDS-PAGE gel and probed for actin, total TM (CH1) and phosphorylated Tm. ImageQuant 5.1 was used to quantify expression. Phosphorylation of Tm was normalized to actin.

Gravimetrics

Animals were sacrificed using cervical dislocation at 3, 6, 9 and 12 months of age. The whole mouse body was weighed on a Mettler Toledo scale and the weight was noted. The hearts were quickly removed, carefully cleaned of excess tissue and rinsed in ice cold PBS pH 7.4. After the heart was gently blotted using a kimwipe, the hearts were weighed. The weight of the heart in mg was divided by the weight of the animal in g, providing the heart weight to body weight (HW:BW) ratio. After the whole heart was weighed, the atria were carefully removed and the ventricles were weighed, providing a ventricle weight to body weight ratio (VW:BW). Finally, the LV free wall, lumen and the intraventricular septum were dissected out and measured, providing a LV chamber weight to body weight (LVC:BW) ratio.

Lungs were removed and weighed. They were then dried for 7 days in a Napco vacuum oven and then weighed again in order to determine if there was peripheral edema in the lung. The liver was dissected out, weighed and dried for 7 days in a Napco vacuum oven as well. The dried livers were then weighed in order to determine whether there was infiltration of inflammatory cells.

Histology and Pathology

Animals were sacrificed using cervical dislocation at 3, 6, 9 and 12 months of age. Hearts were dissected out, washed and placed overnight in 10% neutral PBS buffered formalin. The hearts were then taken to the Cincinnati Children’s Hospital Medical Center Pathology Core to be sectioned. The hearts were

40 embedded in paraffin and were sectioned longitudinally in 5µm sections. The tissues were stained with hemotoxylin and eosin (H&E) for evaluation of myocyte organization and Masson’s Trichrome to evaluate for the presence of fibrosis. Images were taken on an Olympus BX40 and a Nikon SM2-ZT dissecting microscope with a Nikon DS-Ri1 12 megapixel digital camera. Unstained sections were stained with wheat germ agglutinin (WGA) conjugated to Texas Red (Sigma Aldrich). Images were taken at 40X with a fluorescent camera mounted on a Zeiss Axioscope in order to measure cardiomyocyte size.

Cardiomyocyte Cross-Sectional Areas

After staining heart sections with wheat germ agglutinin (WGA) conjugated to Texas Red, at least 10 random images were taken from similar regions of the LV free wall at 40X. ImageJ (NIH) software was used to analyze cell size; the scale bar was measured as a standard and at least 10 cells from every image were measured. Preference was shown for regular, cobble-stoned shaped cardiomyocytes, especially if the nucleus was visible. All measurements from a single heart were averaged and were considered an n of 1. At least 6 animals from each group being studied were measured for cardiomyocyte cross-sectional areas.

Echocardiographic Analysis

Echocardiographic measurements were performed utilizing a 30MHz high-resolution transducer (Vevo

770 High Resolution Imaging System) after anesthetization of 3 month old mice as previously described

[20]. Echocardiographic dimensions and thicknesses were taken from 2D-guided M-Mode from the parasternal long axis view in triplicate on NTG and TG 12-16 week old mice. Fractional shortening (FS, in %) was obtained by the formula: 100 x (LVIDd – LVIDs) / LVIDd; where LVIDd and LVIDs are left ventricular (LV) internal dimensions in diastole and systole, respectively. The relative wall thickness indices (RWT) were calculated by the formula: (LVAW + LVPW) / LVIDd; where LVAW and LVPW indicates anterior and posterior wall thicknesses, respectively; and LVIDd is the LV diastolic internal dimension. The LV outflow tract (LVOT) diameter (D, in cm) was measured in order to calculate the

LVOT cross-sectional area [LVOT CSA (cm2) = π(D/2)2]. The velocity-time integral (VTI, in cm) was

41 calculated by integrating the Doppler velocities in the LVOT. The product LVOT CSA (cm2) x VTI (cm) is the LV stroke volume (cm3), which multiplied by heart rate gives us the cardiac output (mL/min).

Transaortic Constriction and Echocardiography

Male NTG and α-Tm S283A TG animals between 12-16 weeks of age underwent transaortic constriction

(TAC) as described previously [140, 141]. Since only male α-Tm S283A TG mice exhibit cardiomyocyte hypertrophy via histological and echocardiographic analyses, only male animals were used in this study.

In brief, animals are anesthetized with 2% isoflurane at 0.5-1.0 L/min O2, intubated and connected to a rodent ventilator. Throughout surgery, degree of anesthesia is verified using toe pinch. Partial thoracotomy is performed and the sternum is retracted using a surgical chest retractor. Forceps are used to gently separate the thymus and remove any fatty tissue. The aortic arch is identified and a 27 guage needle is laid along the aorta. A length of 7.0 suture silk is threaded under the aorta and tied tightly around the aorta and needle with two separate knots and the silk is clipped close to the knot. The needle is then gently removed, along with the chest retractor. Sham operated animals undergo the entire procedure with the exception of tying off the aorta. The chest is closed using two knots of 9.0 suture silk and the skin is closed with suture glue. The animal is then given analgesic (buprenorphine 0.1mg/kg) IP and is also given saline solution to prevent dehydration [142].

Echocardiographic measurements were taken in M-Mode in triplicate on NTG and TG 12-16 week old mice. Twelve to sixteen week old male mice of both genotypes were subject to a TAC or sham operation as previously described [141]. Pressure gradients across the constriction were measured using Doppler echocardiography as previously described [143]. Two weeks post-surgery, mice were again subject to echocardiography and sacrificed for gravimetric and histological studies.

Measurements of Ca2+ Dependent Activation of Tension

Fiber bundles from papillary muscles of 5 month old male α-Tm S283A TG, α-Tm 180-S283A DMTG and appropriate NTG littermate control hearts were detergent-extracted in high relaxing buffer as

42 described previously [26] and were mounted between a force transducer and a micromanipulator. The sarcomeric length was adjusted to 2.0 µm and 2.2 µm using laser diffraction patterns and isometric tension was measured. Fiber bundles were then subjected to sequential Ca2+ solutions (pCa) and isometric tension was again measured. All experiments were carried out at 22°C.

Isolated Work-Performing Heart Model

Three month old male NTG and α-Tm S283A TG were anesthetized and treated with heparin to prevent microthrombi as previously described [144]. The aorta was cannulated, preserving the aortic valve and the coronary artery. In order to measure intraventricular systolic and diastolic pressures, an intraventricular catheter was inserted into the left ventricle. A cannula was also inserted into the left pulmonary vein, allowing the direction of the perfusate to be switched from retrograde (Langendorff) to anteriograde (working). COBE pressure transducers were utilized to measure aortic pressure, atrial pressure and LV pressure and were recorded using a Grass polygraph and digital acquisition system.

Blood Pressure Studies

Seven male NTG and 9 male α-Tm S283A TG animals from 8-12 weeks of age were subjected to blood pressure analysis using the CODA 8-Channel High Throughput Non-Invasive Blood Pressure System which allows measurement of the blood pressures of 8 mice at the same time. The machine is set up in a quiet room and the warming platforms are turned on. Animals are placed in the holders and are placed gently on the warming platform for 5 minutes. The tail is threaded through the occlusion cuff and the

VPR cuff and the animals are covered with a bench pad in order to darken the room and lessen animal anxiety. 20 measurements are taken during each experiment; 5 acclimation measurements and 15 ‘real’ measurements. Animals are acclimated in this way for 5-7 days prior to beginning to collect data. The experiment then runs for 5-7 days. The systolic blood pressure of a specific animal is averaged for every reading taken on specific days and are then averaged to all blood pressures for a specific animal over the length of the experiment resulting in an n of 1. Finally, the systolic blood pressure values for all NTG animals and TG animals are analyzed [145].

Calcineurin/ Protein Phosphatase Activity Assay

Calcineurin activity (CnA), also known as protein phosphatase 2B (PP2B), was measured using a calcineurin/PP2B activity kit (CalBioChem). Cardiac homogenates from 3 month old male NTG and α-

Tm S283A TG hearts were used. CnA activity is measured as the rate of dephosphorylation of a synthetic peptide in the presence and absence of EGTA, okadaic acid, and EGTA with okadaic acid. Phosphate release was measured by the colorimetric Green Reagent (CalBiochem).

Statistics

All statistics are presented as mean±SEM. Where appropriate, paired and unpaired t-tests, ANOVA with

Bonferroni correction and ANOVA with repeated measures were used to evaluate significance.

Significance was set at p<0.05.

Generation of Recombinant α-Tm, β-Tm and γ-Tm NTG and Mutant Proteins

α-Tm, β-Tm and γ-Tm cDNA was PCR amplified from the pBS+ vector with a start primer that added an ala-ser dipeptide necessary to functionally compensate for lack of acetylation in bacterially expressed Tm

[16] and a stop primer that did not include the 3’UTR (Table 8). The cDNAs were then TA-tail cloned into the pGEM-T easy vector (Promega) (Figure 10). The size of the pGEM-T easy vector including the cDNA was appropriate for SDM (Agilent Technologies), allowing the introduction of the S283A and

S283D mutations into the α-Tm cDNA (Table 9). β-Tm and γ-Tm in pGEM-T easy (Promega) were subjected to SDM (Agilent) utilizing primers listed in Table 9 in order to generate β-Tm (Ser158Ala and

Ser179Ala) and γ-Tm (Val18Ala and Thr179Ala) mutant proteins that more closely resemble α-Tm in the number of alanines present in alanine clusters (Figure 4). Upon sequencing of the NTG and mutant cDNAs, the cDNA was again PCR amplified and was TA-tail cloned into the pET-SUMO vector (Life

Technologies) to allow for bacterial expression of the NTG and mutant proteins (Figure 10). The pET-

SUMO vector includes a 6X HIS tag and a SUMO cleavage moiety to allow easy purification of protein.

The NTG and mutant proteins were expressed and purified according to the Champion pET-SUMO

Expression System protocol (Life Technologies).

The α-Tm mutants will be utilized in MALDI-TOF analysis detailed above and, in the future, may be subjected to a variety of biochemical experiments. The mutants in β-Tm and γ-Tm have been generated individually as well as together in the appropriate Tm isoform in order to determine whether one of these mutations plays a more dominant role in possible differences in biochemical properties. These proteins will be used in biochemical studies to assess whether the number of alanines found in alanine clusters contributes to differential isoform function.

Table 8. Primers used to clone striated muscle Tm isoforms into pGEM-T easy and pET-Sumo Primer Name Primer Sequence ATM ala-ser Forward 5’- GCT TCT ATG GAC GCC ATC A -3’ ATM Stop Reverse 5’- TTA TAT GGA AGT CAT ATC GT -3’ BTM ala-ser Forward 5’- GCT TCT ATG GAC GCC ATC AA -3’ BTM Stop Reverse 5’- TCA GAG GGA AGT GAT GTC A -3’ GTM ala-ser Forward 5’- GCT TCT ATG GAG GCC ATC A -3’ GTM Stop Reverse 5’- TTA TAT AGA GGT CAT GTC A -3’

Figure 10. Plasmids involved in generating Tm recombinant proteins. A. pGEM-T Easy (Promega). B. pET-SUMO (Life Technologies).

Table 9. Primers used to generate α-Tm, β-Tm and γ-Tm mutants Primer Name Primer Sequence αTM S283A Stop Reverse 5’- TTA TAT GGC AGT CAT ATC GT -3’ αTM S283D Stop Reverse 5’- TTA TAT GTC AGT CAT ATC GT -3’ βTM S158A SDM Forward 5’- GCT GAG GAC GCA GAC CGC AAA -3’ βTM S158A SDM Reverse 5’- TTT GCG GTC TGC GTC CTC AGC -3’ βTM S179A SDM Forward 5’- TGG AGC GCG CGG AAG AGA GA -3’ βTM S179A SDM Reverse 5’- TCT CTC TTC CGC GCG CTC CA -3’ γTM V18A SDM Forward 5’- GAC AAA GAG AAT GCT CTG GAC CGA GC -3’ γTM V18A SDM Reverse 5’- GCT CGG TCC AGA GCA TTC TCT TTG TC -3’ γTM T179A SDM Forward 5’- GAC TTG GAA CGC GCA GAG GAA CGT G -3’ γTM T179A SDM Reverse 5’- CAC GTT CCT CTG CGC GTT CCA AGT C -3’

Actin Binding Assay

Actin affinity of Tm and Tm mutants is measured through cosedimentation assays in the presence of bovine cardiac F-actin (Cytoskeleton, INC) and in the presence and absence of the Tn complex. To measure actin-Tm affinity, Tm (0.0-10.0µM) is combined with 5µM actin, incubated in 200 mM NaCl, 10 mM Tris-HCL pH 7.5, 2 mM MgCl2 and 1mM TCEP (Fisher Scientific) for 1 hour at 25°C. The samples are then cosedimented at 100,000 x g in a Sorvall RC M120GX micro-ultra centrifuge for 1 hour at 25°C.

The pellet and supernatant are analyzed via 10% SDS-PAGE gels stained with Coomassie Blue and quantitated using ImageQuant5.1 (Molecular Probes). The concentration of free Tm in the supernatant was calculated using standard curves prepared using known amounts of the same isoform/mutant Tm used in the cosedimentation experiment. Bovine serum albumin (BSA) is included in the sample buffer as an internal control to correct for variability in sample loading. Apparent binding constants (Kapp) and Hill

Coefficients (αH) can be determined using GraphPad to fit the data to the following equation:

αH αH αH αH v = n[TM] * Kapp /1 + [Tm] * Kapp

Human Tn complex (generous gift from Brian Biesiadecki, PhD) facilitates the binding of Tm to actin.

Additional cosedimentation experiments can also be carried out in the presence of 2µM of Tn complex

[146].

Circular Dichroism Measurements

Recombinant proteins were dialyzed into 500mM NaCl, 10mM NaPO4 pH 7.5 and 1 mM EDTA overnight at 4°C. Immediately prior to experimentation, 1mM of TCEP was added directly to the sample to reduce disulfide bonds in the Tm protein and was incubated at room temperature for 10 minutes. Tm unfolding occurs with 3 transitions, but if the protein is not completely reduced, Tm unfolding shows up to 4 distinct transitions, complicating the analysis of protein unfolding [147]. Circular dichroism measurements were made by following the ellipticity as a function of temperature. Data were collected on an Aviv Model at 222nm. The apparent TM is defined as the temperature at which the ellipticity at 222 nm is at the midpoint between the value found at 0°C where all NTG Tm recombinant proteins and all Tm mutant recombinant proteins were fully folded and 95°C, where all proteins were fully unfolded. The Tm proteins were subjected to a cooling procedure in an attempt to refold the protein after the thermal melt.

CHAPTER 3: TROPOMYOSIN DEPHOSPHORYLATION RESULTS IN COMPENSATED

HYPERTROPHY

Emily M. Schulz, Richard N. Correll, Hajer N. Sheikh, Marcos S. Lofrano-Alves, Patti L. Engel, Gilbert Newman, Jo El J. Schultz, Jeffery D. Molkentin, Beata M. Wolska, R. John Solaro, David F. Wieczorek.

Summary

Phosphorylation of tropomyosin (Tm) has been shown to vary in mouse models of cardiac hypertrophy.

Little is known about the in vivo role of Tm phosphorylation. This study examines the consequences of

Tm dephosphorylation in the murine heart. TG mice were generated with cardiac specific expression of

α-Tm with serine 283, the phosphorylation site of Tm, mutated to alanine. Echocardiographic analysis and cardiomyocyte cross-sectional area measurements show that α-Tm S283A TG mice exhibit a hypertrophic phenotype at basal levels. Interestingly, there are no alterations in cardiac function, myofilament calcium (Ca2+) sensitivity, cooperativity or response to β-adrenergic stimulus. Studies of

Ca2+ handling proteins show significant increases in SR ATPase (SERCA2a) protein expression and an increase in phospholamban (PLN) phosphorylation at serine 16, similar to hearts under exercise training.

Compared to controls, the decrease in phosphorylation of α-Tm results in greater functional defects in TG animals stressed by transaortic constriction to induce pressure overload-hypertrophy. This is the first study to investigate the in vivo role of Tm dephosphorylation under both normal and cardiac stress conditions, documenting a role for Tm dephosphorylation in the maintenance of a compensated or physiological phenotype. Collectively, these results suggest that modification of the Tm phosphorylation status in the heart, depending upon the cardiac state/condition, may be a potential strategy for prevention or treatment of cardiac hypertrophy.

Introduction

Tropomyosin (Tm) is an alpha helical coiled coil protein involved in the Ca2+ dependent regulation of the thin filament of the sarcomere. Upon binding of Ca2+ to the troponin complex, a conformational change occurs which allows the Tm filament to move away from the myosin-head binding site on the sarcomeric actin filament. Previous and recently published studies show striated muscle α-Tm is phosphorylated at one site, the penultimate amino acid, serine 283, by several potential kinases including tropomyosin kinase, protein kinase A (PKA) and protein kinase C zeta (PKCζ) [110, 116-118, 148-151]. During fetal development, 70% of cardiac α-Tm in rat hearts is phosphorylated which decreases to approximately 30% post-natally [109]. In vitro studies investigating the functional role of Tm phosphorylation indicate that low phosphorylation levels decrease the ability of α-Tm to polymerize in a head-to-tail fashion; conversely, increasing phosphorylation enhances the interaction between the C and N-terminal ends of adjoining Tm molecules. Additionally, changes in α-Tm phosphorylation status seem to alter sarcomeric function, as shown by differential function of the actin-activated myosin S1-ATPases [118, 119]. Taken together, these in vitro data suggest that altering phosphorylation status affects the ability of Tm to cooperatively activate the thin filament upon binding of Ca2+ to troponin (Tn).

In recent years, in vivo studies performed on animal models indicate that changes in the phosphorylation status of sarcomeric proteins such as troponin I (TnI), myosin binding protein C (MyBPC) and the regulatory myosin light chain (MLC) result in alterations in Ca2+ sensitivity of the myofilament, changes in cardiac function and may play a role in the development of cardiac disease [90, 152-155].

Investigation of a DCM TG mouse model bearing a human α-Tm mutation (E54K) shows that phosphorylation levels of Tm decrease relative to NTG littermates [112, 113]. Additionally, phosphorylation is increased in FHC -Tm N175D mice generated by this laboratory indicating a link between striated muscle Tm phosphorylation, sarcomeric function and cardiac disease [115] (Sheikh and

Wieczorek, unpublished). In order to investigate the in vivo effect of decreased or ablated Tm phosphorylation, we substituted serine 283 with an alanine (S283A), removing the phosphorylation site

50 and effectively inhibiting the ability of α-Tm to be phosphorylated. Several TG mouse lines expressing this α-Tm S283A mutation were generated and analyzed. These TG hearts show no changes in functional parameters when investigated by echocardiography, myofilament Ca2+-tensionrelations or in studies of work-performing heart during β-adrenergic stimulus. However, these animals do have sex specific differences in heart morphology, likely due to the cardioprotective effects of estrogen which has been described previously [156, 157]. Male TG mice show a hypertrophic phenotype as measured by echocardiography and supported by cardiomyocyte cross-sectional area measurements. These mice also show significant modifications in proteins controlling Ca2+ fluxes such as increases in the expression of the SERCA2a and phosphorylation of phospholamban (PLN). Thus, phosphorylation of α-Tm may be part of a signaling cascade which results in changes in Ca2+ handling protein levels and may explain the tight regulation of α-Tm phosphorylation levels. Additionally, when male TG animals are subject to pressure-overload via transaortic constriction (TAC), they exhibit a significant increase in hypertrophy as well as functional defects including a striking decrease in fractional shortening compared to NTG litter mates. This is the first investigation to show that alterations in the phosphorylation status of a thin filament protein, namely α-Tm, can cause a moderate hypertrophic response and increase SERCA2a expression and PLN phosphorylation. Taken with our previous findings in disease models, these results firmly establish that α-Tm phosphorylation is necessary for an appropriate cardiac response during cardiac disease.

Results

Generation of α-Tm S283A TG Mice

To determine the functional significance of Tm phosphorylation, we generated TG mice in which the Tm phosphorylation site (serine residue 283) was replaced with a non-phosphorylatable alanine residue

(S283A). The transgene construct used to generate α-Tm S283A TG mice is shown in Figure 11A.

Multiple TG lines were generated and studied. Line 2 has the highest TG mRNA expression and the second highest copy number of all transgenic animals generated (17 copies) determined by genomic

Southern blot analysis. Forward and reverse sequencing of the construct indicates no mutations or deletions in the transgene.

Cardiac α-Tm S283A Protein Expression and Phosphorylation in Transgenic Mice

Often, mutations in Tm isoforms lead to differential migration on SDS-PAGE gels. However, because serine is only 16 daltons larger than alanine and has nearly an identical isoelectric point, expression levels of TG and endogenous protein cannot be separated using traditional methods. Instead, myofibrillar protein preparations of age matched NTG and TG mouse hearts, as well as recombinantly expressed NTG or TG protein, are resolved by a combination of SDS-PAGE and matrix assisted laser desorption/ionization time of flight (MALDI-TOF) analyses (Figure 11B). The ratio of serine containing peptides (endogenous Tm) and alanine containing peptides (TG Tm) is calculated after normalization and background subtraction. As an additional control, the peptides corresponding to both the serine and alanine profiles are further fractionated to ensure that the proper tryptic peptide is being analyzed. Line 2 has approximately 93.7 percent TG protein expression, Line 25 has approximately 86% TG protein expression and Line 97 has approximately 88% TG protein expression (Figure 11C), with a concomitant decrease in NTG protein, maintaining total Tm levels at 100%. Investigation of the cytosolic fraction shows no significant accumulation of either endogenous or TG Tm, indicating that the TG protein is

Figure 11. α-Tm S283A TG animal characterization. A. α-TM S283A construct. The α-MHC promoter drives cardiac specific expression of the striated muscle α-Tm cDNA with an encoded substitution at amino acid 283 (Ser283Ala). B. MALDI-TOF profiles of NTG and TG recombinant α-Tm and NTG and TG Line 2, Line 25 and Line 97 α-Tm S283A from myofibrillar preparations 3 month old mouse hearts. C. Quantification of B. D. Myofibrillar preparations from each of the 3 TG lines studied. Note that Line 97 samples were run on a separate gel.

53 properly incorporated into the myofibril (data not shown). Additionally, myofibrillar protein preparations run on an SDS-PAGE gel show that all myofibrillar proteins are present in the proper ratio in all three TG lines, indicating that the myofibrils are being properly assembled and there is no change in total Tm levels

(Figure 11D).

Tm Phosphorylation in NTG and α-Tm S283A Mouse Hearts

In order to study Tm phosphorylation in TG mice, it was necessary to establish the basal level of Tm phosphorylation in NTG hearts using 2-D IEF-PAGE. Results show an unphosphorylated and a single phosphorylated species of Tm appears in NTG heart samples (Figure 12A). Upon CIP treatment, the phosphorylated Tm protein species is lost. These results are in agreement with previously published studies that identify Ser283 as the phosphorylation site in striated muscle Tm [109, 110, 113, 151].

Further analysis shows a trend of decreasing Tm phosphorylation from 6 weeks to 5 months of age with an average of approximately 30%. At 15 months of age, animals show a significant increase in Tm phosphorylation, indicating a possible return to fetal gene programs due to senescence (Figure 12B) [158-

160].

In order to determine the phosphorylation status of Tm in TG myofibrillar preparations, we generated a

Tm Ser283 phosphorylation specific antibody. As seen in Figures 12C and 12D, there is a clear decrease in the phosphorylation status of Tm in the TG myofibrillar preparations compared to the NTG preparations. As TG Line 2 had the greatest decrease in phosphorylation and exhibited the same phenotype in comparison to the other TG lines, we chose to focus on Line 2 TG mice. When considering the phosphorylation status of these S283A TG mice, it is important to remember that endogenous Tm in

NTG mice is phosphorylated at 30%. Line 2 has approximately 5 fold less (or 80% less) phosphorylation then NTG littermates, corresponding to 6% endogenous Tm available for phosphorylation in this line.

We believe Line 25 has more endogenous Tm phosphorylation because of its lower level of transgene

54 expression. These results suggest that most, if not all, of the endogenous Tm in the TG mice is being phosphorylated.

Figure 12. Phosphorylation status of NTG and α-Tm S283A TG animals. A. 2-D IEF-PAGE gels show that Tm has one phosphorylation site indicated by the arrow (upper panel) which can be removed after calf intestinal phosphatase (CIP) treatment (lower panel). Cardiac myofibrillar protein preparations were probed with the CH1 striated muscle Tm antibody. B. Percent of total Tm phosphorylated in NTG mice measured using 2-D IEF on myofibrillar preparations taken at 1.5, 3 and 5 months of age. C. Western blot analysis of in α-Tm phosphorylation from hearts at 3 months of age. n=3. D. Quantification of phosphorylation levels of α-Tm found in panel C.

Gravimetrics and Cardiac Morphology of α-Tm S283A TG Hearts

Morphological analyses of the LV wall shows a very mild increase in cardiomyocyte disarray and disorganization as indicated by centrally located nuclei and partial loss of the typical cobblestone shape of the cardiomyocyte (Figure 13A). Staining the membrane with wheat germ agglutinin (WGA) and measurement of cross-sectional area shows a significant increase in TG cardiomyocyte area (445.5 ± 17.4

µm2 vs 686.9 ± 66.9 µm2 p<0.05, NTG and TG respectively) (Figure 13B). Gravimetric analysis was performed on TG animals from 1 month to 9 months of age. Interestingly, results show no changes in heart weight to body weight ratios, likely due to the moderate nature of this hypertrophy (Figure 13C).

There are no differences in the survival of NTG and TG mice.

Figure 13. Histology and Gravimetrics. A. Immunohistochemistry of 3 month α-Tm S283A TG hearts stained with hematoxylin and eosin (i) and wheat germ agglutinin (WGA) (ii). B. Cardiomyocyte cross- sectional area measurements. *p<0.05, TG:n=4, NTG:n=4. C. Heart weight to body weight ratios of 3 month old mice TG:n=6, NTG:n=6.

Cardiac Function of α-Tm S283A TG Hearts

To determine whether the relationship between Ca2+ concentration and force-tension development is altered in myofilaments at the sarcomeric level in TG hearts with significantly decreased phosphorylation of Tm, we analyzed skinned fiber bundles from the papillary muscle of 5 month old hearts. No significant changes in absolute tension or normalized tension in NTG versus TG mice are found (Table 9,

Figure 14A, 14B). Additionally, there are no significant differences in pCa50 or the Hill coefficient (nH), a measure of the cooperative activation of the thin filament of the sarcomere.

Table 10. Parameters involved in Ca2+-tension relations in skinned fiber bundles Group pCa50 nH n NTG 5.280 ± 0.19 1.64 ± 0.09 4 TG 5.251 ± 0.01 1.81 ± 0.07 4

The work-performing heart model was utilized to determine ex vivo functional effects of the decrease in

Tm phosphorylation status. These measurements were performed in mice at three months of age. At basal levels, there are no changes in contraction and relaxation parameters in TG hearts. Additionally, when isoproterenol is administered in order to determine whether the β-adrenergic response is impaired, there are no significant differences in contraction and relaxation (Figure 14C, 14D).

To assess whether decreasing the phosphorylation level of α-Tm has an effect on in vivo cardiac function, we performed echocardiographic analysis on 3 month old NTG and TG mice. There are no physiological changes in heart function between the NTG and TG mice as shown by fractional shortening, cardiac output or ejection fraction (Table 2). However, there are sex specific differences in cardiac morphology.

Male TG animals show significant increases in LV mass, LV anterior wall thickness, LV posterior wall thickness and LV relative wall thickness index, indicating that TG mice have a hypertrophic phenotype without attendant functional defects. Female TG mice show no changes when compared to female or male NTG hearts. Differences in the development of cardiac hypertrophy between sexes has been previously noted [157, 161]. Thus, the increase in cardiomyocyte area and LV hypertrophy with no

58 change in heart weight to body weight ratio or female cardiac enlargement, demonstrates the moderate nature of this hypertrophic phenotype.

Figure 14. Ca2+ sensitivity and work performing heart parameters. A. Normalized force in skinned fiber bundles in NTG and TG hearts. B. Ca2+ - tension relations in NTG and TG skinned fiber bundles. C&D. Isoproterenol dose-response curves in TG and NTG mouse heart subjected to isolated heart analyses.

Table 11. Cardiac function of NTG and α-TM S283A TG mice at 3 months of age as assessed by M- Mode echocardiography Parameters NTG (n=4) TG (n=4) LVIDd, cm 0.40 ± 0.11 0.42 ± 0.13 LVISd, cm 0.28 ± 0.14 0.29 ± 0.24 LV mass, mg 82.71 ± 3.34 112.5 ± 9.99* LVAW, cm 0.07 ± 0.04 0.08 ± 0.01* LVPW, cm 0.07 ± 0.014 0.09 ± 0.06* RTW, cm 0.03 ± 0.02 0.04 ± 0.02* CO, ml/min 25.13 ± 2.48 25.33 ± 2.29 HR, bpm 469.5 ± 24.82 453.3 ± 24.88 EF 58.39 ± 1.99 61.49 ± 6.75 FS % 30.54± 1.29 33.50 ± 4.65 LVIDd indicates LV internal diastolic dimension; LVISd, LV internal systolic dimension; LVAW, LV anterior wall dimension, LVPW, LV posterior wall dimension; RTW, relative wall dimension; CO, cardiac output; HR, heart rate; EJ, ejection fraction. *p<0.05.

Gene Expression and Protein Changes in α-Tm S283A TG Hearts

Given that histological and echocardiographic analyses indicate that TG mice exhibit a moderate hypertrophic phenotype at 3 months of age, altered gene expression was determined in α-Tm S283A TG hearts. Real time RT-PCR analysis of the RNA isolated from ventricular tissue indicate a trend toward an increase in β-MHC, brain natriuretic peptide (BNP) and atrial natriuretic peptide (ANP) without significant statistical increases (Figure 15A).

Genes involved in cardiomyocyte Ca2+ handling were also examined. Interestingly, there are no changes

2+ in gene expression of SERCA2a, the L-type Ca channel, NCX, PLN or RyR2 (Figure 15B). However, there is a significant increase (p<0.05) in the gene expression of MCIP1, a protein involved in modulating

CnA activity in vivo [162].

To determine whether real time RT-PCR levels of Ca2+ handling genes correlate with corresponding protein expression in the TG myocardium, protein expression from whole hearts was determined by

Western blot analysis (Figure 15C, 15D). Results indicate that the phosphorylation site at PLN Ser16 and

SERCA2a protein expression are increased by more than 30% over NTG levels. There are no changes in

60 total PLN, phosphorylation at PLN Thr17, TnI, phosphorylation at TnI 23/24 or CnA. The lack of increased SERCA2a gene expression by real time RT-PCR analysis compared to the significant increase in protein expression suggests that increased protein stability or translation may be operative. MICP1 gene expression is utilized as a marker of alterations in CnA protein expression or activity. Although

MICP1 gene expression is increased, there is no increase in CnA expression. However, it is possible that

CnA activity may be altered without altering protein expression.

Calcineurin Activity Assay

As MIC1P can be both a facilitator and an inhibitor of CnA activity, it is necessary to determine if changes in mic1p gene expression alters CnA activity, as CnA is an important regulator in the Ca2+ handling process. A CnA/PP2B activity assay was performed (CalBioChem) on whole heart preparations from 3 month old mouse hearts. There are no significant changes in CnA activity (Figure 15E).

Figure 15. Cardiomyopathy markers, changes in Ca2+ flux proteins and calcineurin activity assay in α-Tm S283A TG animals. A. qRT-PCR analysis of cardiomyopathy marker genes in 3 month old mouse hearts. There is no significant difference in gene expression for any of the genes included in this profile. NTG n=12, TG n=10. B. qRT-PCR analyses of cardiac gene expression normalized to GAPDH. SERCA2a, MCIP1 NTG n=12, TG n=10. NCX, PLN, L-Type, RyR2 NTG: n=6, TG: n=5. C. Western blot analysis of Ca2+ handling proteins in NTG and TG hearts. D. Quantification of Ca2+ handling protein levels. PLN Ser16, PLN Thr17 phosphosphorylation levels were normalized to total PLN. All other protein expression was normalized to actin. n=8. E. Phosphatase activity in mouse heart extract. NTG: n=3, TG: n=3. *p< 0.05, **p<0.01.

Cardiac Function in α-TM S283A Pressure Overload Mouse Hearts

To determine the effect of decreased α-TM phosphorylation during cardiac stress and disease, 12-16 week old TG mouse hearts were subject to transaortic constriction (TAC), along with NTG littermates and sham operated animals from both genotypes.

Animals were subject to echocardiography prior to surgery as well as 2 weeks after TAC and were then sacrificed for histology and gravimetric studies. NTG and TG TAC operated animals show significantly increased pressure gradients at 2 weeks, indicating the efficacy of the pressure overload model (Figure

16A). LVIDs and LVIDd are the only parameters to show significant alterations in function between

NTG TAC and TG TAC operated animals (Figure 16B, 16C). Interestingly, although NTG TAC operated hearts have a greater pressure gradient, TG TAC operated hearts are the only group that experiences a significant decrease in percent fractional shortening (Figure 16D). NTG sham, TG sham and NTG TAC all have fractional shortening at 33 percent, while the TG TAC operated group has fractional shortening at 24 percent, indicating impairment in cardiac function in that group. These data indicate that significantly decreasing the phosphorylation status of α-Tm impairs the ability of the myocardium to properly respond to stress.

Gravimetrics and Cardiac Morphology in α-TM S283A TAC and Sham Operated Animals

NTG TAC operated animals have a significant increase in heart weight to body weight when compared to

NTG sham operated animals (p<0.0001) (Figure 16E). Additionally, TG TAC operated animals show an increase in heart weight to body weight ratios compared to TG sham operated animals (p<0.05).

H&E staining of NTG sham operated hearts shows no changes in cardiomyocyte morphology, though TG sham operated hearts show a mild increase in disorganization. Masson’s Trichrome staining of both NTG and TG sham operated hearts show no significant increases in the deposition of fibrotic tissue in the LV free wall (Figure 16G i, 16G ii). NTG TAC and TG TAC operated heart sections stained with H&E show cardiomyocyte disorganization and centrally located nuclei (Figure 16G iii). Both NTG and TG TAC

63 operated hearts stained with Masson’s Trichrome show increases in fibrosis. Cardiomyocyte cross- sectional analyses demonstrate significant increases in TG sham, NTG TAC and TG TAC operated hearts

(p<0.001) compared to NTG sham mice. TG TAC cardiomyocytes show greater increases in size compared to both TG sham and NTG TAC cardiomyocytes (p<0.0001, p<0.01) (Figure 16F).

Figure 16. Echocardiographic analyses, gravimetrics and histology of NTG sham, TG sham, NTG transaortic constriction operated (TAC) and TG TAC hearts from 12-16 week old mice. A. Pressure gradients in NTG sham, TG sham, NTG TAC, TG TAC hearts. B&C. Diastolic and systolic LV internal dimensions (LVIDd,s), respectively. D. fractional shortening (%FS). *p<0.05, **p<0.01, ***p<0.001. E. Heart weight to body weight ratio of NTG and TG sham operated hearts and NTG and TG TAC operated hearts. F. Cardiomyocyte cross-sectional area measurements. n=6 for all groups. G. Tissue sections from NTG sham, TG sham, NTG TAC and TG TAC operated hearts stained with hemotoxylin and eosin (H&E) (i), Masson’s Trichrome (ii) and wheat germ agglutinin (WGA) (iii). All images taken at 40X. Scalebar indicates 50 µm.

Discussion

Post translational modifications, such as alterations in the phosphorylation status of sarcomeric and Z-disc proteins can result in altered cardiac contractility with progression to disease and death [90, 152-154].

This is the first in vivo study investigating the functional role of cardiac α-Tm phosphorylation. To address this, the single Tm phosphorylation site, serine 283, was changed to an alanine and TG animals were generated for study. In vivo assessment of basal cardiac function of α-Tm S283A TG mice shows that the hearts exhibit a moderate compensated hypertrophic phenotype with an increase in myocyte size due to a stimulus initiated by decreased Tm phosphorylation. It is possible that the increase in cardiomyocyte size occurs in response to mechanical defects induced by autophagy or apoptosis, two cell death processes involved in the transition from compensated to decompensated hypertrophy [163, 164].

However, the data suggest this compensatory response is an attempt to normalize LV wall stress and preserve pump function, which may point toward the role that autophagy places in maintaining cell and tissue homeostasis [165]. Increases in LV mass, LVAW, LVPW, RTW and increases in cardiomyocyte cross-sectional area, coupled with a lack of functional defects in contractility, indicate the TG hearts are in a compensated or adaptive state of hypertrophy in response to decreased Tm phosphorylation. Further characterization of contractile function assayed by skinned fiber preparations indicate that TG myofibers develop force-tension relations that are similar to NTG controls. Also, TG myofibers lacking Tm phosphorylation do not exhibit changes in Ca2+ sensitivity or cooperative activation of the thin filament.

Tm phosphorylation has long been speculated to play a role in the modulation of the structural and functional properties of the thin filament given that the site of phosphorylation is serine 283 and is located in the 7-11 amino acid head-to-tail overlap region between neighboring Tm molecules. Although cooperativity is not entirely understood, studies suggest that Tm head-to-tail interactions between contiguous Tm dimers play a role. A study by Gaffin et al indicates that substitution of negatively charged amino acids at the C-terminus causes a significant change in the distance between Tm monomer strands and possibly alterations in contiguous Tm molecule interactions [166]. In vitro studies

66 investigating the striated muscle Tm phosphorylation site indicate that changing the phosphorylation status of α-Tm alters the head-to-tail interaction between neighboring Tm molecules [118]. Contrary to expectation, removing the phosphorylation of α-Tm at Ser283 in an in vivo system does not result in any alterations in cooperative activation of the thin filament. Thus, phosphorylation may not be a major modulator of cooperative spread of activation in the myofilament lattice and may be more significantly related to the actin filament independent of Tm head-to-tail interactions [167]. NMR studies of the interaction between the N and C- terminal dimers indicate that the last 2 to 5 amino acids at the C- terminus are flexible [168]. The fact that the very last C-terminal residues are mildly disordered may offer some explanation as to why loss of additional negative charges in the form of phosphorylated serine has no effect on cooperativity or Tm head-to-tail interactions. The phosphorylated Ser283 may be in an area too flexible to allow for strong interaction with residues in the N-terminus of the subsequent Tm molecule.

The lack of change in cardiac function, myofilament cooperativity and Ca2+ sensitivity, coupled with the development of compensated hypertrophy in the TG animals, warranted an investigation into the possible mechanisms involved in the hypertrophic response. The gene expression profile of the α-Tm S283A TG found that mcip1 significantly increases. Mcip1 gene expression is utilized as a marker for calcineurin

(CnA) activity and has been alternatively shown as both a facilitator and inhibitor of CnA activity in vivo

[162, 169, 170]. Increases in CnA activity have been shown to induce cardiac hypertrophy in mouse models and conversely, the development of a hypertrophic phenotype can be prevented via CnA inhibition [171, 172]. Interestingly, in the α-TM S283A TG animals, the increase of mcip1 mRNA did not result in changes of CnA expression or activity, indicating another effector downstream of Tm phosphorylation loss may be responsible for the hypertrophic phenotype.

As numerous studies have demonstrated the importance of Ca2+ in the modulation of cardiac hypertrophy, we examined whether alterations in expression of Ca2+ proteins occurred in TG mice. Although there are no changes in total TnI expression or TnI phosphorylation at amino acids (aa) 23 or 24, there are

67 alterations in SR proteins. Increasing SERCA2a protein expression and/or activity can rescue multiple disease phenotypes and improve myofibrillar efficiency and contractile parameters both in human cardiomyocytes and rodent hearts [51, 66, 173, 174]. Conversely, in animal models of cardiac disease, as well as human patients with HF, SERCA2a protein levels and activity often decrease [38, 175].

Surprisingly, SERCA2a protein levels are increased in the α-Tm S283A TG hearts by approximately 30% over NTG levels. Additionally, the 30% increase in PLN phosphorylation at Ser16 indicates that further restriction on SERCA2a activity has been released, as PLN in an unphosphorylated state results in inhibition of the pump [68, 176, 177]. Similar to increasing SERCA2a expression and activity, phosphorylation of PLN Ser16 results in a hypercontractile heart. However, there are no changes in cardiac contractility in the α-Tm S283A mice. Rather, normal cardiac function as measured by multiple methods is preserved in the TG hearts rather than enhanced. This indicates that the increase in SERCA2a protein levels, as well as the increase in PLN Ser16 phosphorylation, may be necessary to maintain normal cardiac function in a heart in which Tm has largely been dephosphorylated. It is possible, therefore, that if SERCA2a activity was inhibited, a greater degree of hypertrophy and/or a progression to decompensated cardiomyopathy and heart failure would result.

Increased SERCA2a protein and activity levels are associated with physiological hypertrophy. In most exercise trained models, PLN protein levels are unchanged, though changes in PLN phosphorylation are seen [74, 178]. This is similar to the results found with the α-Tm S283A mice. Additionally, in animals that have compensated or physiological hypertrophy due to exercise training, there are no changes in gene expression of common cardiomyopathy markers, identical to what is seen in the TG mice investigated here [179]. Exercise training improves cardiomyocyte contractility and calcium handling and often improves disease in both animal and human models of cardiac disease [75, 76, 180]. Although the mechanisms responsible for the development of physiological hypertrophy during exercise training are not well elucidated, we speculate that ablating the phosphorylation site of α-Tm results in a similar signaling cascade that occurs in response to exercise training. In a physiologically exercised heart, the

68 signaling pathways that are activated result in improved cardiac function, while in response to Tm dephosphorylation, the TG hearts are able to function normally with a mild hypertrophic phenotype.

Increases in the level of SERCA2a appear to be responsible for cardiac dysfunction in the TG TAC operated animals compared to NTG TAC operated animals. Previous work indicates that SERCA2a upregulation by approximately 20% did not result in increased energy consumption by the heart at basal levels [53]. However, when those animals were subjected to pressure overload hypertrophy, SERCA2a over-expressing TG mice show significant decreases in contractile force and free energy, leading to increased morbidity. The adaptive response involving SERCA2a upregulation and increased PLN Ser16 phosphorylation ensures that Tm dephosphorylated hearts remain compensated in a physiological state and do not progress to cardiomyopathy and heart failure under normal conditions. However, this increased SERCA2a expression and PLN phosphorylation most likely leads to cardiac dysfunction and pathology after TAC operation. Animals were euthanized 2 weeks after TAC but we speculate, based on the increase in hypertrophy in the TG TAC operated animals, that this group would exhibit increases in hypertrophic markers as well as increased lethality.

This is the first study indicating that dephosphorylating a sarcomeric protein can result in maintenance of a compensated or physiological hypertrophic phenotype. Additionally, to our knowledge, this is the first study in which a TG animal with alterations in a sarcomeric protein result in increases in SERCA2a protein expression and PLN Ser16 phosphorylation. Tm phosphorylation appears to be involved in the development of compensated or physiological hypertrophy, possibly through proteins involved in signaling at the z-disc. Novel protein kinase C (PKC)δ and PKCε are two molecules shown to promote physiological hypertrophy. Both molecules translocate to the z-disc upon cardiomyocyte stimulation

[181, 182]. PKCε, specifically, has been shown to associate with the myofilament and bind strongly to actin, resulting in constitutively active PKCε [183]. Previous studies indicate that dephosphorylated Tm binds actin differentially from phosphorylated Tm and it is possible that the replacement of Ser283 with an Ala residue is affecting nearest neighbor interactions in the sarcomere, allowing PKCε greater access

69 to the binding site on actin [118, 119]. Mice expressing a PKCε specific activator exhibit normal cardiac function and a compensated hypertrophic phenotype indicating that PKCε can be a positive modulator of compensatory cardiac hypertrophy [184]. Additionally, activated PKCε can activate MEK1 through

Raf1, which has been shown to also result in compensated hypertrophy indicating that the PKCε-Raf1-

MEK1-ERK1/2 pathway may be playing a role in the S283A TG mouse phenotype [185-187]. Studies examining the potential signaling pathways activated by Tm dephosphorylation are currently in progress.

CHAPTER 4: INVESTIGATION OF Tm PHOSPHORYLATION AND α-Tm S283A MICE

Striated Muscle Tm Ser283A Specific Antibody

Generation of a reliable phosphorylation-specific antibody targeting the sole known phosphorylation site of striated muscle α-Tm at Ser283 was considered of the utmost importance as other available methods of determining phosphorylation of Tm were lengthy, involved and not always reliable. Yenzyme generated two α-Tm Ser283 phospho-specific antibodies for our laboratory as detailed in the Materials and Methods section. The more specific antibody generated by Yenzyme was extensively tested against both recombinant α-Tm protein, myofibrillar preparations and myofibrillar preparations subjected to calf intestinal phosphatase (CIP) (New England Biolabs) treatment (Figure 17 A,B). There is no reaction to recombinant α-Tm, which, of course, is not phosphorylated as it is generated in bacteria. There is additionally no reaction with the CIP treated myofibrillar Tm; the antibody does bind to the untreated myofibrillar Tm, indicating specificity to phosphorylated Tm. Interestingly, the pTm283 antibody also reacts with phosphorylated striated muscle β and γ-Tm. Generation of this striated muscle Tm phosphorylation specific antibody will facilitate evaluation of Tm phosphorylation status in a variety of mouse models and experimental conditions.

Figure 17. Validating the Tm phosphorylation specific antibody. A. Western blots analysis of recombinant α-Tm and myofibrillar preparations probed with the pTm283 antibody. B. Western blot analysis of NTG, α-Tm S283A Line 2 and Line 25 myofibrillar preparations probed with sarcomeric actin as a loading control, total Tm, and pTm283 antibodies. Samples were either left untreated or treated overnight with calf intestinal phosphatase (CIP).

Phosphorylation of Human Cardiac Tm

Studying phosphorylation of cardiac specific Tm in murine models is relevant if human cardiac Tm is phosphorylated. To that end, human biopsy samples from the right ventricular (RV) free wall, the intraventricular septum (IVS) and the LV free wall were evaluated for the presence of Tm phosphorylation (Figure 18A). A number of normal and failing human heart biopsy samples were also evaluated in order to determine whether there were differences in Tm phosphorylation between normal and failing hearts (Figure 18B, 18C). Interestingly, human cardiac Tm is, in fact, phosphorylated.

However, the failing human heart samples did not show significant differences in Tm phosphorylated, which could be attributed to multiple causes. Tissue handling, such as length of time between biopsy and deep freezing, can lead to alterations in Tm phosphorylation. The types of cardiac disease, ongoing therapeutic treatment and age of the patients varied as well, which could also possibly impact Tm phosphorylation.

Figure 18. Phosphorylation of Tm in human heart samples. A. Western blot analysis of human heart biopsy homogenate from RV, IVS and LV probed with actin as a loading control, Total Tm and pTm. B. Western blot analysis of human heart biopsies from non-failing hearts (n=3) and failing hearts (n=5). Sarcomeric actin was probed as a loading control, Total Tm and pTm. C. Quantification of samples from panel B.

Tm Phosphorylation is Sensitive to Method of Euthanasia

As the kinase and phosphatase involved in striated muscle Tm phosphorylation are not currently known, it was imperative to study the effect of multiple methods of euthanasia on Tm phosphorylation in order to determine the most reliable method to use for study of Tm phosphorylation. Multiple methods were tested as described in the Materials and Methods section. Interestingly, Avertin and cervical dislocation

(CD) had the least effect on cardiac Tm phosphorylation while hearts from animals euthanized using isofluorane and CO2 increased Tm phosphorylation compared to Avertin and CD (Figure 19A-C). Thus, the use of CD followed by removal of the heart, was chosen as the appropriate method for study of Tm phosphorylation in our lab.

Figure 19. Evaluating the effect of method of euthanasia on Tm phosphorylation. A. Western blot analysis Tm phosphorylation in myofibers from animals sacrificed using Avertin or cervical dislocation. B. Western blot analysis Tm phosphorylation in myofibers from animals sacrificed using CO2 or isoflurane. Actin was used as a loading control. C. Quantification of A and B.

Early Studies on the α-Tm S283A Mice

Multiple TG lines were generated expressing the α-Tm S283A transgene (Figure 11A). Rapid characterization of these animals was necessary. Northern blot analyses indicated a range of transgenic

RNA expression levels in the 8 lines studied (Figure 20A-C) without a dramatic decrease in endogenous

α-Tm RNA expression. The lines were sorted into low expression, moderate expression and high expression groups based on the Northern blot analyses.

Figure 20. Northern blot analyses of 8 TG lines generated from the cardiac specific α-Tm S283A construct. A. Blots probed for α-Tm S283A TG mRNA, endogenous α-Tm mRNA and GAPDH mRNA. B. Quantification of α-Tm S283A TG mRNA normalized to GAPDH. C. Quantification of endogenous α-Tm expression normalized to GAPDH.

Three lines were further investigated: lines 2, 25 and 97. Of these lines, Line 2 had the highest TG mRNA expression, followed by Line 25 and Line 97. The α-Tm S283A TG mRNA expression is significantly different between each of these lines (Figure 21A, 21B). However, there are no significant differences in the endogenous α-Tm mRNA expression.

Figure 21. Northern blot analyses of α-Tm S283A Line 2, Line 25 and Line 97. A. α-Tm S283A TG mRNA expression normalized to GAPDH. B. Endogenous α-Tm mRNA expression normalized to GAPDH. n=3. * p<0.05, ** p<0.01.

Southern blots were performed on digested genomic DNA isolated from NTG and Line 2, Line 25 and

Line 97 α-Tm S283A TG animals. Interestingly, Line 25 showed a large number of copies, while Line 2 and Line 97 both had fewer copies than Line 25 (Figure 22A, 22B). Line 25 copy numbers do not correspond the TG mRNA expression level or the protein expression (Figure 11B, 11C, Figure 21A), suggesting that the transgene may have integrated into a less-transcriptionally active region of the DNA or integrated into multiple sites. Line 2 and Line 97 appear to have similar copy numbers.

Figure 22. Southern blot analysis of α-Tm S283A TG lines. A. Southern blot of EcoRI digested genomic DNA radiolabeled with a probe directed against the α-Tm 3’UTR. B. Quantification of samples from panel A. n=3.

Additional Studies of Line 2 α-Tm S283A TG Animals

Many studies were performed on Line 2 α-Tm S283A TG animals. These animals were chosen for primary study in this project given the high level of protein expression and the lack of appreciable differences in phenotype between Line 2, 25 and 97. In addition to HW:BW ratios, ventricle weight to body weight (VW:BW) and LV chamber weight (LVCW:BW) ratios were determined (Figure 23A, 23B).

There were no significant differences in these parameters. Additionally, there were no significant differences in body weight between NTG and TG mice, meaning that increased body weight was not masking the alterations in the above measured parameters (Figure 23C). Lungs from 4 month and 6 month NTG and TG littermates were isolated and weighed under both wet and dry conditions to determine if pulmonary edema, a hallmark of heart failure, was present in these animals. There was no pulmonary edema present (Figure 23D). Finally, livers were isolated at 4 and 6 months of age to determine whether there was evidence of an influx of inflammatory cells into the liver, which can occur in some cases of heart failure. There were no significant differences between NTG and TG animals at either time point, indicating a lack of inflammatory cell migration to the liver (Figure 23E).

Figure 23. Gravimetric analysis of α-Tm S283A Line 2 mice. A. Ventricle weight to body weight ratio. B. LV chamber weight to body weight ratio. C. Gross body weight. D. Ratio of wet to dry lung weight at 4 months and 6 months. E. Ratio of wet to dry liver weight at 4 and 6 months.

Gene expression changes were studied in Line 2 α-Tm TG animals (Figure 15). Of particular interest was the increase in MCIP1, a protein that has been shown to have an effect on CnA activity. Although studies on CnA, also known as protein phosphatase 2b (PP2b) showed that there were no significant differences in phosphate release between NTG and Line 2 α-Tm TG mice, the assay also allowed tandem measurement of the activity of protein phosphatase 1 (PP1) and protein phosphatase 2a (PP2a).

Interestingly, there was a significant increase in phosphate released in TG animals when both of these phosphatases are measured together. This indicates that PP1 or PP2a may be playing some role in the pathway involved in linking α-Tm dephosphorylation to compensated cardiomyopathy and the increase in

SERCA2a expression and PLN Ser16 phosphorylation (Figure 24).

Figure 24. Phosphate release in NTG and α-Tm S283A TG heart homogenate by both PP1 and PP2a. n=3. *p<0.05.

As blood pressure (BP) can influence the size of the left ventricle, and the predominant changes at basal level in the α-Tm S283A TG animals are on the size of the left ventricle and LV cardiomyocytes (Figure

13, Table 10), we measured systolic blood pressure in male NTG and α-Tm S283A TG animals between

12-16 weeks of age. There were no significant differences in systolic blood pressure between groups studied (Figure 25).

Figure 25. Systolic blood pressure in 12-16 week old male NTG and TG animals. NTG n=7, TG n=9.

Careful examination was made of the apex of the heart, the LV free wall and the whole heart in control and TAC operated hearts (Figure 26A-C). There do not appear to be significant differences between

NTG sham and TG sham operated hearts with regard to fibrosis. There also did not appear to be a significant visual differences in deposition of fibrotic tissue in NTG TAC and TG TAC operated hearts.

Figure 26. Masson’s Trichrome staining of whole heart sections from control and TAC operated hearts. A. Apex of the heart. B. LV free wall of the heart. C. Whole heart images. All sections are stained with Masson’s Trichrome. i. NTG sham hearts. ii. TG sham hearts. iii. NTG TAC hearts. iv. TG TAC hearts.

The Mystery of α-Tm S283A Line 25

As seen in the early studies of the α-Tm S283A TG animals, Line 25 was unusual because it had a large number of transgene copies of the transgene but exhibited moderate TG mRNA expression and lower TG protein expression than Line 2 (Figure 22B, 21A). Also, a subset of line 25 mice also exhibited an extreme phenotype. About 10% of Line 25 animals began exhibiting physical evidence of disease by one month of age; lack of grooming, small size and failure to thrive. The remaining 90% of Line 25 TG animals did not exhibit these characteristics. The Line 25 animals exhibiting the extreme phenotype were called extreme Line 25 and the typical Line 25 animals were labeled typical Line 25. Typical Line 25 animals did not exhibit a significantly different life span from their NTG littermates but the extreme Line

25 animals died by three months of age (Figure 27).

Figure 27. Kaplan Meier Survival Curve for NTG, typical Line 25 and extreme Line 25 TG animals.

When the hearts were removed from of NTG, typical Line 25 and extreme Line 25 TG animals, there were some striking differences. Compared to NTG hearts, those from extreme Line 25 TG animals had hypertrophied left atria with visible calcifications (Figure 28Ai-iv). Whole heart sections stained with

Masson’s Trichrome showed that the typical Line 25 TG hearts had slightly enlarged atria (Figure 28Bi,

28Bii). However, the extreme Line 25 TG heart had a massive increase in atrial size. Note the section shown in Figure 28Biii is a composite as the dissecting microscope could not capture the entirety of the heart. The left ventricle shows dilation and the left atria shows excessive deposition of fibrotic tissue

(Figure 28Biii). Sections of the LV free wall show that typical line 25 cardiomyocytes seem to be enlarged, compared to NTG littermates, but do not seem to exhibit the degree of myocyte disarray seen in extreme Line 25 TG LV freewall (Figure 28Ci-Ciii).

Figure 28. Whole hearts and sections from Line 25 typical and extreme and control mice. A. i, NTG whole heart. ii, extreme Line25 whole heart. iii anterior view of extreme Line 25 left atria. iv posterior view of extreme Line 25 left atria. B. Masson’s Trichrome stained whole heart sections. i, NTG whole heart. ii, Typical Line 25 TG heart. iii, Extreme Line 25 TG heart. This image is a composite of two images. C. H&E staining of LV tissue sections. i, NTG. ii, typical Line 25 TG. iii, extreme Line 25 TG. A&B, taken on a Nikon SM2-ZT dissecting microscope. C. taken on an Olympus BX40 compound microscope at 40X objective, scale bar indicates 50µm.

Gravimetrics were performed on the typical Line 25 and extreme Line 25 TG animals. As expected, given the histological analysis of the hearts seen in Figure 28, the extreme Line 25 TG animals exhibit an increase in HW:BW ratio compared to both NTG littermates and typical Line 25 TG animals (Figure

29A). NTG and typical Line 25 TG animals do not exhibit any significant differences in HW:BW ratios.

NTG, typical Line 25 and extreme Line 25 mice were followed for 5 weeks and their gross body weights were measured as, early on, the extreme TG animals were noted to be smaller and failed to thrive. As expected, the extreme TG animals were smaller than both NTG and typical Line 25 TG animals throughout the period during which the animals survived (Figure 29B). There was no significant difference in body weight between NTG and typical Line 25 TG animals.

Figure 29. Gravimetrics of NTG, Line 25 typical and Line 25 extreme mice. A, Heart weight to body weight ratios of NTG, α-Tm S283A Line 25 Typical and α-Tm S283A Line 25 Extreme TG hearts at 3 months of age. B, Body weights of NTG, α-Tm S283A Line 25 Typical and α-Tm S283A Line 25 Extreme TG hearts. * p<0.05, ** p<0.01.

The extremely large size of the left atria of the extreme Line 25 TG animals, the high copy number and the knowledge that the α-MHC promoter is on earlier in the atria than it is in the ventricles suggested the possibility that there may be alterations in the expression of α-Tm S283A transgenic mRNA between typical Line 25 and extreme Line 25 in either the ventricles or atria. Real Time RT-PCR was utilized to determine the amounts of TG mRNA in the atria and ventricles of these animals. Surprisingly, there was no significant difference between typical Line 25 and extreme Line 25 TG mRNA expression in either region of the heart (Figure 30A, 30B). The extreme Line 25 TG animals were eventually bred out of the

TG line, as they themselves did not reproduce and the F1 breeding pairs producing the extreme phenotype animals died. As the discrepancies between typical and extreme Line 25 could not be easily attributable to any parameters studied, it is likely that the extremely large number of copies may have disrupted some crucial gene that resulted in a more severe phenotype in the extreme Line 25 animals.

Figure 30. Real Time RT-PCR analysis of typical and extreme Line 25. A. RT-PCR analysis of α- Tm S283A transgene expression in the atria of typical Line 25 and extreme TG animals. B. RT-PCR analysis of α-Tm S283A transgene expression in the ventricles of typical Line 25 and extreme TG animals. n=6.

Conclusions

The Tm phosphorylation specific antibody generated for the sole phosphorylation site on striated muscle

α-Tm is an extremely useful tool for the rapid detection and quantification of Tm phosphorylation status.

Validating the effect of method of sacrifice was also important in this study of Tm phosphorylation status.

Utilizing a method that resulted in altered levels of Tm phosphorylation could have masked alterations in phosphorylation occurring in response to mutation, disease, stress and could have rendered future studies of Tm phosphorylation inexplicable.

Verifying that human cardiac Tm is phosphorylated serves the purpose of allowing scientists to place, in context, the studies performed in animals. Even though the study of heart failure patients did not show significant differences in Tm phosphorylation, it is possible that more careful collection, categorization and study of specific human disease tissue explants might show significant differences in Tm phosphorylation, and ultimately determine some linkage of Tm phosphorylation state to disease state.

Generating the cardiac specific α-Tm S283A non-phosphorylatable TG animals and studying lines with various levels of mRNA and protein expression permitted careful selection of a TG line expressing a high level of TG protein, and therefore a significant decrease in Tm phosphorylation. There weren’t significant differences in the phenotype of Line 2, Line 25 (the typical animals) and Line 97, which supports our choice to focus on Line 2 since neither variations in copy number nor protein toxicity appear to effect the phenotype.

Interestingly, similar to many of the studies included in Chapter 3, Line 2 TG animals do not differ significantly from NTG littermates in VW:HW, LVCW:HW, wet/dry lung and wet/dry liver ratios. These animals also do not differ significantly in gross body weight indicating, in fact, that these TG animals exhibit a very mild hypertrophic phenotype. The NTG sham and TG sham operated animals do not differ significantly in the apex, LV free wall or whole heart in terms of fibrotic deposition. This is also

87 true of the NTG TAC and TG TAC operated animals. The increase in size of the left ventricle is uniform and not confined or centralized to a specific area.

Of interest, however, is the fact that there are increases in PP1 and PP2a activity measured together in TG hearts compared to NTG hearts. In the future, it would be ideal if PP1 and PP2a activity could be measured separately. Further study of these phosphatases may allow facilitate elucidation of the link between Tm dephosphorylation and the maintenance of a compensated hypertrophic phenotype with possible increases in SERCA2a activity.

α-Tm S283A Line 25 is unusual in its phenotype. Although the altered gene(s) causing the extreme phenotype seems to have been bred out of the line, it is interesting that 10% of animals had an extreme phenotype. It has been speculated that concatemerized copies of the transgene may have inserted into more than one place in the genome, possibly into a less transcriptionally active area. Additionally, although inbred mouse lines are supposed to be almost completely identical, it’s possible that some animals bore a gene variant or mediator that resulted in a more severe phenotype, typical to what is seen in human beings who sometimes have variable outcomes with the same genetic mutations. Finally, it is possible that the extreme Line 25 TG mice were extremely susceptible to stress or were low in the social structure of mice, which, given the huge number of transgene copies, could possibly have made the animals vulnerable to stress induced disease. The studies reported above do not conclusively point to any of these outcomes and, as stated earlier, the extreme phenotype has been bred out of the line. However, it would have been edifying to have been able to identify the contributors to this differential phenotype within the same TG line.

CHAPTER 5: TROPOMYOSIN DEPHOSPHORYLATION RESCUES TROPOMYOSIN

INDUCED FAMILIAL HYPERTROPHIC CARDIOMYOPATHY

Emily M. Schulz, Tanganyika Wilder, Shamim A. Chowdhury, Beata M. Wolska, R. John Solaro, David F. Wieczorek

Summary

Recent studies have indicated that tropomyosin (Tm) phosphorylation status can vary in different mouse models of cardiac disease. Investigation of Tm phosphorylation in basal and acute cardiac function utilizing a mouse model expressing a cardiac specific α-Tm protein that cannot be phosphorylated

(Ser283Ala) shows a compensated hypertrophic phenotype with significant increases in the expression of

SERCA2a and phosphorylation of phospholamban (PLN) Ser16 (Schulz et al, in press). Upon acute, extrinsic stress, (transaortic constriction (TAC)), α-Tm S283A TG hearts fail sooner than their NTG littermates, possibly due to the increased energetic demand placed on the failing heart by increased

SERCA2a activity.

Given the upregulation of SERCA2a and the increase in PLN Ser16 phosphorylation in the α-TmS283A

TG animals, we hypothesized that α-Tm phosphorylation may be beneficial in the context of a chronic, intrinsic stressor. To test this hypothesis, we exploited the fact that an α-Tm 180 (Glu180Gly) mutation is responsible for the development of a severe FHC phenotype in humans and mice [48, 49, 127, 128].

These FHC hearts are characterized by increases in heart weight:body weight ratios, increases in left atrial size, excessive deposition of fibrotic tissue, increased Ca2+ sensitivity of the myofilament and contractile defects. The FHC α-Tm 180 mice often die by 6-8 months of age. By examining mice expressing both the E180G mutation and the S283A mutation on the same molecule, we surprisingly found the α-Tm 180 phenotype is corrected. α-Tm 180-S283A double mutant transgenic (DMTG) animals exhibit no signs of hypertrophy, measured by biochemical markers or physiological criteria. The DMTG animals have increased phosphorylation of PLN Ser16 and Thr17 compared to the α-Tm 180 mice. DMTG mice have

Ca2+ sensitivity of the myofilament similar to NTG animals and echocardiographic studies show a rescue in all parameters in which α-Tm 180 animals show dysfunction. Moreover, the DMTG hearts are hypercontractile. Thus, the inability to phosphorylate S283 plays a major role in the rescue of hypertrophic cardiomyopathy caused by the E180G mutation in α-Tm.

Introduction

Tropomyosin (Tm) is an α-helical coiled-coil protein involved in the Ca2+ dependent regulation of the thin filament of the sarcomere. Once Ca2+ binds to troponin C, a conformational change occurs, allowing the

Tm filament to move away from the myosin-head binding site on the sarcomeric actin filament, resulting in muscle contraction. α-Tm is the predominant isoform found in cardiovascular muscle, making up approximately 95% of total myofibrillar Tm [21]. All striated muscle isoforms, including α-Tm, are phosphorylated at a single site, serine 283 by several potential kinases [110, 111, 116, 117, 148, 150].

Recent studies indicated that mice expressing transgenic (TG) α-Tm with Ser283 mutated to an alanine

(S283A) show no major alterations in cardiac function at basal levels with striking increases in the expression of the SERCA2a and an increase in phospholamban (PLN) Ser16 phosphorylation (Schulz et al, in press). Additionally, these mice maintain a compensated hypertrophic phenotype throughout their lifetime.

Previous studies demonstrate that Tm mutations that lead to cardiac disease show changes in Tm phosphorylation [66, 113, 114]. It is well known that increasing SERCA2a expression and/or activity by decreasing PLN expression or increasing PLN phosphorylation at Ser16 and/or Thr17 can rescue not only

FHC or HF in mice. In addition, SERCA2a gene therapy has been shown to be efficacious in human patients [38, 51, 66, 188]. The α-Tm S283A hearts exhibit a compensated hypertrophic phenotype with many characteristics that closely mimic changes seen in exercise trained animals. We hypothesized that the S283A mutation may mitigate the well-characterized and severe pathology found in human and mouse hearts carrying the E180G mutation. That is, the S283A mutation might result in improved cardiac function compared to mice expressing the α-Tm 180 mutation alone. To that end, we generated DMTG mice expressing Tm with the α-Tm E180G and S283A mutations on the same molecule; several TG mouse lines expressing these mutations were generated and analyzed. Interestingly, the DMTG hearts do not show the dramatic increase in SERCA2a expression seen in the α-Tm S283A hearts. DMTG animals have no change in heart weight: body weight ratios and no deposition of fibrotic material characteristic of

91 the α-Tm 180 phenotype. Echocardiographic studies indicate that DMTG animals rescue cardiac function with improved performance over NTG mice. This is the first study investigating the role of Tm phosphorylation on the development of cardiac disease caused by sarcomeric mutation.

Results

Generation of α-Tm 180-S283A DMTG Mice

To determine the role of α-Tm phosphorylation in the context of a progressive, genetic cardiac disease, we generated a DMTG mouse model that expressed a non-phosphorylatable alanine at aa 283, the location of the sole known phosphorylation site in striated muscle Tm [110, 111, 150]. On the same construct, we introduced the α-Tm E180G FHC mutation [48]. The transgene construct used to generate these DMTG mice is shown in Figure 31A. Four DMTG lines were generated and studied.

Figure 31. Characterizing the α-Tm 180-S283A DMTG animals. A. α-Tm 180-S283A DMTG construct. The α-MHC promoter drives cardiac specific expression of the striated muscle α-Tm with encoded substitutions at amino acid 180 (E180G) and 283 (S283A) on the same molecule. B. Myofibrillar preparations resolved on an SDS-Page gel and stained with Coomassie. Note the shift in migration in the 180 and DMTG samples. Arrow indicates DMTG endogenous and TG protein. C. Quantification of endogenous protein remaining in NTG, α-Tm 180, α-Tm S283A and DMTG samples.

Cardiac α-Tm 180-S283A DMTG Protein Expression

DMTG protein expression was determined as percent of total Tm protein expression in all four DMTG lines. These lines had a range of DMTG 50-64% protein expression of total Tm protein. Line 325 and

Line 335 (64 and 50% TG protein expression respectively) did not differ in their the morphological and physiological results, for that reason, we focused on Line 325. As seen in Figure 31B, the 180 mutation confers differential migration to the DMTG protein when visualized by SDS-PAGE which allows quantification of endogenous and TG protein levels (Figure 31C). Of particular interest is a feedback mechanism present in cardiac expressed striated muscle Tms which shows the endogenous Tm protein level decreases concomitantly with increasing DMTG Tm protein expression, with no changes in total Tm protein [22, 25]. Examination of the cytoplasmic preparations from these DMTG mice show no significant accumulation of DMTG protein in the cytoplasm, indicating that the DMTG protein is being properly incorporated into the sarcomere (data not shown).

Phosphorylation Status of α-Tm 180 and α-Tm 180-S283A DMTG Mice

Utilizing a Tm phosphorylation specific antibody (Schulz et al, in press), we investigated both the phosphorylation status of the α-Tm 180 and DMTG cardiac myofibers. Interestingly, at 3 months of age,

α-Tm 180 hearts show a significant increase in Tm phosphorylation, whereas Line 325 DMTG hearts exhibit a significant decrease in phosphorylation (Figure 32). Additionally, the phosphorylation status of the DMTG hearts is very similar to the phosphorylation status of the α-Tm S283A mouse hearts.

Figure 32. Phosphorylation status of Tm in NTG, α-Tm 180, α-TM S283A and α-Tm 180-S283A DMTG hearts. A. Immunoblot of phosphorylation status of α-Tm 180 hearts at 3 months of age. B. Immunoblot of phosphorylation status of α-Tm S283A hearts at 3 months of age. C. Immunoblot of phosphorylation of DMTG hearts at 3 months of age. D. Quantification of A-C, n=6. * p<0.05.

Histopathological, Gravimetric and Cardiomyocyte Cross-Sectional Area Analyses of α-Tm 180 and α-

Tm 180-S283A DMTG Mice

Morphological analysis of the LV free wall of DMTG animals at 3 months of age shows a phenotype very similar to age-matched NTG mice with no cardiomyocyte disarray, enlargement and excessive fibrosis found in α-Tm 180 hearts (Figure 33A). Wheat germ agglutinin (WGA) staining of cell the membrane of the cardiomyocytes show a significant increase in the cardiomyocyte cross-sectional area of the α-Tm 180 and α-Tm S283A cardiomyocytes compared to both the NTG and DMTG (Figure 33B).

There is no significant difference in cardiomyocyte cross-sectional area between NTG and DMTG cells.

Interestingly, there is no increase in heart weight to body weight (HW:BW) ratios between NTG and

DMTG animals (Figure 33C). There are also no differences in HW:BW ratios in α-Tm S283A animals compared to both NTG and DMTG. In contrast, α-Tm 180 animals exhibit significant increases in

HW:BW ratios compared to NTG littermates [48-50, 134]. The presence of the S283A mutation is likely responsible for normal cardiac size in the DMTG animals. Additionally, DMTG animals maintain the same HW:BW ratios as NTG littermates for at least 13 months of age. Importantly, while α-Tm 180 animals do not typically survive past 6-8 months of age [50], the addition of the S283A mutation considerably extends the life expectancy of the these mice to that of NTG animals (data not shown).

Figure 33. Histology and gravimetrics. A. Histological studies of NTG, α-Tm 180 and DMTG hearts at 3 months of age stained with hemotoxylin and eosin, Masson’s trichrome and wheat germ agglutinin (WGA). Images taken at 40X and scale bar indicates 50 µm. B. Cardiomyocyte cross-sectional area measurements n=6. ** p<0.01, *** p<0.001. C. Heart weight to body weight ratios of 3 month old mice, n=6

Cardiac Function in α-Tm 180 and α-Tm 180-S283A DMTG animals

To assess whether decreasing phosphorylation of α-Tm in a genetic model of FHC alters cardiac function, we performed echocardiographic analysis on 3 month old NTG, α-Tm 180 and DMTG mice (Table 11).

Interestingly, the DMTG mice do not exhibit alterations in size parameters previously reported in the α-

Tm S283A animals (Schulz et al, in press). DMTG mice do show a slight increase in the size of the left atria; however, the increase in size is not as large as the increase in atrial size typically found in α-Tm 180 animals [48, 49, 134]. Most striking is the significant increase in ejection fraction (EF), velocity of circumferential fiber shortening (Vcf) and percent fractional shortening (FS%) in the DMTG animals compared to both the NTG and α-Tm 180 animals, indicating that the DMTG hearts have improved function and are hypercontractile compared to NTG littermates. Additionally, when parameters measuring diastolic function (E/Em, E/A ratio) are examined, DMTG animals show rescue of function, similar to NTG levels, from the extreme diastolic dysfunction seen in the α-Tm 180 animals (Table 11).

In short, the DMTG animals rescue the dysfunctional cardiac phenotype seen in the α-Tm 180 FHC model and actually improve functional parameters involved in cardiac contraction.

Table 12. Cardiac function of NTG, α-Tm 180 and α-Tm E180G-S283A DMTG at 3 months of age as assessed by echocardiography Parameters NTG n=9 α-Tm 180 n=5 α-Tm 180-S283A n=13 Left atrium 1.93 ± 0.09 3.88 ± 0.23*** 2.16 ± 0.13*** LVIDd (mm) 4.08 ± 0.13 3.97 ± 0.10 3.90 ± 0.06 LV mass (mg) 92.39 ± 5.78 95.29 ± 10.84 85.21 ± 2.56 EF (ejection fraction) 68.96 ± 1.37 66.08 ± 4.38 80.52 ± 1.81***## FS%(fractional 38.41 ± 1.09 36.33 ± 3.14 49.33 ± 1.92***## shortening) Vcf 6.99 ± 0.30 6.33 ± 0.85 8.13 ± 0.26**# IVRT 12.97 ± 0.38 13.13 ± 1.38 13.73 ± 0.57 DT 22.10 ± 1.59 26.04 ± 0.84 24.2 ± 0.77 e/a Ratio 1.67 ± 0.14 6.72 ± 1.03*** 2.07 ± 0.14### E/Em Ratio 38.54 ± 3.67 64.6 ± 3.17** 38.42 ± 1.95### LV (Left Ventricle), LVIDd (LV Internal Dimension in diastole), EF (Ejection Fraction), FS (Fractional Shortening) , Vcf (Velocity of circumferential shortening), IVRT (Iso-Volumetric Relaxation Time), DT (Deceleration Time), E/A ratio (ratio of Early-to-late ventricular filling velocities), E/Em ratio (ratio of Early Pulse-Doppler filling velocity to Early Tissue Doppler velocity). * vs. NTG , p<0.05; # vs. α-Tm 180, *p<0.05, **p<0.01, ***p<0.001.

To determine whether the relationship between Ca2+ concentration and force-tension development is altered in myofilaments of DMTG mice, we analyzed skinned, detergent extracted fiber bundles from the papillary muscle of 3 month old hearts. The DMTG animals show an intermediate phenotype compared to both the NTG and α-Tm 180 myofibers when tension and percent maximum force are measured

(Figure 34). DMTG myofibers exhibit a pCa50 of 5.940 ± 0.01, versus NTG myofibers (5.807 ± 0.02) and the α-Tm 180 myofibers (6.012 ± 0.03) (Table 12). The Hill Coefficient (nH) of DMTG myofibers

(3.44 ± 0.34) is not significantly lower from that of NTG myofibers (3.58 ± 0.47), while the α-Tm 180 nH is significantly different than both the NTG and DMTG myofibers at 2.65 ± 0.42. The cooperative activation of the thin filament is significantly impacted in α-Tm 180 TG myofibers, as has been previously reported ; however, sarcomeric dysfunction is rescued to normal levels in DMTG myofibers.

Figure 34. Parameters involved in Ca2+ tension relations. A. Ca2+ tension relations in skinned fiber bundles from NTG, α-Tm 180 and DMTG hearts. B. Ca2+ force relations in skinned fiber bundles in NTG, α-Tm 180 and DMTG hearts.

Table 13. Parameters involved in Ca2+ tension relations Group pCa50 nH n NTG 5.807 ± 0.02 3.58 ± 0.47 12 α-Tm E180G 6.012 ± 0.03 2.65 ± 0.42 5 α-Tm E180G-S283A 5.940 ± 0.01 3.44 ± 0.34 10

Gene Expression Changes in α-Tm 180 and α-Tm 180-S283A DMTG Hearts

Given the apparent improvement in cardiac function, Ca2+ sensitivity and cooperativity in the DMTG hearts, it was of interest to determine whether the rescued hearts also show improvements at the molecular level by assessing of gene expression of cardiomyopathy markers. Real time RT-PCR analysis of RNA isolated from ventricular tissue shows that α-Tm 180 mice exhibit significant increases in β myosin heavy chain (βMHC), brain natriuretic peptide (BNP) and atrial natriuretic peptide (ANP), as expected given the level of hypertrophy present in the hearts of those animals (Figure 35A-C).

Interestingly, DMTG hearts show a significant decrease in βMHC compared to both NTG and α-Tm 180 hearts. Moreover, DMTG hearts exhibit similar levels of ANP and BNP in comparison to NTG hearts and significant decreases when compared to α-Tm 180 hearts while the α-Tm S283A hearts do not significantly differ from normal.

100

Figure 35. Real Time RT-PCR analysis of cardiomyopathy markers. A. qRT-PCR analysis of expression of the cardiomyopathy marker β-MHC in 3 month old NTG, α-Tm 180, α-Tm S283A and DMTG hearts. B. qRT-PCR analysis of expression of the cardiomyopathy marker BNP in 3 month old NTG, α-Tm 180, α-Tm S283A and DMTG hearts. C. qRT-PCR analysis of expression of the cardiomyopathy marker ANP in 3 month old NTG, α-Tm 180, α-Tm S283A and DMTG hearts. NTG n:12, 180 n=8, DMTG n=10. * vs NTG, # vs α-Tm 180, @ vs α-Tm S283A. * p<0.05, **p<0.01, ***p<0.001.

Changes in Proteins Involved in Ca2+ Handling in α-Tm 180 and α-Tm 180-S283A DMTG Hearts

Proteins involved in Ca2+ fluxes were evaluated in NTG, α-Tm 180, α-Tm S283A and DMTG hearts

(Figure 36A). Significant decreases in phosphorylated PLN were seen in the α-Tm 180 and DMTG animals compared to NTG controls. However, the DMTG samples show significant increases in phosphorylation at both Ser16 and Thr17 compared to α-Tm 180 samples (Figure 36B). The α-Tm

S283A animals exhibit a significant increase in PLN phosphorylation at Ser16 as previously reported

(Schulz et al, in press). Somewhat surprisingly, there was no increase in SERCA2a expression in the

101

DMTG animals, unlike that seen in the α-Tm S283A hearts. There was an increase in total PLN expression in both the α-Tm 180 and DMTG hearts (Figure 36C). With the increase in PLN phosphorylation, it is possible that there is a relief of the inhibition by PLN on SERCA2a, thus increasing

Ca2+ resequestration into the SR and helping rescue the E180G phenotype.

We also examined phosphorylation of troponin I (TnI), as this protein can be considered the master regulator of sarcomeric function and Ca2+ sensitivity. There are no significant differences in troponin I

(TnI) expression or TnI phosphorylation at aa 23 and 24; although there is an upward trend in TnI23/24 phosphorylation in the DMTG mice (Figure 36D, 36E). The absence of alterations indicates that the return of DMTG myofiber Ca2+ sensitivity to NTG levels is not mediated by TnI phosphorylation, but occurs through some other means.

102

Figure 36. Alterations in Ca2+ flux proteins. A. Western blots of sarcoplasmic Ca2+ flux proteins in NTG, α-Tm 180, α-Tm S283A and DMTG hearts. B. Quantification of phosphorylation levels of PLN Ser16 and PLN Thr17 from A. C. Quantification of expression of PLN and SERCA2a expression from A. D. Western blots of sarcomeric Ca2+ flux proteins in NTG, α-Tm 180, α-Tm S283A and DMTG hearts. E. Quantitation of D. * vs NTG, # vs α-Tm 180, @ vs α-Tm S283A. *p<0.05, **p<0.01, ***p<0.001. n=8.

103

Discussion

In recent years, it has become increasingly obvious that post-translational modifications of sarcomeric and z-disc proteins can lead to altered cardiac function. Our previous study (Schulz et al, in press) investigating the role of Tm dephosphorylation during basal cardiac function showed that α-Tm S283A

TG mice exhibit a mild, compensated hypertrophic phenotype with a striking increase in SERCA2a protein expression and PLN Ser16 phosphorylation. However, upon extrinsic, acute stimulus to induce pressure-overload via TAC, the α-Tm S283A hearts show greater cardiac dysfunction than NTG littermates, likely because of the increased energetic demand placed on the heart by the increase in

SERCA2a expression [53]. We hypothesized that dephosphorylation of α-Tm in the context of a chronic, intrinsic stressor at the level of the sarcomere may result in a more beneficial effect as studies have indicated that phosphorylation of Tm in the α-Tm 175 FHC model increases compared to NTG [114].

To test this hypothesis, we generated a cardiac specific Tm construct expressing both the α-Tm E180G mutation responsible for the development of a severe form of FHC and the mutation responsible for preventing Tm phosphorylation S283A [48] (Schulz et al, in press). The α-Tm 180 animals exhibit increased Ca2+ sensitivity, significant increases in heart weight to body weight ratios, severe cardiac dysfunction and typically only survive between 6-8months of age [49, 50]. We were interested in the phosphorylation status of these Tm 180 mice. Contrary to a DCM model which shows decreased Tm phosphorylation, the α-Tm 180 animals exhibit a significant increase in phosphorylation, suggesting that

α-Tm phosphorylation may vary in response to the mechanism of cardiomyopathic growth [113].

Four TG lines were generated expressing both the E180G and S283A mutations on the same α-Tm molecule. All TG lines showed alterations in α-Tm phosphorylation without significant differences in cardiac phenotype. Double mutant transgenic (DMTG) Line 325 was chosen for further study given the similarity in phosphorylation status of the α-Tm S283A animals. We have previously reported that α-Tm

180 TG animals have a shorter life span compared to NTG controls [48-50]. Interestingly, DMTG animals had a longer life span, measured out to 18 months of age, indicating a significant improvement in

104 overall cardiac health of the DMTG animals. While the DMTG hearts did not exhibit the striking increases in SERCA2a expression and PLN Ser16 phosphorylation evident in the α-Tm S283A mice, the

DMTG animals have significant increases in PLN phosphorylation at Ser16 and Thr17 compared to the α-

Tm 180 animals, suggesting that the DMTG mice may have more active SERCA2a. Investigation of cardiac function of NTG, α-Tm 180 and DMTG mice shows that the α-Tm 180 mice exhibit a severe contractile phenotype, specifically in relaxation parameters such as E/Em and e/a ratio, possibly due to alterations in SERCA2a activity and PLN expression and phosphorylation that can lead to a reduction in the rate of Ca2+ transit decay, known to contribute to a slower relaxation rate [189]. Interestingly, the introduction of the S283A dephosphorylation mutation returns E/Em and e/a ratio to normal or near normal levels, rescuing the relaxation defect found in α-Tm 180 animals, similar to that seen in TG mice expressing a chimeric α-/β-Tm protein and the α-Tm 180 protein [136]. The most striking result shown in the echocardiographic studies is the significant increase in ejection fraction (EF), velocity of circumferential fiber shortening (Vcf) and fractional shortening (%FS) in the DMTG animals compared to

NTG littermates, indicating that the DMTG hearts exhibit enhanced contractile parameters, possibly due to an increase in SERCA2a activity. The results of our skinned fiber study indicate that much of the rescue occurs at the level of the sarcomere. Because of the near normal Ca2+ sensitivity levels in DMTG mice compared to the NTG and α-Tm 180 animals, as well as the restoration of nH, indicating a rescue in thin filament cooperative activation, we hypothesize that the carboxy terminus (amino acids 258-284) of

Tm plays a crucial role in the rescue of the FHC phenotype.

Additionally, recent studies have indicated that the E180G mutation results in a strikingly more flexible striated muscle α-Tm compared to WT α-Tm [133]. Moreover, it has been suggested that the E180G mutation leads to an increase in local flexibility, likely partially unwinding or relaxing the coiled-coil around the 180th amino acid [131, 132]. The greater global flexibility of mutant α-Tm indicates that a lower concentration of Ca2+ is necessary to induce conformational changes needed to move Tm off the myosin head binding sites on actin, leading to increased Ca2+ sensitivity of the thin filament [190]. It is

105 possible that the S283A mutation at the carboxyl terminus restores α-Tm to WT levels of flexibility, which may account for the near complete rescue at the level of the sarcomere, although this possibility demands further study.

While we hypothesize that the rescue of the FHC phenotype in the DMTG animals involves a signaling process that occurs through nearest neighbor interactions involving actin and/or the troponin complex, or a restoration of Tm to proper flexibility, previous studies have indicated that normalization of Ca2+ flux dynamics and altered Ca2+ uptake by the SR may also be playing a role. While there is no significant increase in SERCA2a expression in the DMTG animals, unlike the S283A animals, there are alterations in the phosphorylation status of PLN, which may reduce the inhibition of SERCA2a by PLN and result in increased activity. It is also possible that the decrease in SERCA2a in DMTG animals compared to the

S283A animals may be related to the presence of the disease-causing E180G mutation, as previous studies have shown decreases in SERCA2a expression in the α-Tm 180 animals at different ages [49].

Indeed, previous studies altering proteins involved in Ca2+ fluxes in the context of the α-Tm 180 mice have shown that knocking out PLN in the 180 TG animals can rescue the severe cardiomyopathic phenotype. Similar to the DMTG animals studied here, PLNKO/Tm180 animals show reversals in hypertrophic marker genes, compared to the significant increase seen in α-Tm 180 animals. The

PLNKO/Tm180 animals also showed rescue of relaxation parameters by echocardiography. On the other hand, the PLNKO/Tm180 animals did not display the hypercontractile phenotype seen in the DMTG animals [66]. Additionally, rescue of the α-Tm 180 phenotype via adenoviral delivery of SERCA2a normalizes heart weight to body weight ratios and improves cardiac function, similar to our observations in the DMTG animals [51].

Despite multiple studies being conducted, the mechanisms by which α-Tm dephosphorylation improves function in the context of the α-Tm 180 mutation are not known. Gaffin et al [66] has shown that α-Tm

180 animals exhibit significant decreases in ERK 1/2 phosphorylation that are normalized when the

106 animals are crossed with PLNKO mice. ERK1/2 is one of the MAPK pathways involved in cardiac hypertrophy and MEK1, an upstream effector of ERK1/2 has been shown to result in increased levels of

ERK1/2 phsophorylation and a compensated hypertrophic phenotype, similar to what is seen in the α-Tm

S283A animals [186]. It is possible that the alterations occurring at the level of the sarcomere, specifically changes in Tm nearest neighbor interactions, may result in alterations in the activation of proteins such as protein kinase C ε (PKCε), which binds actin and is an upstream activator of the MEK1-

ERK1/2 pathway via c-Raf [181, 183, 185]. These questions require careful consideration and experimental design and are beyond the scope of this paper.

The pathways by which the α-Tm 180 mutation promotes a pathological phenotype have not been well elucidated, but this and other FHC models exhibit increased Ca2+ sensitivity and diastolic dysfunction, possibly leading to stress-sensitive pathways of hypertrophic growth. It has been suggested that decreasing myofilament sensitivity would be a straightforward approach to treating FHC although no current pharmacological interventions are available at this time. Targeting Tm phosphorylation specifically, given the rescue of the FHC phenotype shown in this study, may provide a starting point for research into possible therapeutics. To conclude, our data show that expression of a α-Tm protein expressing both the E180G FHC mutation and the dephosphorylating S283A mutation rescues the α-Tm

180 mediated hypertrophic phenotype. Changes in Ca2+ handling proteins may additionally be responsible for the hypercontractile phenotype found in the DMTG hearts. Changes in local flexibility of the Tm molecule conferred by the replacement of Ser283 with an Ala residue and the significant loss of phosphorylation may be responsible for the restoration of Tm to proper flexibility. Though alterations in actin-Tn-Tm interactions could play a vital role, the precise mechanisms whereby DMTG protein rescues hypertrophy remain to be further elucidated.

107

CHAPTER 6: α-Tm 180-S283A LINE 335 TG ANIMAL STUDIES AND α-Tm 180 CROSSED

WITH α-Tm S283A/D TG ANIMAL STUDIES

α-Tm 180-S283A Line 335 Double Mutant Transgenic Mouse Studies

α-Tm DMTG Line 335 TG Protein Expression and Phosphorylation Status

In addition to the studies performed on α-Tm 180-S283A double mutant transgenic (DMTG) Line 325 discussed in Chapter 5, extensive studies were also performed on α-Tm 180-S283A DMTG Line 335.

Line 335 had a lower expression of DMTG protein than did Line 325 and higher expression of endogenous α-Tm protein (Figure 37A, 37B). The doublet band of endogenous and DMTG protein is clearly visible in Line 335 TG animal cardiac myofibrillar preparations visualized on 10% SDS-PAGE gels stained with Coomassie Blue. All proteins found in the myofilament are present in each TG line in the proper ratios. Of additional interest is the fact that DMTG Line 335 expresses a similar amount of transgenic protein to the α-Tm 180 animals studied in Chapter 5 (Figure 32). In the context of a similar level of TG protein expression in the α-Tm 180 TG animals which results in a severe hypertrophic phenotype, DMTG Line 335 is a line that is very useful in the study of the effect of replacement of

Ser283 with an Ala on disease development and progression [48, 49].

The decrease in expression of TG protein in DMTG Line 335 hearts compared with DMTG Line 335 means that more endogenous α-Tm is available for phosphorylation. Interestingly, there is an increase in

Line 335 DMTG phosphorylation compared to both NTG and DMTG Line 325 animals (Figure 37C,

37D). The increase in Tm phosphorylation in DMTG Line 335 hearts is very similar to the increase in

Tm phosphorylation seen in the α-Tm 180 hearts (Figure 32), possibly because of the similar level of TG protein expressed bearing the E180G mutation.

108

Figure 37. Characterization of α-Tm 180-S283A DMTG Line 325 and Line 335. A. Myofibrillar preparations from NTG, DMTG Line 325 and DMTG Line 335 visualized on 10% SDS-PAGE gels stained with Coomassie Blue. B. Quantification of TG endogenous Tm protein expression from A. B. Immunoblot of actin and phosphorylated Tm from NTG, DMTG Line 325 and DMTG Line 335. D. Quantification of samples in C normalized to sarcomeric actin. * vs NTG, # vs DMTG Line 325. **p<0.01, ***p<0.001. n=3.

109

Gravimetrics and Cardiac Morphology of α-Tm DMTG Line 335 Hearts

Morphological analysis of the LV free wall shows a mild increase in cardiomyocyte disarray in DMTG

Line 335 heart sections compared to NTG and Line 325 (Figure 38Ai). Additionally, sections stained with Masson’s Trichrome seem to indicate an increase in the deposition of fibrotic tissue in the left ventricle (Figure 38Aii). WGA staining and measurement of cardiomyocyte cross-sectional area shows a small but significant increase in the size of the cardiomyocytes in DMTG Line 335 hearts compared to both NTG and DMTG Line 325 animals (Figure 38Aiii, 38B). While there is a trend toward an increase in

DMTG Line 335 in HW:BW ratios at 3 months of age, there is no significant difference between NTG,

DMTG Line 325 and DMTG Line 335 (Figure 38C), indicating that rescue of the α-Tm 180 phenotype is occurring due to the expression of the S283A mutation.

110

Figure 38. Histology and gravimetrics. A. Histology and morphology of NTG, DMTG Line 325 and DMTG Line 335 LV heart sections. i. H&E staining. ii. Masson’s Trichrome staining. iii. WGA staining. Scale bar indicates 50 µm. All images taken at 40X. B. Cardiomyocyte cross-sectional area measured using Aiii. C. HW:BW ratios. *p<0.05. NTG n=6, DMTG-Line 325 n=6, DMTG-Line 335 n=4.

111

Gene Expression Changes in α-Tm DMTG Line 335 Hearts

Given the interesting changes in mRNA expression of cardiomyocyte markers in the DMTG Line 325 hearts, compared to NTG and α-Tm 180, we also studied gene expression changes in DMTG Line 335.

Given the small increase in cardiomyocyte cross-sectional area coupled with the lack of increase in

HW:BW ratio, it was surprising to find that there were no significant increases in any cardiomyocyte markers measured (Figure 39). Although DMTG Line 325 hearts showed a significant decrease in β-

MHC expression compared to NTG expression, DMTG Line 335 expression was unchanged compared to

NTG, but increased in comparison to DMTG Line 325 (Figure 39A). DMTG Line 325 showed no significant alterations in ANP expression compared to NTG controls; however, Line 335 showed a significant decrease compared to both NTG and DMTG Line 325 (Figure 39B). Finally, there were no significant changes in BNP expression between NTG and DMTG Line 325, however, DMTG Line 335 showed significant decreases in ANP expression (Figure 39C).

112

Figure 39. Real Time RT-PCR analysis of cardiomyopathy markers. A. Real Time RT-PCR analysis of β-MHC expression. B. Real Time RT-PCR analysis of ANP expression. C. Real Time RT-PCR analysis of BNP expression. * vs NTG, # vs DMTG Line 325. *p<0.05, **p<0.01, ***p<0.001. DMTG Line 335 n=6.

113

Alterations in Expression and Phosphorylation of Proteins Involved in Ca2+ Fluxes in α-Tm DMTG

Line 335

There are significant changes in the expression and phosphorylation of sarcomeric proteins involved in

Ca2+ fluxes in both α-Tm S283A and DMTG Line 335 hearts. Given the differences in TG protein expression and Tm phosphorylation, examination of proteins involved in Ca2+ fluxes in DMTG Line 335 hearts were of interest. Interestingly, in α-Tm DMTG Line 335 hearts, there are no significant differences in PLN or SERCA2a expression (Figure 40A, 40B) compared to NTG control. There are no significant differences in the phosphorylation status of PLN either at Ser16 or Thr17 (Figure 40A, 40C). There are also no differences in the expression of TnI or the phosphorylation of TnI at Ser23/24 (Figure 40A, 40D).

114

Figure 40. Alterations in Ca2+ flux proteins. A. Western blot analysis of proteins involved in Ca2+ fluxes. B. Quantification of PLN and SERCA2a expression hearts from A in NTG and DMTG Line 335. C. Quantification of PLN phosphorylation at Ser16 and Thr17 hearts from A in NTG and DMTG Line 335 hearts. D. Quantification of total TnI expression and phosphorylation of TnI 23/24 in hearts from A in NTG and DMTG Line 335 hearts. n=4

115

Conclusions from Studies Performed on α-Tm 180-S283A DMTG Line 335 Hearts

Study of the α-Tm 180-S283A DMTG Line 335 mice yielded interesting results. The increase in phosphorylation of α-Tm in DMTG Line 335 compared to NTG and Line 325 samples seems to indicate that the remaining phosphorylatable Tm protein is being phosphorylated. This is interesting because these DMTG animals don’t exhibit any overt pathology, suggesting that perhaps the α-Tm phosphorylation occurs in a specific spatial or temporal pattern and that expressing a non- phosphorylatable Tm perturbs that pattern. When NTG tissue sections were stained with the pTm specific antibody and observed at 100X objective, the pTm appeared to be evenly distributed throughout the sarcomere. Perhaps the increase in Tm phosphorylation seen in the Line 335 mice is an attempt to evenly distribute the remaining phosphorylation.

While the Line 335 DMTG animals don’t show increases in HW:BW ratios, staining with Texas Red conjugated WGA (Figure 38) indicates a mild increase in cardiomyocyte cross-sectional area. These data seem to indicate that the rescue of the Line 335 DMTG animals from the α-Tm 180 phenotype is perhaps not as complete as the rescue found in Line 325. Additionally, I speculate that the Ca2+ sensitivity of the myofilaments and nH would not return to NTG levels. I hypothesize that the rescue at the level of the sarcomere depends on a conformational change that the mutation from Ser283 to Ala confers on the local and global flexibility of the Tm molecule; the lower level of TG protein expression in Line 335 DMTG would not confer the same overwhelming change in the thin filament.

Of particular interest in this DMTG line are the alterations in gene expression of common cardiomyocyte markers. There were significant decreases in ANP and BNP mRNA expression in Line 335 hearts, which is not seen in the Line 325 DMTG hearts. Conversely, Line 325 showed a significant decrease in β-MHC mRNA expression while Line 335 shows no significant change from NTG. These data, coupled with the histological analysis indicate that there is a significant degree of rescue from the α-Tm 180 phenotype occurring in the DMTG Line 335 mice as well as the DMTG Line 325 mice. It would be of interest to determine whether ANP and BNP protein expression was altered in the Line 335 or whether the decrease

116 was only seen at the mRNA level. There is conflicting data with regard to the role of ANP and BNP protein in the maintenance, modulation and progression of heart failure and hypertrophy. However, knocking out ANP and BNP protein seems to result in a greater degree of hypertrophy [191, 192] rather than a rescue, as seen in the DMTG animals.

Of particular interest is the result indicating no changes in any proteins involved in Ca2+ fluxes in the

DMTG Line 335 hearts compared to NTG, indicating rescue of PLN phosphorylation levels to NTG in this line of DMTG animals. Both α-Tm S283A and DMTG Line 325 hearts showed significant differences in the expression and phosphorylation of these proteins and it was surprising not to see similar differences in the DMTG Line 335 hearts. It is, however, possible that a certain threshold of expression must be reached, similar to what was seen in studies of β-Tm over-expressing mice [22, 27, 28]. At a certain percent expression of TG protein, the β-Tm TG animals sickened and died quickly from heart failure, unlike a lower percent expression line of the same animals which showed interesting differences in cardiac function but no pathology.

Additionally, it is possible that the lack of significant decrease in Tm phosphorylation seen in these Line

335 DMTG animals may be responsible for the lack in alterations in Ca2+ flux protein expression and phosphorylation. It is possible that to have the maximum beneficial effect, both individual Tm peptides in the dimer must be dephosphorylated. With the greater decrease in phosphorylation in Line 325, there is greater likelihood that both peptides will be dephosphorylated, compared to Line 335 which shows an increase in phosphorylation, decreasing the likelihood of having two dephosphorylated peptides in the same dimeric Tm. Both the α-Tm S283A and DMTG Line 325 hearts showed significant decreases in

Tm phosphorylation while DMTG Line 335 hearts showed an increase. It could, therefore, be speculated that Tm phosphorylation must decrease in order to see the resulting changes in Ca2+ flux proteins. I would speculate that given the lack of changes in proteins involved in Ca2+ fluxes, if this line was to be studied using echocardiography, it is unlikely that DMTG Line 335 animals would have a hypercontractile phenotype, as exists in the DMTG Line 325 animals.

117

α-Tm 180 TG Animals Bred With Tm Phosphorylation Mutants

Crossing α-Tm 180 TG animals with S283D and S283A TG animals

While generating cardiac specific α-Tm 180-S283A DMTG animals with both mutations on the same Tm molecule, TG α-Tm 180 animals were bred with both the TG α-Tm S283A and the TG α-Tm S283D mice. These crossed mice (denoted hereafter as E180GxS283A and E180GxS283D) would provide animals with the mutations on two separate Tm molecules and would allow us to quickly study the effects of both increasing and decreasing Tm phosphorylation on the progression of cardiac disease instigated by the E180G mutation.

Genotyping the 180xPhosphorylation Crosses

Both the α-Tm E180G construct and the α-Tm S283D/A constructs are expressed under the cardiac specific α-MHC promoter [193]. Site directed mutagenesis was utilized to make single nucleotide changes in the respective constructs, resulting in the desired amino acid residue changes (Table 1,

Jagatheesan and Wieczorek, unpublished). Genotyping for each of the three groups of TG animals was done utilizing a primer which lays down in the α-MHC promoter and another primer which lays down in the center of the α-Tm cDNA. Unfortunately, this method only allowed determination of whether the animal carried any of the three transgenes. Determination of whether both desired transgenes were present was imperative.

Originally, a high-throughput method of genotyping was developed for this laboratory; copy number by real time RT-PCR. Genomic DNA from tail clips was isolated, purified and quantitated. Two primers, appropriate for real time RT-PCR were generated and results were normalized to GAPDH expression.

Unfortunately, even though E180G, S283D and S283A TG mice were used as controls, it was extremely difficult to determine which animals carried both transgenes (Figure 41A). For example, although the α-

Tm 180 animals and the α-Tm S283A TG animals seemed to have similar copy numbers of the respective transgenes, sample 180xA-1 examined by copy number by Real Time RT-PCR does not seem to have

118 have the number of copies of α-Tm 180 added to α-Tm S283A copies. Looking at individual samples does not allow normalization of variability, increasing the difficulty in determining which of the double transgenic crosses carry both or one of the transgenes.

A secondary method of genotyping was attempted. Southern blot analyses were performed on the same samples utilized for RT-PCR analysis. Again, although the RT-PCR and Southern blot analyses seemed to agree, there was still difficulty in determining which animals carried both transgenes (Figure 41B,

41C). Again, the issue in examining the copy number of one animal does not allow for normalization of variability seen between samples, preventing accurate identification of which animals carried copies of two different TG lines.

119

Figure 41. Copy numbers of 180 mice bred with phosphorylation mutants. A. Copy number by real time RT-PCR. All samples normalized to GAPDH. B. Southern blot analysis of 180xS283A and 180xS283D crosses. C. Quantification of samples in panel A.

120

Finally, protein was isolated from these TG hearts, in order to determine whether genotype could be identified by Western blot analyses. α-Tm E180G protein undergoes differential migration under normal

SDS-PAGE conditions. Both urea and 10% SDS-acrylamide gels were run with E180G, S283D and

S283A run as controls. Although the expected separation occurred, there was, again, no certain way to determine which hearts had both transgenes given the lack of separation in the α-Tm S283A mutant

(Figure 42A, 42B).

Figure 42. Western blots of 180 mice bred with phosphorylation mutants. A. Western blot of myofibrillar preparations separated on 10% SDS-PAGE gel. B. Western blot analysis of myofibrillar preparations separated on 3.4M urea 10% SDS-PAGE gel. Both gels are probed with CH1 Tm specific antibody.

121

Conclusions on α-Tm 180 TG Animals Bred with Tm Phosphorylation Mutants

A study performed by Jagatheesan et al [31] showed that when TG animals expressing different Tm isoforms were crossed, the level of TG β and γ-Tm protein decreased, with each isoform (including the endogenous α-Tm) making up approximately 33% of the total Tm expressed in the myofiber. It is therefore likely that something similar would have happened in the case of these double transgenic crosses. 30% expression of either α-Tm S283A or α-Tm S283D protein would mean that more than 60% of the remaining Tm (TG and endogenous) would be available for phosphorylation. This outcome may have led to animals that did not show a significant phenotype that could be attributable to the alteration in phosphorylation.

While the copy number by RT-PCR and Southern analysis did not quite agree or lead to easy differentiation between TG and double TG animals, there was a benefit. Copy number by RT-PCR is an easy, high-throughput method of determining the copy number of a TG line that is trustworthy. This protocol is now standardized for use in the laboratory when investigating TG animals expressing Tm mutants. The lack of ability to account for variability between individual samples in the disease and phosphorylation mutant crosses is normalized when studying multiple samples from the same transgenic lines as the n number can be large enough to allow statistical analysis of copy number. Additionally, this method will lead to more rapid early characterization of new TG lines being produced, allowing reduction in the number of TG lines maintained.

Of interest to future scientists studying animals similar to these TG crosses, it was suggested that the

DNA from tail clips utilized for genotyping could be PCR amplified and sent to a gene sequencing service (Genewiz) to determine which animals carried both transgenes. While this method would be time consuming and expensive for some, it would be a reliable method of determining whether animals carried both transgenes.

122

CHAPTER 7: INVESTIGATING POSSIBLE BIOCHEMICAL CONTRIBUTORS TO

DIFFERENTIAL TM ISOFORM FUNCTION

Rationale

The striated muscle α, β and γ-Tm isoforms are extremely similar at the amino acid level. α and β-Tm share approximately 86% amino acid identity while α and γ-Tm share approximately 93% amino acid identity [18]. However, when α-Tm is replaced in the murine heart as the predominant isoform by overexpressing either β or γ-Tm, cardiac function is significantly altered as detailed in Chapter 1: Striated

Muscle Tropomyosins.

As explained previously in Chapter 1: Tropomyosin Protein, Tm is a coiled-coiled dimeric protein with amino acids arranged in a heptad motif (a, b, c, d, e, f, g) where amino acids in the a and d position interact hydrophobically and amino acids in the e and g position interact ionically. A major difference between the striated muscle isoforms is the number of d position alanines found in alanine clusters detailed in Chapter 1: Tropomyosin Protein. It has been suggested that the presence of the alanines allow narrowing or kinking of the coiled-coil dimer, facilitating Tm in wrapping around the actin filament. α-

Tm has the greatest number of d position alanines present in the 7 alanine clusters while β and γ-Tm have fewer alanines. We proposed that mutating β and γ-Tm to express more alanines in the d position alanine clusters, making the numbers of alanines present in the mutants identical to the number of alanines present in α-Tm, would allow determination of whether the number of d position alanines play a role in differential isoform function. Genetically designed β and γ-Tm proteins, expressing one of two single mutations or both mutations together would be subjected to actin binding assays in the presence and absence of Tn complex and those results would be to compared to WT mutant α, β and γ-Tm (Figure 43).

It is our hypothesis that the β and γ-Tm mutant proteins encoding both mutations would behave most similarly to WT α-Tm.

123

Figure 43. Heptad motif of WT Tm isoforms. A. α-Tm amino acid sequence. B. β-Tm amino acid sequence. C. γ-Tm amino acid sequence. Alanines highlighted in red type. Residues mutated to alanines in β and γ-Tm mutant proteins highlighted with red circles.

124

We had previously found that, when α, β and γ-Tm were subjected to a temperature melt via circular dichroism, they exhibit different flexibilities and different melting points [31]. It was our hypothesis that mutating β and γ-Tm to have the same number of alanines present in the alanine clusters as α-Tm would result in the same flexibility and melting point as α-Tm. There is some interest as to whether one mutation in either β or γ-Tm contributes more to the flexibility and actin binding ability of the isoform, but we believe that mutant recombinant proteins expressing both mutations will show the greatest similarities to α-Tm.

The studies proposed will provide significant insight on the biochemical properties of Tm isoforms and may elucidate the relationship between protein structure and physiological function. It is hoped that these studies will expand the body of knowledge on specific Tm isoforms and the interactions of those isoforms with actin, the Tn complex and myosin.

125

Generation and Purification of Recombinant Proteins

126

Actin Binding Assays

Prior to carrying out actin binding assays, it is necessary to generate a standard curve composed of the specific Tm isoform/mutant used in the actin binding assay. The same concentrations of Tm used in the binding assay must be run on an SDS-PAGE gel along with BSA as a loading control. Later, that information will allow calculation of the amount of free Tm versus bound Tm seen in the supernatant and pellet fractions of the actin binding assay respectively.

Additionally, it is crucial to test a wide range of concentrations (0.0 µM – 10.0 µM Tm has been previously reported [6, 12-14, 16, 168, 194, 195]) to ensure that Tm-actin binding is saturated and a sigmoidal curve can be calculated from available data. Despite multiple trials with WT α, β and γ-Tm, a good standard curve with an acceptable R2 value calculated via linear regression could not be generated

(Figure 44).

Figure 44. Standard curve of 0.0 µM α-Tm to 20.0 µM α-Tm.

127

Although it was not possible to calculate the apparent binding affinity of Tm for actin (kapp) as there was no reliable standard curve, the parameters of the actin binding assay were worked out and detailed in

Chapter 2. These parameters resulted in actin-Tm cosedimentation into the pellet fraction. Free Tm that did not bind actin is found in the supernatant fraction, along with a small amount of non-polymerized actin. BSA is added as a loading control (Figure 45).

Figure 45. Images of supernatant and pellet fractions from actin binding assay. A. Actin binding assay from 0.0 µM to 3.5 µM recombinant α-Tm. B. Acting binding assay from 0.0 µM to 3.5 µM recombinant β-Tm.

128

Circular Dichroism

Circular dichroism (CD) was performed with WT α, β and γ-Tm. While the global melting temperature

(Tm) can be calculated from the temperature melts shown in Figure 46A, it should be noted that Tm is a protein that undergoes multiple transitions while unfolding [147]. Interestingly, when the WT proteins were subjected to the data analysis for a three transition protein, the data gathered from the CD experiments would not properly fit the calculated curves, preventing the determination for the Tm for each transition, as should be properly reported (Figure 46B). Easy visualization of the three transitions can be seen when the first derivative of each thermal denaturation curve is calculated (Figure 46C). The approximate melting points of the individual transitions can be estimated based on the peaks shown in the graph. Interestingly, this data seems to indicate that γ-Tm exhibits two transitions that are more stable than either β-Tm or α-Tm. Additionally, while α- and β-Tm have the same transitions, the relative amounts of protein that are unfolding are inflected: more β-Tm transitions at an lower temperature than α-

Tm which transitions more at a later temperature.

129

Figure 46. Circular Dichroism analyses of recombinant WT Tm isoforms. A. Fraction unfolded of WT Tm isoforms. Global Tm is provided. B. Attempted fit of β and γ-Tm data to the equation for a three transition protein. C. Smooth of First Derivative of WT Tm Isoforms.

Given that the WT recombinant Tm isoforms did not fit the equation for a three transition protein, as it should, indicated that perhaps the protein was incompletely reduced. Incompletely reduced recombinant

Tm would introduce a fourth transition to the unfolding of the Tm protein. Therefore, each of the WT Tm isoforms were run under reducing and non-reducing conditions (Figure 46A). As seen in the Coomassie stained 10% SDS-PAGE gel, the proteins examined by CD were incompletely reduced (Figure 46A, non- reducing conditions). Therefore, a study was performed in order to determine how much tris(2- carboxyethyl)phosphine (TCEP) would be appropriate to reduce recombinant WT γ-Tm as much as possible (Figure 46B). Dithiothreitol (DTT) is not an appropriate reducing agent for use in CD studies as the chemical adds background signal and can skew results. For future experiments, it was decided that 10 mM TCEP added to the recombinant protein sample approximately 10 minutes prior to initiation of CD was effective to properly reduce the recombinant Tms for study.

130

Figure 47. Tm recombinant proteins. A. Recombinant WT proteins run on a 10% SDS-PAGE gel under reducing conditions and recombinant WT proteins run under non-reducing conditions. B. γ-Tm run under reducing conditions and γ-Tm run under non-reducing conditions with an increasing concentration of TCEP added to the recombinant protein sample.

Conclusions

Although the experiments detailed above were not completed in their entirety, important information was generated. All WT and mutant Tm recombinant proteins have been generated and efficiently express protein after induction with IPTG. These generated constructs will allow future scientists to progress quickly through the project.

Additionally, experimental procedures were worked out and standardized. The protein production and purification methodology, actin binding assay and circular dichroism procedures have been carefully tested and optimized. This work will allow future scientists to complete these important studies.

Information of particular value from the development of these protocols was the importance of completely reducing Tm recombinant protein. It was unknown, in this laboratory, that Tm was a multiple transition protein and that the presence or absence of disulfide bonds could perturb analysis of

CD-generated temperature melts meant to be analyzed to determine Tm flexibility. The failure to properly fit the CD data to the equations for a 3 transition protein led to the greater understanding of the

131 importance of fully reducing Tm recombinant protein prior to melting the temperature. Future data generated using reduced Tm in temperature melts via CD will properly fit the equations cited in

Greenfield’s seminal work on CD analysis and allow determination of Tm flexibility [147].

After the discovery that recombinant Tm is not completely reduced, TCEP was added to WT Tm isoforms prior to the initiation of actin binding assays. Compared to actin binding assays performed prior to increased reduction of Tm, the actin binding assays with reduced Tm were far more reliable and effective.

Information learned from the process of developing the actin binding assay or the CD methodology often resulted in more positive results or developments in the alternate assay.

It is of hope that the discoveries made during the optimization and standardization of these assays may help facilitate future work. The data that these experiments would yield are crucial to the enrichment of the field of Tm biochemistry. Greater understanding of biochemical contributors to differential Tm isoform functions would aid in the understanding of how Tm regulates Ca2+ sensitivity and contraction of the thin filaments in the striated muscle groups in which the protein exists.

132

CHAPTER 8: CONCLUSIONS AND FUTURE DIRECTIONS

Many previous studies regarding the function of Tm phosphorylation have been carried out in vitro.

These studies suggested that Tm phosphorylation increases the head-to-tail polymerization of neighboring

Tm coiled-coils when compared to polymerization of non-phosphorylated Tm. This increase in head-to- tail polymerization upon Tm phosphorylation has been speculated to result in increased cooperative activation of the thin filament which may alter myofilament function and by extension, cardiac function

[118-120].

During the development of disease or cardiac stress, there are two rapid mechanisms through which modulation of cardiac function can take place. Isoform switching to forms of proteins that are more sensitive to Ca2+ is one compensatory mechanism. However, isoform switching takes more time than the second mechanism; altering the phosphorylation status of sarcomeric and myofilament proteins. Studies of sarcomeric and myofilament proteins have indicated the importance of phosphorylation of proteins such as MyBPC, TnI, titin and MLC in the regulation and control of cardiac contractility [84, 91, 97,

106]. Taken together with the alterations in Tm phosphorylation in a variety of cardiac disease models discussed above, investigation into the role of α-Tm phosphorylation during both basal cardiac function and during cardiac disease was undertaken.

The work presented in this dissertation is primarily concerned with in vivo cardiovascular function of unphosphorylated α-Tm and the effects of reducing Tm phosphorylation under acute and chronic conditions of heart disease. α-Tm is the predominant Tm isoform expressed in the heart [21]. α-Tm is phosphorylated at a sole reported site, the penultimate amino acid, serine 283 [109-111]. Phosphorylation of cardiac Tm reaches approximately 70% during fetal development and drops to approximately 30% after birth. Throughout the lifetime of FVB/N mice, phosphorylation levels of Tm remains around 30% of total Tm, suggesting close regulation of Tm phosphorylation status by a currently unknown kinase (s)

133

[114]. Several cardiomyopathy models generated by this lab exhibit alterations in Tm phosphorylation status; the α-Tm E54K DCM model results in decreased Tm phosphorylation and the α-Tm 175 FHC model results in increased Tm phosphorylation [112, 113, 115]. Tm phosphorylation is altered in cardiac disease that is not caused by Tm mutations as well. Myocardial infarction and alterations in myofilament length result in increases in Tm phosphorylation. However, activated p38 MAPK depressed cardiac contractility results in decreased Tm phosphorylation, further indicating that Tm phosphorylation may be involved in the development or modulation of many types of cardiac disease [196-198].

In order to study the role of Tm phosphorylation in cardiac function in an in vivo system, two different

TG mouse model systems were developed. To mimic α-Tm phosphorylation, a cardiac specific Tm phosphorylation mimetic (Ser283Asp) was generated. Interestingly, there were no alterations in Ca2+ sensitivity or cooperative activation of the thin filament and there was no obvious cardiac dysfunction other than a mild decrease in the maximum rate of relaxation (-dP/dt). With high expression (>95% TG protein), there was a severe DCM phenotype that resulted in lethality by 1 month of age. Moderate expression mice demonstrated a mild DCM phenotype starting at 6 months of age [Jagatheesan and

Wieczorek, unpublished].

To further elucidate the function of cardiac α-Tm phosphorylation, animals expressing the α-Tm cDNA with serine 283 mutated to an alanine (α-Tm S283A) under the α-MHC promoter were generated.

Consistent with the previous studies on the α-Tm S283D TG animals, α-Tm S283A TG animals exhibit normal cardiac function, efficient contractility and relaxation under basal conditions and under β- adrenergic stimulation (Chapter 3). These results were puzzling considered in the context of previously published in vitro studies, the tight regulation on cardiac Tm phosphorylation levels and the alterations of

Tm phosphorylation levels in cardiac disease.

As many of the studies looking at Tm phosphorylation did so in the context of cardiac disease or stress, the α-Tm S283A TG animals were subjected to TAC operation. The TG animals failed more quickly and

134 exhibited a more severe cardiac phenotype compared to NTG controls, indicating that decreased cardiac

Tm phosphorylation in the context of an acute, extrinsic stressor is deleterious. Studies on Ca2+ handling proteins suggested a possible cause behind the rapid failure in TAC operated TG animals – increases in

SERCA2a expression and phosphorylation of PLN at Ser16 in the α-Tm S283A TG hearts likely results in an increase in SERCA2a activity. Interestingly, although clinical trials in which heart failure patients are provided with AAV delivered SERCA2a are ongoing and seem promising, other work has indicated that overexpression of SERCA2a results in more rapid failure after TAC [52, 53, 174, 188]. This increase in cardiac disease seen in both SERCA2a overexpressing animals as well as the α-Tm S283A TG animals undergoing TAC operation is possibly due to increases in energetic demand although Ca2+ handling studies must be performed to confirm increased Ca2+ transport. Energetic demand increases during heart disease and heart failure and SERCA2a is the second most energetically demanding pump in the myocardium (after the myosin Mg-ATPase), likely contributing to the poor prognosis after TAC.

A possible mechanism linking dephosphorylation to alterations in SERCA2a activity may be altered Ca2+ buffering. Ca2+ buffering is one of the mechanisms used to control cytoplasmic Ca2+ concentrations and therefore, Ca2+ signaling [199]. Specifically, Ca2+ buffering is the rapid binding of Ca2+ ions to different cellular binding sites in the cytoplasm. This buffering occurs by binding of Ca2+ to immobile buffers: molecules that are of high molecular weight or are anchored to intracellular structures, such as the RLC,

TnC, SERCA2a or the mitochondrial uniporter. Mobile buffers are small, soluble proteins or organic anions such as ATP [200, 201]. It was possible, in our studies, that the increase in SERCA2a could be an effect of altered Ca2+ buffering, meaning that more intracellular Ca2+ was present in the cytoplasm and must be buffered properly by SERCA2a to prevent altered signaling cascades. However, the link between

Tm dephosphorylation and SERCA2a in the context of Ca2+ buffering would lie in alterations in the Ca2+ binding ability of the two sarcomeric proteins that serve as Ca2+ buffers: the RLC and TnC [202, 203].

However, as the skinned fiber bundle studies indicate that the S283A fibers exhibit no change in Ca2+ sensitivity, there is no indication that Ca2+ buffering is altered at the level of the sarcomere, suggesting

135 that Ca2+ buffering through the sarcomere does not link Tm dephosphorylation and SERCA2a increase or the more rapid heart failure in the TAC operated TG animals. Whether Ca2+ buffering employing other mechanisms is operative with Tm dephosphorylation is yet to be determined.

Interestingly, when α-Tm dephosphorylation is coupled to a sarcomeric mutation responsible for the development of severe human and murine hypertrophic cardiomyopathy – the mutation of amino acid 180 in α-Tm from a glutamic acid to a glycine (α-Tm 180-S283A DMTG), cardiac function improved compared to the disease model (Chapter 5). All TG lines expressing the double mutant transgene showed significant increases in lifespan, similar to NTG and α-Tm S283A TG lifespans; α-Tm 180 animals die between 6-8 months of age [48, 49]. In fact, TG animals expressing very little phosphorylated α-Tm

(DMTG Line 325) exhibit a near complete morphological and physiological rescue of the α-Tm 180 phenotype, with no significant differences in myofilament Ca2+ sensitivity between DMTG and NTG animals, indicating that the rescue takes place at the level of the sarcomere. Additionally, DMTG Line

325 animals exhibit a hypercontractile phenotype when evaluated via echocardiography, indicating that decreasing Tm phosphorylation levels has benefits beyond the level of the sarcomere.

An interesting observation in this research is that the α-Tm 180-S283A DMTG Line 325 hearts do not exhibit significant differences in SERCA2a expression compared to NTG and the α-Tm 180 TG hearts.

Additionally, compared to both NTG and α-Tm S283A, the α-Tm 180-S283A DMTG Line 325 TG hearts also express less phosphorylated PLN at both Ser16 and Thr17. However, compared to α-Tm 180, the

DMTG Line 325 TG hearts express more phosphorylated PLN at both Ser16 and Thr17, which may account for the increase in contractility seen in these DMTG animals.

Another observation worth noting is the differential effects of variable levels of Tm phosphorylation in the DMTG animals. α-Tm 180-S283A Line 325 and Line 335 animals both exhibit normal lifespans and improvement/rescue of the severe α-Tm 180 phenotype. While Line 325 Tm is phosphorylated at very low levels, Line 335 Tm is phosphorylated at levels above NTG. It is possible that the mild but

136 significant increase in cardiomyocyte cross-sectional area seen in Line 335 is a function of this increase in

Tm phosphorylation. Additionally, I speculate that similar to the α-Tm S283D TG animals, Line 335 may exhibit a mild decrease in the maximum rate of relaxation.

The information presented above raises the question of how do changes in Tm phosphorylation result in alteration in the response of the murine heart to acute, extrinsic stress as well as chronic stress intrinsic to the sarcomere. These studies remain to be initiated, however, it is possible to speculate. Recent studies

[130-132] of Tm molecules expressing either the D175N or E180G mutations have indicated that both mutations result in a more flexible Tm molecule. Both the local flexibility of the molecule around the mutation increases, with a partial unwinding of the coiled-coil occurring in the vicinity of the mutation as well as an overall increase in the flexibility of the molecule. These alterations in flexibility of Tm molecules can affect both the head-to-tail polymerization of the molecules and the interaction with nearest neighbor molecules (Tn complex, actin, etc), and this impacts the ability of the thin filament to be cooperatively activated as seen by the alterations in the Hill coefficient in detergent extracted fiber bundles [48, 49, 115].

Replacement of a serine with an alanine, or the overall loss of phosphorylation in the α-Tm S283A TG animals might result in alterations of the local and overall flexibility of the Tm molecule. If actin or Tn complex binding is altered in these α-Tm S283A TG animals, it is possible that binding of other proteins that occur in the sarcomere may be altered as well. For example, PKCε is a novel PKC isoform that translocates to the z-disc [181]. PKCε, upon binding to actin, becomes constitutively activated [183] and studies have shown that PKCε is a positive modulator of compensatory cardiac hypertrophy, resulting in a similar phenotype to the α-Tm S283A TG animals [150]. Whether or not α-Tm dephosphorylation alters the ability of PKCε to interact with actin is yet to be shown and is an interesting question to follow-up on in the future.

137

Continuing with speculation regarding the rescue of the α-Tm 180 phenotype in the α-Tm 180-S283A

DMTG animals, I speculate that the introduction of the S283A mutation might result in a less flexible

Tm molecule compared to the very flexible molecule generated by the D175N or E180G mutation. A return to a level of flexibility similar to NTG in the α-Tm 180-S283A molecule could explain why, when detergent extracted myofibers are tested, the rescue at the level of the sarcomere is complete. A return to normal flexibility and Ca2+ sensitivity then results in a rescue of the phenotype at the level of the whole heart, returning DMTG Line 325 HW:BW ratios and cardiomyocyte cross-sectional areas to NTG levels.

Surprisingly, it seems that introduction of a lower amount of DMTG protein, such as that seen in the Line

335 TG mice, is sufficient for a near complete rescue of the α-Tm 180 phenotype as well. Although the

DMTG Line 335 TG animals exhibit a slight increase in cardiomyocyte cross-sectional area, there are no changes in HW:BW ratios and lifespan is normalized to NTG levels.

The S283A mutation is found at the very C-terminal end of the Tm molecule, which is known to be relatively unordered [168, 194, 204, 205]. However, the disorder of the Tm molecule seems to mainly be contributed to the splaying apart of the coiled-coil at the C-terminus in order to allow the previous Tm molecule to interact with the next N-terminus of Tm (Figure 48). It seems that the helical structure of the individual Tm peptides seems to be maintained due to the chemical nature of the heptad motif.

138

Figure 48. Model of the Ser283 residues at the C-terminus of tropomyosin. A. Adaptation of [168]. B. C-terminal Tm α-1 peptides are shown in brown and green, troponin T is purple, and the N-terminal Tm α-1 peptide is shown in gray (adapted from PDB code 2Z5H, [7]). Ser283 residues are shown as yellow sticks, highlighted by the black oval. Residues in the Tm α-1 N-terminal peptide that are positively (blue) or negatively (red) charged are highlighted. Of the two C-terminal TM α-1 peptides in the 2Z5H structure, one shows electron density straight through Ser283 (green helix), while the other only shows density through Met281 (brown helix). Thus, the six C-terminal residues from the complete helix were superimposed on the incomplete helix to illustrate the putative position of Ser283 on this chain (tan extension). Note that in this model, only 1 TnT molecule is present. Figure generated by Andrew Herr, Ph.D.

139

If the heptad motif is conserved in the individual Tm peptides, altering the amino acid sequence can play an important role in changing the structure of the C-terminus and the overall protein. Of equal possibility, however, is that the interaction of the splayed C-terminal Tm peptides may differentially interact with the

N-terminus of the following Tm molecule. As seen in Figure 48A, there seems to be close interaction of the C and N termini of subsequent Tm molecules. If there is a significant lack of Tm phosphorylation, the binding may be altered, even though the cooperative activation has been shown not to be altered.

Finally, although the Murakami [7] structure has been acknowledged to not be as correct as the

Greenfield [168] structure, Figure 48B, adapted from the Murakami structure allows visualization of the

Tm overlap and the interaction with TnT, although only one molecule is present in the structure. TnT binds directly over the Tm overlap and the lack of Tm phosphorylation may also disrupt Tm-TnT binding.

This disruption in TnT binding doesn’t impact cooperativity, as seen in Chapter 3, but there may be other, long stretching effects throughout the sarcomere.

Although there are significant alterations in both the expression and phosphorylation of SR proteins involved in Ca2+ handling in α-Tm S283A and α-Tm 180-S283A DMTG Line 325 hearts, there are no significant alterations in these proteins seen in the α-Tm 180-S283A DMTG Line 335 hearts. I speculate it is the significant decrease in Tm phosphorylation, which is similar in both α-Tm S283A and α-Tm 180-

S283A DMTG Line 325 DMTG hearts that results in alterations in SERCA2a activity caused by alterations in SERCA2a expression and PLN phosphorylation. The linkage between the significant decrease in Tm phosphorylation and the alteration in Ca2+ flux proteins is, as yet, unknown.

As with much basic scientific research, the question must be asked: can this information regarding the role of Tm phosphorylation be translated to care of human beings with cardiomyopathic disease? The increase in SERCA2a expression and possible increase in SERCA2a activity seen in the α-Tm S283A TG animals calls to mind similar results found in exercise trained mice [72, 74]. Additionally, humans suffering cardiac disease who are exercise trained do see a mild to moderate improvement in cardiac

140 function. However, these results must be taken with a grain of salt, as approximately 10% of humans who undergo extensive exercise training exhibit idiopathic arrhythmias, likely attributable to alterations in

Ca2+ handling [75].

Another possible therapeutic option is the delivery of dephosphorylated α-Tm to the heart via AAV.

However, much care must be taken in perturbing the balance of proteins in the sarcomere. In some cases,

TG mice overexpressing mutant Tms in the heart have shown severe phenotypes due to both the amount of TG protein incorporated into the sarcomere, indicating a threshold level for cardiac disease, as well as excessive deposition of mutant Tm into the cytoplasm of cardiomyocytes [28, 112].

Finally, the full effect of altering Tm phosphorylation in the context of intrinsic cardiac disease has not been elucidated. Hypertrophic cardiomyopathy and dilated cardiomyopathy occur through different, incompletely defined pathways. While one of the studies detailed above and focused on in this dissertation explores the role of decreased phosphorylation in the context of hypertrophic cardiomyopathy

(α-Tm 180-S283A) and begins to examine the effect of increasing phosphorylation to an extent in the

DMTG Line 335 TG hearts, it is crucial to better understand the effect of increasing phosphorylation of

Tm in the context of hypertrophic cardiomyopathy (α-Tm 180-S283D). Additionally, determining the effect of altering phosphorylation in the context of dilated cardiomyopathy must also be carried out (α-Tm

E54K-S283A, α-Tm E54K-S283D) as studies have shown that different treatments have differential effects in different models of cardiac disease. In the α-Tm E54K TG hearts, Tm phosphorylation actually decreases [113], counter to what occurs in the α-Tm 180 hearts. Therefore, the effect of decreasing Tm phosphorylation in the context of dilatation may result in negative effects, opposite to what is seen in the context of hypertrophic cardiomyopathy.

In conclusion, α-Tm phosphorylation seems to play an important role in the development of both acute and chronic cardiac disease. Many of the interesting results, especially results indicating possible alterations in Ca2+ handling, warrant further work to elucidate the underlying molecular mechanisms

141 involved in translating Tm dephosphorylation to altered SR protein phosphorylation and expression. Of special interest is the data that seems to suggest that decreasing Tm phosphorylation in the context of an acute extrinsic stressor (TAC) may be negative. While more work must be done to evaluate the role of

Tm phosphorylation status in the context of many different types of cardiac disease, these are the first studies examining the role of Tm dephosphorylation at basal levels, under acute cardiac stress and under chronic cardiac stress.

142

REFERENCES

1. Perry, S.V., Vertebrate tropomyosin: distribution, properties and function. Journal of muscle research and cell motility, 2001. 22(1): p. 5-49. 2. Wieczorek, D.F., C.W. Smith, and B. Nadal-Ginard, The rat alpha-tropomyosin gene generates a minimum of six different mRNAs coding for striated, smooth, and nonmuscle isoforms by alternative splicing. Molecular and cellular biology, 1988. 8(2): p. 679-94. 3. Gunning, P.W., et al., Tropomyosin isoforms: divining rods for actin cytoskeleton function. Trends in cell biology, 2005. 15(6): p. 333-41. 4. Lin, J.J., T.E. Hegmann, and J.L. Lin, Differential localization of tropomyosin isoforms in cultured nonmuscle cells. The Journal of cell biology, 1988. 107(2): p. 563-72. 5. Wieczorek, D.F., Regulation of alternatively spliced alpha-tropomyosin gene expression by nerve extract. The Journal of biological chemistry, 1988. 263(21): p. 10456-63. 6. Greenfield, N.J., et al., Solution NMR structure and folding dynamics of the N terminus of a rat non-muscle alpha-tropomyosin in an engineered chimeric protein. Journal of molecular biology, 2001. 312(4): p. 833-47. 7. Murakami, K., et al., Structural basis for tropomyosin overlap in thin (actin) filaments and the generation of a molecular swivel by troponin-T. Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(20): p. 7200-5. 8. Cohen, C. and D.A. Parry, Alpha-helical coiled coils and bundles: how to design an alpha-helical protein. Proteins, 1990. 7(1): p. 1-15. 9. Conway, J.F. and D.A. Parry, Structural features in the heptad substructure and longer range repeats of two-stranded alpha-fibrous proteins. International journal of biological macromolecules, 1990. 12(5): p. 328-34. 10. Stewart, M., Structural basis for bending tropomyosin around actin in muscle thin filaments. Proceedings of the National Academy of Sciences of the United States of America, 2001. 98(15): p. 8165-6. 11. Wolska, B.M. and D.M. Wieczorek, The role of tropomyosin in the regulation of myocardial contraction and relaxation. Pflugers Archiv : European journal of physiology, 2003. 446(1): p. 1- 8. 12. Brown, J.H., et al., Deciphering the design of the tropomyosin molecule. Proceedings of the National Academy of Sciences of the United States of America, 2001. 98(15): p. 8496-501. 13. Singh, A. and S.E. Hitchcock-DeGregori, Local destabilization of the tropomyosin coiled coil gives the molecular flexibility required for actin binding. Biochemistry, 2003. 42(48): p. 14114- 21. 14. Singh, A. and S.E. Hitchcock-DeGregori, Dual requirement for flexibility and specificity for binding of the coiled-coil tropomyosin to its target, actin. Structure, 2006. 14(1): p. 43-50. 15. McLachlan, A.D., M. Stewart, and L.B. Smillie, Sequence repeats in alpha-tropomyosin. Journal of molecular biology, 1975. 98(2): p. 281-91. 16. Singh, A. and S.E. Hitchcock-DeGregori, Tropomyosin's periods are quasi-equivalent for actin binding but have specific regulatory functions. Biochemistry, 2007. 46(51): p. 14917-27. 17. Pieples, K. and D.F. Wieczorek, Tropomyosin 3 increases striated muscle isoform diversity. Biochemistry, 2000. 39(28): p. 8291-7. 18. Lees-Miller, J.P. and D.M. Helfman, The molecular basis for tropomyosin isoform diversity. BioEssays : news and reviews in molecular, cellular and developmental biology, 1991. 13(9): p. 429-37. 19. Denz, C.R., et al., Expression of a novel cardiac-specific tropomyosin isoform in humans. Biochemical and biophysical research communications, 2004. 320(4): p. 1291-7. 20. Rajan, S., et al., Molecular and functional characterization of a novel cardiac-specific human tropomyosin isoform. Circulation, 2010. 121(3): p. 410-8.

143

21. Muthuchamy, M., et al., Developmental analysis of tropomyosin gene expression in embryonic stem cells and mouse embryos. Molecular and cellular biology, 1993. 13(6): p. 3311-23. 22. Muthuchamy, M., et al., Molecular and physiological effects of overexpressing striated muscle beta-tropomyosin in the adult murine heart. The Journal of biological chemistry, 1995. 270(51): p. 30593-603. 23. Jagatheesan, G., S. Rajan, and D.F. Wieczorek, Investigations into tropomyosin function using mouse models. Journal of molecular and cellular cardiology, 2010. 48(5): p. 893-8. 24. Ayscough, K.R., In vivo functions of actin-binding proteins. Current opinion in cell biology, 1998. 10(1): p. 102-11. 25. Schevzov, G., et al., Divergent regulation of the sarcomere and the cytoskeleton. The Journal of biological chemistry, 2008. 283(1): p. 275-83. 26. Wolska, B.M., et al., Correlation between myofilament response to Ca2+ and altered dynamics of contraction and relaxation in transgenic cardiac cells that express beta-tropomyosin. Circulation research, 1999. 84(7): p. 745-51. 27. Palmiter, K.A., et al., Exchange of beta- for alpha-tropomyosin in hearts of transgenic mice induces changes in thin filament response to Ca2+, strong cross-bridge binding, and protein phosphorylation. The Journal of biological chemistry, 1996. 271(20): p. 11611-4. 28. Muthuchamy, M., et al., Beta-tropomyosin overexpression induces severe cardiac abnormalities. Journal of molecular and cellular cardiology, 1998. 30(8): p. 1545-57. 29. Gunning, P., et al., Differential control of tropomyosin mRNA levels during myogenesis suggests the existence of an isoform competition-autoregulatory compensation control mechanism. Developmental biology, 1990. 138(2): p. 443-53. 30. Pieples, K., et al., Tropomyosin 3 expression leads to hypercontractility and attenuates myofilament length-dependent Ca(2+) activation. American journal of physiology. Heart and circulatory physiology, 2002. 283(4): p. H1344-53. 31. Jagatheesan, G., et al., Striated muscle tropomyosin isoforms differentially regulate cardiac performance and myofilament calcium sensitivity. Journal of muscle research and cell motility, 2010. 31(3): p. 227-39. 32. Boussouf, S.E., et al., Role of tropomyosin isoforms in the calcium sensitivity of striated muscle thin filaments. Journal of muscle research and cell motility, 2007. 28(1): p. 49-58. 33. Smith, D.A. and M.A. Geeves, Cooperative regulation of myosin-actin interactions by a continuous flexible chain II: actin-tropomyosin-troponin and regulation by calcium. Biophysical journal, 2003. 84(5): p. 3168-80. 34. Lehrer, S.S. and M.A. Geeves, The muscle thin filament as a classical cooperative/allosteric regulatory system. Journal of molecular biology, 1998. 277(5): p. 1081-9. 35. Hasenfuss, G., et al., Relation between myocardial function and expression of sarcoplasmic reticulum Ca(2+)-ATPase in failing and nonfailing human myocardium. Circulation research, 1994. 75(3): p. 434-42. 36. Hasenfuss, G., et al., Relationship between Na+-Ca2+-exchanger protein levels and diastolic function of failing human myocardium. Circulation, 1999. 99(5): p. 641-8. 37. Drexler, H., G. Hasenfuss, and C. Holubarsch, Signaling pathways in failing human heart muscle cells. Trends in cardiovascular medicine, 1997. 7(5): p. 151-60. 38. Schmidt, U., et al., Human heart failure: cAMP stimulation of SR Ca(2+)-ATPase activity and phosphorylation level of phospholamban. The American journal of physiology, 1999. 277(2 Pt 2): p. H474-80. 39. Ji, Y., et al., Disruption of a single copy of the SERCA2 gene results in altered Ca2+ homeostasis and cardiomyocyte function. The Journal of biological chemistry, 2000. 275(48): p. 38073-80. 40. Periasamy, M., et al., Impaired cardiac performance in heterozygous mice with a null mutation in the sarco(endo)plasmic reticulum Ca2+-ATPase isoform 2 (SERCA2) gene. The Journal of biological chemistry, 1999. 274(4): p. 2556-62.

144

41. Shull, G.E., et al., Physiological functions of plasma membrane and intracellular Ca2+ pumps revealed by analysis of null mutants. Annals of the New York Academy of Sciences, 2003. 986: p. 453-60. 42. Schultz Jel, J., et al., Accelerated onset of heart failure in mice during pressure overload with chronically decreased SERCA2 calcium pump activity. American journal of physiology. Heart and circulatory physiology, 2004. 286(3): p. H1146-53. 43. MacLennan, D.H. and E.G. Kranias, Phospholamban: a crucial regulator of cardiac contractility. Nature reviews. Molecular cell biology, 2003. 4(7): p. 566-77. 44. Baker, D.L., et al., Targeted overexpression of the sarcoplasmic reticulum Ca2+-ATPase increases cardiac contractility in transgenic mouse hearts. Circulation research, 1998. 83(12): p. 1205-14. 45. Vetter, R., et al., Transgenic overexpression of the sarcoplasmic reticulum Ca2+ATPase improves reticular Ca2+ handling in normal and diabetic rat hearts. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 2002. 16(12): p. 1657-9. 46. He, H., et al., Overexpression of the rat sarcoplasmic reticulum Ca2+ ATPase gene in the heart of transgenic mice accelerates calcium transients and cardiac relaxation. The Journal of clinical investigation, 1997. 100(2): p. 380-9. 47. Suarez, J., et al., Doxycycline inducible expression of SERCA2a improves calcium handling and reverts cardiac dysfunction in pressure overload-induced cardiac hypertrophy. American journal of physiology. Heart and circulatory physiology, 2004. 287(5): p. H2164-72. 48. Prabhakar, R., et al., A familial hypertrophic cardiomyopathy alpha-tropomyosin mutation causes severe cardiac hypertrophy and death in mice. Journal of molecular and cellular cardiology, 2001. 33(10): p. 1815-28. 49. Prabhakar, R., et al., A mouse model of familial hypertrophic cardiomyopathy caused by a alpha- tropomyosin mutation. Molecular and cellular biochemistry, 2003. 251(1-2): p. 33-42. 50. Al Moamen, N.J., et al., Loss of the AE3 anion exchanger in a hypertrophic cardiomyopathy model causes rapid decompensation and heart failure. Journal of molecular and cellular cardiology, 2011. 50(1): p. 137-46. 51. Pena, J.R., et al., Neonatal gene transfer of Serca2a delays onset of hypertrophic remodeling and improves function in familial hypertrophic cardiomyopathy. Journal of molecular and cellular cardiology, 2010. 49(6): p. 993-1002. 52. Gwathmey, J.K., A.I. Yerevanian, and R.J. Hajjar, Cardiac gene therapy with SERCA2a: from bench to bedside. Journal of molecular and cellular cardiology, 2011. 50(5): p. 803-12. 53. Pinz, I., et al., Compromised myocardial energetics in hypertrophied mouse hearts diminish the beneficial effect of overexpressing SERCA2a. The Journal of biological chemistry, 2011. 286(12): p. 10163-8. 54. Maack, C. and B. O'Rourke, Excitation-contraction coupling and mitochondrial energetics. Basic research in cardiology, 2007. 102(5): p. 369-92. 55. Kadambi, V.J., et al., Cardiac-specific overexpression of phospholamban alters calcium kinetics and resultant cardiomyocyte mechanics in transgenic mice. The Journal of clinical investigation, 1996. 97(2): p. 533-9. 56. Tada, M., Molecular structure and function of phospholamban in regulating the calcium pump from sarcoplasmic reticulum. Annals of the New York Academy of Sciences, 1992. 671: p. 92- 102; discussion 102-3. 57. Brittsan, A.G., et al., Maximal inhibition of SERCA2 Ca(2+) affinity by phospholamban in transgenic hearts overexpressing a non-phosphorylatable form of phospholamban. The Journal of biological chemistry, 2000. 275(16): p. 12129-35. 58. Brittsan, A.G., et al., Chronic SR Ca2+-ATPase inhibition causes adaptive changes in cellular Ca2+ transport. Circulation research, 2003. 92(7): p. 769-76.

145

59. Cornea, R.L., et al., Mutation and phosphorylation change the oligomeric structure of phospholamban in lipid bilayers. Biochemistry, 1997. 36(10): p. 2960-7. 60. Kimura, Y., et al., Phospholamban inhibitory function is activated by depolymerization. The Journal of biological chemistry, 1997. 272(24): p. 15061-4. 61. Zvaritch, E., et al., The transgenic expression of highly inhibitory monomeric forms of phospholamban in mouse heart impairs cardiac contractility. The Journal of biological chemistry, 2000. 275(20): p. 14985-91. 62. Luo, W., et al., Targeted ablation of the phospholamban gene is associated with markedly enhanced myocardial contractility and loss of beta-agonist stimulation. Circulation research, 1994. 75(3): p. 401-9. 63. Hoit, B.D., et al., In vivo echocardiographic detection of enhanced left ventricular function in gene-targeted mice with phospholamban deficiency. Circulation research, 1995. 77(3): p. 632-7. 64. Delling, U., M.A. Sussman, and J.D. Molkentin, Re-evaluating sarcoplasmic reticulum function in heart failure. Nature medicine, 2000. 6(9): p. 942-3. 65. Freeman, K., et al., Alterations in cardiac adrenergic signaling and calcium cycling differentially affect the progression of cardiomyopathy. The Journal of clinical investigation, 2001. 107(8): p. 967-74. 66. Gaffin, R.D., et al., Long-term rescue of a familial hypertrophic cardiomyopathy caused by a mutation in the thin filament protein, tropomyosin, via modulation of a calcium cycling protein. Journal of molecular and cellular cardiology, 2011. 51(5): p. 812-20. 67. Zhong, Y., P.J. Reiser, and M.A. Matlib, Gender differences in myosin heavy chain-beta and phosphorylated phospholamban in diabetic rat hearts. American journal of physiology. Heart and circulatory physiology, 2003. 285(6): p. H2688-93. 68. Chu, G., et al., A single site (Ser16) phosphorylation in phospholamban is sufficient in mediating its maximal cardiac responses to beta -agonists. The Journal of biological chemistry, 2000. 275(49): p. 38938-43. 69. Hoshijima, M., et al., Chronic suppression of heart-failure progression by a pseudophosphorylated mutant of phospholamban via in vivo cardiac rAAV gene delivery. Nature medicine, 2002. 8(8): p. 864-71. 70. Pluim, B.M., et al., The athlete's heart. A meta-analysis of cardiac structure and function. Circulation, 2000. 101(3): p. 336-44. 71. Bernardo, B.C., et al., Molecular distinction between physiological and pathological cardiac hypertrophy: experimental findings and therapeutic strategies. Pharmacology & therapeutics, 2010. 128(1): p. 191-227. 72. Kemi, O.J., et al., Myocardial sarcoplasmic reticulum Ca2+ ATPase function is increased by aerobic interval training. European journal of cardiovascular prevention and rehabilitation : official journal of the European Society of Cardiology, Working Groups on Epidemiology & Prevention and Cardiac Rehabilitation and Exercise Physiology, 2008. 15(2): p. 145-8. 73. Wisloff, U., et al., Increased contractility and calcium sensitivity in cardiac myocytes isolated from endurance trained rats. Cardiovascular research, 2001. 50(3): p. 495-508. 74. Kemi, O.J., et al., Aerobic interval training enhances cardiomyocyte contractility and Ca2+ cycling by phosphorylation of CaMKII and Thr-17 of phospholamban. Journal of molecular and cellular cardiology, 2007. 43(3): p. 354-61. 75. Lu, L., et al., Exercise training normalizes altered calcium-handling proteins during development of heart failure. Journal of applied physiology, 2002. 92(4): p. 1524-30. 76. Wisloff, U., et al., Aerobic exercise reduces cardiomyocyte hypertrophy and increases contractility, Ca2+ sensitivity and SERCA-2 in rat after myocardial infarction. Cardiovascular research, 2002. 54(1): p. 162-74. 77. Pina, I.L., et al., Exercise and heart failure: A statement from the American Heart Association Committee on exercise, rehabilitation, and prevention. Circulation, 2003. 107(8): p. 1210-25.

146

78. Cannistra, L.B., et al., Moderate-high intensity exercise training after myocardial infarction: effect on left ventricular remodeling. Journal of cardiopulmonary rehabilitation, 1999. 19(6): p. 373-80. 79. Yamasaki, R., et al., Protein kinase A phosphorylates titin's cardiac-specific N2B domain and reduces passive tension in rat cardiac myocytes. Circulation research, 2002. 90(11): p. 1181-8. 80. Kruger, M., et al., Protein kinase G modulates human myocardial passive stiffness by phosphorylation of the titin springs. Circulation research, 2009. 104(1): p. 87-94. 81. Kruger, M. and W.A. Linke, Protein kinase-A phosphorylates titin in human heart muscle and reduces myofibrillar passive tension. Journal of muscle research and cell motility, 2006. 27(5-7): p. 435-44. 82. Linke, W.A., Sense and stretchability: the role of titin and titin-associated proteins in myocardial stress-sensing and mechanical dysfunction. Cardiovascular research, 2008. 77(4): p. 637-48. 83. Borbely, A., et al., Cardiomyocyte stiffness in diastolic heart failure. Circulation, 2005. 111(6): p. 774-81. 84. Granzier, H. and S. Labeit, Structure-function relations of the giant elastic protein titin in striated and smooth muscle cells. Muscle & nerve, 2007. 36(6): p. 740-55. 85. Colucci, W.S., ed. Atlas of Heart Failure: Cardiac Function and Dysfunction. 4th ed. 2004, Current Medicine Group. 86. Richard, P., et al., Hypertrophic cardiomyopathy: distribution of disease genes, spectrum of mutations, and implications for a molecular diagnosis strategy. Circulation, 2003. 107(17): p. 2227-32. 87. van Dijk, S.J., et al., A piece of the human heart: variance of protein phosphorylation in left ventricular samples from end-stage primary cardiomyopathy patients. Journal of muscle research and cell motility, 2009. 30(7-8): p. 299-302. 88. Sadayappan, S., et al., Cardiac myosin-binding protein-C phosphorylation and cardiac function. Circulation research, 2005. 97(11): p. 1156-63. 89. Tong, C.W., et al., Acceleration of crossbridge kinetics by protein kinase A phosphorylation of cardiac myosin binding protein C modulates cardiac function. Circulation research, 2008. 103(9): p. 974-82. 90. Sadayappan, S., et al., Cardiac myosin binding protein C phosphorylation is cardioprotective. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(45): p. 16918-23. 91. Kuster, D.W., et al., Cardiac myosin binding protein C phosphorylation in cardiac disease. Journal of muscle research and cell motility, 2012. 33(1): p. 43-52. 92. Rottbauer, W., et al., Cardiac myosin light chain-2: a novel essential component of thick- myofilament assembly and contractility of the heart. Circulation research, 2006. 99(3): p. 323-31. 93. Soderling, T.R. and J.T. Stull, Structure and regulation of calcium/calmodulin-dependent protein kinases. Chemical reviews, 2001. 101(8): p. 2341-52. 94. Olsson, M.C., et al., Basal myosin light chain phosphorylation is a determinant of Ca2+ sensitivity of force and activation dependence of the kinetics of myocardial force development. American journal of physiology. Heart and circulatory physiology, 2004. 287(6): p. H2712-8. 95. Patel, J.R., et al., Phosphorylation of myosin regulatory light chain eliminates force-dependent changes in relaxation rates in skeletal muscle. Biophysical journal, 1998. 74(1): p. 360-8. 96. Scruggs, S.B., et al., Ablation of ventricular myosin regulatory light chain phosphorylation in mice causes cardiac dysfunction in situ and affects neighboring myofilament protein phosphorylation. The Journal of biological chemistry, 2009. 284(8): p. 5097-106. 97. Levine, R.J., et al., Myosin light chain phosphorylation affects the structure of rabbit skeletal muscle thick filaments. Biophysical journal, 1996. 71(2): p. 898-907. 98. Jideama, N.M., et al., Phosphorylation specificities of protein kinase C isozymes for bovine cardiac troponin I and troponin T and sites within these proteins and regulation of myofilament properties. The Journal of biological chemistry, 1996. 271(38): p. 23277-83.

147

99. Montgomery, D.E., et al., Transgenic incorporation of skeletal TnT into cardiac myofilaments blunts PKC-mediated depression of force. American journal of physiology. Heart and circulatory physiology, 2001. 280(3): p. H1011-8. 100. Perry, S.V., Troponin T: Genetics, properties and function. Journal of muscle research and cell motility, 1998. 19(6): p. 575-602. 101. Sumandea, M.P., et al., Regulation of myofilament activation by PKC phosphorylation of cardiac TnT at Thr206. Biophysical journal, 2003. 84(2): p. 319a-319a. 102. Sumandea, M.P., et al., Identification of a functionally critical protein kinase C phosphorylation residue of cardiac troponin T. Journal of Biological Chemistry, 2003. 278(37): p. 35135-35144. 103. Wolska, B.M., et al., Expression of slow skeletal troponin I in adult transgenic mouse heart muscle reduces the force decline observed during acidic conditions. The Journal of physiology, 2001. 536(Pt 3): p. 863-70. 104. Varian, K.D. and P.M. Janssen, Frequency-dependent acceleration of relaxation involves decreased myofilament calcium sensitivity. American journal of physiology. Heart and circulatory physiology, 2007. 292(5): p. H2212-9. 105. Takimoto, E., et al., Frequency- and afterload-dependent cardiac modulation in vivo by troponin I with constitutively active protein kinase A phosphorylation sites. Circulation research, 2004. 94(4): p. 496-504. 106. Pyle, W.G., et al., Troponin I serines 43/45 and regulation of cardiac myofilament function. American journal of physiology. Heart and circulatory physiology, 2002. 283(3): p. H1215-24. 107. Burkart, E.M., et al., Altered signaling surrounding the C-lobe of cardiac troponin C in myofilaments containing an alpha-tropomyosin mutation linked to familial hypertrophic cardiomyopathy. Journal of molecular and cellular cardiology, 2003. 35(10): p. 1285-93. 108. Solaro, R.J. and T. Kobayashi, Protein phosphorylation and signal transduction in cardiac thin filaments. The Journal of biological chemistry, 2011. 286(12): p. 9935-40. 109. Heeley, D.A., A.J. Moir, and S.V. Perry, Phosphorylation of tropomyosin during development in mammalian striated muscle. FEBS letters, 1982. 146(1): p. 115-8. 110. Mak, A., L.B. Smillie, and M. Barany, Specific phosphorylation at serine-283 of alpha tropomyosin from frog skeletal and rabbit skeletal and cardiac muscle. Proceedings of the National Academy of Sciences of the United States of America, 1978. 75(8): p. 3588-92. 111. Yuan, C. and R.J. Solaro, Myofilament proteins: From cardiac disorders to proteomic changes. Proteomics. Clinical applications, 2008. 2(6): p. 788-99. 112. Rajan, S., et al., Dilated cardiomyopathy mutant tropomyosin mice develop cardiac dysfunction with significantly decreased fractional shortening and myofilament calcium sensitivity. Circulation research, 2007. 101(2): p. 205-14. 113. Warren, C.M., et al., Use of 2-D DIGE analysis reveals altered phosphorylation in a tropomyosin mutant (Glu54Lys) linked to dilated cardiomyopathy. Proteomics, 2008. 8(1): p. 100-5. 114. Sheikh, H.N., Tropomyosin Phosphorylation in Cardiac Health and Disease, in Molecular Genetics, Biochemistry and Microbiology2009, University of Cincinnati: Cincinnati. p. 76. 115. Muthuchamy, M., et al., Mouse model of a familial hypertrophic cardiomyopathy mutation in alpha-tropomyosin manifests cardiac dysfunction. Circulation research, 1999. 85(1): p. 47-56. 116. Montgomery, K. and A.S. Mak, In vitro phosphorylation of tropomyosin by a kinase from chicken embryo. The Journal of biological chemistry, 1984. 259(9): p. 5555-60. 117. deBelle, I. and A.S. Mak, Isolation and characterization of tropomyosin kinase from chicken embryo. Biochimica et biophysica acta, 1987. 925(1): p. 17-26. 118. Heeley, D.H., et al., Effect of phosphorylation on the interaction and functional properties of rabbit striated muscle alpha alpha-tropomyosin. The Journal of biological chemistry, 1989. 264(5): p. 2424-30. 119. Heeley, D.H., Investigation of the effects of phosphorylation of rabbit striated muscle alpha alpha-tropomyosin and rabbit skeletal muscle troponin-T. European journal of biochemistry / FEBS, 1994. 221(1): p. 129-37.

148

120. Sano, K., et al., The effect of single residue substitutions of serine-283 on the strength of head-to- tail interaction and actin binding properties of rabbit skeletal muscle alpha-tropomyosin. Journal of biochemistry, 2000. 127(6): p. 1095-102. 121. Tardiff, J.C., Sarcomeric proteins and familial hypertrophic cardiomyopathy: linking mutations in structural proteins to complex cardiovascular phenotypes. Heart failure reviews, 2005. 10(3): p. 237-48. 122. Bashyam, M.D., et al., Molecular genetics of familial hypertrophic cardiomyopathy (FHC). Journal of human genetics, 2003. 48(2): p. 55-64. 123. Maron, B.J., Hypertrophic cardiomyopathy: a systematic review. JAMA : the journal of the American Medical Association, 2002. 287(10): p. 1308-20. 124. Nakajima-Taniguchi, C., et al., Novel missense mutation in alpha-tropomyosin gene found in Japanese patients with hypertrophic cardiomyopathy. Journal of molecular and cellular cardiology, 1995. 27(9): p. 2053-8. 125. Yamauchi-Takihara, K., et al., Clinical implications of hypertrophic cardiomyopathy associated with mutations in the alpha-tropomyosin gene. Heart, 1996. 76(1): p. 63-5. 126. Jaaskelainen, P., et al., The cardiac beta-myosin heavy chain gene is not the predominant gene for hypertrophic cardiomyopathy in the Finnish population. Journal of the American College of Cardiology, 1998. 32(6): p. 1709-16. 127. Coviello, D.A., et al., Clinical features of hypertrophic cardiomyopathy caused by mutation of a "hot spot" in the alpha-tropomyosin gene. Journal of the American College of Cardiology, 1997. 29(3): p. 635-40. 128. Thierfelder, L., et al., Alpha-tropomyosin and cardiac troponin T mutations cause familial hypertrophic cardiomyopathy: a disease of the sarcomere. Cell, 1994. 77(5): p. 701-12. 129. Golitsina, N., et al., Effects of two familial hypertrophic cardiomyopathy-causing mutations on alpha-tropomyosin structure and function. Biochemistry, 1997. 36(15): p. 4637-42. 130. Bing, W., et al., Effects of two hypertrophic cardiomyopathy mutations in alpha-tropomyosin, Asp175Asn and Glu180Gly, on Ca2+ regulation of thin filament motility. Biochemical and biophysical research communications, 1997. 236(3): p. 760-4. 131. Ly, S. and S.S. Lehrer, Long-Range Effects of Familial Hypertrophic Cardiomyopathy Mutations E180G and D175N on the Properties of Tropomyosin. Biochemistry, 2012. 51(32): p. 6413-20. 132. Li, X.E., et al., The flexibility of two tropomyosin mutants, D175N and E180G, that cause hypertrophic cardiomyopathy. Biochemical and biophysical research communications, 2012. 424(3): p. 493-6. 133. Loong, C.K., H.X. Zhou, and P. Bryant Chase, Familial hypertrophic cardiomyopathy related E180G mutation increases flexibility of human cardiac alpha-tropomyosin. FEBS letters, 2012. 134. Evans, C.C., et al., Altered hemodynamics in transgenic mice harboring mutant tropomyosin linked to hypertrophic cardiomyopathy. American journal of physiology. Heart and circulatory physiology, 2000. 279(5): p. H2414-23. 135. Jagatheesan, G., et al., Functional importance of the carboxyl-terminal region of striated muscle tropomyosin. The Journal of biological chemistry, 2003. 278(25): p. 23204-11. 136. Jagatheesan, G., et al., Rescue of tropomyosin-induced familial hypertrophic cardiomyopathy mice by transgenesis. American journal of physiology. Heart and circulatory physiology, 2007. 293(2): p. H949-58. 137. Michele, D.E., et al., Cardiac dysfunction in hypertrophic cardiomyopathy mutant tropomyosin mice is transgene-dependent, hypertrophy-independent, and improved by beta-blockade. Circulation research, 2002. 91(3): p. 255-62. 138. Olson, T.M., et al., Mutations that alter the surface charge of alpha-tropomyosin are associated with dilated cardiomyopathy. Journal of molecular and cellular cardiology, 2001. 33(4): p. 723- 32. 139. Pfaffl, M.W., A new mathematical model for relative quantification in real-time RT-PCR. Nucleic acids research, 2001. 29(9): p. e45.

149

140. Wilkins, B.J., et al., Targeted disruption of NFATc3, but not NFATc4, reveals an intrinsic defect in calcineurin-mediated cardiac hypertrophic growth. Molecular and cellular biology, 2002. 22(21): p. 7603-13. 141. Wilkins, B.J., et al., Calcineurin/NFAT coupling participates in pathological, but not physiological, cardiac hypertrophy. Circulation research, 2004. 94(1): p. 110-8. 142. deAlmeida, A.C., R.J. van Oort, and X.H. Wehrens, Transverse aortic constriction in mice. Journal of visualized experiments : JoVE, 2010(38). 143. Nakayama, H., et al., alpha1G-dependent T-type Ca2+ current antagonizes cardiac hypertrophy through a NOS3-dependent mechanism in mice. The Journal of clinical investigation, 2009. 119(12): p. 3787-96. 144. Liao, S., et al., The cardioprotective effect of the low molecular weight isoform of fibroblast growth factor-2: the role of JNK signaling. Journal of molecular and cellular cardiology, 2007. 42(1): p. 106-20. 145. Daugherty, A., et al., Measuring blood pressure in mice using volume pressure recording, a tail- cuff method. Journal of visualized experiments : JoVE, 2009(27). 146. Hammell, R.L. and S.E. Hitchcock-DeGregori, Mapping the functional domains within the carboxyl terminus of alpha-tropomyosin encoded by the alternatively spliced ninth exon. The Journal of biological chemistry, 1996. 271(8): p. 4236-42. 147. Greenfield, N.J., Using circular dichroism spectra to estimate protein secondary structure. Nature Protocols, 2007. 1(6): p. 2876-2890. 148. Reddy, Y.S., et al., Phosphorylation of cardiac native tropomyosin and troponin: inhibitory effect of actomyosin and possible presence of endogenous myofibrillar-located cyclic-AMP-dependent protein kinase. Journal of molecular and cellular cardiology, 1973. 5(5): p. 461-71. 149. Watson, M.H., et al., Phosphorylation of alpha alpha- and beta beta-tropomyosin and synthetic peptide analogues. Biochemistry, 1988. 27(12): p. 4506-12. 150. Wu, S.C. and R.J. Solaro, Protein kinase C zeta. A novel regulator of both phosphorylation and de-phosphorylation of cardiac sarcomeric proteins. The Journal of biological chemistry, 2007. 282(42): p. 30691-8. 151. Yuan, C., et al., Quantitative comparison of sarcomeric phosphoproteomes of neonatal and adult rat hearts. American journal of physiology. Heart and circulatory physiology, 2008. 295(2): p. H647-56. 152. Moss, R.L. and D.P. Fitzsimons, Myosin light chain 2 into the mainstream of cardiac development and contractility. Circulation research, 2006. 99(3): p. 225-7. 153. Flashman, E., et al., Cardiac myosin binding protein C: its role in physiology and disease. Circulation research, 2004. 94(10): p. 1279-89. 154. Lu, Q.W., et al., Phosphorylation of cardiac troponin I at protein kinase C site threonine 144 depresses cooperative activation of thin filaments. The Journal of biological chemistry, 2010. 285(16): p. 11810-7. 155. van der Velden, J., et al., Functional effects of protein kinase C-mediated myofilament phosphorylation in human myocardium. Cardiovascular research, 2006. 69(4): p. 876-87. 156. Haines, C., P. Harvey, and L.A. Leinwand, Estrogens Mediate Cardiac Hypertrophy in a Stimulus-Dependent Manner. Endocrinology, 2012. 157. Donaldson, C., et al., Estrogen attenuates left ventricular and cardiomyocyte hypertrophy by an estrogen receptor-dependent pathway that increases calcineurin degradation. Circulation research, 2009. 104(2): p. 265-75, 11p following 275. 158. Ball, K.L. and R.J. Solaro, Discoordinate regulation of contractile protein gene expression in the senescent rat myocardium. Journal of molecular and cellular cardiology, 1994. 26(4): p. 519-25. 159. O'Neill, L., et al., Progressive changes from young adult age to senescence in mRNA for rat cardiac myosin heavy chain genes. Cardioscience, 1991. 2(1): p. 1-5.

150

160. Izumo, S., B. Nadal-Ginard, and V. Mahdavi, Protooncogene induction and reprogramming of cardiac gene expression produced by pressure overload. Proceedings of the National Academy of Sciences of the United States of America, 1988. 85(2): p. 339-43. 161. Gardner, J.D., et al., Cardioprotection in female rats subjected to chronic volume overload: synergistic interaction of estrogen and phytoestrogens. American journal of physiology. Heart and circulatory physiology, 2008. 294(1): p. H198-204. 162. Sanna, B., et al., Modulatory calcineurin-interacting proteins 1 and 2 function as calcineurin facilitators in vivo. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(19): p. 7327-32. 163. Khoynezhad, A., Z. Jalali, and A.J. Tortolani, A synopsis of research in cardiac apoptosis and its application to congestive heart failure. Texas Heart Institute journal / from the Texas Heart Institute of St. Luke's Episcopal Hospital, Texas Children's Hospital, 2007. 34(3): p. 352-9. 164. Fidzianska, A., et al., Autophagy in transition from hypertrophic cardiomyopathy to heart failure. Journal of electron microscopy, 2010. 59(2): p. 181-3. 165. Cuervo, A.M., et al., Autophagy and aging: the importance of maintaining "clean" cells. Autophagy, 2005. 1(3): p. 131-40. 166. Gaffin, R.D., et al., Charged residue alterations in the inner-core domain and carboxy-terminus of alpha-tropomyosin differentially affect mouse cardiac muscle contractility. The Journal of physiology, 2004. 561(Pt 3): p. 777-91. 167. Moraczewska, J., K. Nicholson-Flynn, and S.E. Hitchcock-DeGregori, The ends of tropomyosin are major determinants of actin affinity and myosin subfragment 1-induced binding to F-actin in the open state. Biochemistry, 1999. 38(48): p. 15885-92. 168. Greenfield, N.J., et al., Solution NMR structure of the junction between tropomyosin molecules: implications for actin binding and regulation. Journal of molecular biology, 2006. 364(1): p. 80- 96. 169. Hilioti, Z. and K.W. Cunningham, The RCN family of calcineurin regulators. Biochemical and biophysical research communications, 2003. 311(4): p. 1089-93. 170. Hilioti, Z., et al., GSK-3 kinases enhance calcineurin signaling by phosphorylation of RCNs. Genes & development, 2004. 18(1): p. 35-47. 171. Molkentin, J.D., et al., A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell, 1998. 93(2): p. 215-28. 172. Sussman, M.A., et al., Prevention of cardiac hypertrophy in mice by calcineurin inhibition. Science, 1998. 281(5383): p. 1690-3. 173. del Monte, F., et al., Restoration of contractile function in isolated cardiomyocytes from failing human hearts by gene transfer of SERCA2a. Circulation, 1999. 100(23): p. 2308-11. 174. Kawase, Y., et al., Reversal of cardiac dysfunction after long-term expression of SERCA2a by gene transfer in a pre-clinical model of heart failure. Journal of the American College of Cardiology, 2008. 51(11): p. 1112-9. 175. Arai, M., H. Matsui, and M. Periasamy, Sarcoplasmic reticulum gene expression in cardiac hypertrophy and heart failure. Circulation research, 1994. 74(4): p. 555-64. 176. Luo, W., et al., Phospholamban gene dosage effects in the mammalian heart. Circulation research, 1996. 78(5): p. 839-47. 177. Mattiazzi, A., et al., The importance of the Thr17 residue of phospholamban as a phosphorylation site under physiological and pathological conditions. Brazilian journal of medical and biological research = Revista brasileira de pesquisas medicas e biologicas / Sociedade Brasileira de Biofisica ... [et al.], 2006. 39(5): p. 563-72. 178. Jiao, Q., et al., Sarcalumenin is essential for maintaining cardiac function during endurance exercise training. American journal of physiology. Heart and circulatory physiology, 2009. 297(2): p. H576-82.

151

179. van Deel, E.D., et al., Exercise training does not improve cardiac function in compensated or decompensated left ventricular hypertrophy induced by aortic stenosis. Journal of molecular and cellular cardiology, 2011. 50(6): p. 1017-25. 180. Rolim, N.P., et al., Exercise training improves the net balance of cardiac Ca2+ handling protein expression in heart failure. Physiological genomics, 2007. 29(3): p. 246-52. 181. Robia, S.L., et al., Localization and kinetics of protein kinase C-epsilon anchoring in cardiac myocytes. Biophysical journal, 2001. 80(5): p. 2140-51. 182. Gu, X. and S.P. Bishop, Increased protein kinase C and isozyme redistribution in pressure- overload cardiac hypertrophy in the rat. Circulation research, 1994. 75(5): p. 926-31. 183. Prekeris, R., et al., Molecular analysis of the interactions between protein kinase C-epsilon and filamentous actin. The Journal of biological chemistry, 1998. 273(41): p. 26790-8. 184. Wu, G., et al., Epsilon protein kinase C in pathological myocardial hypertrophy. Analysis by combined transgenic expression of translocation modifiers and Galphaq. The Journal of biological chemistry, 2000. 275(39): p. 29927-30. 185. Hamilton, M., et al., Constitutive association of c-N-Ras with c-Raf-1 and protein kinase C epsilon in latent signaling modules. The Journal of biological chemistry, 2001. 276(31): p. 29079-90. 186. Bueno, O.F., et al., The MEK1-ERK1/2 signaling pathway promotes compensated cardiac hypertrophy in transgenic mice. The EMBO journal, 2000. 19(23): p. 6341-50. 187. Xuan, Y.T., et al., Role of the protein kinase C-epsilon-Raf-1-MEK-1/2-p44/42 MAPK signaling cascade in the activation of signal transducers and activators of transcription 1 and 3 and induction of cyclooxygenase-2 after ischemic preconditioning. Circulation, 2005. 112(13): p. 1971-8. 188. del Monte, F., R.J. Hajjar, and S.E. Harding, Overwhelming evidence of the beneficial effects of SERCA gene transfer in heart failure. Circulation research, 2001. 88(11): p. E66-7. 189. Hasenfuss, G. and B. Pieske, Calcium cycling in congestive heart failure. Journal of molecular and cellular cardiology, 2002. 34(8): p. 951-69. 190. Loong, C.K., H.X. Zhou, and P.B. Chase, Persistence length of human cardiac alpha- tropomyosin measured by single molecule direct probe microscopy. PloS one, 2012. 7(6): p. e39676. 191. Nishikimi, T., N. Maeda, and H. Matsuoka, The role of natriuretic peptides in cardioprotection. Cardiovascular research, 2006. 69(2): p. 318-28. 192. Langenickel, T., et al., Differential regulation of cardiac ANP and BNP mRNA in different stages of experimental heart failure. American journal of physiology. Heart and circulatory physiology, 2000. 278(5): p. H1500-6. 193. Subramaniam, A., et al., Tissue-specific regulation of the alpha-myosin heavy chain gene promoter in transgenic mice. The Journal of biological chemistry, 1991. 266(36): p. 24613-20. 194. Greenfield, N.J. and S.E. Hitchcock-DeGregori, The stability of tropomyosin, a two-stranded coiled-coil protein, is primarily a function of the hydrophobicity of residues at the helix-helix interface. Biochemistry, 1995. 34(51): p. 16797-805. 195. Greenfield, N.J. and S.E. Hitchcock-DeGregori, Conformational intermediates in the folding of a coiled-coil model peptide of the N-terminus of tropomyosin and alpha alpha-tropomyosin. Protein science : a publication of the Protein Society, 1993. 2(8): p. 1263-73. 196. Rao, V.S., et al., Alterations to myofibrillar protein function in nonischemic regions of the heart early after myocardial infarction. American journal of physiology. Heart and circulatory physiology, 2007. 293(1): p. H654-9. 197. Vahebi, S., et al., p38-MAPK induced dephosphorylation of alpha-tropomyosin is associated with depression of myocardial sarcomeric tension and ATPase activity. Circulation research, 2007. 100(3): p. 408-15.

152

198. Monasky, M.M., B.J. Biesiadecki, and P.M. Janssen, Increased phosphorylation of tropomyosin, troponin I, and myosin light chain-2 after stretch in rabbit ventricular myocardium under physiological conditions. Journal of molecular and cellular cardiology, 2010. 48(5): p. 1023-8. 199. Gilabert, J.A., Cytoplasmic calcium buffering. Advances in experimental medicine and biology, 2012. 740: p. 483-98. 200. Koivumaki, J.T., et al., Modelling sarcoplasmic reticulum calcium ATPase and its regulation in cardiac myocytes. Philosophical transactions. Series A, Mathematical, physical, and engineering sciences, 2009. 367(1896): p. 2181-202. 201. Drago, I., et al., Mitochondrial Ca2+ uptake contributes to buffering cytoplasmic Ca2+ peaks in cardiomyocytes. Proceedings of the National Academy of Sciences of the United States of America, 2012. 109(32): p. 12986-91. 202. Creazzo, T.L., J. Burch, and R.E. Godt, Calcium buffering and excitation-contraction coupling in developing avian myocardium. Biophysical journal, 2004. 86(2): p. 966-77. 203. Szczesna-Cordary, D., et al., Myosin regulatory light chain E22K mutation results in decreased cardiac intracellular calcium and force transients. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 2007. 21(14): p. 3974-85. 204. Greenfield, N.J., L. Kotlyanskaya, and S.E. Hitchcock-DeGregori, Structure of the N terminus of a nonmuscle alpha-tropomyosin in complex with the C terminus: implications for actin binding. Biochemistry, 2009. 48(6): p. 1272-83. 205. Greenfield, N.J., et al., The structure of the N-terminus of striated muscle alpha-tropomyosin in a chimeric peptide: nuclear magnetic resonance structure and circular dichroism studies. Biochemistry, 1998. 37(21): p. 7834-43.

153