ABSTRACT Title of Dissertation: INTRACELLULAR REGULATION OF ATRIAL EXCITATION CONTRACTION COUPLING IN NORMAL AND ARRHYTHMOGENIC HEARTS Libet Garber, PhD. 2017 Dissertation directed by: W. Jonathan Lederer, MD, PhD Director of the Center for Biomedical Engineering and Technology Professor of Physiology University of Maryland School of Medicine Atrial fibrillation (AF) is the most common arrhythmia with a prevalence of 1-2% of the US population and it is the most important single risk factor for an ischemic stroke. Despite decades of research, successful termination of the arrhythmia remains difficult. The challenge is in part due to our incomplete understanding of atrial myocyte Ca2+ signaling and underlying disease mechanisms. In the atria, like all cardiac tissue, the conducted action potential (AP) underlies triggering of the [Ca2+]i transient, which is responsible for activating contraction. The process that links electrical activity to Ca2+ signaling and contraction is known as excitation-contraction coupling (ECC). The objective of this dissertation is to understand the mechanism of excitation contraction coupling in atrial myocytes. To achieve this goal, we (1) developed tools to specifically study atrial cell biology, (2) we studied the role of altered Ca2+ buffering on ionic membrane currents and Ca2+ signaling, (3) we investigated the role that reactive oxygen species (ROS) plays in altered Ca2+ signaling and the morphology of the AP and (4) we measured intracellular sodium concentration ([Na+]i ) and studied Na+ and Ca2+ signaling in a transgenic murine model of AF. This work includes mathematical modeling of atrial cell electrical and Ca2+ signaling to define our quantitative understanding of the processes involved. Our results indicate that increased Ca2+ buffering plays a major role in speeding the inactivation of the L type Ca2+ current (ICa,L ). This work also shows that low concentrations of H2O2 for a brief period increases atrial Ca2+ spark rate, changes spark characteristics and decreases the duration of the AP. We quantified for the first time the [Na+]i in murine atrial cells both at rest and during field stimulation in control and transgenic mice. Our results indicate that [Na+]i is significantly lower in atrial myocytes in comparison to their ventricular counterparts, which reveal important differences in how [Na+]i is regulated in atrial cells. Moreover, our work demonstrates that [Na+]i and [Ca2+]i homeostasis are profoundly affected during AF. The results further our understanding of mechanisms that modulate excitation-contraction coupling in atrial myocytes in normal and pathophysiological conditions. INTRACELLULAR REGULATION OF ATRIAL EXCITATION CONTRACTION COUPLING IN NORMAL AND ARRHYTHMOGENIC HEARTS By Libet Garber Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2017 Advisory Committee: Professor W. Jonathan Lederer, MD, PhD., Chair Professor William Bentley, PhD. Assistant Professor Steven Jay, PhD. Associate Professor Isabel Lloyd, PhD., (Dean’s Representative) Adjunct Professor Richard Gray, PhD. © Copyright by Libet Garber 2017 ii Dedication To my loving husband Steve: Your unwavering support has encouraged me to run this race to the finish line. To my little daughter Maria Ester: Your joyful and exploring spirit moves me to listen to the whispers of beauty God has hidden for us to find. To my expected little son: Although only visible to the One who forms you, I can feel your gentle moves that are pulling my heartstrings. Not to us, O LORD, not to us, But to Thy name give glory Because of Thy loving kindness, because of Thy truth. Psalm 115:1, NASB iii Acknowledgements I want to thank my advisor, Dr. W. Jonathan Lederer for the opportunity to be part of the Lederer lab. I appreciate the knowledge and insight he has shared and also the passion and joy to always keep pursuing something new and challenging. I would like to thank my colleague and mentor Dr. Maura Greiser for sharing and providing direction through her knowledge and expertise of atrial cell biology and Dr. George S.B. Williams for his mentorship and guidance in the development of the mathematical modeling components used in this dissertation. I want to thank my committee members for all of their support and comments, especially Dr. Gray for introducing me to the field of cardiac mathematical modeling, Dr. Lloyd for always having a listening and caring personality and for sharing her material science expertise to help me find the aluminum oxide fibers, Dr. Bentley for supporting my doctorate career since I enrolled at the University of Maryland and Dr. Jay for his challenging and insightful feedback along my research path. I would like to thank Dr. Humberto Joca for his useful discussions, experimental support and his contribution to the whole cell Ca2+ epifluorescence experiments. I am grateful to Dr. Moradeke Bamboye, Dr. Patrick Robinson, Dr. Guiling Zhao and Dr. Brian Hagen for their help in teaching ventricular cell isolation and patch clamping. I would like to acknowledge the generous funding I have received, in particular the Sloan Foundation Graduate Scholarship, the Interdisciplinary iv Training Program in Muscle Biology Training Grant from NIH (T32 AR007592) and the Cardiovascular Disease Training Grant (2T32HL007698-22A1) from NIH. v Table of Contents Dedication .......................................................................................................... ii Acknowledgements ........................................................................................... iii Table of Contents .............................................................................................. v List of Tables .................................................................................................... xi List of Figures .................................................................................................. xii List of Abbreviations ....................................................................................... xiv Chapter 1 : Introduction and Background .......................................................... 1 1. Overview ................................................................................................ 1 Heart Structure .......................................................................................... 1 Electrical Propagation and Contraction in the Heart ................................. 2 Electrical Contraction Coupling in a Cardiac Myocyte .............................. 4 Distinctiveness of Atrial Myocyte Structure ............................................... 8 Atrial Myocyte Calcium Transient ........................................................... 11 Ca2+ Buffering Effects on ECC ................................................................ 13 ROS Effects on Ca2+ Signaling ............................................................... 13 Atrial Myocyte Na+ Signaling ................................................................... 14 2. Current Techniques ............................................................................. 15 Myocyte Sources .................................................................................... 15 Ca2+ Imaging ........................................................................................... 16 Action Potential and Current Measurements .......................................... 18 3. Atrial Fibrillation ................................................................................... 20 Chaotic Rhythm ...................................................................................... 20 vi Significance of Studying AF Mechanism ................................................. 20 Current Hypothesis ................................................................................. 21 Cellular Modulators of Atrial Fibrillation .................................................. 22 Ca2+ Signaling Silencing ......................................................................... 24 4. Computational Modeling ...................................................................... 25 5. Conclusion ........................................................................................... 26 Chapter 2 : A Novel Approach to Isolate, Position and Stabilize Atrial Myocytes for Biological Experiments ............................................................................... 27 1. Introduction .......................................................................................... 27 Single Atrial Myocytes Isolation .............................................................. 27 Stabilization of Atrial Myocytes ............................................................... 29 2. Materials and Methods ......................................................................... 34 Atrial Cell Isolation from Mice ................................................................. 34 Atrial Cell Isolation from Rabbit ............................................................... 35 Procedure for Verifying Fiber Stiffness ................................................... 38 Validation of Fiber Suitability to Stabilize Atrial Myocytes ....................... 39 Statistical Analysis .................................................................................. 40 3. Results ................................................................................................. 42 Cell Isolation Quality ............................................................................... 42 Rod Material Verification ......................................................................... 45 Sapphire Rod Validation ......................................................................... 47 4. Discussion ............................................................................................ 49 5. Conclusion ........................................................................................... 51 vii Chapter 3 : Buffering Effects on the L-Type Ca2+ Channel Current and Ca2+ Dynamics in Atrial Cardiomyocytes. ................................................................ 52 1. Introduction .......................................................................................... 52 2. Methods ............................................................................................... 54 Cell Isolation ........................................................................................... 54 Electrophysiology Experiments ............................................................... 55 Mathematical Compartmental Modeling ................................................. 56 Mathematical Compartmental Stochastic Modeling ................................ 57 Mathematical Spatial Modeling ............................................................... 58 Statistical Analysis .................................................................................. 58 3. Results ................................................................................................. 61 Buffering Effects on the Rabbit Atria LCC Current .................................. 61 Compartmental Modeling Simulations of the LCC Current ..................... 64 Simulations Using Compartmental Stochastic Model of CICR ............... 67 Computational Model Simulation of Spatiotemporal Ca2+ Dynamics ...... 69 4. Discussion ............................................................................................ 71 5. Conclusions ......................................................................................... 72 Chapter 4 : Effects of Reactive Oxygen Species on Calcium Induced Calcium Release and the Action Potential ..................................................................... 74 1. Introduction .......................................................................................... 74 2. Methods ............................................................................................... 77 Cell Isolation ........................................................................................... 77 Confocal Ca2+ imaging ............................................................................ 77 viii Ca2+ Spark Measurements under Controlled H2O2 Delivery .................... 77 Action Potential Measurements .............................................................. 80 Mathematical Compartmental Modeling ................................................. 81 Mathematical Spatial Modeling ............................................................... 82 Statistical Analysis .................................................................................. 83 3. Results ................................................................................................. 84 Reactive Oxygen Effects in Ca2+ Spark Rate .......................................... 84 Reactive Oxygen Effects in Ca2+ Spark Characteristics ......................... 88 Ca2+ Spark Characteristics Comparison Between Atrial and Ventricular Rabbit Myocytes ..................................................................................... 92 Reactive Oxygen Effects in Action Potential Duration ............................ 94 4. Discussion .......................................................................................... 100 5. Conclusion ......................................................................................... 101 Chapter 5 : Arrhythmogenic Na+/Ca2+ Imbalance in a Murine Transgenic Model of AF .............................................................................................................. 103 1. Introduction ........................................................................................ 103 2. Methods ............................................................................................. 106 Transgenic Murine Model of AF ............................................................ 106 Cell Isolation ......................................................................................... 107 Intracellular Sodium Indicator Cell Loading .......................................... 107 [Na+]i Measurements and Instrumentation ............................................ 107 [Na+]i Calibration ................................................................................... 110 Anemonia Sulcata Toxin (ATX-II) Experiments .................................... 112 ix [Na+]i Extrusion in Atrial Myocytes ........................................................ 112 [Ca2+]i Measurements ........................................................................... 113 Modeling of [Na+]i Homeostasis in Atrial Myocytes ............................... 114 3. Results ............................................................................................... 116 Imaging Parameters Validation ............................................................. 116 [Na+]i in Atrial vs. Ventricular Myocytes ................................................. 119 Frequency-dependent Increase in [Na+]i ............................................... 121 Late INa Effect on [Na+]i in Atrial Myocytes ............................................ 123 [Na+]i Extrusion in Atrial Myocytes ........................................................ 126 [Na+]i Homeostasis in Murine Atrial Myocytes from a Transgenic Model of AF ......................................................................................................... 128 [Ca2+]i Homeostasis in Murine Atrial Myocytes from a Transgenic Model of AF ..................................................................................................... 132 4. Discussion .......................................................................................... 136 5. Conclusion ......................................................................................... 139 Chapter 6 : Discussion and Future Direction ................................................. 141 Introduction ........................................................................................... 141 Novel Atrial Myocytes Biological Techniques ....................................... 141 Effect of Ca2+ Buffering on Atrial Myocyte Pathophysiology ................. 141 Effect of ROS Signaling on Electrical Contraction Coupling ................. 142 Experimental Observations in a Murine Model of AF ............................ 143 Future Work .......................................................................................... 144 Conclusion ............................................................................................ 145 x Bibliography ................................................................................................... 146 Curriculum Vitae ............................................................................................ 162 xi List of Tables Table 1.1. Major Cardiac Membrane Currents ................................................. 7 Table 2.1. Mouse Cardiac Cell Isolation Buffer .............................................. 37 Table 2.2 Tyrode solution used for storage and experimentation .................. 37 Table 2.3. Theoretical Young’s modulus of rod materials. ............................. 39 Table 3.1. ICa,L measurement solutions. ......................................................... 56 Table 4.1. Solution for AP recordings. .......................................................... 81 Table 4.2. Spatial rabbit Ca2+ spark modeling parameters. ........................... 83 Table 5.1. [Na+]i Calibration Solutions ......................................................... 110 Table 5.2. Sodium Extrusion Solutions ........................................................ 113 Table 5.3. Na+,K+ ATPase Pump Modeling Parameters .............................. 115 Table 5.4. Murine Atrial Myocyte Modeling Parameters .............................. 116 xii List of Figures Figure 1.1. Cardiac anatomy and cardiac conduction system ......................... 3 Figure 1.2. Excitation Contraction Coupling Representation. .......................... 6 Figure 1.3. Myocyte Structure ........................................................................ 10 Figure 1.4. TT and [Ca2+]i Transient in Atrial vs. Ventricular Myocytes. ......... 12 Figure 1.5. Confocal Imaging of Ca2+ Sparks. ................................................ 17 Figure 1.6. Whole Cell Patch Clamp Technique ............................................ 19 Figure 2.1. Rod Stabilization .......................................................................... 33 Figure 2.2. Verification Protocol ..................................................................... 41 Figure 2.3. Cell Viability ................................................................................. 43 Figure 2.4. Cell Properties ............................................................................. 44 Figure 2.5. Rod Stiffness ................................................................................ 46 Figure 2.6. Murine Sarcomere Length Validation ........................................... 48 Figure 3.1. Spatial Model Geometry ............................................................. 60 Figure 3.2. Experimental ICa,L Under Different Buffering Conditions. ............. 62 Figure 3.3. Experimental ICa,L Statistics . ........................................................ 63 Figure 3.4. Computationally Modeled ICa,L ..................................................... 65 Figure 3.5. Computationally Simulated ICa,L. ................................................. 66 Figure 3.6 Stochastically Simulated CICR under EGTA ................................ 68 Figure 3.7 Spatial Modeling Simulation of Ca2+ Diffusion ............................... 70 Figure 4.1. Spatiotemporal Control of H2O2 Delivery. .................................... 79 Figure 4.2. Ca2+ Spark Rates in Cells Treated with 20 µM H2O2 ................... 85 Figure 4.3 Spatially Modeled Ca2+ Sparks ..................................................... 87 xiii Figure 4.4. Rabbit Atrial Ca2+ Spark Characteristics ...................................... 90 Figure 4.5. Spatially Modeled Ca2+ Spark Characteristics ............................. 91 Figure 4.6. Ca2+ Spark Characteristics in Rabbit Ventricular and Atrial Myocytes ................................................................................................. 93 Figure 4.7. ROS Effect on the AP of Rabbit Myocytes ................................... 95 Figure 4.8. Computational IK,ATP in Atrial and Ventricular Model. ................... 98 Figure 4.9. Computationally Modeled ROS Effects on the AP ....................... 99 Figure 5.1. Light Path Schematic Used in [Na+]i Measurements .................. 109 Figure 5.2. In situ SBFI Calibration. ............................................................. 111 Figure 5.3. Imaging Set-up Validation .......................................................... 118 Figure 5.4. [Na+]i in Quiescent Atrial and Ventricular Myocytes. .................. 120 Figure 5.5. Frequency-dependent Increase in [Na+]i .................................... 122 Figure 5.6. Late INa Effect on [Na+]i in Atrial Myocytes ................................. 125 Figure 5.7. Na+ Extrusion in Control Murine Atrial Myocytes ....................... 127 Figure 5.8. [Na+]i in Atrial Myocytes from a Transgenic AF Murine Model ... 130 Figure 5.9. Computationally Modeled [Na+]i in Atrial Myocytes ................... 131 Figure 5.10. Experimental [Ca2+]i in Atrial Cells from an AF Murine Model . 134 Figure 5.11. Computationally Modeled [Ca2+]i in Atrial Myocytes ................ 135 xiv List of Abbreviations AF Atrial fibrillation ADP Adenosine Di-phosphate AP Action Potential APD Action Potential Duration ATP Adenosine Tri-Phosphate ATX-II Anemonia sulcata toxin AV Atrioventricular [Ca2+]i Intracellular calcium concentration CaM Calmodulin CRU Ca2+ release unit ECC Excitation-contraction coupling FLAG-F1759A-NaV1.5 Double transgenic murine AF model FDHM Full duration at half maximum FWHM Full width at half maximum GPa Giga Pascal ICa,L L type Ca2+ current INa Voltage gated sodium channel current (NaV 1.5) INCX Sarcolemmal Na+/Ca2+ exchanger current INa,Late Persistent INa IK,ATP ATP sensitive potassium current IK1 Inward rectifier potassium current [K+]i Intracellular potassium concentration xv [K+]o Extracellular potassium concentration LTCCs L-type Ca2+ channels µN micro Newton [Na+]i Intracellular sodium concentration NAC N-acetylcysteine NaV1.5 Voltage gated sodium channel, cardiac type NCX Sarcolemmal Na+/Ca2+ exchanger NKA Sodium-Potassium ATPase NOX2 NADPH oxidase, type 2 ROS Reactive oxygen species RyR2s Ryanodine receptors, type 2 RAP Rapid atrial pacing SAN Sinoatrial node SR Sarcoplasmic reticulum SERCA Sarcoplasmic reticulum Ca2+ ATPase SL Sarcolemma TTs Transverse tubules 1 Chapter 1 : Introduction and Background 1. Overview Through the entire lifespan of an individual the heart rhythmically contracts and relaxes to keep blood flowing through the entire cardiovascular system. This complex organ is an electro-mechanical pump, in which electrical signals are translated into mechanical contraction through a process called excitation contraction coupling (ECC). The processes that lead to disturbances of heart rhythms or arrhythmogenesis in the heart are not fully understood, and this thesis will be looking at both normal heart function and dysfunction. We will be focused on the cellular function of the atrial tissue of the heart, which is located in the heart’s upper chambers. The objective of this dissertation is to understand the mechanism of excitation contraction coupling in atrial myocytes than when deregulated would lead to Atrial fibrillation (AF), a common arrhythmia. Heart Structure The heart consists of 4 chambers; the two upper chambers are called the atria and the two lower chambers are called the ventricles. The right atrium and ventricle work to pump de-oxygenated blood to the lungs to be oxygen replenished. The oxygenated blood is then returned to the left atrium and ventricle of the heart to be pumped into the whole systemic circulation through the aorta. 2 The phase in which the heart is relaxed and its chambers are filling with blood is termed diastole, followed by an active phase when the heart contracts and pumps its blood termed systole. Electrical Propagation and Contraction in the Heart The event that originates excitation-contraction coupling is the action potential (AP) (1). In healthy hearts electrical impulse formation is initiated in the sinoatrial (SA) node, the heart’s primary pacemaker. SA nodal cells depolarize spontaneously due to the activation of specific ion currents during diastole. The AP that is thus initiated propagates to neighboring cells through gap junctions (2) triggering the [Ca2+]i transient which then activates contraction. Figure 1.1 depicts the electrical propagation from the SA node to atrial tissue. Since the electrical propagation reaches the atrial tissue first, the atria contract before the ventricles. This is named atrial kick because it causes the filling ventricles to increase their blood volume by 15 % which increases the pumping effectiveness of the heart (3). After reaching the atrial tissue, the electrical propagation is conducted to the atrioventricular (AV) node and then to ventricular tissue through the fast conducting Purkinje fiber network. 3 Figure 1.1. Cardiac anatomy and cardiac conduction system (A) Normal cardiac propagation. RA indicates right atrium; TV, tricuspid valve; MV, mitral valve; LV, left ventricle; LA, left atrium. (B) Representation of cardiac conduction during AF. (Image modified from Munshi et al.(4), with permission from Wolters Kluwer Health, Inc.) 4 Electrical Contraction Coupling in a Cardiac Myocyte Cardiomyocytes, or cardiac cells, are the basic contractile unit that makes up the heart muscle. Figure 1.2 shows the process of excitation contraction coupling in a cardiac myocyte. During systole an AP activates the voltage gated sodium channels current (INa), which causes the cell to quickly depolarize (5). The L-type Ca2+ channels (LTCCs) are located in the sarcolemma (SL) and open as a result of the depolarization that occurs during the AP. The Ca2+ influx comes primarily into a restricted space between the membrane and the junctional sarcoplasmic reticulum (SR), known as the "subspace", which is only about 12 to 20 nm wide (6). The small volume of the subspace means that the small Ca2+ influx into the subspace through the LTCC current (ICa,L) produces a large change in [Ca2+]subspace (7-9). On the other side of the subspace, opposite to the LTCC, are the ryanodine receptors (RyR2s), which are the intracellular calcium release channels located in the SR, arranged in clusters called a Ca2+ release unit (CRU)(10,11). The CRU is triggered to open by the large [Ca2+]subspace signal. This constitutes the Ca2+ signaling amplification mechanism known as Calcium-Induced Calcium Release (CICR) (12). The myofilaments are the contractile machinery of the cell and when triggered by Ca2+ they are responsible for converting chemical energy into mechanical energy for cell contraction. The myofilaments are organized in the sarcomere, which is the fundamental contractile unit, whose boundaries are the Z-lines, where the actin or thin filaments attach. The spatiotemporal 5 summation of individual Ca2+ releases give rise to the cell-wide [Ca2+]i transient. This increase in cytosolic Ca2+ allows the Ca2+ in the cytoplasm to bind to cardiac troponin-C, which moves the troponin complex away from actin-binding sites of the myosin filaments. This removal of the troponin complex frees the actin to be bound by myosin and to slide to bring the Z- lines closer together, decreasing the sarcomere length and initiating contraction (13). Once Ca2+ is released from the myofilaments relaxation occurs, and Ca2+ is quickly removed from the cytosol by four main mechanisms: 1) the sarcoplasmic reticulum Ca2+ ATPase (SERCA) that sequesters Ca2+ back to the SR (14), 2) sarcolemmal Na/Ca exchanger (NCX) that is responsible for extruding most of the Ca2+ from the cell (15) with a very small contribution by 3) the sarcolemmal Ca2+ ATPase, and finally there is a small amount of Ca2+ that enters the mitochondria through 4) the Ca2+ uniport system (13,16). Meanwhile, in the cell membrane the potassium channels activate and give rise to outward currents that are responsible for repolarization, or the returning of the membrane potential to its negative baseline level (17). Table 1.1 shows the major cardiac plasma membrane currents that are involved in generating the action potential. 6 Figure 1.2. Excitation Contraction Coupling Representation. After an AP, INa is activated which leads to membrane depolarization which activates the ICa,L, allowing Ca2+ to enter the cell. This Ca2+ influx activates the RyR2s leading to SR Ca2+ release into the cytosol. Ca2+ binds to the myofilaments to initiate contraction. Relaxation occurs once Ca2+ is released from the myofilaments and sequestered back to the SR through SERCA and extruded from the cell mainly through NCX. 7 Current Name Channel Protein INa Na+ current Nav1.5 (voltage gated Na+ channel) ICa Ca2+ current Cav1.2 (L-Type Ca2+ channel) IK Repolarizing IKr Repolarizing IKs Ito G-protein activated ATP-sensitive HERG + miRP1* KvLQT1 +minK* Kv4.3 (voltage gated K+ channels) GIRK* Kir6.2 + SUR2A = KATP* If (Na+ + K+) Pacemaker current HCN4 Table 1.1. Major Cardiac Membrane Currents *These are heteromultimeric channels HERG, human ether-a-go-go-related gene (related to the Kv family of K+ channels); GIRK, G-protein activated inwardly rectifier K+ channel; HCN4, Hyperpolarization-activated and cyclic nucleotide–gated channel 4 (Modified from Lederer (18), used with author’s permission) 8 Distinctiveness of Atrial Myocyte Structure Atrial myocytes possess many of the structural properties found in ventricular myocytes. As shown in panel A of Figure 1.3, both atrial and ventricular myocytes have Z-lines, which form the boundary of the sarcomere or contractile unit. The mitochondria are also located between the Z-lines; which provides the local proximity to the adenosine triphosphate (ATP) energy source needed for contraction. The RyR2 or Ca2+ release channels of the sarcoplasmic reticulum are also located near the Z lines in both cells types (13). In adult ventricular myocytes, the L-type Ca2+ channels are located in the sarcolemma (SL). They are also located in the transverse tubules (TTs), which are invaginations of the SL organized along the Z lines. Structural studies have shown that LTCCs face the CRU, which has longitudinal, or Z- line spacing, of 1.87 µm and axial spacing of about 1.05 µm (19). This means that the Ca2+ release from the CRUs inside the ventricular myocytes can be synchronized by the activation of the LTCC during the depolarization that is conducted axially through the TTs (20,21). Various studies have reported the presence of TTs in atrial myocytes isolated from sheep (22-24). However, they are less abundant than in their ventricular counterparts. Additionally, Lenaerts et al. demonstrated that TTs density is further reduced in atrial myocytes isolated from a sheep model of atrial fibrillation (23). While Richards et al. reported the presence of t-tubules in atrial cells of large mammals like sheep, cow and horse, their study found 9 that the presence of t-tubules in human atrial cells depended on cell size and in many small human atrial cells they did not find any TTs (25). This is consistent with imaging studies that have shown that TTs are sparser in atrial myocytes in comparison to their ventricular counterparts in many species, including rabbits, and for many species atrial myocytes do not have any TTs. Figure 1.3 depicts this spatial organization in an atrial myocyte that lacks organized TTs and shows that the LTCCs are located in the SL membrane at the cell periphery. Although sparser than in the ventricle, it is worth noting that murine atrial myocytes do posses a significant amount of TTs as shown in our analysis presented in panel A of Figure 1.4 and as has been demonstrated in the literature (26,27). Thus in this dissertation, we have conducted experiments using both rabbit and mouse atrial myocytes to capture the different aspects of Ca2+ signaling that can be revealed by the structural differences between the atrial cells in these two species. 10 Figure 1.3. Myocyte Structure (A) Scanning electron micrograph images showing the relevant structures in a cardiac myocytes. (Modified from Kostin et al. and Frank JS (10,11), used with permission from Springer) (B) Di-8 ANEPP membrane staining portrays the lack of T-tubules in rabbit atrial myocytes. Despite the lack of these membrane invaginations, atrial myocytes have the organized RyR2 along the Z-lines as shown by the RyR2 antibody staining. Right schematics: Green elements represent LTCCs; pink elements, the RyR2s; blue region, the SR. (Modified from Greiser et al. (28,29), used with permission from Oxford University Press). 11 Atrial Myocyte Calcium Transient The lack of an abundant and well-defined atrial TTs in many species like cat (20), rabbit and humans (30,31) has implication for the spatio-temporal characteristics of the [Ca2+]i transient in these cells. The reduction or lack of TTs in atrial cells causes the [Ca2+]i transient to rise first at the cell periphery and then spread to the center of the cell. As shown in panel B of Figure 1.4, the amplitude of the central [Ca2+]i transient is equivalent to the subsarcolemmal transient. This indicates that this process does not rely solely on passive diffusion, but involves additional Ca2+ releases as the [Ca2+]i transient propagates spatially and activates inner CRUs inside the atrial myocytes through CICR (20,32). Interestingly, recent atrial myocyte structural studies seem to indicate that the axial spacing of the CRU in atrial cells is less than 600 nm and that CRUs are spaced closer in murine atrial cells than ventricular cells, which could enhance CICR (27). Moreover, a new mechanism of enhanced CICR was discovered in rabbit atrial myocytes that involves sensitization of the RyR2 by SERCA uptake at the [Ca2+]i transient wavefront, causing the local SR [Ca2+]SR to increase thereby lowering the activation threshold for RyR2 to open in response to the activating cytosolic [Ca2+]i. This mechanism was termed “Fire-Diffuse-Uptake-Fire” to signify both the RyR2 activation and sensitization by cytosolic and luminal SR calcium respectively (32). 12 Figure 1.4. TT and [Ca2+]i Transient in Atrial vs. Ventricular Myocytes. (A) Our experimental data showing that rabbit atrial myocytes posses an insignificant number of TTs. Although TTs are present in mice atrial cells, there are sparser than in their ventricular counterparts. (B) This image was obtained by a transverse line scan with a confocal microscope showing the delay in the Ca2+ transient in the atrial cell center (figure modified from (13) and (20) with permission from John Wiley and Sons, Inc.) Rabbit Atria Mouse Atria Mouse Ventricle 0 20 40 60 % T T A R EA A B 13 Ca2+ Buffering Effects on ECC ICa,L is the most important membrane current involved in ECC. It is also known that this current’s amplitude is reduced during atrial fibrillation, although the molecular mechanisms are not fully understood (e.g., protein expression, phosphorylation, phosphatase effects) (29,33). ICa,L has two modes of inactivation that act as a negative feedback mechanism to stop the influx of Ca2+ into the cell (34). The slower one is voltage inactivation and the faster acting is a Calmodulin (CaM) mediated Ca2+-dependent inactivation (CDI) (35-37). Ca2+ buffering affects CDI directly by limiting the Ca2+ ions available to bind to CaM. Moreover, Ca2+ buffering affects the availability and diffusion of Ca2+ in the cytosol. Michailova et al. used experimental data to develop a mathematical model of Ca2+ buffering and diffusion in an atrial cardiomyocytes with inhibited SR. Their model highlighted the importance of physiological Ca2+ buffers in the propagation of the [Ca2+]i transient (38). ROS Effects on Ca2+ Signaling The effect of ROS on Ca2+ signaling is a well studied but controversial field (39,40). Many studies have reported the harmful effects of increase oxidative stress during atrial fibrillation (40-43) and others have found that physiological levels of ROS play an important function in normal Ca2+ signaling (40,44-46). The novel mechanism discovered by Prosser et al. is an example of physiological levels of stretch-dependent ROS sensitizing and fine-tuning excitation contraction coupling (45). 14 Moreover, ROS modulates enzymes that are critical to the regulation of cellular Ca2+ (40). For example Guo et al. showed that the peak ICa,L was increased after exposure to 100 µM H2O2 and Viola et al. observed a similar effect after exposure to 30 µM H2O2 (47,48). However, other studies have not found any effect of ROS on the ICa,L parameters (40,49), but instead have found that exposure to 200 µM H2O2 led to significant reduction of the SR Ca2+ content. H2O2 was chosen as the experimental ROS chemical in this project based on its cell permeability and stability (40,50). Atrial Myocyte Na+ Signaling Intracellular sodium concentration ([Na+]i) is an important regulator of intracellular Ca2+ and provides insight into the activation of the sarcolemmal sodium calcium exchanger (NCX) and the behavior of voltage gated Na+ channels (NaV1.5) and the Na+,K+-ATPase. There are a number of membrane channels and transporters that affect [Na+]i (51), but those mentioned above are the primary ones. In the contracting myocyte, the major Na+ influx pathways are NCX and the voltage gated sodium channel (NaV1.5), while the major Na+ extrusion pathway is the Na+,K+-ATPase (NKA). Upon the arrival of an action potential NaV1.5 channels activate allowing the fast inward INa current to quickly depolarize the cell membrane. After opening briefly during the upstroke of the AP, each NaV1.5 promptly inactivates until repolarization is concluded. In contrast, NCX is an ATP independent exchanger that plays an important role in extruding Ca2+ from 15 the cell. Because it exchanges one Ca2+ ion for 3 Na+ ions, it is electrogenic, producing a depolarizing current, INCX (52). NKA maintains Na+ and K+ gradients across the cell membrane by actively extruding 3 Na+ ions from the cell in exchange for 2 K+ ions into the cell by utilizing the energy derived from the hydrolysis of one ATP molecule (53,54). This is an important role because the K+ gradient is the most significant determinant of the cell’s membrane potential, and the Na+ gradient is the driving force behind many crucial ion-exchange processes, which include the extraction of calcium via (NCX) and the extraction of H+ from the cytosol via the Na+/H+ exchanger to maintain the cell’s pH (53,55,56). 2. Current Techniques Myocyte Sources We require cells that have the ECC characteristics of adult cardiac myocytes to adequately represent the biological process. Currently there are no cell lines that continue to actively contract over time; even short-term cultures quickly de-differentiate and loose this ability. Other possible sources include human embryonic stem cell (hESC)-derived atrial CMs (57) and human induced pluripotent stem cell (iPSC)-derived cardiomyocytes (58). However, various studies that attempted to characterize these cells have found numerous limitations with many factors affecting their morphological and molecular characterization, as wells as their ionic current and contractile properties (59-61). Human sources of fresh cells are naturally hard to come 16 by, so therefore, at the moment, freshly isolated single atrial myocytes from animal cardiac tissue offers the best representation of physiological function in cardiac investigations. Single myocytes from animal models are thus an important resource for biological investigation of cardiac disease. Ca2+ Imaging The synchronized opening of clustered RyR2s forming the CRU results in elevations of local [Ca2+] known as ‘‘Ca2+ sparks,’’ which were first discovered in cardiac myocytes and are the elementary Ca2+ release events underlying ECC (62). During diastole and in unstimulated cardiac myocytes, there is no LTCC Ca2+ influx, but spontaneous Ca2+ sparks can be observed due to the stochastic nature of RyR2 openings. An increase in spontaneous RyR2 openings results in SR Ca2+ leak, or loss Ca2+ from the SR, which can lead to Ca2+ signaling instabilities promoting arrhythmogenesis in cardiac myocytes (63). In this dissertation we have used Ca2+ spark rate as a metric of diastolic Ca2+ release. Figure 1.5 depicts how a Ca2+ spark and [Ca2+]i transient can be imaged using confocal laser scanning microscopy. The laser beam is focused on a line that moves back and forth across the selected sample. The lines are then integrated into an image that represents the spatial profile in the y-axis as time progressed in the x-axis. This technique has the advantage of providing useful spatiotemporal information about Ca2+ signaling. 17 Figure 1.5. Confocal Imaging of Ca2+ Sparks. (A) Two-dimensional confocal images of Ca2+ sparks in an unstimulated cardiac myocyte (scan rate 1.0 s/frame). (B) Line scan confocal images of an action potential (AP)-elicited [Ca2+]i transient (top) and a spontaneous spark (bottom) (scan rate 2.0 ms/line). (Images from Cheng and Lederer(64), Physiological Reviews allows reuse in thesis and dissertations without special permission). 18 Action Potential and Current Measurements The patch clamp technique was first developed by Sakmann and Neher (65) to measure single channel currents. It was modified to the conventional whole cell patch clamp technique by the same group as described by Hamill et at. (66), which can also measure AP. Figure 1.6 depicts the whole cell patch clamp technique; briefly, a clean glass micropipette can fuse with the cell membrane to create a seal with resistance in the giga-ohm range. The seal is created by gently approaching the cell membrane with a pipette and applying suction until a 2 GΩ seal is achieved. Access into the cell is then gained by gently rupturing the cell membrane with rapid suction (17). The solution inside the pipette is made to resemble the ionic composition of the cytosol and the bath solution is made to resemble the extracellular environment. The patch pipette acts as an electrode connected to the patch clamp amplifier, which is also connected to the bath electrode. In the voltage clamp mode, the amplifier controls the voltage, and the whole cell currents flowing between the patch pipette electrode and the bath electrode are recorded. In the current clamp mode, the membrane potential is allowed to vary freely and action potentials are measured by injecting the cell with a depolarizing current signal. 19 Figure 1.6. Whole Cell Patch Clamp Technique A glass micropipette is fused with the cell membrane to create a seal with high resistance. Access into the cell is gained by rupturing the cell membrane with gentle suction. Currents or membrane potentials are measured with the amplifier patch clamp system. (Image modified from (67) and (68) and used with permission from Discovery Medicine) 20 3. Atrial Fibrillation Chaotic Rhythm During atrial fibrillation (AF), the normal propagation from the SA node to the atria is overridden by a rapid and abnormal chaotic electrical propagation in the atrial tissue. This disordered electrical propagation prevents the atrial tissue from contracting in synchrony leading the atria to fibrillate. This unsynchronized contraction has two adverse consequences. First, there is a loss of the atrial kick, reducing the pumping effectiveness of the heart (69). Second and most significant, the fibrillating atria lead to blood stagnation, which increases the risk of coagulum formation and thromboembolism, forerunners to stroke (70). Significance of Studying AF Mechanism AF is the most common cardiac arrhythmia with a prevalence of 1-2% in the general population, increasing to 10% in people older than 80 years of age (71,72). Demographic studies project that there will be over 12 million cases of AF by 2030 with an annual growth rate of 4.3% for AF prevalence and 4.6% for AF incidence over the period of 2010 to 2030 (73). This is a huge socioeconomic burden, with a cost of $6 billion per year in the Unites States (74,75). The high prevalence and incidence of AF is a major healthcare concern because AF is also the most important single risk factor for ischemic stroke. Despite decades of research, successful termination of the arrhythmia 21 remains difficult. Current treatments encompass a range of non-invasive to highly invasive options and include medical treatments, endovascular catheters and surgical variants (76). Available antiarrhythmic medical therapies for the treatment of AF are suboptimal due to modest efficacy, as well as proarrhythmic and noncardiovascular toxicities of current antiarrhythmic drugs (77). The surgical and catheter based ablation are only recommended for use when medical therapy has failed, due to increased safety risks of invasive procedures and their limited long term success (78). The above treatment challenges are in part due to our incomplete understanding of normal atrial myocyte Ca2+ signaling and any underlying disease mechanisms (79). Current Hypothesis There are two widely accepted theories on the mechanism of AF. The first involves a reduction of the AP duration, which enables reentrant electrical activity within the atrial tissue. The second one involves instable calcium signaling, which increases the likelihood of delayed after-depolarizations (DADs) that can lead to triggered activity that promotes AF (80). DADs form when there is an increase diastolic calcium release from the SR, this additional calcium triggers NCX to extrude each Ca2+ ion for three Na+ ions, creating an inward current that raises the membrane potential. If the membrane potential crosses the depolarization threshold, the DAD can trigger an action potential that is not in synchrony with the normal SA node propagation. 22 Cellular Modulators of Atrial Fibrillation In this work, we are interested in three modulators of Ca2+ signaling in atrial cells: Ca2+ buffering, reactive oxygen species and intracellular sodium concentration ([Na+]i). Ca2+ buffering ICa,L is the most important membrane current involved in ECC. It is also known that this current’s amplitude is reduced during atrial fibrillation, although the molecular mechanisms are not fully understood (e.g., protein expression, phosphorylation, phosphatase effects) (29,33). Greiser and colleagues showed that Ca2+ buffering plays an important role in the cellular adaption that occurs under rapid atrial pacing in rabbit atrial myocytes (31). In particular, the results show that fast Ca2+ buffering strength was increased as one of the adaptation mechanism that occurs to bring Ca2+ signaling stabilization (81). Reactive oxygen species Recently, the Lederer laboratory discovered a novel mechanism linking cellular stretch in ventricular myocytes to the facilitation of intracellular Ca2+ release (45). This mechanism, "X-ROS signaling", depends on NOX2 (NADPH oxidase) in the sarcolemmal and TT membranes to generate reactive oxygen species (ROS). ROS appear to oxidize the nearby Ca2+ release channels of the intracellular Ca2+ storage organelle and increase their 23 sensitivity to [Ca2+]i. This suggests that this mechanism may also be operational in atrial cells, since other researchers have reported an increase in ROS during AF, and conditions like ischemia and oxidative stress are associated with an increased incidence of AF (82-86). Intracellular sodium homeostasis In cardiac diseases such as atrial fibrillation (AF), intracellular Ca2+ signaling is profoundly altered (79,87-90). While the mechanisms underlying altered intracellular Ca2+ homeostasis are well characterized, the role that [Na+]i may play and its dysregulation in cardiac disease are less well understood. For example, Greiser at al. have previously shown that high rate (10 Hz) atrial activation for 5 days in a rabbit model leads to a significant reduction in [Na+]i (31), while Willis et al. showed that increased [Na+]i in Purkinje fibers promoted arrhythmogenesis (91). There are indications that some membrane currents exchangers involved in [Na+]i homeostasis are altered during AF. For example, in some studies of AF, the voltage gated sodium channel (NaV1.5) remained activated after the AP upstroke and created a small but persistent current that remained throughout the plateau phase of the cardiac AP (86,92), termed INa,Late. During this phase the membrane resistance is high and even a small increase of an inward current can cause AP prolongation that leads to rhythm instabilities (93,94). Additionally, in a sheep model of AF, the rate of Ca2+ extrusion by NCX was increase in comparison to the cells isolated from 24 control sheep (23). Ca2+ Signaling Silencing Recent work by Greiser and collaborators have demonstrated that the specific molecular and cellular adaptations that occurs during AF also involve a novel mechanism called “calcium signaling silencing”, which is a distinct cellular and molecular adaptive response to rapid cardiac activation. In this study, they found that in atrial cells that underwent rapid atrial pacing (RAP) to simulate the effect of rapid activation and its role in the initiation of AF, the intracellular Ca2+ buffering strength was increased, while intracellular sodium and calcium concentration were reduced ([Na+]i and [Ca2+]i respectively) (95). 25 4. Computational Modeling Cardiac ionic models were developed on the foundational work of Hodgkin and Huxley (HH) published in 1952, which provided mathematical representations of the ionic currents in the giant squid axon (5). Since then, there have been many atrial myocyte biophysical models, including three that incorporate spatial diffusion (38,96-98). Additionally, there are several compartmental ionic models of the atrial myocyte that take into account calcium handling (99,100). For the modeling aspects of this dissertation, we used the atrial myocyte modeling formalism published by Grandi et al.(99) and updated it to incorporate Ca2+ buffering effects on the LTCC current, according to our experimental results. Spatial aspects of Ca2+ signaling under different Ca2+ buffering conditions were also modeled using a simple geometry of atrial and ventricular mycytes developed with the COMSOL Multiphysics Software linked to the Matlab solution of the Grandi et al. ionic current equations. We used the Grandi model in this dissertation because the specific differences in membrane currents and Ca2+ handling between atrial and ventricular myocytes are taken into account in their mathematical formulation. For modeling predictions in which stochastic RyR2 gating was necessary, we used the modleing framework of a mouse ventricular myocyte developed by Williams et al.(63) and Wescott et al. (101) and updated it to account for our murine or rabbit atrial myocyte parameters. We also used the 3 dimensional (3D) spatial implementation in the Wescott et al. local control 26 model that represents a transverse subsection of a cardiomyocyte to simulate the experimental atrial myocyte Ca2+ line scans. The Williams and Wescott models are state-of-the-art biophysical models that have been experimentally validated to demonstrate the details of RyR2 Ca2+ release. This integrative modeling approach fine-tuned mathematical formalisms and incorporated verification and validation to ensure that our model predictions reflect experimental results. 5. Conclusion Our goal is to understand the mechanism of excitation contraction coupling in atrial myocytes than when deregulated would lead to AF. In particular to understand the role of altered Ca2+ buffering, ROS and [Na+]i on the AP and Ca2+ signaling. To achieve this goal we (1) developed specific tools to study atrial cell biology, (2) we studied the role of altered Ca2+ buffering on ionic membrane currents and intracellular calcium, (3) we investigated the role that ROS play in altered Ca2+ signaling and the morphology and duration of the action potential (AP), and finally (4) we studied Na+ and Ca2+ signaling in a transgenic murine model of AF. The work presented here involves biophysical experiments used to investigate atrial cell biology and mathematical modeling used to quantify and describe atrial cell behavior. 27 Chapter 2 : A Novel Approach to Isolate, Position and Stabilize Atrial Myocytes for Biological Experiments 1. Introduction In this chapter we discuss the development, verification and validation of the techniques used to perform experimentation in atrial myocytes. In particular we focus on two goals: (1) to improve existing cardiac myocytes isolation techniques to obtain healthy single atrial myocytes from mice and rabbit and (2) to determine a better small diameter rod shaped material that can withstand the atrial myocyte force of contraction and provide stability during biological experimentation. Single Atrial Myocytes Isolation We require cells that have the ECC characteristics of adult cardiac myocytes to adequately represent the biological process under investigation. Currently there are no cell lines that continue to actively contract over time; even short-term cultures quickly de-differentiate and loose this ability. Other possible sources include human embryonic stem cell (hESC)-derived atrial CMs (57) and human induced pluripotent stem cell (iPSC)-derived cardiomyocytes (58). However, various studies that attempted to characterize these cells have found numerous limitations with many factors affecting their morphological and molecular characterization, as wells as their ionic current and contractile properties (59-61). Therefore, at the moment, freshly isolated 28 single atrial myocytes from animal cardiac tissue offers the best representation of physiological function in cardiac investigations. Single myocytes from animal models are thus an important resource for biological investigation of cardiac disease. Because of technical and economic considerations, murine models have the advantage that they can be genetically modified to study specific mechanisms of various cardiovascular diseases (102). Murine atrial cardiomyocytes contain a denser transverse and axial membrane tubular system (TATS) than human (26,31) or rabbit atrial myocytes (31,37). Thus, it often necessary to conduct biological experiments on rabbit atrial myocytes that lack a high density of tubular membrane and have a closer ionic membrane profile to the human atrial cardiomyocytes . Nevertheless, the development of a method to isolate murine atrial cells of high quality is a useful tool. The specific proteins that are expressed in the extracellular matrix are spatially controlled in the heart, which leads to atrial and ventricular tissue differences (103). Therefore, cell isolation techniques need to be fine-tuned depending on the species and the region of the heart one wants to isolate cells from. There are numerous published methodologies for ventricular (104- 106) and atrial myocyte isolation (26,107-111) from different species. Since the atrial tissue volume is much smaller than the ventricular tissue, it is extremely challenging to isolate an adequate number of healthy atrial cells, especially from a small species like mouse. We have found that a 29 modification of the isolation methodology published by Shioya et al. (105) in combination with the isolation technique used by Wagner and colleagues (26) results in the highest yield of rod shaped murine atrial myocyte able to respond to field stimulation up to 3 Hz. For the rabbit atrial myocytes isolation, we found that the protocol used by Greiser et al. is sufficient to obtain a high yield of rabbit atrial myocytes with adequate electrophysiology properties (31). Stabilization of Atrial Myocytes Because of the small size of atrial cells, it would be an advantage to be able to stabilize cells physically, so that their contractions do not interfere with patch clamp and other manipulations. By stabilizing a cell, we should be able to generate a higher signal to noise ratio, desirable in many electrophysiology and fluorescence microscopy experiments because the signal from a single cell is very small. Although the hardware and software tools available maximize the signal and filter out undesirable noise, when measuring signals during electrical stimulation, the cell will contract an average of 10% of its baseline length. This motion artifact can affect the electrophysiology signal and also any imaging that is been acquired under stimulation. Therefore, being able to stabilize the cell to reduce the motion artifact due to contraction is desirable in many physiological experiments. Previously, stabilization was established biochemically by inhibition of contraction. 2,3-butanedione monoxime (BDM) is a negative inotropic agent that seems to inhibit cardiac contraction at the myofibril level (112), and it is frequently used in many cardiac studies to stop contraction. However, many 30 studies in ventricular cardiac myocytes have demonstrated that BDM affects the gating of many ionic currents. For example in an investigation of the ICa,L inactivation in ventricular rabbit cardiac myocytes, Altamirano and colleagues found that BDM itself modified ICa,L gating, even at low [Ca2+ ]i. (113). In another study by Watanabe and colleagues in guinea-pig cardiac ventricular cells, it was found that BDM inhibits both directions of the sodium-calcium exchanger current (INCX ) in a concentration-dependent manner. The inhibition occurred at concentrations of BDM between 1 and 30 mM, which is similar to that for suppressing cardiac muscle contraction (114). Therefore our goal was to develop another stabilization technique that does not involve chemicals that could affect the atrial myocyte physiological properties. To achieve this objective we investigated relative recent techniques used to stretch ventricular myocytes. Iribe and Kohl used a pair of carbon fibers attached to opposite ends of guinea pig ventricular myocytes to stretch these cells and study the effect of changing axial length cell on SR Ca2+ dynamics (115). Prosser and colleagues developed a special biological adhesive called MyoTak and attached borosilicate glass rods of 25 µm diameter to different ends of rat ventricular myocytes to study the effect of stretch dependent ROS production on Ca2+ signaling (45). However, these studies were designed to work with contractions, and rod flexibility was required. For these techniques to be used in our atrial cell biological application the rods need to meet two additional requirements: 1) they need to have small diameters to leave an adequate atrial myocyte area uncovered 31 for experimentation and 2) be able to withstand the attached atrial myocyte contractile force with minimal bending. For our application this requirement is defined as able to bend less than 1 µm under a contractile force of 4 µN. This force requirement is based on the results of Bluhm and colleagues, which measured the active force of different types of cardiac muscles (116). As demonstrated in the careful imaging study performed by Walden and colleagues (117), atrial myocytes have smaller diameter and length than ventricular myocytes. Therefore smaller diameter rods would be more suitable for atrial myocyte manipulation. An ideal rod will have a diameter less than 10 µm. However, the stiffness of a rod is related to the fourth power of the radius, therefore reducing the diameter by 60 % implies reducing the stiffness by 97.4 % when all other variables are held constant. Prosser et al. used borosilicate glass. One of the challenges with using 25 µm diameter borosilicate glass is illustrated in Figure 2.1. If the length of an atrial cell is approximately 80 µm and two 25 µm rods cover 50 µm total, then only 30 µm would be left for cellular manipulations, and as there is always overhang at the ends, not even 30 µm would be available. This is in addition to the issue of stiffness of glass rods. The Young’s modulus is a measure of the stiffness of a solid material and is defined as the ratio of stress (force per unit area) to strain (material deformation) and measured in units of gigapascal (GPa). Given our experimental requirements, if cells are attached with a rod of length of 500 µm and 10 µm diameter, the Young’s modulus to withstand a 4 µN force while 32 bending less than 1 µm would have to be greater than 339.5 GPa. Theoretically, borosilicate glass has a Young’s modulus of only 60 GPa and carbon fibers, as used by Iribe and Kohl (115), has a Young’s modulus of 242 GPa. Therefore, we investigated, verified and validated the use of aluminum oxide (sapphire) rods, which have a theoretical Young’s modulus of 373 GPa. 33 Figure 2.1. Rod Stabilization (A) Ventricular myocyte attached to two 25 µm diameter borosilicate glass rods (Modified from (45) and used with permission from The American Association for the Advancement of Science). (B) Atrial myocyte attached to two 10 µm diameter sapphire fiber rods. (C) Schematics portraying the enhanced suitability of smaller diameter rods for physiological investigations in atrial myocytes. 34 2. Materials and Methods Atrial Cell Isolation from Mice Atrial myocytes are isolated from adult C57BL/6 mice. Each mouse is placed in a precision vaporizer and induction chamber gassed with isoflurane in 100% oxygen. The isoflurane flow is set to 1% until the animal is unresponsive before giving an intraperitoneal (IP) heparin injection (1-1.25 U/g) 15 minutes before starting the isolation. The mouse is then deeply anesthetized by increasing the isoflurane anesthesia to 5% (which is confirmed by foot pinch). As in (105), a thoracotomy is quickly performed to remove the heart, which is placed in ice-cold Ca2+-free Cell Isolation Buffer (CIB) solution to decrease its metabolic demand (Solution components are listed in Table 2.1). The heart is then cannulated through the aorta and tied using a silk suture under a light microscope with 3X resolution in CIB. It is then confirmed that the cannula is actually above the aortic sinuses by delivering the same solution through the cannula connected to a syringe and observing the coronary artery perfusion under the microscope. This step is important to ensure proper perfusion to the atrial tissue because it has been shown that the coronary artery anatomy is highly variable in mice (118). The heart is then mounted to a Langendorff perfusion apparatus and perfused at 37°C with the Ca2+-free CIB solution for 5 minutes to wash out the blood. Subsequently, the heart is perfused with CIB containing collagenase (0.8 mg/ml, Worthington), trypsin (0.06 mg/ml, Fisher Scientific), protease (0.06 mg/ml, Sigma-Aldrich) 35 and 0.1 mM CaCl2 for 3-5 minutes or until the atrial tissue is soft. The right and left atria are excised and transferred to a small beaker containing the same CIB solution perfused in the step above but with 0.15 mM CaCl2 and placed in an incubator at 37°C for 5 minutes. The atria are then transferred to Tyrode’s solution containing 15 mM Bovine Serum Albumin (BSA, Sigma-Aldrich) and 30 mM 2,3-Butanedione monoxime (BDM, Sigma-Aldrich) (Tyrode’s components listed in Table 2.2). The atrial tissue is gently agitated and cut to remove the cells. The cell suspension is then strained through a 200 µM filter and CaCl2 is added three times every 10 minutes to a final concentration of 0.3 mM. Experiments are conducted within 4 hours of cell isolation in Tyrode’s solution free of BDM and BSA and with 1.8 mM CaCl2 to replicate physiological conditions. It is worth noting that after Tyrode’s perfusion, our experience has been that BDM washes out of the cells, and the cells recover their full contractile properties, as was also observed by Yu et al. (119). Atrial Cell Isolation from Rabbit The rabbit cell isolation performed is similar to the methodology used by Greiser and colleagues (31). Briefly, each rabbit was given an intramuscular injection of xylazine (5 mg/kg) and ketamine (50 mg/kg) to sedate and anesthetize it. The rabbit was then given 1 ml of heparin (3600 U/ml) intravenously (iv) to avoid formation of any arterial coagulum. After 10 minutes, the rabbit is euthanized with an iv injection of pentobarbital (100 36 mg/kg). Once the rabbit expires (i.e., measured by cessation of respiration or deep plane of anesthesia), the heart is quickly removed through thoracotomy. The heart is then placed in ice-cold Ca2+-free CIB solution (Table 2.2) to decrease its metabolic demand and the pericardial sac is separated from the heart. The heart is manually pumped to remove blood and cannulated through the aorta on a Langendorff perfusion apparatus kept at 37 °C. The heart is perfused with Ca2+-free Tyrode’s solution for 5 minutes to wash out the remaining blood. The heart is then perfused with Ca2+-free Tyrode’s solution containing collagenase (0.8 mg/ml, Worthington) and BSA (1g/ml) for 6-8 minutes or until the atria tissue is soft and the aortic valve is ruptured. The atria are excised and transferred to the Tyrode’s solution containing 15 mM BSA and 30 mM BDM. The atrial tissue is then gently agitated and cut to remove the cells. The cell suspension is then strained through a 200 µM filter and Ca2+ is added five times every 10 minutes to a final concentration of 180 µM. Experiments are conducted within 4 hours of cell isolation using Tyrode’s solution free of BDM and BSA and containing 1.8 mM CaCl2 to simulate physiological conditions. 37 Cell Isolation Buffer Concentration (mM) NaCl 130 KCI 5.4 MgCl2.6H2O * 0.5 NaH2PO4 0.33 Glucose 16 HEPES 25 Taurine 6 EGTA 0-0.4* CaCl2 0-0.15* Digestive Enzymes See text* Table 2.1. Mouse Cardiac Cell Isolation Buffer *See text for specific concentrations at each step Tyrode’s solution Concentration (mM) NaCl 133 KCl 5 MgCl2+6H2O 2 KH2PO4 1.2 Taurine 6 Creatinine 6 Glucose 10 2,3-Butanedione monoxime (BDM) 0 - 30* Bovine Serum Albumin (BSA) 0 - 15* CaCl2 0 - 1.8* Table 2.2 Tyrode solution used for storage and experimentation *See text for specific concentrations at each step 38 All animal procedures were performed in accordance with the standards set forth by the University of Maryland School of Medicine Institutional Animal Care and Use Committee and by the National Institutes of Health. The University of Maryland Baltimore is an AALAC accredited facility. Procedure for Verifying Fiber Stiffness The stiffness of borosilicate glass rods (10 µm and 25 µm diameter, Fiber Optics Technology Inc.), carbon fiber rods (7 µm diameter, Zoltek Corporation) and sapphire rods (10 µm diameter, 3M Corporation) were empirically verified. Table 2.3 shows the theoretical Young’s modulus of these materials. The equipment used to determine the stiffness consisted of a confocal microscope (LSM 510), a piezo electric-length microcontroller (Mad City Labs) driven by a variable voltage output source, a force transducer (Aurora Scientific Inc.), an analog to digital (A/D) converter (Aurora Scientific Inc.) and a digital controller (Aurora Scientific Inc.). The following methodology was designed to test the fibers’ stiffness experimentally. To test each fiber a soldering iron was used to melt wax and attach 1 mm of the rod to the force transducer. The rods were then placed on the force transducer right above the optical aperture of the LSM confocal microscope, while the length controller was placed in contact with the tip of the rod. Once the length controller and force transducer were connected to the digital controller and A/D converter, the length deflection in the protocol shown in Figure 2.2 was applied and the force exerted by the rod from the force transducer was recorded. 39 Material Modulus of elasticity (GPa) Borosilicate glass 60 Carbon fibers 242 Aluminum Oxide (Sapphire) 373 Table 2.3. Theoretical Young’s modulus of rod materials. Validation of Fiber Suitability to Stabilize Atrial Myocytes Once an appropriate material was verified (see Results), the fiber with an adequate empirical Young’s modulus was also validated to ensure that it could perform its intended function, which is to stabilize atrial myocytes in biological experiments. To conduct this validation one micro-rod is connected to the force transducer and another to the length controller (WPI) driven by a variable voltage output source. The rods are then coated with the biological adhesive MyoTak (Ionoptix Inc.). Isolated murine atrial myocytes are placed in glass cover slip inside a custom cell chamber and mounted on an Eclipse Ti inverted confocal microscope (Nikon Instrument Inc.). The chamber is equipped with a pair of platinum field electrodes that are connected to a Myopacer cell stimulator (Ionoptix Inc.). The rods are then positioned on opposite sides above a selected myocyte using motorized micromanipulators (Siskiyou Corporation). The cell is attached at both ends by gently pressing down with the MyoTak-coated micro-rods. Sarcomere length is measured by outputting the transmitted light to a camera connected to sarcomere length measurement software (Aurora Scientific Inc.). Sarcomere length is measured on the micro-rods attached atrial myocyte during rest and when subjected to 1 Hz field stimulation. These results for attached cells are then compared to the 40 sarcomere length measured in unattached cells, and the percentage in movement reduction between attached and unattached cells is calculated. Statistical Analysis Results are reported as means ± SEM for the indicated number of cell measurements. Statistical analyses were performed using two-tailed Student t test for unpaired values. Values of p<0.05 were considered significant. 41 Figure 2.2. Verification Protocol (A) Length deflection input to displace the rod attached to the force transducer. (B) Representative microscopy image of the deflection-force protocol. 0 10 20 30 40 50 0 25 50 75 100 Time (s) Le ng th D is pl ac em en t, Δ L (µ m ) ΔL= 5 µm ΔL= 10 µm ΔL= 20 µm ΔL= 25 µm ΔL= 35 µm ΔL= 50 µm Force Transducer Length Controller A B 42 3. Results Cell Isolation Quality Figure 2.3 shows isolated mouse atrial myocytes with healthy striation. These cells were able to respond to field stimulation up to 3 Hz at room temperature (25 °C) and undergo both confocal imaging and electrophysiology experimental protocols. Viability of the cells, which was defined as the percentage of relaxed, striated and rod-shaped cells in total cell count, was 65.63 ± 2.4 % for rabbit atrial myocytes (n=8) and 53.13 ± 3.89 % for mouse atrial myocytes (n=8). To test the cell quality of our mouse and rabbit cell isolations we measured the L-type Ca2+ channels current (ICa,L) amplitude using the voltage clamp technique. Figure 2.4 depicts ICa,L during a voltage clamp step to 0 mV. Voltage clamp technique can only be done in cells with preserved membrane integrity. The amplitude of the current in both rabbit and mouse atrial myocytes was large and it showed the adequate inactivation profile of the channel for both cell types. These results are in agreement with ICa,L parameters reported in the literature (87,88,113). 43 Figure 2.3. Cell Viability (A) Representative murine atrial cells taken with a 40X demonstrating healthy atrial myocytes with striation. (B) Cell viability after isolation of rabbit and mouse atrial myocytes. Rabbit Mouse 0 20 40 60 80 C el l V ia bi lit y (% ) A B 44 Figure 2.4. Cell Properties (A) Representative ICa,L from a mouse atrial myocyte with a Voltage step to 0 mV. (B) Representative ICa,L from a rabbit atrial myocyte with a Voltage step to 0 mV. (C) Statistics of the maximum amplitude of ICa,L in mouse and rabbit atrial myocytes (n = 11 cells). -10 -5 0 I C a, L (p A /p F) 20 ms 400 ms -50 mV 0 mV -10 -5 0 I C a, L (p A /p F) 200 ms -50 mV 0 mV 50 ms Rabbit Mouse -20 -15 -10 -5 0 I C a, L (p A /p F) A B C 45 Rod Material Verification The flexural stiffness of a beam relates the deflection of the beam under a certain force (120) and is given by the following formula: EI = !!!!"! , Where E is the Young’s modulus; I is the second moment of area (which for a circular beam is given by I = !!!! ); l is the length of the beam, Δ𝐿 is the displacement of the beam and F is the applied force. As mentioned above, to meet our experimental requirements, the Young’s modulus necessary to withstand a 4 µN force while bending less than 1 µm would have to be greater than 339.5 GPa. Table 2.3 shows the theoretical value of three materials that can be manufactured as 10 µm diameter fibers. Figure 2.5 depicts the forces obtained under the same length deflections for the different fibers tested. The force is proportional to the diameter of the fiber and the stiffness of the material. Using the flexural stiffness equation, we were able to measure the empirical Young’s modulus of the available fibers. This figure also shows these values, which deviate from the theoretical values due to differences in the manufacturing process and material composition of each company. According to these data, sapphire fibers have sufficient flexural stiffness for atrial myocyte positioning and experimentation. 46 Figure 2.5. Rod Stiffness (A) Force-length relationship for the different rods. (B) Experimental Young’s modulus calculated from the force-deflection protocol. 0 20 40 60 80 100 0 50 100 ΔL (µm) Fo rc e (µ N ) Borosilicate 10 µm Borosilicate 25 µm Sapphire 10 µm Sapphire Clean 10 µm Carbon 7 µm Bo ro sil ica te 10 µm Bo ro sil ica te 25 µm Sa pp hir e 1 0 µ m Sa pp hir e C lea n 1 0 µ m Ca rb on 7 µm 0 200 400 600 Ex pe rim en ta l E (G Pa ) A B 47 Sapphire Rod Validation Since only the sapphire rods have theoretical and empirical Young’s modulus greater than the requirement of 339.5 GPa, we only used this material in our biological validation. To validate that the movement artifact was reduced in cells attached to the sapphire rods, we conducted sarcomere length measurements in atrial cells that were field stimulated at 1 Hz. We used mouse atrial myocytes in these experiments because sarcomere length can be validated using either cell source, and mouse cells are a more economical source than rabbit cells. Figure 2.6 shows the results of the sarcomere length validation. Panel A shows a representative example of the sarcomere length traces in attached and free murine atrial myocytes under field stimulation. In free atrial myocytes that are allowed to contract freely, the first contraction is much larger than subsequent contractions. It is important to notice that the baseline sarcomere length did not change between free and attached atrial myocytes, which indicated that other physiological and technical variables were the same in both experimental groups. Under field stimulation the sarcomere length in unattached atrial myocytes diminished by 0.19 ± 0.01 µm in comparison with a reduction of only 0.03 ± 0.0013 µm in myocytes attached to the 10 µm diameter sapphire rods. This indicated that the percent change in cell length during contraction was reduced from 11.76 ± 0.75 to 1.67 ± 0.07 % (p value <0.0001), demonstrating that the movement artifact due to cellular contraction can be reduced by 85 % using sapphire rods. 48 Figure 2.6. Murine Sarcomere Length Validation (A) Representative example of a murine atrial myocyte sarcomere length measurement recorded under 1Hz field stimulation with and without sapphire rods attachment. (B) Resting sarcomere length is the same in attached and free atrial cells; n = 7 and 8 free and attached myocyte measurements respectively. (C) Maximum sarcomere length change under pacing in free and attached atrial cells; n = 86 and 36 free and attached cell measurements respectively. (D) Cell length reduction under 1 Hz field stimulation in attached and free atrial myocytes; n = 86 and 36 free and attached myocyte measurements respectively. Data are presented as mean ± SEM. ****P < 0.0001; t test. 0 2 4 6 8 1.0 1.2 1.4 1.6 1.8 2.0 Time (s) Sa rc om er e Le ng th (µ m ) Rods 1Hz No Rods 1Hz No Rods 1Hz Rods 1Hz 0.00 0.05 0.10 0.15 0.20 0.25 Δ S ar co m er e Le ng th (µ m ) **** No Rods 0 Hz Rods 0 Hz 0.0 0.5 1.0 1.5 2.0 R es tin g Sa rc om er e Le ng th (µ m ) ns No Rods 1Hz Rods 1Hz 0 5 10 15 C el l L en gt h R ed uc tio n (% ) **** A B C D 49 4. Discussion Firstly, we have described a method to isolate murine atrial cells and stabilize atrial myocytes for physiological experiments. Our technique to isolate healthy heart cells from mouse models is similar to that described by Shioya and colleagues (105). However, we have modified several aspects of the technique to be able to tailor the isolation to atrial myocytes, for example we added taurine to the cell isolation buffer and used a different Tyrode’s solution to store the cells before experimentation. We also confirmed that the cannula was actually above the aortic sinuses by delivering CIB solution through the cannula connected to a syringe and observing the coronary artery perfusion under 3X magnification light microscopy. This step is important to ensure proper perfusion to the atrial tissue because coronary sinuses have been shown to have diverse anatomies in murine models (121). As shown above the cell viability with our isolation techniques is high in both mouse and rabbit atrial myocytes. The cells have excellent structural and electrophysiology properties that provide assurance that our isolation is not affecting the cellular membrane and its ionic channels characteristics. Secondly, we determined which material for a rod could withstand the active force of an atrial myocyte to reduce its movement artifact during experimentation. Active force experiments with different types of cardiac muscle suggest that 4 µN is an acceptable requirement for the ideal rod to withstand with minimal bending (116). Our verification and validation results demonstrate that the sapphire rods can attach and stabilize atrial myocytes 50 and reduce the movement due to cellular contraction by 85 % when compared to unattached cells placed in a glass chamber. This rod attachment stabilization method offers a non-chemical solution to the side effects of chemicals that stabilize cells by inhibiting myocyte contraction, especially BDM. Because BDM is a negative inotropic agent that inhibits muscle contraction at the myofibril level, it is often used in many cardiac studies to stop contraction. However, as explained above, many studies in cardiac myocytes physiology have shown that BDM affects the electrophysiology properties of the cardiac myocytes, thereby calling into question some results. Our new sapphire rod stabilization methodology does not have unwanted effects on the electrophysiological properties of the cell, thus enabling the study of atrial physiology without introducing unwanted experimental variability. The limitation with this methodology is the length of times it takes to position and attach each cell to the sapphire rods, in our experience attaching the rods to the length controller and force transducer and coating them can initially take an average of 45 minutes. Moreover, positioning and attaching each myocyte to the rod can add another 15 minutes to the experimental time. However, as one gains experience the length of time to position cells is significantly reduced. 51 5. Conclusion In this chapter we have demonstrated two new methodologies, one to isolate healthy atrial myocytes and another to perform biological experimentations with reduced motion artifacts due to myocyte contraction. The first shows that murine atrial cells can be used with good viability. The second significantly eliminates problems with biochemical inhibition of contraction. The advantage of this micro rod sapphire fiber stabilization methodology is that it is non-chemical and does not affect the cell function. The limitation with this methodology is the length of times it takes to position and attach each cells to the sapphire rods. However, this is offset by the increase experimental quality and reduced signal to noise ratio that can be achieved with this approach. The methods described here will be applied to subsequent chapters. 52 Chapter 3 : Buffering Effects on the L-Type Ca2+ Channel Current and Ca2+ Dynamics in Atrial Cardiomyocytes. 1. Introduction Several studies have demonstrated that atrial cells have unique electrophysiological properties due to differences in ion channels currents, expression and function (29,87,117,122). ICa,L is the most important membrane current involved in ECC. It is also known that this current’s amplitude is reduced during atrial fibrillation, although the molecular mechanisms are not fully understood (e.g., protein expression, phosphorylation, phosphatase effects) (29,33). LTCC opening during depolarization is followed by inactivation to stop the current (i.e., ICa,L) been conducted through this channel. ICa,L has two modes of inactivation: the slower one is voltage inactivation and the faster acting one is a Calmodulin (CaM) mediated Ca2+-dependent inactivation (CDI) (35-37). These act as a negative feedback mechanisms to stop the influx of Ca2+ into the cell (34). When arrhythmias are triggered, the cells adapts to the new regimen by modifying a variety of signals. Greiser and colleagues showed that rabbits that underwent rapid atrial pacing (RAP) to simulate the onset of AF, had their peak ICa,L current density decreased by 70%. In another set of experiments, Greiser et al. measured Ca2+ buffering by calculating the slope of the relationship between total [Ca2+]i (measured by integrating the total amount of calcium that is extruded through NCX after applying caffeine to the cell) and 53 the free [Ca2+]i (measured using whole cell epifluorescence simultaneously with the NCX current measurements). They found that this slope was steeper in the RAP cells, demonstrating an increased Ca2+ buffering strength. The results suggest that increased fast Ca2+ buffering strength is one possible adaptation that occurs to modulate and stabilize Ca2+ signaling with arrhythmias (31). Despite many experimental and mathematical modeling data showing the modulating effects of Ca2+ buffering on Ca2+ signaling (123-127), many patchclamp experiments measuring ICa,L used 10 mM of ethylene glycol tetraacetic acid (EGTA) in the patch pipette, which is 20 times higher than physiological buffering levels (29,31,128,129). The reason EGTA is used in the pipette is that it chelates Ca2+ ions, reducing cell contraction and subsequent instability in current recording. However, the effect of buffering due to the rapid pacing adaptation cannot be distinguished from the EGTA buffering artifact in experiments to measure ICa,L. Therefore the objective of this study is to characterize ICa,L at more physiological [Ca2+]i and, in particular, to elucidate the effect of buffering on ICa,L amplitude and CDI. We also use mathematical modeling of atrial cells to bring a quantitative understanding of the effect of buffering on the ICa,L properties, as well as other aspects of Ca2+ signaling. 54 2. Methods Cell Isolation For this project, only rabbit cells were used. The method is described in detail in chapter 2 and also used by Greiser and colleagues (31). For specific solution details please refer to the rabbit cell isolation section in chapter two. Briefly, each rabbit was sedated, anesthetized and heparinized. After 10 minutes, the rabbit was euthanized with an iv injection of pentobarbital (100mg/kg). Once the rabbit expired, the heart was removed quickly through thoracotomy. The heart was placed in ice-cold Ca2+-free Tyrode’s solution to decrease its metabolic demand and the pericardial sac was separated from the heart. The heart was manually pumped to remove blood and cannulated through the aorta on a Langendorff perfusion apparatus kept at 37 °C. The heart was perfused with Ca2+- free Tyrode’s solution for 5 minutes to wash out the remaining blood. The heart was then perfused with Ca2+-free Tyrode ‘s solution containing collagenase (0.8 mg/ml, Worthington) and BSA (1g/ml) for 6-8 minutes or until the atrial tissue was soft and the aortic valve was ruptured. The left atrium was excised and transferred to the Tyrode’s solution containing 15 mM BSA and 30 mM BDM. The tissue was gently agitated and cut to remove the cells. Cell suspension was then strained through a 200 µM filter and Ca2+ was added five times every 10 minutes to a final concentration of 180 µM. Experiments were conducted within 4 hours of cell isolation using Tyrode’s solution free of BDM and BSA and containing 1.8 mM CaCl2 to simulate physiological conditions. 55 Electrophysiology Experiments Electrophysiological measurements were done by applying the techniques described in Walden et al. (117) and Greiser et al (31).The patch- clamp technique was used in conventional whole-cell mode with an Axopatch 200B amplifier (Molecular Devices) and a computer equipped with the Clampex and Clampfit software (Molecular Devices). Briefly, pipettes with resistance of 1.5 - 2.5 MΩ were made with the micropipette puller (Model P- 97 Flaming/Brown) available in our laboratory. Then the micropipette was fused to the cell membrane to create a seal with resistance in the giga-ohm range. Gently approaching the cell membrane with the pipette and applying suction until a 2 GΩ seal was achieved, as determined via the software, created the seal. Access into the cell was then gained by gently rupturing the cell membrane with rapid suction (17). For ICa,L measurements the whole-cell configuration was established with access resistances < 10 MΩ. The cells were allowed to equilibrate until all fast inward currents and all outward currents had disappeared, and ICa,L had reached a stable plateau, as assessed by voltage ramps from -90 to +60 mV over 200 ms every 45 s. A pre-pulse was given by depolarizing the cells to 0 mV at a frequency of 1 Hz for 10 s to bring the cells to steady state. After pre-pulse, cells were slowly (500 ms) ramped up to a voltage of -50 mV to inactivate Na+ channels and prevent contamination of the Ca2+ current by Na+. Cells were held at -50 mV for 50 ms before each desired voltage step. 56 Experiments were performed at room temperature (21-23°C). Table 3.1 depicts the solutions that were used for ICa,L measurements. Pipette solution Concentration (mM) CsCl 140 MgCl2 1.7 Mg-ATP 0.3 HEPES 10 EGTA 0.05 or 10 (see text) pH adjusted to 7.2 with CsOH Bath Solution Tetraethyl ammonium chloride (TEA-Cl) 136 CaCl2 1.8 MgCl2 1.8 HEPES 10 pH adjusted to 7.4 with TEA-OH Table 3.1. ICa,L measurement solutions. Mathematical Compartmental Modeling We used the compartmental model of the human atrial myocyte developed by Grandi et al. (99). The ICa,L mathematical formulation is based on the Goldman-Hodgkin- Katz equation and deterministic gating variables: 𝑰𝑪𝒂,𝑳 = 𝑑 ∙ 𝑓 ∙ 𝑓!" ∙ 𝑃!" ∙ 4 ∙ 𝑉𝐹!𝑅𝑇 ∙ 𝛾!"# ∙ [𝐶𝑎!!]! ∙ 𝑒 !!"!" − 𝛾!"# ∙ [𝐶𝑎!!]!𝑒 !!"!" − 1 The gating variables for the activation, voltage dependent inactivation and Ca2+ dependent inactivation are represented by the following ODEs: 𝑑𝑑𝑑𝑡 = 𝑑! − 𝑑𝜏! ; 𝑑! = 11+ 𝑒(!!!!".!)/! ; 𝜏! = 1− 𝑒(!!!!".!)/!0.035(𝑉 + 143.5) 57 𝑑𝑓𝑑𝑡 = 𝑓! − 𝑓𝜏! ; 𝑓! = 11+ 𝑒!!!".!"!.! + 0.61+ 𝑒!"!!!" ; 𝜏! = 10.0197 ∙ 𝑒!(!.!""#(!!!".!)! + 0.02 𝑓!" = 1− 𝑓!"# ; 𝑑𝑓!"#𝑑𝑡 = 1.7 𝐶𝑎!! ! 1− 𝑓!"# − 11.9𝑓!"# To compute the buffering effects seen in our experiments, we have incorporated the following rate equations for EGTA and BAPTA to the Grandi model, which already included other relevant physiological Ca2+ buffers developed in the Shannon et al. rabbit model and were left unchanged (130). 𝐸𝐺𝑇𝐴 ! + [𝐶𝑎!!]! !!" !!"" [𝐸𝐺𝑇𝐴 ∙ 𝐶𝑎!!]! 𝑑[𝐸𝐺𝑇𝐴 ∙ 𝐶𝑎!!]!𝑑𝑡 = 𝑘!" 𝐶𝑎!! ! 𝐵! − 𝐸𝐺𝑇𝐴 ∙ 𝐶𝑎!! ! − 𝑘!"" 𝐸𝐺𝑇𝐴 ∙ 𝐶𝑎!! ! For EGTA, 𝑘!" = 1.5 µ𝑀!! 𝑠!! , 𝑘!"" = 0.3 s!!, 𝐵! = 10 mM or 0.05 mM For BAPTA, 𝑘!" = 600 µ𝑀!! 𝑠!! , 𝑘!"" = 100 s!!, 𝐵! = 10 mM or 0.05 mM The rest of the ionic currents were modeled as in Grandi et al. (99). Mathematical Compartmental Stochastic Modeling To determine the effect that buffering may play in CICR, we integrated the buffering formalism shown above into the ventricular myocyte modeling framework developed by Williams et al.(63) and Wescott et al. (101). We updated this model by reducing the number of RyR2 in a CRU from 50 to 35, as it has been shown in the literature that CRU clusters are smaller in atrial than in ventricular cells (131). We then simulated the effects of changing 58 EGTA concentrations, under a 200 ms voltage step to 0 mV, on CICR by incorporating the buffering rates shown in the equations above. Mathematical Spatial Modeling The results of the compartmental model simulation were incorporated into a simple spatiotemporal diffusion model to take into account the effect of the TTs in model atrial and ventricular cells in which the diffusivity of the Ca2+ ions was changed depending on the buffering power. Figure 3.1 shows the model geometry, which is based on the published experimental data of Chen- Izu et al. (19) and Parfenov et al. (132). The following governing equations and boundary conditions were incorporated into the spatial simulation: ![!"!!]!!" = ∇ ∙ 𝐷!"!! ∙ ∇[𝐶𝑎!!]! , ∇[𝐶𝑎!!]! = !!"!!!!"!! ,𝑎𝑡 𝑡 = 0 [𝐶𝑎!!]! = 100 𝑛𝑀 𝐽!"!! = 𝐽!"## + 𝐽!"# + 𝐽!"#$ + 𝐽!"# , 𝐶𝑎!!𝑓𝑙𝑢𝑥 𝑡ℎ𝑟𝑜𝑢𝑔ℎ 𝑖𝑜𝑛 𝑐ℎ𝑎𝑛𝑛𝑒𝑙, 𝑒𝑥𝑐ℎ𝑎𝑛𝑔𝑒𝑠 𝑎𝑛𝑑 𝑝𝑢𝑚𝑝 Values of Ca2+ diffusivity under different buffering conditions were obtained from the work of Kushmerick et al.(133). The spatial model geometric mesh and equations were solved using the Comsol Multiphysics Software. Statistical Analysis Results are reported as means ± SEM for the indicated number of cells or measurements. Statistical analyses were performed using two-tailed 59 Student t test for unpaired values. Values of p < 0.05 were considered significant. 60 Figure 3.1. Spatial Model Geometry Spatial geometry of the model of a cardiac myocyte with and without a t- tubular network. Dimensions are derived from typical geometries of cardiac myocytes (117). The T-tubule spacing and dimensions are based on the published experimental data of Chen-Izu et al. (19) and Parfenov et al. (132). 61 3. Results Buffering Effects on the Rabbit Atria LCC Current ICa,L was measured under different buffering conditions to study the impact of physiological versus high buffering on ICa,L amplitude and inactivation. The experimental results are depicted in Figure 3.2 and Figure 3.3. Measured ICa,L at physiological buffering conditions (i.e., 0.05 mM EGTA) shows a time to 70% recovery of 11.1 ± 0.9 ms, in comparison to 28.5 ± 22.2 ms when ICa,L was measured at high buffering conditions (i.e., 10 mM EGTA). This demonstrates a 61% increase in calcium dependent inactivation or CDI at low buffering conditions. These LTCC results in rabbit atria are in agreement with observations seen in experiments in ventricular myocytes; which is expected given that there is no evidence that the LTCC has different isoforms between the atria and ventricular tissue (34,113). Increasing the intracellular Ca2+ buffering did not affect the peak current density of ICa,L, which was -9.2 ± 0.4 pA/pF and -8.3 ± 1.2 pA/pF in 10 mM and 0.05 mM EGTA, respectively. This result was somewhat surprising, but it could be explained due to the fast nature of ICa,L inactivation in comparison with the slower binding and unbinding kinetics of EGTA to calcium ions. 62 Figure 3.2. Experimental ICa,L Under Different Buffering Conditions. (A) Representative ICa,L current traces measured with a voltage step to 0 mV under physiological levels of buffering (EGTA 0.05 mM) and with large amounts of EGTA (EGTA 10 mM). (B) Current voltage relationship displaying the maximum current at each of the indicated voltage steps. 63 Figure 3.3. Experimental ICa,L Statistics . (A) Time of reduction to 70% of peak value was significantly increased in currents measured with 10 mM of EGTA in comparison with physiological levels of EGTA (n = 8 cells in each group). (B) Peak current measured with a 0 mV step did not have significant changes under these two calcium buffering conditions. (n = 8 cells at 10 mM EGTA and n = 6 at 0.05 mM EGTA). EGTA 0.05 mM EGTA 10 mM 0 10 20 30 40 Ti m e to 7 0% in ac tiv at io n (m s) **** EGTA 0.05 mM EGTA 10 mM -15 -10 -5 0 M ax im um I C a, L (p A /p F) ns A B 64 Compartmental Modeling Simulations of the LCC Current Mathematical modeling was implemented to simulate ICa,L under the buffering formalism using the Grandi et al. atrial myocyte model (99). We first verified that the compartmental model could adequately quantify the changes observed in the EGTA experiments. Figure 3.4 and Figure 3.5 depict the modeling results. The model predicted a maximum current of -8.75 pA/pF and -8.74 pA/pF with 10 mM and 0.05 mM EGTA, respectively, which deviated from the experimental results by only 5 %. Moreover, the model predicted a time to 70 % recovery of inactivation of 26.5 ms when using 10 mM EGTA vs. 13.6 ms when using the physiological equivalent of 0.05 mM EGTA. This represents an 8 % deviation from the experimental observations. Thus, the compartmental model demonstrated high fidelity in capturing the effects of Ca2+ buffering on the ICa,L properties. Therefore, it was used to make predictions on the effect of the Ca2+ chelator BAPTA, which has binding kinetics 400 times greater than those of EGTA. Simulations with 10 mM BAPTA did not significantly change the maximum amplitude of ICa,L, which was 8.78 pA/pF. However, BAPTA was able to bind to Ca2+ ions faster in the vicinity of the channel and slow CDI more effectively than EGTA. This resulted in a time to 70 % recovery of inactivation of 40.1 ms, which is almost 40 % slower than the inactivation simulated with 10 mM EGTA. 65 Figure 3.4. Computationally Modeled ICa,L (A) Representative current traces modeled with a simulated voltage step to 0 mV and with different buffering formulations representative of the binding kinetics of BAPTA and EGTA. (B) Peak current voltage relationship as simulated in the mathematical model. -50 0 50 -10 -5 0 Voltage (mV) I C a, L (p A /p F) EGTA (0.05 mM) EGTA (10 mM) Bapta (10 mM) A B 60 mV -50 mV 200 ms -50 mV 0 mV 200 ms 66 Figure 3.5. Computationally Simulated ICa,L. (A) Time of ICa,L reduction to 70% of peak value is considerably changed under different simulated buffering conditions. (B) Peak current simulated with 0 mV step did not have significant changes in the model as corroborated by the experiments in Figure 3.3. EGTA 0.05 mM EGTA 10 mM Bapta 10 mM 0 10 20 30 40 Ti m e to 7 0% in ac tiv at io n (m s) EGTA 0.05 mM EGTA 10 mM Bapta 10 mM -10 -8 -6 -4 -2 0 M ax im um I C a, L (p A /p F) A B 67 Simulations Using Compartmental Stochastic Model of CICR To predict the effects that buffering may play in other CICR parameters, we integrated the buffering equations shown in the method section into the ventricular myocyte modeling framework developed by Williams et al.(63) and Wescott et al. (101), which includes a state of the art mathematical formalism that captures the stochastic gating properties of the RyR2s. Then using our rabbit atrial myocyte parameters, we simulated the effect of changing EGTA concentrations on [Ca2+]i, [Ca2+]SR, SL shortering, SERCA function and INCX. The stochastic Ca2+ release simulation, under 0.05 mM EGTA to represent physiological conditions and 10 mM EGTA to simulate high buffering conditions, indicates that under high buffering the [Ca2+]i transient is significantly reduced. This is expected as EGTA chelates the Ca2+ ions in the cytosol. The reduced [Ca2+]i transient then leads to less Ca2+ reuptake through the SERCA ATPase, resulting in lower [Ca2+]SR. Lower [Ca2+]i also leads to less Ca2+ extrussion and significantly inhibits contraction, as shown by the insignificant SL shortening under 10 mM EGTA. These results are shown in Figure 3.6. 68 Figure 3.6 Stochastically Simulated CICR under EGTA (A) Representative ICa,L traces under a simulated voltage step to 0 mV. (B) Ca2+ reuptake into the SR through the SERCA ATPase pump. (C) Ca2+ extrusion through NCX. (D) Cytosolic [Ca2+]i transient. (E) SR Ca2+ concentration during the [Ca2+]i transient. (F) Sarcomere Length change is significantly inhibited under 10 mM EGTA. 0.00 0.05 0.10 0.15 0.20 0.25 -10 -8 -6 -4 -2 0 Time [s] I C a, L (p A /p F) EGTA (0.05 mM) EGTA (10mM) 200 ms-50 mV 0 mV 0.00 0.05 0.10 0.15 0.20 0.25 0 100 200 300 400 Time [s] J S ER C A (n M /s ) EGTA (0.05 mM) EGTA (10mM) 200 ms-50 mV 0 mV 0.00 0.05 0.10 0.15 0.20 0.25 0.0 0.2 0.4 0.6 0.8 Time [s] I N C X (p A /p F) EGTA (0.05 mM) EGTA (10 mM) 200 ms-50 mV 0 mV 0.00 0.10 0.20 0.30 0.40 0.50 0.0 0.2 0.4 0.6 Time [s] [C a2 + ] i ( nM ) EGTA (0.05 mM) EGTA (10 mM) 200 ms-50 mV 0 mV 0.00 0.10 0.20 0.30 0.40 0.50 400 500 600 700 800 900 Time [s] [C a2 + ] SR (n M ) EGTA (0.05 mM) EGTA (10 mM) 200 ms-50 mV 0 mV 0.00 0.10 0.20 0.30 0.40 0.50 1.6 1.7 1.8 1.9 2.0 Time [s] Sa rc om er L en gt h (µ m ) EGTA (0.05 mM) EGTA (10mM) 200 ms-50 mV 0 mV A B C D E F 69 Computational Model Simulation of Spatiotemporal Ca2+ Dynamics We also simulated the spatial effects of Ca2+ diffusion in rabbit atrial cells, which lack an organized tubular membrane system in contrast with rabbit ventricular cardiac myocytes. Figure 3.7 displays the simulation results, indicating that in atrial cells the effect of lower diffusivity is more pronounced than in ventricular cells. The effect on the centripetal [Ca2+]i transient was also more pronounced in the atrial cell representation. This simulation was limited to Ca2+ diffusion in a simplified atrial myocyte model with Ca2+ depleted intracellular stores to capture the effect of extracellular Ca2+ sources and diffusivity in the absent of CICR. Our results seem to indicate that low diffusivity plays an important role in decreasing the [Ca2+] amplitude in the center of the model geometry. Figure 3.7 (B) shows Ca2+ diffusion with similar conditions in a model that lacks TTs. Our spatial modeling results suggest that the t-tubular network plays an important role in synchronizing and speeding Ca2+ propagation (compare A to B). The simulation also predicted that the effect of buffering and SR re-uptake in slowing Ca2+ diffusion would be more pronounced in atrial myocytes, since they lack the membrane invaginations that provide an extracellular Ca2+ source to the center of the cell. 70 Figure 3.7 Spatial Modeling Simulation of Ca2+ Diffusion (A) Representation of rabbit ventricular cell geometry with organized TTs un- buffered conditions (left panel) and buffered condition (right panel). (B) Representation of rabbit atrial cell geometry without organized TTs, un- buffered conditions (left panel) and buffered conditions (right panel). 71 4. Discussion Based on these experimental results, we expected that the increased buffering strength observed in rabbit RAP myocytes would decrease CDI, which would increase the total amount of Ca2+ entering the cell and available to trigger CICR, and which could at least partially balance the reduction in the peak ICa,L current density observed during RAP. Our experimental and modeling results, shown in Figure 3.2 and Figure 3.4, respectively, demonstrated that buffering plays a major role in the inactivation profile of ICa,L. However, increases in buffering had insignificant effects in the amplitude of ICa,L as shown by both experimental and compartmental modeling results. As expected, our stochastic modeling predicts that high buffering reduces [Ca2+]i transient amplitude. Reduced cytosolic [Ca2+]i then leads to less Ca2+ reuptake through the SERCA ATPase resulting in lower [Ca2+]SR. Lower [Ca2+]i also significantly inhibits contraction and leads to less Ca2+ extrussion through NCX. Our spatial modeling simulations predicted that limited Ca2+ diffusion into atrial cells that have increased buffering power should result in much lower [Ca2+ ]i concentrations in the center of these cells, specially when relying solely on extracellular calcium flux without the regenerative effect of CICR. These results seem to confirm the observation of Ca2+ signaling silencing, in which increased Ca2+ buffering strength contributes to the failed propagation of the [Ca2+]i signal to the myocyte center, both in patients with AF and in the rapid atrial paced rabbit model (31). In particular, the 72 experiments performed by Greiser et al. demonstrated that pharmacological reduction of ICa,L did not interrupt the centripetal intracellular Ca2+ wave, but which was interrupted when they increased Ca2+ buffering strength by using BAPTA. Moreover, other groups have shown that due to the regenerative effect of CICR, the Ca2+ transient in the central region of the cell is relative independent of ICa,L (134,135). These effects play an important role in electrical contraction coupling in atrial myocytes. For example, it seems that atrial myocytes may compensate for the lack or reduction of TTs by having reduced CRU axial spacing. An atrial structural study performed by Brandenburg and colleagues indicates that the inter-CRU spacing was reduced from 1 µm in ventricular myocytes to less than 600 nm in atrial myocytes (27). 5. Conclusions As demonstrated by the mathematical modeling simulation, Ca2+ buffering plays an important role in the spatial profile of the Ca2+ signal and, therefore, may be an important mechanism in the molecular and cellular adaptation that occurs during AF. As shown in the simulation results, high buffering contributes to the reduction of the [Ca2+]i signal in the center of atrial cell. This is in agreement with the calcium silencing mechanism described by Greiser et al., in which high buffering was a major feature of this process and contributed to the failed propagation of the [Ca2+]i signal in the myocyte center of the RAP rabbit model and in patients with AF. 73 Our experimental measurements and compartmental modeling results also indicate that high concentrations of EGTA should be avoided in patch clamp experiments if one is characterizing the physiological kinetic properties of ICa,L. However, if Ca2+ chelation cannot be avoided, our results indicate that EGTA is preferable to BAPTA, due to its slower binding kinetics and its lack of effects on the current amplitude. The next chapter will continue to explore Ca2+ signaling using reactive oxygen species. 74 Chapter 4 : Effects of Reactive Oxygen Species on Calcium Induced Calcium Release and the Action Potential 1. Introduction A major goal of this project is to investigate the role of reactive oxygen species (ROS) in excitation contraction coupling in atrial myocytes and in doing so to replicate some of the adaptations that occur during atrial fibrillation. In particular we are interested in two main areas, the effects of ROS on Calcium Induced Calcium Release (CICR) and the role of ROS in modifying the morphology of the action potential (AP). Understanding these effects is important because experimental studies have reported increases in ROS sources during AF (41,42,82,136,137). Therefore, focusing on these two ROS effects will contribute not only to our understanding of normal atrial function, but also ECC adaptations during atrial fibrillation, which can lead to the development of new and more effective therapeutics. In the first part of this chapter, we focus on characterizing ROS effects on diastolic Ca2+ release from the sarcoplasmic reticulum. As explained in the introduction, increases in spontaneous RyR2 openings result in SR Ca2+ leak, which can lead to inward diastolic INCX currents, which in turn can then generate delayed after depolarizations that sustain the chaotic electrical propagation observed in AF (80,110,138). Prosser et al. discovered a ROS mediated mechanism linking cellular stretch in ventricular myocytes to facilitation of intracellular Ca2+ release (45). ROS appears to oxidize the 75 ryanodine receptors of the sarcoplasmic reticulum and increase their sensitivity to [Ca2+]i. We hypothesized that this mechanism may also be operational in atrial cells, since both atrial and ventricular myocytes contain the same cardiac ryanodine receptor isoform (i.e., RyR2), which are located along the Z-lines in both cell types (28,135), albeit having some spatial distribution differences such as the number of RyR2 per CRU and the spacing between CRUs (27,31). In the second part of this chapter we focus on ROS effects during systole. In particular, the effect that increasing ROS has on AP morphology. We hypothesized that ROS reduces the Action Potential Duration (APD) that is known to enable reentrant wavelets to propagate in the atrial tissue to maintain the AF rhythm (80,139,140). The repolarization phase of the action potential is initiated through the activation of potassium channels when the cell is depolarized and contributes significantly to the APD. When K+ channels in the cell membrane activate following depolarization, they produce an outward K+ current that is responsible for repolarizing the cell towards the resting membrane potential (17). Na+,K+-ATPase uses ATP to maintain a large K+ gradient across the cell membrane (i.e., [K+]o is typically 5.4 mM while [K+]i is typically 140 mM) needed for this repolarization to occur. However, the role of ROS in potassium channels and potassium currents is still unclear. ROS studies on different potassium channels do not seem to indicate that ROS directly enhances the outward potassium currents through the 76 diverse potassium channels present in the cardiac cell membrane (141), except for a moderate effect in enhancing the inward rectifying potassium current or IK1 (142). However, there is evidence that oxidative stress and ROS leads to the loss of the mitochondrial membrane potential, causing an increase in Adenosine Diphosphate to Adenosine Triphosphate ratio (ADP/ATP) in the cytosol (143,144). The increase in the ADP/ATP ratio then leads to the activation of the ATP sensitive potassium channels, which causes the cell membrane to depolarize faster and shortens the APD (145- 150). We believe that studying these two effects of ROS on diastolic Ca2+ release and systolic APD will enhance our understanding of both normal ROS signaling and ROS signaling during AF. In these experiments we have chosen to study the effects of hydrogen peroxide (H2O2) because it is more stable than other reactive oxygen species like superoxide, it can penetrate the sarcolemma and it is an endogenous signaling molecule in many cell types (40,151). 77 2. Methods Cell Isolation Rabbit myocyte isolation was done as described in earlier chapters. For this project, only rabbit cells were used. Confocal Ca2+ imaging Atrial [Ca2+]i transients and sparks were measured using Ca2+ sensitive dyes. The myocytes were incubated for 20 minute with 10 µM Fluo-4 AM and 20% Pluronic ® F127 (a poloxamer made by BASF, Florham Park, N.J., USA) in a dimethyl sulfoxide (DMSO) solvent. Cells were imaged using an inverted laser confocal microscope (LSM 510 Carl Zeiss). A wavelength of 488 nm from the argon laser in confocal line-scan mode scanned the cell at a rate of 1.92 ms/line. Automated analysis of line-scan images for Ca2+ spark location and properties were performed using the open source Sparkmaster software which can be downloaded as a plugin of Image J, freely provided by the National Institutes of Health (NIH) (152). Ca2+ Spark Measurements under Controlled H2O2 Delivery Using a pneumatic pico-pump (Model PV8630, World Precision Instruments (WPI)) the 20 µM H2O2 in Tyrode’s solution was delivered around the cell through a picospritzer. The signals were controlled with high temporal resolution using a Pulsemaster (Model A300, WPI), which allows the precise control of the timing and duration of the pico-pump. The in/out trigger functions of the LSM 510 confocal microscope were used to control and 78 create precise markers when the solution was been delivered. 10 µM sulforhodamine was added to the Tyrode’s solution to validate the spatiotemporal location of the H2O2 solution around the cell. Sulforhodamine has a different excitation-emission spectrum than Fluo-4 AM; therefore both indicators were imaged and tracked using different lasers and optical filters in the LSM microscope multitrack mode. Figure 4.1 portrays this methodology and validation. 79 Figure 4.1. Spatiotemporal Control of H2O2 Delivery. (A) Field image during H2O2 and sulforhodamine (red) exposure. This figure shows the spatial control of H2O2 around the atrial myocyte. (B) Multitrack line-scan displaying spatiotemporal delivery control. When the pump is turned on through the confocal microscope timer, the picospritzer is able to deliver the H2O2 solution with exact timing, validated using the sulforhodamine track. 80 Action Potential Measurements Electrophysiological measurements were done by applying the techniques described in Walden et al. and Greiser et al. (31,117) .The patch- clamp technique in conventional whole-cell mode was used with an Axopatch 200B amplifier (Molecular Devices) and a computer equipped with the Clampex and Clampfit software (Molecular Devices). Briefly, a clean glass micropipette was fused to the cell membrane to create a seal with resistance in the giga-ohm range. The seal was created by gently approaching the cell membrane with a pipette and applying suction until a 2 GΩ seal was achieved. Access into the cell is then gained by gently rupturing the cell membrane with rapid suction (17). Pipettes with resistance of 1.5 - 2.5 MΩ were made with a micropipette puller (Model P-97 Flaming/Brown). Table 4.1 depicts the solutions that were used for the AP measurements. AP measurements were initiated by depolarizing current pulses (1 ms duration, 1.5 x threshold) at the desired rate under whole cell current clamp conditions. A minimal current was injected to hold the cell at -70 mV at rest. Experiments were performed at room temperature (21-23°C). To measure the effect of ROS on the APD, 100 µM H2O2 was added to the Tyrode’s perfusion solution (Table 2.2) once the action potentials were recorded at steady state with normal Tyrode’s perfusion. After 5 minutes of H2O2 perfusion, action potentials were then recorded again for comparison. 81 Pipette solution Concentration (mM) KCl 11 K-aspartate 120 Mg-ATP 5 NaCl 5 MgCl 0.5 GTP 0.5 EGTA 0.05 HEPES 10 pH adjusted to 7.2 with KOH Table 4.1. Solution for AP recordings. Mathematical Compartmental Modeling To simulate the effect of ROS on the action potential we used the compartmental model that includes ionic currents and Ca2+ handling based on the Grandi et al models of the human ventricular myocyte (153) and the human atrial myocyte (99). We updated the Grandi atrial and ventricular model by incorporating the mathematical formalism for IK,ATP developed by Ferrero et al. (154). This IK,ATP formalism had been integrated by Zhou et al. in a computational model of excitation-contraction coupling linked to mitochondrial bioenergetics. The Zhou model also incorporates mitochondrial ROS-induced ROS release with coupling between the mitochondrial energy state to calculate the ATP and ADP concentrations that control the activation of IK,ATP (144): 𝑰𝒌,𝑨𝑻𝑷 = 𝜎 ∙ 𝑝! ∙ 𝑔! ∙ 𝑓!"# ∙ (𝑉! − 𝐸!) 82 𝑓!"# = 11+ ([𝐴𝑇𝑃]! 𝐾!)! 𝐾! = 35.8+ 17.9 ∙ ([𝐴𝐷𝑃]!)!!,!"# 𝐻 = 1.3+ 0.74 ∙ 𝑒(!!!"#$ ∙[!"#]!) Where, 𝑓!"# (Fraction of activated channels) σ = 0.6 channels/µm! (Channel density) 𝑔! = 30.95 (Unitary conductance) 𝑝! = 0.91 (Maximum channel open probability) 𝐸! = −84 𝑚𝑉 (KATP channel reverse potential) 𝐻!"# = −0.001 (Hill Coefficient) 𝐾!,!"# = 0.25 (Km Constant) The rest of the ionic currents were modeled as in the atrial and ventricular model of Grandi et al. Mathematical Spatial Modeling To model the Ca2+ spark rate appropriately, stochastic RyR2 gating is necessary; therefore we used the modeling framework of a mouse ventricular myocyte developed by Williams et al.(63) and Wescott et al. (101) and updated it to account for our rabbit atrial myocyte parameters as shown in Table 4.2. We also used the 3 dimensional (3D) spatial implementation in the Wescott et al. local control model that represents a transverse subsection of a cardiomyocytes to simulate the experimental atrial myocyte Ca2+ confocal line scans before and after exposure to ROS. 83 The effect of ROS on the rabbit atrial mycocyte Ca2+ spark was simulated by increasing the RyR2 opening rate by 20 %. The modified parameters were based on the experimental results of Brandenburg et al., which demonstrated that in atrial cells the transverse inter-CRU distance is shorter than in ventricular myocytes (27), and Boknik et. al., which compared the SERCA pump parameters in rabbit atrial and ventricular myocytes (155). The model confocal Ca2+ line scans were then analyzed using the Sparkmaster module in Image J with the same thresholds as those used in the experimental Ca2+ line scans analysis. Table 4.2. Spatial rabbit Ca2+ spark modeling parameters. N Number of CRUs 20 000 nRyR Number of RyR2s per CRU 35 dCRUy Transverse inter-CRU distance 0.52 µm Ap SERCA density 110 channels/ µm2 Statistical Analysis Results are reported as means ± SEM for the indicated number of cell measurements. Statistical analyses were performed using two-tailed Student t test for unpaired values. Values of p<0.05 were considered significant. All statistical analyses were completed using Prism 7. 84 3. Results Reactive Oxygen Effects in Ca2+ Spark Rate Experimental Results We imaged the Ca2+ release from quiescent rabbit atrial cells during 30 second laser confocal line scans. At 10 seconds we turned on the pressure connected to the picospritzer pipette filled with 20 µM H2O2, and directed it towards the myocyte undergoing imaging. We then turned off the pressure after 10 seconds to understand ROS washout effects. We found that in rabbit atrial cells the baseline Ca2+ spark rate, which is a metric of diastolic Ca2+ release, was 0.69 ± 0.2 sparks.s-1 .100µm-1. Under 20 µM H2O2 the spark rate almost double and increased to 1.15 ± 0.3 sparks.s-1 .100µm-1. Once the H2O2 delivery was stopped the Ca2+ spark rate went down to almost baseline levels 0.8 ± 0.3 sparks.s-1 .100µm-1. The experimental results shown in Figure 4.2 support our hypothesis that ROS sensitizes the RyR2s in atrial myocytes to increase their open probability. These experiments also validate our methodology to deliver pharmacologic agents with high spatiotemporal control in our cellular experiments. The next section explores the quantitative evaluation of ROS mediated increases in the RyR2s sensitivity. 85 Figure 4.2. Ca2+ Spark Rates in Cells Treated with 20 µM H2O2 (A) Representative line-scan of a single rabbit atrial myocyte before, during and after treatment with 20 µM H2O2 for 10 seconds. (B) Ca2+ spark rate before, during and after exposure to H2O2 (above) and associated pooled Ca2+ spark count (below) (n = 14 cells). Pre H2O2 Post 0.0 0.5 1.0 1.5 2.0 2.5 C a2 + S pa rk R at e (s -1 1 00 µ m -1 ) * 2 6 10 14 18 22 26 30 0 5 10 15 20 Time (s) C a2 + Sp ar k C ou nt A B Pre 20 µM H2O2 Post 86 Stochastic Spatial Modeling Results To analyze the amount of RyR2 sensitization by ROS that is necessary to reproduce the increase in Ca2+ spark rate observed in our experiments, we simulated 20 confocal line scans in the 3D stochastic spatial model and analyzed the sparks characteristics using Image J with the same parameters as those used to analyze experimental Ca2+ line scans. We found that increasing the RyR2 open probability by 20 % provides adequate agreement and a quantitative measure of the sensitization that occurs during ROS exposure as can be observed in Figure 4.3. 87 Figure 4.3 Spatially Modeled Ca2+ Sparks (A) Representative simulated line-scan before, during and after increasing the RyR2 open probability by 20%. (B) Ca2+ spark rate before, during and after changing RyR2 open probability (above) and associated Ca2+ spark count (bellow) (n = 20 simulated scans). A B Pre H2O2 Post 0.0 0.5 1.0 1.5 2.0 C a2 + S pa rk R at e (s -1 1 00 µm -1 ) *** * 2 6 10 14 18 22 26 30 0 5 10 15 20 Time (s) C a2 + Sp ar k C ou nt Pre RyR2 Po (120%) Post 88 Reactive Oxygen Effects in Ca2+ Spark Characteristics Experimental Results Using the publically available sparkmaster plugin for Image J, we analyzed the characteristics of each Ca2+ spark. Our first parameter to analyze was Ca2+ spark amplitude, which is associated with the amount of Ca2+ released in each spark. When the Ca2+ signal is not calibrated against known Ca2+ concentrations, Ca2+ spark amplitude is confounded with the quality and variability of the Ca2+ indicator loading, as well as laser power used during the confocal line recording. However, because we compared Ca2+ spark rate in the same cell and the same confocal line scan with the same laser power during baseline and exposure to ROS, these confounding variables do not affect our assessment, and the ratio of the amplitude of the spark compared to baseline values (F/Fo) is an adequate representation of the amount of Ca2+ released in each spark. We found that the Ca2+ spark amplitude increased by 40 % when the cell was exposed to ROS, and it recovered to baseline values after we stopped exposure as can be observed in panel A of Figure 4.4. We also measured the full width at half maximum (FWHM) for sparks, as can be seen in Figure 4.4 B. We found that rabbit atrial sparks at baseline have a FWHM of 1.9 ± 0.1 µm but increased by 44% to 2.73 ± 0.17 µm when the cell was exposed to H2O2 and returned to baseline values after exposure. The spark mass, which is a volumetric estimate of the Ca2+ spark, was calculated using the equation validated by Hollingworth et al (156) and shown 89 in panel C of Figure 4.4. The results show that spark mass was significantly greater during exposure to H2O2, which is expected since spark mass is related to both the amplitude and FWHM which showed similar trends. Finally we analyzed the full duration at half maximum (FDHM) of the Ca2+ sparks, as shown in panel D of Figure 4.4. At baseline the FDHM was 46.52 ± 4.06 ms and nearly doubled to 79.09 ± 6.3 ms during exposure to H2O2. Unlike the rest of the Ca2+ sparks characteristics, the FDHM remained elevated after washout. Modeling Simulation Results We then analyzed the spark characteristics of the spatial model simulated line scans and found that the model agrees with the trends observed in the experimental results. We analyzed the model sparks with Image J using the same threshold parameters used in the experimental analyzes for equivalent comparison. The modeled Ca2+ sparks characteristics are shown in Figure 4.5. 90 Figure 4.4. Rabbit Atrial Ca2+ Spark Characteristics (A) Ca2+ Spark amplitude before, during and after ROS exposure. (B) Ca2+ Spark FWHM before, during and after ROS exposure. (C) Ca2+ Spark mass calculates as Amplitude*1.206 *(FWHM)3 . (D) Ca2+ Spark FDHM before, during and after ROS exposure. n = 110 Ca2+ Sparks, * p value < 0.05, **p value < 0.01, *** p value < 0.001, ****is p value < 0.0001. Pre H2O2 Post 0.0 0.5 1.0 1.5 2.0 Sp ar k A m pl itu de (F /F o) Spark Amplitude * *** Pre H2O2 Post 0 1 2 3 4 FW H M (µ m ) FWHM **** ** Pre H2O2 Post 0 50 100 150 200 Sp ar k M as s (F /F ο *µ m 3 ) Spark Mass ** ** Pre H2O2 Post 0 20 40 60 80 100 FD H M (m s) **** FDHM A C B D 91 Figure 4.5. Spatially Modeled Ca2+ Spark Characteristics (A) Ca2+ Spark amplitude before, during and after ROS exposure. (B) Ca2+ Spark FWHM before, during and after ROS exposure. (C) Ca2+ Spark mass calculates as Amplitude*1.206 *(FWHM)3 . (D) Ca2+ Spark FDHM before, during and after ROS exposure. n = 43 sparks Pre H2O2 Post 0.0 0.5 1.0 1.5 2.0 Sp ar k A m pl itu de (F /F o) Spark Amplitude Pre H2O2 Post 0.0 0.5 1.0 1.5 2.0 FW H M (µ m ) FWHM Pre H2O2 Post 0 2 4 6 8 10 Sp ar k M as s (F /F ο *µ m 3 ) Spark Mass Pre H2O2 Post 0 20 40 60 80 FD H M (m s) FDHM A C B D 92 Ca2+ Spark Characteristics Comparison Between Atrial and Ventricular Rabbit Myocytes To understand any differences that might occur in the Ca2+ spark characteristics between atrial and ventricular myocytes, we also performed Ca2+ confocal line scan imaging on rabbit ventricular myocytes. We found that the spatial characteristics were equivalent between atrial and ventricular myocytes, as shown by the amplitude, FWHM and spark mass comparison in Figure 4.6. However, we found a statistically significant longer Ca2+ spark FDHM in rabbit ventricular myocytes when compared to rabbit atrial myocytes. 93 Figure 4.6. Ca2+ Spark Characteristics in Rabbit Ventricular and Atrial Myocytes (A) Ca2+ Spark amplitude in atrial and ventricular myocytes. (B) Ca2+ Spark FWHM in atrial and ventricular myocytes. (C) Ca2+ Spark mass in atrial and ventricular myocytes. (D) Ca2+ Spark FDHM in atrial and ventricular myocytes. n = 110 atrial sparks, n= 136 ventricular sparks, ****is p value < 0.0001. Atria Ventricle 0.0 0.5 1.0 1.5 Sp ar k A m pl itu de (F /F o) Spark Amplitude ns Atria Ventricle 0.0 0.5 1.0 1.5 2.0 2.5 FW H M (µ m ) FWHM ns Atria Ventricle 0 20 40 60 Sp ar k M as s (F /F ο * µ m 3 ) Spark Mass ns Atria Ventricle 0 20 40 60 80 100 FD H M (m s) **** FDHM A B C D 94 Reactive Oxygen Effects in Action Potential Duration Experimental Results We measured the AP at steady state during normal Tyrode’s perfusion and after adding 100 µM H2O2 to the bath solution in both atrial and ventricular myocytes. We found that the baseline APD in rabbit atrial cells was 80.8 ± 5.5 ms and that it was reduced by almost 50 % after 5 minute of perfusion with H2O2 to a new AP duration of 41 ± 2.4 ms. Although not as pronounced, we found a significant reduction in APD in the same experiments conducted on rabbit ventricular myocytes, which had a baseline value of 227.9 ± 9.7 ms that decrease by 28% to a new APD of 164.1 ± 7.7 ms. These results are shown in Figure 4.7. 95 Figure 4.7. ROS Effect on the AP of Rabbit Myocytes (A) Representative steady state APs in an atrial myocyte before and after 5 minutes of 100 µM H2O2 bath perfusion. (Right) Statistics of the APD before and after 100 µM H2O2 perfusion (n = 80 control measurements, n =75 control measurements). (B) Representative steady state APs in a ventricular myocyte before and after 5 minutes of 100 µM H2O2 bath perfusion. (Right) Statistics of the APD before and after 100 µM H2O2 perfusion (n = 40 control measurements, n = 40 control measurements) ****is p value < 0.0001. Compartmental Modeling Results 0 100 200 300 400 500 -100 -50 0 50 100 Time (ms) Vo lta ge (m V) Control 100 µM H2O2 0 100 200 300 400 500 -100 -50 0 50 100 Time (ms) Vo lta ge (m V) Control 100 µM H2O2 Pre 100 µM H2O2 0 20 40 60 80 100 A PD 90 (m s) **** Pre 100 µM H2O2 0 50 100 150 200 250 A PD 90 (m s) **** Atrial Ventricle A B 96 We used our updated versions of the Grandi atrial and ventricular models with the incorporation of the mathematical formalism for IK,ATP of Ferrero et al. and Zhou et al. to test whether increasing the physiological levels of ADP 10X to simulate ROS effects on the loss of mitochondrial membrane potential would account for the decrease in the AP duration shown in the experiments above. We ran both the atrial and ventricular models maintaining the ATP levels at 5 mM (i.e., to mimic our experimental conditions where ATP was clamped to 5 mM in the patch pipette solution as shown above in Table 4.1). We chose 35 µM as the physiological ADP concentration based on the experimental literature (13) and compared it with an increase to 350 µM to simulate ROS effects. Upon ADP rise in the atrial computational simulation, the peak of the IK,ATP current increased 3X from 0.86 pA/pF to 2.65 pA/pF, and the full profile of this current measured with an AP clamp is shown in panel A of Figure 4.8. We also simulated increasing the concentration of ADP 10X in the ventricular model and found a similar effect, with the IK,ATP current increasing from 0.78 pA/pF to 2.56 pA/pF as shown in panel B in Figure 4.8. The model also showed that the increase in the IK,ATP current has a profound impact on the APD as seen in the experiments. As can be observed in Figure 4.9, the atrial model predicted the near 50 % decrease in APD from 183.3 ms to near 106 ms when ADP was increased 10X in the simulations. The ventricular model predicted a slightly more modest APD decrease from 185 ms to 120 ms upon increasing the ADP/ATP ratio. 97 98 Figure 4.8. Computational IK,ATP in Atrial and Ventricular Model. (A) Representative atrial IK,ATP recordings under an AP clamp at simulated ADP concentrations of 35 µM and with 350 µM to simulated ROS effect. (B) Representative ventricular IK,ATP recordings under an AP clamp at simulated ADP concentrations of 35 µM and 350 µM. A B 0 100 200 300 400 500 0 1 2 3 Time (ms) I K ,A TP (p A /p F) ADP 35 µM ADP 350 µM 0 100 200 300 400 500 0 1 2 3 Time (ms) I K ,A TP (p A /p F) ADP 35 µM ADP 350 µM Atrial IK,ATP Ventricle IK,ATP 99 Figure 4.9. Computationally Modeled ROS Effects on the AP (A) Representative atrial modeled APs in the presence of ADP concentrations of 35 µM (control) and with 350 µM (simulated ROS effects). (Right) APD comparison under different concentrations of ADP. (B) Representative ventricular modeled APs in the presence of ADP concentrations of 35 µM and with 350 µM. (Right) APD comparison under different concentrations of ADP. 0 100 200 300 400 500 -100 -50 0 50 Time (ms) Vo lta ge (m V) ADP 35 µM ADP 350 µM ADP 35 µM ADP 350 µM 0 50 100 150 200 A PD 90 (m s) ADP 35 µM ADP 350 µM 0 50 100 150 200 A PD 90 (m s) Atrial Computation Ventricular Computation A B 0 100 200 300 400 500 -100 -50 0 50 Time (ms) Vo lta ge (m V) ADP 35 µM ADP 350 µM 100 4. Discussion In this chapter we have validated a methodology to deliver a precise concentration of a chemical with high spatiotemporal control to a myocyte being confocal imaged. Using this methodology, we were able to study the effects of ROS, delivered as 20 µM H2O2, on the diastolic Ca2+ release. Many ROS studies on ventricular myocytes isolated from different species have shown that ROS sensitizes the RyR2, which leads to increases in diastolic Ca2+ release (44,45,157,158). In agreement with those studies, we found that ROS also increases the diastolic Ca2+ release in rabbit atrial myocytes, measured here as the increase in Ca2+ spark rate. Our results were quantitatively explained using a stochastic mathematical model of the RYR2 that has been previously validated against experimental data (63,101). The model showed that the 20 % increase in the RyR2 open probability is able to capture the ROS effects on Ca2+ spark rate seen experimentally. We also observed experimentally that ROS increases the width, amplitude and duration of Ca2+ sparks. This spatiotemporal increase in spark dimension would activate additional nearby CRUs, which would then trigger arrhythmogenic Ca2+ waves (45). This is relevant because the increase in various sources of ROS during atrial fibrillation (82) could cause an increase in diastolic Ca2+ release, and this excess of diastolic Ca2+ would trigger NCX to extrude Ca2+, creating an inward current as 3 Na+ ions come into the cell for each of the Ca2+ ions that gets extruded. This inward current can then depolarized the cell outside the normal heart rate timing, causing 101 ectopic activity that promotes the chaotic electrical propagation in atrial tissue (80,159). The second part of this project focused on ROS effects during systole, especially the changes that ROS causes to the action potential. Our observation that ROS, perfused as 100 µM H2O2, shortens the rabbit action potential duration by half is consistent with the decrease in the effective refractory period observed during AF. Shorter APs allow for reentrant waves to propagate in the atrial tissue, thereby maintaining the AF chaotic rhythm (80,160). A mathematical model in which ROS leads to mitochondrial membrane potential decoupling, increasing the concentration of ADP in the cell, which in turn activates the outward ATP sensitive potassium channel current, leading to action potential shortening, theoretically explained these experimental observations. Although further experiments are needed to demonstrate this hypothesis, this theory is supported by the experimental work on potassium ATP channels by Nichols and Lederer (146,147). Moreover, mitochondrial dysfunction, which included decrease mitochondria respiration with higher concentrations of ADP, has been observed during AF (161). 5. Conclusion The results in this chapter demonstrate that ROS plays an important role in the mechanism that leads to Ca2+ signaling dysfunction during AF. First, ROS increases sensitization of the RyR2 channels, which can contribute to increase diastolic Ca2+ release leading to Ca2+ waves and 102 arrhythmogenesis. Moreover, ROS seems to affect mitochondria function, which causes a myocyte energetic imbalance that affects the ATP sensitive outward potassium current and leads to AP shortening, which could then promote reentrant chaotic activity in the atrial tissue, maintaining the AF. In the next chapter we will explore Ca2+ and Na+ signaling imbalances in a murine model with sustained atrial fibrillation. 103 Chapter 5 : Arrhythmogenic Na+/Ca2+ Imbalance in a Murine Transgenic Model of AF 1. Introduction There are two main theories on the mechanism of AF. The first one was addressed in the previous chapter and involves a reduction in AP duration, which would enable reentrant electrical activity within the atrial tissue. The second one involves unstable calcium signaling that could lead to delayed-depolarizations as reviewed by Heijman et al (80). One cause of unstable calcium signaling may involve [Na+]i. Intracellular sodium concentration ([Na+]i) is an important regulator of intracellular Ca2+ signaling (51). However, little is known about its role in excitation contraction coupling in atrial cells. Because intracellular Ca2+ signaling is profoundly altered during AF and sodium and calcium homeostasis in cardiac cells are intertwined, the role [Na+]i may play and its dysregulation in disease in atrial myocytes could be significant. We will explore this topic using a transgenic mouse model of AF, which has been found to have sustained AF. The first part of this chapter will focus on validation of the methodology to measure [Na+]i reliably in mouse atrial cells, isolated as described in Chapter 2. Measuring [Na+]i reliably has been challenging due to several limitations of available techniques. For example, Na+ sensitive microelectrodes (162) are limited to large cells, like purkinje fibers (163), 104 because they required the penetration of two electrode tips, which can permit extracellular Na+ to leak into the cell. Intracellular fluorescence indicators are, therefore, preferable for measuring Na+ in atrial myocytes. The only available sodium ratio-metric dye is sodium-binding benzofuran isophthalate (SBFI), available in two forms, as a salt and as an acetoxymethyl (AM) ester (164). A salt version needs to be delivered directly into the cell (i.e., patch pipette), but the AM ester form is cell permeable and once in the cytosol its AM ester group is cleaved leaving the indicator trapped inside the cell (165). This AM version has the advantage in that it can measure [Na+]i non-invasively, avoiding the challenges of modifying the intracellular cytosolic milieu with the solution in the patch pipette, which is necessary to deliver the SBFI salt version inside the cell. However, SBFI-AM has slower binding kinetics, variability in de-esterification and low signal to noise ratio that have to be taken into account in our imaging set up to allow for accurate [Na+]i measurements in murine atrial myocytes. The second part of this chapter involves the application of the [Na+]i measurement techniques to study [Na+]i homeostasis during control and drug application in a transgenic murine model of AF. We used transgenic mice expressing a human NaV1.5 variant with a mutation in the anesthetic-binding site (F1759A-NaV1.5), which causes these mice to have a persistent late INa current. These mice were a generous contribution from Dr. Steve Marx, Columbia University, who has demonstrated that Na+ channel inactivation is sufficient to drive structural alterations, including atrial and ventricular 105 enlargement, fibrosis and electrophysiological dysfunctions, that together lead to spontaneous and prolonged episodes of AF in these mice (93). As mentioned in Chapter 1, upon the arrival of an action potential, the voltage gated sodium channels (NaV1.5) activate, allowing the fast inward INa current to quickly depolarize the cell membrane (17). After opening briefly during the upstroke of the AP, each NaV1.5 promptly inactivates until repolarization is concluded. NaV1.5 channels that open after the AP upstroke create a small but persistent current that remains throughout the plateau phase of the cardiac AP. During this phase the membrane resistance is high and even a small increase of an inward current can cause AP prolongation (93). Late INa has been shown to be increased in ischemia, oxidative stress and heart failure. Many of these conditions are associated with an increased incidence of AF (83-86). Moreover, late INA has been shown to be increased in AF (84,166,167). Therefore, our goal in the second half of this project is to understand how [Na+]i homeostasis is regulated in this transgenic murine model of AF with persistent late INa (INa,Late). [Na+]i can also provide insight into the activation of the sarcolemmal sodium calcium exchanger (NCX) and the behavior of both Na+ channels and the Na+,K+-ATPase. Our purpose is to understand the role that [Na+]i may play in AF and whether its adaptation in disease is stabilizing as in the rabbit rapid atrial pace model of AF. For instance, Greiser et al. has shown that a high rate (10 Hz) of atrial activation for 5 days in a rabbit model leads to a 106 significant reduction in [Na+]i (31). This reduction in [Na+]i is one of the hallmarks of a novel mechanism called “calcium signaling silencing”, which is a distinct cellular and molecular adaptive response to rapid cardiac activation (81). This has never been demonstrated in a mouse model of AF. Studying [Na+]i homeostasis and its role in Ca2+ signaling in this transgenic mouse model would also enhance our understanding of sodium and calcium signaling adaptations in AF, especially because this model has sustained episodes of AF. Our hypothesis is that these adaptation mechanisms during AF involve both calcium-signaling instabilities promoting the AF rhythm and compensatory calcium signaling silencing to stabilize the atrial tissue. 2. Methods Transgenic Murine Model of AF The murine model of AF used in these experiments was a double transgenic murine line that contained an alanine substitution in the voltage gated sodium channel gene at position F1759 with a modified tetracycline inducible promoter (rtTA). This mouse line was designated FLAG-F1759A- NaV1.5. Mice that were only rtTA positive but did not have the F1759A mutation were used as littermate controls and were designated WT-rtTA. Only the double transgenic mouse presented the AF phenotype in an extensive characterization performed by Wan and colleagues (93). 107 Cell Isolation Only murine atrial cells were used in this project. The mice cell isolation was conducted with the protocol described in Chapter 2, with minor adjustments in isolation time in the transgenic murine model due to increased atrial size and fibrosis due to the sustained atrial fibrillation in this model. Intracellular Sodium Indicator Cell Loading We used the cell permeant acetoxymethyl (AM) ester variant of the fluorescent indicator sodium-binding benzofuran isophthalate (SBFI-AM) (Molecular Probes). To aid in the dispersion of the dye and to achieve homogenous cell loading, the optimized pluronic surfactant polyol called PowerLoad (Thermo Fisher Scientific) was also used along with diamethyl sulfoxide (DMSO). A 50 µg vial of the SBFI-AM (MW 1127) was first diluted in 8.8 µL of DMSO and 35 µL of PowerLoad (1mM stock solution). Then we used 11 µL of this dye stock per ml of cell suspension for a final concentration of 10 µM. The cells were incubated for 90 minutes on a rocker at room temperature. After washout of the extra dye, SBFI-AM de-esterification was allowed to continue for 30 minutes before proceeding for the [Na+]i measurements. [Na+]i Measurements and Instrumentation Figure 5.1 depicts the light path schematics used in our experimental configuration. A rapid switching DG5-plus illuminator (Sutter Instrument) using a 300 W xenon light source with wide field imaging provided dual UV 108 excitation (340 nm and 380 nm) with a frame time of 80 ms. The excitation light was attenuated using a neutral density (ND) filter with an optical density of 2 and also by reducing DG5 intensity to 33%. Fluorescence intensities measured at 340 and 380 nm wavelengths (F340 and F380) were collected at 510 ± 40 nm by an EMCCD camera (Princeton Instruments) connected to an inverted confocal microscope (Nikon Instruments). After 30 minutes of indicator de-esterification, freshly isolated atrial myocytes were concentrated and a cell suspension volume of 25 µL was placed as a droplet on a laminin coated glass cover slip to maximize the number of cells possible above the field of view of microscope 40X objective. 109 Figure 5.1. Light Path Schematic Used in [Na+]i Measurements (A) DG5-plus illuminator contains a full spectrum Xenon Lamp and fast switching mirrors that can direct the light to any of the available 5- excitation filter positions (we used two positions with filters 340 nm and 380 nm). Fluorescence images (F340 and F380) were collected at 510 ± 40 nm by an EMCCD camera connected to an inverted microscope. (B) SBFI spectral profile and excitation and emission filter transmission. 110 [Na+]i Calibration Because variability in de-esterification is a known limitation of SBFI- AM, we performed a calibration in each cell to accurately determine the relationship between the ratio of the SBFI-AM signals from the 340 nm and 380 nm excitation wavelength (F340/380) and the [Na+]i. After each physiological measurement, [Na+]i was calibrated by exposing the SBFI- loaded myocytes to a series of calibration solutions ([Na+]o, Table 5.1), in the presence of 10 µM gramicidin D, which forms a monovalent cation channel in the cell membrane, and 100 µM strophanthidin, which is a Na+,K+ ATPase pump blocker. The solutions of [Na+]o were made by mixing, in different proportions, two solutions of equal ionic strength as depicted in Table 5.1. An example of the calibration is shown in Figure 5.2. Na+ Solution Concentration (mM) K+ Solution Concentration (mM) HEPES 10 HEPES 10 Glucose 10 Glucose 10 EGTA 2 EGTA 2 NaCl 30 KCl 30 Na Gluconate 115 K Gluconate 115 pH adjusted to 7.2 Trisbase [Na+] (mM) K+ Solution (ml)/20ml Na+ Solution (ml)/20 ml 0 0 20 5 0.79 19.31 10 1.38 18.62 15 2.07 17.93 20 2.76 17.24 Table 5.1. [Na+]i Calibration Solutions 111 Figure 5.2. In situ SBFI Calibration. (A) Representative experiment in an atrial myocyte during Tyrode’s perfusion and using various [Na+]o solutions in the presence of 10 µmol/l gramicidin D and 100 µmol/l-strophanthidin. (B) Linear fit of this calibration demonstrating high fidelity and a direct relationship between the F340/380 signal and [Na+]i. (C) Atrial Myocyte F340/380 signal intensity increases during exposure to calibration solutions of increasing [Na+]o. 0 10 20 30 40 50 0.6 0.7 0.8 time (minutes) F 34 0/ 38 0 Tyrode Gramicidin/Strophanthidin 0 5 0 10 15 20 0 5 10 15 20 0.5 0.6 0.7 0.8 0.9 [Na]i (mM) F 34 0/ 38 0 0 mM [Na+] 5 mM [Na+] 10 mM [Na+] 15 mM [Na+] 20 mM [Na+] A B C 112 Anemonia Sulcata Toxin (ATX-II) Experiments In order to separate the effect of increasing the late INa from other changes that occur in our transgenic AF mice, we performed experiments in control mice exposed to the INa,Late enhancer ATX-II. ATX-II (Alomone Labs, Israel) was prepared by diluting this toxin in deionized water. ATX-II has a MW of 4935, and a 50 µg vial was diluted in 5 ml H2O for a stock concentration of 2 µM. This stock was frozen in 35 µL aliquots to be used each experimental day. During experiments, 30 µL of the ATX-II stock were added to 20 ml of Tyrode’s solution for an experimental concentration of 3 nM. Once the cells were de-esterified in normal Tyrode’s solution for 30 minutes, the bath perfusion was switched to 3 nM ATX-II Tyrode’s for 10 minutes before starting [Na+]i measurements. In the presence of ATX-II, [Na+]i was recorded for 5 minutes in quiescent cells, followed by 3 minutes of 1 Hz stimulation, and 2 minutes of 2 Hz and 3 Hz stimulations respectively. [Na+]i Extrusion in Atrial Myocytes To determine the relationship between [Na+]i and Na +,K+ ATPase pump activity, we measured the time course of Na+ efflux in quiescent myocytes loaded with SBFI as described by Despa and colleagues (168). After measuring baseline F340/380 levels for 5 minutes, the cells were perfused with a K+ free solution to stop the Na+,K+ ATPase pump, thereby loading the cells with Na+. Once the F340/380 had increased 3X from its baseline value, a Na+ free unloading solution was used to perfuse the cells and measure the rate of 113 decline of the [Na+]i before proceeding to the F340/380 [Na+]i calibration. The rate of [Na+]i decline was then plotted as a function of [Na+]i, and we fitted these data with a Hill expression: JNAK = Vmax/(1 + (Km/[Na+]i)n), where Vmax is maximal Na+ transport rate, Km is the [Na+]i for half-maximal stimulation and n is the Hill coefficient. The loading and unloading solutions used in the [Na+]i extrusion experiments are depicted in Table 5.2 Na+ Loading Solution Concentration (mM) Na+ Unloading Solution Concentration (mM) HEPES 10 HEPES 10 Glucose 10 Glucose 10 EGTA 2 EGTA 2 NaCl 145 TEACL 140 KCL 4 pH adjusted to 7.4 Trisbase Table 5.2. Sodium Extrusion Solutions [Ca2+]i Measurements Because sodium and calcium homeostasis in cardiac cells are intertwined, we also measured the [Ca2+]i in cells isolated from littermate controls and AF mice. Atrial myocytes were loaded with 1 µM of the ratio metric Ca2+ indicator Indo-1 (ThermoFisher Scientific), and intracellular Ca2+ transients were recorded with a 350 nm excitation light provided by the illuminator DG4 (Sutter Inc.). The dual fluorescence emissions of Indo-1 (i.e., 405 nm Ca2+ bound, 485 nm Ca2+ free) were collected with two 114 photomultiplier tubes (Aurora Scientific). Cells were maintained in normal Tyrode’s solution and [Ca2+]i was recorded at rest and during stimulation up to 3 Hz. Once the physiological measurements were completed, each cell underwent a calibration by exposing the cells to 2 µM of the ionophore ionomycin (Sigma-Aldrich) with a known minimum [Ca2+]i (i.e., 0 mM, 10 mM EGTA) and a known maximum [Ca2+]i (i.e., 1.8 mM). The following equation was used to convert the ratio of the signals (F405/F485) to the [Ca2+]i : [𝐶𝑎!!]! = 𝐾! ∙ 𝑄 ∙ 𝐹!"# 𝐹!"# − 𝑅!"#𝑅!"# − 𝐹!"# 𝐹!"# where, 𝑅!"# = 𝐹!"# 𝐹!"# 𝑤𝑖𝑡ℎ 𝑚𝑖𝑛𝑖𝑚𝑢𝑛 𝑘𝑛𝑜𝑤𝑛[𝐶𝑎!!]! 𝑑𝑢𝑟𝑖𝑛𝑔 𝑐𝑎𝑙𝑖𝑏𝑟𝑎𝑡𝑖𝑜𝑛 𝑅!!" = 𝐹!"# 𝐹!"# 𝑤𝑖𝑡ℎ 𝑚𝑎𝑥𝑖𝑚𝑢𝑛 𝑘𝑛𝑜𝑤𝑛[𝐶𝑎!!]! 𝑑𝑢𝑟𝑖𝑛𝑔 𝑐𝑎𝑙𝑖𝑏𝑟𝑎𝑡𝑖𝑜𝑛 𝐾! 𝐼𝑛𝑑𝑜_1 = 260 𝑛𝑀 𝑄 = 𝐹!"#,!"#𝐹!"#,!"# Modeling of [Na+]i Homeostasis in Atrial Myocytes To model [Na+]i homeostasis, we used the mathematical formulation of the Na+,K+ ATPase pump developed by Oka and colleagues (54). This formulation is a mechanistic model which reduced the 15 state reaction cycle of the Na+,K+ ATPase pump to 4 states, based on the mathematical formulation of Smith and Crampin (169). The 4 states represent intracellular binding, extracellular binding and two occlusion states (i.e., for Na+ and K+ transport). The cycle rate of the Na+,K+ ATPase pump is derived using the 115 King-Altman methods from all the transition states in the 4 state model. We reproduced the model in Matlab with our murine atrial myocytes dimensions and experimental parameters as shown in Table 5.3. Murine atrial cell dimensions were modeled as 80 µm in length, 10 µm in width, 10 µm in height based on our imaging experiments. Inward Na+ was assumed negligible because experimental [Na+]o was 0 mM during sodium unloading. Non Na+/K+ ATPase outward Na+ leak was assumed negligible based on the results of Despa et al. 2002 (168). dNKA Na+/K+ ATPase density 2000 channels/µm2 Vcytosol Atrial Cytosol Volume 8 pL Cm,atria Atrial Cell capacitance 80 pF Vm Membrane Potential -70 mV Temperature T 25 °C [Na+]o Extracellular Na+ concentration 0 mM [K+]o Extracellular K+ concentration 4 mM Table 5.3. Na+,K+ ATPase Pump Modeling Parameters We then integrated this Na+/K+ ATPase formalism into the mouse ventricular myocyte modeling framework developed by Williams et al. (63) and Wescott et al. (101), updated it to account for our mouse atrial myocyte parameters as shown in Table 5.4, and verified that steady state [Na+]i was balanced at rest before simulating pacing conditions. 116 Parameter Control AF Number of RyR2s per CRU 35 35 CRUs facing an LTCC 40 % 40 % ICa,peak 100 % 30 % INKA,peak 100 % 125 % Table 5.4. Murine Atrial Myocyte Modeling Parameters 3. Results Imaging Parameters Validation Because SBFI has slower binding kinetics, variability in de- esterification and low signal to noise ratio, we validated our loading and imaging methodology before proceeding to [Na+]i measurements. The first step in the validation was to ensure that the auto fluorescence emission from murine atrial myocytes was characterized so that the light emitted from the SBFI signal was not confounded. To do this, we measured the F340 and F380 from fresh isolated atrial myocytes without SBFI loading. The signal was acquired with the same imaging settings as those used in the sodium measurements experiments (e.g., light intensity, filter specifications). Before acquiring the UV signal, we located the ROI occupied by the cell using the microscope transmitted light as shown in the left panned of Figure 5.3 (A). Both the F340 and F380 were unchanged between the background ROI and the cell occupied ROI, indicating that at the excitation wavelength used, the cells do not emit any significant light. 117 We also characterized any harmful effects of the UV excitation light by using the antioxidant N-acetylcysteine (NAC) to determine if there was a change in the signal due to any oxidation produced by UV excitation of the SBFI loaded atrial myocytes. We first electrically stimulated the cell briefly with 1 HZ field pacing until the [Na+]i increased to its new steady state as demonstrated by the increase in the F340/380 signal. We then applied NAC to see if there was any signal reduction if the UV light was causing a rise in reactive oxygen species. As shown in Figure 5.3 (B) there was no change in the F340/380 signal, assuring that the steps taken to reduce the UV intensity on the cells were effective in maintaining the cell’s physiological functions. 118 Figure 5.3. Imaging Set-up Validation (A) Left panel shows the bright field image used to locate the cell occupied ROI while the right panel shows the ROI under the UV [Na+]i measurement parameters used in experiments. The results demonstrate there is not auto florescence signal from atrial myocytes under UV excitation. (B) Application of NAC did not change the F340/380 signal demonstrating no oxidation damage or change caused the UV excitation. F340 F380 0 200 400 600 800 In te ns ity ROIbackground ROIcell 0 5 10 15 20 0.95 1.00 1.05 1.10 1.15 1.20 1.25 Time (minutes) F 34 0/ 38 0 Tyrode 1Hz NAC A B 119 [Na+]i in Atrial vs. Ventricular Myocytes Although the objective of this project is to study atrial myocyte Na+ and Ca2+ signaling, we measured the [Na+]i in ventricular myocytes as a way to validate our [Na+]i measurements, since various groups have published the murine ventricular myocyte [Na+]i at rest. Our results show that murine ventricular myocytes have a resting [Na+]i of 10.74 ± 1.54 mM, which is comparable to the 11.1 ± 1.8 mM measured by Despa et al. (170) and also the 12 ± 1 mM measured by Correll et al. (171) in murine ventricular cells. Resting [Na+]i in mouse atrial myocytes was significantly different than ventricular myocytes, with a baseline value of 7.6 ± 0.22 mM, which is 30 % lower than in ventricular myocytes. This suggests a significant difference between excitation-contraction coupling between these two cell types. 120 Figure 5.4. [Na+]i in Quiescent Atrial and Ventricular Myocytes. (A) From left to right, images showing the simultaneous recording of 340 nm excitation, 380 nm excitation image and F340/380 image of atrial myocytes loaded with SBFI perfused with normal Tyrode’s solution. (B) Representative transmitted light images of an atrial (left) and a ventricular (right) cardiac myocyte. (C) [Na+]i in quiescent atrial and ventricular cardiac myocytes. **p < 0.01; t tests. Atria Ventricles0 5 10 15 [N a+ ] i (m M ) N = 26 59 cells N = 3 5 cells ** 10 μm 10 μm 10 μm A B C 121 Frequency-dependent Increase in [Na+]i We report here for the first time the steady states of [Na+]i in murine atrial cells during pacing, using frequencies of up to 3 Hz. As shown and estimated in ventricular myocytes (150,168), sodium influx increases during stimulation, due to the influx through the Nav1.5 channels during depolarization and via NCX during relaxation, as 1 Ca2+ ion is extruded from the cytosol in exchange for 3 Na+ ions entering the cell. At each pacing frequency, [Na+]i in ventricular myocytes reaches a new steady state where the increased Na+ influx is balanced by the Na+ efflux through the Na+/K+ ATPase. We expect atrial cells to show the same changes in steady states with changes in frequency. Figure 5.5 (A) shows a representative experiment with mouse atrial myocytes, where we measured [Na+]i every 10 seconds during stimulation and perfusion with normal Tyrode’s solution. We then calibrated this signal, obtaining a linear relationship between the F340/380 signal and the known [Na+]o perfused in the presence of 10 µM gramicidin D and 100 µM strophanthidin. Similar to rat ventricular myocytes, we found that [Na+]i rises linearly with pacing frequency. In murine atrial cells, [Na+]i increased with a slope of 1.6 mM/Hz and an R2 value of 0.96 as shown Figure 5.5 B. The measured [Na+]i values were 7.6 ± 0.22 mM, 9.7 ± 0.34 mM, 11.4 ± 0.67 mM and 12.2 ± 0.7 mM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. When the stimulation was stopped, [Na+]i returned to resting levels in most of the experiments. 122 Figure 5.5. Frequency-dependent Increase in [Na+]i (A) Representative F340/380 measurement in a control mouse under field stimulation. (Inset) Liner fit of the F340/380 calibration. (Right) Calibrated experiment showing the corresponding [Na+]i values. (B) Frequency response of [Na+]i demonstrating a significant rise in [Na+]i as pacing rate is increased. (Right) Linear regression of the Frequency response of [Na+]i. Data are presented as mean ± SEM. ≠ p < 0.05 compared to previous pacing rate, ****p < 0.0001 compared with quiescent values using Ordinary One-way Anova with multiple comparisons. 0 5 10 15 0.90 0.95 1.00 Time (minutes) F 34 0/ 38 0 Field Stimulation Frequency 0 Hz 1 Hz 0 Hz2 Hz 3 Hz 0 1 2 3 0 5 10 15 [N a+ ] i (m M ) Stimulation Frequency (Hz) **** **** **** ≠ ≠ N = 26 59 cells N = 8 26 cells N = 11 36 cells N = 4 12 cells 0 5 10 15 5 10 15 Time (minutes) [N a+ ] i (m M ) Field Stimulation Frequency 0 Hz 1 Hz 0 Hz2 Hz 3 Hz 0 1 2 3 6 8 10 12 14 [N a+ ] i (m M ) Stimulation Frequency (Hz) 0 5 10 15 20 0.8 0.9 1.0 1.1 [Na]i (mM) F 34 0/ 38 0 R2 = 0.96 A B 123 Late INa Effect on [Na+]i in Atrial Myocytes In order to separate the effect of increasing the late INa from other changes that occur in our transgenic AF model, we first studied the frequency dependent modifications of [Na+]i in our control C57/BL6 mice using ATX-II. The effects of ATX-II in cardiac myocytes ionic currents have been extensively studied for the past 30 years. Based on the study of the effect of ATX-II on ionic membrane currents, Isenberg and colleagues proposed that ATX-II creates a modification in the Na+ channel and loads the cells with 0.5 mM Na during a prolongation of the AP of 2 s (86). It is now well established that ATX-II modulates voltage-gated Na+ channel gating kinetics by delaying its inactivation and prolonging the action potential in many species and tissue types (92,172). Figure 5.6 A shows Anemonia sulcata from where the toxin is derived and an INa current trace from Kornyeyev et al., showing the persistent late INa in the presence of ATX-II (94). In our experiments we measured the [Na+]i in C57/BL6 mice during rest and during pacing at frequencies up to 3 Hz in the presence of ATX-II. We found that, as in the untreated cells (i.e., in the absence of ATX-II), the [Na+]i increased with pacing frequency. However, the sodium increased with a slope of 3.03 ± 0.06 mM/Hz, which is almost double that in untreated cells. The relationship between the [Na+]i and pacing rate was also linear, with a linear regression R2 value of 0.999 as shown in Figure 5.6. The measured [Na+]i values were 12.21 ± 1.05 mM, 15.03 ± 1.1 mM, 18.11 ± 0.99 mM and 21.29 ± 0.7 mM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing 124 frequency, respectively. These values show that much more sodium enters the cells through the ATX-II modulated Nav1.5 channel. We also suggest that as previously shown, the increase AP duration leads to additional cytosolic calcium through ICa,L, and the removal of this additional Ca2+ through NCX may also contribute to the additional rise of [Na+]i (94). 125 Figure 5.6. Late INa Effect on [Na+]i in Atrial Myocytes (A) Representative NaV1.5 current (INa) in cardiomyocytes showing an increased persistent current (INa,Late) in the presence of ATX-II. Current trace from Kornyeyev and Belardinelli (94) (AJP Heart allows graphic reuse in dissertations without special permission). (Right) Anemonia sulcata from where ATX-II is derived. (B) [Na+]i in atrial myocytes during stimulation in the presence of 3 nM ATX-II significantly increases in comparison to control at all frequencies. Data are presented as mean ± SEM. **p < 0.01, ****p < 0.0001; multiple t tests. A B 0 1 2 3 0 5 10 15 20 25 Stimulation Frequency (Hz) [N a+ ] i (m M ) Control ATX-II 3 nM **** **** **** ** 0 1 2 3 0 5 10 15 20 25 [N a+ ] i (m M ) Stimulation Frequency (Hz) Control ATX-II 3 nM R2 = 0.99 m = 3 mM/Hz R2 = 0.96 m = 1.6 mM/Hz 126 [Na+]i Extrusion in Atrial Myocytes Since the Na+/K+ ATPase pump is the main contributor to Na+ extrusion in cardiac myocyte, we measured the Na+-dependent Na+ extrusion behavior in murine atrial cells. We also used our experimental parameters to constrain the Oka 2010 mathematical formulation of the Na+/K+ ATPase. We then compared the simulation results with our Na+-dependent Na+ extrusion experimental data. We found that the physiological Vmax , or maximum rate of the pump, in control murine atrial myocytes is 8 ± 1.8 mM/min experimentally and 7.35 mM/min in the mathematical model prediction. The Km, or [Na+]i at a rate equal half Vmax, was 7.5 ± 0.73 mM experimentally and 10.7 mM according to the mathematical predictions. The pump rate at a resting [Na+]i of 7.6 mM was 2.41 mM/min experimentally and 2.014 mM/Min according to the model prediction. These results are shown in Figure 5.7. We also plotted the modeling results alongside the experimental values and they corroborate, for the most part, the predictions of the Oka 2010 Na+/K+ ATPase pump mathematical model when simulated with our experimental parameters. 127 Figure 5.7. Na+ Extrusion in Control Murine Atrial Myocytes (A) Representative example of Na+ extrusion experiment (left) and calibrated result (right) in an atrial myocyte. (B) Combined Na+ extrusion data and mathematical model prediction showing the rate of the pump at resting [Na+]i (left). Experimental and Modeling Vmax and Km of the Na+/K+ ATPase pump at physiological conditions. (n = 5 myocytes in experiments) 0 10 20 30 40 0.8 1.0 1.2 1.4 Time (minutes) F 34 0/ 38 0 Gramicidin/Strophanthidin 0 5 15 Tyrode 0K+ 0 Na+ 0 5 10 15 20 25 0 2 4 6 8 [Na+ ]i (mM) J N A K (m M /m in ) Experiments Oka Model 0 5 10 15 20 25 0 5 10 15 [Na+]i (mM) -d [N a+ ] i/ dt (m M /m in ) Exp. Model 0 2 4 6 8 10 V m ax (m M /m in ) Exp. Model 0 5 10 15 K m (m M ) A B 128 [Na+]i Homeostasis in Murine Atrial Myocytes from a Transgenic Model of AF Experimental Results We measured the steady stated [Na+]i in murine atrial cells from double transgenic AF mice (i.e., FLAG-F1759A-NaV1.5) and the single transgenic control littermates (WT-rtTa) at rest and during pacing frequencies of up to 3 Hz. As expected, Figure 5.8 shows that the single transgenic control littermates have equivalent [Na+]i concentrations at all frequencies as the C57BL/6 mice. The measured [Na+]i values in the littermate control mice were 8.4 ± 0.5 mM, 11.01 ± 0.8 mM, 12.4 ± 0.6 mM and 13.1 ± 0.8 mM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. The slope of the frequency dependent increase in [Na+]i in the control littermate was 1.55 mM/Hz, which compares well with the 1.6 mM/Hz observed in normal C57BL/6 mice. Very surprisingly, the [Na+]i concentration was significantly reduced at all frequencies in the AF double transgenic mice, by an average of 4.25 mM. This was unexpected given that these mice have a mutation in the voltage gated sodium channel which causes a persistent late INA, which we demonstrated increases the [Na+]i concentration at all pacing frequencies in the ATX-II experiments shown above. The measured [Na+]i values in the double transgenic mice were 5.12 ± 0.7 mM, 6.5 ± 0.6 mM, 7.53 ± 0.36 mM and 8.6 ± 0.5 mM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing frequency, 129 respectively. Moreover, the slope of the frequency dependent increase in [Na+]i was 1.15 mM/Hz, which is 26 % lower than in the control littermates. Mathematical Modeling Results To simulate the experimental observations we tested whether an up- regulation of the Na+/K+ ATPase, which was modeled as a 25% increase in peak current, and a reduction of the L-type calcium current (ICa,L), which was modeled as a 70% reduction in peak current, would lead to the [Na+]i reduction observed in the AF transgenic mice experiments. Our mathematical modeling results corroborating a decrease in [Na+]i at all pacing frequencies in the AF modeling simulation are shown in Figure 5.9. 130 Figure 5.8. [Na+]i in Atrial Myocytes from a Transgenic AF Murine Model (A) [Na+]i in atrial myocytes during stimulation in Tyrode’s bath solution in transgenic (FLAG-F1759A-Nav1.5) and their control littermates (WT-rtTA). (B) Linear regression of the frequency response of [Na+]i in both transgenic and control myocytes. Data are presented as mean ± SEM. ****p < 0.0001; multiple t tests. (n = 13 wild type cells, n = 8 transgenic cells) A B R2 = 0.99 m = 1.15 mM/Hz R2 = 0.93 m = 1.55 mM/Hz 0 1 2 3 0 5 10 15 Stimulation Frequency (Hz) [N a+ ] i (m M ) WT-rtTa FLAG-F1759A-NaV1.5 **** **** **** **** 0 1 2 3 0 5 10 15 [N a+ ] i (m M ) Stimulation Frequency (Hz) WT-rtTa FLAG-F1759A-NaV1.5 131 Figure 5.9. Computationally Modeled [Na+]i in Atrial Myocytes (A) [Na+]i in control and AF mathematically modeled conditions. (B) Linear regression of the frequency response of [Na+]i in the mathematical simulation. R2 = 0.99 m = 1.86 mM/Hz R2 = 0.98 m = 2.45 mM/Hz 0 1 2 3 5 10 15 Stimulation Frequency (Hz) [N a+ ] i (m M ) Control AF 0 1 2 3 5 10 15 [N a+ ] i (m M ) Stimulation Frequency (Hz) Control AF A B 132 [Ca2+]i Homeostasis in Murine Atrial Myocytes from a Transgenic Model of AF Experimental Results The [Na+]i reduction at all frequencies in the AF double transgenic mice suggests that despite the increase in the persistent component of the voltage gated sodium current, the AF mice seem to have undergone a stabilizing adaptation that actually lowers the intracellular sodium concentration. To understand if this adaptation also involves [Ca2+]i reduction, we measured the [Ca2+]i using whole cell epifluorescence during field stimulation up to 3 Hz. Our [Ca2+]i measurement results are shown in Figure 5.10 and indicate that [Ca2+]i during diastole and systole is lower in the transgenic mice than in the littermate as pacing frequency is increased. The baseline diastolic measurements in the control littermates (WT-rtTA) were 47.5 ± 2.7 nM, 56.7 ± 2.9 nM, 87.6 ± 5.2 nM and 144.9 ± 16.5 nM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. However, these values were decreased in the FLAG-F1759A-Nav1.5 AF mice, where the baseline diastolic measurements were 40.7 ± 3.9 nM, 47.5 ± 3.96 nM, 60 ± 4.38 nM and 78.2 ± 5.3 nM at 0 Hz, 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. During systole the [Ca2+]i transient amplitude was also decreased in the AF mice. The peak [Ca2+]i during systole in control mice were 217.4 ± 16.15 nM, 306.1 ± 23.4 nM and 495.1 ± 43.68 nM at 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. These values were significantly lower in the 133 AF mice and were, 200.7 ± 16.9 nM, 240.3 ± 15.95 nM and 304.1 ± 21.1 nM at 1 Hz, 2 Hz and 3 Hz pacing frequency, respectively. Mathematical Modeling Results Our mathematical modeling in which Na+/K+ ATPase was up-regulated while ICa,L current peak was reduced to model the changes predicted in AF, also reflected a reduction of the [Ca2+]i transient amplitude during systole and lower diastolic [Ca2+]i as can be observed Figure 5.11. 134 Figure 5.10. Experimental [Ca2+]i in Atrial Cells from an AF Murine Model (A) Representative [Ca2+]i measurements under pacing frequencies up to 3 Hz in control littermates (left, WT-rtTa) and transgenic mice (right, FLAG- F1759A-Nav1.5). (B) (Left) Frequency response of [Ca2+]i during diastole, demonstrating significantly lower baseline [Ca2+]i in transgenic mice with respect to controls as pacing rate is increased. (Right) Frequency response of the [Ca2+]i transient during systole, demonstrating it is significantly lower in transgenic myocytes with respect to controls as pacing rate is increased, **p < 0.01, ****p < 0.0001, multiple t test, n= 16 littermate cells, n=15 transgenic cells. 135 Figure 5.11. Computationally Modeled [Ca2+]i in Atrial Myocytes (A) Representative [Ca2+]i measurements under pacing frequencies up to 3 Hz in control and AF mathematically modeled conditions. (B) (Left) Frequency response of [Ca2+]i during diastole demonstrating significantly lower baseline [Ca2+]i in the AF mathematical model conditions. (Right) Frequency response of the [Ca2+]i transient during systole showing lower [Ca2+]i during AF. 136 4. Discussion In this chapter, we show for the first time the role [Na+]i in normal and AF animals. We also demonstrate the difference between atrial and ventricular sodium parameters. To begin to elucidate the role of sodium, we first verified a reliable technique to measure [Na+]i in murine atrial myocytes in several ways. First, we demonstrated that our UV excitation light was sufficiently reduced to prevent any changes driven by an increase in oxidation, yet with enough excitation power to acquire a reliable signal. Second, we demonstrated that there was not a significant contribution from auto-florescence sources inside the cell. Third, we demonstrated that our calibrated signal had a linear relationship to the [Na+]i concentration, which is an intrinsic property of the SBFI signal and has been shown in other studies (91,168). Furthermore, we measured the [Na+]i in ventricular myocytes as a way to validate our [Na+]i measurements, since various groups have published the murine ventricular myocyte [Na+]i at rest. Our results show equivalent results as those previously measured by Despa et al (170) and Correll et al. (171) in murine ventricular cells. Furthermore, these data demonstrate the differences between sodium handling in ventricular myocytes and atrial myocytes, suggesting that there may be significant differences in ECC coupling that need to be explored in the future. We are the first to measure and show [Na+]i in quiescent and paced murine atrial myocytes. We found that [Na+]i in atrial myocytes is lower than in 137 ventricular myocytes. We believe this finding is supported by a calcium signaling study performed by Walden and colleagues, comparing rat atrial and ventricular cardiac myocytes (117). They found that SR Ca2+ content was much higher in rat atrial cells than in ventricular cells. This suggests that in atrial cells the balance is shifted so that during relaxation, there is much more calcium reuptake into the SR than calcium extruded through NCX. Since NCX transports three sodium ions into the cytosol for each Ca2+ that is extruded, this could mean that less sodium is coming in through NCX in atrial myocytes. Therefore the steady state [Na+]i would be lower in atrial myocytes in comparison to ventricular myocytes, as our data show. Our measurement of [Na+]i under stimulation frequencies up to 3 Hz indicated that sodium influx increases during stimulation. This is expected due to the Na+ influx through the Nav1.5 channels during depolarization and via NCX during relaxation as 1 Ca2+ ion is extruded in exchange for 3 Na+ ions entering the cell. At each pacing frequency, [Na+]i reached a new steady state where the increased Na+ influx was balanced by the Na+ efflux through the Na+/K+ ATPase. We are the first to confirm this in mouse atrial cells. Moreover, the Na+-dependent Na+ extrusion behavior in murine atrial cells as characterized in our experiments were corroborated by the Oka 2010 Na+/K+ ATPase mathematical formulation, when updated with our murine atrial cell parameters. To isolate the effect of the voltage gated sodium channel mutations from the adaptations that occur during AF, we measured [Na+]i in control cells 138 exposed to ATX-II and found [Na+]i significantly increased under ATX-II at all pacing frequencies . Our results are in agreement with the findings of Hoey et al. who studied the effect of ATX-II perfusion in rat and guinea pig ventricular myocytes. They found that at a pacing rate of 0.2 Hz in cells perfused with ATX-II, [Na+]i increased by 1.9 and 2.25 mM in rat and guinea-pig cells, respectively. They also found that these increases were associated with enhanced cell shortening of 195 ± 31 % and 317 ± 250 %, respectively (85). More recently, Kornyeyev et al. demonstrated that treatment with ATX-II increased late INa 8.7-fold in rabbit ventricular myocytes (94). On the other hand, when the voltage gated sodium channel was mutated in the transgenic model of AF used in this study, it lead to an increased INa,Late that was not followed by increases in [Na+]i. On the contrary, [Na+]i was significantly reduced in this AF mouse model at all pacing frequencies, indicating that adaptive mechanisms during AF must have contributed to compensatory changes that actually reduced [Na+]i. These finding are coupled with our measured decreases in [Ca2+]i, both in diastole and systole. Based on the mathematical model simulation we used and the sodium and calcium signaling changes others have observed during AF, we believe that this finding is due to three main adaptive mechanisms. First, there seems to be up regulation of the Na+/K+ ATPase in the AF mice because we find that their [Na+]i at rest is lower than in control littermates and Na+/K+ ATPase is the main source of Na+ extrusion. However, in a study comparing Na+/K+ ATPase function in cells isolated from patients 139 undergoing cardiac surgery with and without AF, Workman and colleagues did not find any difference in the sensitivity of the pump or in its voltage dependence (173). It is worth noting that one limitation with cells derived from patients undergoing cardiac surgery is that although they might not have the AF rhythm, they usually have other cardiovascular comorbidities, which limits their usefulness as true controls of physiological function. Moreover, the pump is modulated by short term factors that were not accounted for in this study (174). The second factor that we suggest plays a role in the decrease of the [Na+]i and [Ca2+]i in the AF mice is the reduction of the L-type calcium current ( ICa,L). It is well established that ICa,L is significantly reduced during AF (29,87- 89). We suggest that this reduced Ca2+ influx leads to a smaller cytosolic Ca2+ transient. This leads to smaller Ca2+ efflux through NCX, which in turn results in lower [Na+]i. The last factor that we believe plays a role is an increased Ca2+ buffering, which was studied in Chapter 3 of this dissertation. As mentioned before, increased Ca2+ buffering also contributes to a smaller Ca2+ transient (31,81), which also leads to a reduction of the [Na+]i influx through NCX. 5. Conclusion These reductions of [Na+]i and [Ca2+]i found here are in agreement with the “calcium signaling silencing” mechanism described by Greiser and colleagues (31,81). Our observations of [Na+]i and Ca2+ signaling in this transgenic mouse model have increased our understanding of sodium and 140 calcium signaling adaptations in AF. We have demonstrated that adaptation mechanisms during AF involve both calcium-signaling instabilities promoting the AF rhythm and compensatory calcium-signaling silencing to stabilize the atrial tissue. 141 Chapter 6 : Discussion and Future Direction Introduction The experimental and modeling work presented in this thesis highlights the unique features of Ca2+ signaling in atrial cells both during physiological conditions and during the mechanisms that lead to Atrial Fibrillation. In this chapter, we will integrate the findings of the previous chapters. Novel Atrial Myocytes Biological Techniques In the second chapter of this dissertation we described a method to isolate murine atrial cells and stabilize atrial myocytes for physiological experiments. Our results demonstrated that our isolation technique produced a high yield of healthy atrial cardiomyocytes from both mice and rabbits. The isolated cells had excellent structural and electrophysiological properties. We also determined and characterized a sapphire rod that could withstand the active force of atrial myocyte contractions to reduce the contraction artifact during experimentation, thereby increasing the experimental quality and reducing the signal to noise ratio that can be achieved in our experiments. By using this rod method, we could eliminate chemicals used to stop contractions that had been shown to be problematic for normal electrophysiological measurements. Effect of Ca2+ Buffering on Atrial Myocyte Pathophysiology In chapter three, we studied the effect of Ca2+ buffering in Ca2+ signaling and ECC in rabbit atrial myocytes that lack an organized T-tubular 142 system, similar to humans and different from the more well-studied ventricular myocytes. We demonstrated that Ca2+ buffering plays an important role in the spatial profile of the Ca2+ signal, and that it is an important mechanism in the molecular and cellular adaptation that occurs during atrial fibrillation. We also observed that increased buffering strength decreased the calcium dependent inactivation of the L type Ca2+ channel, which then increased the total amount of Ca2+ entering the cell and available to trigger CICR. We believe that this is an adaptive mechanism to partially balance the reduction in the peak ICa,L current density observed during AF. Additionally, our modeling simulations predict that increased Ca2+ buffering strength contributes to the failed propagation of the [Ca2+]i signal to the myocyte center, which is part of the Ca2+ signaling silencing observed in patients with AF (81). Effect of ROS Signaling on Electrical Contraction Coupling After validating a methodology to deliver a precise chemical concentration with high spatiotemporal control to a myocyte during confocal imaging, in chapter four we were able to study the effects of ROS on Ca2+ signaling. In agreement with studies performed in ventricular myocytes, we found that ROS also increases the diastolic Ca2+ release in rabbit atrial myocytes. Moreover, we were able to quantify this effect and use mathematical modeling to determine that this experimental RyR2 sensitization is equivalent to a 20% increase in the RyR2 open probability. This ROS- dependent RyR2 sensitization contributes to increase diastolic Ca2+ release leading to Ca2+ waves and arrhythmogenesis. Moreover we observed that 143 exposure to 100 µM H2O2 for 5 minutes, shortens the rabbit action potential duration by half. Our modeling simulation predicts that this observation is consistent with ROS affecting mitochondria function, which causes an energetic imbalance that affects the ATP sensitive outward potassium current and leads to AP shortening, which subsequently could then promote reentrant chaotic activity in the atrial tissue, maintaining the AF. Experimental Observations in a Murine Model of AF In chapter five, using a transgenic mouse model of AF, we characterized the intracellular sodium and calcium adaptive changes that occur during atrial fibrillation. Despite the voltage gated sodium channel mutation that leads to increased INa,Late, we found that [Na+]i and [Ca2+]i were significantly reduced in this AF mouse model at all pacing frequencies. Based on our integrated work in this dissertation, we suggest that there are mechanisms during AF that contribute to stabilizing compensatory changes. We believe that this finding is due to three main adaptive mechanisms. First, there seems to be an up regulation of the Na+/K+ ATPase in AF because we found that [Na+]i at rest is lower than in control littermates. Second, there seems to be a reduction of the L-type calcium current (ICa,L) that results in lower [Ca2+]i , which leads to lower NCX Na+ flux, which then lowers [Na+]i during stimulation. Third, we believe that increased Ca2+ buffering also contributes to a smaller Ca2+ transient (31,81), which also leads to a reduction of [Ca2+]i and thereby a reduction of Na+ influx through NCX. We believe that the reductions of [Na+]i and [Ca2+]i found in our murine 144 AF model are in agreement with the calcium signaling silencing mechanism described by Greiser and colleagues (31,81). Our results seem to indicate that in this murine model of AF, calcium signaling instabilities promoting the AF rhythm are counteracted by calcium signaling silencing to stabilize the atrial tissue. Future Work Although our experimental findings and modeling work have brought much understanding of the mechanism of Ca2+ signaling in healthy and disease atrial tissue, it has also generated important hypotheses that need to be experimentally tested in the future. Also, the new approaches developed in this work open many opportunities for novel experimentation. Our new sapphire rods can also be used to study contractile force in atrial myocytes from control and AF mice. It will be informative to know if the isometric force generated by the myocytes is reduced in the murine AF mouse model, since contractile dysfunction is a known characteristic of AF. Given our results of reduced [Ca2+]i during AF, it will also be important to expand the Ca2+ buffering study to measure and compare Ca2+ buffering strength in this murine AF model with controls, which we predict will be increased in AF to agree with our experimental findings. Our experimental results showed that exposure to 100 µM H2O2 (a ROS) for 5 minutes leads to significant reduction of the action potential duration. Our mathematical simulation predicts that this is due to the effect of ROS on activating KATP channels through mitochondrial decoupling. Energetic 145 experiments measuring the exact effect of ROS on the ATP/ADP ratio would need to be conducted to test this hypothesis. Finally the three proposed mechanisms that lead to reduced [Na+]i and [Ca2+]i in our murine model of AF would need to be further explored by studying LTCC and Na+/K+ ATPase expression and function in these mice. Conclusion Atrial Fibrillation is a complex disease involving at least two kinds of adaptation mechanisms. On one hand, there are maladaptive Ca2+ signaling instabilities promoting the AF rhythm, and on the other hand there are compensatory Ca2+ signaling silencing adaptations to stabilize the atrial tissue. Our results indicate that Ca2+ buffering, ROS and changes in [Na+]i are all important factors in this AF adaptation. 146 Bibliography 1.Fozzard, H.A. 1977. Heart: excitation-contraction coupling. Annu Rev Physiol 39:201-220. 2.Veenstra, R.D. and R.L. DeHaan. 1986. Measurement of single channel currents from cardiac gap junctions. Science 233:972-974. 3.Mehrzad, R., M. Rajab, and D.H. Spodick. 2014. The three integrated phases of left atrial macrophysiology and their interactions. Int J Mol Sci 15:15146-15160. 4.Munshi, N.V. 2012. Gene regulatory networks in cardiac conduction system development. Circulation research 110:1525-1537. 5.Hodgkin, A.L. and A.F. Huxley. 1952. A quantitative description of membrane current and its application to conduction and excitation in nerve. The Journal of physiology 117:500-544. 6.Walker, M.A., G.S. Williams, T. Kohl, S.E. Lehnart, M.S. Jafri, J.L. Greenstein, W.J. Lederer, and R.L. Winslow. 2014. Superresolution modeling of calcium release in the heart. Biophys J 107:3018-3029. 7.Lederer, W.J., E. Niggli, and R.W. Hadley. 1990. Sodium-calcium exchange in excitable cells: fuzzy space. Science 248:283. 8.Lederer, W.J., B.M. Hagen, and G. Zhao. 2012. Cell biology. Superresolution subspace signaling. Science 336:546-547. 9.Stern, M.D. 1992. Theory of excitation-contraction coupling in cardiac muscle. Biophys J 63:497-517. 10.Kostin, S., D. Scholz, T. Shimada, Y. Maeno, H. Mollnau, S. Hein, and J. Schaper. 1998. The internal and external protein scaffold of the T-tubular system in cardiomyocytes. Cell Tissue Res 294:449-460. 11.JS, F. 1990. Ultrastructure of the unfixed myocardial sarcolemma and cell surface, in Calcium and the Heart. Raven Press, New York. 12.Fabiato, A. and F. Fabiato. 1979. Calcium and cardiac excitation- contraction coupling. Annu Rev Physiol 41:473-484. 13.Bers, D.M. 2001. Excitation-Contraction Coupling and Cardiac Contractile Force. Springer Science & Business Media. 14.Bassani, R.A., J.W. Bassani, and D.M. Bers. 1992. Mitochondrial and sarcolemmal Ca2+ transport reduce [Ca2+]i during caffeine contractures in rabbit cardiac myocytes. The Journal of physiology 453:591-608. 147 15.Bassani, R.A. and D.M. Bers. 1994. Na-Ca exchange is required for rest- decay but not for rest-potentiation of twitches in rabbit and rat ventricular myocytes. Journal of molecular and cellular cardiology 26:1335-1347. 16.Boyman, L., A.C. Chikando, G.S. Williams, R.J. Khairallah, S. Kettlewell, C.W. Ward, G.L. Smith, J.P. Kao, and W.J. Lederer. 2014. Calcium movement in cardiac mitochondria. Biophys J 107:1289-1301. 17.Hille, B. 2001. Ion channels of excitable membranes. Sinauer, Sunderland, Mass. 18.Lederer, W.J. 2016. Cardiac Electrophysiology and the Electrocardiogram. In Medical Physiology. Elsevier Saunders Publisher. 19.Chen-Izu, Y., S.L. McCulle, C.W. Ward, C. Soeller, B.M. Allen, C. Rabang, M.B. Cannell, C.W. Balke, and L.T. Izu. 2006. Three-dimensional distribution of ryanodine receptor clusters in cardiac myocytes. Biophys J 91:1-13. 20.Huser, J., S.L. Lipsius, and L.A. Blatter. 1996. Calcium gradients during excitation-contraction coupling in cat atrial myocytes. The Journal of physiology 494 ( Pt 3):641-651. 21.Song, L.S., E.A. Sobie, S. McCulle, W.J. Lederer, C.W. Balke, and H. Cheng. 2006. Orphaned ryanodine receptors in the failing heart. Proceedings of the National Academy of Sciences of the United States of America 103:4305-4310. 22.Dibb, K.M., J.D. Clarke, M.A. Horn, M.A. Richards, H.K. Graham, D.A. Eisner, and A.W. Trafford. 2009. Characterization of an extensive transverse tubular network in sheep atrial myocytes and its depletion in heart failure. Circ.Heart Fail. 2:482-489. 23.Lenaerts, I., V. Bito, F.R. Heinzel, R.B. Driesen, P. Holemans, J. D'Hooge, H. Heidbuchel, K.R. Sipido, and R. Willems. 2009. Ultrastructural and functional remodeling of the coupling between Ca2+ influx and sarcoplasmic reticulum Ca2+ release in right atrial myocytes from experimental persistent atrial fibrillation. Circulation research 105:876-885. 24.Richards, M.A., J.D. Clarke, P. Saravanan, N. Voigt, D. Dobrev, D.A. Eisner, A.W. Trafford, and K.M. Dibb. 2011. Transverse tubules are a common feature in large mammalian atrial myocytes including human. Am J Physiol Heart Circ.Physiol 301:H1996-H2005. 25.Richards, M.A., J.D. Clarke, P. Saravanan, N. Voigt, D. Dobrev, D.A. Eisner, A.W. Trafford, and K.M. Dibb. 2011. Transverse tubules are a common feature in large mammalian atrial myocytes including human. 148 American journal of physiology. Heart and circulatory physiology 301:H1996- 2005. 26.Wagner, E., S. Brandenburg, T. Kohl, and S.E. Lehnart. 2014. Analysis of tubular membrane networks in cardiac myocytes from atria and ventricles. J Vis Exp:e51823. 27.Brandenburg, S., T. Kohl, G.S. Williams, K. Gusev, E. Wagner, E.A. Rog-Zielinska, E. Hebisch, M. Dura, et al. 2016. Axial tubule junctions control rapid calcium signaling in atria. The Journal of clinical investigation 126:3999-4015. 28.Greiser, M., W.J. Lederer, and U. Schotten. 2011. Alterations of atrial Ca(2+) handling as cause and consequence of atrial fibrillation. Cardiovasc.Res 89:722-733. 29.Greiser, M., C.R. Halaszovich, D. Frechen, P. Boknik, U. Ravens, D. Dobrev, A. Luckhoff, and U. Schotten. 2007. Pharmacological evidence for altered src kinase regulation of I (Ca,L) in patients with chronic atrial fibrillation 7. Naunyn.Schmiedebergs.Arch Pharmacol 375:383-392. 30.Legato, M.J. 1973. Ultrastructure of the atrial, ventricular, and Purkinje cell, with special reference to the genesis of arrhythmias. Circulation 47:178- 189. 31.Greiser, M., B.G. Kerfant, G.S. Williams, N. Voigt, E. Harks, K.M. Dibb, A. Giese, J. Meszaros, et al. 2014. Tachycardia-induced silencing of subcellular Ca2+ signaling in atrial myocytes. J Clin.Invest. 32.Maxwell, J.T. and L.A. Blatter. 2016. A novel mechanism of tandem activation of ryanodine receptors by cytosolic and SR luminal Ca2+ during excitation-contraction coupling in atrial myocytes. The Journal of physiology. 33.Christ, T., P. Boknik, S. Wohrl, E. Wettwer, E.M. Graf, R.F. Bosch, M. Knaut, W. Schmitz, et al. 2004. L-type Ca2+ current downregulation in chronic human atrial fibrillation is associated with increased activity of protein phosphatases. Circulation 110:2651-2657. 34.Hadley, R.W. and J.R. Hume. 1987. An intrinsic potential-dependent inactivation mechanism associated with calcium channels in guinea-pig myocytes. The Journal of physiology 389:205-222. 35.Dick, I.E., M.R. Tadross, H. Liang, L.H. Tay, W. Yang, and D.T. Yue. 2008. A modular switch for spatial Ca2+ selectivity in the calmodulin regulation of CaV channels. Nature 451:830-834. 149 36.Tadross, M.R., I.E. Dick, and D.T. Yue. 2008. Mechanism of local and global Ca2+ sensing by calmodulin in complex with a Ca2+ channel. Cell. 133:1228-1240. 37.Bers, D.M. 2002. Cardiac excitation-contraction coupling. Nature 415:198- 205. 38.Michailova, A., F. DelPrincipe, M. Egger, and E. Niggli. 2002. Spatiotemporal features of Ca2+ buffering and diffusion in atrial cardiac myocytes with inhibited sarcoplasmic reticulum. Biophys J 83:3134-3151. 39.Hool, L.C. 2006. Reactive oxygen species in cardiac signalling: from mitochondria to plasma membrane ion channels. Clin Exp Pharmacol Physiol 33:146-151. 40.Bamgboye, M. Acute ROS in Cardiac Calcium Signaling, University of Maryland Baltimore. University of Maryland Baltimore. 41.Mihm, M.J., F. Yu, C.A. Carnes, P.J. Reiser, P.M. McCarthy, D.R. Van Wagoner, and J.A. Bauer. 2001. Impaired myofibrillar energetics and oxidative injury during human atrial fibrillation. Circulation 104:174-180. 42.Cai, W., Q.D. Gao, L. Zhu, M. Peppa, C. He, and H. Vlassara. 2002. Oxidative stress-inducing carbonyl compounds from common foods: novel mediators of cellular dysfunction. Mol Med 8:337-346. 43.Dudley, S.C., Jr., N.E. Hoch, L.A. McCann, C. Honeycutt, L. Diamandopoulos, T. Fukai, D.G. Harrison, S.I. Dikalov, and J. Langberg. 2005. Atrial fibrillation increases production of superoxide by the left atrium and left atrial appendage: role of the NADPH and xanthine oxidases. Circulation 112:1266-1273. 44.Prosser, B.L., C.W. Ward, and W.J. Lederer. 2010. Subcellular Ca2+ signaling in the heart: the role of ryanodine receptor sensitivity. J Gen.Physiol 136:135-142. 45.Prosser, B.L., C.W. Ward, and W.J. Lederer. 2011. X-ROS signaling: rapid mechano-chemo transduction in heart. Science 333:1440-1445. 46.Kimura, S., G.X. Zhang, A. Nishiyama, T. Shokoji, L. Yao, Y.Y. Fan, M. Rahman, T. Suzuki, et al. 2005. Role of NAD(P)H oxidase- and mitochondria-derived reactive oxygen species in cardioprotection of ischemic reperfusion injury by angiotensin II. Hypertension 45:860-866. 47.Guo, J., W.R. Giles, and C.A. Ward. 2000. Effect of hydrogen peroxide on the membrane currents of sinoatrial node cells from rabbit heart. American journal of physiology. Heart and circulatory physiology 279:H992-999. 150 48.Viola, H.M., P.G. Arthur, and L.C. Hool. 2007. Transient exposure to hydrogen peroxide causes an increase in mitochondria-derived superoxide as a result of sustained alteration in L-type Ca2+ channel function in the absence of apoptosis in ventricular myocytes. Circulation research 100:1036-1044. 49.Greensmith, D.J., D.A. Eisner, and M. Nirmalan. 2010. The effects of hydrogen peroxide on intracellular calcium handling and contractility in the rat ventricular myocyte. Cell Calcium 48:341-351. 50.Thannickal, V.J. and B.L. Fanburg. 2000. Reactive oxygen species in cell signaling. Am J Physiol Lung Cell Mol Physiol 279:L1005-1028. 51.Bers, D.M., W.H. Barry, and S. Despa. 2003. Intracellular Na+ regulation in cardiac myocytes. Cardiovascular research 57:897-912. 52.Kimura, J., T. Ono, K. Sakamoto, E. Ito, S. Watanabe, S. Maeda, Y. Shikama, M.S. Yatabe, and I. Matsuoka. 2009. Na+ -Ca2+ exchanger expression and its modulation. Biol Pharm Bull 32:325-331. 53.Lewalle, A., S.A. Niederer, and N.P. Smith. 2014. Species-dependent adaptation of the cardiac Na+/K+ pump kinetics to the intracellular Na+ concentration. The Journal of physiology 592:5355-5371. 54.Oka, C., C.Y. Cha, and A. Noma. 2010. Characterization of the cardiac Na+/K+ pump by development of a comprehensive and mechanistic model. J Theor Biol 265:68-77. 55.Vaughan-Jones, R.D. and K.W. Spitzer. 2002. Role of bicarbonate in the regulation of intracellular pH in the mammalian ventricular myocyte. Biochem Cell Biol 80:579-596. 56.Crampin, E.J. and N.P. Smith. 2006. A dynamic model of excitation- contraction coupling during acidosis in cardiac ventricular myocytes. Biophys J 90:3074-3090. 57.Devalla, H.D., V. Schwach, J.W. Ford, J.T. Milnes, S. El-Haou, C. Jackson, K. Gkatzis, D.A. Elliott, et al. 2015. Atrial-like cardiomyocytes from human pluripotent stem cells are a robust preclinical model for assessing atrial-selective pharmacology. EMBO Mol Med 7:394-410. 58.Karakikes, I., M. Ameen, V. Termglinchan, and J.C. Wu. 2015. Human induced pluripotent stem cell-derived cardiomyocytes: insights into molecular, cellular, and functional phenotypes. Circulation research 117:80-88. 59.Du, D.T., N. Hellen, C. Kane, and C.M. Terracciano. 2015. Action potential morphology of human induced pluripotent stem cell-derived cardiomyocytes does not predict cardiac chamber specificity and is dependent on cell density. Biophys J 108:1-4. 151 60.Liu, J., D.K. Lieu, C.W. Siu, J.D. Fu, H.F. Tse, and R.A. Li. 2009. Facilitated maturation of Ca2+ handling properties of human embryonic stem cell-derived cardiomyocytes by calsequestrin expression. Am J Physiol Cell Physiol 297:C152-159. 61.Lundy, S.D., W.Z. Zhu, M. Regnier, and M.A. Laflamme. 2013. Structural and functional maturation of cardiomyocytes derived from human pluripotent stem cells. Stem Cells Dev 22:1991-2002. 62.Cheng, H., W.J. Lederer, and M.B. Cannell. 1993. Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science 262:740-744. 63.Williams, G.S., A.C. Chikando, H.T. Tuan, E.A. Sobie, W.J. Lederer, and M.S. Jafri. 2011. Dynamics of calcium sparks and calcium leak in the heart. Biophys J 101:1287-1296. 64.Cheng, H. and W.J. Lederer. 2008. Calcium sparks. Physiol Rev. 88:1491-1545. 65.Neher, E., B. Sakmann, and J.H. Steinbach. 1978. The extracellular patch clamp: a method for resolving currents through individual open channels in biological membranes. Pflugers Archiv : European journal of physiology 375:219-228. 66.Hamill, O.P., A. Marty, E. Neher, B. Sakmann, and F.J. Sigworth. 1981. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Archiv : European journal of physiology 391:85-100. 67.Clare, J.J. 2010. Targeting ion channels for drug discovery. Discov Med 9:253-260. 68.Li, C. 2008. A reliable whole cell clamp technique. Adv Physiol Educ 32:209-211. 69.Koizumi, R., K. Funamoto, T. Hayase, Y. Kanke, M. Shibata, Y. Shiraishi, and T. Yambe. 2015. Numerical analysis of hemodynamic changes in the left atrium due to atrial fibrillation. J Biomech 48:472-478. 70.Ceornodolea, A.D., R. Bal, and J.L. Severens. 2017. Epidemiology and Management of Atrial Fibrillation and Stroke: Review of Data from Four European Countries. Stroke Res Treat 2017:8593207. 71.Kannel, W.B., P.A. Wolf, E.J. Benjamin, and D. Levy. 1998. Prevalence, incidence, prognosis, and predisposing conditions for atrial fibrillation: population-based estimates. The American journal of cardiology 82:2N-9N. 152 72.Claes, N., C. Van Laethem, M. Goethals, P. Goethals, G. Mairesse, B. Schwagten, D. Nuyens, W. Schrooten, and J. Vijgen. 2012. Prevalence of atrial fibrillation in adults participating in a large-scale voluntary screening programme in Belgium. Acta Cardiol 67:273-278. 73.Colilla, S., A. Crow, W. Petkun, D.E. Singer, T. Simon, and X. Liu. 2013. Estimates of current and future incidence and prevalence of atrial fibrillation in the U.S. adult population. Am J Cardiol. 112:1142-1147. 74.Kumar, N., R. Khera, and N. Garg. 2016. Atrial fibrillation associated hospitalizations in patients with end-stage renal disease in the United States, 2003-2012. Heart Rhythm 13:2027-2033. 75.Vaughan Sarrazin, M.S., M. Jones, A. Mazur, P. Cram, P. Ayyagari, and E. Chrischilles. 2017. Cost of Hospital Admissions in Medicare Patients With Atrial Fibrillation Taking Warfarin, Dabigatran, or Rivaroxaban. Journal of the American College of Cardiology 69:360-362. 76.Cardiac Electrophysiology Device Branch; Garber, L. 2013. Clinical Study Designs for Surgical Ablation Devices for Treatment of Atrial Fibrillation, p. 18. In Federal Registry (Ed.). Federal Registry Food and Drug Administration; Silver Spring, MD. 77.Zimetbaum, P. 2012. Antiarrhythmic drug therapy for atrial fibrillation. Circulation 125:381-389. 78.Calkins, H., K.H. Kuck, R. Cappato, J. Brugada, A.J. Camm, S.A. Chen, H.J. Crijns, R.J. Damiano, Jr., et al. 2012. 2012 HRS/EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, endpoints, and research trial design. Europace 14:528-606. 79.Schotten, U., S. Verheule, P. Kirchhof, and A. Goette. 2011. Pathophysiological mechanisms of atrial fibrillation: a translational appraisal. Physiol Rev 91:265-325. 80.Heijman, J., N. Voigt, S. Nattel, and D. Dobrev. 2014. Cellular and Molecular Electrophysiology of Atrial Fibrillation Initiation, Maintenance, and Progression. Circulation research 114:1483-1499. 81.Greiser, M. 2017. Calcium signalling silencing in atrial fibrillation. The Journal of physiology 595:4009-4017. 82.Reilly, S.N., R. Jayaram, K. Nahar, C. Antoniades, S. Verheule, K.M. Channon, N.J. Alp, U. Schotten, and B. Casadei. 2011. Atrial sources of reactive oxygen species vary with the duration and substrate of atrial 153 fibrillation: implications for the antiarrhythmic effect of statins. Circulation 124:1107-1117. 83.Ward, C.A. and W.R. Giles. 1997. Ionic mechanism of the effects of hydrogen peroxide in rat ventricular myocytes. The Journal of physiology 500 ( Pt 3):631-642. 84.Fischer, T.H., J. Herting, F.E. Mason, N. Hartmann, S. Watanabe, V.O. Nikolaev, J.U. Sprenger, P. Fan, et al. 2015. Late INa increases diastolic SR-Ca2+-leak in atrial myocardium by activating PKA and CaMKII. Cardiovascular research 107:184-196. 85.Hoey, A., S.M. Harrison, M.R. Boyett, and U. Ravens. 1994. Effects of the Anemonia sulcata toxin (ATX II) on intracellular sodium and contractility in rat and guinea-pig myocardium. Pharmacol Toxicol 75:356-365. 86.Isenberg, G. and U. Ravens. 1984. The effects of the Anemonia sulcata toxin (ATX II) on membrane currents of isolated mammalian myocytes. The Journal of physiology 357:127-149. 87.Van Wagoner, D.R., A.L. Pond, M. Lamorgese, S.S. Rossie, P.M. McCarthy, and J.M. Nerbonne. 1999. Atrial L-type Ca2+ currents and human atrial fibrillation. Circulation research 85:428-436. 88.Skasa, M., E. Jungling, E. Picht, F. Schondube, and A. Luckhoff. 2001. L-type calcium currents in atrial myocytes from patients with persistent and non-persistent atrial fibrillation 2. Basic.Res Cardiol. 96:151-159. 89.Christ, T., P. Boknik, S. Wohrl, E. Wettwer, E.M. Graf, R.F. Bosch, M. Knaut, W. Schmitz, et al. 2004. L-type Ca2+ current downregulation in chronic human atrial fibrillation is associated with increased activity of protein phosphatases 20. Circulation 110:2651-2657. 90.Greiser, M. and U. Schotten. 2013. Dynamic remodeling of intracellular Ca(2)(+) signaling during atrial fibrillation. J Mol Cell.Cardiol. 58:134-142. 91.Willis, B.C., S.V. Pandit, D. Ponce-Balbuena, M. Zarzoso, G. Guerrero- Serna, B. Limbu, M. Deo, E. Camors, et al. 2016. Constitutive Intracellular Na+ Excess in Purkinje Cells Promotes Arrhythmogenesis at Lower Levels of Stress Than Ventricular Myocytes From Mice With Catecholaminergic Polymorphic Ventricular Tachycardia. Circulation 133:2348-2359. 92.Chahine, M., E. Plante, and R.G. Kallen. 1996. Sea anemone toxin (ATX II) modulation of heart and skeletal muscle sodium channel alpha-subunits expressed in tsA201 cells. J Membr Biol 152:39-48. 93.Wan, E., J. Abrams, R.L. Weinberg, A.N. Katchman, J. Bayne, S.I. Zakharov, L. Yang, J.P. Morrow, et al. 2016. Aberrant sodium influx causes 154 cardiomyopathy and atrial fibrillation in mice. The Journal of clinical investigation 126:112-122. 94.Kornyeyev, D., N. El-Bizri, R. Hirakawa, S. Nguyen, S. Viatchenko- Karpinski, L. Yao, S. Rajamani, and L. Belardinelli. 2016. Contribution of the late sodium current to intracellular sodium and calcium overload in rabbit ventricular myocytes treated by anemone toxin. American journal of physiology. Heart and circulatory physiology 310:H426-435. 95.Greiser, M., B.G. Kerfant, G.S. Williams, N. Voigt, E. Harks, K.M. Dibb, A. Giese, J. Meszaros, et al. 2014. Tachycardia-induced silencing of subcellular Ca2+ signaling in atrial myocytes. The Journal of clinical investigation 124:4759-4772. 96.Koivumaki, J.T., T. Korhonen, and P. Tavi. 2011. Impact of sarcoplasmic reticulum calcium release on calcium dynamics and action potential morphology in human atrial myocytes: a computational study. PLoS Comput.Biol 7:e1001067. 97.Li, Q., S.C. O'Neill, T. Tao, Y. Li, D. Eisner, and H. Zhang. 2012. Mechanisms by which cytoplasmic calcium wave propagation and alternans are generated in cardiac atrial myocytes lacking T-tubules-insights from a simulation study. Biophys J 102:1471-1482. 98.Thul, R., S. Coombes, H.L. Roderick, and M.D. Bootman. 2012. Subcellular calcium dynamics in a whole-cell model of an atrial myocyte. Proc.Natl.Acad.Sci U.S.A 109:2150-2155. 99.Grandi, E., S.V. Pandit, N. Voigt, A.J. Workman, D. Dobrev, J. Jalife, and D.M. Bers. 2011. Human atrial action potential and Ca2+ model: sinus rhythm and chronic atrial fibrillation. Circ.Res 109:1055-1066. 100.Maleckar, M.M., J.L. Greenstein, N.A. Trayanova, and W.R. Giles. 2008. Mathematical simulations of ligand-gated and cell-type specific effects on the action potential of human atrium. Prog.Biophys Mol Biol 98:161-170. 101.Wescott, A.P., M.S. Jafri, W.J. Lederer, and G.S. Williams. 2016. Ryanodine receptor sensitivity governs the stability and synchrony of local calcium release during cardiac excitation-contraction coupling. Journal of molecular and cellular cardiology 92:82-92. 102.James, J.F., T.E. Hewett, and J. Robbins. 1998. Cardiac physiology in transgenic mice. Circulation research 82:407-415. 103.Lockhart, M., E. Wirrig, A. Phelps, and A. Wessels. 2011. Extracellular matrix and heart development. Birth Defects Res A Clin Mol Teratol 91:535- 550. 155 104.Hein, S., S. Kostin, A. Heling, Y. Maeno, and J. Schaper. 2000. The role of the cytoskeleton in heart failure. Cardiovascular research 45:273-278. 105.Shioya, T. 2007. A simple technique for isolating healthy heart cells from mouse models. J Physiol Sci 57:327-335. 106.Mitra, R. and M. Morad. 1985. A uniform enzymatic method for dissociation of myocytes from hearts and stomachs of vertebrates. Am J Physiol 249:H1056-1060. 107.Hancox, J.C., A.J. Levi, C.O. Lee, and P. Heap. 1993. A method for isolating rabbit atrioventricular node myocytes which retain normal morphology and function. Am J Physiol 265:H755-766. 108.Pawloski-Dahm, C.M., G. Song, D.L. Kirkpatrick, J. Palermo, J. Gulick, G.W. Dorn, 2nd, J. Robbins, and R.A. Walsh. 1998. Effects of total replacement of atrial myosin light chain-2 with the ventricular isoform in atrial myocytes of transgenic mice. Circulation 97:1508-1513. 109.Wu, J.Y., J. Vereecke, E. Carmeliet, and S.L. Lipsius. 1991. Ionic currents activated during hyperpolarization of single right atrial myocytes from cat heart. Circulation research 68:1059-1069. 110.Li, N., D.Y. Chiang, S. Wang, Q. Wang, L. Sun, N. Voigt, J.L. Respress, S. Ather, et al. 2014. Ryanodine receptor-mediated calcium leak drives progressive development of an atrial fibrillation substrate in a transgenic mouse model. Circulation 129:1276-1285. 111.Voigt, N., X.B. Zhou, and D. Dobrev. 2013. Isolation of human atrial myocytes for simultaneous measurements of Ca2+ transients and membrane currents. J Vis Exp:e50235. 112.Li, T., N. Sperelakis, R.E. Teneick, and R.J. Solaro. 1985. Effects of diacetyl monoxime on cardiac excitation-contraction coupling. The Journal of pharmacology and experimental therapeutics 232:688-695. 113.Altamirano, J. and D.M. Bers. 2007. Effect of intracellular Ca2+ and action potential duration on L-type Ca2+ channel inactivation and recovery from inactivation in rabbit cardiac myocytes. American journal of physiology. Heart and circulatory physiology 293:H563-573. 114.Watanabe, Y., T. Iwamoto, I. Matsuoka, S. Ohkubo, T. Ono, T. Watano, M. Shigekawa, and J. Kimura. 2001. Inhibitory effect of 2,3- butanedione monoxime (BDM) on Na(+)/Ca(2+) exchange current in guinea- pig cardiac ventricular myocytes. Br J Pharmacol 132:1317-1325. 115.Iribe, G. and P. Kohl. 2008. Axial stretch enhances sarcoplasmic reticulum Ca2+ leak and cellular Ca2+ reuptake in guinea pig ventricular 156 myocytes: experiments and models. Progress in biophysics and molecular biology 97:298-311. 116.Bluhm, W.F., A.D. McCulloch, and W.Y. Lew. 1995. Active force in rabbit ventricular myocytes. J Biomech 28:1119-1122. 117.Walden, A.P., K.M. Dibb, and A.W. Trafford. 2009. Differences in intracellular calcium homeostasis between atrial and ventricular myocytes. J Mol Cell.Cardiol. 46:463-473. 118.Icardo, J.M. and E. Colvee. 2001. Origin and course of the coronary arteries in normal mice and in iv/iv mice. J Anat 199:473-482. 119.Yu, Z.B. and F. Gao. 2005. [Non-specific effect of myosin inhibitor BDM on skeletal muscle contractile function]. Zhongguo Ying Yong Sheng Li Xue Za Zhi 21:449-452. 120.Etnier, S.A. 2001. Flexural and torsional stiffness in multi-jointed biological beams. Biol Bull 200:1-8. 121.ICARDO, J.M. 2001. Origin and course of the coronary arteries in normal mice and in iv/iv mice. 122.Olson, T.M., A.E. Alekseev, X.K. Liu, S. Park, L.V. Zingman, M. Bienengraeber, S. Sattiraju, J.D. Ballew, et al. 2006. Kv1.5 channelopathy due to KCNA5 loss-of-function mutation causes human atrial fibrillation. Hum Mol Genet 15:2185-2191. 123.Berlin, J.R., J.W. Bassani, and D.M. Bers. 1994. Intrinsic cytosolic calcium buffering properties of single rat cardiac myocytes. Biophys J 67:1775-1787. 124.Bovo, E., S.R. Mazurek, M. Fill, and A.V. Zima. 2015. Cytosolic Ca buffering determines the intra-SR Ca concentration at which cardiac Ca sparks terminate. Cell Calcium. 125.Trafford, A.W., M.E. Diaz, and D.A. Eisner. 1999. A novel, rapid and reversible method to measure Ca buffering and time-course of total sarcoplasmic reticulum Ca content in cardiac ventricular myocytes. Pflugers Archiv : European journal of physiology 437:501-503. 126.Jafri, M.S. and J. Keizer. 1995. On the roles of Ca2+ diffusion, Ca2+ buffers, and the endoplasmic reticulum in IP3-induced Ca2+ waves. Biophys J 69:2139-2153. 157 127.Smith, G.D., J. Wagner, and J. Keizer. 1996. Validity of the rapid buffering approximation near a point source of calcium ions. Biophys J 70:2527-2539. 128.Yuan, W., K.S. Ginsburg, and D.M. Bers. 1996. Comparison of sarcolemmal calcium channel current in rabbit and rat ventricular myocytes. The Journal of physiology 493 ( Pt 3):733-746. 129.Dun, W., T. Yagi, M.R. Rosen, and P.A. Boyden. 2003. Calcium and potassium currents in cells from adult and aged canine right atria. Cardiovascular research 58:526-534. 130.Shannon, T.R., F. Wang, J. Puglisi, C. Weber, and D.M. Bers. 2004. A mathematical treatment of integrated Ca dynamics within the ventricular myocyte. Biophys J 87:3351-3371. 131.Wagner, E., M.A. Lauterbach, T. Kohl, V. Westphal, G.S. Williams, J.H. Steinbrecher, J.H. Streich, B. Korff, et al. 2012. Stimulated emission depletion live-cell super-resolution imaging shows proliferative remodeling of T-tubule membrane structures after myocardial infarction. Circ.Res 111:402- 414. 132.Parfenov, A.S., V. Salnikov, W.J. Lederer, and V. Lukyanenko. 2006. Aqueous diffusion pathways as a part of the ventricular cell ultrastructure. Biophys J 90:1107-1119. 133.Kushmerick, M.J. and R.J. Podolsky. 1969. Ionic mobility in muscle cells. Science 166:1297-1298. 134.Sheehan, K.A. and L.A. Blatter. 2003. Regulation of junctional and non- junctional sarcoplasmic reticulum calcium release in excitation-contraction coupling in cat atrial myocytes. The Journal of physiology 546:119-135. 135.Hatem, S.N., A. Benardeau, C. Rucker-Martin, I. Marty, P. de Chamisso, M. Villaz, and J.J. Mercadier. 1997. Different compartments of sarcoplasmic reticulum participate in the excitation-contraction coupling process in human atrial myocytes. Circulation research 80:345-353. 136.Van Wagoner, D.R. 2008. Oxidative stress and inflammation in atrial fibrillation: role in pathogenesis and potential as a therapeutic target. J Cardiovasc Pharmacol 52:306-313. 137.Roosimaa, M., T. Podramagi, L. Kadaja, A. Ruusalepp, K. Paju, R. Puhke, M. Eimre, E. Orlova, et al. 2013. Dilation of human atria: increased diffusion restrictions for ADP, overexpression of hexokinase 2 and its coupling to oxidative phosphorylation in cardiomyocytes. Mitochondrion 13:399-409. 158 138.Kirchhof, P., E. Marijon, L. Fabritz, N. Li, W. Wang, T. Wang, K. Schulte, J. Hanstein, et al. 2013. Overexpression of cAMP-response element modulator causes abnormal growth and development of the atrial myocardium resulting in a substrate for sustained atrial fibrillation in mice. International journal of cardiology 166:366-374. 139.Ravelli, F. and M. Allessie. 1997. Effects of atrial dilatation on refractory period and vulnerability to atrial fibrillation in the isolated Langendorff- perfused rabbit heart. Circulation 96:1686-1695. 140.Lee, A.M., A. Aziz, J. Didesch, K.L. Clark, R.B. Schuessler, and R.J. Damiano, Jr. 2013. Importance of atrial surface area and refractory period in sustaining atrial fibrillation: testing the critical mass hypothesis. J Thorac Cardiovasc Surg 146:593-598. 141.Sahoo, N., T. Hoshi, and S.H. Heinemann. 2014. Oxidative modulation of voltage-gated potassium channels. Antioxid Redox Signal 21:933-952. 142.Gomez, R., R. Caballero, A. Barana, I. Amoros, E. Calvo, J.A. Lopez, H. Klein, M. Vaquero, et al. 2009. Nitric oxide increases cardiac IK1 by nitrosylation of cysteine 76 of Kir2.1 channels. Circulation research 105:383- 392. 143.Aon, M.A., S. Cortassa, E. Marban, and B. O'Rourke. 2003. Synchronized whole cell oscillations in mitochondrial metabolism triggered by a local release of reactive oxygen species in cardiac myocytes. The Journal of biological chemistry 278:44735-44744. 144.Zhou, L., S. Cortassa, A.C. Wei, M.A. Aon, R.L. Winslow, and B. O'Rourke. 2009. Modeling cardiac action potential shortening driven by oxidative stress-induced mitochondrial oscillations in guinea pig cardiomyocytes. Biophys J 97:1843-1852. 145.Lederer, W.J., C.G. Nichols, and G.L. Smith. 1989. The mechanism of early contractile failure of isolated rat ventricular myocytes subjected to complete metabolic inhibition. The Journal of physiology 413:329-349. 146.Nichols, C.G. and W.J. Lederer. 1990. The regulation of ATP-sensitive K+ channel activity in intact and permeabilized rat ventricular myocytes. The Journal of physiology 423:91-110. 147.Nichols, C.G., C. Ripoll, and W.J. Lederer. 1991. ATP-sensitive potassium channel modulation of the guinea pig ventricular action potential and contraction. Circulation research 68:280-287. 148.Weiss, J.N., N. Venkatesh, and S.T. Lamp. 1992. ATP-sensitive K+ channels and cellular K+ loss in hypoxic and ischaemic mammalian ventricle. The Journal of physiology 447:649-673. 159 149.Venkatesh, N., J.S. Stuart, S.T. Lamp, L.D. Alexander, and J.N. Weiss. 1992. Activation of ATP-sensitive K+ channels by cromakalim. Effects on cellular K+ loss and cardiac function in ischemic and reperfused mammalian ventricle. Circulation research 71:1324-1333. 150.Shivkumar, K., N.A. Deutsch, S.T. Lamp, K. Khuu, J.I. Goldhaber, and J.N. Weiss. 1997. Mechanism of hypoxic K loss in rabbit ventricle. The Journal of clinical investigation 100:1782-1788. 151.Chance, B., H. Sies, and A. Boveris. 1979. Hydroperoxide metabolism in mammalian organs. Physiol Rev 59:527-605. 152.Picht, E., A.V. Zima, L.A. Blatter, and D.M. Bers. 2007. SparkMaster: automated calcium spark analysis with ImageJ. Am J Physiol Cell Physiol 293:C1073-1081. 153.Grandi, E., F.S. Pasqualini, and D.M. Bers. 2010. A novel computational model of the human ventricular action potential and Ca transient. Journal of molecular and cellular cardiology 48:112-121. 154.Ferrero, J.M., Jr., J. Saiz, J.M. Ferrero, and N.V. Thakor. 1996. Simulation of action potentials from metabolically impaired cardiac myocytes. Role of ATP-sensitive K+ current. Circulation research 79:208-221. 155.Boknik, P., C. Unkel, U. Kirchhefer, U. Kleideiter, O. Klein-Wiele, J. Knapp, B. Linck, H. Luss, et al. 1999. Regional expression of phospholamban in the human heart. Cardiovascular research 43:67-76. 156.Hollingworth, S., J. Peet, W.K. Chandler, and S.M. Baylor. 2001. Calcium sparks in intact skeletal muscle fibers of the frog. J Gen Physiol 118:653-678. 157.Terentyev, D., I. Gyorke, A.E. Belevych, R. Terentyeva, A. Sridhar, Y. Nishijima, E.C. de Blanco, S. Khanna, et al. 2008. Redox modification of ryanodine receptors contributes to sarcoplasmic reticulum Ca2+ leak in chronic heart failure. Circulation research 103:1466-1472. 158.Hidalgo, C., P. Aracena, G. Sanchez, and P. Donoso. 2002. Redox regulation of calcium release in skeletal and cardiac muscle. Biol Res 35:183- 193. 159.Sood, S., M.G. Chelu, R.J. van Oort, D. Skapura, M. Santonastasi, D. Dobrev, and X.H. Wehrens. 2008. Intracellular calcium leak due to FKBP12.6 deficiency in mice facilitates the inducibility of atrial fibrillation. Heart Rhythm 5:1047-1054. 160 160.Wakili, R., N. Voigt, S. Kaab, D. Dobrev, and S. Nattel. 2011. Recent advances in the molecular pathophysiology of atrial fibrillation. The Journal of clinical investigation 121:2955-2968. 161.Montaigne, D., X. Marechal, P. Lefebvre, T. Modine, G. Fayad, H. Dehondt, C. Hurt, A. Coisne, et al. 2013. Mitochondrial dysfunction as an arrhythmogenic substrate: a translational proof-of-concept study in patients with metabolic syndrome in whom post-operative atrial fibrillation develops. Journal of the American College of Cardiology 62:1466-1473. 162.Thomas, R.C. 1978. Comparison of the mechanisms controlling intracellular pH and sodium in snail neurones. Respir Physiol 33:63-73. 163.Eisner, D.A., W.J. Lederer, and R.D. Vaughan-Jones. 1983. The control of tonic tension by membrane potential and intracellular sodium activity in the sheep cardiac Purkinje fibre. The Journal of physiology 335:723-743. 164.Minta, A. and R.Y. Tsien. 1989. Fluorescent indicators for cytosolic sodium. The Journal of biological chemistry 264:19449-19457. 165.Harootunian, A.T., J.P. Kao, B.K. Eckert, and R.Y. Tsien. 1989. Fluorescence ratio imaging of cytosolic free Na+ in individual fibroblasts and lymphocytes. The Journal of biological chemistry 264:19458-19467. 166.Sossalla, S., B. Kallmeyer, S. Wagner, M. Mazur, U. Maurer, K. Toischer, J.D. Schmitto, R. Seipelt, et al. 2010. Altered Na(+) currents in atrial fibrillation effects of ranolazine on arrhythmias and contractility in human atrial myocardium. Journal of the American College of Cardiology 55:2330- 2342. 167.Maier, L.S. and S. Sossalla. 2013. The late Na current as a therapeutic target: where are we? Journal of molecular and cellular cardiology 61:44-50. 168.Despa, S., M.A. Islam, S.M. Pogwizd, and D.M. Bers. 2002. Intracellular [Na+] and Na+ pump rate in rat and rabbit ventricular myocytes. The Journal of physiology 539:133-143. 169.Smith, N.P. and E.J. Crampin. 2004. Development of models of active ion transport for whole-cell modelling: cardiac sodium-potassium pump as a case study. Progress in biophysics and molecular biology 85:387-405. 170.Despa, S., A.L. Tucker, and D.M. Bers. 2008. Phospholemman- mediated activation of Na/K-ATPase limits [Na]i and inotropic state during beta-adrenergic stimulation in mouse ventricular myocytes. Circulation 117:1849-1855. 161 171.Correll, R.N., P. Eder, A.R. Burr, S. Despa, J. Davis, D.M. Bers, and J.D. Molkentin. 2014. Overexpression of the Na+/K+ ATPase alpha2 but not alpha1 isoform attenuates pathological cardiac hypertrophy and remodeling. Circulation research 114:249-256. 172.Brand, S., T. Seeger, and C. Alzheimer. 2000. Enhancement of persistent Na+ current by sea anemone toxin (ATX II) exerts dual action on hippocampal excitability. Eur J Neurosci 12:2387-2396. 173.Workman, A.J., K.A. Kane, and A.C. Rankin. 2003. Characterisation of the Na, K pump current in atrial cells from patients with and without chronic atrial fibrillation. Cardiovascular research 59:593-602. 174.Carmeliet, E. 2003. Does the Na+,K+ pump current undergo remodeling in atrial fibrillation? Cardiovascular research 59:536-537. 162 Curriculum Vitae Libet Garber, Ph.D. Biomedical Engineer Scientist Libet.Garber@fda.hhs.gov EDUCATION AND TRAINING University of Maryland, College Park, Ph.D. in Bioengineering Research Area: Cardiac Electrophysiology, Calcium Signaling and Computational Modeling Principal Investigator: Dr. W. Jonathan Lederer GPA: 4.0 Johns Hopkins University, Baltimore, MD M.S.E in Biomedical Engineering, October 2006 Research Area: Cardiac Electrophysiology Thesis Committee: Dr. Leslie Tung, Dr. Alexander Spector, Dr. Henry Colecraft B.S. in Biomedical Engineering, May 2005 Concentration: Electrical Engineering GPA: 3.7 Johns Hopkins University Applied Physics Laboratory, Laurel, MD Postgraduate Courses in Electrical Engineering, 2008 Mathematical Methods in Physics and Engineering Principles of Modern Radars RELEVANT CLASSES Cardiac Electrophysiology, General Physiology, Computational Modeling, Robot Sensors and Actuators; Biomedical Instrumentation; Signals and Systems; Control Systems; Medical Imaging Systems; Microfabrication Laboratory; Electrical and Computer Laboratory, Chaos Theory Dynamical System, Transport, Bio-molecular Rate Process RESEARCH EXPERIENCE 9/2011-Present PhD student University of Maryland, College Park, MD § Studied Intracellular Sodium Regulation and Calcium Signaling in an animal model of Atrial Fibrillation using confocal microscopy. § Measured Calcium buffering effects in atrial myocytes through whole cell patch clamping and mathematical modeling. § Simulated the impact that the L-type calcium channel gating has on the Ventricular Tachycardia to Ventricular Fibrillation transition. 163 § Developed Homology models of the Voltage gated L-type Calcium channel and designed protocols for the functional verification and validating testing. § Mentored first year Ph.D. students by providing project goals and guidance on how to complete each research aim. § Collaborate with faculty to advance computational modeling projects that will enhance FDA ability to make scientifically based regulatory decisions in the future. § Taught Regulatory Science graduate Courses on translating research findings into Biomedical Products. 2/03-10/06 Research Assistant, Cardiac Bioelectric Laboratory, Johns Hopkins University, Baltimore, MD § Created scientific protocols to investigate electro-cardiac mini-reentries. § Designed, built and validated a microfluidic chamber able to deliver two different solutions to a 2D monolayer of cardiac cells during optical mapping. § Performed data processing and data analysis using LabView and MATLAB software. § Built a 2.5 cm diameter LED array with a filter that emits 530 ± 25 nm wavelength light with high accuracy. 1/00-6/01 Research Assistant, Environmental Engineering Laboratory, University of Miami, Miami, FL § Developed tests, analyzed data and presented results looking at bacterial levels of lake waters in Miami. § Worked to find a staining solution to test for arsenic in treated wood and protect public health. WORK EXPERIENCE 8/09-present Scientific Reviewer Food and Drug Administration, Office of Device Evaluation § Lead reviews of complex pre-IDE, IDEs, PMAs and 510(k) of multiple types of devices including ablation catheters, bedside monitors, electrocardiograms (ECG), mapping systems, trans-myocardial revascularization and optical coherence tomography. § Made technical recommendations to the Center for Medicare and Medicaid Services (CMS) to determine the appropriate use of Fourier transform for ECG devices and their reimbursement qualifications. § Served as acting branch chief on several occasions, this included assigning work, reviewing work, leading discussions, facilitating decisions, coordinating with other offices and reporting to senior division management. § Implemented a cardiac computational modeling project with the Office of Science and Engineering Laboratories (OSEL) to determine the factors affecting ventricular tachycardia to fibrillation (VT/VF) transitions and to inform our regulatory scientific decisions on the adequate timing of defibrillator shocks to treat sudden cardiac death. 3/07-6/08 Electrical Engineer Johns Hopkins University Applied Physics Laboratory, Laurel, MD 164 § Developed and implemented software verification and validation testing protocols to ensure the accuracy of the radar technology used in ballistic missile defense. § Conducted data analysis and prepared summary reports to integrate the results and interpretation of trajectory data collected from ballistic missiles exercises. § Presented data reports to Navy officials and provided expert advice related to the adequacy of software changes in radar technology. COMMUNITY SERVICE § Key worker for the federal campaign to raise money for charities. § Presenter in the FDA minority student science conferences. § Spanish translator for inner city schools. § Leader of the hospitality and prayer ministry in a multiracial church community. JOB AWARDS § ODE Surgical Ablation Review Award (2017) § ODE Cardiac Ablation Summit Award (2014) § CDRH Special Recognition -Cardiac Electrophysiology and Monitoring Devices Branch (2012) § Certificate of Appreciation (2010) § HHS Commissioner Letter of Appreciation (2010) PARTICIPATION IN MEETINGS CONFERENCE AND WORKSHOPS § Heart Rhythm Society (2013, 2014, 2015, 2017) § Biophysical Society Meeting (2014, 2015, 2017) § NIH Multi-scale Modeling Consortium Meeting (2012) § National Academy of Science Verification, Validation and Uncertainty (2012) § Center of Excellence in Regulatory Science meeting (2012) § Atrial Fibrillation Symposium (2010, 2011, 2012) § Transcatheter Cardiovascular Therapeutics (2010) § Computational Modeling Workshop (2010) MEMBERSHIPS § Heart Rhythm Society § Biophysical Society § American Physical Society § Tau Beta Pi Engineering Honor Society § Biomedical Engineering Honor Society § Golden Key Honor Society ORAL PRESENTATIONS § University of Maryland Student Seminar (December, 2017) Role of Intracellular Sodium Signaling in Atrial Myocytes 165 § University of Maryland Muscle Biology Retreat (April, 2017) Arrhythmogenic Na+/Ca2+ Imbalance in Atrial Myocytes § University of Maryland Student Seminar (December, 2015) Relationship between Action Potential and Excitation Contraction Coupling in Atrial Myocytes § University of Maryland Student Seminar (April, 2015) Ca2+ Signaling in Atrial Myocytes: an Experimental and Mathematical Modeling Approach § University of Maryland Pharmacy School (2015) Clinical Trial for Medical Devices § Johns Hopkins University Regulatory Lecture Series (2012, 2013, 2014) The 510(k) Process § University of Maryland Computational Group (2012) Modeling Electrophysiology impact of the L-type Calcium Channel § FDA Science Minority Symposium (2012) Regulation of Medical Devices § Division of Cardiovascular Device Clinical Rounds (2011) Trial Design Options for Confirmatory Clinical Trial Studies § FDA Science Hispanic Symposium (2011) Getting to Market with a Medical Device ABSTRACTS AND PUBLICATIONS § Garber L, Lederer WJ, Greiser M. EMCCD Camera-based ratiometric measurements of intracellular [Na+] in isolated atrial myocytes from mouse hearts. J Vis Exp, in preparation. § Garber L, Greiser M, Humberto J, Hagen B, Lederer WJ. Non- Chemical methodology to stabilize atrial myocyte for biological experiments. Biotechniques, in preparation. § Garber L, Humberto J, Lederer WJ, Greiser M. Stabilizing Na+/Ca2+ Signaling in Atrial Fibrillation. JCI, 2018 in preparation. § Garber L, Lederer WJ, Greiser M. Characterization of Intracellular Sodium Homeostasis in Murine Atrial Myocytes. Biophysical Journal, Vol. 112, Issue 3, p96a, 2017. § Garber L, Greiser M, Willaims GS, Lederer WJ. Buffering Effects on the LCC Current and Spatiotemporal Ca2+ Dynamics. Biophysical Journal, Vol. 108, Issue 2, p105a–106a, 2015. § Garber L, Willaims GS, Lederer WJ. Solving Calcium spatiotemporal Diffusion Using Comsol Multi physics. COMSOL Conference Poster Presentation, October 2014. 166 § Garber L, Williams GS, Gray AR, Lederer WJ. The Role of Structure, Buffering and LCC Gating in EC coupling: A simulation study comparing atrial and ventricular myocytes. Muscle Biology Scientific Retreat Poster Presentation, May 2014. § Garber L, Gray AR. The Role of the L-type Calcium Current (ICaL) in Ventricular Fibrillation: A Simulation Study. Heart Rhythm Society Abstract Poster Presentation, May 2013. § Garber L, Gray AR, Matysiak S. Modeling the structure and function of L-type Calcium channel (Cav1.2) to understand its effect in cardiac propagation. Biophysical Journal, Vol. 104, Issue 2, p460a, 2013. § Gray RA, Garber L, Colasky T. Development of a Whole Heart Electromechanical Model to Simulate Normal and Abnormal Phenomenon for Multi-Center FDA Use. FDA Poster Presentation, Critical Path Initiative 2011. § Lin JW, Garber L, Qi YR, Chang MG, Cysyk J, Tung L. Region of slowed conduction acts as core for spiral wave reentry in cardiac cell monolayers. Am J Physiol Heart Circ Physiol 294: H58-H65, 2008. § Garber L. Dual superfusion chamber led to the discovery of cardiac mini-reentries and s-shaped spiral waves (Master's thesis). Baltimore, MD: The Johns Hopkins University, 2006