Modulation of the Arrhythmia Substrate in Cardiovascular Disease

Home , Cardiac action potential, Rolofylline

DISSERTATION

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

Victor Paul Long III

Graduate Program in Pharmaceutical Sciences

The Ohio State University

2016

Dissertation Committee:

Dr. Cynthia A. Carnes, Advisor

Dr. Sandor Györke

Dr. Kari Hoyt

Dr. Peter J Mohler

Victor Paul Long III

2016

Abstract

Heart failure remains a leading cause of morbidity and mortality in the United

States. Many of the deaths attributed to heart failure are sudden, presumably due to lethal arrhythmia. It is a combination of structural and electrical remodeling within the failing heart that promotes the abnormalities of normal rhythm that lead to arrhythmia. This remodeling can occur in the ventricle, resulting in tachyarrhythmia or the sinus node, where it can cause either brady- or tachyarrhythmias. Potassium currents mediate the repolarization phase of ventricular action potential, as well as the diastolic phase of the sinoatrial node action potential. One of the purposes of the research described in this dissertation is to understand, from the standpoint of cellular electrophysiology, how alterations of potassium currents play a role in heart-failure induced arrhythmia. The second purpose is to determine if management of serum potassium levels by pharmacists is an effective strategy in patients to minimize proarrhythmia risk in patients taking antiarrhythmic medications.

We found that heart failure duration is very important in the progressive reduction of the repolarization reserve of K+ currents in the ventricle. Our results differ from other models, as we were able to identify IKr reduction in chronic heart failure compared to short-term duration heart failure. As a consequence of depleted repolarization reserve, chronic heart failure resulted in a high frequency of early afterdepolarizations (cellular

ii arrhythmia). We also found increased ventricular tissue fibrosis in chronic heart failure, a hallmark of human end stage heart failure which is often absent in short-term pacing models.

Our chronic heart failure model was also used to investigate the role of adenosine- induced sinus node dysfunction in heart failure. Failing sinoatrial node cells had slower intrinsic firing rates versus normal control cells. We were able to demonstrate an increase in the sensitivity of the rate slowing effects of adenosine in failing sinoatrial node cells.

These results are due to an increase in adenosine A1 receptor (A1R) and G-coupled inward rectifier (GIRK) signaling. The negative chronotropic effects of adenosine were abolished through the use of A1R and GIRK inhibitors.

Finally, we tested the performance of an algorithm designed to allow pharmacists to manage patient serum potassium levels in patients taking antiarrhytmic medications.

Due to the high risk of proarrhythmia among this population, we established a conservative range of serum potassium to prevent low potassium from affecting patient therapy. Patients undergoing the protocol reached desired potassium levels more quickly than standard of therapy, and were able to maintain these levels upon a follow up visit occurring months later.

We propose that these studies provide insights into the cellular bases for the development of heart failure-induced arrhythmias, as well as optimizing management of arrhythmia in a clinical setting.

iii

Dedication

This document is dedicated to my loving wife Tammy, and my two greatest

achievements, Isaac and Zander.

Acknowledgments

The dissertation that follows is the combined product of caffeine-driven determination, all-nighters in the lab, and a little bit of luck. None of this would have been possible without the help of colleagues that helped and assisted me during my graduate school experience.

I would like to foremost thank my advisor, Dr. Carnes for taking me into her laboratory in the first place. Her unwavering support kept me moving forward, especially during the frustrating period of starting sinoatrial cell isolation. She seemed to always make time for me if I needed anything at all. She was also very understanding of my unique circumstances (having two kids) and provided unbelievable flexibility when it was needed. I am forever grateful for her guidance and input into my work. Without her,

I certainly would not have made it to this point.

I also thank my committee members, Dr. Sandor Gyorke, Dr. Peter Mohler, and

Dr. Kari Hoyt. In our committee meetings, their insights and suggestions really advanced my work and thinking.

I’d like to thank our collaborators and their laboratory members for helping me and allowing me to work on their projects. These include Dr. Vadim Fedorov, Dr. Loren

Wold, Dr. Sandor Gyorke, Dr. Thomas Hund, Dr. Melissa Snider, and Dr. Peter Mohler.

I’d also like to thank all of the present and past members of our laboratory: Jeanne Green,

Jon Vecchiet, Jae Yoo, Stephen Baine, and Pedro-Vargas Pinto. I feel I need to give a special acknowledgment to Ingrid Bonilla, who really demonstrated a lot of patience in teaching me how to patch clamp, troubleshoot equipment, and perform data analysis. In addition, she is a good friend and helped keep the mood light during the many stressful times of this endeavor.

Finally, I’d like to thank my wife for allowing me to jump into all this, especially when we had a comfortable life beforehand. This experience was something I had to do, and somehow she understood that. I acknowledge her sacrifice when I’m away for long hours (sometimes overnight) while she works full-time, goes to graduate school part- time, and raises two wonderful kids. She is the most exceptional person I’ve met in my life and I just want to say thank you for letting me do all this.

Vita

1996...... West Chester East High School

2000...... B.S. Biochemistry, Virginia Tech

2007...... PharmD University of Maryland at

Baltimore

2011 to present ...... Graduate Research Associate,

Pharmaceutical Sciences, The Ohio State

University

Publications

1. Long VP, III, Carnes CA. Treating cocaine cardiotoxicity: Does receptor subtype

matter (editorial)? Trends Cardiovasc.Med. 2015 Jan 12.

2. Smith SA, Sturm AC, Curran J, Kline CF, Little SC, Bonilla IM, Long VP,

Makara M, Polina I, Hughes LD, et al. Dysfunction in the betaII spectrin-

dependent cytoskeleton underlies human arrhythmia. Circulation 2015 Feb

24;131(8):695-708.

3. Long VP, III, Bonilla IM, Vargas-Pinto P, Nishijima Y, Sridhar A, Li C, Mowrey

K, Wright P, Velayutham M, Kumar S, et al. Heart failure duration progressively

vii

modulates the arrhythmia substrate through structural and electrical remodeling.

Life Sci. 2015 Feb 15;123:61-71.

4. Bonilla IM, Long VP, III, Vargas-Pinto P, Wright P, Belevych A, Lou Q, Mowrey

K, Yoo J, Binkley PF, Fedorov VV, et al. Calcium-activated potassium current

modulates ventricular repolarization in chronic heart failure. PLoS.One.

2014;9(10):e108824.

5. Bonilla IM, Vargas-Pinto P, Nishijima Y, Pedraza-Toscano A, Ho HT, Long VP,

III, Belevych AE, Glynn P, Houmsse M, Rhodes T, et al. Ibandronate and

ventricular arrhythmia risk. J.Cardiovasc.Electrophysiol. 2014 Mar;25(3):299-

306.

6. Bonilla IM, Belevych AE, Sridhar A, Nishijima Y, Ho HT, He Q, Kukielka M,

Terentyev D, Terentyeva R, Liu B, Long VP, et al. Endurance exercise training

normalizes repolarization and calcium-handling abnormalities, preventing

ventricular fibrillation in a model of sudden cardiac death. J.Appl.Physiol (1985.)

2012 Dec 1;113(11):1772-83.

Fields of Study

Major Field: Pharmaceutical Sciences

Specialization: Translational Science

viii

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... vii

List of Tables ...... xv

List of Figures ...... xvi

Chapter 1 : Introduction ...... 1

A Brief History of Cardiovascular Physiology and Electrophysiology ...... 1

Early studies of the Heart ...... 1

The Heart as an Electric Organ: The roots of electrophysiology ...... 4

The Cardiac Conduction System ...... 6

The Study of Excitable Cells ...... 8

Cardiac Action Potentials are Orchestrated by Ion Currents ...... 10

The Canine Action Potential...... 10

The Ventricular Action Potential ...... 11

The Ventricular Action Potential and its Relationship to the ECG ...... 13

The SAN Pacemaker Cell has a Unique Morphology ...... 13

Ion Channel Structure and Function ...... 15

Ion Channel Structure of a K+ channel ...... 15

Ion Channel Gating ...... 15

Potassium Channels of the Ventricular Action Potential ...... 18

IK1 is carried through inward rectifying Kir channels ...... 18

Ito is carried through Kv channels and initiates early repolarization ...... 19

The delayed rectifier Kv channels are responsible for the bulk of repolarization ..... 21

The K+ channels of the ventricular action potential provide a repolarization reserve 23

K+ Channels and Ventricular Arrythmogenesis ...... 24

Action potential beat-to-beat variability can be used to predict arrhythmogenesis .. 25

Diastolic Potassium-Conducting Channels of the SAN Action Potential ...... 26

The If channel activates upon hyperpolarization ...... 26

IKAch/IKAdo are carried through ligand gated Kir channels ...... 28

Diastolic SAN K+ Channels and Arrhythmogenesis ...... 29

Chapter 2 : Heart failure duration progressively modulates the arrhythmia substrate through structural and electrical remodeling ...... 41

Introduction ...... 42

Materials and Methods ...... 43

Heart failure canine model ...... 43

Myocyte Isolation ...... 44

Immunoblots ...... 46

Real-time PCR for gene expression...... 46

Electron paramagnetic resonance (EPR) spectroscopy ...... 47

Data Analysis ...... 48

Chemicals ...... 49

Results ...... 49

Chronic tachypacing is accompanied by an increase in left ventricular mass, a

prolonged QTc interval, and interstitial fibrosis ...... 49

Protein and mRNA expression of K Channel Subunits does not correlate with

function ...... 52

EPR spectroscopy ...... 52

Discussion ...... 53

Limitations ...... 57

Conclusion ...... 58

Chapter 3 : Chronic Heart Failure Increases Negative Chronotropic Effects of Adenosine in Sinoatrial Pacemaker Cells via increased A1R-mediated If block and activation of

GIRK-mediated IKAdo ...... 68 xi

Introduction: ...... 69

Materials and Methods: ...... 71

Heart failure canine model ...... 71

Cell Isolation...... 72

Quantification of Cardiac adenosine levels ...... 73

Real-time PCR for gene expression...... 75

Electrophysiological recordings ...... 76

Chemicals ...... 77

Statistical analysis...... 77

Results: ...... 78

Right atrial cardiac adenosine levels are lower in HF and are associated with

upregulation of SAN A1R and ENT mRNA ...... 78

HF at baseline results in a slower intrinsic firing rate of SAN cells and a decrease in

pacemaker current, If ...... 78

HF increases SAN cell sensitivity to the negative chronotropic actions of

adenosine ...... 79

SDD and MDP correlate with adenosine-induced negative chronotopic effects in

CTL and HF SAN cells, respectively ...... 80

Adenosine’s negative chronotropic effects on SAN cells in both control and HF are

mediated by A1R ...... 81 xii

HF augments adenosine-induced reduction of If at physiological voltages ...... 82

HF augments GIRK current, IKAdo response to adenosine ...... 82

GIRK inhibition by, Tertiapin Q: increased in HF SAN cells ...... 83

Discussion: ...... 83

Decreased right atrial cardiac adenosine levels are accompanied by A1R

upregulation ...... 84

SAN HF cells demonstrate reduced intrinsic firing and If...... 86

Upregulation of A1R and GIRK in HF augments IKAdo and adenosine-mediated If

block ...... 86

Clinical Implications...... 88

Limitations ...... 89

Conclusions ...... 89

Chapter 4 : Evaluation of a pharmacist-managed electrolyte protocol to achieve and maintain optimal potassium concentrations in outpatients on antiarrhythmic medications ...... 98

Introduction ...... 99

Patients and Methods ...... 100

Inclusion Criteria ...... 100

Study Design...... 101

Hypokalemia Protocol ...... 102 xiii

Results ...... 103

Discussion ...... 104

Limitations ...... 106

Conclusions ...... 107

Chapter 5 : Summary and Conclusions ...... 111

Reference List ...... 115

xiv

List of Tables

Table 1.1 Species-dependent ion current contributions to ventricular action potential .... 40

Table 2.1 Echocardiographic and Electrocardiogram parameters in 4 M HF animals ..... 59

Table 4.1 Patient demographics and medical history ...... 109

List of Figures

Figure 1.1 The cardiac conduction system...... 32

Figure 1.2 Ionic basis of the ventricular action potential ...... 33

Figure 1.3 The ECG and the ventricular action potential ...... 34

Figure 1.4 Ionic basis for the sinoatrial node action potential ...... 35

Figure 1.5 Structure and Proposed Gating Mechanism of K+ Channels ...... 36

Figure 1.6 Inactivation Gating of Kv Channels ...... 37

Figure 1.7 GIRK and HCN signaling in SAN cells ...... 38

Figure 1.8 Mechanism for a re-entrant arrhythmia ...... 39

Figure 2.1 Interstital Fibrosis is Increased in Chronic HF ...... 60

Figure 2.2 Progressive action potential prolongation and cellular arrhythmias during heart failure ...... 61

Figure 2.3HF decreases Ito ...... 62

Figure 2.4 Ito kinetics are altered in chronic HF ...... 63

Figure 2.5 Outward IK1, but not inward IK1 is reduced in chronic but not short-term HF 64

Figure 2.6 IKs and IKr are reduced in chronic HF ...... 65

Figure 2.7 Protein expression and mRNA levels of K+ channel subunits ...... 66

Figure 2.8 Heart Failure increases ventricular oxidative stress ...... 67

xvi

Figure 3.1 Decreases in right atrial adenosine levels are associated with upregulation of

A1R and ENT mRNA ...... 90

Figure 3.2 Intrinsic firing and If are decreased in SAN HF cells ...... 91

Figure 3.3 SAN HF cells demonstrate increased sensitivity to negative chronotropic effects of adenosine...... 92

Figure 3.4 SDD and MDP correlate with adenosine-induced intrinsic rate slowing in CTL and HF SAN cells, respectively ...... 93

Figure 3.5 Adenosine's negative chronotropic effects of SAN cells in both CTL and HF are mediated by A1R ...... 94

Figure 3.6 Adenosine dose-dependently decreases If in CTL and HF SAN cells with greater sensitivity of inhibition observed in HF ...... 95

Figure 3.7 Adenosine dose-dependently increases IKAdo in CTL and HF SAN cells with greater sensitivity observed in HF ...... 96

Figure 3.8 GIRK blockade prevents adenosine-induced rate slowing ...... 97

Figure 4.1Pharmacist-managed potassium electrolyte algorithm ...... 108

Figure 4.2 Flow chart of study ...... 110

xvii

Chapter 1 : Introduction

A Brief History of Cardiovascular Physiology and Electrophysiology

Early studies of the Heart

Our recognition of the importance of the heartbeat in sustaining life can be traced to prehistoric cave paintings from as far back as 18,000 years ago. In 1908, after exploring El Pindal cave in Spain, French archeologist Henri Breuil reproduced an outline of a wooly mammoth with what appears to be an anatomically positioned, red colored heart-shape within.1 The original drawing has been generally accepted as a training guide for prehistoric hunters on where to aim weapons to ensure a kill. The discovery of the Rosetta Stone in 1799 and subsequent translation of the ancient text, the

Ebers Papyrus (1534 BC), suggests that it was the Egyptians had a profound interest in cardiovascular physiology2. The Eber Papyrus describes vessels that extend from the heart and “speak,” or what we now refer to as having a pulse.3 Furthermore, the papyrus indicated that the ancient Egyptians could link abnormal or weak peripheral pulse measurement with advancing disease. The Egyptian’s believed the vessels connected to the heart were called metw, and carried not only blood, but semen, feces, and evil spirits.

Hundreds of years later, early Chinese medicine incorporated pulse measurement for the 1 diagnosis and treatment of disease. The Nei ching (300-500 BC) documents six organ- specific pulses in the wrist and over 200 different pulse variations.4 The pulse, to the early Chinese physicians was, was the body playing a stringed instrument and its measurement could reflect disharmony (bad health).5 The Nei ching describes blood as

“under control of the heart” and it flows “continuously in a circle and never stops,” but stops short of providing any rationale or evidence for this primitive description of circulation.4

The emergence of Greek philosophy and their introduction of scientific dissection were monumental in advancing the understanding of the heart as a part of a larger circulatory system. Alceomon of Cronos (500 BC) made the first general distinction between arteries and veins. He concluded that veins withdrew pneuma, or spirit of air, from the brain during sleep but was brought back via arteries to wake the body up6. The

Greek philosopher, Aristostle (350 BC), through dissection of over 500 types of animals7, erroneously described the heart as having three chambers, or ventricles, and as the source of all nerves and blood vessels in the body8. The purpose of other organs, according to

Aristotle, was to cool the heart because it was the seat of intelligence, motion, sensation and the center of vitality in the body.9 Human cadaver dissection at the Alexandria

School (300 BC) by the Greek physician, Herophilus and his contemporary, Erasistratus led to a more accurate portrait of the heart. Herophilus confirmed Alceomon’s previous observation on the difference between arteries and the veins, noting that the former had thicker walls.9 He is credited as the first to recognize the atria as a distinct chamber.10

Herophilus associated the heartbeat with pulse of arteries by timing it with a water clock

2 and described pulse variations in terms of volume, rate, and rhythm.11 Furthermore, he distanced himself from Aristotle in his assertion that nerves did not originate in the heart, but rather the brain.10 Like Herophilus and Aristotle, Erasistratus believed the arteries transported only pneuma through passive pulsations with each heartbeat. Erasistratus described cardiac valves and their function, and associated arteries with the left ventricle and veins with the right ventricle. He postulated that air enters the lungs and passes through the pulmonary vein to the left ventricle, where it is changed into “vital spirit.” 12

The vital spirit was then transported throughout the body by the arteries.

Claudius Galen (160 AD) drew greatly upon the work of Herophilus and

Erasistratus, but is often credited with their achievements. With human dissection no longer acceptable at the time, Galen based his cardiovascular studies on the Barbary ape, a species indigenous to his home in Pergamon.12 In his landmark writing, On the

Usefulness of Body Parts, Galen described the heart as a strong muscle that works harder than any other organ.13 He described the heart as having a “coat,” or what we refer to now as the pericardium. He proved arteries contained blood, not just pneuma as described by Erasistratus. To Galen, the heart’s purpose was still primarily to transport pneuma and the movement of vital spirit was almost identical to that of Erasistratus. Unlike

Erasisitratus, Galen believed blood was generated by ingested food and drink and became

“natural spirit” before moving to the left ventricle to become “vital spirit.”12 Galen believed nerves connected the heart and blood vessels, such that when the heart contracted, vessels dilated to allow inward flow. Galen’s theories endured until the

Middle Ages, owing largely to them being directly in line with the Catholic Church. It

3 wasn’t until William Harvey published the Exercitatio anatomica de motu cordis et sanguinis in animaibus in 1628 that Galen’s views were disproved. Through experiments in animals, along with the use of human cadavers, Harvey deduced that the enormous amount of blood pumped by the ventricle could not have come from food and drink alone as Galen theorized.12 He concluded that blood pumped through the aorta had to come from veins, and thus blood movement was both circular and contained. He described the heart as having four chambers where the ventricles contracted after the atria. He concluded that blood pumped from the right ventricle passes through the lungs, where it is mixed with air, and then returns through the left atria where it is pumped throughout the body as pneuma. Harvey’s discovery of the circulation of blood was critical in shaping both cardiovascular physiology and modern medicine.

The Heart as an Electric Organ: The roots of electrophysiology

While the foundations of cardiovascular physiology were in place at the end of the 17th century, it wasn’t until a century later that the view of the heart as an electric organ began to form. In 1791, Luigi Galvani reported that cutting the legs of a frog anchored to a brass hook with a metal scalpel caused them to twitch.14 German scientists

Rudolf Albert von Kolliker and Heinrich Muller built upon Galvani’s discovery of bioelectricity to conduct a seminal experiment in cardiovascular electrophysiology. In

1856, they demonstrated that placing the ventricle of a beating frog’s heart on the exposed sciatic nerve of a dismembered frog’s leg also caused it twitch.15 Soon afterwards, investigators began working on ways to record the heart’s electrical activity.

Augustus Waller, a British physiologist, published the electrocardiogram (ECG) from the body surface of a human being in 1887. Using a Lippman capillary electrometer attached through electrodes to his subject, Waller was able to project the movement of the mercury meniscus produced by the heart’s electrical contractions onto photosensitized

16 paper. He identified “V1” and “V2” deflections that corresponded to ventricular depolarization (contraction) and repolarization (relaxation), respectively. Because of the poor quality of the recording, Waller himself thought the ECG would be of little use to physicians.17

It was the Dutch scientist William Einthoven in 1913 who improved upon

Waller’s ECG by using a string galvanometer instead of the Lippman electrometer.16 The string galvanometer was a silver-quartz wire that passed between two electromagnets.

Leads were connected from the patient’s extremities (placed in electrolyte solution) to the wire. The electric current through the wire produced movement and was projected onto scrolling photosensitive paper marked with graduations allowing Einthoven to calculate the precise timing of cardiac events. Einthoven employed three strings to create a three lead system (Einthoven’s triangle) that would generate positive (upward) deflections on the ECG. Both atrial and ventricle depolarization moves downward and to the left, and he found if he placed a positive electrode on the left foot, he could get positive deflections relative to electrodes from the left arm (lead III) and the right arm (lead II). A

5 positive electrode on the right arm relative to the left arm (lead I) also produced a positive deflection. From recordings from each lead of this triangle, Einthoven named the deflections (waves) that are still used clinically today. Einthoven’s triangle of bipolar electrodes served as the basis for the inclusion of unipolar chest wall leads by Frank

Wilson in 1933 augmented limb leads by Emanuel Goldberger in 1942 to create the 12- lead electrocardiogram used today.

The Cardiac Conduction System

While the ECG was being refined, several key developments in basic research suggested that the heart’s electrical system was complex. Between 1879 and 1883, British physiologist Walter Gaskell studied the initiation of the heartbeat and its conduction through cardiac tissue. By observing a slowly beating turtle heart, Gaskell demonstrated that contraction moved as a wave that began at the sinus venous, through the atrium to” ventricle-auricle” (atrium near AV groove), and ended in the ventricle.18 Removal of sections of atrial tissues at the AV groove imparted delays of ventricular contraction and complete transection caused ventricle to stop contracting before resuming at its own, independent, slow rate. From these studies, Gaskell concluded that the AV groove slows impulse conduction from the atrium to the ventricle. In his 1893 work, Embryonic

Cardiac Activity and Its Significance for Adult Heart Movement Theory, Wilheim His Jr built upon the work of Gaskell.19 He discovered a muscular bundle within the connective tissue of the AV groove (later named the His Bundle) that forked into right and left limbs along the upper border of the ventricular septum. His never severed his bundle in a 6 beating heart, so he wasn’t convinced it was responsible for transmitting impulses from the atria to the ventricle. A more complete picture of the cardiac conduction system was proposed by the pathologist Sunoa Tawara in 1906. In his histological examination of over 150 hearts, Tawara was able to find that His Bundle actually originated from a node at the base of the atrial septum (AV node). He followed the His Bundle in the downward direction and noticed that limbs extending from the fork ended in a network of fibers around the ventricular subendocardium, discovered by Jan Evangelista Purkinje in 1839.

Tawara had, in effect, discovered a conduction pathway that began at the AV node, travelled through the His Bundle, and forked into branches that ended in what are now known as Purkinje fibers. The last major piece of the cardiac conduction puzzle was placed by the collaboration of Arthur Keith and Martin Flack in 1906, who were deeply involved in studying the recently discovered conduction system described by Tawara.

Flack, under the mentorship of Keith, described a fibrous structure similar to the AV node, but superior to it, located at the junction of the superior vena cava and right atrium.

This fibrous matrix was also innervated and located at the sinus venous, where Gaskell had previously noted contraction to begin. In their 1907 publication, The form and nature of the muscular connections between the primary divisions of the vertebrate heart, Keith and Flack declared this region was the origin of the heart’s rhythm and were given credit as the founders of the “sinoauricular node” (later termed SA node).20 With the SA node now identified as the origin of the heartbeat, the modern view of the cardiac conduction pathway was in place at the beginning of the 20th century (FIGURE 1.1).

The Study of Excitable Cells

While the anatomy of the cardiac conduction system was well established in the early 1900’s, the question remained of what mechanisms allowed for impulse propagation through the myocardium. The collaboration of biophysicists Alan Hodgkin and Andrew Huxley, in the late 1940’s, led to our current understanding of the initiation and propagation of action potentials. Utilizing a fine capillary electrode filled with sea water impaled into a squid giant axon, along with a reference electrode in the outside bath

(both attached to a voltage amplifier and oscilloscope), Hodgkin and Huxley stimulated the nerve with current and published the first action potential.21 By lowering the sodium

(Na+) concentration in the external bathing solution, they discovered they could reduce the amplitude and slope of the rising phase of the action potential.22 Experiments by other groups had demonstrated the squid axon’s membrane potential dependence on

+ extracellular potassium concentration [K ]o. Hodgkin and Huxley next set out to not only measure the currents elicited at a particular voltage, but which ions carry them.

Hodgkin and Huxley’s use of the “voltage clamp” technique allowed for the recording of current while holding the potential of the axon constant (Ohm’s law).

Membrane potential was recorded similarly to their action potential setup, however the voltage amplifier was also inserted into a “feedback” amplifier. Connected to the feedback amplifier was a device with a command potential, used by the experimenter to set a desired membrane potential. The feedback amplifier amplified the difference between recorded membrane and desired voltage, and through a second intracellular electrode impaled in the axon, injected the difference current back through the

8 membrane. This current was measured relative to the bath ground electrode where it was recorded as a representation of currents flowing across the membrane. By measuring

+ currents at difference membrane voltage steps, and altering [Na o] in the external bath,

23 Hodgkin and Huxley identified two major currents, INa and IK. Furthermore, it was noticed that manipulation of voltage steps could affect the conduction of both currents, a term later called activation and inactivation. Hodgkin and Huxley deduced that these currents were controlled by a gating process, and incorporated this into a mathematical model of the action potential.24 Their work essentially described the behavior of ion channels before the channels were identified, as well as their contribution to action potentials.

Arriving at Cambridge University in 1948, to work with Alan Hodgkin, was a

Swiss physiologist named Silvio Weidmann. Using similar techniques as his mentor,

Weidmann recorded the first transmembrane action potential of a false tendon from the right ventricular endocardium of a dog heart.25 After this breakthrough, Hodgkin famously told Weidmann, “You can now rediscover the whole field of cardiac electrophysiology.”26. Indeed, Weidmann continued his studies and is credited for the demonstrating that the sodium current was responsible for excitation and conduction, and could be blocked by local anesthetics27,28. He noticed that during the plateau of the action potential, before repolarization began, the resistance of membrane was increased29

– a finding that would late be discovered to have major implications in the development of arrhythmias. Finally, tracking radioactive potassium flux across the intercalated disks

9 of sheep ventricle fibers, he recognized that cardiac cells communicate with one another intercellularly30 far in advance of knowledge of the existence of gap junctions.

Weidmann’s contributions spanned the 1960’s, and during that time other groups recorded the unique action potentials of various components of the cardiac conduction system such as the SA node31, atria32, AV node33, and His Bundle34. However, it was the development of the patch clamp in 1972 by Erwin Neher and Bert Sakmann, which brought cardiac electrophysiology into a new era. By finishing a clean glass pipette with a tip diameter of 1-3 μm, it was possible to make a high resistance seal (gigaseal) with a

“patch” of an individual cell membrane. This technique made it possible to obtain low- noise recordings of single channel currents, as well as single cell action potential recordings. Furthermore, manipulation of this technique, such as the inside-out patch, allowed researchers for the first time to manipulate the intracellular components of a cell to identify their effects on ion current. The patch clamp became an invaluable tool for the study how cardiac ion channels are modulated as well as their effects on the action potential. The remainder of this chapter will focus on present day knowledge of the ion channels that contribute to the cardiac action potential.

Cardiac Action Potentials are Orchestrated by Ion Currents

The Canine Action Potential

Cardiac action potentials vary with regards to morphology across species as well as across location in the myocardium. Since they are the primary focus of this dissertation, only the canine ventricle cardiomyocyte and sinoatrial node cell will be 10 discussed. The canine action potential was chosen for our studies due to the similarities between resting heart rate and cardiovascular ion current complement between dogs and humans. As experimentation with other species are critical for the advancement of science and are also referred to frequently in this dissertation, TABLE 1.1 defines the major ion currents in mouse and guinea pig in the ventricular action potential.

The Ventricular Action Potential

The resting membrane potential of a ventricular cardiomyocyte is approximately

-85 mV. This is very similar to EK (Potassium equilibrium potential) based on normal intracellular and extracellular K+ concentrations. Indeed, at rest or Phase 4, an inward

+ rectifying current IK1, is actively conducting K into the cell. Ventricular cells are primarily connected through intercalated disks at their long axes, which contain gap junctions that allow electrical impulse propagation from cell-to-cell. When a sufficient amplitude impulse reaches a ventricular cardiomyocyte, sodium channels are activated and a rapid influx of sodium (INa) that depolarizes the cell (Phase 0), activating voltage-

2+ , gated L-type Ca (ICa,L) transient outward current (Ito), and delayed rectifying currents

(IKr and IKs) in the process. At Phase, 1, about 90% of the sodium channels rapidly inactivate and Ito contributes to early repolarization before quickly inactivating.

35 Interestingly, there is an Ito gradient across the ventricular myocardium, with cells in the middle of the myocardium (mid-myocardium) demonstrating a prominent notch prior to the Phase 2 plateau, attributed to greater transient outward current36.

The depth of the Phase 1 notch and the phase 2 plateau are critical modulators of calcium handling and subsequent contraction of the ventricle.37 At the end of Phase 1, the membrane bound sodium-calcium exchange current (NCX) acts in reverse mode,

+ 2+ removing 3 Na from the cell in exchange for 1 Ca . Along with the influx of ICa,L and the remaining activated “late sodium” current, this generates a net depolarizing force.

This force is balanced by repolarizing forces from outward IKr and IKs currents, resulting in very little movement of ions across the membrane, during this high resistance part of the action potential. The influx of Ca2+ binds to ryanodine receptors on the sarcoplasmic reticulum (SR) and induces calcium-induced calcium release. Here, the concentration of

Ca2+ in the cytosol increases from 500 nM to as high as 1 uM. The Ca2+ increase allows for binding to actin- troponin c sites and subsequent coupling to the S1 heads of myosin filaments. Cardiac contraction is then initiated with binding of ATP to the myosin head via a “power stroke” where myosin filaments slide between actin filaments. and initiates cardiac contraction. Towards the end of Phase 2, the sarco-endoplasmic reticulum Ca2+-

ATPase (SERCA pump) sequesters cytosolic Ca2+, removing the ion from actin-troponin c sites and ending contraction. In addition, the NCX exchanger works in forward mode and also helps to remove the systolic calcium from the cytosol. Phase 3 repolarization is characterized by the increase in activity of IKs, and outward IK1 as well as the inactivation of late sodium current. This repolarization returns the cardiomyocyte to Phase 4, or resting membrane potential. An illustration of the ventricular action potential is illustrated in Figure 1.2.

The Ventricular Action Potential and its Relationship to the ECG

Prior to the patch clamp, researchers used microelectrode recordings to record action potentials of the heart. Simultaneous recordings of these transmembrane action potentials and surface ECGs demonstrate a distinct relationship between the phases of the ventricular action potential and the segments on the ECG (Figure 1.3).38 The QRS interval on the ECG corresponds to Phase 0 depolarization of all ventricular cardiomyocytes. The ST segment corresponds to the flat, Phase 2 plateau and early phase of repolarization. The QT interval corresponds to ventricular repolarization (Phase 2 and

3) and is of clinical importance in determining risk for the development of proarrhythmia39. The ability to patch clamp multiple cells isolated from a heart therefore has particular translational value when trying to demonstrate either disease or drug effects on the electrophysiology of the myocardium.

The SAN Pacemaker Cell has a Unique Morphology

As opposed to the ventricular cardiomyocyte, there is very little information regarding the ion currents that contribute to SAN cell action potential of higher mammals. Instead, most of the studies are from mouse or rabbit models, and the existence of several currents has yet to be functionally demonstrated in humans or large mammals. The resting membrane potential of SAN cells is around -60 mV. This depolarized potential is best explained by the observation that SAN cells lack prominent

40 IK1 . Instead, the cells exhibit a diastolic hyperpolarization-activated current, If, that is thought to be primarily responsible for the spontaneous Phase 4 depolarization (Figure 13

1.4). There are two other currents, ICa,T and ICa,L, that are also activated during this diastolic window, and may contribute to Phase 4 depolarization. Unfortunately, due to significant overlap of voltage activation between the two currents and the lack of pharmacologically selective agents, it is difficult to determine the specific roles these currents contribute in native SAN cells.

Although not formally a phase of the action potential, the nonlinear region between Phase 4 and Phase 0 is heavily debated as the primary mechanism for cardiac pacemaking41. Briefly, spontaneous low-voltage activated Ca2+ release from the SR have been observed in SAN cells42. The rise in cytoplasmic Ca2+ forces NCX to exchange

Ca2+ for Na+, which drives the voltage up to activate the upstroke (Phase 0) of SAN cells, generated by inward ICa,L. This is often referred to as the “calcium clock” mechanism. It remains controversial as voltage clamp studies have reported that depolarization is required for these local Ca2+ releases (LCR’s) to occur, and holding the membrane potential at -70 mV abolishes the spontaneous release43. Phase 0 is followed by a Phase 2 plateau phase where inward L-type calcium and the NCX exchange activity is balanced by repolarizing delayed recitifer K+ currents. As L-type calcium inactivates and and the intracellular calcium content wanes, NCX activity is reduced and the repolarization phase

(Phase 3) begins to dominate with outward K+ driven out of the cell. In general, Phase 2 and 3 repolarization can be carried through IKur, IKs, or IKr, the contribution of which appears to be species dependent.

Ion Channel Structure and Function

Ion Channel Structure of a K+ channel

While Hodgkin and Huxley accounted for voltage-induced gating mechanisms of

INa and IK, they did not know that these currents were passed through specialized pore- forming proteins spanning the cell membrane, or what we call ion channels. Today, we know that there are at least 300 genes that code for subunits of voltage-gated ion channels44. The first crystal structure of a bacteria KcsA (K+) (Streptomyces lividans)

Channel was reported in 199845 and revealed four identical transmembrane subunits

(tetramers) with an aqueous pore in the middle (Figure 1.5, PANEL A). For KcSA channels, negatively charged amino acids line the intracellular and extracellular entrances to the channel, ensuring passage of only cations through the channel. The selectivity filter is the narrowest linkage of the vestibule to the outside of cell, and is lined with polar carbonyl groups pointing inward, ensuring that only appropriately size K+ ions can pass through (as opposed to Ca2+, or other ions). This protein sequence that makes up the selectivity filter contains a G-Y-G sequence that is conserved among all voltage gated K+ channels46. Finally, the vestibule cavity opens up into cytoplasm of the cell.

Ion Channel Gating

It is the passage of ions through the ion channel (across the sarcolemma) that generates a current. Since diffusion is the ultimate driving force for ion movement, one would think the extracellular and intracellular concentrations of passing ions would

15 predict the magnitude and direction of current. However, because ions are “charged,” the difference between the intracellular and extracellular charge (potential, or voltage difference) must also be considered. Therefore, a net electro-chemical gradient exists between the inside and outside of an excitable cell that determines movement of ions through ion channels. The Nernst equation describes the calculation for equilibrium voltage of a single ion (where no current flows in or out of the cell) based on its

47 intracellular and extracellular concentration. Using the Nernst equation , EK (equilibrium

+ + + potential for K ) is -84 mV, assuming an [K i] of 140 mM and [K o] of 5 mM. This is strikingly similar to the resting membrane potential of a ventricular cardiomyocyte.

Indeed, at rest all excitable cells have negative resting potential, because K+ channels are open to a much greater extent than the other two principal ions in the body, Na+ and

Ca2+.48 This illustrates an important concept that ion channels are not always open, but rather exist in one of three states: closed, open, or inactivated. The closed conformation is

+ thought to exist in voltage gated K channels (Kv) similar to the bacterial KcsA channel, where the inner and outer transmembrane helices block entrance to the cytosolic pore49

Ion channels can be activated by voltage or ligands. As an example, Kv channels have each of their four subunits further divided into 6 α-helices (S1-S6) that span across the membrane. These alpha subunits are considered to be the “pore forming” subunits of the ion channel. The S4 segment contains positively charged side chains, at every third position, and is thought act as voltage sensor50. The link between the S5 and S6 segments, or p loop, enters the membrane as a hairpin and forms the extracellular pore in Figure 1.5,

PANEL B. The p loop contains the GYG motif that is that thought to confer ion

16 selectivity. The S5 segment is covalently linked to the S4 segment (S4-S5 linker). When the channel is closed, the p loop occludes the lumen from the extracellular side. Upon voltage activation of a Kv channel, S4 remains in a transmembrane position, but pushes outward and rotates, causing a pull on the S4-S5 linker of all 4 subunits. This “pull” is thought to bend the S6 segment enough to open the pore from the intracellular end, and allow outward K+ current (Figure 1.5, PANEL C)

An inactivated state is a nonconducting state of the ion channel separate from the closed state. The purpose of inactivation can be to either rapidly stop conduction or to prepare a channel for precisely timed conductance. As will be discussed in the next section, in a ventricular action potential the transient outward current, Ito carried through

+ Kv 4.3, is important in early phase repolarization (outward K flow). It is important for this channel to quickly inactivate or the action potential may terminate prematurely, disrupting proper excitation-contraction coupling. In the opposite scenario, the rapid delayed rectifier current, IKr carried through Kv11.1 (HERG), is critical for late phase repolarization. If it was fully active in the earlier phase of repolarization, premature action potential termination could also occur. Not surprisingly, quick inactivation processes have been identified for both these voltage-gated channels51-53. There are several mechanisms for inactivation, but the most well established are derived from studies of Shaker K+ channels54 are N and C-type inactivation55. N-type activation is commonly referred to as a “ball and chain model” where positive charges along the cytoplasmic amino terminus (ball) bind to negative charges in both the S4-S5 linker and vestibule, effectively stopping conduction. N-type inactivation is fast inactivation,

17 occurring within milliseconds. C-type inactivation, is less well understood, but is believed to involve conformational changes along the extracellular pore that prevent conduction. C-type inactivation contributes to a slow inactivation process. Figure 1.6

PANEL A and B illustrate how these two inactivation processes may work.

While the above discussion pertains to a majority of the voltage-gated ion channels, it doesn’t account for all structure or gating mechanisms (i.e. inward rectifier channels) seen in the myocardium. Furthermore, ion channels may coassemble with beta- subunits that can alter ion channel gating kinetics. For the sake of brevity, the next two sections will only cover relevant ion channels, gating, and subunits discussed in the experiments of this dissertation.

Potassium Channels of the Ventricular Action Potential

IK1 is carried through inward rectifying Kir channels

As described earlier, IK1 is a primary deterimant of the resting membrane potential of the ventricular cell. The structure of the channel that carries IK1 is similar to KcSA, in that it consists of four domains, each of which contain only 2 transmembrane segments connected with a p loop.56. The protein subunits that form the channel are from the Kir2.x family (Kir = inward rectifying channel), with Kir2.1 being the predominant pore- forming subunit in the myocardium. The inward rectifier gene KCNJ2 codes for Kir2.157.

There is no S4 voltage sensor associated with Kir channels, rather gating is dependent on extracellular [K+]. In addition, in Kir channels, a large water-filled cytosplasmic pore that from an addition gate (G-loop) between β-strands that form the walls of the pore. This G 18 loop is capable of constriction and thought to have involvement in inward rectifiction58.

Inward rectification refers to the channel’s ability to conduct ions preferentially inward than outward. At potentials negative to the resting membrane, IK1 is inward, but as depolarization occurs extracellular [K+] is increased and the outward current diminishes.

Intracellular polyamines59 and Mg2+ 60 “plug” the channel from the inside to prevent further outward current, thus causing rectification. The role of outward IK1 in terminal repolarization of the ventricular action potential may differ between species. Isolated ventricular canine cells blocked by 10 uM BaCl2 demonstrated an 18% in increase in their action potential duration at 90% repolarization (APD90), compared to only a 5% increase observed in human ventricular mycocytes.61

Ito is carried through Kv channels and initiates early repolarization

The cardiac transient outward current, Ito is a voltage-gated ion channel, with its

α-subunit formed by Kv4.3 (encoded by KCND3) and Kv1.4 (encoded by KCNA4).62

Each protein subunit demonstrates unique inactivation properties63,64, resulting in two components of Ito current, Ito,f and Ito,s. Upon depolarization, Ito,f activates and inactivates

65 rapidly (fast), and also recovers rapidly upon repolarization. Ito,s activates rapidly as well, however is much slower to inactivate and recover from steady-state inactivation

(Figure 1.6, PANEL C). The attenuation of Ito,f by antisense oligo-deoxy nucleotides

66 against Kv4.3, implicated this protein as the carrier of Ito,f. Kv4 channels appear not to inactivate via the classic N and C type inactivation. Deletion of the amphipathic region on the N-terminus of Kv4.1 removes the fast component of recovery from inactivation in 19

Xenopus oocytes expressing the protein.67 However, deletion of arginine residues on the

C-terminus of Kv4.1 removed this component as well68 These studies suggest the inactivation mechanism of Kv4.3 is more complex than Shaker type channels, where C- type inactivation is thought to be involved only in slow inactivation kinetics. Kv1.4 channels, meanwhile, demonstrate canonical N- and C-type inactivation processes that are coupled to each other, similar to Shaker K+ channels69.

In addition to the two pore forming subunits, Ito channels may have accessory β-

70 subunits. In particular, KChiP2 and DPP6 have been demonstrated to impact Ito kinetics .

KChiP2 (K+ Channel Interacting Protein), when co-expressed with Kv in CHO cells, increased peak Ito, slowed inactivation of the current, and the current recovered more quickly from inactivation71. The role of DPP6 in healthy ventricle cardiomyocytes is less clear. Co-expression of DPP6, Kv4.3, and KChiP2 appears to be required to recapitulate

72 normal Ito function. However, in Purkinje cells overexpression of DPP6 increased Ito density by almost 50%, these effects were not observed in ventricle cells with similar overexpression.73

In human and canine ventricle, Kv4.3 is the predominant pore-forming subunit74, and therefore Ito,f is the major contributor of Ito. Ito’s role in early repolarization is due to direct alteration of the phase 1 voltage, which can modulate ICa,L and subsequent calcium handling. For instance, augmentation of Ito not only increases the phase 1 notch, but increases the driving force of ICa,L by moving ECa away from its equilibrium potential

75 (~134 mV). However, reduced Ito prolongs the time to peak ICa,L and also reduces peak

37 2+ 2+ ICa,L amplitude. This “desynchronizes” SR Ca release and leads to lower Ca

2+ transients, suggesting an important role of Ito in physiological Ca handling.Due to the lack of a specific Ito blocker, the contribution of this current to the late phase of ventricular repolarization is still debated. Transfection of Kv4.3 into guinea pig ventricular myocytes, an animal model that lacks endogenous Ito, produced Ito density- dependent APD50 and APD90 reductions, suggesting a role for Ito in both early and late

76 phase repolarization in Guinea pigs . Alternatively, a subset of ventricular myocytes isolated MI-induced canine model demonstrated a 33% reduction in Ito slope conduction,

+ with no other changes in repolarizing K currents, and APD90 was unchanged from

77 controls. Computer aided simulation studies suggest Ito can prolong or shorten action potential duration depending on baseline levels of the current.78 In these studies, increases in Ito at high baseline levels, will overcome the counterbalance of ICa,L and result in early action potential termination. Increasing Ito at low levels of the current may result in action potential prolongation, mediated by an increase in ICa,L.

The delayed rectifier Kv channels are responsible for the bulk of repolarization

There at least two delayed rectifier currents (IKr and IKs) present in the human and canine ventricle, with the existence of a third current (IKur) still under debate. At one time,

IKr and IKs were thought to be one current, IK. It wasn’t until 1990, when a benzenesulfomide antiarrythmic agent E-4031 identified an E-4031-sensitive component

(IKr) that demonstrated rapid activation and rectification compared to an non-rectifying,

79 slowly activating insensitive component, IKs .

The pore-forming alpha-subunit that conducts IKr is Kv11.1, is also known as hERG, and is encoded by KCNH2 mRNA. Like Kir channels, Kv11.1 exhibits inward rectification, however it requires a depolarization step to activate the channel to produce inward current and its rectification does not depend on intracellular block by Mg2+ or polyamines.80 Kv11.1, requiring depolarization to activate (half-activation voltage of -

21.5 mV79), is rapidly deactivated such that the channel is mostly inactive during

+ depolarization, but repolarization allows recovery of the channel and outward K can contribute to Phase 3 of the ventricular action potential81. Inactivation of Kv11.1 is reportedly prevented by application of external tetraethylammonium (TEA) and a mutation at position 631 on the outer pore of the channel enhanced inactivation – both results suggesting C-type inactivation82 Kv11.1 includes a cyclic nucleotide binding domain (CNBD) on the C-terminus. Direct cAMP binding to the domain shifts the

83 voltage dependent activation of IKr to more negative potentials. Interestingly, protein kinase A (PKA) phosphorylation on intracellular sites of the N and C termini, opposes cAMP effects, causing a decrease in IKr amplitude and voltage shifting of activation to more positive potentials.83 Coassembly of Kv11.1 with Mink-related peptide 1 (MiRP1) causes the cAMP-mediated effects to predominate.83 Expression of MiRP1 outside the cardiac conduction system is relatively low,84 leaving an uncertain role for MiRP1 in modulation of IKr in the ventricle.

A tetramer of Kv7.1 (KCNQ1) requires coassembly with the β-subunit minK to generate IKs. As opposed to Kv11.1, this complex exhibits delayed activation that begins

79 around -40 mV and has a voltage for half activation of +15.7 mV. IKs also demonstrates

22 fast inactivation kinetics relative to the diastolic phase of the action potential85, suggesting a limited role in repolarization. Indeed, in ventricular cells isolated from both healthy canine and human subjects, inhibition of IKs by the selective blocker HMR-1556, failed to prolong APD at 1 Hz (60 beats per minute)61. The inactivation mechanism of

Kv7.1 is still unclear, but is thought to involve modulation by extracellular K+.86

62,87 The existence of an ultrarapid Kv current (IKur) in the ventricle is still debated .

This current, well established in the atrium, is formed by Kv1.5 (KCNA5) tetramers,

88 containing two transmembrane segments each . IKur demonstrates activation kinetics that

89 are twice as fast as IKr and activates around -30 mV. However, blockage of IKur with a specific Kv1.5 inhibitor, XEN-D0101, failed to prolong human ventricular action potential at 1 Hz in tissue from general donors.90.

The K+ channels of the ventricular action potential provide a repolarization reserve

The ventricular action potential has at least four K+ currents contributing to repolarization. This redundancy in repolarizing currents is a safeguard that allows repolarization to occur in the absence of one more K+ currents. For instance, blockage of

IK1 alone does not prolong the APD of a human ventricular myocyte, while blockage of

61 IKr alone prolongs it by 26%. In the same study, blockage of IK1 and IKr together produce a prolongation of 73%. This illustrates the concept has been termed “repolarization

91 reserve.” . IKr is the most important contributor to repolarization reserve and while fractional blockage of IKr alone may not result in APD prolongation, any other insult to a

+ repolarizing current combined with IKr block likely will. Disease states causing K 23 channel downregulation, congenital mutations in channels, and pharmacological inhibition of K+ channels all can have profound effects on reducing the repolarization reserve. Failure of repolarization can foster the development of early afterdepolarizations

(EADs), a trigger of ventricular arrhythmias.92

K+ Channels and Ventricular Arrythmogenesis

Prolonged ventricular repolarization predisposes ventricular cells to EADs.93

EADs can arise when the action potential plateau is extended and ICa,L channels reopen, producing a positive voltage oscillation.94 This secondary calcium influx may increase

+ [Cai], causing NCX to exchange one calcium for three Na ions, producing another mechanism for voltage oscillation95. In either case, the end result of an oscillation of sufficient amplitude is an early “triggered” beat. Monophasic action potential recordings of patients with long QT syndrome (LQTS), a genetic repolarization abnormality of multiple etiologies, revealed the presence of EADs in the right ventricle at baseline96. In the same study, EADs were also observed in acquired Long QT syndrome (LQTS), where patients developed prolongation abnormalities through hypokalemia and pharmacological inhibition of repolarizing currents by amiodarone. In congestive heart

+ failure, there is a downregulation of K channels with reported decreases in IK1, IKs, Ito,

97-99 and IKr, predisposing this population to EADs as well .

EAD formation triggers lethal arrhythmias such torsade de pointes (TdP). It is unlikely that one ventricular myocyte having an EAD will successfully conduct a depolarizing current to its neighbors, triggering their activity as well. The reason is 24 because one cell is coupled to at least 11 other cells via gap junctions100, and unless most of them have the same EAD timing, the depolarizing current from the EAD cell will be insufficient to depolarize surrounding cells due to a source-sink-mismatch.95 Computer simulation studies have suggested that up to 700,000 coupled ventricular cells with an identically timed EAD are needed to propagate a premature ventricular contraction

(PVC) .101 An island of midmyocardial cells (M cells) with prolonged APDs relative to surrounding tissue, has been proposed to be large enough to produce a synchronous trigger that can propagate through the myocardium.102 This could result in a focal tachycardia, or due to heterogeneity of the APD between layers of the myocardium, the

EAD produced impulse could encounter an area of refractory tissue (unidirectional block) and the conditions for a re-entrant arrhythmia103, such as TdP, would be set104. Other studies suggest that the chaotic nature of EADs themselves, creates small “islands” of

EAD producing cells that shift location from beat-to-beat95. Because these small EAD islands are next to normal cells, APD dispersion is created such that conduction block and subsequent re-entry can occur.

Action potential beat-to-beat variability can be used to predict arrhythmogenesis

QT prolongation alone is insufficient to predict the development of TdP.105 A major breakthrough in the identification of a biomarker for TdP susceptibility came about

106 through the treatment of chronically AV blocked dogs with d-Sotalol (IKr blocker). The dog model was used since it demonstrated an impaired repolarization reserve that was exacerbated by adding the IKr blocker. Dogs were dosed serially, and with the low dose of 25

2 mg/kg, 25% of the dogs developed TdP versus 75% of the animals infused with the high dose. The QTc interval prolongation was similar between both doses. By measuring the APD of 30 consecutive beats of a left ventricular monophasic action potential, the differences between each beat was assessed. Increases and decreases in APD between successive beats were assessed as a short term variability (STV) marker that was associated with TdP. Furthermore, cells isolated from these dogs, reflected similar STV when exposed to high doses of sotalol, ultimately resulting in EADs. Thus, patch clamp measurements of beat-to-beat variability are validated as a surrogate marker to predict arrhythmogenesis in vivo.

Diastolic Potassium-Conducting Channels of the SAN Action Potential

The If channel activates upon hyperpolarization

If is carried through HCN (Hyperpolarization-activated Cyclic-Nucleotide Gated) channels, part of the superfamily of Kv channels. These channels also have four domains, each consisting of six α-helices. HCN channels are not K+ specific conducting channels, as they also carry Na+, in a ratio of 1 Na+ to 4 K+. The reason for the lack of ion selectivity is unclear since HCN channels have the same GYG sequence of all Kv channels107. One explanation may be that GYG motif complex within the tetramer is not as rigid, thereby allowing ions of different sizes to pass through the channel.108

HCN channels in the SAN can form heterotetromers between HCN4 (the major subunit in canine SAN)109 and HCN2. HCN2 homomeric channels have faster activation kinetics compared to HCN4110, however the functional consequence of 26 heterotetramization on If is not simply summative, suggesting complex current modulation which warrants further experimentation111. A critical component of HCN structure is a cyclic-nucleotide binding domain (CNBD), which is an intracellular domain consisting of three α-helices (A-C) with a β-roll between the first two helices (A-B)112.

The CNBD is connected to the C-terminus of S6 through a C-linker region containing six

α-helices. cAMP binding to the CNBD induces a conformational change in the C-helix, which may loosen the normal rigidity of the C-linker, in turn altering S5 and S6 regions113. The functional consequence of cAMP binding to HCN is an increase in

114 activation and amplitude of If.

HCN channels are open at the maximum diastolic potential (MDP) of an SAN cell

(~ -60 mV). Because they are activated and generate an inward current at potentials negative to -45 mV, they have been considered to be the primary determinant of the diastolic depolarization (phase 4) of SAN cells115. HCN channels are open at diastole, as opposed to other Kv channels, do not inactive but rather close in response to depolarization, where the S4-S5 linker decouples from S6 116. This rapid inactivation is removed when the voltage lowers to more negative potentials. Because of its pivotal role in diastolic depolarization, funny current inhibition would be expected to have dramatic negative chronotopic effects on the SAN. However, blockage of funny current with the specific If blocker, ivabridine, at a dose (3 μM) that blocks up to over 50% of the current, results in only a 10% reduction in firing rate.117 On the other hand, silencing HCN4 in the hearts of mutant mice resulted in a 4-fold reduction in basal heart rate118. These results

27 suggest the funny current reserve must be greatly diminished prior to the observation of negativity chronotropy.

IKAch/IKAdo are carried through ligand gated Kir channels

Both IKAch (current evoked by acetylcholine) and IKAdo (current evoked by adenosine) are carried through GIRK (G-protein coupled inward rectifier) channels. In the SAN, GIRK channels can form heterotetrameric complexes between Kir3.4 (GIRK4) and Kir 3.1 (GIRK1),119 although there is evidence that Kir3.4 can form homotetromers.120 As inward rectifier channels, they have a similar structure to bacterial

KcSA channels, with four domains, each containing two transmembrane helices as well as a cytoplasmic domain with a G-loop. These channels are G-protein coupled as agonist binding to a G-coupled protein receptor (GPCR) releases a Gβγ subunit that associates with the cytoplasmic domain of GIRK. Binding of adenosine to adenosine 1 receptors

(A1R) or acetylcholine to muscarinic acetylcholine receptors (M2) are both processes that release Gβγ from the G-protein complex. Upon association with Gβγ, the cytoplasmic domain of GIRK rotates counterclockwise, opening the inner transmembrane helices that

121 are closed in the absence of Gβγ. The result is the generation of an inward rectifying current with properties similar to IK1. Fast desensitization of GIRK current, where there is a partial current decay lasting for a few milliseconds, is a property of these channels.122

Interestingly, Kir3.4 homomers do not desensitize, and may conduct a larger current than heteromers.120 GIRK current can be specifically blocked by tertiapin Q, a peptide that binds to the extracellular pore opening, between the M1 and M2 transmembrane linker.123 28

In the SAN, activation of GIRK decreases firing rate, however the mechanism is unclear. Studies have reported that both higher doses of adenosine124 and acetylcholine125 can hyperpolarize the SAN membrane by as much as 10 mV via generation of outward

GIRK current, and cause heart rate reductions of 75 beats per minute in rabbits126. This hyperpolarization is thought to increase the time until initiation of Phase 0, and hence increase the cycle length between spontaneous action potentials. Interestingly, there is evidence that adenosine and acetylcholine may exert their effects using different subunits

127 of the G-protein. In this study, A1R preferentially coupled to Gα subunit, which modulates cAMP and If ,and tertiapin Q had no effect on adenosine-induced rate slowing in human embryonic stem cells. These results are consistent with studies in rabbits, that

128 demonstrated that adenosine blocks If with negligible effects on IKAdo. Because If and

IKAch/IKAdo are active at similar diastolic voltages, more studies will be needed in determining the relative contribution of each current in SAN rate slowing. An overview of SAN signaling as it pertains to GIRK and If is presented in Figure 1.7.

Diastolic SAN K+ Channels and Arrhythmogenesis

SAN dysfunction accounts for one half of all pacemaker implants.129 SAN dysfunction can result from heart failure, aging, ischemia, genetic abnormalities, or pharmacologic interventions.130 Left untreated, SAN dysfunction can result in bradyarrhythmias, atrial fibrillation or other forms of tachy-brady arrythmias, all of which are associated with high morbidity and mortality.131

HCN4 has been implicated in sinus node dysfunction. At least five mutations in

HCN4 underlying SAN dysfunction have been identified in humans.132 One of the mutations, hHCN4-573X, was found to cause sinus bradycardia with intermittent atrial fibrillation.133 This same study also confirmed the mutation renders the HCN channel insensitive to cAMP. Other mutations in HCN4 that produced bradycardia in patients,

134,135 either shifted If activation out of the diastolic voltage range for SAN cells , or altered intracellular HCN4 trafficking.135,136

Altered HCN4 also is linked to SAN dysfunction and HF. Both HF patients137 and animal heart failure models109,138 have demonstrated a reduction in intrinsic SAN firing rate. HF models also demonstrate a downregulation of HCN4109 , and in isolated SAN

139 109 cells, If as well . Interestingly in HF, atrial HCN4 upregulation combined with increased fibrosis140, may provide the “ectopic” trigger and substrate for a re-entrant atrial tachyarrhythmia. Furthermore atrial sensitivity to adenosine may also be increased

140 in HF , resulting in shortened APD via IKAdo that may also facilitate atrial fibrillation and tachyarrhythmia in the presence of SAN dysfunction.

As of now, GIRK channels have do not have a clear role in clinical SAN dysfunction. GIRK4 knockout mice, however, demonstrated a blunted bradycardic response to methoxamine (stimulation of vagal reflex) compared to wild-type.141 In addition, GIRK4 mice were also not able to adapt their heart rate rapidly (<2 s) in response to an A1R agonist or methoxamine. These results suggest that GIRK is important in mediating the bradycardic effects induced by both parasympathetic and adenosine stimulation.

The next two chapters will expand upon what is currently known about ion channel regulation in heart failure. Chapter 2 examines how the duration of heart failure regulates repolarizing K+ currents and its effect on ventricular action potentials. Chapter 3 provides a mechanistic insight as to how upregulation of the A1R receptor contributes to

SAN dysfunction. Chapter 4, is more clinically oriented, and addresses the importance of maintaining optimal serum K+ levels in patients taking on antiarrhythmic therapy.

Figure 1.1 The cardiac conduction system.

Note the AV bundle is the same as the His Bundle described in the text. Conduction speed is indicated in parenthesis and units are in seconds. The sinus node, where the impulse begin is at time = 0 s (Reprinted by permission of Macmillan Publishers Ltd: Nat Med 2000;6:969)

Figure 1.2 Ionic basis of the ventricular action potential

Current magnitude and direction are printed in blue. As repolarizing K+ currents are one of the main focuses of this dissertation, they are outlined in red (Reprinted with permission by Physiological Society: Physiol Rev 2005;85:1207.)

Figure 1.3 The ECG and the ventricular action potential

The relationship between a surface electrocardiogram (ECG) and the ventricular action potential. Arrows are for identification of currents related to action potential phases and do not represent direction of current. (Reprinted from the Lancet Infect Dis, Vol. 7, Cardiotoxicity of antimalarial drugs, White N, 549, 2007, with permission from Elsevier)

Figure 1.4 Ionic basis for the sinoatrial node action potential

DD = Diastolic Depolarization. IK = overall contribution of IKur, IKs, and IKr. (Reprinted with permission by Physiological Society: Physiology 2013;28:75)

Figure 1.5 Structure and Proposed Gating Mechanism of K+ Channels

PANEL A is a representative crystal structure of bacterial KcSA channel with ions entering extracellularly (Reprinted by permission by MacMillan Publishers Ltd: Nature 2006;440:327) Panel B illustrates the surface membrane topology of a monomer of a Kv channel (Reprinted by permission of Walters Kluwer Health, Inc: From Rasmussen R, Morales M, Wang S, Liu S, Campbell D, Brahmajothi N, Strauss H, Circulation, Inactivation of Voltage-Gated Cardiac K Channels 1998;82(7):743. Panel C shows open and closed conformations of a Kv1.2 channel. (Modified from Long SB, Campbell EB, MacKinnon R. Voltage sensor of Kv1.2: structural basis of electromechanical coupling. Science 2005;309:905. Reprinted with permission of AAAS.)

Figure 1.6 Inactivation Gating of Kv Channels

PANEL B illustrates the mechanism for C-type inactivation. (PANEL A and B reprinted by permission of Walters Kluwer Health, Inc: From Rasmussen R, Morales M, Wang S, Liu S, Campbell D, Brahmajothi N, Strauss H, Circulation, Inactivation of Voltage-Gated Cardiac K Channels 1998;82(7):741.) PANEL C illustrates the fast and slow type inactivation of Ito from Kv4.3 and Kv1.4 heteromers.

Figure 1.7 GIRK and HCN signaling in SAN cells

Both M2 and A1 agonists promote the dissociation of the G protein complex. The Gα subunit inhibits adenylyl cyclase (AC), preventing cAMP formation and reducing HCN current. The Gβγ subunit activates GIRK current. These effects are antagonized by beta- adrenergic signaling. PKA dependent phosphorylation of Cav1.3 augments L-type current, increasing slope of diastolic depolarization. PKA dependent phosphorylation of

ERG1 (hERG) reduces IKr and shifts activation to more positive potentials, but direct cAMP binding to hERG opposes this effect by shifting the activation to more negative potentials. Ist is a sustained inward current whose molecular correlate has yet to be identified, and function has yet to be determined. It is active in the entire range of the SAN action potential and beta-adrenergic stimulation increases its amplitude. (Modified with permission by Physiological Society: Physiol Rev 2008;88:956.) 38

Figure 1.8 Mechanism for a re-entrant arrhythmia

Roman numerals represent Zone 1 and Zone 2 of adjacent myocardium along a conduction pathway (inset above). A premature beat (arising from an EAD) in Zone 2 will be blocked in one direction because Zone I is still not repolarized (refractory) from the last impulse. However, the premature beat can travel in another direction around an obstacle (i.e. fibrosis) and if appropriately timed (after the refractory period), will generate an action potential (AP) in Zone 1. The red line extended from Zone 2 to Zone 1 along the AP traces represents the EAD and AP “c” is an early AP. This is essentially a “short-circuit” causing rapid heart rates. (Reprinted by permission of Macmillan Publishers Ltd: Nat Rev Drug Discov 2006;5:1038)

Table 1.1. Species-dependent ion current contributions to ventricular action potential

(Action potential figures modified with permission by Physiol Genomics 2008; 35:265).

Chapter 2 : Heart failure duration progressively modulates the arrhythmia substrate through structural and electrical remodeling

Reprinted from Life Sciences, Vol 123, Long VP, Bonilla IM, Vargas-Pinto P, Nishijima

Y, Sridhar S, Li C, Mowrey K, Wright P, Velayutham M, Kumar S, Lee NY, Zweier JL,

Mohler PJ, Gyorke S, Carnes CA. Heart failure duration progressively modulates the arrhythmia substrate through structural and electrical remodeling, 61-71, 2015, reprinted with permission from Elsevier

Introduction

Heart failure (HF) is a leading contributor to morbidity and mortality. In 2010, the number of deaths in the US attributed to HF was ~279,000, and HF was noted in 1 of 9 death certificates 142. Sudden death due to lethal ventricular arrhythmias is six- to nine- fold higher in HF patients than in the general population 143,144, and accounts for up to

50% of deaths in HF patients143.

HF is associated with both structural and electrical remodeling that transforms the normal myocardium into a substrate susceptible to arrhythmogenesis. Left ventricular dilation, hypertrophy, and fibrosis are all examples of compensatory changes that are initially adaptive in the failing heart 145. These changes may progress to become maladaptive resulting in further deterioration of heart function, and have been linked to the development of arrhythmia and/or sudden death 146-148. At the cellular level, cardiomyocytes of the failing heart display electrical remodeling, including a signature prolongation of the action potential (AP) 149,150.

We previously reported that chronic (four or more months) canine tachypacing

HF becomes irreversible and emulates multiple aspects of chronic human HF 151. In the present study, we tested the hypothesis that dogs paced into chronic HF (4 M HF) would demonstrate greater structural and electrophysiological remodeling than dogs paced into acute HF (1 M HF), providing a substrate for ventricular arrhythmias. We used serial echocardiography and electrocardiograms to assess ventricular function and electrical activity, respectively. Patch clamp recordings were used to measure action potentials and

K+ currents in ventricular myocytes. Real-time polymerase chain reaction (RT-PCR) was used to quantify ion channel subunit mRNA.

Materials and Methods

Heart failure canine model

All animal procedures were approved by the Institutional Animal Care and Use

Committee of the Ohio State University. A total of 63 adult mixed breed dogs of either sex (2- 5 years of age) weighing between 8 and 20 kg with normal cardiac function were used. Dogs were verified to have normal cardiac function by routine electrocardiograms and echocardiographic examinations during butorphanol tartrate (0.5 mg kg -1 intramuscularly) sedation. Dogs had a RV pacemaker lead implanted in the RV apex, and

HF was induced (n=16) by tachypacing for four months as previously described 152. To assess time dependence during the progression of HF, echocardiograms and electrocardiogram were measured at baseline, after 1 month, and 4 months of pacing in the 4 M HF group. A second group of dogs was RV tachypaced for 1 month at 180 bpm

(n=17) and echocardiograms were measured at baseline and at the end of the pacing protocol as previously reported. 151,153. An age matched group of 30 healthy dogs were used as controls and studied in parallel. Transmural samples of left ventricular tissue were formalin fixed and embedded in paraffin and sectioned to 5 µm thickness, using standard procedures. Tissue sections were stained with Masson’s Trichrome to define the percentage area of fibrosis, as previously described 153

Myocyte Isolation

On the day of the terminal procedure, the dogs were anesthetized with pentobarbital sodium (50 mg/kg IV). The heart was rapidly removed and perfused with cold cardioplegia solution containing the following in mM: NaCl 110, CaCl2 1.2, KCl 16,

MgCl2 16 and NaHCO3 10. Cannulation of the left circumflex artery was used to perfuse the left ventricle, as previously described.152,154 Adjacent tissue samples were collected and snap frozen for protein analyses. Tyrode’s solution (mM) containing NaCl 130, KCl

5.4, MgCl2 3.5, NaH2PO4 0.5, Glucose 10, HEPES 5 and taurine 20, was used as the initial perfusate. During the cell isolation process the heart was perfused with three different solutions (36ºC). First the heart was perfused for 10 minutes with Tyrode’s solution with 0.1 mM EGTA; followed by perfusion with Tyrode’s solution containing

0.3 mM Calcium, 0.12 mg/ml of Trypsin Inhibitor (NIBCO) and 1.33mg/ml of collagenase (Type II, Worthington), for a maximum of 45 minutes. Following enzymatic digestion, the heart was perfused with normal Tyrode’s solution for five minutes to remove residual enzyme. After digestion, the cells were re-suspended in incubation buffer. This isolation procedure typically yields 40-60% rod shaped ventricular myocytes.

All myocyte electrophysiology experiments were conducted within 10 hours of isolation.

Electrophysiological recordings

To assess myocyte electrophysiology, Amphotericin-B perforated patch clamp techniques with a bath temperature of 36 ± 0.5ºC were used. The myocytes were placed in a laminin coated cell chamber (Cell Microcontrols, Norfolk, VA) and superfused with bath solution containing (in mM): 135 NaCl, 5 MgCl2, 5 KCl, 10 glucose, 1.8 CaCl2, and

5 HEPES with pH adjusted to 7.40 with NaOH. For current measurements the calcium in the bath solution was reduced to 1.0mM in addition, 2 µM nifidepine was added to the bath solution to avoid contamination with L-type Ca2+ current. Borosilicate glass micropipettes with tip resistance of 1.5-4.5 MΩ, were filled with pipette solution containing the following (in mM): 100 K-aspartate, 40 KCl, 5 MgCl, 5 EGTA, 5 HEPES, pH adjusted to 7.2 with KOH.

Action potentials (AP) were recorded in a train of 25 traces at 0.5, 1 and 2 Hz.

The average of the last 10 traces (i.e. from trace 16-25) was used to calculate the action potential duration (APD). APD was calculated at 50 and 90 percent of repolarization

(APD50 and APD90). The standard deviation of the APD90 for the last 10 traces not exhibiting EADs was used to evaluate repolarization variability 106. EADs were defined as positive oscillations occurring in Phase 2 or Phase 3 of the APD.

For current recordings, only recordings with an access resistance <20 MΩ were included in the analyses. Transient outward potassium current (Ito), potassium inward rectifying current (IK1), and delayed rectifying currents (IKr + IKs) were elicited as

77,155 previously described. To assess steady state inactivation kinetics of Ito, a series of

500 ms prepulses were clamped to voltages between -80 mV and +10 mV (holding

156 potential of -60 mV) followed by a 300 ms step to +50 mV (peak current) . Ito elicted during each prepulse was normalized to peak current and plotted against the respective prepulse potential. The AUC for the window current was defined as the area beneath the intersection of the normalized Ito activation and inactivation curves, in the voltage range between -50 to +10 mV. To investigate Ito recovery from inactivation, a two-step protocol

45 was used in which two 200 ms pulses from a holding potential of -60 mV to + 40 mV were separated by a variable interpulse interval of 20 to 1000 ms 156. Rectification ratio was calculated as previously described 157,158.

Data was collected with a low noise data acquisition system Digidata 1440A

(Molecular devices, Sunnyvale, CA), Clampex software and an Axopatch 200A amplifier

(Axon Instruments, Sunnyvale, CA).

Immunoblots

Following protein quantification, tissue lysates were analyzed on Mini-

PROTEAN tetra cell (BioRad) on a 4-15% precast TGX gel (BioRad) in

Tris/Glycine/SDS Buffer (BioRad). Gels were transferred to a nitrocellulose membrane using the Mini-PROTEAN tetra cell (BioRad) in Tris/Glycine buffer with 20% methanol

(v/v, BioRad). Membranes were blocked for 1 hour at room temperature using a 3% BSA solution and incubated with primary antibody overnight at 4°C. Antibodies were KChiP2

(Alomone, Santa Cruz), Kv4.3 (Covance), and GAPDH (Fitzgerald). Donkey anti-rabbit-

HRP (Jackson Laboratories) was used as the secondary antibody. Densitometry was performed using Image lab software and all data was normalized to GAPDH levels present in each sample.

Real-time PCR for gene expression

Total RNA was extracted from the cardiac tissues with Trizol reagent

(Invitrogen), and 2 µg RNA was then converted to cDNA by using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Gene expression was quantified

46 by real-time RT-PCR (Light Cycler 96, Roche Applied Science) using SYBR green assay reagent and gene-specific primer listed in Table1. Relative amplification was quantified by normalizing the gene-specific amplification to that of 18s rRNA in each sample.

-ΔΔC 159,160 Changes in mRNA abundance were calculated using 2( T) method . qPCR reactions were run in triplicates. Primers used were as follows (5’-3’): Kv11.1(forward:

CCTGCTGCTGGTCATCTACA; reverse: TCCTCGTTGGCATTGACATA), KCNE2

(forward: GAACACGACAGCTGAGCAAG; reverse:

ACTGGTGGTAGGGGTCATTG), KChiP2 (forward: GCTGGTTTGTCGGTGATTCT; reverse: AAGAAGCTCTCCACGTGCTC), KCNQ1 (forward:

CTTTACCTGCCAGGGGTACA; reverse: ACCACATACTCCGTCCCAAA), DPP6

(forward: CCCATCGAGTGTCAGCACTA; reverse: GATGGATCGGTACAGGTGCT),

Kv4.3 (forward: GTTTGAGCAGAACTGCATGG; reverse:

GTGGATGGTGCTGAGCTCTT), KCNJ2 (forward: TATCAACGTTGGGTTCGACA; reverse: AAATCAGTTATGGTTCCTTTGGT); 18s (forward:

GCTCTAGAATTACCACAGTTATC; reverse: AAATCAGTTATGGTTCCTTTGGT)

Electron paramagnetic resonance (EPR) spectroscopy

The ventricular tissue samples were flash frozen and stored in liquid nitrogen prior to electron paramagnetic resonance (EPR) analysis to measure radical and paramagnetic species. Both semi-quinone radical and Fe-S centers were quantified using previously described methods161-163 Each EPR sample was prepared by transferring the frozen heart tissue (235–570 mg) into a ceramic mortar pre-chilled with liquid nitrogen. 47

The tissue was then crushed in liquid nitrogen using a pestle. The tissue in liquid N2 was then loaded into a finger Dewar containing liquid nitrogen. Low temperature, 77 K, EPR spectra were recorded with a Bruker ESP 300E spectrometer (Bruker BioSciences,

Billerica, MA, USA) operating at X-band with 100 KHz modulation frequency and a

164 TM110 cavity as described previously . The finger Dewar containing heart tissue samples in liquid nitrogen was placed within the EPR spectrometer cavity. All spectra were recorded with the following parameters: receiver gain = 1 × 105, modulation amplitude = 2 G (4 G for Fe-S signals), time constant = 164 ms, scan time = 60 s, microwave power = 1 mW (20 mW for Fe-S signals), and number of scans = 10.

Data Analysis

Cellular electrophysiology data were analyzed using Clampfit 10.3 software

(Axon Instruments) and Origin 9.0 software (OriginLab, Northampton, MA, USA). One way repeated measured ANOVA was used to analyze differences within groups, while comparison between groups was analyzed by one-way ANOVA with post hoc least significant difference testing or Students t-test (OriginPro 8.6, OriginLab). Normality was tested via Kolmogorov-Smirnov. Non parametric analysis was performed using Chi-

Square. All data are presented as mean ± SE (or SD for PCR quantification) and p<0.05 was the criterion for statistical significance for all comparisons.

Chemicals

All chemicals used were purchased from Sigma-Aldrich (St. Louis, MO, USA) and Fisher Scientific (Pittsburgh, PA, USA), unless otherwise noted. All buffers and solutions were prepared daily.

Results

Chronic tachypacing is accompanied by an increase in left ventricular mass, a prolonged

QTc interval, and interstitial fibrosis

RV tachypacing resulted in significantly impaired contractility, LV chamber dilation and increased LV mass at 4 M HF compared to 1 month HF and baseline

(TABLE 1). The corrected QT interval was significantly prolonged at 4 M HF versus baseline.

Ventricular tissue samples (control, 1 M HF, and 4 M HF) had significantly increased fibrosis in the 4 M HF group (p<0.05 vs control) (Figure 1). Collectively, these results suggest that time-dependent decreases in cardiac function with continued RV tachypacing are associated with the development of structural remodeling (cardiac hypertrophy and fibrosis) in chronic (4 M) HF.

Chronic tachypacing results in a prolonged AP associated with early afterdepolarizations, increased myocyte size, and downregulation of repolarizing K+ currents

Heart failure prolonged ventricular cardiomyocyte action potentials (Figure 2). In addition to AP prolongation, a disappearance of the prominent phase one “notch” occurred with HF. APD50 was significantly prolonged at all rates tested (0.5, 1, and 2 Hz) 49 in both 1 M HF and 4 M HF groups compared with control myocytes (p<0.05) (Figure

2B). APD90 was significantly increased in 4 M HF compared to both baseline and 1 M

HF at all rates (p<0.05) (Figure 2C). Rate-adaptation of the APD was maintained in each group (p<0.05). The beat-to-beat variability of APD90, a marker of proarrhythmic potential 165(Figure 2D) was significantly increased in the 4 M HF group at all rates versus control (p<0.05), and at 1 Hz versus 1 M HF (p<0.05). Consistent with these findings, EADs were significantly more frequent in the 4 M HF group compared to both control and 1 M HF (p<0.05) (Figure 2E). No change in the resting membrane potential was found between the groups. Collectively, this data suggests that HF duration mediates action potential prolongation and the development of a proarrhythmic ventricular substrate.

Membrane capacitance was used as a measure of ventricular cell hypertrophy.

Consistent with the LV mass findings, 4 M HF ventricular myocytes had a significantly larger capacitance (239.5 ± 16.6, n=23, p<0.05) compared to controls (159.0 ± 6.6 pF,

(n=52) and 1 M HF (163.7 ± 9.4 pF,n=29)

Transient outward current (Ito) density and the corresponding slope conductance was significantly decreased in both 1 M and 4 M HF groups (p<0.05 vs Control) (Figure

3B and 3C). Inactivation of Ito was best fitted as the sum of two exponentials, a rapidly inactivating Ito,fast (Ito,f) and a slowly inactivating Ito,slow (Ito,s). There was no significant difference in either time constant between groups. Mean Ito,f and Ito,s at +50 mV was 18.4

± 1.9 ms and 87.3 ± 21.1 ms, 11.3 ± 0.80 ms and 43.1 ± 10.2 ms, and 15.8 ± 1.5 ms and

68.9 ± 21.5 ms for control (n=23), 1 M HF (n=5), and 4 M HF (n=10), respectively. The

50 rapidly inactivating component comprised 80.6 ± 2.6% of the decay current in the control group, 73.7 ± 6.2% in 1 M HF, and 74.8± 7.1% in 4 M HF (p=NS). A plot of the steady state inactivation of Ito against current activation revealed no voltage shift in kinetics

(Figure 4A). However, there was a decrease in the overall current available (AUC) in both 1 and 4 M HF compared to controls (AUC values were reduced by 97% for 1 M HF and 63% for 4 M HF). This “window” current peaked around -35 mV, representing approximately 5% of peak Ito (Figure 4B). The time for Ito recovery from activation was assessed using a two-step protocol (Figure 4C). The time for 50% recovery from inactivation was prolonged in the 4 M HF group vs. controls (185.5±16.9 ms vs.

107.0±17.3 ms, p<0.05).

HF-dependent changes in the inward rectifier current, IK1 are shown in Figure 5.

There were no significant differences in inward slope conductance between groups. Peak

IK1 outward current was significantly decreased in 4 M HF compared to controls (p<0.05,

Figure 5C). The rectification ratio was significantly increased in 4 M HF (p<0.05 vs control, Figure 5D).

HF-induced alterations in the delayed rectifier currents, IKr and IKs are shown in

Figure 6. IKs amplitude was significantly decreased in 4 M HF at all test potentials compared to both 1 M HF and controls (Figure 6B), and IKs slope conductance was decreased in 4 M HF (p<0.05 vs control) (Figure 6C). IKr was significantly decreased in

4 M HF compared to 1 M HF and control (p<0.05) (Figure 6D).

Protein and mRNA expression of K Channel Subunits does not correlate with function

Ito in the canine ventricle is conducted through channel pore-forming subunits

Kv4.3 and Kv1.4, with Kv4.3 being the main alpha subunit contributing to the current 166

Kv4.3 can assemble with multiple accessory proteins including K+ channel interacting protein 2 (KChiP2)which can modulate Kv4.3 gating and function 167. Dipeptidyl- aminopeptidase-like protein 6 (DPP6) also coassembles with Kv4.3, and has been shown

168 to alter Ito kinetics . Biochemical analysis showed no significant difference between protein expression of Kv4.3 and KChiP2 in the control, 1 M HF, and 4 M HF group

(Figure 7A and 7B). In contrast, notable HF-dependent changes were observed in transmural gene expression of K+ channel subunits (Figure 7C). The mRNA of the Kv4.3 and Kv1.4 subunits demonstrated no change at 1 M HF, but increased gene expression at

4 M HF (p<0.05 vs control and 1 M HF). Similarly the mRNA levels for KChiP2 and

DPP6 increased at 4 M HF. KCNJ2 encodes the Kir2.1 channel, the pore-forming subunit of IK1. At 1 M HF, KCNJ2 gene expression was significantly elevated (p<0.05 vs control), and continued to increase at 4 M HF (p<0.05 vs control, p<0.05 vs 1 M HF). IKs is carried via a complex which includes the pore-forming KvLTQ1 (KCNQ1) and minK

(KCNE1). Both KCNQ1 and KCNE1 mRNA are significantly upregulated at 4 M HF compared to both control (p<0.05) and 1 M HF (p<0.05). IKr is carried through Kv11.1 channels, and no differences in Kv11.1 mRNA were found between the three groups.

EPR spectroscopy

EPR spectroscopy revealed a modest increase in semiquinone radicals that did not reach significance in left ventricular tissue among either of the HF groups compared to 52 control (P=NS) (Figure 8B). However, Fe-S center signals were significantly increased in both 1 M HF and 4 M HF (p<0.05 vs control) (Figure 8C).

Discussion

Large animal models of dilated cardiomyopathy are accepted as a surrogate of the pathology of human HF 169. In the present study, we found that the 4 month duration of

HF overcomes a limitation of the short-term canine tachypacing model - the lack of structural remodeling. We observed an increase in LV mass at 4 M HF (≈1.7 fold), coinciding with an increase in myocyte membrane capacitance (≈1.5 fold) – both indicators of hypertrophy. These results are consistent with a previous report of increased

LV mass after 7 to 10 months of chronic RV pacing 151. We also report an increase in interstitial fibrosis at 4 M HF. The appearance of increased fibrosis in 4 to 6 week canine

HF models varies between studies 170-172. A key difference between the current model and previous reports is that our chronic pacing rates are not as rapid (180 bpm vs 240 bpm) resulting in a sustained hypertrophic response not seen in more rapid, short-term pacing models. This notion is supported by our previous work, where we observed increases in

LV mass after 10 months of pacing, 6 months of which was maintained at 160 bpm 151. It is well known that hypertrophy induces the structural remodeling of the collagen matrix necessary for fibrosis. 173 The lack of fibrosis in rapid pacing models may be responsible for their “reversible” nature.

A hallmark of HF is downregulation of repolarizing K+ currents and subsequent

AP prolongation. The most studied and consistently downregulated K+ current in both

98,166,174 canine HF models and human HF is Ito , and we found reductions in both 1 M HF and 4 M HF groups. The role of Ito in early repolarization is evident as the loss of the prominent phase one “notch” during HF. Ito may contribute to the prolonged APD50 in

37 both HF groups, as altered Ito can alter ICa . The role of Ito in late repolarization (APD90) is less clear. The peak of the “window current” we observed is in agreement with findings

175 from other groups , and suggests Ito may contribute repolarizing current during the plateau of the AP.

The molecular mechanisms for downregulation of Ito in HF are still not fully elucidated. In both short-term canine tachypacing models and limited human HF studies, reductions in Kv4.3 protein have been reported 176,177. The relatively consistent finding of

176- reduced Kv4.3 mRNA in canine HF models is a plausible mechanism for decreased Ito

178. However, gene expression studies in human HF are conflicting, demonstrating decreases 179 or increases 180 in Kv4.3 mRNA. Here, we report an increase in Kv4.3 mRNA at 4 M HF with no change in its protein expression. Consistent with our findings, other animal models of tachypacing-induced HF have reported no change in total Kv4.3

181 mRNA or protein, yet have also found reduced Ito . One potential explanation for the reduced current could a change in the accessory subunit, KChiP2. However, we found no change in KChIP2 protein expression despite a significant increase in KChiP2 mRNA at

4 M HF. This finding is contradictory to reports of KChIP2 mRNA being downregulated in failing human hearts 180,182

Interestingly, we report a prolonged time of Ito recovery from inactivation in 4 M

HF compared to controls. Such results were not seen in other canine HF studies using

54 midmyocardial cells183, but were reported in epicardial cells184. DPP6, an ancillary subunit of Kv4.3, has been shown to slow recovery from inactivation in CHO cells when expressed with Kv4.3 168. Here, we report that gene expression of DPP6 is increased in 4

M HF. More studies will be needed to determine the precise role of DPP6 in modulating cardiac Ito, as well as whether Ito inactivation contributes to APD90 prolongation at faster heart rates (cycle length<500 ms) due to the slow recovery from inactivation.

185 Outward IK1 is a modulator of terminal repolarization . We observed unchanged inward current with a decrease in the peak outward current (-60 mV) reflecting altered IK1 rectification 157,158. Because rectification factors, such as polyamines and intracellular

186 magnesium, are involved in the normal gating of IK1 , our data suggests a possible role for these factors in decreasing the outward portion of this current in HF. Given the role of

187 IK1 in the pathogenesis of arrhythmias , additional study of IK1 in HF will be required to clarify the variable results.

The delayed rectifier currents, IKr and IKs, play a prominent role in phase 2 and

188 phase 3 repolarization . Studies in both humans and canines have suggested that IKs reduction only prolongs the APD in the setting of downregulation of other K+ currents

(reduced repolarization reserve) 189-191. Consistent with a previous 4 to 6 week canine

98 model of HF , we found a reduction in IKs at 4 M HF. IKr is a dominant modulator of ventricular repolarization, with reductions evident as both AP and QT prolongation 192.

We found a reduction in IKr in 4 M HF, which does not occur in 4-6 week tachypacing models 98 or in 1 month HF in the present study. The HF duration-dependent reduction may be explained by hypertrophic modulation of IKr as this was only observed in the 4 M

HF group. This is consistent with reduced IKr observed in a proarrhythmic chronic AV

193 nodal block canine model of biventricular hypertrophy . While not measuring IKr directly, human studies using E-4031 (a specific IKr blocker) have demonstrated reduced response to IKr blockade in failing human hearts compared to controls. This study also found Kv11.1 hERG 1a protein expression was reduced compared to controls194, suggesting reduced functional expression of IKr may be one mechanism contributing to

HF-induced IKr downregulation.

Since we found reductions in multiple repolarizing currents, this suggests a role

190 for reduced repolarization reserve as the basis of EAD formation. Reduction of IKr , specifically, has been demonstrated to produce repolarization instability, which in itself can predict proarrhythmia 195. Our study showed that chronic HF results in both AP and

QTc prolongation with concurrent increased beat-to-beat variability of cellular repolarization and the appearance of EADs. In normal canine ventricle, combined

175 inhibition of IKr, IKs, and Ito block can induce EADs , suggesting these reductions as a mechanism for our observed EADs in HF.

A recent report suggests a role for reactive oxygen species (ROS) in the

196 regulation of IKr function Increased oxidative stress has been implicated in the pathophysiology of several forms of HF, included non-ischemic dilated cardiomyopathy

197. While we found no increases in superoxide anion formation, we had previously demonstrated a generalized increase in ventricular ROS in this 4 M HF model 198. Nitric oxide synthase-dependent ROS are suggested to have a role in the regulation of several repolarizing K+ currents 199. In this study, we found increases in Fe-S signal intensity in

56 both 1 M HF and 4 M HF, consistent with increased reduction of these centers that could arise secondary to an increase in nitric oxide (NO) that can block distal electron transport

164,200 201 . NO has been found to reduce Ito in human cardiomyocytes and therefore the observed HF-induced current decreases may be in part attributable to altered NO signaling.

Limitations

Our model is a model of non-ischemic dilated cardiomyopathy, and does not necessarily reflect all etiologies of HF. We also limited our study to ventricular myocytes from the midmyocardial layer which may not fully reflect all transmural changes. We did not assess in vivo arrhythmias, but have previously reported increased premature ventricular contractions in chronic canine HF202. The present study did not evaluate the contribution of other arrhythmogenic mechanisms; previous work in this model has shown that changes in calcium handling attributable in part to posttranslational modifications of ryanodine receptors by ROS occur 198and may also contribute to arrhythmogenesis. Our study also did not evaluate other potentially affected ion currents

(e.g. late sodium or sodium-calcium exchanger) which may also contribute to arrhythmogenesis 203,204.

We did not assess protein expression of all K+ subunits. Furthermore, differences in mRNA expression also exist between this model and previous reports 176,177. The literature on ion channel gene and protein expression in heart failure is highly variable and sometimes contradictory 92,180. The variability may reflect the complexity of the system which is affected by modulation downstream of mRNA transcription, such as 57 micro-RNA targeted degradation, post-translational modifications, and/or changes in protein trafficking can determine ion channel function. Therefore, alterations in gene expression may be dynamic and not always translate to equivalent protein expression.

Conclusion

Electrophysiological modeling during human HF is poorly defined due to the reliance on end-stage HF (explanted hearts from transplant recipients) who are treated with multiple drugs which can elicit their own electrophysiologic effects 205and limited access to true normal controls 206. In this paper, we present a chronic canine HF model which emulates many of the alterations seen in human HF more accurately than other short-term canine tachypacing models. The downregulation of IKr, not seen in other pacing models, along with other K+ currents provides a rational mechanism for EAD formation. We present data suggesting that duration of HF produces progressive electrical remodeling, resulting in proarrhythmic potential at the cellular and organ level.

Further studies are warranted to elucidate the relationships between ion channel subunit gene, protein and function during heart failure.

Baseline 1 Month HF 4 Month HF

Fractional shortening (%) 29.6 ± 1.49 15.2 ± 0.97 * 11.2 ± 0.84 *, #

Left Ventricle Dimension (cm)

Diastole 3.46 ± 0.13 4.33 ± 0.15 * 5.28 ± 0.21 *,#

Systole 2.41 ± 0.07 3.67 ± 0.12 * 4.68 ± 0.18 *,#

Left Ventricle mass (g) 91.6 ± 9.15 121.2 ± 11.0 157.7 ± 16.1 *,#

ECG parameters

PR (ms) 110.7 ± 5.24 108.1 ± 5.01 107.9 ± 3.85

QRS (ms) 45.8 ± 2.47 50.1 ± 3.03 54.2 ± 4.10*

RR (ms) 578.4 ± 35.4 520.2 ± 53.3 511.2 ± 69.4

QT (ms) 201.1 ± 6.10 199.6 ± 7.40 198.9 ± 8.38

* QTcf (ms) 242.5 ± 3.35 250.1 ± 4.59 252.4 ± 4.33

Table 2.1. Echocardiographic and Electrocardiogram parameters in 4 M HF animals

N=7-10 per observation; *p<0.05 vs baseline; #p<0.05 vs 1 M HF. Values are means ±

SE.

Figure 2.1 Interstital Fibrosis is Increased in Chronic HF

Representative Masson’s Trichrome staining of LV tissue. A. Control. B. 1 M HF. C. 4 M HF. D. Summary data (* p<0.05 vs control).

Figure 2.2 Progressive action potential prolongation and cellular arrhythmias during heart failure

A. Representative action potentials at 1 Hz. B. APD50 is prolonged at 1 and 4 months of

HF (p<0.05) . C. APD90 is significantly prolonged in 4 M HF compared to 1 M HF and control (p<0.05). D. Beat-to-beat variability (representative, top) was significantly increased in 4 M HF (* p<0.05 vs. control, # p<0.05 vs. 1M HF). E. Early afterdepolarizations (representative, top) were more frequent in 4 M HF (*p<0.05 vs control).

Figure 2.3HF decreases Ito

A. Representative Ito current tracings from each group; voltage protocol shown in the inset. B. I-V curves (*p<0.05 vs. control) C. Ito slope conductance is decreased in HF (*p<0.05).

Figure 2.4 Ito kinetics are altered in chronic HF

A. Representative steady state inactivation traces of Ito recorded with the voltage protocol displayed in the inset. B. Steady state inactivation and activation curves of Ito fit to

Boltzmann functions, demonstrates a “window” current, that is reduced as the heart fails

C. Representative traces elicited by two-step protocol in control and 4 M HF; voltage protocol in inset. D. Summary data of recovery from inactivation; HF significantly prolongs recovery to 50% of total Ito current (p<0.05 vs control).

Figure 2.5 Outward IK1, but not inward IK1 is reduced in chronic but not short-term HF

A. Representative IK1 current tracing, voltage protocol displayed in the inset. B. I-V curves; inset shows expanded I-V curve of outward IK1 C. Peak outward IK1 is significantly reduced in 4 M HF vs. control (p<0.05). D. Rectification ratio is significantly increased in 4 M HF vs control (p<0.05).

Figure 2.6 IKs and IKr are reduced in chronic HF

A. Representative IKs tail currents (defined as the sotalol-insensitive current) from each group; inset: voltage protocol B. I-V curves (*p<0.05 vs. control; # p<0.05 vs. 1 M HF)

C. IKs slope conductance is reduced in 4 M HF (p<0.05 vs control). D. Representative IKr tail currents (defined as the sotalol-sensitive current) recorded with the same voltage protocol in (A). For ease of viewing, only traces recorded -20 mV, +30 mV, and +60 mV are depicted. E. IKr I-V curves (*p<0.05 vs control).

Figure 2.7 Protein expression and mRNA levels of K+ channel subunits

A. Representative Western Blots of Ito subunits Kv4.3 and KChiP2. B. Summary data of

Kv4.3 and KChIP2 protein expression (p=NS) C. mRNA for K+ channel subunits.

(*p<0.05 vs. control; # p<0.05 vs. 1 M HF).

Figure 2.8 Heart Failure increases ventricular oxidative stress

A. Representative EPR spectra of tissue homogenates measured at 77 K in control, 1 M

HF, and 4 M HF. B. Summary data of semi-quinone radical centeres. (C) Summary data of Fe-S centers. (p< 0.05 vs. control).

Chapter 3 : Chronic Heart Failure Increases Negative

Chronotropic Effects of Adenosine in Sinoatrial Pacemaker

Cells via increased A1R-mediated If block and activation of

GIRK-mediated IKAdo

Introduction:

Heart failure (HF) results in over 300,000 deaths per year in the United States207, with up to half of these deaths attributed to cardiac arrhythmia208. Patients with HF frequently have abnormalities of the pacemaker and conduction system137, and sudden death in HF due to bradycardia accounts for approximately 40% of these deaths209,210. In addition, up to 50% of patients with sinus node (SAN) dysfunction can also experience tachy-brady syndrome as a contributor to atrial tachyarrhythmias; this sequelae of sick sinus syndrome carries an increased risk of embolic stroke.211 Unfortunately, pacemaker therapy has not been shown to improve survival in patients with bradycardia secondary to

SAN dysfunction212,213, and there remains an unmet need for identification of strategies to improve outcomes in patients with concomitant HF and SAN dysfunction.

Adenosine is an endogenous metabolite of the heart, and is known to cause negative chronotropic actions on the SAN214,215. Myocardial adenosine levels increase in response to metabolic stressors and cell injury such as hypoxia216 and ischemia217. These high workload states result in increased free adenosine diphosphate ([ADP]) to drive oxidative phosphorylation218, resulting in adenosine triphosphate ([ATP]) and adenosine monophosphate ([AMP]) formation via the myokinase reaction. Breakdown of [AMP] by cytosolic 5’-nucleotidase to adenosine and its subsequent release into the extracellular space through equilabrative nucleoside transporters (ENTs)219 is the primary mechanism of formation of extracellular adenosine under cardiac stress, however extracellular production via 5’ ecto-nucleotidase may also be a source220,221. Interestingly, while adenosine levels have been reported to be increased approximately 3-fold in the

69 pericardial fluid in canine HF222, other models of left ventricular dysfunction demonstrate an 4-fold net decrease in cardiac adenosine during decompensated HF.223

Adenosine exerts its negative chronotropic actions through its effects on adenosine A1 receptors (A1Rs) in the SAN 224, however the electrophysiological mechanism that induces rate slowing remains controversial. A1Rs are G-coupled receptors (GPCRs), and upon activation, GPCRs activate G protein-coupled inward rectifying potassium channels (GIRK) by Gβγ subunits. The Gαi/o subunit, meanwhile, can inhibit adenylyl cyclase to reduce intracellular cAMP levels and PKA activity,

225 2+ 226 putatively inhibiting funny current, If , and L-type Ca current, ICa,L , respectively. In rabbit sinoatrial node cells, the adenosine-induced GIRK current, IKAdo, was suggested to be primarily responsible for hyperpolarization and rate slowing in rabbits – with If and

227 ICa,L inhibition only observed in the presence of β-agonism . However another report

228 229 found that adenosine induced block of If without GIRK activation . The role of A1R signaling in SAN myocytes in HF has yet to be addressed.

We have previously demonstrated that A1R receptors are upregulated in a chronic model of canine HF 140. Furthermore, dual-sided intramural optical mapping in this model demonstrated that pathophysiological concentrations of adenosine are associated with enhanced depression of SAN function and facilitation of atrial fibrillation in HF. The primary purpose of this study was to examine the role of A1R signaling in HF within pacemaker cells at physiological concentrations. We hypothesized that HF SAN myocytes would demonstrate increased sensitivity to the negative chronotropic effects of adenosine. To test this hypothesis we used the perforated patch clamp technique in

70 isolated SAN myocytes from this canine HF model and measured action potentials and currents in the presence of adenosine. To assess A1R and GIRK contribution to adenosine effects, action potentials were also measured in the presence of 8-cyclopentyl-

1,3-dipropylxanthine (DPCPX), an A1-selective antagonist, or tertiapin-Q (TPQ), a selective GIRK antagonist. We also investigated the effect of HF on cardiac adenosine levels. Adenosine tissue content was quantified through the use of LC-MS. Real-time polymerase chain reaction (RT-PCR) was used to quantify ion channel subunits, A1R, and ENTs.

Materials and Methods:

Heart failure canine model

All animal procedures were approved by the Institutional Animal Care and Use

Committee of the Ohio State University, conforming to the NIH Guide for the Care and

Use of Laboratory animals. A total of 40 adult mixed breed dogs with normal cardiac function were used. Dogs were verified to have normal cardiac function echocardiographic examinations during butorphanol tartrate (0.5 mg kg -1 intramuscularly) sedation. Dogs had a RV pacemaker lead implanted in the RV apex, and

HF was induced (n=19, weight pre-pacing 20.9±1.4 kg) by tachypacing for four months as previously described152 . Echocardiograms were measured at baseline and 4 months of pacing. Left ventricular dilatation was significantly increased in HF dogs and left ventricular fractional shortening was reduced from 36.8±1.4% pre-pacing to 16.9±1.3%

(p<0.01) post-pacing. An age matched group of 22 healthy dogs (weight, 21.4±1.2 kg) were used as controls (CTL) and studied in parallel.

Cell Isolation

On the day of the terminal procedure, anesthesia was induced with pentobarbital sodium (50 mg/kg IV) and maintained with 2% isoflurane. Once the chest cavity was opened and the beating heart exposed, a section of the right atrial appendage was snap frozen in vivo with customized tongs which were immersed in liquid nitrogen immediately before use, and tissue was immediately stored in liquid nitrogen for subsequent cardiac adenosine assays. Afterwards, the heart was rapidly removed and perfused with cold cardioplegia solution containing the following in mM: NaCl 110,

CaCl2 1.2, KCl 16, MgCl2 16 and NaHCO3 10. The right atrium and surrounding tissue were removed and placed in ice-cold Tyrode’s solution (mM) containing NaCl 130, KCl

5.4, MgCl2 3.5, NaH2PO4 0.5, Glucose 10, HEPES 5 and taurine 20 with pH adjusted to

7.4. SAN tissue was excised along the borders of the SAN artery, and cut into perpendicular strips approximately 2 mm x 2 mm. Select dogs had SAN tissue strips snap frozen in liquid nitrogen and stored at -80ºC for subsequent mRNA analyses. For cell isolation, chunks were placed in Tyrode’s solution with 0.2 mM Ca2+ and gently shaken at 37ºC for 5 minutes. The supernatant was replaced by enzyme solution: Tyrode’s solution containing 0.2 mM Ca2+, 250 U/mL collagenase (Type II, Worthington), 0.96

U/mL protease (Sigma), 2.25 U/mL elastase (Worthington). The tissue was enzymatically digested while gently shaking for 20 minutes at 37ºC. Supernatant was removed and chunks were transferred to 5 mL of KB solution containing (mM): Glutamic Acid 100, 72

HEPES 5, Glucose 20, KCl 25, K Aspartate 10, MgSO4 2, KH2PO4 10, Taurine 20,

Creatine 5, EGTA 0.5, BSA 1mg/mL, with pH adjusted to 7.2 with KOH. Chunks were gently triturated with a glass pipette for 3 minutes, and supernatant was examined under phase-contrast microscopy for spindle-shaped SAN cells. Enzymatic digestion was repeated in intervals of 10 minutes until the presence of cells were confirmed. Cells were stored at 4º C until use. This isolation procedure yields between 5-30% spontaneously beating SAN myocytes. All cellular electrophysiology experiments were conducted within 12 hours of isolation.

Quantification of Cardiac adenosine levels

Adenosine from 10 mg of freeze dried ground RAA tissue was extracted using boiling water following a previously described protocol230,231; Three nmol of 2- chloroadenosine was added as an internal standard at the time of the extraction to each sample.

After lyophilization, extracts were resuspended in 300 L ultrapure water, vortexed and loaded on an Amicon 3kDA filtering device. The samples were centrifuged at 14,500 x g for 60 min at 4°C, and then, transferred to LC-MS/MS glass vials. The analysis of adenosine and 2-chloroadenosine was carried out using a UHPLC (Ultra High

Pressure Liquid Chromatography) 1290 from Agilent Technologies, Inc (Santa Clara,

CA) coupled to a hybrid Triple Quadrupole/Ion trap mass spectrometer QTRAP 5500 from AB Sciex (Framingham, MA). The extracts were placed in the autosampler kept at 4

ºC. Twenty L of extract was diluted in 980 L of double distilled water and 2 L of sample was injected onto the column. The liquid chromatography analysis was carried 73 out at 30 ºC. Adenosine and 2-chloroadenosine were separated using a Zorbax column

(4.6 x 100mm) with a 0.3 mm frit from Agilent Technologies, Inc (Santo Clara, CA). The gradient used to separate the metabolites consisted of methanol (solvent A) and 10 mM ammonium acetate in water (solvent B). The total LC-MS/MS run was 12 minutes with a flow rate of 750 l/min. The gradient was as follows: A= 0-0.5 minutes 20 %, 0.5-5 minutes 50 %, 5-5.1 minutes 90 %, 5.1-8 min 90 %, 8-8.1 minutes 20%, and 8.1-12 minutes 20%. The mass spectra were acquired using Turbo Spray Ionization of 4000V in positive ion mode and multiple reaction monitoring. The curtain gas (nitrogen), CAD

(Collision Activated Dissociation), nebulizing and heating gas were set to 35 psi, medium, 65 psi and 55 psi, respectively. The mass spectrometer was set to have a dwell time of 100 msec. The parameters for each metabolite were set as follows: DP:

Declustering Potential, EP: Entrance Potential, CE: Collision Energy, CXP: Collision cell

Exit Potential, are all shown from each metabolite.

Metabolites

Retention Parent Daughter DP EP CE CXP Metabolite Time Mass Mass (volts) (volts) (volts) (volts) (min)

Adenosine 268.1 136.0 2.7 50 10 23 16

2-chloroadenosine 302.0 170.0 3.9 50 10 25 20

LC-MS/MS data were acquired and processed using Analyst 1.6.1 software.

Real-time PCR for gene expression

Total RNA was extracted from the cardiac tissues with Trizol reagent

(Invitrogen), and 2 µg RNA was then converted to cDNA by using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Gene expression was quantified by real-time RT-PCR (Light Cycler 96, Roche Applied Science) using SYBR green assay reagent and gene-specific primer listed in Table1. Relative amplification was quantified by normalizing the gene-specific amplification to that of 18s rRNA in each sample.

-ΔΔC 159,160 Changes in mRNA abundance were calculated using 2( T) method . qPCR reactions were run in triplicates. Primers used were as follows (5’-3’): KCNJ3 (forward:;

CGTCCCCTTTAATAGCACCA; reverse: GGCAAATCTCCCAAGCTGTA), KCNJ5

(forward: GCTGGATCAAGAGGAGTTCG; reverse: TGTTGGTCTCGTAGGTGTCG),

HCN4 (forward: GGATTGTGGTGGAGGACAAC; reverse:

GCTGAGGATCTTGGTGAAGC), HCN2 (forward: TGAGCGTGGACAACTTCAAC; reverse: GCGGTCGTACTTGACGATCT), ADORA1 (forward:

GTGATCTGGGCAGTGAAGGT; reverse: GAGCTCTGGGTGAGGATGAG), NT5E

(forward: GGCACAATTACCTGGGAGAA; reverse:

CAGGTTTTCGGGAAAGATCA), ENT1 (forward: TGGCTACCTCGCCAGTCTCT; reverse: TCGGCTGGCTTCACTTTCTT); ENT2 (forward:

CGTGCTGGTCTTCACAGTCACT; reverse: GGTCACCATGGCCGTGAT); ENT3

(forward: TGTGCCTCATGGCGAACT; reverse: CAGCACTCGGACATGAATCG);

ENT4 (forward: CGACATGAGCCTCACGTAATC; reverse:

GTCTCTCCACCAGCGCGTTA); 18s (forward: GCTCTAGAATTACCACAGTTATC; reverse: AAATCAGTTATGGTTCCTTTGGT)

Electrophysiological recordings

To assess SAN cell electrophysiology, Amphotericin-B perforated patch clamp techniques with a bath temperature of 36 ± 0.5ºC were used. The myocytes were placed in a laminin coated cell chamber (Cell Microcontrols, Norfolk, VA) and superfused with bath solution containing (in mM): 135 NaCl, 5 MgCl2, 5 KCl, 10 glucose, 1.8 CaCl2, and

5 HEPES with pH adjusted to 7.40 with NaOH. For current measurements the calcium in the bath solution was reduced to 1.0 mM with the addition of 2 µM nifedipine to avoid contamination with L-type Ca2+ current. Borosilicate glass micropipettes with tip resistance of 2-5.5 MΩ, were filled with pipette solution containing the following (in mM): 100 K-aspartate, 40 KCl, 5 MgCl, 5 EGTA, 5 HEPES, pH adjusted to 7.2 with

KOH. Action potentials (AP) were characterized by spontaneous firing rate (bpm), action potential duration at 50 and 90% repolarization (APD50 and APD90), maximal diastolic potential (MDP), action potential over shoot, maximum conduction velocity (dV/dtmax), and slope of slow diastolic depolarization rate (SDD, measured as the linear of slope of

10% and 50% between MDP and 50% of depolarization). Only stable baseline recordings with positive overshoots and MDP equal to or more negative than -48 mV were used for analysis232. Action potential parameters from 10 consecutive APs were averaged using a custom written MATLAB program. Cell capacitance was measured and compensated for using the MultiClamp auto whole cell compensation. For current recordings, only recordings with an access resistance <20 MΩ were included in the analysis. If was 76 elicited with voltage steps of 1.5 seconds ranging between -130 mV and +50 mV from a holding potential of -40 mV. IKAdo was elicited by a 7 second voltage ramp from -120 to

0 mV from a holding potential of -40 mV. Data was collected with a low noise data acquisition system Digidata 1440A (Molecular devices, Sunnyvale, CA), Clampex software and a Multiclamp 700B amplifier (Axon Instruments, Sunnyvale, CA).

Chemicals

The If inhibitor, ZD7288 and GIRK blocker, Tertiapin Q were purchased from

Tocris (Ellisville, MI). Adenosine, internal standard 2-chloroadenosine, and the selective

A1AR receptor antagonist DPCPX, were purchased from Sigma Aldrich (St Louis, MO).

Statistical analysis

Cellular electrophysiology data were analyzed using Clampfit 10.3 software

(Axon Instruments, Sunnyvale, CA). MATLAB (Mathworks, Natick, MA), GraphPad

Prism (GraphPad, La Jolla CA) and Origin 9.0 software (OriginLab, Northampton, MA,

USA). For action potentials and currents with adenosine, two-way repeated measured

ANOVA was used to analyze differences between and within groups, with post hoc least significant difference testing. One sample Student’s t-test was used to compare effects at baseline between HF and CTL. Effects of A1R or GIRK antagonists with adenosine were compared vs. adenosine alone with one way ANOVA. Non-parametric data were analyzed via the Mann-Whitney test. All data are presented as mean ± SE (or SD for PCR quantification) and p<0.05 was the a priori criterion for statistical significance for all comparisons.

Results:

Right atrial cardiac adenosine levels are lower in HF and are associated with upregulation of SAN A1R and ENT mRNA

We first determined whether or not right atrial adenosine cardiac tissue content was increased in HF relative to control. Figure 1A shows the right atrial appendage, which was snap frozen and removed in vivo prior to heart extraction, in relation to the

SAN. Chromatographs and bar graphs in Figure 1B and 1C illustrate the ~3-fold reduction in adenosine tissue content in HF versus control. We next investigated whether the lower adenosine levels observed in HF were associated with compensatory changes in

A1R and ENTs. To do this, we measured the gene expression of these proteins in the

SAN in both HF and control. Figure 1D demonstrates that the expression of ADORA1

(A1R) was increased almost 3-fold and in Figure 1E, all four equilabrative transporters

(ENTs) are increased 1.5 to 2 fold (p<0.05 vs CTL)

HF at baseline results in a slower intrinsic firing rate of SAN cells and a decrease in pacemaker current, If

Figure 2A illustrates the representative morphology of two types of SAN cells observed in this study. Cellular capacitance was 50.5 ± 2.8 pA/pF in control cells (n=46) and 52.5 ± 2.9 pA/pF in HF (n=29, p=NS vs controls). Representative action potentials recorded from a HF and a control cell in Figure 2B depicts the slower intrinsic firing rate

78 of a HF cell with a reduction of SDD. Action potential parameters are summarized in

Figure 2C. Notably, with the exception of significant reductions in SDD and intrinsic firing rate, all other action potential characteristics were unchanged between HF and controls. Representative If tracings are presented in Figure 2D. The If inhibitor, ZD7288 was used to confirm the presence of funny current, If. The current-voltage (I-V) curve in

Figure 2E demonstrates a clear reduction of inward current in HF cells versus controls.

Importantly, these reductions occurred at voltages around the MDP (Figure 2F).

HF increases SAN cell sensitivity to the negative chronotropic actions of adenosine

To determine if reduced adenosine concentrations and increased gene expression of A1Rs resulted in increased functional response to adenosine, we exposed both HF and control SAN cells to 1 and 10 μM adenosine. Adenosine reversibly slowed the intrinsic firing rate in both HF and control cells at both concentrations (Figure 3A). Figure 3B and

Figure 3C display representative action potentials of control and HF SAN cells exposed to 1 μM adenosine. Compared to controls, HF cells demonstrated an increased negative chronotropic response to 1 μM adenosine, with increased hyperpolarization and greater reduction of SDD. Adenosine-induced action potential parameter changes are shown in

Figure 3D. Compared to baseline, the spontaneous rate of control SAN cells were reduced 16.8±3.4 bpm with 1 μM adenosine compared to 43.5±10.1 bpm in HF cells

(p<0.05 vs CTL). In HF, this rate slowing was associated with similar increases in MDP hyperpolarization as well as SDD reduction relative to controls. In the CTL group, cells that hyperpolarized more than 0.75 mV (n=7) had an average reduction of 24.3±5.2 bpm 79

(Figure 4A). Exposure to 10 uM adenosine caused further rate slowing in both HF and controls, although the response was similar in both groups. Furthermore, there was no further MDP hyperpolarization compared to 1 μM in either group.SDD reduction between groups and adenosine concentration showed a statistically significant between factor-interactions. Thus, within each group there was a significant adenosine dose effect, and at 1 μM the SDD reduction is greater in HF than control – indicating greater sensitivity. However, there was no difference in SDD reduction between groups at the higher concentration, and the net change between 1 μM and 10 μM adenosine is significantly larger in control versus HF. Taken together, the SDD changes observed are likely to reach a “saturated” effect at the 10 μM dose. Interestingly, maximum conduction velocity (dV/dtmax) and overshoot were initially increased from baseline during 1 μM adenosine exposure, but both parameters were significantly decreased from baseline during 10 μM adenosine exposure.

SDD and MDP correlate with adenosine-induced negative chronotopic effects in CTL and HF SAN cells, respectively

We next determined which APD parameters correlated with adenosine-induced rate slowing in each group. Figure 4A illustrates SDD, MDP, and overshoot reduction (x- axis) plotted against firing rate reduction (y-axis) in both CTL and HF SAN cells exposed to 1 μM adenosine. In CTL cells, but not HF, SDD reduction was the only determinant to demonstrate a positive correlation with firing rate (p=0.051). MDP reduction was not

80 significantly correlated with rate in CTL cells, but demonstrated a moderate-to-high positive correlation (r=0.68 1μM ADO, r=0.79 10 μM ADO) in HF cells (p<0.05). These effects were also observed at the 10 μM adenosine dose (Figure 4B). Overshoot reduction failed to correlate with rate reduction in either group at either dose. in HF, dV/dtmax showed a moderate (r=0.65, p<0.05) positive correlation to intrinsic firing rate at 1 μM adenosine. However, at 10 μM adenosine, where dV/dtmax was depressed, there was no correlation (r=0.18, p=0.58). Given the interaction of rate with APD, APD50 and APD90 were not evaluated in the context of adenosine-induced negative chronotropy.

Adenosine’s negative chronotropic effects on SAN cells in both control and HF are mediated by A1R

To evaluate the role of A1Rs in the negative chronotropic effects of adenosine, we then exposed SAN cells to DPCPX (an A1R-selective antagonist) prior to exposing cells to 1 and 10 µM adenosine. Representative action potentials in Figure 5A and 5B illustrate that A1R antagonism prevented adenosine-induced rate slowing in both controls and HF, respectively. DPCPX prevented (Figure 5C) adenosine-induced changes in action potential parameters. In control cells exposed to 10 µM adenosine, overshoot and dV/dtmax reductions were significantly reduced with coadministration of DPCPX, suggesting that A1Rs mediate the effects as well.

HF augments adenosine-induced reduction of If at physiological voltages

We next evaluated the role of If in the adenosine mediated effects in HF. We measured gene expression of HCN2 and HCN4, with the results displayed Figure 6A. No difference was observed in gene expression in HCN2 isoform, however HCN4 increased by 1.5 fold in HF. Raw traces of If recordings from SAN cell at physiologic diastolic voltages and I-V curves in control and HF SAN cells are illustrated in Figure 6B and 6C, respectively. Adenosine exhibited dose-dependent If inhibition in both control and HF, with HF demonstrating sensitivity to If block at lower adenosine concentrations (Figure

6D).

HF augments GIRK current, IKAdo response to adenosine

Figure 7A illustrates similar HF-induced increases in gene expression of the two major GIRK subunits in the SAN, Kir3.1 and Kir3.4. To evaluate whether the HF- induced increase in A1R mRNA expression was associated with increased GIRK- mediated IKAdo, we assayed IKAdo in SAN cells via patch clamp. Tertiapin Q (TPQ), a selective GIRK antagonist, was used to block the GIRK-sensitive current induced by adenosine. Figure 7B illustrates the ramp-votage protocol used to elicit IKAdo as well as the raw current traces obtained with 1µM adenosine. As seen in Figure 7C and 7D, the 1

µM adenosine concentration produced a greater peak inward current in HF versus controls, consistent with the HF- associated increase in GIRK gene expression. 10 µM adenosine produced similar IKAdo responses in both HF and CTL.

GIRK inhibition by, Tertiapin Q: increased in HF SAN cells

To see if blocking the increased GIRK-mediated IKAdo prevented negative chronotropy, we measured SAN action potentials in the presence of TPQ. Figure 8A and

8B illustrate action potentials from both control and HF cells in the presence of TPQ and both doses of adenosine. Figure 8C displays action potential characteristic changes between the two groups. In HF, TPQ abolished negative chronotropy when 1 µM adenosine was applied to the cells, and prevented rate slowing by 67% with 10 µM adenosine. In control cells, TPQ only prevented rate slowing at 10 µM, by 67%. In both groups, TPQ prevented rate reduction by preventing hyperpolarization and SDD reduction.

Discussion:

While upregulation of SAN A1Rs in HF has been implicated in the induction of bradyarrythmia140, the effects of adenosine on altered A1R signaling in pacemaker cells from failing hearts have remained poorly defined. Surprisingly, reports of increased circulating levels of adenosine in human HF233,234 are not consistent with the previously observed A1R upregulation in our preclinical HFmodel140, as increased ligand is typically associated with a reduction in receptors. Consistent with this concept, in the present study adenosine effects are increased in HF, and are associated with decreases in right atrial tissue adenosine levels and an upregulation of A1Rs. Through pharmacological inhibition of GIRK channels, we were able to abolish rate slowing at physiological concentrations

83 of adenosine in HF. Thus, the primary effect of adenosine on SAN cells in HF is GIRK- mediated rate slowing via hyperpolarization, likely exacerbated by a reduction of If.

Decreased right atrial cardiac adenosine levels are accompanied by A1R upregulation

Our first major finding was that right atrial appendage adenosine levels were decreased by 3-fold in chronic HF. A similar finding was also observed in mice overexpressing TNFα and exhibiting ventricular dysfunction223. Nonetheless, both results are in stark contrast to both canine volume overload 222 and human HF studies223,234, where adenosine levels are increased in the myocardium and plasma, respectively. One explanation for the disparity may be our chronic HF model is well past the acute decompensation phase of remodeling235, the point in disease progression where accumulation of adenosine in the coronary effluent was previously observed in a pressure-overload hypertrophy rat model. 236. In the same model, adenosine levels returned to baseline after compensation, suggesting that the HF-induced fluctuation of adenosine may be transient.

A second explanation may be that while peripheral plasma adenosine levels are increased in chronic HF, they are not increased in the myocardium. Consistent with this explanation, a study in NYHA III and NYHA IV HF patients demonstrated higher concentrations of peripheral adenosine levels relative to the coronary sinus237.

Furthermore, infusion of dipyridamole (an inhibitor of cellular reuptake of endogenous adenosine primarily produced by the myocardium) was observed to have a blunted effect on the increase of peripheral plasma levels in patients with idiopathic dilated 84 cardiomyopathy compared to non-HF patients238. Taken together, these studies suggest that in HF, myocardial adenosine levels are not only lower relative to the periphery, but myocardial adenosine reserves may be reduced after initial cardiac decompensation.

Indeed, myocardial ATP and total adenine nucleotides have been demonstrated to be reduced in both human end stage heart failure239 and 7 to 9 week extended canine tachypacing HF models240.

We have previously reported that in our HF model that A1Rs are upregulated in the SAN in HF140, corresponding with the present observation that A1R mRNA is increased in HF. Low myocardial adenosine levels were reported to be accompanied by

A1R upregulation in failing hearts223. Notably, adenosine A1R agonism and antagonism reciprocally regulates A1R expression in cerebellar granule cells241, supporting a reciprocal relationship between adenosine signaling and receptor expression. These results suggest a feedback mechanism by which local adenosine levels proximal to the membrane bound A1R receptors play a critical role in A1R regulation. We did not analyze interstitial adenosine levels in this study. Although it remains to be determined if the upregulation of ENT mRNA in HF we observed correlates with SAN cell adenosine transport, increased ENTs at the cell membrane combined with low intracellular adenosine levels would theoretically reduce extracellular adenosine levels242, providing a substrate for A1R upregulation.

SAN HF cells demonstrate reduced intrinsic firing and If

At baseline, HF SAN cells have been previously reported to have reduced intrinsic firing rate109,137,139 and reduced SDD139. The SDD occurs across the diastolic voltage range of the SAN action potential, and in agreement with other studies139, we found the inward pacemaker current If to be reduced at these voltages in HF. While our finding of HCN4 mRNA expression upregulation in HF has been reported in failing SAN rat tissue as well243, it suggests posttranscriptional modulation play a role in reduction of

41 If. While the role and contribution of If in SAN pacemaking is controversial and outside the scope of this study, it should be noted that a 34% reduction of peak If over the physiologic membrane potential range corresponded to a 13% reduction in firing rate in rabbit SAN cells117. We found similar current and firing rate reduction between our two groups at baseline, suggesting If reduction is a plausible mechanism for these observations. Interestingly, transgenic mice overexpressing A1R had slower intrinsic firing rates than controls244, supporting a role for upregulated A1R signaling in baseline

SAN firing rate depression.

Upregulation of A1R and GIRK in HF augments IKAdo and adenosine-mediated If block

The second major finding of this study was that HF increases SAN pacemaker cell sensitivity to the negative chronotropic effects of adenosine. The physiological levels of adenosine in the pericardial fluid is approximately 1 μM245,246 and was used as one of the tested concentrations in this study. To test effects of dose-dependence, as well as mimic the reported increased extracellular adenosine (in some patients up to seven- 86 fold233) we used 10 μM adenosine. In addition to A1R upregulation, GIRK4 protein was upregulated in the SAN our HF model140. An increase in both A1R and GIRK subunit expression in HF should amplify downstream IKAdo and its rate-slowing hyperpolarization effects on the SAN action potential. There was a clear, exaggerated rate slowing response to 1 μM adenosine in HF, with concurrent MDP hyperpolarization that was absent in controls. Furthermore, peak inward IKAdo was increased with 1 μM adenosine in HF, where the blockage of IKAdo by TPQ prevented rate slowing. Although there was an exaggerated firing rate response to 1 μM adenosine in HF versus CTL, the response was only incremental when increased from1 μM to 10 uM. This pattern was also observed in

MDP and SDD action potential parameters, as well as IKAdo. These results suggest a

“saturated” response to the higher concentration of adenosine. Because MDP hyperpolarization was not as predictable at the 10 μM dose in either group, these effects may also reflect desensitization of either GIRK or A1R. Previous studies, though, did not report adenosine-induced desensitization using much higher concentrations of adenosine

227 (50 μM) . Interestingly, IKAdo was similar in both groups when measured with 10 μM adenosine, yet CTL cells tested at this concentration failed to demonstrate any hyperpolarization. For IKAdo to hyperpolarize the MDP, it must produce an outward current over the diastolic range at which inward If is also active. It is possible that CTL cells maintain enough If reserve even in the presence of adenosine that IKAdo effects would be negligible. This explanation may also explain the lack of hyperpolarization seen in normal rabbit SAN cells exposed to similar concentrations of adenosine used in this study. 228

Our results are consistent with previous findings of GIRK inhibition as a strategy to alleviate SAN dysfunction in non-HF models. In mice, CaV1.3 generates an L- type calcium current active during the diastolic depolarization phase. Mice lacking

Cav1.3 demonstrate the phenotype of sick sinus syndrome, including the development of

247 248 - - bradycardia and tachybrady syndrome. Genetic inactivation of GIRK4 in Cav1.3 / prevented N6-cyclopentyl adenosine (CCPA, A1 agonist) induced bradycardia seen in

- - 118 - - Cav1.3 / and wild-type mice. Cav1.3 / may reduce diastolic inward current reserve, allowing for A1 agonist-induced hyperpolarization that can be blocked by GIRK inactivation, a finding consistent with our observations this study.

Clinical Implications

In clinical trials, antagonism of the A1R receptor has been demonstrated to prevent deterioration to HF in patients with sick sinus syndrome249. In addition, blockage of A1R prevented adenosine-induced SAN dysfunction and subsequent development of atrial fibrillation in right atrial preparations from chronic HF dogs140. Unfortunately, the expression and critical role of A1Rs in the central nervous system, makes its global inhibition undesirable.250 Furthermore, adenosine and partial adenosine agonists have been proposed as strategies to attenuate chronic heart failure due to their inhibition of catecholamine release, anti-hyperproliferative effects, and contribution to angiogenesis251. In the present study, we report that blockage of GIRK channels may be an alternative strategy to prevent adenosine-induced SAN dysfunction and/or arrhythmia.

Because GIRK channels are also present in a variety of tissues, targeting only local 88 expression of these channels may not only prevent bradycardia in HF, but also allow for future adenosine therapies to be considered to attenuate HF.

Limitations

At the higher doses of adenosine used in this study, approximately 30% of total rate reduction appears not to be due to IKAdo. This could be due to incomplete block by TPQ.

252 It is also possible that adenosine and/or HF could have effects on ICa,L (Cav1.2 or Cav

248 253 1.3 ), or other aspects of calcium handling . In addition, because adenosine blocks If as well as activate IKAdo, it is difficult to completely discern the contribution of each current to rate slowing. We plan computational modeling experiments to elucidate the role of adenosine on these potential contributors to its observed effects.

Conclusions

In HF, A1R and GIRK upregulation, combined with reduction of If reserve, leads to increased adenosine-induced block of If along with enhanced IKAdo Adenosine-induced hyperpolarization and SAN rate slowing are greater in HF than CTL, and can be prevented by GIRK inhibition. Thus, GIRK blockade may be a useful strategy to mitigate bradycardia in HF.

Figure 3.1 Decreases in right atrial adenosine levels are associated with upregulation of A1R and ENT mRNA

A. Anatomic location of right atrial appendage (RAA) and SAN. B. Representative LC- MS chromatograms of RAA tissue adenosine content in CTL and HF . C. Quantification of RAA adenosine content in CTL and HF D. A1R mRNA expression in CTL and HF. E. ENT mRNA expression in CTL and HF

Figure 3.2 Intrinsic firing and If are decreased in SAN HF cells

A. Morphology of representative SAN cells. B. Representative CTL and HF SAN action potentials C. APD parameters CTL and HF SAN cells D. Representative step voltage If tracings in CTL, CTL + 5 μM ZD7288, and HF E. Current voltage relationship of If in

CTL and HF SAN cells. F. If density recorded at diastolic SAN cell voltages (right)

Figure 3.3 SAN HF cells demonstrate increased sensitivity to negative chronotropic effects of adenosine

A. Adenosine (ADO) effects are reversible upon washout B. Representative action potential SAN cell with 1 μM ADO in CTL and C. HF. D. Action potential parameters changed from baseline (BL) in CTL and HF SAN cells.

Figure 3.4 Examination of AP parameters as contributors to rate changes reveals that SDD and MDP correlate with adenosine-induced intrinsic rate slowing in CTL and HF SAN cells, respectively

A. Scatterplots of select APD parameters (x-axis) versus firing rate reduction (y-axis) after exposure to 1 μM ADO. r= Pearson’s coefficient B. Scatterplots of select APD parameters (x-axis) versus firing rate reduction (y-axis) after exposure to 10 μM ADO.

Figure 3.5 Adenosine's negative chronotropic effects of SAN cells in both CTL and HF are mediated by A1R

Representative SAN action potential in A. CTL and B. HF exposed to ADO in the presence of A1R antagonist, DPCPX. C. Firing rate reduction, D. SDD reduction, and E. MDP reduction from BL compared in presence of ADO DPCPX compared to ADO alone.

Figure 3.6 Adenosine dose-dependently decreases If in CTL and HF SAN cells with greater sensitivity of inhibition observed in HF

A. mRNA expression of pore forming subunit of HCN channels B. Representative tracings of If at -50 mV and -70 mV in CTL and HF in presence of ADO C. Current voltage relationships of If in CTL (left) and HF (right) exposed to ADO. SDD reduction, and D. If density at diastolic SAN voltages in CTL and HF exposed to ADO.

Figure 3.7 Adenosine dose-dependently increases IKAdo in CTL and HF SAN cells with greater sensitivity observed in HF

A. mRNA expression of pore forming subunit of GIRK channels B. Representative tracings of IKAdo CTL and HF at 1 μM ADO C. Current voltage relationships of IKAdo in

CTL and HF at 1μM ADO (left) and 10 μM ADO (right). D. Peak IKAdo density at diastolic SAN voltages in CTL and HF exposed to ADO.

Figure 3.8 GIRK blockade prevents adenosine-induced rate slowing

A. Representative action SAN potential in CTL and B. HF in exposed to ADO in the presence of GIRK antagonist TPQ. C. Firing rate reduction, D. SDD reduction, and E. MDP reduction from BL compared in presence of ADO DPCPX compared to ADO alone.

Chapter 4 : Evaluation of a pharmacist-managed electrolyte protocol to achieve and maintain optimal potassium concentrations in outpatients on antiarrhythmic medications

Introduction

Hypokalemia ([K+]< 3.6 mmol/L254) is frequently encountered in cardiac patients, often occurring as an adverse effect of diuretic therapy255, or increased circulating catecholamines256,257 in settings of high adrenergic tone such as post-myocardial infarction258 or heart failure259,260Monitoring of potassium levels is critical in patients receiving antiarrhythmic therapy since hypokalemia can not only promote arrhythmogenesis through modulation of sodium and potassium currents at the cellular level261-263, but can promote proarrhythmia by adversely altering the effects of antiarrhythmic agents264. In addition, many patients who present with cardiac rhythm disorders have pathophysiologic abnormalities that predispose them to proarrhythmia.190

Therapeutic monitoring of potassium levels is therefore essential to optimize the safety and efficacy of antiarrhythmic therapy.

Currently, there is a lack of consensus on the desired [K+] range in the setting of concomitant antiarrhythmic medication use254,265. The most current recommendations suggest a more conservative potassium target concentration range for patients with underlying heart disease ([K+] ≥4.0 mmol/L)254. The Ohio State University Medical

Center (OSUMC) Antiarrhythmia Medications Clinic has a previously established therapeutic monitoring program run by pharmacists which has been reported to improve the quality of care in patients receiving chronic antiarrhythmic medications, with a net economic benefit to our patients and institution266,267. This clinic has over 500 patients and utilizes a collaborative practice agreement policy as well as privileging of pharmacy

99 staff. The collaborative practice agreement allows pharmacists to review medications for drug interaction review, provide patient education, and manage the ordering and subsequent review of objective testing, In 2009, in collaboration with electrophysiologists, we established a pharmacist-managed potassium electrolyte protocol for clinic patients designed to maintain a [K+] of 4.0 to 5.0 mmol/L. The protocol utilizes dietary advice, as well as magnesium and potassium supplementation to normalize [K+] without making additional changes to a patient’s medication regimen. The objective of this study was to evaluate the effectiveness of our pharmacist-managed monitoring and treatment protocol in maintaining [K+] in the desired concentration range.

Patients and Methods

Inclusion Criteria

The study was approved by the Ohio State University Institutional Review Board and the requirement for informed consent was waived. Enrolled adult outpatients in the

OSUMC Antiarrhythmic Medications Clinic between June 2009 and July 2013 served as the sample population for this study. Patients were referred to the clinic for monitoring at time of hospital discharge or during an outpatient appointment with a cardiologist. Those who had a [K+] level below the desired range (4.0-5.0 mmol/L) at an antiarrhythmic medication clinic appointment were eligible for and included in this study. Patients were excluded if they were prisoners, pregnant, , or were 89 years or older at the time of their appointment. Patients were treated regardless of reversible causes such as potassium- lowering medications and renal function. Patients who were subsequently discontinued 100 from antiarrhythmic medications or had symptomatic atrial fibrillation and transferred to inpatient care were not included in follow up analyses. Enrolled patients seen between

October 2008 and May 2009 (the time period prior to institution of the protocol) served as a control group (CTL).

Study Design

This single-center, retrospective chart review utilized computerized medical records to identify patients with [K+<4.0 mmol/L] from laboratory values obtained at clinic appointments. A patient encounter was defined as the combination of an initial visit to the clinic, subsequent lab draws, and follow up visit. A single patient could have multiple encounters if potassium fell out of the desired range at a later date and was eligible to reenter the protocol. Interventions (therapy recommendations) were made upon lab results, either drawn at OSUMC prior to or during a clinic visit, or weekly thereafter until the patient’s potassium concentration returned to the desired range.Pharmacists called patients to ensure follow-up lab draws and conducted review of lab results. Adherence to the recommended treatment, and arrhythmia symptoms were assessed through patient interview. If present, arrhythmia symptoms were recorded as palpitations, shortness of breath, fatigue, pre-syncope or syncope. After the electrolyte protocol intervention, maintenance of potassium levels was assessed at the next visit to the clinic, scheduled 3 (for dofetilide treated paients) or 6 months after the initial visit.

The primary study endpoints were 1) response to intervention, measured by the percentage of patient encounters achieving the desired potassium levels after intervention and 2) percentage of encounters remaining at the desired potassium level upon follow to 101 the clinic. Secondary endpoints included the number of interventions and time (in days) required to reach desired [K+] levels.

Hypokalemia Protocol

The pharmacist-managed electrolyte algorithm used for this study is illustrated in

Figure 1. The first intervention was correction of [Mg2+] to ≥ 2.0 mmol/L (if warranted) through 400 or 500 mg BID magnesium oxide supplementation along with dietary advice. Furthermore, since [Mg2+] was drawn with [K+], magnesium supplementation could be instituted at any point of the protocol. Dietary advice included increasing intake of potassium-rich foods. A pamphlet, or verbal phone review was used to inform patients of potassium-rich foods for patients to add to their diet, although no target mg of potassium per day was supplied. Patients who initially presented with more severe hypokalemia ([K+] <3.5 mmol/L) were initially started on KCl supplementation, as opposed to dietary modifications alone. All interventions recommended the patients have labs redrawn at1 week intervals until potassium was in the desired range. Patients that remained below the desired [K+] range after potassium chloride (KCl) supplementation reached 40 mEq/day, or that became hyperkalemic (([K+] >5.2 mmol/L)) at any point in the study were referred to their physician for management. Patients were instructed to remain on supplementation indefinitely.

Statistical Analysis

Numeric data was compared between treatment groups via Mann Whitney.

Categorical data was compared via Chi-Square Analysis with Fisher exact test.

102

Results

Table 1 depicts the characteristics of the two patient populations at the initial clinic visit used in this study. There were no significant differences between the two populations. The populations were unbalanced because the protocol was instituted approximately one year after the clinic opened, and there were a limited number of patients that met the inclusion criteria. The median age was 62 years (23-88) and 196 patients (82%) were on a Class III antiarrhythmic. The median [K+] was 3.8 mmol/L

(3.1-3.9) and [Mg2+] was 2.1 mmol/L (1.3-2.5).

In the pre-protocol group, 20 of 41 subsequent encounters (49%) had a desired

[K+] and the median time to a documented target [K+] was 146 days (7-285). In only 5 of the 41 encounters was serum potassium actively managed, and treatment strategies were generally comparable to the protocol, although 1 patient had their loop diuretic discontinued. Of the 8 patients experiencing arrhythmia symptoms with [K+<4.0 mmol/L], 3 reported resolution of symptoms upon potassium correction (38%). In the post-protocol group, all encounters were actively managed. 191 encounters (152 patients) were evaluated, and161 of 191 encounters (84.2%, p<0.01 vs CTL) resulted in target [K+] with a median time for this group to reach target was 14 days (7-203). The majority of protocol managed patients (80%) reached the desired [K+] with one intervention, resulting in a median time of 12 days (3-203) after the first visit. Dietary modification resulted in 49.7% of total successful interventions, KCl supplementation in 38.5%, with magnesium supplementation combined with diet (5.6%) or KCl (5.6%) accounting for the remainder. Of the 124 encounters that reached the desired [K+] with intervention and

103 returned for a subsequent up visit, 76% remained at the target level. The median time for follow up clinic visit was 161 days (18-161)) after last intervention. A breakdown of interventions revealed 40% of patients were maintained with dietary modifications alone,

30% with KCl supplementation, and 6% with magnesium oxide supplementation alone.

Of the 29 patients who had arrhythmia symptoms in the post-protocol group, 13 reported symptom resolution upon reaching target [K+], (44%, p>0.05, n.s. vs CTL). There were no instances of treatment-induced hyperkalemia in either group.

Discussion

Our results demonstrate the electrolyte protocol corrected [K+] levels more rapidly than the standard of care (control), and maintained those levels in 76% of the patients seen at follow up. In addition, there were no reported incidents of hyperkalemia and only 1 patient discontinued the protocol due to an adverse effect of therapy. The reduction of patient-reported arrhythmia symptoms upon reaching the desired [K+] in both groups were approximately 40%, confirming the importance of electrolyte management in this population. This study presents a novel algorithm that can be utilized by pharmacists to achieve and maintain [K+] in a population at high risk for serious serum potassium-related events.

Patients within the protocol were brought to corrected potassium by either magnesium supplementation, diet, or KCl supplementation. Loss of [Mg2+] can impair proper function of renal membrane Na+-K+ ATPase and increase distal renal [K+] secretion18, both of which contribute to [K+] depletion. Correction of [Mg2+], therefore

104 was the first intervention in the protocol. Approximately 10% of patients in the study had

[K+] corrected to desired levels through magnesium supplementation with either diet advice or in addition to KCl supplementation. In the presence of normal magnesium, dietary advice was recommended prior to initiation of KCl supplementation, a strategy to avoid prescription treatment if possible. Dietary advice alone proved enough to correct potassium in almost 50% of the post-protocol encounters, with KCl supplementation effective in correcting 40% of encounters. Although the argument can be made that the median [K+] of patients at the post-protocol evaluation was in close proximity to our target and may have demonstrated correction with repeat testing (without need for intervention), the pre-protocol group had similar ranges without interventions and had a much lower success rate. The potential for adverse effects of KCl supplementation was a concern in this study, but only 1 patient had to discontinue due to intolerable side effects, demonstrating the safety of the protocol. Importantly, the protocol was designed to manage potassium levels without interruption of a patient’s current medication regimen.

This is of particular concern in a population where almost one third of the patients are taking PPI’s or ACE inhibitors (TABLE 1).

Previous guidelines recommend maintaining serum potassium levels at or above

4.0 mmol/L for patients with concomitant organic heart disease254. This population is prone to hypokalemia due to 1) increased sodium retention/potassium excretion by activation of the renin-angiotensin-aldosterone system 2) increased catecholamines driving potassium into cells and 3) management of the condition with diuretics. Older patients (>60 yrs) with structural heart abnormalities such as myocardial infarction or

105 heart failure, who are also managed with potassium-depleting diuretics, appear particularly prone to hypokalemia-induced ventricular arrhythmia261. We previously reported that approximately 45% of our clinic patients have cardiomyopathy or congestive heart failure266, which is consistent with the 30% seen in this study - thus justifying the use of a more conservative [K+] target for the developed protocol vs. the usual normal range of [K+}

It has been estimated that by 2030, there will be 12.1 million patients in the U.S. diagnosed with atrial fibrillation (AF), with an annual increase in incidence of 4.6%.268

Antiarrhythmic drugs remain first line therapy for maintaining sinus rhythm in patients with AF.269 A survey of electrophysiologists revealed that there is an anticipated increase in workload primarily driven by an aging population and increased patient access to healthcare.270 As a result, there exists both a need and opportunity for other healthcare providers to alleviate physician burden and optimize physician time for patient care and procedures. In addition to demonstrating a clinical benefit to patients, this study presented a realistic example of how pharmacists can assume nontraditional roles to improve healthcare delivery.

Limitations

Outpatient adherence to interventions could not be assessed and quantified, but rather depended on patient interview at visit or telephone conversation. Outpatient dropout likely impacted the assessment of the pre-protocol effectiveness as well as the post-protocol [K+] maintenance analysis. Both may be attributed to the long time between 106 lab draws. There was a limited pool of pre-protocol patients, primarily because these were among the first patients of the clinic and the electrolyte protocol was instituted one year later. The pre-protocol period did not include scheduled lab draws and this likely overestimated the time to reach the desired potassium target. Finally, arrhythmias are stochastic events, may not always be symptomatic, and may also be confounded by the presence of other conditions (i.e. heart failure), limiting the reliability of patient-reported symptoms as a surrogate of arrhythmia frequency.

Conclusions

A pharmacist-managed electrolyte protocol, implemented as part of a comprehensive antiarrhythmic monitoring service, effectively achieves and maintains desired [K+] concentrations.

107

Figure 4.1Pharmacist-managed potassium electrolyte algorithm

108

Pre-Protocol group Post-Protocol p value (n=45 patients) (n=196 patients) Demographics Age (years), median(range) 62 (23-82) 62.5 (23-88) n.s. Sex (male) 25 (56%) 112 (62%) n.s. Medical History Cardiomyopathy 12 (27%) 56 (29%) n.s. Previous ablation 16 (35.6%) 89 (45%) n.s. Previous cardioversion 15 (33.3%) 88 (45%) n.s Implanted device 21 (47%) 91 (46%) n.s. Arrhythmia symptoms 22 (49%) 74 (61%) n.s. present? Drug Treatment Class I antiarrhythmic 5 (11%) 45(22%) n.s. Class III antiarrhythmic 41 (91%) 157(80%) n.s. Diuretic (loop) 13 (29%) 58 (30%) n.s. Diuretic (thiazide) 8 (18%) 37 (19%) n.s. Potassium-sparing diuretic 5 (11%) 11 (6%) n.s. PPI 10 (29%) 70 (36%) n.s. ACE inhibitor or ARB 21 (46%) 66 (34%) n.s. Serum electrolyte concentration Potassium (mmol/L), median 3.8 (3.3-3.9) 3.8 (3.1-3.9) n.s. (range) Magnesium (mmol/L), median 2.1 (1.5-2.4) 2.1(1.3-2.5) n.s. (range) Table 4.1 Patient demographics and medical history

109

Figure 4.2 Flow chart of study

110

Chapter 5 : Summary and Conclusions

The studies presented in this dissertation mainly focus on alterations of K+ currents as they pertain to cardiac action potential abnormalities in HF. In patch clamping of isolated myocytes, cellular cardiac action potentials are taken as surrogates of in vivo behavior. In the ventricle, K+ current downregulation leads to prolonged action potentials and increased susceptibility to cellular arrhythmia (EADs). In the sinoatrial node, downregulation of diastolic K+ carrying currents can reduce firing frequency (an indicator of SAN dysfunction). K+ current blockade, however, is sometimes warranted, particularly in patients taking antiarrhythmic medications for indications such as atrial fibrillation.

Endogenous mediators and electrolytes can have a profound effect on ion channel function, particularly in patients taking antiarrhythmic medications. Combined, the studies presented in this dissertation not only provide novel insights into the development of a substrate for arrhythmogenesis in HF, but also into the management of patients with arrhythmias susceptible to proarrhythmia.

The second chapter of this dissertation emphasizes the critical role of HF duration as it pertains to structural and electrical remodeling the ventricle. It is important because short-term (2 to 4 week) models of canine tachypacing are often employed to produce

HF.98,183 While these models are economical and have the benefit of being reversible, 111 they lead to often inconsistent conclusions regarding the roles of HF-induced ventricular

K+ current downregulation92 and fibrosis170-172. For instance, in human end stage HF, the primary contributor of repolarization reserve, IKr has been demonstrated to be

99 downregulated in the ventricle. Furthermore, in human ventricle, blockage of IK1 and

189 IKs appear to have action potential prolonging effects only in the presence of IKr block.

These results are not consistent with patch clamping experiments in short-term tachypacing models demonstrating no change in IKr, which at the same time have a high frequency of EADs.98 These short-term models also fail to consistently demonstrate fibrosis, yet cardiac fibrosis is a hallmark of end stage human HF. Our chronic (4 month tachypacing) HF model overcomes both limitations of the short-term pacing models, and emulates chronic human HF. Our data demonstrate that all repolarizing currents are downregulated by the end four months, providing a logical explanation for action potential prolongation and EAD development. In addition, fibrosis is progressive, and at the end of pacing we demonstrated a 2 -fold increase in ventricular interstitial fibrosis versus normal tissue – consistent with human studies 271

SAN dysfunction is an important contributor to bradyarrhythmia and death in HF, but the mechanisms accounting for its presence remain to be elucidated. In Chapter 3, we examine effects of adenosine as a possible mediator of SAN dysfunction in cells isolated from the SAN. Adenosine produces exaggerated rate-slowing in HF, and appears to be due to upregulation of both A1R receptors and GIRK signaling140. Enhanced A1R signaling in HF decreases If and increase IKAdo, resulting in hyperpolarization and rate slowing. These effects were abolished by A1R blockade, but also GIRK blockade. GIRK

112 blockade may provide a useful strategy in mitigating adenosine-induced bradycardia, especially as there exists current interest in using adenosine or partial adenosine agonists to improve ventricular function.

The final chapter pertains to the management of [K+] in patients taking

262 antiarrhythmic medications. Low potassium levels have can have inhibit IKr , which is of particular concern in patients taking antiarrhythmic medications that may be blocking

K+ currents. Furthermore, patients prone to arrhythmias often have HF, where the repolarization reserve has been demonstrated to be diminished in Chapter 2. The management of potassium is therefore critical in preventing proarrhythmia in this population. In Chapter 4, we evaluated the performance of a pharmacist-managed electrolyte protocol for potassium management and compared it to standard of therapy

(pre-initiation of protocol). We demonstrated that potassium levels were corrected to desired levels more rapidly than standard of care and these levels were maintained in

75% of patients. Furthermore, management of potassium led to arrhythmia symptom reduction, there were no reports of hyperkalemia, and only 1 adverse effect was reported in the study.

In summary, HF and HF-induced arrhythmia will remain a leading cause of mortality in the U.S. for the foreseeable future. Strategies to both predict and prevent sudden death in heart failure remain inadequate.272 This is due in part to limitations of the animal models used in experimental studies.169 In this dissertation, we presented a chronic model of HF that demonstrates the electrical remodeling seen in K+ carrying currents of the ventricle and the SAN in human heart failure. We then used our model to

113 identify the upregulation of an A1R-GIRK signaling pathway that promotes negative chronotropic responses in SAN cells mediated by alterations in If and IKAdo. Patch clamping studies in isolated cells from our models have several limitations, the greatest of which is due to transmural and central/peripheral action potential heterogeneity existing in the ventricle and SAN tissue, respectively. Despite these limitations, our work provides powerful insights into the how alterations of K+ currents can generate a substrate for arrhythmia in both the ventricle and SAN. We also addressed how the use of pharmacists can improve health care delivery to minimize arrhythmias in high-risk patients. While the last study was limited by its retrospective nature, it provides a novel algorithm for potassium management in the setting of antiarrhythmic medication use and offers a potential solution for easing the burden of electrophysiologists as the incidence of arrhythmia continues to rise.

114

Reference List

(1) Osborn HF. Men of the old stone age their environment, life and art. 2nd ed ed. New York: C. Scribner; 1916.

(2) Sigerist HE. A history of medicine. 1951;no. 27 [etc.].

(3) Nunn JF. Ancient Egyptian medicine. Norman: University of Oklahoma Press; 1996.

(4) FIELDS A. The pulse in ancient Chinese medicine. Calif Med 1947;66:304.

(5) Wang YY, Wang SH, Jan MY, Wang WK. Past, Present, and Future of the Pulse Examination ( mai zhen). J Tradit Complement Med 2012;2:164-185.

(6) ATKINSON MH. MAN'S CHANGING CONCEPTS OF THE HEART AND CIRCULATION. Can Med Assoc J 1964;91:596-601.

(7) Crivellato E, Ribatti D. A portrait of Aristotle as an anatomist: historical article. Clin Anat 2007;20:447-485.

(8) Van PR, Van PS. Aristotle's "triventricular" heart and the relevant early history of the cardiovascular system. Chest 1983;84:462-468.

(9) Bestetti RB, Restini CB, Couto LB. Development of anatomophysiologic knowledge regarding the cardiovascular system: from Egyptians to Harvey. Arq Bras Cardiol 2014;103:538-545.

(10) Bay NS, Bay BH. Greek anatomist herophilus: the father of anatomy. Anat Cell Biol 2010;43:280-283.

(11) Ghasemzadeh N, Zafari AM. A brief journey into the history of the arterial pulse. Cardiol Res Pract 2011;2011:164832.

(12) ATKINSON MH. MAN'S CHANGING CONCEPTS OF THE HEART AND CIRCULATION. Can Med Assoc J 1964;91:596-601.

(13) Galen, May MT. Galen on the usefulness of the parts of the body Peri chreias morion. (romanized form) De usu partium. Ithaca, N.Y: Cornell University Press; 1968.

115

(14) Galvani L. Commentary on the effects of electricity on muscular motion. Norwalk, Conn: Burndy Library; 1953.

(15) Breathnach CS, Westphal W. Early detectors of the heart's electrical activity. Pacing Clin Electrophysiol 2006;29:422-424.

(16) Rivera-Ruiz M, Cajavilca C, Varon J. Einthoven's string galvanometer: the first electrocardiograph. Tex Heart Inst J 2008;35:174-178.

(17) Sykes AH. A D Waller and the electrocardiogram, 1887. Br Med J (Clin Res Ed) 1987;294:1396-1398.

(18) Silverman ME, Grove D, Upshaw CB, Jr. Why does the heart beat? The discovery of the electrical system of the heart. Circulation 2006;113:2775-2781.

(19) HIS W, Jr. The activity of the embryonic human heart and its significance of the understanding of the heart movement in the adult. J Hist Med Allied Sci 1949;4:289- 318.

(20) Keith A, Flack M. The Form and Nature of the Muscular Connections between the Primary Divisions of the Vertebrate Heart. J Anat Physiol 1907;41:172-189.

(21) Hodgkin AL, Huxley AF. Resting and action potentials in single nerve fibres. J Physiol 1945;104:176-195.

(22) HODGKIN AL, KATZ B. The effect of sodium ions on the electrical activity of giant axon of the squid. J Physiol 1949;108:37-77.

(23) Hodgkin AL, Huxley AF. Currents carried by sodium and potassium ions through the membrane of the giant axon of Loligo. J Physiol 1952;116:449-472.

(24) Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol 1952;117:500-544.

(25) Coraboeuf E, Weidmann S. *Potentiel de Repos et Potentiels Daction du Muscle Cardiaque, Mesures A Laide Delectrodes Internes. Comptes Rendus des Seances de la Societe de Biologie et de Ses Filiales 1949;143:1329-1331.

(26) Surawicz B. Brief history of cardiac arrhythmias since the end of the nineteenth century: part I. J Cardiovasc Electrophysiol 2003;14:1365-1371.

(27) Weidmann S. The effect of the cardiac membrane potential on the rapid availability of the sodium-carrying system. J Physiol 1955;127:213-224.

(28) WEIDMANN S. Effects of calcium ions and local anesthetics on electrical properties of Purkinje fibres. J Physiol 1955;129:568-582.

116

(29) Weidmann S. Effect of current flow on the membrane potential of cardiac muscle. J Physiol 1951;115:227-236.

(30) Weidmann S. The diffusion of radiopotassium across intercalated disks of mammalian cardiac muscle. J Physiol 1966;187:323-342.

(31) WEST TC. Ultramicroelectrode recording from the cardiac pacemaker. J Pharmacol Exp Ther 1955;115:283-290.

(32) HOFFMAN BF, SUCKLING EE. Cellular potentials of intact mammalian hearts. Am J Physiol 1952;170:357-362.

(33) Cranefield PF, Hoffman BF, DE CARVALHO AP. Effects of acetylcholine on single fibers of the atrioventricular node. Circ Res 1959;7:19-23.

(34) ALANIS J, GONZALEZ H, LOPEZ E. The electrical activity of the bundle of His. J Physiol 1958;142:127-140.

(35) Nabauer M, Beuckelmann DJ, Uberfuhr P, Steinbeck G. Regional differences in current density and rate-dependent properties of the transient outward current in subepicardial and subendocardial myocytes of human left ventricle. Circulation 1996;93:168-177.

(36) Antzelevitch C, Sicouri S, Litovsky SH, Lukas A, Krishnan SC, Di Diego JM et al. Heterogeneity within the ventricular wall. Electrophysiology and pharmacology of epicardial, endocardial, and M cells. Circ Res 1991;69:1427-1449.

(37) Sah R, Ramirez RJ, Backx PH. Modulation of Ca(2+) release in cardiac myocytes by changes in repolarization rate: role of phase-1 action potential repolarization in excitation-contraction coupling. Circ Res 2002;90:165-173.

(38) Surawicz B, Gettes LS, Ponce-Zumino A. Relation of vulnerability to ECG and action potential characteristics of premature beats. Am J Physiol 1967;212:1519-1528.

(39) Cubeddu LX. QT prolongation and fatal arrhythmias: a review of clinical implications and effects of drugs. Am J Ther 2003;10:452-457.

(40) Guo J, Mitsuiye T, Noma A. The sustained inward current in sino-atrial node cells of guinea-pig heart. Pflugers Arch 1997;433:390-396.

(41) Lakatta EG, DiFrancesco D. What keeps us ticking: a funny current, a calcium clock, or both? J Mol Cell Cardiol 2009;47:157-170.

(42) Lakatta EG, Maltsev VA, Vinogradova TM. A coupled SYSTEM of intracellular Ca2+ clocks and surface membrane voltage clocks controls the timekeeping mechanism of the heart's pacemaker. Circ Res 2010;106:659-673.

117

(43) Huser J, Blatter LA, Lipsius SL. Intracellular Ca2+ release contributes to automaticity in cat atrial pacemaker cells. J Physiol 2000;524 Pt 2:415-422.

(44) Wermuth CG. The practice of medicinal chemistry. Amsterdam: Elsevier; 2015.

(45) Doyle DA, Morais CJ, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL et al. The structure of the potassium channel: molecular basis of K+ conduction and selectivity. Science 1998;280:69-77.

(46) Heginbotham L, Lu Z, Abramson T, MacKinnon R. Mutations in the K+ channel signature sequence. Biophys J 1994;66:1061-1067.

(47) Nernst W, Lehfeldt RA. Theoretical chemistry from the standpoint of Avogadro's rule & thermodynamics. Rev. in accordance with the 4th German ed ed. London: Macmillan; 1904.

(48) Hille B. Ion channels of excitable membranes. 3rd ed ed. Sunderland, Mass: Sinauer; 2001.

(49) Perozo E, Cortes DM, Cuello LG. Structural rearrangements underlying K+-channel activation gating. Science 1999;285:73-78.

(50) Guy HR, Seetharamulu P. Molecular model of the action potential sodium channel. Proc Natl Acad Sci U S A 1986;83:508-512.

(51) Jerng HH, Shahidullah M, Covarrubias M. Inactivation gating of Kv4 potassium channels: molecular interactions involving the inner vestibule of the pore. J Gen Physiol 1999;113:641-660.

(52) Wang S, Liu S, Morales MJ, Strauss HC, Rasmusson RL. A quantitative analysis of the activation and inactivation kinetics of HERG expressed in Xenopus oocytes. J Physiol 1997;502 ( Pt 1):45-60.

(53) Wang S, Morales MJ, Liu S, Strauss HC, Rasmusson RL. Time, voltage and ionic concentration dependence of rectification of h-erg expressed in Xenopus oocytes. FEBS Lett 1996;389:167-173.

(54) Hoshi T, Zagotta WN, Aldrich RW. Two types of inactivation in Shaker K+ channels: effects of alterations in the carboxy-terminal region. Neuron 1991;7:547-556.

(55) Rasmusson RL, Morales MJ, Wang S, Liu S, Campbell DL, Brahmajothi MV et al. Inactivation of voltage-gated cardiac K+ channels. Circ Res 1998;82:739-750.

(56) Kubo Y, Baldwin TJ, Jan YN, Jan LY. Primary structure and functional expression of a mouse inward rectifier potassium channel. Nature 1993;362:127-133.

118

(57) Plaster NM, Tawil R, Tristani-Firouzi M, Canun S, Bendahhou S, Tsunoda A et al. Mutations in Kir2.1 cause the developmental and episodic electrical phenotypes of Andersen's syndrome. Cell 2001;105:511-519.

(58) Pegan S, Arrabit C, Zhou W, Kwiatkowski W, Collins A, Slesinger PA et al. Cytoplasmic domain structures of Kir2.1 and Kir3.1 show sites for modulating gating and rectification. Nat Neurosci 2005;8:279-287.

(59) Lopatin AN, Makhina EN, Nichols CG. Potassium channel block by cytoplasmic polyamines as the mechanism of intrinsic rectification. Nature 1994;372:366-369.

(60) Matsuda H, Saigusa A, Irisawa H. Ohmic conductance through the inwardly rectifying K channel and blocking by internal Mg2+. Nature 1987;325:156-159.

(61) Jost N, Virag L, Comtois P, Ordog B, Szuts V, Seprenyi G et al. Ionic mechanisms limiting cardiac repolarization reserve in humans compared to dogs. J Physiol 2013;591:4189-4206.

(62) Nerbonne JM, Kass RS. Molecular physiology of cardiac repolarization. Physiol Rev 2005;85:1205-1253.

(63) Wang S, Bondarenko VE, Qu YJ, Bett GC, Morales MJ, Rasmusson RL et al. Time- and voltage-dependent components of Kv4.3 inactivation. Biophys J 2005;89:3026-3041.

(64) Jiang X, Bett GC, Li X, Bondarenko VE, Rasmusson RL. C-type inactivation involves a significant decrease in the intracellular aqueous pore volume of Kv1.4 K+ channels expressed in Xenopus oocytes. J Physiol 2003;549:683-695.

(65) Xu H, Guo W, Nerbonne JM. Four kinetically distinct depolarization-activated K+ currents in adult mouse ventricular myocytes. J Gen Physiol 1999;113:661-678.

(66) Guo W, Li H, Aimond F, Johns DC, Rhodes KJ, Trimmer JS et al. Role of heteromultimers in the generation of myocardial transient outward K+ currents. Circ Res 2002;90:586-593.

(67) Jerng HH, Covarrubias M. K+ channel inactivation mediated by the concerted action of the cytoplasmic N- and C-terminal domains. Biophys J 1997;72:163-174.

(68) Hatano N, Ohya S, Muraki K, Clark RB, Giles WR, Imaizumi Y. Two arginines in the cytoplasmic C-terminal domain are essential for voltage-dependent regulation of A- type K+ current in the Kv4 channel subfamily. J Biol Chem 2004;279:5450-5459.

(69) Baukrowitz T, Yellen G. Modulation of K+ current by frequency and external [K+]: a tale of two inactivation mechanisms. Neuron 1995;15:951-960.

119

(70) Radicke S, Cotella D, Graf EM, Ravens U, Wettwer E. Expression and function of dipeptidyl-aminopeptidase-like protein 6 as a putative beta-subunit of human cardiac transient outward current encoded by Kv4.3. J Physiol 2005;565:751-756.

(71) An WF, Bowlby MR, Betty M, Cao J, Ling HP, Mendoza G et al. Modulation of A-type potassium channels by a family of calcium sensors. Nature 2000;403:553-556.

(72) Radicke S, Cotella D, Sblattero D, Ravens U, Santoro C, Wettwer E. The transmembrane beta-subunits KCNE1, KCNE2, and DPP6 modify pharmacological effects of the antiarrhythmic agent tedisamil on the transient outward current Ito. Naunyn Schmiedebergs Arch Pharmacol 2009;379:617-626.

(73) Xiao L, Koopmann TT, Ordog B, Postema PG, Verkerk AO, Iyer V et al. Unique cardiac Purkinje fiber transient outward current beta-subunit composition: a potential molecular link to idiopathic ventricular fibrillation. Circ Res 2013;112:1310-1322.

(74) Dixon JE, Shi W, Wang HS, McDonald C, Yu H, Wymore RS et al. Role of the Kv4.3 K+ channel in ventricular muscle. A molecular correlate for the transient outward current. Circ Res 1996;79:659-668.

(75) Sun X, Wang HS. Role of the transient outward current (Ito) in shaping canine ventricular action potential--a dynamic clamp study. J Physiol 2005;564:411-419.

(76) Hoppe UC, Marban E, Johns DC. Molecular dissection of cardiac repolarization by in vivo Kv4.3 gene transfer. J Clin Invest 2000;105:1077-1084.

(77) Sridhar A, Nishijima Y, Terentyev D, Terentyeva R, Uelmen R, Kukielka M et al. Repolarization abnormalities and afterdepolarizations in a canine model of sudden cardiac death. Am J Physiol Regul Integr Comp Physiol 2008;295:R1463-R1472.

(78) Greenstein JL, Wu R, Po S, Tomaselli GF, Winslow RL. Role of the calcium-independent transient outward current I(to1) in shaping action potential morphology and duration. Circ Res 2000;87:1026-1033.

(79) Sanguinetti MC, Jurkiewicz NK. Two components of cardiac delayed rectifier K+ current. Differential sensitivity to block by class III antiarrhythmic agents. J Gen Physiol 1990;96:195-215.

(80) Spector PS, Curran ME, Zou A, Keating MT, Sanguinetti MC. Fast inactivation causes rectification of the IKr channel. J Gen Physiol 1996;107:611-619.

(81) Ono K, Ito H. Role of rapidly activating delayed rectifier K+ current in sinoatrial node pacemaker activity. Am J Physiol 1995;269:H453-H462.

(82) Smith PL, Baukrowitz T, Yellen G. The inward rectification mechanism of the HERG cardiac potassium channel. Nature 1996;379:833-836.

120

(83) Cui J, Melman Y, Palma E, Fishman GI, McDonald TV. Cyclic AMP regulates the HERG K(+) channel by dual pathways. Curr Biol 2000;10:671-674.

(84) Pourrier M, Zicha S, Ehrlich J, Han W, Nattel S. Canine ventricular KCNE2 expression resides predominantly in Purkinje fibers. Circ Res 2003;93:189-191.

(85) Gintant GA. Two components of delayed rectifier current in canine atrium and ventricle. Does IKs play a role in the reverse rate dependence of class III agents? Circ Res 1996;78:26-37.

(86) Larsen AP, Steffensen AB, Grunnet M, Olesen SP. Extracellular potassium inhibits Kv7.1 potassium channels by stabilizing an inactivated state. Biophys J 2011;101:818- 827.

(87) Sridhar A, da Cunha DN, Lacombe VA, Zhou Q, Fox JJ, Hamlin RL et al. The plateau outward current in canine ventricle, sensitive to 4-aminopyridine, is a constitutive contributor to ventricular repolarization. Br J Pharmacol 2007;152:870-879.

(88) Dobrzynski H, Rothery SM, Marples DD, Coppen SR, Takagishi Y, Honjo H et al. Presence of the Kv1.5 K(+) channel in the sinoatrial node. J Histochem Cytochem 2000;48:769-780.

(89) Wang Z, Fermini B, Nattel S. Sustained depolarization-induced outward current in human atrial myocytes. Evidence for a novel delayed rectifier K+ current similar to Kv1.5 cloned channel currents. Circ Res 1993;73:1061-1076.

(90) Ford J, Milnes J, Wettwer E, Christ T, Rogers M, Sutton K et al. Human electrophysiological and pharmacological properties of XEN-D0101: a novel atrial- selective Kv1.5/IKur inhibitor. J Cardiovasc Pharmacol 2013;61:408-415.

(91) Roden DM. Taking the "idio" out of "idiosyncratic": predicting torsades de pointes. Pacing Clin Electrophysiol 1998;21:1029-1034.

(92) Nattel S, Maguy A, Le BS, Yeh YH. Arrhythmogenic ion-channel remodeling in the heart: heart failure, myocardial infarction, and atrial fibrillation. Physiol Rev 2007;87:425-456.

(93) Cranefield PF, Aronson RS. Torsade de pointes and other pause-induced ventricular tachycardias: the short-long-short sequence and early afterdepolarizations. Pacing Clin Electrophysiol 1988;11:670-678.

(94) Zeng J, Rudy Y. Early afterdepolarizations in cardiac myocytes: mechanism and rate dependence. Biophys J 1995;68:949-964.

(95) Weiss JN, Garfinkel A, Karagueuzian HS, Chen PS, Qu Z. Early afterdepolarizations and cardiac arrhythmias. Heart Rhythm 2010;7:1891-1899.

121

(96) Bonatti V, Rolli A, Botti G. Monophasic action potential studies in human subjects with prolonged ventricular repolarization and long QT syndromes. Eur Heart J 1985;6 Suppl D:131-143.

(97) Nabauer M, Kaab S. Potassium channel down-regulation in heart failure. Cardiovasc Res 1998;37:324-334.

(98) Li GR, Lau CP, Ducharme A, Tardif JC, Nattel S. Transmural action potential and ionic current remodeling in ventricles of failing canine hearts. Am J Physiol Heart Circ Physiol 2002;283:H1031-H1041.

(99) Holzem KM, Gomez JF, Glukhov AV, Madden EJ, Koppel AC, Ewald GA et al. Reduced response to I blockade and altered hERG1a/1b stoichiometry in human heart failure. J Mol Cell Cardiol 2015.

(100) Hoyt RH, Cohen ML, Saffitz JE. Distribution and three-dimensional structure of intercellular junctions in canine myocardium. Circ Res 1989;64:563-574.

(101) Xie Y, Sato D, Garfinkel A, Qu Z, Weiss JN. So little source, so much sink: requirements for afterdepolarizations to propagate in tissue. Biophys J 2010;99:1408-1415.

(102) Akar FG, Yan GX, Antzelevitch C, Rosenbaum DS. Unique topographical distribution of M cells underlies reentrant mechanism of torsade de pointes in the long-QT syndrome. Circulation 2002;105:1247-1253.

(103) Allessie MA, Bonke FI, Schopman FJ. Circus movement in rabbit atrial muscle as a mechanism of tachycardia. III. The "leading circle" concept: a new model of circus movement in cardiac tissue without the involvement of an anatomical obstacle. Circ Res 1977;41:9-18.

(104) Surawicz B. Electrophysiologic substrate of torsade de pointes: dispersion of repolarization or early afterdepolarizations? J Am Coll Cardiol 1989;14:172-184.

(105) Belardinelli L, Antzelevitch C, Vos MA. Assessing predictors of drug-induced torsade de pointes. Trends Pharmacol Sci 2003;24:619-625.

(106) Thomsen MB, Verduyn SC, Stengl M, Beekman JD, de PG, van OJ et al. Increased short- term variability of repolarization predicts d-sotalol-induced torsades de pointes in dogs. Circulation 2004;110:2453-2459.

(107) Kaupp UB, Seifert R. Cyclic nucleotide-gated ion channels. Physiol Rev 2002;82:769- 824.

(108) Wahl-Schott C, Biel M. HCN channels: structure, cellular regulation and physiological function. Cell Mol Life Sci 2009;66:470-494.

122

(109) Zicha S, Fernandez-Velasco M, Lonardo G, L'Heureux N, Nattel S. Sinus node dysfunction and hyperpolarization-activated (HCN) channel subunit remodeling in a canine heart failure model. Cardiovasc Res 2005;66:472-481.

(110) Ludwig A, Zong X, Stieber J, Hullin R, Hofmann F, Biel M. Two pacemaker channels from human heart with profoundly different activation kinetics. EMBO J 1999;18:2323-2329.

(111) Zhang Q, Huang A, Lin YC, Yu HG. Associated changes in HCN2 and HCN4 transcripts and I(f) pacemaker current in myocytes. Biochim Biophys Acta 2009;1788:1138-1147.

(112) Zagotta WN, Olivier NB, Black KD, Young EC, Olson R, Gouaux E. Structural basis for modulation and agonist specificity of HCN pacemaker channels. Nature 2003;425:200- 205.

(113) Craven KB, Zagotta WN. CNG and HCN channels: two peas, one pod. Annu Rev Physiol 2006;68:375-401.

(114) DiFrancesco D, Tortora P. Direct activation of cardiac pacemaker channels by intracellular cyclic AMP. Nature 1991;351:145-147.

(115) DiFrancesco D. A study of the ionic nature of the pace-maker current in calf Purkinje fibres. J Physiol 1981;314:377-393.

(116) Bahring R, Covarrubias M. Mechanisms of closed-state inactivation in voltage-gated ion channels. J Physiol 2011;589:461-479.

(117) Yaniv Y, Maltsev VA, Ziman BD, Spurgeon HA, Lakatta EG. The "funny" current (I(f)) inhibition by ivabradine at membrane potentials encompassing spontaneous depolarization in pacemaker cells. Molecules 2012;17:8241-8254.

(118) Mesirca P, Alig J, Torrente AG, Muller JC, Marger L, Rollin A et al. Cardiac arrhythmia induced by genetic silencing of 'funny' (f) channels is rescued by GIRK4 inactivation. Nat Commun 2014;5:4664.

(119) Wickman K, Krapivinsky G, Corey S, Kennedy M, Nemec J, Medina I et al. Structure, G protein activation, and functional relevance of the cardiac G protein-gated K+ channel, IKACh. Ann N Y Acad Sci 1999;868:386-398.

(120) Bender K, Wellner-Kienitz MC, Inanobe A, Meyer T, Kurachi Y, Pott L. Overexpression of monomeric and multimeric GIRK4 subunits in rat atrial myocytes removes fast desensitization and reduces inward rectification of muscarinic K(+) current (I(K(ACh))). Evidence for functional homomeric GIRK4 channels. J Biol Chem 2001;276:28873- 28880.

123

(121) Whorton MR, MacKinnon R. X-ray structure of the mammalian GIRK2-betagamma G- protein complex. Nature 2013;498:190-197.

(122) Kurachi Y, Nakajima T, Sugimoto T. Short-term desensitization of muscarinic K+ channel current in isolated atrial myocytes and possible role of GTP-binding proteins. Pflugers Arch 1987;410:227-233.

(123) Dobrev D, Carlsson L, Nattel S. Novel molecular targets for atrial fibrillation therapy. Nat Rev Drug Discov 2012;11:275-291.

(124) Belardinelli L, Giles WR, West A. Ionic mechanisms of adenosine actions in pacemaker cells from rabbit heart. J Physiol 1988;405:615-633.

(125) Campbell GD, Edwards FR, Hirst GD, O'Shea JE. Effects of vagal stimulation and applied acetylcholine on pacemaker potentials in the guinea-pig heart. J Physiol 1989;415:57- 68.

(126) Verkerk AO, Geuzebroek GS, Veldkamp MW, Wilders R. Effects of acetylcholine and noradrenalin on action potentials of isolated rabbit sinoatrial and atrial myocytes. Front Physiol 2012;3:174.

(127) Fu Y, Huang X, Zhong H, Mortensen RM, D'Alecy LG, Neubig RR. Endogenous RGS proteins and Galpha subtypes differentially control muscarinic and adenosine- mediated chronotropic effects. Circ Res 2006;98:659-666.

(128) Zaza A, Rocchetti M, DiFrancesco D. Modulation of the hyperpolarization-activated current (I(f)) by adenosine in rabbit sinoatrial myocytes. Circulation 1996;94:734-741.

(129) Lamas GA, Lee K, Sweeney M, Leon A, Yee R, Ellenbogen K et al. The mode selection trial (MOST) in sinus node dysfunction: design, rationale, and baseline characteristics of the first 1000 patients. Am Heart J 2000;140:541-551.

(130) Choudhury M, Boyett MR, Morris GM. Biology of the Sinus Node and its Disease. Arrhythm Electrophysiol Rev 2015;4:28-34.

(131) Mangoni ME, Nargeot J. Genesis and regulation of the heart automaticity. Physiol Rev 2008;88:919-982.

(132) Nof E, Glikson M, Antzelevitch C. Genetics and Sinus Node Dysfunction. J Atr Fibrillation 2009;1:328-336.

(133) Schulze-Bahr E, Neu A, Friederich P, Kaupp UB, Breithardt G, Pongs O et al. Pacemaker channel dysfunction in a patient with sinus node disease. J Clin Invest 2003;111:1537- 1545.

124

(134) Milanesi R, Baruscotti M, Gnecchi-Ruscone T, DiFrancesco D. Familial sinus bradycardia associated with a mutation in the cardiac pacemaker channel. N Engl J Med 2006;354:151-157.

(135) Nof E, Luria D, Brass D, Marek D, Lahat H, Reznik-Wolf H et al. Point mutation in the HCN4 cardiac ion channel pore affecting synthesis, trafficking, and functional expression is associated with familial asymptomatic sinus bradycardia. Circulation 2007;116:463-470.

(136) Ueda K, Nakamura K, Hayashi T, Inagaki N, Takahashi M, Arimura T et al. Functional characterization of a trafficking-defective HCN4 mutation, D553N, associated with cardiac arrhythmia. J Biol Chem 2004;279:27194-27198.

(137) Sanders P, Kistler PM, Morton JB, Spence SJ, Kalman JM. Remodeling of sinus node function in patients with congestive heart failure: reduction in sinus node reserve. Circulation 2004;110:897-903.

(138) Vermeulen JT, Tan HL, Rademaker H, Schumacher CA, Loh P, Opthof T et al. Electrophysiologic and extracellular ionic changes during acute ischemia in failing and normal rabbit myocardium. J Mol Cell Cardiol 1996;28:123-131.

(139) Verkerk AO, Wilders R, Coronel R, Ravesloot JH, Verheijck EE. Ionic remodeling of sinoatrial node cells by heart failure. Circulation 2003;108:760-766.

(140) Lou Q, Hansen BJ, Fedorenko O, Csepe TA, Kalyanasundaram A, Li N et al. Upregulation of adenosine A1 receptors facilitates sinoatrial node dysfunction in chronic canine heart failure by exacerbating nodal conduction abnormalities revealed by novel dual-sided intramural optical mapping. Circulation 2014;130:315-324.

(141) Wickman K, Nemec J, Gendler SJ, Clapham DE. Abnormal heart rate regulation in GIRK4 knockout mice. Neuron 1998;20:103-114.

(142) Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Blaha MJ et al. Heart disease and stroke statistics--2014 update: a report from the American Heart Association. Circulation 2014;129:e28-e292.

(143) Tomaselli GF, Zipes DP. What causes sudden death in heart failure? Circ Res 2004;95:754-763.

(144) Thom T, Haase N, Rosamond W, Howard VJ, Rumsfeld J, Manolio T et al. Heart disease and stroke statistics--2006 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2006;113:e85- 151.

125

(145) Cohn JN, Ferrari R, Sharpe N. Cardiac remodeling--concepts and clinical implications: a consensus paper from an international forum on cardiac remodeling. Behalf of an International Forum on Cardiac Remodeling. J Am Coll Cardiol 2000;35:569-582.

(146) Hsia HH, Marchlinski FE. Electrophysiology studies in patients with dilated cardiomyopathies. Card Electrophysiol Rev 2002;6:472-481.

(147) Iles L, Pfluger H, Lefkovits L, Butler MJ, Kistler PM, Kaye DM et al. Myocardial fibrosis predicts appropriate device therapy in patients with implantable cardioverter- defibrillators for primary prevention of sudden cardiac death. J Am Coll Cardiol 2011;57:821-828.

(148) Haider AW, Larson MG, Benjamin EJ, Levy D. Increased left ventricular mass and hypertrophy are associated with increased risk for sudden death. J Am Coll Cardiol 1998;32:1454-1459.

(149) Beuckelmann DJ, Nabauer M, Kruger C, Erdmann E. Altered diastolic [Ca2+]i handling in human ventricular myocytes from patients with terminal heart failure. Am Heart J 1995;129:684-689.

(150) Akar FG, Rosenbaum DS. Transmural electrophysiological heterogeneities underlying arrhythmogenesis in heart failure. Circ Res 2003;93:638-645.

(151) Nishijima Y, Feldman DS, Bonagura JD, Ozkanlar Y, Jenkins PJ, Lacombe VA et al. Canine nonischemic left ventricular dysfunction: a model of chronic human cardiomyopathy. J Card Fail 2005;11:638-644.

(152) Sridhar A, Nishijima Y, Terentyev D, Khan M, Terentyeva R, Hamlin RL et al. Chronic heart failure and the substrate for atrial fibrillation. Cardiovasc Res 2009;84:227-236.

(153) Nishijima Y, Sridhar A, Viatchenko-Karpinski S, Shaw C, Bonagura JD, Abraham WT et al. Chronic cardiac resynchronization therapy and reverse ventricular remodeling in a model of nonischemic cardiomyopathy. Life Sci 2007;81:1152-1159.

(154) Bonilla IM, Sridhar A, Nishijima Y, Gyorke S, Cardounel AJ, Carnes CA. Differential effects of the peroxynitrite donor, SIN-1, on atrial and ventricular myocyte electrophysiology. J Cardiovasc Pharmacol 2013;61:401-407.

(155) Bonilla IM, Vargas-Pinto P, Nishijima Y, Pedraza-Toscano A, Ho HT, Long VP, III et al. Ibandronate and Ventricular Arrhythmia Risk. J Cardiovasc Electrophysiol 2013.

(156) Nabauer M, Beuckelmann DJ, Erdmann E. Characteristics of transient outward current in human ventricular myocytes from patients with terminal heart failure. Circ Res 1993;73:386-394.

126

(157) Carnes CA, Dech SJ. Effects of dihydrotestosterone on cardiac inward rectifier K(+) current. Int J Androl 2002;25:210-214.

(158) Lopatin AN, Shantz LM, Mackintosh CA, Nichols CG, Pegg AE. Modulation of potassium channels in the hearts of transgenic and mutant mice with altered polyamine biosynthesis. J Mol Cell Cardiol 2000;32:2007-2024.

(159) Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-408.

(160) Kumar S, Pan CC, Bloodworth JC, Nixon A, Theuer C, Hoyt DG et al. Antibody-directed coupling of endoglin and MMP-14 is a key mechanism for endoglin shedding and deregulation of TGF-beta signaling. Oncogene 2013.

(161) Nishijima Y, Sridhar A, Bonilla I, Velayutham M, Khan M, Terentyeva R et al. Tetrahydrobiopterin depletion and NOS2 uncoupling contribute to heart failure- induced alterations in atrial electrophysiology. Cardiovasc Res 2011;91:71-79.

(162) Zweier JL, Kuppusamy P, Williams R, Rayburn BK, Smith D, Weisfeldt ML et al. Measurement and characterization of postischemic free radical generation in the isolated perfused heart. J Biol Chem 1989;264:18890-18895.

(163) Zweier JL, Flaherty JT, Weisfeldt ML. Direct measurement of free radical generation following reperfusion of ischemic myocardium. Proc Natl Acad Sci U S A 1987;84:1404- 1407.

(164) Zweier JL, Wang P, Kuppusamy P. Direct measurement of nitric oxide generation in the ischemic heart using electron paramagnetic resonance spectroscopy. J Biol Chem 1995;270:304-307.

(165) Oosterhoff P, Oros A, Vos MA. Beat-to-beat variability of repolarization: a new parameter to determine arrhythmic risk of an individual or identify proarrhythmic drugs. Anadolu Kardiyol Derg 2007;7 Suppl 1:73-78.

(166) Akar FG, Wu RC, Deschenes I, Armoundas AA, Piacentino V, III, Houser SR et al. Phenotypic differences in transient outward K+ current of human and canine ventricular myocytes: insights into molecular composition of ventricular Ito. Am J Physiol Heart Circ Physiol 2004;286:H602-H609.

(167) Deschenes I, DiSilvestre D, Juang GJ, Wu RC, An WF, Tomaselli GF. Regulation of Kv4.3 current by KChIP2 splice variants: a component of native cardiac I(to)? Circulation 2002;106:423-429.

(168) Radicke S, Cotella D, Graf EM, Ravens U, Wettwer E. Expression and function of dipeptidyl-aminopeptidase-like protein 6 as a putative beta-subunit of human cardiac transient outward current encoded by Kv4.3. J Physiol 2005;565:751-756.

127

(169) Houser SR, Margulies KB, Murphy AM, Spinale FG, Francis GS, Prabhu SD et al. Animal models of heart failure: a scientific statement from the American Heart Association. Circ Res 2012;111:131-150.

(170) Burashnikov A, Di Diego JM, Sicouri S, Doss MX, Sachinidis A, Barajas-Martinez H et al. A temporal window of vulnerability for development of atrial fibrillation with advancing heart failure. Eur J Heart Fail 2014;16:271-280.

(171) Akar FG, Spragg DD, Tunin RS, Kass DA, Tomaselli GF. Mechanisms underlying conduction slowing and arrhythmogenesis in nonischemic dilated cardiomyopathy. Circ Res 2004;95:717-725.

(172) Hanna N, Cardin S, Leung TK, Nattel S. Differences in atrial versus ventricular remodeling in dogs with ventricular tachypacing-induced congestive heart failure. Cardiovasc Res 2004;63:236-244.

(173) Weber KT, Janicki JS, Shroff SG, Pick R, Chen RM, Bashey RI. Collagen remodeling of the pressure-overloaded, hypertrophied nonhuman primate myocardium. Circ Res 1988;62:757-765.

(174) Beuckelmann DJ, Nabauer M, Erdmann E. Alterations of K+ currents in isolated human ventricular myocytes from patients with terminal heart failure. Circ Res 1993;73:379- 385.

(175) Virag L, Jost N, Papp R, Koncz I, Kristof A, Kohajda Z et al. Analysis of the contribution of I(to) to repolarization in canine ventricular myocardium. Br J Pharmacol 2011;164:93-105.

(176) Akar FG, Wu RC, Juang GJ, Tian Y, Burysek M, Disilvestre D et al. Molecular mechanisms underlying K+ current downregulation in canine tachycardia-induced heart failure. Am J Physiol Heart Circ Physiol 2005;288:H2887-H2896.

(177) Zicha S, Xiao L, Stafford S, Cha TJ, Han W, Varro A et al. Transmural expression of transient outward potassium current subunits in normal and failing canine and human hearts. J Physiol 2004;561:735-748.

(178) Kaab S, Dixon J, Duc J, Ashen D, Nabauer M, Beuckelmann DJ et al. Molecular basis of transient outward potassium current downregulation in human heart failure: a decrease in Kv4.3 mRNA correlates with a reduction in current density. Circulation 1998;98:1383-1393.

(179) Borlak J, Thum T. Hallmarks of ion channel gene expression in end-stage heart failure. FASEB J 2003;17:1592-1608.

128

(180) Soltysinska E, Olesen SP, Christ T, Wettwer E, Varro A, Grunnet M et al. Transmural expression of ion channels and transporters in human nondiseased and end-stage failing hearts. Pflugers Arch 2009;459:11-23.

(181) Rose J, Armoundas AA, Tian Y, DiSilvestre D, Burysek M, Halperin V et al. Molecular correlates of altered expression of potassium currents in failing rabbit myocardium. Am J Physiol Heart Circ Physiol 2005;288:H2077-H2087.

(182) Radicke S, Cotella D, Graf EM, Banse U, Jost N, Varro A et al. Functional modulation of the transient outward current Ito by KCNE beta-subunits and regional distribution in human non-failing and failing hearts. Cardiovasc Res 2006;71:695-703.

(183) Kaab S, Nuss HB, Chiamvimonvat N, O'Rourke B, Pak PH, Kass DA et al. Ionic mechanism of action potential prolongation in ventricular myocytes from dogs with pacing-induced heart failure. Circ Res 1996;78:262-273.

(184) Cordeiro JM, Calloe K, Moise NS, Kornreich B, Giannandrea D, Di Diego JM et al. Physiological consequences of transient outward K+ current activation during heart failure in the canine left ventricle. J Mol Cell Cardiol 2012;52:1291-1298.

(185) Lopatin AN, Nichols CG. Inward rectifiers in the heart: an update on I(K1). J Mol Cell Cardiol 2001;33:625-638.

(186) Lopatin AN, Makhina EN, Nichols CG. Potassium channel block by cytoplasmic polyamines as the mechanism of intrinsic rectification. Nature 1994;372:366-369.

(187) Dhamoon AS, Jalife J. The inward rectifier current (IK1) controls cardiac excitability and is involved in arrhythmogenesis. Heart Rhythm 2005;2:316-324.

(188) Nerbonne JM, Kass RS. Molecular physiology of cardiac repolarization. Physiol Rev 2005;85:1205-1253.

(189) Jost N, Virag L, Comtois P, Ordog B, Szuts V, Seprenyi G et al. Ionic mechanisms limiting cardiac repolarization reserve in humans compared to dogs. J Physiol 2013;591:4189-4206.

(190) Roden DM. Taking the "idio" out of "idiosyncratic": predicting torsades de pointes. Pacing Clin Electrophysiol 1998;21:1029-1034.

(191) Varro A, Balati B, Iost N, Takacs J, Virag L, Lathrop DA et al. The role of the delayed rectifier component IKs in dog ventricular muscle and Purkinje fibre repolarization. J Physiol 2000;523 Pt 1:67-81.

(192) Tseng GN. I(Kr): the hERG channel. J Mol Cell Cardiol 2001;33:835-849.

(193) Volders PG, Sipido KR, Vos MA, Spatjens RL, Leunissen JD, Carmeliet E et al. Downregulation of delayed rectifier K(+) currents in dogs with chronic complete 129

atrioventricular block and acquired torsades de pointes. Circulation 1999;100:2455- 2461.

(194) Holzem KM, Glukhov AV, Efimov IR. The role of IKr in transmural repolarization abnormalities in human heart failure . Circulation 124. 2011. Ref Type: Generic

(195) Thomsen MB, Volders PG, Beekman JD, Matz J, Vos MA. Beat-to-Beat variability of repolarization determines proarrhythmic outcome in dogs susceptible to drug-induced torsades de pointes. J Am Coll Cardiol 2006;48:1268-1276.

(196) Shu L, Zhang W, Su G, Zhang J, Liu C, Xu J. Modulation of HERG K+ channels by chronic exposure to activators and inhibitors of PKA and PKC: actions independent of PKA and PKC phosphorylation. Cell Physiol Biochem 2013;32:1830-1844.

(197) Sam F, Kerstetter DL, Pimental DR, Mulukutla S, Tabaee A, Bristow MR et al. Increased reactive oxygen species production and functional alterations in antioxidant enzymes in human failing myocardium. J Card Fail 2005;11:473-480.

(198) Terentyev D, Gyorke I, Belevych AE, Terentyeva R, Sridhar A, Nishijima Y et al. Redox modification of ryanodine receptors contributes to sarcoplasmic reticulum Ca2+ leak in chronic heart failure. Circ Res 2008;103:1466-1472.

(199) Bonilla IM, Sridhar A, Gyorke S, Cardounel AJ, Carnes CA. Nitric oxide synthases and atrial fibrillation. Front Physiol 2012;3:105.

(200) Cleeter MW, Cooper JM, Darley-Usmar VM, Moncada S, Schapira AH. Reversible inhibition of cytochrome c oxidase, the terminal enzyme of the mitochondrial respiratory chain, by nitric oxide. Implications for neurodegenerative diseases. FEBS Lett 1994;345:50-54.

(201) Gomez R, Nunez L, Vaquero M, Amoros I, Barana A, de PT et al. Nitric oxide inhibits Kv4.3 and human cardiac transient outward potassium current (Ito1). Cardiovasc Res 2008;80:375-384.

(202) Kubalova Z, Terentyev D, Viatchenko-Karpinski S, Nishijima Y, Gyorke I, Terentyeva R et al. Abnormal intrastore calcium signaling in chronic heart failure. Proc Natl Acad Sci U S A 2005;102:14104-14109.

(203) Valdivia CR, Chu WW, Pu J, Foell JD, Haworth RA, Wolff MR et al. Increased late sodium current in myocytes from a canine heart failure model and from failing human heart. J Mol Cell Cardiol 2005;38:475-483.

(204) Sipido KR, Bito V, Antoons G, Volders PG, Vos MA. Na/Ca exchange and cardiac ventricular arrhythmias. Ann N Y Acad Sci 2007;1099:339-348.

130

(205) Haverkamp W, Breithardt G, Camm AJ, Janse MJ, Rosen MR, Antzelevitch C et al. The potential for QT prolongation and pro-arrhythmia by non-anti-arrhythmic drugs: clinical and regulatory implications. Report on a Policy Conference of the European Society of Cardiology. Cardiovasc Res 2000;47:219-233.

(206) Hearse DJ, Sutherland FJ. Experimental models for the study of cardiovascular function and disease. Pharmacol Res 2000;41:597-603.

(207) Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M et al. Heart Disease and Stroke Statistics-2016 Update: A Report From the American Heart Association. Circulation 2015.

(208) Stevenson WG, Sweeney MO. Arrhythmias and sudden death in heart failure. Jpn Circ J 1997;61:727-740.

(209) Luu M, Stevenson WG, Stevenson LW, Baron K, Walden J. Diverse mechanisms of unexpected cardiac arrest in advanced heart failure. Circulation 1989;80:1675-1680.

(210) Faggiano P, d'Aloia A, Gualeni A, Gardini A, Giordano A. Mechanisms and immediate outcome of in-hospital cardiac arrest in patients with advanced heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardiol 2001;87:655-1.

(211) Semelka M, Gera J, Usman S. Sick sinus syndrome: a review. Am Fam Physician 2013;87:691-696.

(212) Simon AB, Janz N. Symptomatic bradyarrhythmias in the adult: natural history following ventricular pacemaker implantation. Pacing Clin Electrophysiol 1982;5:372- 383.

(213) Lamas GA, Lee KL, Sweeney MO, Silverman R, Leon A, Yee R et al. Ventricular pacing or dual-chamber pacing for sinus-node dysfunction. N Engl J Med 2002;346:1854-1862.

(214) Drury AN, Szent-Gyorgyi A. The physiological activity of adenine compounds with especial reference to their action upon the mammalian heart. J Physiol 1929;68:213- 237.

(215) West GA, Belardinelli L. Sinus slowing and pacemaker shift caused by adenosine in rabbit SA node. Pflugers Arch 1985;403:66-74.

(216) Headrick JP, Matherne GP, Berne RM. Myocardial adenosine formation during hypoxia: effects of ecto-5'-nucleotidase inhibition. J Mol Cell Cardiol 1992;24:295-303.

(217) Borst MM, Schrader J. Adenine nucleotide release from isolated perfused guinea pig hearts and extracellular formation of adenosine. Circ Res 1991;68:797-806.

131

(218) Sparks HV, Jr., Bardenheuer H. Regulation of adenosine formation by the heart. Circ Res 1986;58:193-201.

(219) Chaudary N, Naydenova Z, Shuralyova I, Coe IR. Hypoxia regulates the adenosine transporter, mENT1, in the murine cardiomyocyte cell line, HL-1. Cardiovasc Res 2004;61:780-788.

(220) Darvish A, Pomerantz RW, Zografides PG, Metting PJ. Contribution of cytosolic and membrane-bound 5'-nucleotidases to cardiac adenosine production. Am J Physiol 1996;271:H2162-H2167.

(221) Deussen A. Metabolic flux rates of adenosine in the heart. Naunyn Schmiedebergs Arch Pharmacol 2000;362:351-363.

(222) Newman WH, Grossman SJ, Frankis MB, Webb JG. Increased myocardial adenosine release in heart failure. J Mol Cell Cardiol 1984;16:577-580.

(223) Funakoshi H, Zacharia LC, Tang Z, Zhang J, Lee LL, Good JC et al. A1 adenosine receptor upregulation accompanies decreasing myocardial adenosine levels in mice with left ventricular dysfunction. Circulation 2007;115:2307-2315.

(224) Koeppen M, Eckle T, Eltzschig HK. Selective deletion of the A1 adenosine receptor abolishes heart-rate slowing effects of intravascular adenosine in vivo. PLoS One 2009;4:e6784.

(225) DiFrancesco D. Pacemaker mechanisms in cardiac tissue. Annu Rev Physiol 1993;55:455-472.

(226) Mery PF, Abi-Gerges N, Vandecasteele G, Jurevicius J, Eschenhagen T, Fischmeister R. Muscarinic regulation of the L-type calcium current in isolated cardiac myocytes. Life Sci 1997;60:1113-1120.

(227) Belardinelli L, Giles WR, West A. Ionic mechanisms of adenosine actions in pacemaker cells from rabbit heart. J Physiol 1988;405:615-633.

(228) Zaza A, Rocchetti M, DiFrancesco D. Modulation of the hyperpolarization-activated current (I(f)) by adenosine in rabbit sinoatrial myocytes. Circulation 1996;94:734-741.

(229) Karmouty-Quintana H, Zhong H, Acero L, Weng T, Melicoff E, West JD et al. The A2B adenosine receptor modulates pulmonary hypertension associated with interstitial lung disease. FASEB J 2012;26:2546-2557.

(230) Cocuron JC, Anderson B, Boyd A, Alonso AP. Targeted metabolomics of Physaria fendleri, an industrial crop producing hydroxy fatty acids. Plant Cell Physiol 2014;55:620-633.

132

(231) Cocuron JC, Alonso AP. Liquid chromatography tandem mass spectrometry for measuring (1)(3)C-labeling in intermediates of the glycolysis and pentose phosphate pathway. Methods Mol Biol 2014;1090:131-142.

(232) Woods WT, Urthaler F, James TN. Spontaneous action potentials of cells in the canine sinus node. Circ Res 1976;39:76-82.

(233) Funaya H, Kitakaze M, Node K, Minamino T, Komamura K, Hori M. Plasma adenosine levels increase in patients with chronic heart failure. Circulation 1997;95:1363-1365.

(234) Varani K, Laghi-Pasini F, Camurri A, Capecchi PL, Maccherini M, Diciolla F et al. Changes of peripheral A2A adenosine receptors in chronic heart failure and cardiac transplantation. FASEB J 2003;17:280-282.

(235) Long VP, III, Bonilla IM, Vargas-Pinto P, Nishijima Y, Sridhar A, Li C et al. Heart failure duration progressively modulates the arrhythmia substrate through structural and electrical remodeling. Life Sci 2015;123:61-71.

(236) Meyer TE, Chung ES, Perlini S, Norton GR, Woodiwiss AJ, Lorbar M et al. Antiadrenergic effects of adenosine in pressure overload hypertrophy. Hypertension 2001;37:862-868.

(237) Franceschi F, Deharo JC, Giorgi R, By Y, Monserrat C, Condo J et al. Peripheral plasma adenosine release in patients with chronic heart failure. Heart 2009;95:651-655.

(238) Laghi-Pasini F, Guideri F, Petersen C, Lazzerini PE, Sicari R, Capecchi PL et al. Blunted increase in plasma adenosine levels following dipyridamole stress in dilated cardiomyopathy patients. J Intern Med 2003;254:591-596.

(239) Starling RC, Hammer DF, Altschuld RA. Human myocardial ATP content and in vivo contractile function. Mol Cell Biochem 1998;180:171-177.

(240) Shen W, Asai K, Uechi M, Mathier MA, Shannon RP, Vatner SF et al. Progressive loss of myocardial ATP due to a loss of total purines during the development of heart failure in dogs: a compensatory role for the parallel loss of creatine. Circulation 1999;100:2113-2118.

(241) Hettinger-Smith BD, Leid M, Murray TF. Chronic exposure to adenosine receptor agonists and antagonists reciprocally regulates the A1 adenosine receptor-adenylyl cyclase system in cerebellar granule cells. J Neurochem 1996;67:1921-1930.

(242) Cui M, Bai X, Li T, Chen F, Dong Q, Zhao Y et al. Decreased extracellular adenosine levels lead to loss of hypoxia-induced neuroprotection after repeated episodes of exposure to hypoxia. PLoS One 2013;8:e57065.

133

(243) Yanni J, Tellez JO, Maczewski M, Mackiewicz U, Beresewicz A, Billeter R et al. Changes in ion channel gene expression underlying heart failure-induced sinoatrial node dysfunction. Circ Heart Fail 2011;4:496-508.

(244) Matherne GP, Linden J, Byford AM, Gauthier NS, Headrick JP. Transgenic A1 adenosine receptor overexpression increases myocardial resistance to ischemia. Proc Natl Acad Sci U S A 1997;94:6541-6546.

(245) Rubio R, Berne RM. Release of adenosine by the normal myocardium in dogs and its relationship to the regulation of coronary resistance. Circ Res 1969;25:407-415.

(246) Driver AG, Kukoly CA, Spence PA, Chitwood WR, Jr., Mustafa SJ. Pericardial fluid adenosine in ischemic and valvular heart disease. Chest 1995;107:346-351.

(247) Zhang Z, Xu Y, Song H, Rodriguez J, Tuteja D, Namkung Y et al. Functional Roles of Ca(v)1.3 (alpha(1D)) calcium channel in sinoatrial nodes: insight gained using gene- targeted null mutant mice. Circ Res 2002;90:981-987.

(248) Mesirca P, Bidaud I, Briec F, Evain S, Torrente AG, Le QK et al. G protein-gated IKACh channels as therapeutic targets for treatment of sick sinus syndrome and heart block. Proc Natl Acad Sci U S A 2016.

(249) Alboni P, Menozzi C, Brignole M, Paparella N, Gaggioli G, Lolli G et al. Effects of permanent pacemaker and oral theophylline in sick sinus syndrome the THEOPACE study: a randomized controlled trial. Circulation 1997;96:260-266.

(250) Slawsky MT, Givertz MM. Rolofylline: a selective adenosine 1 receptor antagonist for the treatment of heart failure. Expert Opin Pharmacother 2009;10:311-322.

(251) Kitakaze M, Hori M. Adenosine therapy: a new approach to chronic heart failure. Expert Opin Investig Drugs 2000;9:2519-2535.

(252) Cerbai E, Klockner U, Isenberg G. Ca-antagonistic effects of adenosine in guinea pig atrial cells. Am J Physiol 1988;255:H872-H878.

(253) Verkerk AO, van Borren MM, van Ginneken AC, Wilders R. Ca(2+) cycling properties are conserved despite bradycardic effects of heart failure in sinoatrial node cells. Front Physiol 2015;6:18.

(254) Cohn JN, Kowey PR, Whelton PK, Prisant LM. New guidelines for potassium replacement in clinical practice: a contemporary review by the National Council on Potassium in Clinical Practice. Arch Intern Med 2000;160:2429-2436.

(255) Gennari FJ. Hypokalemia. N Engl J Med 1998;339:451-458.

(256) Brown MJ, Brown DC, Murphy MB. Hypokalemia from beta2-receptor stimulation by circulating epinephrine. N Engl J Med 1983;309:1414-1419. 134

(257) Tveskov C, Djurhuus MS, Klitgaard NA, Egstrup K. Potassium and magnesium distribution, ECG changes, and ventricular ectopic beats during beta 2-adrenergic stimulation with terbutaline in healthy subjects. Chest 1994;106:1654-1659.

(258) Nordrehaug JE, Johannessen KA, von der LG, Myking OL. Circulating catecholamine and potassium concentrations early in acute myocardial infarction: effect of intervention with timolol. Am Heart J 1985;110:944-948.

(259) Cohn JN, Levine TB, Olivari MT, Garberg V, Lura D, Francis GS et al. Plasma norepinephrine as a guide to prognosis in patients with chronic congestive heart failure. N Engl J Med 1984;311:819-823.

(260) Francis GS. The relationship of the sympathetic nervous system and the renin- angiotensin system in congestive heart failure. Am Heart J 1989;118:642-648.

(261) Caralis PV, Materson BJ, Perez-Stable E. Potassium and diuretic-induced ventricular arrhythmias in ambulatory hypertensive patients. Miner Electrolyte Metab 1984;10:148-154.

(262) Sanguinetti MC, Jurkiewicz NK. Role of external Ca2+ and K+ in gating of cardiac delayed rectifier K+ currents. Pflugers Arch 1992;420:180-186.

(263) Parham WA, Mehdirad AA, Biermann KM, Fredman CS. Hyperkalemia revisited. Tex Heart Inst J 2006;33:40-47.

(264) Yang T, Roden DM. Extracellular potassium modulation of drug block of IKr. Implications for torsade de pointes and reverse use-dependence. Circulation 1996;93:407-411.

(265) Macdonald JE, Struthers AD. What is the optimal serum potassium level in cardiovascular patients? J Am Coll Cardiol 2004;43:155-161.

(266) Snider M, Kalbfleisch S, Carnes CA. Initial experience with antiarrhythmic medication monitoring by clinical pharmacists in an outpatient setting: a retrospective review. Clin Ther 2009;31:1209-1218.

(267) Snider M, Carnes C, Grover J, Davis R, Kalbfleisch S. Cost-benefit and cost-savings analyses of antiarrhythmic medication monitoring. Am J Health Syst Pharm 2012;69:1569-1573.

(268) Colilla S, Crow A, Petkun W, Singer DE, Simon T, Liu X. Estimates of current and future incidence and prevalence of atrial fibrillation in the U.S. adult population. Am J Cardiol 2013;112:1142-1147.

(269) January CT, Wann LS, Alpert JS, Calkins H, Cleveland JC, Jr., Cigarroa JE et al. 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: A

135

Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. Circulation 2014.

(270) Deering TF, Clair WK, Delaughter MC, Fisher WG, Garlitski AC, Wilkoff BL et al. A Heart Rhythm Society Electrophysiology Workforce study: current survey analysis of physician workforce trends. Heart Rhythm 2010;7:1346-1355.

(271) Glukhov AV, Fedorov VV, Kalish PW, Ravikumar VK, Lou Q, Janks D et al. Conduction remodeling in human end-stage nonischemic left ventricular cardiomyopathy. Circulation 2012;125:1835-1847. (272) Fishman GI, Chugh SS, Dimarco JP, Albert CM, Anderson ME, Bonow RO et al. Sudden cardiac death prediction and prevention: report from a National Heart, Lung, and Blood Institute and Heart Rhythm Society Workshop. Circulation 2010;122:2335-2348.

136