|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004759
N-glycosylation modulates gating and antibiotic block of the human potassium channel, herg1a
h [electronic resource] /
by Sarah Norring.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Dissertation (PHD)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: Arrhythmias are often caused by aberrant ion channel activity, resulting in remodeling of the cardiac action potential. Two K+ currents, IKs and IKr, contribute to phase III repolarization of the human cardiac action potential. Human ether-a-go-go-related gene 1 (hERG1), a voltage-gated potassium channel, underlies IKr. Alterations in the repolarization phase of the action potential, and in particular IKr, can lead to arrhythmias, long or short QT syndrome, heart disease, and sudden cardiac death. HERG1A has two putative N-glycosylation sites located in the S5-S6 linker region, one of which is N-glycosylated. The aim of the first study was to determine whether and how N-linked glycosylation modifies hERG1A channel function. Voltage-dependent gating and kinetics of hERG1A were evaluated under conditions of full glycosylation, no sialylation, in the absence of complex N-glycans, and following the removal of the full N-glycosylation structure. The hERG1A steady state activation relationship was shifted linearly along the voltage axis by a depolarizing ~9 mV under each condition of reduced glycosylation. Steady state channel availability curves were shifted by a much greater depolarizing 20-30 mV under conditions of reduced glycosylation. There was no significant difference in steady state gating parameters among the less glycosylated channels, suggesting that channel sialic acids are responsible for most of the effect of N-glycans on hERG1A gating. A large rightward shift in hERG1A window current for the less glycosylated channels was caused by the observed depolarizing shifts in steady state activation and inactivation. The much larger shift in inactivation compared to activation leads to an increase in hERG1A window current. Together, these data suggest that there is an increase in the persistent hERG current that occurs at more depolarized potentials under conditions of reduced glycosylation. This would lead to increased hERG1A activity during the AP, effectively increasing the rate of repolarization, and reducing AP duration, as observed through in silico modeling of the ventricular AP. The data describe a novel mechanism by which hERG1A activity is modulated by physiological and pathological changes in hERG1A glycosylation, with increased channel sialylation causing a loss of hERG1A activity that would likely cause an extension of the ventricular AP. The second study was to evaluate possible changes in antibiotic drug block as a result of alterations to N-glycosylation. We determined that N-glycans play a protective role on the hERG1A channel. SMX, Erythromycin, and Penicillin G were assessed individually at three concentrations. The data showed increases in antibiotic block with decreases in N-glycans. In addition, alterations in the voltage-dependence of block with changes in N-glycans were observed. SMX block was voltage-independent at each drug concentration under conditions of reduced sialylation only. Overall, these data indicate a functional role for N-glycosylation in the modulation of hERG1A antibiotic block, suggesting that even small changes in channel N-glycosylation modulate hERG1A block, and thereby likely impact the rate of action potential repolarization. The data from these studies enhances our understanding of the role of N-glycosylation on hERG1A function and drug block, and how that role will impact the cardiac action potential and overall cardiac excitability.
Advisor: Eric S. Bennett, Ph.D.
x Medical Sciences
t USF Electronic Theses and Dissertations.
N Glycosylation Modulates Gating and Antibiotic Block of the Human Potassium Channel, hERG1A by Sarah A. Norring A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Molecular Pharmacology and Physiology College of Medicine University of South Florida Major Professor: Eric S. Bennett, Ph.D Jahanshah Amin, Ph.D Jay B. Dean, Ph.D Andreas G. Seyfang Ph.D. Kay Pong D. Yip, Ph.D Date of Approval: September 30 20 10 Keywords: glycans, K + arrhythmias, sugars drug block Copyright 2010 Sarah Ann Norring
ACKNOWLEDGMENTS First, I would like to thank my family for the support and guidance they have given me during my graduate education. Their love and motivation has always been and always will be unwavering. A big thank you goes out to Eric S. Bennett, PhD. He has been a wonderful my PhD without his help. I also would like to thank my committee members: Jahanshah Amin, PhD, Jay B. Dean, PhD, Kay Pong (Daniel) Yip, PhD, and Andreas G. Seyfang, PhD. Finally, I would like to thank the members of the Bennett lab: Tara A. Schwetz, PhD, Marty L. Montpetit, PhD, Andrew Ednie, BS, and Kofi Kermit Horton, MS. On a special note, Tara was a true friend and colleague.
i TABLE OF CONTENTS LIST OF TABLES v LIST OF FIGURES vi ABSTRACT viii CHAPTER 1 INTRODUCTION 1 Normal heart function 1 Arrhythmias and conduction dysfunction 6 9 Alterations in I Kr can cause arrhythmias 12 HERG1A underlies I Kr 12 HERG1A activity during the ventricular AP 14 HERG1A characteristics 15 HERG1A N glycosylation 19 Known role of glycosylation on hERG1A 20 Glycosylation effects on other channels 20 Glycosylat ion pathway 21 Congenital disorders of glycosylation (CDG) 23 24
ii Glycosylation is differentially expressed and remodeled in the heart 24 Subunit vs cell specific effects of glycosylation 25 Summary 2 5 CHAPTER 2 MATERIALS AND METHODS 2 7 Chinese h amster o vary c ell c ulture and t ransfection 2 7 Vector construction and mutagenesis 27 Electrophysiology and data a nalysis 2 8 Whole c ell recordings in CHO c ells 2 8 Conductance voltage r elationship for CHO cell recordings 29 Steady state i nactivation 31 Window c urrent 32 Time constants for deactivation ( Tn ) 3 2 Computer s imulation of action potential 32 Whole c ell h omogenization 33 Cell surface biotinylation 34 Immunoblots 34 Data analysis and statistics 35 CHAPTER 3 N GLYCANS LIMIT HERG1A ACTIVITY 36 Results 38 HERG1 A is differently glycosylated 38 HERG1A function is modulated by N glycans 41 N glycans promote hERG1A voltage dependent a ctivation 41
iii The hERG1A steady state activation (SSA) relationships are shifted to more depolarized potentials under conditions of reduced glycosylation 4 5 N glycans act to slow closing of hERG1A 4 5 N glycans cause a greater shift in voltage dependence of steady state channel availability 4 7 The large and variable shift in the voltage dependence of activation and inactivation causes a shift to more physiological potentials and an increase in hERG1A window current 4 7 Discussion 51 N Glycans lead to a loss of hERG1A function 51 CHAPTER 4 PENICILLIN BLOCK OF HERG1A 5 5 Results 60 HERG1A is blocked by Penicillin G 60 Discussion 63 CHAPTER 5 N GLYCANS ALTER ANTIBIOTIC BLOCK OF HERG1A 65 Results 79 SMX block is altered by N glycans 79 Erythromycin block is modified by complex N glycan s 81 N glycans alter Penicillin G block of hERG1A 83 N glycans modulate SMX b lock at small depolarizations 85 Complex N glycans impact Erythromycin b lock at small depolarizations 89 Si alic acid and complex N glycans block Penicill in G at small depolarizations 93
iv Discussion 93 CHAPTER 6 FINAL DISCUSSION 102 Physiologic and pathologic consequences 102 The impact of N glycan dependent hERG1A activity on AP waveform is unique among voltage gated K+ c hannels 102 Regulated changes in glycosylation could lead to modulated hERG1A activity 103 Aberrant changes in glycosylation could lead to hERG1A dysfunction 104 HERG1A antibiotic bl ock is red u ced with N glycans 105 Antibiotic block may hap p en through multiple pathways 106 Future Studies 108 Summary 1 08 CHAPTER 7 REFERENCES 111 ABOUT THE AUTHOR End Page
v LIST OF TABLES Table 1 Concentrat ions for antibiotic drug study 30 Table 2 Biophysical parameters for hERG1A as expressed in Pro5, Lec2, Lec1, and following N Glycanase treatment 43
vi LI ST OF FIGURES Figure 1 Normal electrical conduction through the heart 2 Figure 2 Normal ECG 4 Figure 3 Action potential waveforms in the heart 5 Figure 4 Dysfunction s in the heart 8 Figure 5 Long QT disorder list 11 Figure 6 HERG1A activity in the heart 13 Figure 7 HERG1A channel 16 Figure 8 Gating properties of hERG1A 17 Figure 9 Glycosylation pathway 2 2 Figure 10 HERG1A is differently glycosylated 40 Figure 11 HERG1A current is altered by N glycans 42 Figure 12 HERG1A steady state activation curve is shifted to more hyperpolarized potentials with full N glycosylation 44 Figure 13 N glycans act to slow hERG1A closing 4 6 Figure 14 Steady state channel availability is altered by N glycans 4 8 Figure 15 N glycans limit hERG1A window currents with persistent hERG1A activity at more hyperpolarized potentials 50 Figure 16 in silico modeling predicts ventricular action potential is extended by increased N glycosylation 52 Figure 17 Structure of Penicillin G 57
vii Figure 18 Current trace at +20 mV for hERG1A at multiple drug concentrations 6 1 Figure 19 Percent block by P enicillin G 62 Figure 20 ICH S7B guideline 6 7 Figure 21 Potential hERG1A block by antibiotics 69 Figure 22 SMX block of hERG1A 74 Figure 23 Erythromycin blocks hERG1A 76 Figure 24 Penicillin G blocks hERG1A 78 Figure 25 Concentration dependent block by SMX 80 Figure 26 Concentration de pendent block by Erythromycin 82 Figure 27 Concentration de pendent block by Penicillin G 84 Figure 28 50 g/mL SMX block at small depolarizations 8 6 Figure 29 300 g/mL SMX block at small depolarizations 87 Figure 30 600 g/mL SMX block at small depolarizations 88 Figure 31 25 g/mL Erythromycin block at small depolarizations 90 Figure 32 100 g/mL Ery th r omycin block at small depolarizations 91 Figure 33 200 g/mL Erythromycin block at small depolarizations 92 F igure 34 10 g/mL Penicillin G block at small depolarizations 94 Figure 35 50 g/mL Penicillin G block at small depolarizations 95 Figure 36 200 g/mL Penicillin G block at small depolarizations 96 Figure 37 Model of extracellular block of hERG1A 98 Figure 38 Model of intracellular block of hERG1A 99
viii ABSTRACT Arrhythmias are often caused by aberrant ion channel activity, resulting in remodeling of the cardiac action potential. Two K + currents, I Ks and I Kr contribute to phase III repolarization of the human cardiac action potential H uman ether a go go related gene 1 (hERG1 ), a voltage gated potassium channel, underlies I Kr Alterations in the r epolarization phase of the action potential and in particular I Kr can lead to arrhyth mias, long or short QT syndrome heart disease, and sudden cardiac deat h. HERG1A has two putative N glycosylation sites located in the S5 S6 linker region, one of which is N glycosylated. The aim of the first study was to determine whether and how N link ed glycosylation modifies hERG1A channel function Voltage dependent gating and kinetics of hERG1A were evaluated under conditions of full glycosylation, no sialylation, in the absence of complex N glycans, and following the removal of the full N glycosyl ation structure. The hERG1A steady state activation relationship was shifted linearly along the voltage axis by a depolarizing ~9 mV under each condition of reduced glycosylation. Steady state channel availability curves were shifted by a much greater de polarizing 20 30 mV under conditions of reduced glycosylation. There was no significant difference in steady state gating parameters among the less glycosylated channels, suggesting that channel sialic acids are responsible for
ix most of the effect of N gly cans on hERG1A gating A large rightward shift in hERG1A window current for the less glycosylated channels was caused by the observed depolarizing shifts in steady state activation and inactivation. The much larger shift in inactivation compared to activ ation leads to an increase in hERG1A window current. Together, these data suggest that there is an increase in the persistent hERG current that occurs at more depolarized potentials under conditions of reduced glycosylation. This would lead to increased hERG1A activity during the AP, effectively increasing the rate of repolarization, and reducing AP duration, as observed through in silico modeling of the ventricular AP. The data describe a novel mechanism by which hERG1A activity is modulated by physiolo gical and pathological changes in hERG1A glycosylation, with increased channel sialylation causing a loss of hERG1A activity that would likely cause an extension of the ventricular AP The second study was to evaluate possible changes in antibiotic drug b lock as a result of alterations to N glycosylation. We determined that N glycans play a protective role on the hERG1A channel. SMX, Erythromycin, and Penicillin G were assessed individually at three concentrations. The data showed increases in antibiotic block with decreases in N glycans. In addition, alterations in the voltage dependence of block with changes in N glycans were observed. SMX block was voltage independent at each drug concentration under conditions of reduced sialylation only. Over all, these data indicate a functional role for N glycosylation in the modulation of hERG1A antibiotic block suggesting that even small changes in channel N glycosylation modulate hERG1A block and thereby likely
x impact the rate of action potential re polarization. The data from these studies enhance s our understanding of the role of N glycosylation on hERG1A function and drug block, and how that role will impact the cardiac action potential and overall cardiac excitability.
1 CHAPTER 1 INTRODUCTION Normal heart function The heart is a muscular organ that is responsible for pumping bl ood throughout the body by repeated, rhythmic contractions. It is made up of cardiac muscle which is found only within this organ. The average human heart beat s about 72 times per minute The right side of the heart collects the de oxygenated blood from the body in to the right atrium and then pump s it via the right ventricle into the lungs for gas e xchange. Th en the left at rium fills with oxygenated blood before moving to the left ventricle which pumps it out to the body. The ventricles are thicker and stronger than the atria. The muscle wall surrounding the left ventricle is thicker than the wall surrounding the right vent ricle due to the force needed to pump the blood through the body The heart is effectively a syncytium, a meshwork of cardiac muscle cells interconnected by contiguous cytoplasmic bridges 1 3 These muscle cells lin ked together communicate when to contract and relax the heart through the pathway of conduction. Conduction is the electrical signaling pathway through the heart that produces normal cardiac rhythm (see Figure 1). This pathway is comprised
2 Figure 1 Normal electrical conduction through the heart Conduction in the heart begins with the sinoatrial (SA) node generating the electric signals that will spread to the atrioventricular (AV) node resulting in ventricular depolarization followed by repolarization in the reverse direction.
3 of five elements: sino atrial (SA) node atrio ventricular (AV) node bundle of His left and right bundle branches and Purkinje fibers. The SA node is the natural pacemaker of the heart. It releases electrical stimuli at a regular ra te that is dictated by the needs of the body. Each stimulus passes through the myocardial cells of the atria creating a wave of contraction which spreads rapidl y through both atria. Th is stimulus from the SA node reaches the AV node and is delayed briefly so that the contracting atria have enough time to pump all the blood into the ventricles. Once the atria are empty of blood the valves between the atria and ve ntricles close. T he electrical stimulus passes through the AV node and Bundle of His into the Bu ndle branches and Purkinje fiber s. This allows all the cells in the ventricles to receive an electrical stimulus causing them to contract (depolarize) Once the ventricles have contracted, the cells then relax, repolarize, and are then ready for the next stimulus. Figure 2 displays an electrocardiogram (ECG) that is a recording a heart rhythm. The electrical stimulation moving through the heart is propagated by action potentials. The action potential (AP) is the hallmark of electrical communication used by the body and is the rapid transient depolarization and subsequent repolarization produced by a concerted effort of m any different ion channels (see Figure 3). There are two types of APs in the heart: fast and slow. The slow action potential consists of phases 0, 3, and 4, whereas the fast action potential has phases 0, 1, 2, 3, and 4. In phase 0 of the fast AP, voltage gated (Na v )
4 Figure 2. Normal ECG. An ECG is a recording of conduction across the chest A normal ECG illustrates that the heart is working correctly and allows for detection of changes in how the heart beats. Figure adapted from eleceng.dit.ie/tburke/biomed/ecg.png
5 Figure 3 : Action potential waveforms in the heart. The top panel illustrates examples of the variety of AP waveforms found in the heart. The different currents involved in the fast AP are shown in the lower pane. Figure modified from cvphysiology.com
6 channels open and cause the rapid depolarization of the membrane, due to sodium (Na + ) ions moving into the cell down their electrochemical gradient. Na v channels then inactivate quickly while voltage gated potassium (K v ) channels activate to produce a tra nsient outward current (I to ). This results in the small, rapid repolarization seen in phase 1. The plateau phase (phase 2) consists of voltage gated calcium (Ca v ) channels activating producing an inward current that is compensated with an outward potas sium (K + ) current (I Kur ). In phase 3, the outward K + current, I Kr and I Ks starts to exceed the inward calcium current, thereby repolarizing the cell back to its resting membrane potential. The diastolic potential (phase 4) is accomplished by an outward K + current equaling the action of the Na + /K + ATPase pump. Normal heart rhythm occurs when the cardiac muscle cells, the conduction pathway, and the voltage gated ion channels all work properly and in sync with one another. Any disruption can potentially cause dysfunction of the heart. Arrhythmias and conduction dysfunction Cardiac arrhythmia s are considered a large collection of conditions where there is abnormal electrical activity in the heart 4,5 The heart may beat too fast or too slow, and can be regular or irregular. Many arrhythmias are life threatening and can result in cardiac death 2,6 Other arrhythmias may be less threatening and symptoms may not present, but it may predispose an individual toward a s troke
7 or embolus. There are many types of arrhythmias from bradycardias, tachycardias, automaticity, reentrant, and fibrillations (see Figure 4) 7 Bradycardia is a slowing of the heart beat that may be brought on by a slowed signal from the SA node, a loss of activity from the SA node, or by blocking the electrical signal from the atria to the ventricles (AV block). Tachycardia is an increase in heart rate about 100 beats per minute. Tachycardia is not always considered an arrhythmia. The sympathetic nervous system acts on the SA node to increase h eart rate in times of exercise or emotional stress. When tachycardia results from extra abnormal impulses in the cardiac cycle, then it is considered to be pro arrhythmic. Automaticity refers to the potential a cardiac muscle cell firing an action potent ial. All cardiac muscle cells have the ability to initiate an AP, but normal conduction keeps the generation of APs to designated cells, such as the SA node. When a cardiac muscle cell initiates an AP, it is commonly called an ectopic focus. This has th e potential of causing an extra beat or seriously altering the contraction and proper flow of blood throughout the body. An ectopic focus has the possibility of permanently modifying the efficiency of the heart to pump blood. Reentrant, or re entry, arrh ythmias occur when the electrical signal gets stuck in a small part of the heart and travels back and forth in that area. The normal signaling pathway occurs through the whole heart. This dysfunction may occur when a signal is delayed traveling to an are a in the heart. The heart cells have the potential of treating the delayed signal as a
8 Figure 4: Dysfunctions in the heart Alterations in ion channel activity can cause multiple cardiac dysfunctions including Long QT syndrome, torsades de pointes, and ventricular fibrillation. Figure modified from Keating 2001 7
9 still moving through the rest of the cardiac heart cells. This may lead to atrial flutter and ventricular tachycardia. Fibrillation builds on the reentrant phenomenon where the electrical impuls e is stuck and in one area, except it is now stuck in an entire chamber of the heart, i.e. atria or ventricle. The chamber essentially flutters from the chaos of multiple signaling moving in every direction. Although atrial fibrillation is not always con sidered life threatening, ventricle fibrillation (VF) is a medical emergency. VF is considered cardiac arrest since the heart cannot pump blood when the ventricle cannot contract. This can lead to death in a matter of minutes. Sudden Arrhythmia Death Sy ndrome (SADS) is the term used to describe death in a person due to an arrhythmia. The most common form of this is coronary artery disease that kills over 300,000 people every year in the US. Disruption of the conduction in t he heart can occur as a result of many reasons from diseases (such as heart disease and diabetes), aging, smoking, excessive drinking, inherited abnormalities, and alterations in cardiomyocyte action potentials. Cardiac action potential maladies can be c aused by alterations in ion channel function. Figure 4 displays one of these dysfunctions, Long QT Syndrome (LQTS). LQTS is an extension of the QT segment as seen on an
10 electrocardiogram (ECG). The QT segment involves the depolarization and subsequent r epolarization of the ventricles in the heart. The QRS segment shows the depolarization of the ventricle and the T segment illustrates the repolarization of the ventricle. LQTS can be either inherited or acquired. Acquired LQTS typically happens as a res ult of medications that alter ion channel function. Inherited LQTS occurs when there is a mutation of a gene 8 16 Currently, there are twelve known inherited LQTS categories with LQT1, LQT2, and LQT 3 being the most common (see Figure 5). LQT1 and LQT2 make up about 65% of all LQTS cases. LQT1 involves mutations the the KvLQT1 channel. KvLQT1 encodes I Ks one of two major repolarizing currents of the AP. These mutations will delay repolarization o f the action potential and increase action potential duration (APD) leading to elongation of the QT segment. Although LQT1 is the most common form of LQTS, it is the least severe. LQT2 occurs as a result of mutations in the hERG1 channel. HERG1 underli es I Kr the other major repolarizing current of the cardiac AP. Altering the repolarization phase, essentially delaying it, will extend the action potential and lengthen the QT segment. I Kr is the current responsible for ending the AP and normal function of hERG1 could be considered protective against early after depolarizations (EADs). LQT3 entails mutations in SCN5A, the channel that underlies the main sodium current in phase 0 of the action potential. Extension of the depolarization phase of the AP t hrough mutations that cause a decrease in inactivation of the Na v channel will also elongate the QT segment. Again, this has the potential of
11 Figure 5. Long QT disorder list. A list of known LQT disorders and the genes associated with each disorder. Figure from eplabdigest.com
12 causing disruption to normal heart rhythm. Interestingly, Brugada Syndrome can also be caused by mutations in SCN5A 9 These mutations in SCN5A, that underlies I Na cause a loss o f function resulting in elevation of the ST segment and dispersion of repolarization. This will cause a phase 2 reentry that has the potential to trigger ventricular tachycardia. In addition to LQTS, another disease that involves the QT segment is called Short QT Syndrome (SQTS). SQTS is the result of a shortening of the QT segment typically resulting from an increase in potassium current. The main repolarizing potassium currents are I Ks and I Kr 17 19 Increases in these currents can be the consequence of mutations of the channels that underlie each current. Whether the Q T segment is lengthened or shortened, the result can be the same. Any alteration to the QT segment can alter the cardiac AP and can lead to cardiac dysfunction 15,16,18 20 Alter ations in I Kr can cause arrhythmias Alterations in I Kr can lead to cardiac dysfunction 8,11,15,18,20 26 If the current is blocked, then an extension of the QT segment occurs and can result in LQT2. If there is an increa se in I Kr then a shortening of the QT segment transpires and this may still lead to an arrhythmic event 27 HERG1A underlies I Kr H uman ether a go go related gene 1 (hERG1 ) is the voltage gated potassium channel that underlies I Kr 19,28 31 I Kr is one of two major potassium currents that
13 help repolarize the cardiac ventricular action potential (see Figure 6). Phase 3 is Kr is contributing nearly all of the K+ efflux from the cell. This removal of potassium causes a rapid change in the membrane potential from positive to negative potentials. HERG1A activity during the ventricular AP HERG1A has a unique functionality when it comes to the cardiac action potential 19,28 31 When the AP begins at phase 0, the membrane potential moves very quickly from negative to positive potentials. Phase 0 happens in ~10 ms. inactivation rate is faster than its activation rate at positive potentials. By the time phase 0 is complete and the membrane potential is around +40 mV, hERG1A channels have mos tly transitioned to the inactivated state. This allows for hERG1A to be prepared for the final part of the plateau phase and the repolarization phase. As the membrane potential starts to fall from positive potentials, hERG1A transitions for inactive to t he open state, starting the flow of K+ efflux from the cell. As the potential becomes even more negative with K+ efflux, hERG1A channels recover even faster to their open state. This could be considered a positive loop feedback. This large efflux of pot assium from the cell can be seen as the steep slope of phase 3 of the AP. Once the membrane potential nears the resting potential, the channels transition to their closed state, ready to start the next action potential.
14 Figure 6 : HERG1A activity in the heart HERG1A is in an inactivated state for most of phase 0 and 1. During phase 2 repolarization hERG1A transitions back into its open state allowing a large efflux of potassium (phase 2 and 3). Figure from Keating 2001 7
15 HERG1A characteristics HERG1A is a voltage gated potassium channel 18,31 It consists of four subunits coming together to create a pore forming channel. Each subunit consists of six trans membrane domains or segments (S1 6) (see Figure 7). Segments 1 4 are called the voltage sensing domains with segment 4 (S4) having positive charges every three amino acids. These charges sense a change in the membrane potential and will move through the membrane outward allowing the channel to open. Segments 5 and 6 are the pore forming domains. The linker region that connects segments 5 and 6 is called the pore forming loop, or P loop, and shapes the pore region of the channel. Within the P loop is a recognized amino acid sequence that deter mines if it is a K+ channel. This sequence is GYG (glycine tyrosine glycine). This sequence is thought to allow the K+ channel to be selective towards K+ ions 32 However, there are e xceptions to this and phenylalaline glycine), but is still selective for K+ ions. Once hERG1A has four subunits together to form a functional channel, it can then start to allow K+ ions to flow through the pore region. This occurs by the channel transitioning from the closed state to an open state. When the S4 segment senses a change in the membrane potential, the segments (there is one on each subunit or four total) will move outward causing a conformational change in the channel, opening the pore region. K+ ions will start to flow out of the cell. Figure 8 shows typical gating properties of hERG1A. In addition to opening and closing, hERG1A can also
16 Figure 7 : HERG1A channel The hERG1A channel consists of four subunits (one is pictured). The red arrows point to the N glycosylation consensus sequences found on the P loop of hERG1A. Only the site N terminal to the pore is glycosylated. Figure from Sanguinetti 2006 31
17 Figure 8 : Gating properties of hERG1A. The above panel displays the unique gating properties of hERG1A. Note that the rate of inactivation is faster as the membrane potential becomes more positive, essentially transitioning past the open state. At positive potentials, there is a decrease in hERG1A current because most channels are in the inactivated state 33
18 inactivate, but only through C type inactivation. Channel inactivation can happen one of two ways; C type or N type. N type inactivation is also called the ball and chain method where the N terminus moves up and occludes the pore regio n to stop the flow of ions. C type inactivation is different. C type inactivation happens by a constricting of the pore loop area of the channel, squeezing shut and thus stopping the flow of ions. Most C comp letely inactivate. However, C type inactivation for hERG1A is fast and at very positive potentials allows almost no potassium efflux 31 The hERG1A channel has been implicated in both inherited and drug induced arrhythmias 4,7,14,19,22,23,27,28,34 58 Mutations are associated with ventricular fibrillation, torsades de pointes (TdP), and arrhythm ias. Drug induced arrhythmias can be triggered by occlusion of the pore region by certain drugs, and/or by alterations in function of the channel by a mutation in hERG1A. Over two hundred hERG1A mutations have been characterized that cause disruption (re duction) of I Kr and potentially cause long QT syndrome (LQTS) 8,11,15,18,20 26 The resulting LQTS can be caused through multiple hERG1A dysfunctions. For example, mutations can lead to misfolding of the protein causing retention in the endoplasmic reticulum (ER) where rapid degradation will occur. Secondly, the mutation can produce dominant negative suppression or alter gating of the channel when there is co assembly of mutant
19 and wild type subunits. In either case, a loss of function occurs and a reduction in I Kr initiates the prolongation of phase II and III and an extension of the cardiac action potential. The extension of the action potential can lead to early after depolarizations (EADs) and arrhythmias. There is one known mutation in the hERG1A channel that causes a shortening of the QT interval 27 The shortening is due to a gain of function by abolishment of inactivation. The repolarization phase of the AP is more rapid an d ventricular fibrillation and arrhythmias can HERG1A is unique among voltage gated potassium channels in that it can be inhibited by many different types of drugs. HERG1A can be blocked by drugs including antibioti cs, antihistamines, antiarrhythmics, antipsychotics, and antimicrobials, directly leading to induced arrhythmias. Because of the diversity in structures among these drugs, establishing how and where hERG1A is blocked has proven difficult. Drug induced ar rhythmias are widespread and affect individuals of all age groups. Because prolonging the QT interval can lead to TdP, arrhythmias, and even death, pharmaceutical companies screen all potential drug candidates for hERG1A channel inhibition. HERG1A N glyc osylation HERG1A has two extracellular, N linked glycosylation consensus sites located in the S5 S6 linker region of the channel, very near the pore and selectivity filter.
20 The consensus sequence consists of NXS/T with N being Asparagine, X being any amino acid except Proline, and S/T being a Serine or Threonine. Through work completed by others and confirmed by our lab, the N linked glycosylation site at N629 is not glycosylated 44,59,60 It is thought that the N629 site is too close to the S6 segment and is not exposed to the glycosylation machinery to have glycans attached. Thus, the functional hERG1A tetramer has the capacity for four N glycosylation structures to be attached within the pore region. Known role of glycosylation on hERG1A Most research completed about the glycosylation of hERG1A has focused mainly on trafficking c oncerns for the channel. Originally, it was thought that N linked glycosylation was needed for proper trafficking of hERG1A to the surface of the cell. A repeat of the study by Zhou showed that N glycosylation is not needed for correct trafficking to the surface, but removal of N glycosylation altered the stability of the channel in the membrane over the course of twenty four hours 8,44,59,60 Glycosylation effects on other channels Many voltage gated ion channels are heavily glycosylated. All Na v and many K v channels are glycosylated, although in different locations on the channel and with differing amounts and types of glycans. Several studies have suggested that glycosylation, in particular, sialylation, may impact channel function 61 72 Gating
21 of Na v 1.4, K v 1.1, K v 1.2, and the Drosophila ShB channels are shifted to more depolarized potentials with a reduction in N glycosylation. In addition, data suggest that rat neonata l and adult atrial and adult ventricular Na v channels are more sialylated and therefore are more affected by changes in sialylation that neonatal ventricular Na v channels. Glycosylation pathway Glycosylation is a post translational modification. Glycosylation structures attached to mass. The addition and removal of glycans from proteins is completed by glycosidases, glycosyltransferases, and transport proteins (see Figure 9) The Conso rtium for Functional Glycomics recognizes more than 500 human genes involved in glycosylation. With the human genome being composed of approximately 30,000 genes, the genes required for the addition and removal of glycans ma ke up more than 1% of the entire genome 66 Two forms of glycosylation have b een identified: N linked and O linked glycosylation. The addition of glycosylation is not template driven, but it is a highly ordered process 73,74 Each step in the pathway determines the next step. N glycosylation begins in the endoplasmic reticulum (ER). An oligosaccharide precursor, composed of two core N acetylglucosamines (Glc NAc) and five mannose residues, is linked to a dolichol phophate on the cytosolic side of the
22 Figure 9 : Glycosylation pathway N glycosylation is a regulated pathway in which multiple proteins work collaboratively to add/remove sugars. The top panel shows the initial steps in the pathway, and the bottom panel shows the complex glycosylation that occurs in the Golgi. Figure from clemonslab.caltech.edu/n linked.html
23 ER. The lipid linked precursor is then flipped across the bilayer into the lumen of the ER. The precursor is complete with the addition of mannose and glucose structures. A protein complex, named oligosaccharyltransferase (OST), transfers the dolichol p recursor en bloc to an asparagine residue on a newly translated protein. Following the covalent attachment of the precursor to the asparagines residue, glucosidases remove all of the glucose residues. The glycoprotein (newly glycosylated protein) is tran slocated to the cis Golgi where the mannoses are trimmed down leaving only three. Formation of complex glycans takes place in the medial and trans Golgi with sialic acid (SA) often being the terminal glycan structure. Sialic acids are negatively charged at physiologic pH and can be attached to other sialic acids. These sialic acids have the potential of contributing to the negative surface charge of the protein. Congenital disorders of g lycosylation (CDG) Currently, there are twenty six forms of congeni tal disorders of glycosylation (CDG) characterize d 75 78 Of these twenty six forms, sixteen affect N linked glycosylation. CDG is a genetic disorder caused by glycosylation genes that are missing or mutated. These defects can alter multiple steps in the glycosylation pathway. CDG produces glycoproteins with decreased amounts and types of glycans. Multiple systems within the body ar e influenced by CDG including cardiovascular, muscular, and neuronal.
24 Chagas disease American trypanosomiasis is a human parasite disease that has no vaccine and no known cure 79 83 The parasite, named Trypanosoma cruzi infects its host and releases a trans sialidase to assist the transfer of sialic acid residues from the host to the parasite. The host tissue typically is the human heart, specifically cardiomy ocytes. Sialic acid is a terminal residue located on glycosylation structures. Many ion channels could be affected by this alteration in glycosylation. Symptoms from this disease include arrhythmias, myocarditis, cardiomyopathy, cardiac failure, and dea th. Glycosylation is differentially expressed and remodeled in the heart Cardiac glycogene expression i s h ighly r egulated 66 This was determin ed by measuring the relative expression of hundreds of glycogenes using GeneChip microarray analysis of mRNA isolated from neonatal and adult atria and ventricles Microarray data were validated using real time RT PCR (qPCR) analysis The microarray and qPCR data were consistent. The data showed that almost 50% of glycogenes tested were significantly differentially expressed among the four myocyte types studied. Additionally, g lycogene expression between atria and ventricles at a single developmental stage and throughout development of each chamber was a lso highly regulated The result of the removal of just one glycogene was enough to alter the conductance volt age relationship for Na v gating in atrial myocytes in mice.
25 Subunit vs cell specific effects of glycosylation N glycosylation can affect cardiac action potential waveforms in one of two ways: subunit or cell specific. Altering N glycosylation on a specifi c channel can modify the AP. Furthermore, modulating the glycosylation machinery available in a tissue can change the N glycosylation of all channels in the tissue and still have an effect on the AP. Summary It is well accepted that alterations to I Kr th at affect the action potential waveform can lead to arrhythmias. Our previous work with sodium and potassium channels showed that a change in sialylation modulates gating. In addition, certain ac unrest with a change in glycosylation. The hERG1A channel is heavily glycosylated and s tudies have shown glycosylation of hERG1 A may be needed for proper stability in the m embrane However, very little has been done to determine a specific role or mec hanism for N glycans in hERG1A activity. Here, we describe in detail, a role for N glycans on hERG1A gating and kinetics. We find that N glycans limit hERG1A activity, thereby causing an extension of the cardiac AP. In addition, we questioned the probab le modulation, by N glycosylation, of drug inhibition of three well prescribed antibiotics, Sulfamethoxazole (SMX), Erythromycin, and Penicillin G. This study will provide insight into understanding
26 the potential role of glycosylation for inherited, drug induced, and potentially the lives of hundreds of thousands of people every year and with millions of people living with life threatening arrhythmias daily, a greater awareness i s needed.
27 CHAPTER 2 MATERIALS AND METHODS Chinese hamster ovary cell culture and transfection Pro5, Lec2, and Lec1 cells were grown in minimal media and transfected with channel cDNA as previously described 84,85 The cells were plated at ~50% confluence onto 35 mm dishes 24 hours prior to transfect ion with a solution of 1 mL Opti MEM (Invitrogen), 8 L lipofectamine (Invitrogen), and 2.5 g of channel cDNA and incubated at 37 C in a 5% CO 2 humidified incubator. 24 hours post transfection, the transfection media was replaced with growth media, consi sting of alpha and without (Lec2) ribo and deoxyribonucleosides supplemented with 10% fetal bovine serum (FBS; Hyclone), 100 U/mL penicillin, and 100 g/mL streptomycin (Mediatech). Cells w ere incubated at 37 C for another 48 hours prior to electrophysiological recordings. Vector construction and mutagenesis The hERG1A cDNA in a pcD NA3 vector was a gift from Gail A. Robertson (University of Wisconsin). Vector construction and mutagenesis were performed similar to that previously described. The cDNA containing hERG1A open reading
28 frame was inserted into either a pIRES2 DsRED2 vector or a pIRES2 EGFP vector. The IRES2 vectors contain an internal ribosome entry site that allow both the gene of interest and the DsRED2 or EGFP gene to be translated from a single biscistronic mRNA. Electrophysiology and data analysis Whole cell recordings in CHO cells The Pro5/Lec2/Lec1 expression system, cell lines of Chinese Hamster Ovary cells, has been used successfully to determine the effects of glycosylation on channel gating. The Pro5 cell line produces normal N glycosylation and is the parental cell line for both Lec2 and Lec1. The Lec2 cell line is deficient in the CMP sialic acid transporter and produces essentially non sialylated proteins. This cell line serves as a model for CDG type IIf. The Lec1 cell line produces proteins in a mannose rich, or core N glycosylated, state because of a deficiency in the Gl cNAc T1 (N acetylglucosaminyltransferase I) causing a disruption in complex glycans being added to glycoproteins 85 Whole cell current recording s were performed using pulse protocols, solutions, whole cell patch clamp techniques, and data analyses as previously described. An Axon 200B patch clamp amplifier in combination with a CV203BU headstage (Axon Instruments) and a Nikon TE200 inverted micros cope were used. Pulse acquisition software (HEKA) operating on a Pentium III computer (Dell
29 Computers) was utilized for pulse protocol generation. The ensuing analog signals were digitized using an ITC 16 AD/DA converter (Instrutech). All experiments we re conducted at room temperature (~22 C). Drummond capillary tubes were pulled into electrodes with a resistance of 1 2 M using a model P 97 Sutter electrode puller. Series resistance was compensated 95 98%. The extracellular solution was (mM): 65 NaCl, 5 KCl, 1MgCl 2 2 CaCl 2 155 sucrose, 5 glucose, 10 Hepes (pH 7.3). The intracellular solution used was (mM): 70 KCl, 65 KF, 5 NaCl, 1MgCl 2 10 EGTA, 5 glucose, 10 Hepes (pH 7.3). For the drug perfusion studies, the extracellular solution was identical t o that listed above plus the addition of each concentration of drug listed in Table 1. Immediately prior to use, all solutions were filtered with a 0.2 m Gelman filter. To ensure complete dialysis of the intracellular solution, data was collected at lea st 5 minutes after attaining whole cell configuration. Conductance voltage relationship for CHO cell recordings Steady state and kinetic gating parameters were examined through the use of standard pulse protocols and solutions described by our lab and others. Cells were held at 80 mV, stepped to more depolarized potentials ( 80 mV to +50 mV in 10 mV increments) for 4 seconds, then stepped back to 50 mV for another 4 seconds, and returned to the holding potential. Steady state whole cell conductance values (G) were determined by measuring the peak current (I) of the tail current elicited at each test potential (V p ) and predicting a K + Nernst
30 Table 1. Concentrations for antibiotic drug study Drug Name Low Concentration Mid Concentration High Concentra tion Erythromycin 25 g/mL 100 g/mL 200 g/mL Penicillin G 10 g/mL 50 g/mL 200 g/mL Sulfamethaxozole 50 g/mL 300 g/mL 600 g/mL
31 equilibrium potential (E k = p E k )). The maximum conductance generated by each cell was used to normalize the data for each cell to its maximum conductance by fitting the data to a single Boltzmann distribution (equation 1, solving for maximal conductance). These single Boltzmann distri butions were used to determine the average V a SEM and K a SEM values. The normalized data were averaged with those from the other cells and the resulting average G V curve was fit via least squares using the Boltzmann relation below: Fraction of maximal conductance = [1 + exp ((V V a )/K a )] 1 equation (1) where V is the membrane potential, V a is the voltage of half activation, and K a is the slope factor. Steady state inactivation Cells were held at 80 mV before stepping to +20 mV for 3 seconds. Then we stepped to 120 mV to +60 mV (in 10 mV increments) for 30 ms before stepping back to +20 mV for 1 second and then returned to the resting potential. The maximum current generated by each cell was used to normalize the data for eac h cell to its maximum current by fitting the data to a single Boltzmann distribution (equation 2, solving for maximal current), from which the mean V i SEM and K i SEM values were determined. Fraction of maximal current = y o + [1 + exp ((V V i )/K i )] 1 equation (2) where V is the membrane potential, V i is the voltage of half inactivation, K i is the
32 slope factor, and y o is the fraction of current that comes from constitutively active channels. Window Current The product of the two Boltz mann distributions for steady state activation and inactivation. See formulas above. Time constants for deactivation ( T n ) The deactivation protocol was set up to start at the membrane potential of 80 mV and then step to +20 mV for 1.6 seconds before stepping from 40 mV to 100 mV (in 10 mV increments) for 6 seconds each and then returning to the resting potential. Deactiva tion time constants were determined by fitting the current trace described above for each voltage tested. Whole current traces were fitted using the Hodgkin Huxley function of the PulseFit software suite (HEKA). The current trace data were fitted to an e xponential function. Computer simulation of action potential The Tentusscher Noble Noble Panfilov model of human Endocardial ventricular myocyte was implemented to simulate the AP under four experimental conditions, namely Full glycosylation, Reduced sialylation, N Glycan a se treated, and Mannose Rich 1 The extracellular concentration of K + Na + and Ca 2+ are 5 mM, 140mM and 2mM respectively. The initial values of intracellular
33 concentration of K + Na + and Ca 2+ are 138.3 mM, 11.6 mM, and 0.0002 mM. The rapid delayed rectifier current I Kr is described by the following equat ion, where x r 1 is an activation gate and x r 2 is an inactivation gate. K o /5.4 represents the K o dependence o f the current 1 Based upon the V a K a V i K i statistics from our voltage clamp experiments under four HERG +/ sugars treatments, we computationally fit the model steady state activation and inactivation curves. Whole cell homogenization Cells were rinsed with cold PBS and incubated for 5 minutes i n ice cold sodium pyrophosphate buffer with protease inhibitors (PI; 20 mmol/liter tetrasodium pyrophosphate, 20 mmol/liter Na 2 PO 4 1 mmol/liter MgCl 2 0.5 mmol/liter iodacetamide, 1.1 mol/liter leupeptin, 0.7 M pepstatin, 76.8 nM aprotinin). Cells were then homogenized using manual tissue grinders. The homogenates were centrifuged for 10 minutes at 1000 x g in a Beckman bench top centrifuge. The supernatant was centrifuged in an Eppendorf ultracentrifuge for one hour at 50,000 rpm after which, the pell et was resuspended in an appropriate volume of sodium pyrophosphate buffer containing PIs. The lysates were then stored at 80 C and protein levels were determined using the Pierce BCA Protein Assay kit and a Beckman DU 530 spectrophotometer.
34 Cell surfa ce biotinylation Cell surface protein isolation (Pierce) was performed on hERG1A expressed in Pro5, Lec2, and Lec1 cell lines. The cells were grown in T75 flasks until ~95% confluent and then biotinylated for 30 minutes at 4 C. Then the reaction was quen ched, the cells harvested and lysed. Next, the biotinylated proteins were PAGE buffer plus 50mM DTT. The protein sample at the end of the reaction is then run on a SDS PAGE gel (see below ) and probed with hERG1A antibody to detect if the hERG1A channel is found of the surface of the cell. Immunoblots Immunoblot gel shift analysis was performed as previously described 65,84,86,87 Cell homogenates (1 3 g/lane) were combined with one volume of 2x sample buffer (12.5% upper Tris buffer, 10% glyce rol, 5% 2 mercaptoethanol, and 3% sodium dodecyl sulfate) and denatured in boiling water for 3 minutes. Samples were then run on 5 7.5% SDS PAGE gel for 90 110 minutes at 75 110 mV and then transferred on nitrocellulose membranes using a semi dry transfer cell (BioRad). HERG1A was detected using a polyclonal primary antibody, HERG C 20, raised in goat (Santa Cruz Biotechnology). After incubation with primary antibody, the blot was treated with a Swine Anti Goat conjugated to HRP (Southern Biotech), secon dary antibody, and visualized using an enhanced chemiluminescence kit (Pierce). Deglycosylation of homogenates was
35 performed through 2 hour treatments at 37 C with 5 mU of N Glycanase /9 g of protein (PROzyme). Data analysis and statistics Data and stat istical analyses were performed using Pulse/PulseFit (HEKA), Sigma Plot (SSPS inc.), and Microsoft Excel. Student T test, ANOVA, and Tukey test were performed on the data produced by the electrophysiology recordings. Significance was determined by p<0.05 w here applicable.
36 CHAPTER 3 N GLYCANS LIMIT HERG1A ACTIVITY Human ether a go go related gene 1 (hERG1), is the pore forming subunit of the voltage gated potassium channel whose activity is responsible for I Kr a key component of late phase II and phase III of the human ventricular cardiac action potential (AP) 17,18,88 A change in I Kr can alter the repolarization phase of the AP that can l ead to modulated cardiac conduction. This altered conduction would be reflected in an ECG by a change in the QT interval HERG1 dysfunction is responsible for Long QT2 disorders and a Short QT disorder 16,19,89 Abnormal hERG1 activity can lead to arrhythmias, torsades de poin tes, cardiac disease, and sudden cardiac death 7,16,20,21 The cardiac ventricular AP is comprised of 5 phases 3 The rising phase, or phase 0 is produced by the rapid activation of voltage gated Na + channels that then inactivate, shutting down the rapid depolarization. During this phase, hERG1A channels are beginning to transition out of the closed state to either the open or inactiv at e d states. Phase 1 begins with a sharp, brief repolarization that occurs due to a transient outward K + current I to The plateau, or phase 2, is produced by combating K + efflux and Ca ++ influx such that the membrane p otential remains
37 almost constant. During phase 2 depolarization, hERG1A channels start to recover from inactivation by transitioning to their open state. A positive feedback loop begins as K + leaves the cell, which will further repolarize the membrane and lead to the opening of additional hERG1A channels that are recovering from inactivation, thereby further repolarizing the membrane. Evidence for this positive feedback mechanism is observed by the steep slope of Phase 3. Phase 4 is the final recovery to the resting membrane potential before another AP begins. HERG1A channels do not inactivate through fast N type inactivation, but through C type mechanisms 90 Typically, C type inactivation is thought to involve pore constriction and is slower than N type inactivation. However, hERG1A inactivation is fast and voltage dependent and recovery from inactivation typically occurs rapidly with the transition from the inactivated back into the open cha nnel state. This leads to increases in the rate of AP repolarization. Thus, any change in hERG1A voltage dependent gating and/or kinetics would lead to a change in channel activity during the AP, and likely alter the rate of AP repolarization. Ion channel comprised of glycan structures 69,91 94 Our l ab and others describe how changes in glycosylation and sialylation can affect gating and kinetics of voltage gated Na + and K + channels 65,68 71,87,93 97 The addition and removal of glycans from proteins is completed by the activity of >200 glycosidases, glycosyltransferases, and
38 transport proteins (glycogenes). In a r ecent report, we showed that the cardiac glycome, defined as the complete set of glycan structures produced in the heart, varies between atria and ventricles, and changes differentially during development of each cardiac chamber 98 Specifically, we showed that nearly half of the 239 glycogenes tested were significantly differentially expressed among neonatal and adult atrial and ventricular myocytes. The N glycan structures produced among cardiomyocyte types were markedly variable. We went on to show that the regulated expression of a single g lycogene was sufficient to modulate AP waveforms and gating of less sialylated voltage gated Na + channels consistently. These data provide initial insight into the possibility of a newly described mechanism for controlling and modulating cardiac excitabili ty through regulated changes in cardiac glycosylation. HERG1A has two extracellular, N linked glycosylation consensus sites, N598 and N629, which are located in the S5 S6 linker region of the channel (or the pore region) and very near the selectivity filt er. W ork done by others and confirmed by our lab, showed that the N629 glycosylation site is not glycosylated 60 It is believed to be too close to the S6 domain and therefore not exposed to glycosyltransferases as the newly synthesized channel moves through the endoplasmic reticulum and Golgi. However, the N598 glycosylation site is glycosylated Here we question whether and how altera tions in the types
39 and amounts of N glycans attached to hERG1A can modify channel gating and kinetics. Results H ERG1A is differently glycosylated We questioned the impact of four variable glycosylation patterns on hERG1A function. The Pro5, or parental cell line, is essentially fully glycosylated 85 The Lec2 cell line is deficient in a CMP sialic acid transporter and produces glycoproteins with essentially no sialic acid, and serves as a model for CDGIIf 75 77,85 The Lec1 cell line is deficient in GlcNAc T1 (N acetylglucosaminyltransferase I) causing a disruption in the complex N glycans being added to glycoproteins 85 As a result, a mannose rich glycosylation structure is produced. Exemplary N glycan structures that are putatively produced in each cell type are shown in Figure 10 The fourth glycosylation pattern was ach ieved through enzymatic deglycosylation, using N glycanase to cleave the full N glycan structure at the Asparagine residue. Our previous work enzymatic treatments to study channel function under variable levels of channel glycosylation 65,70,71,87,94 96 To demonstrate that hERG1A is glycosylated differently and is expressed on the cell surface under each condition of glycosylation, we performed immunoblot
40 Figure 10. HERG1A is differently glycosylated. A: Immunoblot of hERG1a expressing Pro5 cell lysates N Glycanase treatment. Lane 1: Untreated control. Lane 2: N glycanase treated. B: Surface biotinylated lysates extracted from Pro5 (Lane 1), Lec2 (Lane 2), and Lec1 (Lane 3) cells expressing hERG1A. Typical predicted schematic structures of N glycans produced i n each cell line.
41 analysis. As seen in Figure 10 A, for hERG1A expressed in Pro5 cells, treatment with N glycanase causes a reduction in predicted molecular weight (MW), and a less diffuse banding pattern, suggesting that N glycanase completely removes the N glycan structure. Analysis of surface channel expression shows that altered glycosylation does not affect channel trafficking of hERG1A in any of the cell lines tested. The shifts in MW between Pro5, Lec2, and Lec1 suggest that each cell l ine glycosylates hERG1A differently, and is expressed on the cell surface Further, note the heterogeneous banding pattern of the Pro5 and Lec2 lysates, and the more homogeneous pattern of the Lec1 and N glycanase treated lysates, indicating the variabili ty of glycosylated hERG1A protein expressed on the cell surface. H ERG1A function is modulated by N glycans hERG1A expressed in each cell line and following N glycanase treatment produced typical hERG1A currents as illustrated in Figure 11 Variable glyco sylation had no significant effect on the hERG1A current densities. However, note the distinct characteristics of the current traces among the differently N glycosylated hERG1A channels. As an example for each condition of reduced glycosylation, note the larger and faster (deactivating) tail currents at 50 mV following the four second test pulses. Details of the changes in hERG1A biophysical characteristics are d iscussed below and summarized in Table 2.
42 Figure 11. HERG1A current is altered by N glycans Whole cell current traces for hERG1A expressed in Pro5, Lec2, Lec1, and N Glycanase treated Pro5 cells. Cells were stepped to increasingly depolarized potentials for 4 s from the 80 mV holding potential, and then to 50 mV for 4 s to record the tail current.
43 Table 2. Biophysical parameters for hERG1A as expressed in Pro5, Lec2, Lec1, and following N Glycanase treatment Table 2. Data are the mean SEM. V a : Voltage of half activation. K a : Boltzmann activation slope factor. V i : Voltage of half inactivation. K i : Boltzmann inactivation slope factor. = significance ( p <0.05).
44 N glycans promote hERG1A voltage dependent activation There is a consistent rightward shift in the current voltage (I V) relationships for hE RG1A under conditions of reduced glycosylation (see Figure 12 ). Note that the full I V curves were shifted by a depolarizing 5 10 mV along the voltage axis for hERG1A expressed under conditions of reduced sialylation (Lec2) reduced complex N glycosylatio n (Lec1) and when fully de glycosylated (N glycanase treatment) These data indicate that N glycans shift the activation voltage for hERG1A to more hyperpolarized potentials. The hERG1A steady state activation (SSA) relationships are shifted to more depolarized potentials under conditions of reduced glycosylation To determine more rigorously the impact of N glycans on hERG1A activation, the SSA relationships were determined using tail current analysis (Figures 11 and 12 ). There was a consistent, 8 12 mV depolarizing shift in the SSA relationships for hERG1A expressed under each condition of r educed glycosylation (Figure 12 Table 2). As Table 2 indicates, the voltages of half activation (V a ) measured were also 8 12 mV more depolarized for less glycos ylated hERG1A. The slope factors of the SSA relationships, K a were unaffected. This suggests that N glycans modulate hERG1A activation causing a linear shift in the SSA relationship.
45 Figure 12. H ERG1A steady state activation curve is shifted to more hyperpolarized potentials with full N glycosylation. Current voltage (I V) and steady state Activation curves for hERG1A expressed in Pro5 (square), Lec2 (circle), Lec1 (triangle) and N Glycanase treated Pro5 (upside down triangle) c ells. Data are mean SEM. Top: IV curves. Lines are point to point. Bottom: Steady state Activation (SSA) curves. Lines are fits of the data to single Boltzmann distributions. Vertical, dotted lines project to the V a for each conditi on. n= 11, 11, 6, 3 (See Table 2 ).
46 N glycans act to slow closing of hERG1A Time constants for hERG1A deactivation were determined under each condition of N glycosylation. As observed in Figure 13 the deactivation time constants for the fully glycosylated hERG1A are sign ificantly greater than those observed for the less glycosylated channels. This is most notable at less hyperpolarized voltages, likely caused by the strong voltage dependence of hERG1A deactivation. N glycans cause a greater shift in voltage dependence o f steady state channel availability To determine whether hERG1A channel availability is affected by N glycans, steady state channel inactivation (SSI) curves were measured (Figure 14 Table 2 ). Note the large, significant, depolarizing shift in steady st ate channel availability under conditions of reduced glycosylation (21 30 mV). Specifically, the voltage dependence of half inactivation (V i ) for hERG1A under conditions of reduced sialylation was 21 mV more depolarized than the fully glycosylated hERG1A ( Table 2 ). This depolarization of V i was larg er for hERG1A expressed in a mannose rich state (in Lec1 cells, ~ 30 mV), or followin g full deglycosylation (~22 mV ). The less glycosylated hERG1A channels inactivated at voltages significantly more d epolarized than the fully glycosylated control channels. However, there was no significant difference in V i measured among the less glycosylated channels.
47 Figure 13. N glycans act to slow hERG1A closing Top panel: H ERG1A deactivation time consta nts as expressed in Pro5 (squares), Lec2 (circle), Lec1 (triangle), and N Glycanase treated Pro5 (upside down triangle) cells. Data are the mean SEM deactivation time constant at a membrane potential. Lines are point to point. Inset: Sample whole cell currents measured usin g the deactivation protocol. Lower panel: Mean SEM deactivation times constants at 50 mV for Pro5 (small slanted lines), Lec2 (mesh), Lec1 (large slanted lines), and N Glycanase treated Pro5 (horizontal lines) cel ls. n= 11, 11, 6, 3 (See Table 2 ). Significance (*) was determined by comparing Pro5 with each condition of reduced glycosylation (p<0.05).
48 Figure 14. Steady state channel availability is altered by N glycans. S teady state inactivation (SSI) curves for hERG1A expressed in Pro5 (square), Lec2 (circle), Lec1 (triangle), and N Glycanase treated Pro5 (upside down triangle) cells. Data are mean SEM Lines are fits of the data to single Boltzmann distributions. n= 11, 11, 6, 3 (See Table 1). Inset: Typical whole cell inactivating current traces from a hERG1A expressing Pro5 cell.
49 The large and variable shift in the volta ge dependence of activation and inactivation causes a shift to more physiological potentials and an increase in hERG1A window current HERG1A activation, deactivation, and inactivation voltage dependence were each modulated by changes in the glycosylation state of the channel. For each gating mechanis m studied, N glycans caused a hyper polarizing shift in the voltage dependence of channel gating. However, the magnitude of the shift in steady state activation and inactivation were not uniform; channel availability was shifted about twice as much along the voltage axis as was channel activation. The combined distribution of channels betwee n the active and inactive states at a membrane potential will affect the density of channels that are active at that membrane potential. The area under the overlapping portions of the SSA and SSI curves represents the window current, or the voltages at wh ich hERG 1 channels are active. Figure 15 plots the window currents measured under the four conditions of glycosylation. The bell shaped curves are the product of the two Boltzmann relationships for steady state activation and inactivation, and represent the predicted persistent channel activity (Fig. 15B) Because both the activation and inactivation curves shifted to more depolarized potentials for less glycosylated channels, the window current will also be shifted to more depolarized potentials as obse rved. Note that the peak of the window current is 10 15 mV more depolarized for the less glycosylated hERG1A channels compared to the fully glycosylated channels. In Figure 15C the peak of each
50 Figure 15 N glycans limit hERG1A window currents with persistent hERG1A activity at more hyperpolarized potentials A: Overlapping SSA and SSI curves for hERG1A as expressed in Pro5 (solid line), Lec2 (long dash line), Lec1 (short dash line), and N Glycanase treated Pro5 (d otted line) cells. Curves are the Boltzmann fits from Fig. 12 and Fig. 14. B: Predicted window currents for hERG1A. Lines are as described in panel A. C: Peak predicted normalized window current. Bar graph comparing the relative magnitude (percent active channels) of the peak window current for each condition.
51 curve is greater for the less glycosylated channels, increasing by 1.6 fold for Lec2, 2.4 fold for Lec1, and 1.4 fold for enzymatically deglycosylated cells. This suggests an increase i n window current with a reduction in N glycans. Estimates made to determine the area under the curve predict a 1.3 2.5 fold increase in the amount of potassium moving across the membrane for the less glycosylated hERG1A. To further examine the hERG1A cha nnel persistent current, the steady state currents following a four second pulse to voltages near the I V peak for the less glycosylated channels (0, 10, and 20 mV) showed a 23 80% increase in window current compared to fully glycosylated controls (data no t shown). With reduced levels of N glycosylation and sialylation, channels recovering from inactivation during the AP would occur at smaller repolarizations and in greater numbers, thus increasing the rate of AP repolarization. Simulation of the ventricu lar AP predicted exactly th is change in AP waveform (Fig. 16 ). That is, simulation of the rightward shift and increase in hERG1A window current caused increased phase II/phaseIII repolarization that would lead to reduced AP duration. Thus, additional N g lycans, specifically, sialic acids, would act to limit hERG1A activity during the AP causing an increased AP duration as discussed later.
52 Figure 16. in silico modeling predicts ventricular action potential is extended by increased N glycosylation Predicted ventricular action potential waveforms, assuming hERG1A activity as observed under each condition of glycosylation. Computer simulation of the cardiac AP was done in collaboration with Dr. Hui Yang, Department of Industrial & Management Systems Engineering, USF
53 Discussion N Glycans lead to a loss of hERG1A function Here we show that N glycans, particularly sialic acids attached to N glycans, alter gating and kinetics of hERG1A. Previous work on hERG1A and N glycosylation focused on trafficking and stability of the channel in the membrane 59,60 This work represents the first set of data questioning the role of N glycans on hERG1A gating and kinetics. For all parameters studied, N glycans shifted voltage dependent gating to more negative potentials. The data show hERG1A gating is modulated similarly under three different conditions of reduced glycosylation ranging from a modest reduction of sialic acids to the complete removal of N glycans following N glycanase treatment. The data show that reduced N glycosylation achieved through expression in Lec2 or Lec1 cells, or through enzymatic deglycosylation, each altered channel activation and inactivation similarly, with significant rightward shifts in steady state activation, steady state inactivation, and deactivation kinetics observed for each condition of reduced glycosylation. However, no significant differences in steady state activation or steady state inactivation were observed among the three conditions of reduced glycosylation. This suggests that the sialic ac ids attached to hERG1A N glycans are responsible for the majority of the effect of N glycans on channel function.
54 While the direct effects of changes in channel glycosylation are similar to our previous work on voltage gated Na + and K + channels, the i mpact on AP waveform with such changes in glycosylation are novel channel N glycans limit hERG1A activity, thereby acting to extend the cardiac AP The rightward shifts seen for both steady state activation and inactivation under each condition of reduc ed glycosylation would lead to a depolarizing shift in channel window current (Fig. 15 A). The larger rightward shift in the inactivation curve would also cause an increase in the magnitude of the window current (Fig. 15 B). For l ess glycosylated hERG1A ch annels, a greater number of channels would be re activated at depolarized potentials (less repolarized) compared to the fully glycosylated channels. That is the reduction in hERG1A glycosylation or sialylation leads to a gain of hERG1A function. The pos itive feedback nature by which hERG1A reactivates would lead to the rapid recovery of hERG1A, and further increase K+ efflux at earlier times during the AP, and likely reduce AP duration as observed through simulation of the ventricular AP (Fig. 16 ). As w ill be discussed in detail below, as channel sialylation is regulated, or when it is aberrant, hERG1A activity will change, thereby impacting AP repolarization.
55 CHAPTER 4 PENCILLIN BLOCK OF HERG1A H ERG 1A blockade by a range of drugs is linked to acquired long QT syndrome ( aLQTS) 35,41,46,48 50,99 This result s in an increased risk of sudden death from ventricular arrhythmias. Several medications have been withdrawn from the pharmaceutical market due to causing QT prolongation 100,101 The successful development of potential new drug therapies hinges on finding drugs that do not induce cardiac arrhythmia s. The most common problem in acquired long QT syndrome is blocking hERG1A delay ing cardiac repo larization and increasing the risk of torsades de pointes (TdP). Previously, it was known that some medications that were used to trea t cardiac arrhythmias were also considered proarrhythmic. T he Cardiac Arrhythmia Suppression Trial (CAST) of 1991 inves tigated class I anti arrhythmic agents that block ed Na V channels and reduce d conduction velocity of electrical impuls es in the ventricular myocardium I t became apparent that non cardiac drugs, including antibiotics, antipsychotics, antidepressants and antimalarials can also cause TdP TdP is a twisting of the QRS complex around the isoel ectric line of
56 the body surface e lectrocardiogram (ECG) and leads to an increased incidence of sudden death. Once it was known that non cardiac drugs could block hE RG1A and potentially alter the cardiac AP, a flood of research went into understanding what can block hERG1A and how to deduce new structures that would be anti arrhythmic. Understanding what current drugs on the market block hERG1A and at what concentrati ons is essential to correctly prescribing medications to patients. The discovery of penicillin is attributed to Scottish scientist and Nobel laureate Alexander Fleming in 1928 (see Figure 17) This observation began the modern era of antibiotic s. Penicil lin is a lactam antibiotic It works by inhibiting the formation of peptidoglycan cross links in the bacterial cell wall. The mechanism of action of penicillin at the molecular level starts with the binding of penicillin to penicillin binding proteins (PBPs) which are located in the cell wall. Some PBPs are inhibitors of cell autolytic enzymes that literally eat the cell wall and are most likely necessary during cell division. Other PBPs are enzymes that are involved in the final step of cell wall syn thesis called transpeptidation. These enzymes are outside the cell membrane and link cell wall components together by joining glycopeptide polymers together to form peptidoglycan. The bacterial cell wall owes its strength to layers composed of
57 Figure 17. Structure of Penicillin G. Molecular structure of Penicillin G. Formula weight for Penicillin G is 334 g/mol. pKa is~3.
58 peptidoglycan This is a complex polymer composed of alternating N acetylglucosamine and N acetylmuramic acid a s a backbone from which a set of identical tetrapeptide side chains branch, and a set of identical peptide cross bridges also branch. The tetrapeptide side chains and the cross bridges vary from species to species, but the backbone is the same in all bact erial species. Each peptidoglycan layer of the cell wall is actually a giant polymer molecule because all peptidoglycan chains are cross linked. In gram positive bacteria there may be as many as 40 sheets of peptidoglycan, making up to 50% of the cell wall material. In gram negative bacteria, there are only one or two sheets (about 5 10% of the cell wall material). Penicillin G has high activity against gram positive bacteria and low activity against gram negative bacteria. Penicillin acts by inhibiting peptidoglycan synthesis by blocking the final transpeptidation step in the synthesis of peptidoglycan. It also removes the inactivator of the inhibitor of autolytic enzymes, and the autolytic enzymes then lyse the cell wall, and the bacterium ruptures. T his latter is the final bacteriocidal event. Penicillin G was the first discovered penicillin in what is now a long list of pe nicillins It is absorbed orally with about 2/3 of the dose being degraded by stomach acids. Penicillin G is very susceptible t o beta lactamases, enzymes which inactivate the compound by degrading the beta lactam ring. A ll penicillins
59 are acidic drugs that are acid labile and will degrade in neutral or basic solutions. The degradation opens the beta lactam ring and results in ina ctivation of the penicillin. Penicillins are generally safe, having low toxicity, and can be used in pregnancy and during lactation. Penicillin G is considered a short acting drug After oral administration, penicillin is absorbed mainly from the duodenum and upper jejunum. The extent of absorption depends on the presence of food in the gastrointestinal tract, gastric and intestinal pH and the relative acid stability of the penicill in derivative. Penicillin G is acid labile and should be taken on an empty stomach. Peak serum levels are reached within 30 to 60 minutes. P enicillins are readily distributed into ascitic, synovial, pleural and pericardial fluids. Distribution into tis sues varies widely, with highest amounts in the kidney and lower concentrations in the liver, lungs, skin, intestines and muscle. Small amounts are found in all other body tissues and in the CSF. When the meninges are inflamed, the CSF concentration is a bout 5% of the serum concentration and can be therapeutic against sensitive organisms. Penicillins readily cross the placenta and are distributed into breast milk. In patients with normal kidney function, penicillin is excreted rapidly by filtration and active tubular secretion. The elimination half life is about 30 minutes. HERG1A has been shown to be blocked by numerous antibiotics, including erythromycin, sulfa drugs, and grepafloxacin. Grepafloxacin has been removed
60 from the market due to QT prolong ation and death. With all of the studies being done on hERG1A drug block, no one has published any findings on the very first antibiotic, Penicillin G. Here, we show that Penicillin G blocks hERG1A 102 Results To determine whether Penicillin G blocks hERG1A, hERG1A was expressed in the fully glycosylated Pro5 cell line. Pro5 cells were exposed to three differ ent concentrations; 10 g/mL, 50 g/mL, and 200 g/mL. Penicillin G was perfused at 0.5 mL/min using the typical extracellular solution as the vehicle. Since Penicillin G is readily absorbed, we did not need to add it to the intracellular solution. Five minutes of equilibration time was allowed before recording. All recordings were conducted while perfusing on solution. HERG1A is blocked by Penicillin G Initially, we were using Penicillin G as a potential negative control, since extensive literature sea rches yielded no reports on Penicillin block of hERG1A. What was found was that, at therapeutic concentrations, hERG1A function was blocked by Penicillin G (see Figure 18). Maximal conductance of the 10, 50, and 200 g/mL concentrations were normalized to the zero concentration. Figure 19 is a bar graph reflecting the % block at each concentration.
61 Figure 18. Current trace at +20 mV for hERG1A at multiple drug concentrations Current trace showing block of hERG1A activity by Penicillin G at each concentration tested (30 M, 150 M, and 600 M).
62 Figure 19. Percent block by Penicillin G The total block by Penicillin G increases with increasing concentrations of the drug. The h igh concentration of Penicillin G shows a 35% block of hERG1A current.
63 Discussion For the first time, it was shown that Penicillin G blocks hERG1A function at therapeutic concentrations. With more than 1/3 of hERG1A function blocked at the 200 g/mL con centration, the normal AP could be seriously compromised. This blockade of hERG1A could potentially alter the cardiac action potential and cause an acquired QT prolongation. Although this is the first report that shows Penicillin G blocks hERG1A function we cannot overlook the fact that others have not shown any harmful side effects of Penicillin G on cardiac function. There could be multiple reasons. The first rationale is Penicillin G can be harmful when administered to patients. Either no one has ci ted problems in the literature or these concerns have not been attributed to hERG1A blockade. Although Penicillin G is a good antibiotic, it does have its drawbacks. It is short acting and can easily be inactivated by stomach acid. Normal administration of the drug is by injection. Even then, Penicillin G is not used as readily as other antibiotics (many have an allergy to it). As a result, there may not be enough data to confer whether there are dangerous consequences to Penicillin G use. The second point goes back to remodeling. The body has a wonderful way of adapting to changes and unsafe environments. Penicillin G may block hERG1A,
64 but that does not mean it will automatically cause QT prolongation. There are multiple channels at work during the repolarization phase of the cardiac AP. These additional sodium, calcium, and potassium channels may be up regulated or altered to balance out the blockade of hERG1A. Consequently, hERG1A may be altered itself to change how a drug blocks it o r how it works during the AP. The third concern is that the harmful effects of hERG1A blockade are not seen considering the half life of Penicillin G is around thirty minutes. This may not be adequate time for the drug to reach the heart and cause hERG1A dysfunction. Additionally, it may reach the heart, but the thirty minute timeframe is not sufficient to cause block of the channel. We show that direct application of the drug to the outside of the cell for as little as five minutes is ample time to caus e block even with a low concentration of 10 g/mL. However, cell expression systems do not completely replicate events occurring in vivo. They are merely an excellent tool to begin looking at a question. Additional studies should be conducted to properl y comprehend the blockade of hERG1A by Penicillin G. Any alterations in the cardiac action potential can be harmful. Penicillin G can now be added to the long list of drugs that block hERG1A function. These changes will likely have a detrimental effect on normal cardiac rhythm. This could result in arrhythmias, TdP, and sudden cardiac death.
65 CHAPTER 5 N GLYCANS ALTER ANTIBIOTIC BLOCK OF HERG1A In the last ~ 12 years, significant efforts were made to inform the world of ult, three important guidelines were released named the Points to Consider document, ICH S7A and ICH S7B guidelines The Points to Consider document was released in 1997 by the Committee for Proprietary Medicinal Products (CPMP). It stated facts about th e assessment of QT prolongation by non cardiac medications. This was the beginning of global understanding of potential dangers of new and current drugs that has forced pharmaceutical companies to establish an improved cardiac safety profile. In 2000, the ICH S7A guideline was introduced by the International Conference of SAFETY PHARMACOLOGY STUDIES FOR HUMAN PHARMACEUTICALS thirteen page document whose objective is to protect participants and patients of clinical trials from potentially adverse effects of pharmaceuticals, while avoiding unnecessary resources and animal use. The guideline goes on to dictate the tests needed to be done for all of the physiological systems in the body. The I CH THE NON CLINICAL EVALUATION OF THE POTENTIAL FOR DELAYED VENTRICULAR REPOLARIZATION (QT INTERVAL
66 PROLONGATION) BY HUMAN PHARMACEUTICALS in 2005. This guideline describes a non clinical testing strategy f or assessing the potential of a test substance to delay ventricular repolarization (see Figure 20) This guideline includes information concerning non clinical assays and integrated risk assessments. The ICH S7B expands upon the ICH S7A guideline. The ob jective of the newest guideline is to identify whether a test substance has the potential to delay ventricular repolarization and to correlate the extent of delayed ventricular repolarization to the concentrations of the test substance The results can be used to reveal a mechanism of action and potentially estimate risk for delayed ventricular repolarization and QT interval prolongation in humans. T he ICH S7B is a preclinical safety strategy that all pharmaceutical companies comply with. The first step is an in vitro I Kr assay to test for potential hERG interactions by means of patch clamp studies. Taken as a whole, t hese documents assess and set forth steps that pharmaceutical companies take to est ablish a safety profile regarding cardiac even ts for all new drug candidates and already marketed drugs that have been found to have a dverse clinical events, a new patient population, or a new route of administration raises concerns not previously addresse d Hundreds of drugs have been found to directly affect the v entricular repolarizing current I Kr reduc e K+ efflux, and lead to an arrhythmia 35,54,55,103 105 This toxicity is widespread, as indicated by analysis of a drug monitoring database
67 Figure 20. ICH S7B guideline The ICH S7B guideline dictates strategies for testing pharmaceutical drug candidates for possible arrhythmic events.
68 maintained by the World Health Organizat ion 106 108 This is why the ICH S7B guideline requires an in vitro assay on the main repolarizing current to test for potential drug interactions. It has been found that 2 9% of patients taking quinidine, an antiarrhythmic, produces TdP 46 As a result, there have been certain drugs taken off the market. From 1997 to 2000, three drugs were removed from patient use. The first drug, Te rfenadine was in 1997. It is an antihistamine form erly used for the treatment of allergic conditions. Terfenadine is a prodrug generally completely metaboliz ed to the active form fexofenadine by intestinal CYP3A4. Terfenadine itself, however, has a ca rdiotoxic effect and may be absorbed and reach myocytes if the patient is concurrently taking a CYP3 A4 inhibitor (e.g. erythromycin or grapefruit juice) In 1999, Grepafloxacin, an antibiotic was removed from the market due to increased QT prolongation. This antibiotic i s in a class of drugs called fluoroquinolones and fights bacteria in the body In 2000, Cisapride was withdrawn because of reports of causing Long QT in patients. Cisapride is a parasympathomimetic which acts as a serotonin 5 HT4 agonis t. It is used f or th e symptomatic treatment of patients with nocturnal heartburn due to gastroesophageal reflux disease Cisapride was associated with 270 adverse cardiac effects, specifically cardiac arrhythmias, and there have been 70 fatal reactions. To better understand how I Kr is altered by drug block, we need to understand how hERG1 A is blocke d by a drug. Figure 21 describes the potential
69 Figure 21. Potential hERG1A block by antibiotics Drugs can block hERG1A several ways; 1) from the outside (in the case of SMX) and 2) from the inside, interacting with the aromatic structures (as does Erythromycin). The mechanism that Penicillin G uses to block hERG1A has not yet been elucidated.
70 mechani sms to how a drug can block hERG1A. The most common areas are either outside or inside the cell. Blocking the flow of potassium through hERG1A from the outside of the cell is not fully understood. Most work has been completed with divalent cations, sinc e so many reduce hERG current 109 112 One thought is that the drug (divalent cation) may physically occlude the pore region, essenti ally sitting in the opening of the extracellular side of the channel. Since hERG1A is inactivated by C type inactivation, it is thought that the external side of the pore for hERG1A is larger than other K + channels. Another externally oriented blocking m echanism may involve the acidic residues on the S2 and S3 segments, specifically D226 and D456, along with the first four arginine residues on S4. These residues work together during the activation and deactivation of the channel through an electrostatic interaction. The divalent cations may bind to or screen sites on the hERG1A channel altering the negative electrostatic potential causing alterations to the gating and kinetics of hERG1A, thereby modulating current. This mechanism is called a surface cha rge theory. An additional theory is the voltage dependent block theory where the drug binding site is within the pore and the drug must cross part of the electric field of the membrane to block hERG1A. Because the drug must cross the electric field
71 the binding and unbinding is voltage dependent. Drugs are not limited to utilizing one of these theories, but can use a combination of any of the theories. As for block from the intracellular side, it has been shown that hERG1A has structural features that assist the binding of drugs better than other K + channels. HERG1A channels are blocked by drugs with diverse structures that cover several drug classes, including antiarrhythmic, psychiatric, antimicrobial and antihistamine 48,110,113,114 Because of this unusual susceptibility to blockage by drugs, it is thought that hERG1 A has multiple unique binding site s An Ala scanning mutagenesis approach was used to identify residues of hERG 1A that interact with several drugs inside the pore region Residues within the pore wer e individually mutated and the n evaluated for sensitivity to potent hERG 1A blockers 50,115 Mutation of multiple residues ( Thr 623, Ser 624, Val 625, Gly 648, Tyr 652 and Phe 656) located in the S6 domain o f the hERG1A subunit lowered the affinity of multiple drugs, including MK 499, a potent anti arrhythmic The same residues were found to be important for binding of cisapride, terfenadine and several other drugs from different chemical classes 55 Tyr 652 and Phe 656 have side chains that are essential for drug block. These aromatic residues are turned toward the center of the channel only when the channel is open which is consistent with hERG1A being blocked after the channel has opened. It has been reported that drugs block hERG1A in a state dependent manner and most have been shown to not block the channel in its closed state 116 Conceivably, the
72 multiple aromatic side chains (eight per channel), will accommodate multiple interactions by numerous drugs. Reduction in or reduced I Kr can lead to drug induced arrhythmias. The drugs that block hERG1A encompass almost every class of drugs on the market. These drugs are known to elongate the repolarization phase, including many antibiotics 24,28,35,41,46,49,117 121 We chose to look at antibiotics for many reasons: 1) antibiotics are given to both men and women, 2) they are prescrib ed to patients of all age groups from young to elderly, 3) antibiotics are provided to individuals for many different diseases, and 4) h ERG1A mutations can alter the susceptibility of the channel complex to many drugs. The Sesti group documented an increa se in drug block of the hERG1/MiRP1 channel complex by SMX when the complex was formed with a mutated MiRP1 subunit 119,121 The T to A mutation disrupts the NXS/T consensus site for N linked glycosylation. Although this mutation is not associated with inherited LQTS and it does not alter the formation of the hERG1/MiRP1 complex, the polymorphism is implicated in drug induced arrhythmias. To decide what antibiotics to investigate, we examined how a drug blocks the channel. As described earlier, there are multiple ways to block hERG1A (see Figure 21). Sulfamethoxazole (SMX) was selected since it ma y block hERG1A extracellularly, apparently similarly to how many divalent cations block the
73 hERG1A pore 119 Erythromycin was chosen because it is believed to bloc k the hERG1A channel intracellularly and increasing concentrations of Erythromycin cause a severe decrease in hERG1A current and an alteration in the voltage dependence of channel block 35,41 Penicillin G had not been shown to block the hERG1A channel until our findings (summarized in chapter 4) However, Ampicil lin, an analog to Penicillin, was revealed to prolong the QT interval during anaphylaxis to the prescribed antibiotic 99 The cardiac abnormalities sub sided after a few days. We thought to use Penicillin G to test whether a drug that has no known inhibition of hERG1A can become an inhibitor under condition s of altered glycosylation. Our studies fou nd Penicillin G does inhibit hERG1A function and therefore cannot be used as a negative control. Sulfamethoxazole (SMX) is a sulfonamide bacteriostatic antibiotic (see Figure 22) I ts primary activity is against susceptible forms of Streptococcus, Staphylococcus aureus, Escherichia coli, Haemophilus influenzae and oral anaerobes. It is primarily u sed to treat urinary tract infections and can be used as an alternative to amoxicillin based antibiotics to treat sinusitis. Sulfonamides are structural analogs and competitive antagonists of para aminobenzoic acid (PABA). They inhibit normal bacterial utilization of PABA for the synthesis of f olic acid, an important metabolite in DNA synthesis Folic acid is not synthesized in humans, but is instead a dietary requirement. This allows for the selective toxicity to bacterial cells (or any cell dependent on synthesizing folic acid) over
74 Figure 22. SMX block of hERG1A The top panel displays the molecular structure for SMX. The molecular weight is 290 g/mol and the pKa is ~5.7. The bottom panel is a typical current trace at +20 mV showing a reduction in hERG1A current at the following con centrations: 172 M, 1 mM, and 2 mM.
75 human cells. Bacterial re sistance to SMX is caused by mutations in the folic acid enzyme that inhibit PABA from binding and block folic acid synthesis Sulfamethoxazole is typically prescribed to adults and children, to treat bronchitis, middle ear infections, and urinary tract infections. The dosage for an adult (above 88 lbs.) is 800 mg every 12 hours, while the dosage for a child (weighing less than 88 lbs.) is 9 13 mg per pound every 12 hours. d Erythromycin from the meta bolic products of a strain of Saccharopolyspora erythraea found in soil samples in 1949 (see Figure 23) Lilly filed for patent protection of the compound and U.S. patent 2,653,899 was granted in 1953. The product was launched commercially in 19 52 under the brand name Ilosone. Erythromycin was formerly also called Ilotycin. Erythromycin is a macrolide antibiotic that has an antimicrobial spectrum similar to or sli ghtly wider than that of penicillin, and is often used for people who have an allergy to penicillins. For respiratory tract infections, it has better coverage of atypical organisms, including mycoplasma and Legionellosis Its structure c ontains a 14 memb ered lactone ring with ten asymmetric centers and two sugars (L cladinose and D desoamine), making it a compound very difficult to produce via synthetic methods. Erythromycin inhibi ts the c ytochrome P450 system which has the potential of affecting the metabolism of many different drugs. If CYP3A4 substrates are taken with erythromycin the levels of the substrates of these
76 Figure 23. Erythromycin blocks hERG1A The top panel displays the molecular structure for Erythromycin. The molecular weight is 734 g/mol and pKa is ~8.8. The bottom panel is a typical current trace at +20 mV showing a reduction in hERG1A current at the following concentrations: 34M, 136 M, and 272 M.
77 drugs could increase often causing adverse effects Erythromycin is eas ily inactivated by gastric acid and is normally administered as an enteric coated tablet. It is very rapidly absorbed, and diffuses into most tissues and phagocytes. Due to the high c oncentration in phagocytes, erythromycin is actively transp orted to the site of infection where large co ncentrations of erythromycin can be released. Erythromycin is typically prescribed to adults and ch ildren to treat strep throat and Legionnaires diseas e. The dosage for an adult is 250 500 mg every 6 12 hours, while the dosage for a child is 3.4 5.6 mg per pound every 6 hours or 6.8 11.4 mg per pound every 12 hours. The typical s erum levels range from 4 40 M. Penicillin is a group of antibiotics der i ved from Penicillium fungi (see Figure 24). Penicillin antibiotics are historically significant because they are the first drugs that were effective against many previously serious diseases such as syphilis and Staphylococcus infections. Penicillins are still widely used today, though many type s of bacteria are now resistant. All penicillins are Beta lactam antibiotics and are used to treat bacterial infections The term "penam" is used to describe the core skeleton of a member of a penicillin antibioti c. This skeleton has the molecular formula R C 9 H 11 N 2 O 4 S, where R is a variable side chain. Penicillin is typically prescribed to adults and children, to kill bacteria or prevent bacterial growth. The dosage for an adult is 200,000 500,000 U (125 312 mg)
78 Figure 24. Penicillin G blocks hERG1A The top panel displays the molecular structure for Penicillin G. The bottom panel is a typical current trace at +20 mV showing a reduction in hERG1A current at the following concentrations: 30 M, 150 M, a nd 600 M.
79 every 4 6 hours, while the dosage for a child is 189 13,636 U per pound every 4 8 hours. Our earlier work, summarized in chapter 3, demonstrated that N glycans limit hERG1A activity. We want to better understand how these alterations in N glycans may impact antibiotic block of the channel. Utilizing the CHO cell expression system of Pro5, Lec2, and Lec1 cells, we delivered different concentrations of three antibiotics individually to the hERG1A channel to determine whether and how changes in N glycosylation modulate antibiotic block of hERG1A. The data indicate that N glycans play a protective role on hERG1A antibiotic block 122 Also, hERG1A block was never lessened by conditions of reduced N glycosylation. Results SMX block is altered by N glycans SMX is thought to block hERG1A extracellularly. A competitive inhibition study done with SMX and cadmium showed that SMX blocks hERG1A in a similar fashion to cadmium 119 The hERG1A N glycosylation structures reside in the P loop. With the CHO cell expression system, we tested whether changes in N glycans affected block of hERG1A. The data show decreases in N glycosylation caused an increase in hERG1A block (see Figure 25). If sialic acids are absent from the N glycan structure, an increase in maximum block is seen for the low
80 Figure 25. Concentration dependent block by SMX Top panel: Percent block of hERG1A current by SMX at each concentration tested. Bottom Panel: Dose response curve of SMX. is significant (p < 0.05).
81 (50 g/mL), mid (300 g/mL) and high (600 g/mL) concentrations. Interestingly, when the complex N gl ycosylation is absent, there is also a significant increase in SMX block for every concentration tested. These data suggest that sialic acids play a protective role on hERG1A block by SMX at saturating depolarizations. The data also suggest that complex N glycans may provide additional protection against SMX block. Erythromycin block is modified by complex N glycans Erythromycin has been a heavily researched drug relative to hERG1A block. It has been linked to arrhythmias in individuals and shown to be pro arrhythmic by blocking hERG1A. The studies completed on Erythromycin conclude that the aromatic residues on the S6 segment interact with the drug and block hERG1A from the inside 35,55,105 We questioned whether changes in N glycosylation alters Erythromycin block. The data indicate that alterations in N glycans modify block of hERG1A by Erythromycin (see Figure 26). For channels that are fully glycosylated or missing sialic acids, a similar 10 15 % block was observed at 25 g/mL. Howeve r, the Lec1 cells expressing hERG1A show a 20 % significant block at the low concentration. As the Erythromycin concentration was increased, no significant difference in block was observed among conditions. This suggests that a saturating effect may occu r. That is, regardless, of the N glycans attached, Erythromycin will block hERG1A. This might not be surprising since it is well established that Erythromycin blocks hERG1A from the inside.
82 Figure 26. Concentration dependent block by Erythromycin Top panel: Percent block of hERG1A current by Erythromycin at each concentration tested. Bottom Panel: Dose response curve of Erythromycin. is significant (p < 0.05).
83 Perhaps, the impact of extracellular N glycans on intracellular block i s a minor effect. Keep in mind that there was a small, but significant protective effect of N glycans on Erythromycin block at the low concentration. If the N glycans are in the extracellular pore region, how do they protect against drug block from the i nside? There could be many possible explanations for this that include, 1) N glycan structures can be rather large and the bulky structure may dip down farther into the pore region affecting the surrounding area, and 2) hERG1A N glycosylation may alter th e channel structure within the plasma membrane. If the channel structure is altered by reduced complex N glycans, the aromatic structures might be more or less accessible to drugs that will block hERG1A. N glycans alter Penicillin G block of hERG1A In ch apter 4, we showed that Penicillin G blocks the hERG1A channel in a concentration dependent manner. Here, we asked whether alterations in N glycosylation impact Penicillin G block at the following concentrations: 10 g/mL, 50 g/mL, and 200 g/mL. A decr ease in sialic acids significantly increase the block by Penicillin G at every concentration tested (see Figure 27). The Lec1 cells expressing hERG1A show a greater block than control, but this does not reach significance until the high concentration. Th ese data suggest that sialic acids account for alterations in Penicillin G block on hERG1A. Interestingly, at the high concentration of Penicillin G, we see equal block between the Lec2 and Lec1 suggesting that the sialic acids play a significant role in hERG1A block.
84 Figure 27. Concentration dependent block by Penicillin G Top panel: Percent block of hERG1A current by Penicillin G at each concentration tested. Bottom Panel: Dose response curve of Penicillin G. is significant (p < 0.05).
85 Overall block by SMX, Erythromycin, and Penicillin G is different for each antibiotic and N glycans alter the block uniquely for each drug. This is the measured overall blocking ability of hERG1A at maximized depolarizations. However, achieving maximum de polarizations may not occur for all cells at similar times during the AP and may not even be consistent among ventricular myocytes. Thus, we sought to question whether there is an effect of N glycans on hERG1A block at less depolarizing potentials (less t han maximum depolarizations). In order to do this, drug block at 40, 30, 20, and 10 mV was measured. N glycans modulate SMX block at small depolarizations HERG1A block by SMX at each concentration is shown in figures 28 30. There are several interes ting, significant phenomenon that can be gleaned from this data. When hERG1A is expressed in Lec2 cells, SMX block, at small depolarizations, is greater than when the channel is fully glycosylated. The increase in drug block was seen at every concentrati on of SMX. Note for the Lec2 and Lec1 cell lines, there is no change in block with membrane potential. This is quite different than that observed in the Pro5 cells. That is, apparently, SMX blocks hERG1A in the absence of sialic acids and complex N glyc ans in a voltage independent manner. For the Lec1 expressing hERG1A cells, an increase in SMX block was also observed when compared to fully glycosylated channels. These data suggest that sialic acids have the greater impact
86 Figure 28. 50 g/mL SMX block at small depolarizations % block by SMX at small depolarizations ( 40, 30, 20, and 10 mV) for the 50 g/mL concentration.
87 Figure 29. 300 g/mL SMX block at small depolarizations % block by SMX at small depolarizations ( 40, 30, 20, and 10 mV) for the 300 g/mL concentration.
88 Figure 30. 600 g/mL SMX block at small depolarizations % block by SMX at small depolarizations ( 40, 30, 20, and 10 mV) for the 600 g/mL concentration.
89 on SMX block at small depolarizations than complex N glycosylation. These data suggest that the negative charges alter the mechanism of SMX block, specifically its voltage dependence. Complex N glycans impact Erythromycin block at small depolarizations The previous data showed a small, but significant increase in overall Erythromycin block of hERG1A at the low Eyrthromycin concentration, but only for the Lec1 cells. However, no other significant effect of N glycans on overall Erythromycin block was observed. Interestingly, at small depolarizations ( 40, 30, 20, and 10 mV), Lec1 cells expressing hERG1A show a large and significant increase in Erythromycin block compared to Pro5 cells expressing hERG1 A. In fact, the Pro5 cells show no block from 40 to 20 mV (see Figures 31 33). Pro5 and Lec2 cells expressing hERG1A show the same voltage dependence, and still, the block by Erythromycin is greater in the Lec2 cells then the Pro5 cells. These data ar e consistent with the removal of sialic acids is not enough to alter voltage dependent block of Erythromycin, but an increase in block is still evident. In addition, the removal of complex glycosylation greatly increase the Erythromycin block of hERG1A an d may alter voltage dependent block.
90 Figure 31. 25 g/mL Erythromycin block at small depolarizations % block by Erythromycin at small depolarizations ( 40, 30, 20, and 10 mV) for the 25 g/mL concentration
91 Figure 32. 100 g/mL Erythromycin block at small depolarizations % block by Erythromycin at small depolarizations ( 40, 30, 20, and 10 mV) for the 100 g/mL concentration.
92 Figure 33. 200 g/mL Erythromycin block at small depolarizations % block by Erythromycin at small depolarizations ( 40, 30, 20, and 10 mV) for the 200 g/mL concentration
93 Sialic acid and complex N glycans b lock Penicillin G at small depolarizations Overall, hERG1A Penicillin block is greater when fewer N glyca ns are present (see Figures 34 36). At small depolarizations, fully glycosylated cells show no block at 40 and 30 mV for every concentration of Penicillin G. The Lec2 cells expressing hERG1A show a large and significant block by Penicillin at every vol tage ( 40 to 10 mV) compared to the Pro5 expressing hERG1A cells. Interestingly, the Lec1 cells show a Penicillin G block that shifts with concentration. At the low concentration of drug, the Lec1 cells mimic the Pro5 cells with regard to voltage depend ence, even though the block is still greater in the Lec1 cells. At the high concentration, the Lec1 cells mimic the Lec2 voltage dependence and the block is not different between the Lec2 and Lec1 cells. The sialic acids absent from hERG1A alter block (a t small depolarizations) by Penicillin G greater than any other N glycan structure. The additional absence of complex glycosylation was not effective in altering voltage dependence more than the sialic acids. Discussion Overall, N glycosylation plays a p rotective role for hERG1A against antibiotic block. The data show significantly less antibiotic block when N glycans are present. For all measurements tested (3 antibiotics at 3 concentrations each at more than 5 different voltages and under two differen t conditions of reduced N glycosylation), never was the block under conditions of reduced glycosylation
94 Figure 34. 10 g/mL Penicillin G block at small depolarizations % block by Penicillin G at small depolarizations ( 40, 30, 20, and 10 mV) for the 10 g/mL concentration.
95 Figure 35. 50 g/mL Penicillin G block at small depolarizations % block by Penicillin G at small depolarizations ( 40, 30, 20, and 10 mV) for the 50 g/mL concentration.
96 Figure 36. 200 g/mL Penicillin G block at small depolarizations % block by Penicillin G at small depolarizations ( 40, 30, 20, and 10 mV) for the 200 g/mL concentration.
97 significantly less glycosylation. Interestingly, block at small depolarizations for SMX, Erythromycin, and Penicillin are impacted by changes in glycosylation N glycans may alter the voltage dependence of block in different ways depending on the antibiotic. A lack of sialic acids, and the negative charge associated with it, can render block to be voltage independent for SMX. This is consistent with the thought that SMX blocks hERG1A from outside the cell (see Figure 37). The negative sialic acids may have the ability to interact more with SMX since the N glycans sit in the pore region. The negative sialic acid residues on the terminal ends of the N glycosylation struc tures may impact SMX block, since SMX is negatively charged at physiologic pH, by altering the surface charge (or repelling the drug itself). When sialic acid are absent from hERG1A, SMX may have the ability to more readily bind to its site of action beca use this site is available. Erythromycin block is apparently not affected by changes to the surface charge, but may be affected by removal of complex N glycans. Erythromycin has been shown to interact with the aromatic structures deep in the pore region and probably blocks hERG1A from inside the cell (Figure 38). This mechanism of block may be too distant from the surface charges outside the channel to observe an effect. Nevertheless, the complex N glycans may potentially change a structural element of hERG1A, thereby altering the affinity of a drug for its binding site.
98 Figure 37. Model of extracellular block on hERG1A
99 Figure 38. Model of intracellular block on hERG1A
100 Penicillin G block was altered with changes to glyc osylation, and a concentration dependent, voltage dependence was seen for the Lec1 cells expressing hERG1A. HERG1A expressed in the Lec2 cells showed an increase in block at the small depolarizations. It is not known what mechanism Penicillin G uses to block hERG1A Based on the findings that Penicillin G block is altered by both sialic acids and complex N glycosylation, there is still an uncertainty as to how it blocks hERG1A. However, at physiologic pH, Penicillin G is highly charged and would not transverse the lipid bilayer suggesting that Penicillin G may only block from the outside (Figure 37). The data suggest that sialic acid residues play the biggest role in altering block by Penicillin G in these studies, possibly by repelling ability to interact with hERG1A because of the negative charges on both Penicillin G and sialic acid. This implies that Penicillin G may block hERG1A similarly to SMX utilizing the surface charge theory of block. As discussed in chapter 3, alterations in glycosylation affect gating and kinetics of hERG1A. These changes in N glycan structures also alter antibiotic block of the channel. Changes in hERG1A function (block) will alter I Kr current. If there is modification to a major repolarizing current of the cardiac AP, then the AP could be compromised. The changing blocking characteristics of antibiotics on hERG1A with alterations in N glycans could become relevant in individuals with pathologic changes in glycosylation. One example to illustrate the po ssible impact of alterations in N
101 a potential for a shortened QT interval. Given the data presented here, we might predict likelihood of drug block on hERG1A. This increase in block could lead to an extension of the AP and may result in an adverse cardiac event. However, since ready have a short AP, then the Penicillin G might be used therapeutically to correct the affliction instead of causing one. Our data show that sialic acid impact cardiac function differently when investigating on ck. Multiple diseases affect glycosylation states in the body and are linked to cardiovascular dysfunction. Understanding the probable role of N glycosylation on drug block will lead to safer drugs being put on the market.
102 CHAPTER 6 FINAL DISCUSSION Cardiac a ction potent ial waveforms are produced by the gating of multiple ion channels. A change in ion channel function can lead to arrhythmias, torsades de pointes, cardiac disease, and sudden cardiac death HERG1A, one of these ion channels is partially responsible for the repolarization of the AP. Dysfunction of hERG1A can result in multiple cardiac maladies and is the source of hundreds of types of cardiac arrhythmias. Inherited, acquired, and drug induced arrhythmias can all occur with alterations in hERG1A. Physiologic and pathologic consequences The impact of N glycan dependent hERG1A activity on AP waveform is unique among voltage gated K + channels Here we show a depolarizing shift in hERG1A voltage dependent gating parameters with conditions of reduced glycosylation. We believe that sialic acid play a large role in the micro environment surface charge surrounding many ion channels. This surface charge directly affects normal function of the channels. This molecular mechanism is consistent with some previous studies of K v 1 and K v 4 channel isoforms (although not all), with conclusions made th at reduced
103 sialylation of the affected K v 1 and K v 4 isoforms limits K v channel activity during the AP (loss of function) 65,87,93 95 The negative sialic acids likely impact the shift in voltage dependent gating through elec trostatic mechanisms. In fact, for several of these isoforms, only channel activation was affected by glycosylation. F or K v 1.2 channel inactivation was enhanced by decreased glycosylation. Both of these mechanisms would lead to a further loss of channel activity during the AP with reduced glycosylation 71 ,72 Here we show that a less sialylated hERG1A channel will be more active during the AP than a more heavily sialylated hERG1A channel (see Figure 15) This gain of function with reduced glycosylation, primarily sialylation, is unique among K v channe l isoforms. Regulated changes in glycosylation could lead to modulated hERG1A activity Changes in N linked sugar types and levels, particularly sialic acids, alter hERG1A function. Previously, we showed that the glycogenes whose products are responsible for the addition and removal of glycans, are significantly differentially expressed in the atria versus ventricles and during development of each cardiac chamber 66 Further, our data suggested that cardiomyocyte N glycan structures are remodeled across the developing ventri cle, particularly at the level of complex N glycans. We also showed that the major voltage gated cardiac Na + channel, Na v 1.5, is less sialylated in the neonatal ventricle than in the adult ventricle, consistent with the regulated changes in cardiac ion ch annel sialylation 93 If hERG1A sialylation is similarly regulated throughout ventricle
104 development, then, given our data shown here, we would predict that the more d uring the ventricular AP, leading to an increased AP duration in the adult as reported previously 123,124 Aberrant changes in glycosylation could lead to hERG1A dysfunction It is well accepted that alterations in I Kr often lead to aberrant cardiac AP repolarization and arrhythmias. Currently, there are more than twenty six known forms of Congenital Disorders of Glycosylation (CDG) with a prevalence of 1 in 5000 births 75 78 Of these, most affect N linked glycosylation. CDG are genetic disorders that onset because of missing or mutant glycogenes, primarily glycosyltransferase s. The mutant/missin g glycogene results in proteins and lipids with relatively modest reductions in glycosylation levels. The minimally reduced levels of glycosylation grossly affect multiple systems including the cardiovascular system, with many CDG patients presenting card iomyopathy and/or arrhythmias 75 77 If hERG1A channels of CDG patients are less sialylated then one would predict decreased ventricular AP duration for those patients. disease ( Trypanasoma cruzi T. c ruzi infection) is a human parasite disease that has no vaccine and no known cure and afflicts ~18 million people wordlwide 80,82,83 The mortality rate is ~30%, with nearly all terminal patients experiencing heart failure preceded by ventricular tachycardia 81 T. c ruzi
105 releases a neuraminidase that cleaves sialic acids from the host tissue, which is typically the heart. A recent study measured mouse ECG as a function of time post infection using two strains of T. c ruzi to infect. The data indicated that ~60% of infecte d mice (compared to ~5% of control) showed some conduction abnormality 125 If hERG1A is less sialylated in chagasic patients, then one would predict increased hERG1A activity during the AP and a decrease in AP duration that could contribute to the ventricular arrhythmias experienced by many of these patients. Taken as a whole, alterations in N glycosylation will alter gating and kinetics of hERG1A. This, in turn, may lead to alterations in the AP and po tentially adverse cardiac events. Recognizing that small changes to N glycans can have a large impact on the cardiac AP will facilitate understanding of many cardiac diseases. HERG1A antibiotic block is reduced with N glycans Here, we showed that N glyca ns are protective for hERG1A against antibiotic block. SMX, Erythromycin, and Penicillin G block were each greater with reduced N glycosylation, and block by each antibiotic was uniquely affected by N glycans. The data suggest that the negatively charged sialic acids seemed to alter block by SMX more than Erythromycin suggesting that the sugars affect the surface charge, and thereby alter the mechanism that SMX blocks the channel. However, Erythromycin does not seem to be affected by surface charge chang es and its mechanism of block could be altered by sugars in a voltage dependent
106 manner whereby the membrane potential is altered by the absence of sialic acid on the extracellular surface causing an interior change to the binding site for Erythromycin. Pe nicillin G was affected by both sialic acid and complex N glycosylation suggesting that its mechanism of block may be through surface charge effects, perhaps similarly to the effects of sialic acids on SMX block. Antibiotic block may happen through multip le pathways As discussed, one common theme of this report is that while N glycans limit hERG1A function, they also serve to protect hERG1A from antibiotic block. The N glycans may protect against antibiotic block of hERG1A through several mechanisms. Fir st, the findings from chapter 3 predict that N glycans alter the percentage of channels in a given state at a membrane potential and therefore there should be a change in the percentage of block at a given voltage. That is, we would predict that the rightward shift in voltage dependent activation that occurs with reduced glycosylation should cause a reduction in the percent block at a membrane potential given that block occurs in a state other than closed. However, this cannot be solely responsible f or the variation in block with N glycans reported here. In fact, because the data reported in chapter 5 are a ratio of block comparing the current with drug to current without drug, the N glycan dependent modulation of gating on hERG1A is not measured. I n addition, the overall block would measure the ability of the drug to block all available channels because the percent block is determined as a change in maximal conductance.
107 Therefore, the data suggest an additional N glycan effect on antibiotic block beyond altering the state dependence of the channel at a given voltage. A second possible mechanism by which N glycans protect against antibiotic block may involve structural changes in how hERG1A sits in the membrane or its conformational state. If N gl ycans alter how the channel is inserted into or how it resides in the plasma membrane, then the affinity of a drug for its binding site might be affected by the change in structure or conformation. Perhaps, even access of the drug to its site of block may be altered. This could explain why hERG1A expressed in Lec1 cells has an increase in block by Erythromycin. This potential conformational change to the channel may alter the binding site of Erythromycin giving the drug better access. A third possible m echanism involves a more direct effect in which N glycans form keeping the drug from its site of action. This shield could form a structural hindrance or, perhaps as suggest ed by the Lec2 studies on SMX (and perhaps Penicillin G), a charge effect. Altering the surface charge surrounding the channel could directly affect the affinity of a drug for hERG1A. Further studies need to be done to elucidate how N glycans protect hER G1A from antibiotic block.
108 Future studies The data show that N glycosylation alters gating and kinetics of hERG1A such that hERG1A function is limited by N glycans. Also, the data demonstrate that N glycans can protect hERG1A against antibiotic channel b lock. Both findings are novel, potentially significant and likely occur through distinct mechanisms. However, they represent only a start to our understanding of the full extent that N glycans impact hERG1A. Future questions to be addressed include, 1) identifying how changes in block at small depolarizations ( 40 to 10 mV range) may affect the cardiac AP. Our findings indicate those voltages are relevant during Phase III where hERG1A is most active in the AP. 2) determining the mechanisms by which N glycans affect block by antibiotics, focusing on hERG1A block from either inside or outside of the cell and how N glycans can alter both mechanisms. Our data initially suggest that surface charge may play a role in altering drug block of hERG1A when the drug blocks from the outside, and complex N glycans may modulate drug block from inside the pore region in a structural manner. 3) expand computer simulation of the human ventricle to include the effects of N glycans on hERG1A function and block. Summary The hERG1A channel is essential to normal heart rhythm as hERG1A activity underlies I Kr I Kr is responsible for portions of phase II and phase III repolarization of the cardiac AP The data shown here indicate that hERG1A
109 activity is limited by N glycans attached to the channel, with channel sialic acids responsible for most of the effect. This will effectively extend the ventricular AP. As channel sialylation levels are altered, either physiologically or pathophysiologically, hERG1A activity and AP repolarization will be modulated. Physiological increases in channel sialylation would serve to further extend the AP while physiological and pathological reductions in channel sialylation would lead to increased hERG1A activity and reduced AP durati ons. Either pathway c ould increase susceptibility to ventricular arrhythmia. Thus, we have described a novel mechanism by which hERG1A activity can be modulated by regulated and aberrant changes in glycosylation, particular ly sialylation. Such modulation of hERG1A activity would affect the rate at which ventricular myocytes repolarize during the AP and potentially increase susceptibility to ventricular arrhythmias. Additionally, our findings show that N glycans protect against antibiotic block of hERG1A, likely through mechanisms distinct from the impact of sialic acids on channel gating. HERG1A can be blocked by a drug from outside of inside the cell. These mechanisms could be through surface charge, conforma tional changes, and voltage dependent effects. Total block by each antibiotic was affected differently by N glycosylation. Additionally, voltage dependent antibiotic block of hERG1A was modified by either sialic acids or complex N glycans. Acquired Long QT syndrome can affect anyone, even someone who has no history of cardiac problems. Understanding how glycosylation is modulated in
110 the heart, and how changes in glycosylation affect ion channel function may lead to development of better anti arrhythmic drugs with fewer side effects (such as pro arrhythmic behavior reversing an adverse cardiac event.
111 CHAPTER 7 REFERENCES (1) ten Tusscher KH, Noble D, Noble PJ, Panfilov AV. A model for human ventricular tissue. Am J Physiol Heart Circ Physiol. 2004;286:H1573 H1589. (2) Armoundas AA, Wu R, Juang G, Marban E, Tomaselli GF. Electrical and structural remodeling of the failing ventricle. Pharmacol Ther. 2001;92:213 230. (3) Nerbonne JM, Kass RS. Molecular Physiology of Cardiac Repolarization. Physiol Rev. 2005;85:1205 1253. (4) Olshansky B, Delmar M, Tomaselli GF. The year in arrhythmias 2009 part II. Heart Rhythm. 2010;7:538 548. (5) Rosamond W, Flegal K, Furie K et al. Heart disease and stroke statistics -2008 update: a report from the American Heart Association Statis tics Committee and Stroke Statistics Subcommittee. Circulation. 2008;117:e25 146. (6) Tomaselli GF, Marban E. Electrophysiological remodeling in hypertrophy and heart failure. Cardiovasc Res. 1999;42:270 283. (7) Keating MT, Sanguinetti MC. Molecular and Cellular Mechanisms of Cardiac Arrhythmias. Cell. 2001;104:569 580. (8) Zhou Z, Gong Q, Epstein ML, January CT. HERG Channel Dysfunction in Human Long QT Syndrome. INTRACELLULAR TRANSPORT AND FUNCTIONAL DEFECTS. J Biol Chem. 1998;273:21061 21066.
112 (9) Wilde AAM, Postema PG, Di Diego JM et al. The pathophysiological mechanism underlying Brugada syndrome: Depolarization versus repolarization. Journal of Molecular and Cellular Cardiology. 2010;49:543 553. (10) Wes tenskow P, Splawski I, Timothy KW, Keating MT, Sanguinetti MC. Compound mutations: a common cause of severe long QT syndrome. Circulation. 2004;109:1834 1841. (11) Towbin JA, Wang Z, Li H. Genotype and Severity of Long QT Syndrome. Drug Metab Dispos. 200 1;29:574 579. (12) Tomaselli GF. Cardiac I(to), KCNE2, and Brugada syndrome: promiscuous subunit interactions, or what happens in HEK cells stays in HEK cells? Heart Rhythm. 2010;7:206 207. (13) Teng GQ, Lees Miller JP, Duan Y, Li BT, Li P, Duff HJ. [K +]o dependent change in conformation of the HERG1 long QT mutation N629D channel results in partial reversal of the in vitro disease phenotype. Cardiovascular Research. 2003;57:642 650. (14) Splawski I, Shen J, Timothy KW et al. Spectrum of Mutations in Long QT Syndrome Genes : KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation. 2000;102:1178 1185. (15) Roden DM, Lazzara R, Rosen M, Schwartz PJ, Towbin J, Vincent GM. Multiple Mechanisms in the Long QT Syndrome: Current Knowledge, Gaps, and Future Direct ions. Circulation. 1996;94:1996 2012. (16) Roden DM, Balser JR. A plethora of mechanisms in the HERG related long QT syndrome: Genetics meets electrophysiology. Cardiovascular Research. 1999;44:242 246. (17) Trudeau MC, Warmke JW, Ganetzky B, Robertson GA. HERG, a human inward rectifier in the voltage gated potassium channel family. Science. 1995;269:92 95.
113 (18) Sanguinetti MC, Jiang C, Curran ME, Keating MT. A Mechanistic Link between an Inherited and an Acquired Cardiac Arrhythmia: HERG encodes the IKr Potassium Channel. Cell. 1995;81:299 307. (19) Curran ME, Splawski I, Timothy KW, Vincent GM, Green ED, Keating MT. A Molecular Basis for Cardiac Arrhythmia: HERG Mutations Cause Long QT Syndrome. Cell. 1995;80:795 803. (20) Sanguinetti MC, Curran ME, Spector PS, Keating MT. Spectrum of HERG K+ channel dysfunction in an inherited cardiac arrhythmia. PNAS. 1996;93:2208 2212. (21) Sanguinetti MC, Tristani Firouzi M. hERG potassium channels and cardiac arrhythmia. Nature. 2006;440:463 469. (22) Hu ang FD, Chen J, Lin M, Keating MT, Sanguinetti MC. Long QT syndrome associated missense mutations in the pore helix of the HERG potassium channel. Circulation. 2001;104:1071 1075. (23) Chen J, Zou A, Splawski I, Keating MT, Sanguinetti MC. Long QT syndro me associated mutations in the Per Arnt Sim (PAS) domain of HERG potassium channels accelerate channel deactivation. J Biol Chem. 1999;274:10113 10118. (24) Abbott GW, Sesti F, Splawski I et al. MiRP1 Forms IKr Potassium Channels with HERG and Is Associa ted with Cardiac Arrhythmia. Cell. 1999;97:175 187. (25) Mazhari R, Greenstein JL, Winslow RL, Marban E, Nuss HB. Molecular Interactions Between Two Long QT Syndrome Gene Products, HERG and KCNE2, Rationalized by In Vitro and In Silico Analysis. Circ Res 2001;89:33 38. (26) Ravens U, Cerbai E. Role of potassium currents in cardiac arrhythmias. Europace. 2008;10:1133 1137.
114 (27) Cordeiro JM, Brugada R, Wu YS, Hong K, Dumaine R. Modulation of IKr inactivation by mutation N588K in KCNH2: A link to arrhy thmogenesis in short QT syndrome. Cardiovascular Research. 2005;67:498 509. (28) Sanguinetti MC, Jiang C, Curran ME, Keating MT. A mechanistic link between an inherited and an acquired cardiac arrhythmia: HERG encodes the IKr potassium channel. Cell. 199 5;81:299 307. (29) Sanguinetti MC, Curran ME, Spector PS, Keating MT. Spectrum of HERG K+ channel dysfunction in an inherited cardiac arrhythmia. Proc Natl Acad Sci U S A. 1996;93:2208 2212. (30) Sanguinetti MC, Zou A. Molecular physiology of cardiac delayed rectifier K+ channels. Heart Vessels. 1997;Suppl 12:170 172. (31) Sanguinetti MC, Tristani Firouzi M. hERG potassium channels and cardiac arrhythmia. Nature. 2006;440:463 469. (32) Heginbot ham L, Lu Z, Abramson T, MacKinnon R. Mutations in the K+ channel signature sequence. Biophysical Journal. 1994;66:1061 1067. (33) Shepard PD, Trudeau MC. Emerging roles for ether a go go related gene potassium channels in the brain. J Physiol. 2008;586: 4785 4786. (34) Anson BD, Ackerman MJ, Tester DJ et al. Molecular and functional characterization of common polymorphisms in HERG (KCNH2) potassium channels. Am J Physiol Heart Circ Physiol. 2004;286:H2434 H2441. (35) Antzelevitch P, Sun MPZ Q, Zhang M D, Yan MPG X. Cellular and Ionic Mechanisms Underlying Erythromycin Induced Long QT Intervals and Torsade de Pointes. Journal of the American College of Cardiology. 1996;28:1836 1848. (36) Bennett PB, Yazawa K, Makita N, George AL, Jr. Molecular mechanis m for an inherited cardiac arrhythmia. Nature. 1995;376:683 685.
115 (37) Bennett PB. Long QT syndrome: biophysical and pharmacologic mechanisms in LQT3. J Cardiovasc Electrophysiol. 2000;11:819 822. (38) Brundel BJ, van G, I, Henning RH et al. Alterations in potassium channel gene expression in atria of patients with persistent and paroxysmal atrial fibrillation: differential regulation of protein and mRNA levels for K+ channels. J Am Coll Cardiol. 2001;37:926 932. (39) Brundel BJ, Henning RH, Kampinga H H, van G, I, Crijns HJ. Molecular mechanisms of remodeling in human atrial fibrillation. Cardiovasc Res. 2002;54:315 324. (40) Compton SJ, Lux RL, Ramsey MR et al. Genetically defined therapy of inherited long QT syndrome. Correction of abnormal repolari zation by potassium. Circulation. 1996;94:1018 1022. (41) Daleau P, Lessard E, Groleau MF, Turgeon J. Erythromycin Blocks the Rapid Component of the Delayed Rectifier Potassium Current and Lengthens Repolarization of Guinea Pig Ventricular Myocytes. Circulation. 1995;91:3010 3016. (42) Friederich P, Solth A, Schillemeit S, Isbrandt D. Local anaesthetic sensitivities of cloned HERG channels from human heart: comparison with HERG/MiRP1 and HERG/MiRP1T8A. Br J Anaesth. 2004;92:93 101. (43) Ganapathi SB, Kester M, Elmslie KS. State dependent block of HERG potassium channels by R roscovitine: implications for cancer therapy. Am J Physiol Cell Physiol. 2009;296:C701 C710. (44) Gong Q, Jones MA, Zhou Z. Mechanisms of Pharmacological Rescue of Traffickin g defective hERG Mutant Channels in Human Long QT Syndrome. J Biol Chem. 2006;281:4069 4074. (45) Guo J, Zhan S, Lees Miller JP, Teng G, Duff HJ. Exaggerated block of hERG (KCNH2) and prolongation of action potential duration by erythromycin at temperatu res between 37[degree sign]C and 42[degree sign]C. Heart Rhythm. 2005;2:860 866.
116 (46) Haverkamp W, Breithardt G, Camm AJ et al. The potential for QT prolongation and proarrhythmia by non antiarrhythmic drugs: clinical and regulatory implications. Report on a Policy Conference of the European Society of Cardiology. Eur Heart J. 2000;21:1216 1231. (47) Herzberg IM, Trudeau MC, Robertson GA. Transfer of rapid inactivation and sensitivity to the class III antiarrhythmic drug E 4031 from HERG to M eag channe ls. J Physiol (Lond). 1998;511:3 14. (48) Kamiya K, Niwa R, Mitcheson JS, Sanguinetti MC. Molecular determinants of HERG channel block. Mol Pharmacol. 2006;69:1709 1716. (49) Lee MB KL, Jim MB MH, Tang MB SC, Tai MB YT. QT Prolongation and Torsades de Pointes Associated with Clarithromycin. The American Journal of Medicine. 1998;104:395 396. (50) Mitcheson JS, Chen J, Lin M, Culberson C, Sanguinetti MC. A structural basis for drug induced long QT syndrome. Proc Natl Acad Sci U S A. 2000;97:12329 12333 (51) Olshansky B, Delmar M, Tomaselli GF. The year in arrhythmias -2009: part I. Heart Rhythm. 2010;7:417 426. (52) Phillip J.Stansfeld, Peter Gedeck, Martin Gosling, Brian Cox, John S.Mitcheson, Michael J.Sutcliffe. Drug block of the hERG potassium channel: Insight from modeling. 2007:568 80. (53) Sanchez Chapula JA, Navarro Polanco RA, Culberson C, Chen J, Sanguinetti MC. Molecular determinants of voltage dependent human ether a go go related gene (HERG) K+ channel block. J Biol Chem. 2002;277:235 87 23595. (54) Sanguinetti MC, Bennett PB. Antiarrhythmic drug target choices and screening. Circ Res. 2003;93:491 499. (55) Sanguinetti MC, Chen J, Fernandez D, Kamiya K, Mitcheson J, Sanchez Chapula JA. Physicochemical basis for binding and voltage
117 dependent block of hERG channels by structurally diverse drugs. Novartis Found Symp. 2005;266:159 166. (56) Spector PS, Curran ME, Keating MT, Sanguinetti MC. Class III antiarrhythmic drugs block HERG, a human cardiac delayed rectifier K+ channel. Open c hannel block by methanesulfonanilides. Circ Res. 1996;78:499 503. (57) Yoshida K, Niwa T. Quantitative Structure Activity Relationship Studies on Inhibition of HERG Potassium Channels. J Chem Inf Model. 2006;46:1371 1378. (58) Znnkler BJ. Human ether a go go related (HERG) gene and ATP sensitive potassium channels as targets for adverse drug effects. Pharmacology & Therapeutics. 2006;112:12 37. (59) Petrecca K, Atanasiu R, Akhavan A, Shrier A. N linked glycosylation sites determine HERG channel surfac e membrane expression. J Physiol (Lond). 1999;515:41 48. (60) Gong Q, Anderson CL, January CT, Zhou Z. Role of glycosylation in cell surface expression and stability of HERG potassium channels. Am J Physiol Heart Circ Physiol. 2002;283:H77 H84. (61) Be nnett E, Urcan MS, Tinkle SS, Koszowski AG, Levinson SR. Contribution of sialic acid to the voltage dependence of sodium channel gating. A possible electrostatic mechanism. J Gen Physiol. 1997;109:327 343. (62) Bennett ES. Effects of channel cytoplasmic regions on the activation mechanisms of cardiac versus skeletal muscle Na(+) channels. Biophys J. 1999;77:2999 3009. (63) Bennett ES. Channel cytoplasmic loops alter voltage dependent sodium channel activation in an isoform specific manner. J Physiol. 20 01;535:371 381.
118 (64) Bennett ES. Isoform specific effects of sialic acid on voltage dependent Na+ channel gating: functional sialic acids are localized to the S5 S6 loop of domain I. J Physiol. 2002;538:675 690. (65) Johnson D, Bennett ES. Isoform specific Effects of the beta2 Subunit on Voltage gated Sodium Channel Gating. J Biol Chem. 2006;281:25875 25881. (66) Montpetit ML, Stocker PJ, Schwetz TA et al. Regulated and aberrant glycosylation modulate cardiac electri cal signaling. Proc Natl Acad Sci U S A. 2009;106:16517 16522. (67) Schwetz TA, Norring SA, Bennett ES. N glycans modulate K(v)1.5 gating but have no effect on K(v)1.4 gating. Biochim Biophys Acta. 2010;1798:367 375. (68) Ufret Vincenty CA, Baro DJ, Le derer WJ, Rockman HA, Quinones LE, Santana LF. Role of sodium channel deglycosylation in the genesis of cardiac arrhythmias in heart failure. J Biol Chem. 2001;276:28197 28203. (69) Ufret Vincenty CA, Baro DJ, Santana LF. Differential contribution of sia lic acid to the function of repolarizing K(+) currents in ventricular myocytes. Am J Physiol Cell Physiol. 2001;281:C464 C474. (70) Watanabe I, Zhu J, Recio Pinto E, Thornhill WB. Glycosylation Affects the Protein Stability and Cell Surface Expression of Kv1.4 but Not Kv1.1 Potassium Channels. J Biol Chem. 2004;279:8879 8885. (71) Watanabe I, Zhu J, Sutachan JJ, Gottschalk A, Recio Pinto E, Thornhill WB. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated actio n potentials. Brain Research. 2007;1144:1 18. (72) Zhu J, Recio Pinto E, Hartwig T, Sellers W, Yan J, Thornhill WB. The Kv1.2 potassium channel: The position of an N glycan on the extracellular linkers affects its protein expression and function. Brain R esearch. 2009;1251:16 29.
119 (73) Snider MD, Rogers OC. Intracellular movement of cell surface receptors after endocytosis: resialylation of asialo transferrin receptor in human erythroleukemia cells. J Cell Biol. 1985;100:826 834. (74) Snider MD, Rogers OC. Membrane traffic in animal cells: cellular glycoproteins return to the site of Golgi mannosidase I. J Cell Biol. 1986;103:265 275. (75) Jaeken J. Komrower Lecture. Congenital disorders of glycosylation (CDG): it's all in it! J Inherit Metab Dis. 2003 ;26:99 118. (76) Jaeken J, Carchon H. Congenital disorders of glycosylation: a booming chapter of pediatrics. Curr Opin Pediatr. 2004;16:434 439. (77) Jaeken J. Congenital disorders of glycosylation (CDG): update and new developments. J Inherit Metab D is. 2004;27:423 426. (78) Marquardt T, Denecke J. Congenital disorders of glycosylation: review of their molecular bases, clinical presentations and specific therapies. Eur J Pediatr. 2003;162:359 379. (79) Bern C, Montgomery SP, Herwaldt BL et al. Eva luation and Treatment of Chagas Disease in the United States: A Systematic Review. JAMA. 2007;298:2171 2181. (80) Bestetti RB, Muccillo G. Clinical course of Chagas' heart disease: a comparison with dilated cardiomyopathy. International Journal of Cardiology. 1997;60:187 193. (81) Biolo A, Ribeiro AL, Clausell N. Chagas Cardiomyopathy -Where Do We Stand After a Hundred Years? Progress in Cardiovascular Diseases. 2001;52:300 316. (82) Tribulatti MV, Mucci J, Van Rooijen N, Leguizamon MS, Campetel la O. The trans Sialidase from Trypanosoma cruzi Induces Thrombocytopenia during Acute Chagas' Disease by Reducing the Platelet Sialic Acid Contents. Infect Immun. 2005;73:201 207.
120 (83) Libby P, Alroy J, Pereira ME. A neuraminidase from Trypanosoma cruzi removes sialic acid from the surface of mammalian myocardial and endothelial cells. Journal of Clinical Investigation. 1986;77:127 135. (84) Johnson D, Bennett ES. Gating of the shaker potassium channel is modulated differentially by N glycosylation and sialic acids. Pflugers Arch. 2008;456:393 405. (85) Stanley P, Caillibot V, Siminovitch L. Selection and Characterization of Eight Phenotypically Distinct Lines of Lectin resistant Chinese Hamster Ovary Cell. Cell. 1975;6:121 128. (86) Bennett ES. Cha nnel activation voltage alone is directly altered in an isoform specific manner by Na(v1.4) and Na(v1.5) cytoplasmic linkers. J Membr Biol. 2004;197:155 168. (87) Johnson D, Montpetit ML, Stocker PJ, Bennett ES. The sialic acid component of the beta1 sub unit modulates voltage gated sodium channel function. J Biol Chem. 2004;279:44303 44310. (88) Warmke JW, Ganetzky B. A family of potassium channel genes related to eag in Drosophila and mammals. Proceedings of the National Academy of Sciences of the Unit ed States of America. 1994;91:3438 3442. (89) Cordeiro JM, Brugada R, Wu YS, Hong K, Dumaine R. Modulation of IKr inactivation by mutation N588K in KCNH2: A link to arrhythmogenesis in short QT syndrome. Cardiovascular Research. 2005;67:498 509. (90) S pector PS, Curran ME, Zou A, Keating MT, Sanguinetti MC. Fast inactivation causes rectification of the IKr channel. J Gen Physiol. 1996;107:611 619. (91) Sparks S. Inherited Disorders of Glycosylation. Molecular Genetics and Metabolism. 2006;87:1 7.
121 (9 2) Santacruz Toloza L, Huang Y, John SA, Papazian DM. Glycosylation of shaker potassium channel protein in insect cell culture and in Xenopus oocytes. Biochemistry. 1994;33:5607 5613. (93) Stocker PJ, Bennett ES. Differential Sialylation Modulates Volta ge gated Na+ Channel Gating throughout the Developing Myocardium. J Gen Physiol. 2006;127:253 265. (94) Bennett E, Urcan MS, Tinkle SS, Koszowski AG, Levinson SR. Contribution of Sialic Acid to the Voltage Dependence of Sodium Channel Gating A Possible Electrostatic Mechanism. J Gen Physiol. 1997;109:327 343. (95) Bennett ES. Isoform specific effects of sialic acid on voltage dependent Na+ channel gating: functional sialic acids are localized to the S5 S6 loop of domain I. J Physiol (Lond). 2002;538:6 75 690. (96) Thornhill WB, Wu MB, Jiang X, Wu X, Morgan PT, Margiotta JF. Expression of Kv1.1 Delayed Rectifier Potassium Channels in Lec Mutant Chinese Hamster Ovary Cell Lines Reveals a Role for Sialidation in Channel Function. J Biol Chem. 1996;271:19 093 19098. (97) Schwetz TA, Norring SA, Bennett ES. N glycans modulate Kv1.5 gating but have no effect on Kv1.4 gating. Biochimica et Biophysica Acta (BBA) Biomembranes. 2010;1798:367 375. (98) Montpetit ML, Stocker PJ, Schwetz TA et al. Regulated an d aberrant glycosylation modulate cardiac electrical signaling. PNAS. 2009. (99) Mehta D, Warwick G, Goldberg M. QT Prolongation after Ampicillin Anaphylaxis. British Heart Journal. 1986;55:308 310. (100) Hardman RM, Stansfeld PJ, Dalibalta S, Sutcliff e MJ, Mitcheson JS. Activation gating of hERG potassium channels: S6 glycines are not required as gating hinges. J Biol Chem. 2007;282:31972 31981.
122 (101) Stansfeld PJ, Gedeck P, Gosling M, Cox B, Mitcheson JS, Sutcliffe MJ. Drug block of the hERG potass ium channel: insight from modeling. Proteins. 2007;68:568 580. (102) Norring SA, Schwetz TA, Bennett ES. Penicillin G Blocks hERG1A. Manuscript in preparation. 2010. (103) Belus A, White E. Effects of streptomycin sulphate on ICaL, IKr and IKs in guinea pig ventricular myocytes. European Journal of Pharmacology. 2002;445:171 178. (104) Fernandez D, Ghanta A, Kauffman GW, Sanguinetti MC. Physicochemical features of the HERG channel drug binding site. J Biol Chem. 2004;279:10120 10127. (105) Sanguinetti MC, Mitcheson JS. Predicting drug hERG channel interactions that cause acquired long QT syndrome. Trends Pharmacol Sci. 2005;26:119 124. (106) De Bruin ML, van Staa TP, Belitser SV, Leufkens HG, Hoes AW. Predicting cardiac arrhythmias and sudden death in diabetic users of proarrhythmic drugs. Diabetes Care. 2005;28:440 442. (107) De Bruin ML, Pettersson M, Meyboom RH, Hoes AW, Leufkens HG. Anti HERG activity and the risk of drug induced arrhythmias and sudden death. Eur Heart J. 2005;26:590 597. (108) De Bruin ML, van Hemel NM, Leufkens HG, Hoes AW. Hospital discharge diagnoses of ventricular arrhythmias and cardiac arrest were useful for epid emiologic research. J Clin Epidemiol. 2005;58:1325 1329. (109) Sanchez Chapula JA, Sanguinetti MC. Altered gating of HERG potassium channels by cobalt and lanthanum. Pflugers Arch. 2000;440:264 274. (110) Sanguinetti MC, Mitcheson JS. Predicting drug h ERG channel interactions that cause acquired long QT syndrome. Trends in Pharmacological Sciences. 2005;26:119 124.
123 (111) Ho WK, Kim I, Lee CO, Youm JB, Lee SH, Earm YE. Blockade of HERG channels expressed in Xenopus laevis oocytes by external divalent c ations. Biophys J. 1999;76:1959 1971. (112) Lin MC, Papazian DM. Differences between ion binding to eag and HERG voltage sensors contribute to differential regulation of activation and deactivation gating. Channels (Austin ). 2007;1:429 437. (113) Fari d R, Day T, Friesner RA, Pearlstein RA. New insights about HERG blockade obtained from protein modeling, potential energy mapping, and docking studies. Bioorg Med Chem. 2006;14:3160 3173. (114) Perrin MJ, Subbiah RN, Vandenberg JI, Hill AP. Human ether a go go related gene (hERG) K+ channels: function and dysfunction. Prog Biophys Mol Biol. 2008;98:137 148. (115) Mitcheson J, Perry M, Stansfeld P, Sanguinetti MC, Witchel H, Hancox J. Structural determinants for high affinity block of hERG potassium chan nels. Novartis Found Symp. 2005;266:136 150. (116) Ficker E, Jarolimek W, Kiehn J, Baumann A, Brown AM. Molecular determinants of dofetilide block of HERG K+ channels. Circ Res. 1998;82:386 395. (117) Abbott GW, Goldstein SAN. Potassium Channel Subunit s: The MiRP Family. Mol Interv. 2001;1:95 107. (118) Milberg P, Eckardt L, Bruns HJ et al. Divergent Proarrhythmic Potential of Macrolide Antibiotics Despite Similar QT Prolongation: Fast Phase 3 Repolarization Prevents Early Afterdepolarizations and Tor sade de Pointes. J Pharmacol Exp Ther. 2002;303:218 225. (119) Park KH, Kwok SM, Sharon C, Baerga R, Sesti F. N glycosylation dependent block is a novel mechanism for drug induced cardiac arrhythmia. FASEB J. 2003;03 0577fje.
124 (120) Sanguinetti MC, Jur kiewicz 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. (121) Sesti F, Abbott GW, Wei J et al. A common polymorphism associated with antibio tic induced cardiac arrhythmia. PNAS. 2000;97:10613 10618. (122) Norring SA, Schwetz TA, Ednie AR, Bennett ES. N Glycans Alter Antibiotic Block of hERG1A. Manuscript in preparation. 2010. (123) Krasemann T, Bente K, Burkhardtsmaier G. The corrected QT interval in 24 h ECGs in neonates. Clinical Research in Cardiology. 2010;99:309 314. (124) Taubel J, Naseem A, Harada T et al. Levofloxacin can be used effectively as a positive control in thorough QT/QTc studies in healthy volunteers. Br J Clin Pharma col. 2010;69:391 400. (125) Bustamante JM, Rivarola HW, Fretes R, Paglini Oliva PA. Weekly electrocardiographic pattern in mice infected with two different Trypanosoma cruzi strains. International Journal of Cardiology. 2005;102:211 217.
A BOUT THE AUTHOR Sarah Ann Norring was born December 6, 1978 in Skokie, Illinois. She graduated from Mundelein High School in 1996. Sarah enrolled at the University of Illinois, Urbana Champaign in the Biology program in the College of Liberal Arts & she graduated with a B.S. in Biology. That year, Sarah started working for Pharmacia Corporatio n in the Medicinal Chemistry department for her first year out of undergraduate school. She then moved to Salt Lake City, UT and began work with the Analytical Method Development team at Cephalon, Inc. In 2005, Sarah joined the laboratory of Eric S. Bennett, Ph.D. at the University of South Florida. She was awarded an AHA Pre doctoral Fello wship while at USF. She earned her M.S.M.S. and was accepted as a doctoral candidate in 2007. Sarah was conferred with a Ph.D. in Medical Sciences from the University of South Florida, College of Medicine in 2010.