USF Libraries
USF Digital Collections

Glycosylation modulates cardiac excitability by altering voltage-gated potassium currents

MISSING IMAGE

Material Information

Title:
Glycosylation modulates cardiac excitability by altering voltage-gated potassium currents
Physical Description:
xiv, 182 leaves : ill. ; 29 cm.
Language:
English
Creator:
Schwetz, Tara A
Publication Date:

Subjects

Subjects / Keywords:
Glycosylation   ( mesh )
Ion Channel Gating   ( mesh )
Potassium Channels, Voltage-Gated   ( mesh )
Arrhythmias, Cardiac   ( mesh )
Sialylation
Potassium channel
Cardiac conduction
Arrhythmia
Dissertations, Academic -- Molecular Pharmacology and Physiology -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Neuronal, cardiac, and skeletal muscle electrical signaling is achieved through the highly regulated activity of several types of voltage-gated ion channels to produce an action potential (AP). Voltage-gated potassium (Kv) channels are responsible for repolarization of the AP. Kv channels are uniquely and heavily glycosylated proteins. Previous reports indicate glycosylation modulates gating of some Kv channel isoforms; often, terminal sialic acid residues alter Kv channel gating. Here, we questioned whether alterations in glycosylation impact Kv channel gating, thus altering APs and cardiac excitability. ST3Gal-IV, a sialyltransferase expressed at uniform levels throughout the heart, adds sialic acids to N- and O-glycans through alpha 2-3 linkages. Electrocardiograms (ECGs) suggest that cardiac conduction/rhythm are altered in ST3Gal-IV(-/-) animals, which show an increased incidence of arrhythmic beats.AP waveform parameters and two components of IK, the transient outward, Ito, and the slowly inactivating, IK,slow, were compared in neonatal control versus ST3Gal-IV(-/-) and glycosidase treated atrial and ventricular myocytes. Action potential durations (APDs) measured from ST3Gal-IV(-/-) and glycosidase treated atrial myocytes were lengthened significantly (~25-150%) compared to control; however, ventricular APDs were unaffected by changes in glycosylation. Consistently, atrial Ito and IK,slow activation were shifted to more depolarized potentials (by ~9-17 mV) in ST3Gal-IV(-/-) and glycosidase treated myocytes, while ventricular K+ currents were unaltered. Those channels responsible for producing Ito and IK,slow were examined under conditions of full and reduced glycosylation.Sialylation and N-glycosylation uniquely and differently impact gating of two mammalian Shaker family Kv channel isoforms, Kv1.4 and Kv1.5; Kv1.4 gating was unaffected by changes in channel glycosylation, while N-linked sialic acids, acting through electrostatic mechanisms, fully account for glycan effects on Kv1.5 gating. In addition, sialic acids modulate the gating of three Kv channel isoforms that are not N-glycosylated, Kv2.1, Kv4.2, and Kv4.3, through apparent electrostatic mechanisms. Click chemistry was utilized to confirm that these three isoforms are O-glycosylated and sialylated; thus, O-linked sialylation modulates gating of Kv2.1, Kv4.2, and Kv4.3. This study suggests that regulated or aberrant glycosylation alters the gating of channels producing IK in a chamber-specific manner, thus altering the rate of cardiac repolarization and potentially leading to arrhythmias.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Tara A. Schwetz.
General Note:
Document formatted into pages; contains 182 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002324024
oclc - 668993028
usfldc doi - E14-SFE0003040
usfldc handle - e14.3040
System ID:
SFS0027357:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader ntm 2200361Ka 4500
controlfield tag 001 002324024
005 20101007100558.0
008 101007s2009 xx a ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003040
035
(OCoLC)668993028
040
FHJ
c FHJ
049
FHJA
096
QU 55.7
b S415g 2009
1 100
Schwetz, Tara A.
0 245
Glycosylation modulates cardiac excitability by altering voltage-gated potassium currents /
by Tara A. Schwetz.
260
2009.
300
xiv, 182 leaves :
ill. ;
29 cm.
500
Document formatted into pages; contains 182 pages.
Includes vita.
502
Dissertation (Ph.D.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: Neuronal, cardiac, and skeletal muscle electrical signaling is achieved through the highly regulated activity of several types of voltage-gated ion channels to produce an action potential (AP). Voltage-gated potassium (Kv) channels are responsible for repolarization of the AP. Kv channels are uniquely and heavily glycosylated proteins. Previous reports indicate glycosylation modulates gating of some Kv channel isoforms; often, terminal sialic acid residues alter Kv channel gating. Here, we questioned whether alterations in glycosylation impact Kv channel gating, thus altering APs and cardiac excitability. ST3Gal-IV, a sialyltransferase expressed at uniform levels throughout the heart, adds sialic acids to N- and O-glycans through alpha 2-3 linkages. Electrocardiograms (ECGs) suggest that cardiac conduction/rhythm are altered in ST3Gal-IV(-/-) animals, which show an increased incidence of arrhythmic beats.AP waveform parameters and two components of IK, the transient outward, Ito, and the slowly inactivating, IK,slow, were compared in neonatal control versus ST3Gal-IV(-/-) and glycosidase treated atrial and ventricular myocytes. Action potential durations (APDs) measured from ST3Gal-IV(-/-) and glycosidase treated atrial myocytes were lengthened significantly (~25-150%) compared to control; however, ventricular APDs were unaffected by changes in glycosylation. Consistently, atrial Ito and IK,slow activation were shifted to more depolarized potentials (by ~9-17 mV) in ST3Gal-IV(-/-) and glycosidase treated myocytes, while ventricular K+ currents were unaltered. Those channels responsible for producing Ito and IK,slow were examined under conditions of full and reduced glycosylation.Sialylation and N-glycosylation uniquely and differently impact gating of two mammalian Shaker family Kv channel isoforms, Kv1.4 and Kv1.5; Kv1.4 gating was unaffected by changes in channel glycosylation, while N-linked sialic acids, acting through electrostatic mechanisms, fully account for glycan effects on Kv1.5 gating. In addition, sialic acids modulate the gating of three Kv channel isoforms that are not N-glycosylated, Kv2.1, Kv4.2, and Kv4.3, through apparent electrostatic mechanisms. Click chemistry was utilized to confirm that these three isoforms are O-glycosylated and sialylated; thus, O-linked sialylation modulates gating of Kv2.1, Kv4.2, and Kv4.3. This study suggests that regulated or aberrant glycosylation alters the gating of channels producing IK in a chamber-specific manner, thus altering the rate of cardiac repolarization and potentially leading to arrhythmias.
590
Advisor: Eric S. Bennett, Ph.D.
653
Sialylation
Potassium channel
Cardiac conduction
Arrhythmia
2 650
Glycosylation.
Ion Channel Gating.
Potassium Channels, Voltage-Gated.
Arrhythmias, Cardiac.
690
Dissertations, Academic
z USF
x Molecular Pharmacology and Physiology
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3040



PAGE 1

Glycosylation Modulates Cardiac Exci tability by Altering Voltage-Gated Potassium Currents by Tara A. Schwetz A dissertation submitted in partial fulfillment of the requirement s of the degree of Doctor of Philosophy Department of Molecular P harmacology and Physiology College of Medicine University of South Florida Major Professor: Eric S. Bennett, Ph.D. Javier Cuevas, Ph.D. Jay B. Dean, Ph.D. Craig A. Doupnik, Ph.D. Bernd Sokolowski, Ph.D. Date of Approval: July 10, 2009 Keywords: sialylation, potassium channe l, cardiac conduction, arrhythmia Copyright 2009, Tara A. Schwetz

PAGE 2

DEDICATION In loving memory of my father, Charles H. Munn

PAGE 3

ACKNOWLEDGEMENTS First and foremost, I would like to thank my husband, Brian, for his unwavering patience and support throughout the course of my graduate education. To him, I am truly indebted. Additionally I am immensely grateful to my parents, Charles and Vicki, my step-father, Michael, and my brother, Trevor, for their constant encouragement. I also would like to thank Raymond, Joyce, and Helen Schwetz for always motivating me. I extend the deepest thanks to Eric S. Benne tt (AKA D.E.B.), to whom I am very much indebted. I credit much of my success as a graduate student to him, as he has been a great mentor. I also would li ke to thank my committee members: Javier Cuevas, Ph.D., Jay Dean, Ph.D ., Craig A. Doupnik, Ph.D., and Bernd Sokolowski, Ph.D. I would like to thank the me mbers of the Bennett lab: Marty L. Montpetit, Ph.D., Sarah A. Norring, M.S., Andrew Ednie, B.S., Jeanie Harper J.D., and Barrett McCormick, M.S. Their assistance and advice have been invaluable to me. Additionally, I’d like to include a spec ial thank you to Sarah Norring, whose friendship I have truly cherished (Stera will prevail). Thank you to Barbara Nicholson, Bridget Shields, Franjesca Ja ckson, and Kathryn Zahn for all of their assistance. Finally, a sincere thank you to my friends and extended family, who have provided me with continued support and in spiration: my grandparents, Harold and Marion Munn, Barbara Barrette, and Jerome and Thorma Barrette, and my close friends Stephanie Cunni ngham, Michelle Richardson, Brittany Roberts, Carolyn Robinson, and all thos e I haven’t room to list.

PAGE 4

i TABLE OF CONTENTS LIST OF FIGURES vi LIST OF TABLES viii LIST OF ABBREVIATIONS ix ABSTRACT xiii CHAPTER 1 – INTRODUCTION 1 Action potentials are generated and conducted through the myocardium 1 Remodeling of the myoc ardium is involved in normal processes and pathologies 9 Alterations in ion channel function can impac t excitability, resulting in pathological consequences 10 Members of the diverse potassi um channel family function 12 similarly Delayed rectifiers and KA channels contribute to the generation of voltagegated potassium currents 18 Kv channel isoforms responsible for production of Ito and IK,slow are expressed differentially throughout the myocardium 22 Kv channels are post-translationally modified 24 The impact of glycosylation is significant throughout development and disease 25 Changes in glycosylation can alter the gating of ion channels 31 CHAPTER 2 – MATERIALS AND METHODS 37

PAGE 5

ii Chinese Hamster Ovary Cell Culture and Transfection 37 Vector Construction and Mutagenesis 37 Transgenic Mice 38 Neonatal Cardiomyocyte Isolation fo r Electrophysiology 39 Electrophysiology and Data A nalysis 40 Whole Cell Recordings in CHO Cells 40 Conductance-Voltage Relationship for CHO Cell Recordings 41 Steady-State Inactivation (hinf) 42 Time Constants of Activation ( n) 42 Recovery from Inactivation 43 Cardiomyocyte Potassium Current Recordings 43 Cardiomyocyte Ito and IK,slow G-V Relationships 44 Action Potential Recordings from Cardiomyocytes 44 Action Potential Protocols 45 Whole Cell Homogenization 45 Immunoblots 46 Click-iT Glycoprotein Labeling and Detection 47 Cardiac Tissue Isolation 48 RNA Isolation and Reverse Transcr iption 48 Quantitative RT-PCR 48 Telemetric Electrocardiogram Recordings 50 Data Analysis and Statistics 50

PAGE 6

iii CHAPTER 3 – THE SIALYLTRANSFERASE, ST3Gal-IV, ALTERS CARDIAC ACTION POTENT IAL WAVEFORMS AND IK 51 Results 52 Electrocardiograms suggest slowed heart ra tes and altered waveforms in the ST3Gal-IV(-/-) animal 52 Action potential duration is prolonged in ST3Gal-IV(-/-) atrial myocytes, yet is unaltered in ventricula r myocytes 56 The voltage-dependence of Ito and IK,slow activation is altered in the atria, but is unaffected in the ventri cles of ST3Gal-IV(-/-) animals 62 Discussion 65 CHAPTER 4 – REMOVAL OF GLYCOSYLATION PR OLONGS ACTION POTENTIAL DURATION AND MODULATES IK 70 Results 71 Action potentials are prolonged in neonatal at rial myocytes following removal of glycosylation 71 Ventricular action potentials are unaffected by changes in glycosylation 74 Atrial Ito and IK,slow activation is shifted to more depolarized potentials with removal of glycosylation 77 Glycosylation does not impact ventricular Ito or IK,slow activation 79 Discussion 82 Atrial action potentials are modulated by alterations in glycosylation, while ventricular action potentials ar e unaffected 82 Glyco sylation affects cardiac Ito and IK,slow activation in a chamber-specific manner 83 CHAPTER 5 – UNIQUE MODULATION OF KV1.4 AND KV1.5 GATING BY N-GLYCANS 86 Results 87

PAGE 7

iv N-linked sialic acids account for the full effect of glycans on Kv1.5 gating 89 N-linked glycans impact Kv1.5 gating 89 N-linked sialic acids acc ount for the full effect of sugars on Kv1.5 gating 90 Kv1.5 is N-glycosylated 96 Kv1.5 sialic acids apparent ly modulate channel gating by contributing to the external negative surface potential 96 Kv1.4 gating is not modulated by N-glycosylation or sialylation 99 Sialic acids do not affect Kv1.4 gating 99 N-glycosylation does not alter gating of Kv1.4 103 Kv1.4 is N-glycosylated and sialylated 103 Discussion 105 N-linked sialic acids account for the full effect of glycosylation on Kv1.5 gating—a novel finding for modulation of Shaker pota ssium channel function 105 N-glycosylation does not affect Kv1.4 voltagedependent gating 107 Summary 108 CHAPTER 6 – O-LINKED SIALIC ACIDS IMPACT KV CHANNEL GATING 110 Results 112 Sialic acids modulate Kv4.2 and Kv4.3 activation 114 Kv4.2 and Kv4.3 fast inactivation and recovery from fast inactivation are not altered by changes in channel sialylation 114 Kv2.1 is not N-glycosylated 118 Sialic acids modulate Kv2.1 activation 118 Sialic acids impose an apparent electrostatic effe ct on Kv2.1, Kv4.2, and Kv4.3 channel gating 122

PAGE 8

v Kv2.1, Kv4.2, and Kv4.3 channels are O-glycosylated and sialylated 125 Discussion 128 O-linked sialylation modul ates the gating of Kv2.1, Kv4.2, and Kv4.3 through apparent electrostatic mechanisms 128 Summary 133 CHAPTER 7 – FINAL DISC USSION 134 ST3Gal-IV deficient mice show alterations in ECGs 134 Regulated and aberrant glycosylation modulates atrial, but not ventricular, action potentials 136 With reduced glycosylation, changes in activation of repolarizing Kv currents are consistent with the effects reported for cardiac APs 139 Kv channel activation is modulated by reduc ed glycosylation in isoform-specific manners 141 Gating of two Kv1 isoforms is modulated uniquely by glycosylation 142 O-linked sialic acids alter activation of Kv2.1, Kv4.2, and Kv4.3 143 Physiological and pathological relevance 144 Future Directions 152 Summary 155 REFERENCES 157 ABOUT THE AUTHOR End Page

PAGE 9

vi LIST OF FIGURES Figure 1.1 The action potential waveform is modified throughout the myocardium 2 Figure 1.2 Electrical conduction thr ough the heart 5 Figure 1.3 The human atrial and ventricular action potentials and the ionic basis of their production 7 Figure 1.4 Electrocardiogram recordings of cardiac irregularities 11 Figure 1.5 Diversity of K+ channels 13 Figure 1.6 Schematic of voltage-gated potassium channel structure 15 Figure 1.7 Representation of the voltage-sensor (S4) mo vement 17 Figure 1.8 Human and murine ventricular ac tion potentials are generated by similar currents 19 Figure 1.9 Kv channels can inactivate through tw o mechanisms 21 Figure 1.10 N-glycan biosynt hesis in the Endoplasmic Re ticulum 27 Figure 1.11 Biosynthesis of mamma lian O-glycans 29 Figure 1.12 Glycosyl ation sites of Kv channels are located on the extracellular linkers 33 Figure 3.1 ECGs recorded from conscious, unrestrained control and ST3Gal-IV(-/-) mice suggest altered cardiac rhythm in the knockout 53 Figure 3.2 Action potential durat ion is prolonged in ST3Gal-IV(-/-) atrial myocytes 57 Figure 3.3 Ventricular action potent ial duration is not impacted by ST3Gal-IV 59 Figure 3.4 Atrial Ito and IK,slow activation are altered by loss of ST3Gal-IV 63

PAGE 10

vii Figure 3.5 ST3Gal-IV defici ency does not affect ventricular Ito and IK,slow 66 Figure 4.1 Enzymatic treatment to remove glycosylation 72 Figure 4.2 Glycosidase treatment pr olongs atrial action potential duration 73 Figure 4.3 Ventricular action potentials are not affected by removal of glycosylation 75 Figure 4.4 Deglycosylation shifts steady-state activation of atrial Ito and IK,slow to more depolarized potentials 78 Figure 4.5 Steady-state activation of ventricular Ito and IK,slow is unaffected by treatment with glycosidases 80 Figure 5.1 Kv1.4 and Kv1.5 contain one putative N-glycosylation site 88 Figure 5.2 N-linked sialic acids account for the full effect of glycans on Kv1.5 activation 91 Figure 5.3 Kv1.5 is N-glycosylated 95 Figure 5.4 Sialic acids modulate Kv1.5 gating through apparent electrostatic mechanisms 97 Figure 5.5 Kv1.4 gating is not impacted by sialylat ion or N-glycosylation 100 Figure 5.6 Kv1.4 is N-glycosylated and sialylated 104 Figure 6.1 Sialic acids modulate Kv4.2 and Kv4.3 activation 113 Figure 6.2 Kv4.2 and Kv4.3 fast inactivation and recovery from fast inactivation are not affected by sialic acids 115 Figure 6.3 Kv2.1 is not N-glycosylated 119 Figure 6.4 Sialic acids modulate Kv2.1 activation 120 Figure 6.5 Sialic acids modulate Kv2.1, Kv4.2, and Kv4.3 activation through apparent el ectrostatic mechanisms 123 Figure 6.6 Metabolic incorporation of labeled sugars utilizing Cu(I)-catalyzed cycloaddition 126 Figure 6.7 Kv2.1, Kv4.2, and Kv4.3 channels are O-glycosylated and sialylated 129

PAGE 11

viii LIST OF TABLES Table 1.1 The diversity of Kv channels and Kv currents throughout the mammalian heart 23 Table 3.1 The measured action potential parameters fo r control and ST3Gal-IV(-/-) atria and ventricles 61 Table 3.2 Measured gating parameters for atrial and ventricular Ito and IK,slow from control and ST3Gal-IV(-/-) mice 67 Table 4.1 The measured action potential parameters for control and glycosidase treated atria and ventricles 76 Table 4.2 Measured gating parameters of atrial and ventricular Ito and IK,slow under control and glycosidase treated conditions 81 Table 5.1 Gating parameters measured for Kv1.5 94 Table 5.2 The measured gating parameters for Kv1.4 102 Table 6.1 The measured gating parameters for Kv4.2 and Kv4.3 sialic acid 117 Table 6.2 Gating parameters measured for Kv2.1 121

PAGE 12

ix LIST OF ABBREVIATIONS ECG Electrocardiogram AP Action Potential SA Node Sinoatrial Node Nav Voltage-Gated Sodium Na+ Sodium Kv Voltage-Gated Potassium Ito Transient Outward Current Cav Voltage-Gated Calcium K+ Potassium AV Node Atrioventricular Node BPM Beats Per Minute AF Atrial Fibrillation ICa Calcium Current INa Sodium Current IKur Ultrarapid Delayed Rect ifier Potassium Current LVH Left Ventricular Hypertrophy LQTS Long QT Syndrome HF Heart Failure Tl+ Thallium Rb+ Rubidium NH4 + Ammonium

PAGE 13

x TEA Tetraethylammonium 4-AP 4-Aminopyridine KA A-Type K+ Channels Ito,f Fast Transient Outward Current Ito,s Slow Transient Outward Current CFG Consortium for Functional Genomics Asn/N Asparagine GlcNAc N-Acetylglucosamine ER Endoplasmic Reticulum OST Oligosaccharyltransferase SA Sialic Acid GalNAc N-Acetylgalactosamine CDG Congenital Disorder s of Glycosylation T. cruzi Trypanosoma cruzi G-V Conductance-Voltage Va Voltage of Half-Activation CHO Chinese Hamster Ovary ST Sialyltransferase MEM Alpha-Minimum Essential Medium FBS Fetal Bovine Serum ORF Open Reading Frame DNA Deoxyribonucleic Acid PCR Polymerase Chain Reaction DMEM Dulbecco’s Modified Eagle’s Medium Pro5 Cells Fully Sialylating CHO Cell Line

PAGE 14

xi Lec2 Cells Reduced Sialylating CHO Cell Line G Conductance I Current Vp Peak Voltage EK Equilibrium Potential Ka Activation Slope Factor hinf Steady-State Inactivation Vi Voltage of Half-Inactivation Ki Inactivation Slope Factor n Activation Time Constant Sialidase A Neuraminidase TTX Tetrodotoxin PI Protease Inhibitors Ac4GalNAz N-Azidoacetylgalactosamine Ac4ManNAz N-Azidoacetylmannosamine RNA Ribonucleic Acid HPRT Hypoxanthinephosphoribosyltransferase ST3Gal-IV(-/-) ST3Gal-IV Knockout IK Potassium Current STX ST8Sia2 APD Action Potential Duration APD10 Action Potential Duration at 10% Repolarization APD50 Action Potential Duration at 50% Repolarization APD90 Action Potential Duration at 90% Repolarization EAD Early After Depolarization

PAGE 15

xii HCN Hyperpolarization-Activat ed Cyclic Nucleotide-Gated PNS Parasympathetic Nervous System Kv1.5N290Q Kv1.5 N-Glycosylation Mutant (Asparagine Mutation) Q Glutamine S Serine A Alanine Kv1.5S292A Kv1.5 N-Glycosylation Mut ant (Serine Mutation) MW Molecular Weight Kv1.4N354Q Kv1.4 N-Glycosylation Mutant (Asparagine Mutation) Click Cu(I)-Catalyzed Cycloaddition Kv2.1N283Q Kv2.1 N-Glycosylation Mutant (Asparagine Mutation) VWF Von Willebrand Factor If Pacemaker Current NET Norepinephrine Transporter NE Norepinephrine CDG-Ia Congenital Disorders of Glycosylation Type Ia CNS Central Nervous System ECM Extracellular Matrix CMD Congenital Muscular Dystrophy MD1 Myotonic Muscular Dystrophy Type 1 EDMD Emery-Dreifuss Muscular Dystrophy DMD Duschenne Muscular Dystrophy BMD Becker Muscular Dystrophy IgAN IgA Nephropathy LDL Low Density Lipoprotein

PAGE 16

xiii GLYCOSYLATION MODULATES CARDIAC EXCITABILITY BY ALTERING VOLTAGE-GATED POTASSIUM CURRENTS Tara A. Schwetz ABSTRACT Neuronal, cardiac, and skeletal muscle el ectrical signaling is achieved through the highly regulated activity of severa l types of voltage-gated ion channels to produce an action potential (AP) Voltage-gated potassium (Kv) channels are responsible for repolar ization of the AP. Kv channels are uniquely and heavily glycosylated proteins. Previous reports indicate glycosyl ation modulates gating of some Kv channel isoforms; often, terminal sialic acid residues alter Kv channel gating. Here, we questioned whether alte rations in glycosylation impact Kv channel gating, thus altering APs and ca rdiac excitability. ST3Gal-IV, a sialyltransferase expressed at uniform le vels throughout the heart, adds sialic acids to Nand O-glycans through 2-3 linkages. Electrocardiograms (ECGs) suggest that cardiac conduction/rhyt hm are altered in ST3Gal-IV(-/-) animals, which show an increased incidence of arrhythmic beats. AP waveform parameters and two components of IK, the transient outward, Ito, and the slowly inactivating, IK,slow, were compared in neonatal control versus ST3Gal-IV(-/-) and glycosidase treated atrial and ventricula r myocytes. Action potential durations (APDs) measured from ST3Gal-IV(-/-) and glycosidase treated atrial myocytes

PAGE 17

xiv were lengthened significantly (~25-150% ) compared to control; however, ventricular APDs were unaffected by c hanges in glycosylation. Consistently, atrial Ito and IK,slow activation were shifted to mo re depolarized potentials (by ~917 mV) in ST3Gal-IV(-/-) and glycosidase treated myocytes, while ventricular K+ currents were unaltered. Those c hannels responsible for producing Ito and IK,slow were examined under conditions of full and reduced glycosylation. Sialylation and N-glycosylation uniquely and differently impact gating of two mammalian Shaker family Kv channel isoforms, Kv1.4 and Kv1.5; Kv1.4 gating was unaffected by changes in channel glycosylation, while N-linked sialic acids, acting through electrostatic mechanisms, fully account for glycan effects on Kv1.5 gating. In addition, sialic acids modulate the gating of three Kv channel isoforms that are not N-glycosylated, Kv2.1, Kv4.2, and Kv4.3, through apparent electrostatic mechanisms. Click chemistry was utilized to confirm that these three isoforms are O-glycosylated and sialyl ated; thus, O-linked sialyl ation modulates gating of Kv2.1, Kv4.2, and Kv4.3. This study suggests that regulated or aberrant glycosylation alters the gat ing of channels producing IK in a chamber-specific manner, thus altering the rate of cardia c repolarization and potentially leading to arrhythmias.

PAGE 18

1 CHAPTER 1 INTRODUCTION Diseases of the cardiovascular system ar e abundant in the United States and can be life-threatening. In fact, for seve ral years, cardiovascular disease has been reported as the leading cause of death in the Unit ed States, with an estimated 80.7 million American adults (1 in 3) diagnosed with one or more types of cardiovascular disease. One afflicti on of the cardiovascular system, cardiac arrhythmias, affects more than 5 milli on people nationwide and results in over 400,000 deaths annually1. Arrhythmias, as well as heart failure, myocardial infarction, and other cardiovascular maladi es, are associated with alterations in normal cardiac ion channel function2-12. Thus, processes governing channel gating can potentially impac t cardiac excitability. Action potentials are generated a nd conducted through the myocardium Cardiac conduction and contraction can be investigated by measurement of electrical signals produced across the surface of the heart utilizing an electrocardiogram (ECG). Distinct phases of the cardiac cycle are represented as waves on the ECG. In general, the P wave corresponds to atrial depolarization. Atrial repolarization is masked by the QR S complex, which reflects ventricular depolarization. Ventricular repolarization is represent ed by the T wave (Figure 1.1). Each P-T segment denotes one full heart beat.

PAGE 19

2 Figure 1.1. The action potential w aveform is modified throughout the myocardium. 0 1 2 3 4 03 4 Figure 1.1. Representation of action potent ial waveform variation across different regions of the heart and the correspondi ng electrocardiogram (ECG) waves. Differential expression of ion channels is thought to account for these differences. Slow and fast action pot ential phases are numbered. Action potentials are displaced to signify the sequential firing through the myocardium

PAGE 20

3 and are aligned with the resulting ECG wave SA: sinoatrial; AV: atrioventricular; RV: right ventricle; LV: left ventricl e. Figure modified from Nerbonne and Kass, 200513.

PAGE 21

4 Electrical signals, measured cellularly as action potentials (AP), are conducted through the heart from t he sinoatrial (SA) node, depolarizing through the ventricles, and subsequently repolarizing in an opposite fashion (Figure 1.2). If this signal or path is disr upted at any location, altera tions in normal contraction and conduction can result, potentially leading to arrhythmias. Therefore, proper action potential production is an important component of normal cardiac rhythm. Action potentials, the rapi d depolarization and subseq uent repolarization of the membrane, are generated by the regulated opening and closing of voltage-gated ion channels producing electr ical currents (Figure 1.3) Cardiac action potentials are categorized as fast or slow APs by their waveform and the ion channels underlying their production. In phase 0 of the fast AP, voltage-gated sodium (Nav) channels open and cause a rapid depolariz ation of the membrane, due to movement of sodium (Na+) ions into the cell down their electrochemical gradient. Nav channels inactivate whils t voltage-gated potassium (Kv) channels activate to produce a transient outward current (Ito), resulting in the slight, rapid repolarization observed in phase 1. Duri ng the plateau phase ( phase 2), voltagegated calcium (Cav) channels activate, generating an inward current that is counterbalanced by an outward potassium (K+) current (human IKs). In phase 3, the outward K+ current, corresponding to IKr and IKur in humans and IK,slow in the mouse, begins to exceed the inward calc ium current; the cell is repolarized back toward the resting membrane potential as a result. The ensuing di astolic potential

PAGE 22

5 Figure 1.2. Electrical c onduction through the heart. Figure 1.2. Electrical signals that gener ate the heart beat originate in the sinoatrial (SA) node and depolarize the ventricles via passage through the atrioventricular (AV) node, subsequently re polarizing in the reverse direction. Figure from Kerwin, 2007. http://afibcryoablation.com/types.of.arrhythmia.asp

PAGE 23

6 is produced by the outward K+ current equaling the action of the Na+/K+-ATPase, which pumps K+ ions into the cell (phase 4) (Figure 1.3). Although generated in a similar manner to fast APs, slow APs are produced by different ion channels and consist of three phases. In phase 0, Cav channels depolarize the membrane, activating Kv channels to commence repolarization (phase 3). A constant inward current (If) drives the memb rane toward more positive potentials in phase 4, creating a diastolic depolarization. The diastolic depolarization allows for t he maintenance of consistency and rhythm and is an important characteristic of pacemaker cells (Figure 1.1)14. Variation in AP waveform is due to di fferential expression of numerous ion channel isoforms throughout the heart. Fo r example, fast action potentials, described above, are produced in the purki nje fibers and epicardium. In the sinoatrial and atrioventricular (AV) nodes, slow AP’s predominate (Figure 1.1)15. Furthermore, the AP waveform varies bet ween the atria and ventricles, as demonstrated by the distinct phas e 2 shape (Figures 1.1 and 1.3)16. In fact, differing cardiac AP waveforms can be observed throughout the developing myocardium17-20. Alterations in the type, relati ve density, or activity of ion channels can modify the waveform of the AP through a process often referred to as remodeling15,16,21.

PAGE 24

7 Figure 1.3. Human atrial and ventri cular action potentials and the ionic basis of their production. Phase 0 1 2 3 4

PAGE 25

8 Figure 1.3. Schematic of human atrial (blue) and ventricular (red) action potentials and the correspond ing currents and the known or postulated (?) ion channels underlying the macroscopic curr ents (left and right, respectively). Several of the currents involved in at rial and ventricular action potential production overlap, as indicated in purple. Note the phase in which each current contributes. Figure modifi ed from Pond and Nerbonne, 200116.

PAGE 26

9 Remodeling of the myocardium is involved in normal processes and pathologies The term remodeling refers to a change in cellular processes resulting from altered protein expression and is characte ristic of normal cardiac development and pathologies2-12,17-19,22-30. In fact, human prenatal cardiac rhythm is typically between 120 and 160 beats per mi nute (bpm), while the average adult heart rate is 72 bpm; this slowing of cardiac rhythm is demonstrative of cardiac remodeling throughout development and likely due to aut onomic innervation and ion channel remodeling31-36. Several cardiac pathologies are the cons equence of electrical and structural remodeling. In fact, electrical rem odeling perpetuates and maintains atrial fibrillation (AF), the most prevalent sustained arrhythm ia in the United States37-39. Current densities of calcium (ICa), sodium (INa), and transient outward (Ito) currents are significantly reduced in cani ne models of AF and in patients with permanent AF9,10,40-43. Consistently, parallel reductions in the relative expression levels of mRNA encoding the channel isoforms responsible for producing the aforementioned currents are observed in AF models44. A significant reduction in the current density of t he ultrarapid delayed rectif ier potassium current (IKur) has been shown in myocytes from patients with persistent AF, as well as decreased protein levels of the re sponsible channel isoform10,45.

PAGE 27

10 Additionally, left ventricular hypertrophy (LVH), an adaptive myocardial response to increased load, is associated with prol ongation of the ventricular AP due to remodeling of K+ channels46-48. IK,slow and Iss current densities and amplitudes are lower in mouse models of LVH and mRNA levels of the corresponding Kv channels, Kv1.5, Kv2.1, and TREK1, are altered as well49. Ionic currents and glycosylation, one form of post-translatio nal modification, are altered in heart failure (HF) models, illustrating the prevalence of remodeling in cardiac pathologies50-57. Alterations in ion channel function can impact excitability, resulting in pathological consequences Voltage-gated ion channels contribute to AP initiation and propagation in excitable cells. Thus, changes in the function of voltage-gated ion channels can impact cardiac, neuronal, and skeletal muscle AP waveform and duration. Pathologies such as Long QT Syndrom e (LQTS), deafness, and epilepsy can be produced by changes in ion channel function or distribution58-66. The focus of this study will be cardiac maladies. Mutations in eight different genes are corre lated to LQTS, which is characterized by delayed ventricular repolarizati on or a lengthened QT interval on the electrocardiogram (Figure 1. 4). In most instances, LQTS is generated by loss of potassium channel function or gain of func tion in sodium channels resulting from mutations in the alpha subunit of the channe l protein, beta/accessory subunits, or

PAGE 28

11 Figure 1.4. Electrocardiogram record ings of cardiac irregularities. Figure 1.4. Abnormal ECGs in humans and genetically engine ered mice. LQT: prolonged QT interval; TdP: torsadede-pointe episode; We: Wenckebach atrioventricular (AV) block; Avb: hi gh degree AV block. Fi gure from Nerbonne, et al ., 200167.

PAGE 29

12 modulatory proteins. For example, mutations in the genes coding for KvLQT1 (KCNQ1) and hERG, respectively re sponsible for production of IKs and IKr, can induce a reduction or loss of these current s, thereby decreasi ng the repolarizing current during the AP and leading to LQ TS. Disruption of sodium channel inactivation, brought on by mutations of t he channel protein, is causative of LQTS as well68-70. Furthermore, slight changes in ion channel genes, termed polymorphisms, may play a role in the varied respons e of patients carrying the modified gene to pharmaceutical agent s; acquired or drug-induced Long QT Syndrome can result71. Unlike LQTS, Brugada Syndrome, characterized by right bundle branch block, ST interval elevat ion, and ventricular tachyarrhythmias, results from loss of Na+ channel function and decreased INa 72. Brugada Syndrome can be caused by mutations in the cardiac sodium channel, Nav1.573,74. Members of the diverse potassium channel family function similarly The potassium channel family is a lar ge and varied designation of ion channels (Figure 1.5). In fact, isofo rms of potassium channels exist in bacteria, plants, and metazoan animals75; multiple, distinct K+ channels have been identified in evolutionarily early metazoan animals76. The greatest diversification of K+ channels is among voltage-gated K+ channels. Kv channels differ in many facets, including activation kinetics, ability to inactivate, and modulation by accessory subunits and intracellular messengers.

PAGE 30

13 Figure 1.5. Diversity of K+ channels. Figure 1.5. Potassium channels are a la rge and varied ion channel class. The most general grouping arranges the channe ls by membrane topology (2TM, 4TM, 6TM). Branching reflects the predict ed evolutionary relationships. The red box highlights the main classifi cation of voltage-gated potassium (Kv) channels. Figure from Hille, 200175.

PAGE 31

14 The primary responsibilities of K+ channels are to set the resting membrane potential and repolarize the cell subsequent to a depolarization by allowing the flow of potassium ions out of the cell; in general, these channels stabilize the membrane potential by drawing it toward the potassium equilibri um potential and, thus, away from the firing threshol d. Several subfamilies of voltage-gated potassium (Kv) channels exist and each performs this task (of repolarization) in a unique manner. Kv channels do, however, maintain a consistent, homologous structure. The alpha subunit, or primar y protein sequence encoding the poreforming subunit, is a tetramer of six trans membrane domains that join to form a functional channel. In addition, each monom er contains a voltage-sensor (S1-S4) and a pore-forming loop (S5-S6 linker) (Figure 1.6). The voltage sensor includes the first through fourth tr ansmembrane domains (S1-S4), with the S4 segment carrying much of the gating charge. The S4 segment contains positively charged ami no acids (arginine and lysine) termed gating charges at approximately every th ird residue; the first four arginines account for most of the gati ng charge. Approximately twelve to fourteen electron charges per tetrameric K+ channel (or 3 to 3.5 per subunit) cross the membrane voltage difference and constitute the gating charge77-79. The positive gating charges on the S4 segment are r epelled through the membrane in the extracellular direction when the membr ane is depolarized, allowing the channel to open75,80. Currently, two schools of thought exist as to how this is performed; however, it is widely accepted that amino acid residues of the S4 segment move

PAGE 32

15 Figure 1.6. Schematic of voltagegated potassium channel structure. ABC + + + + +S5 S6S4 Figure 1.6. Depiction of Kv channel structure and tetramerization. A A single subunit of Kv channels. Important structural aspects are noted, including the positively charged voltage sensor (o range) and the pore loop (purple). B An enlarged image of the pore loop, linking the fifth and sixth transmembrane segments (S5 and S6). C Illustration of the tetramerization of Kv channels. Four subunits converge to form a functional chann el. Ions move through the pore and are selected for passage through the channel via the selectivity filters. Figure adapted from Purves, 200481.

PAGE 33

16 towards the extracellular surface during a gating event82. The first theory, the conventional theory, describes the S4 segm ent as a cylindrical structure located within a water-filled column formed by t he other voltage-sensing segments (S1S3). With a depolarization, the S4 segm ent is propelled outward, perpendicular to the membrane (Figure 1.7A). In 2003, the MacKinnon group proposed that the S4 segment forms a paddle-like structur e that rotates th rough the hydrophobic membrane toward the extracellular space, resulting in a conf ormational change of the pore-forming segments; this c onformational change opens the channel (Figure 1.7B). Crystal st ructures of KvAP and Kv1.2 have substantiated this claim83-85. Although the theories shar e distinct similarities (in both models, the S4 segment is repelled in the extracellular direction through the lipid bilayer), much debate exists as to which is corre ct, with published data supporting both claims83-97. In addition to a voltage-sensor, Kv channels contain a pore-forming subunit, composed of the S5 and S6 segments, which line the pore. The extracellular linker between these segments forms a P-lo op, dipping into the pore (Figure 1.6). One distinct and unique feature of most potassium channels is their “signature sequence,” located within t he P-loop. The consensus amino acid sequence is – Thr-X-X-Thr-X-Gly-Tyr-Gly-Glu (TXXT XGYGD), where X is any amino acid98-100. Specifically, the residues Thr-X-Gly-Tyr-Gly (TXGYG) line the sele ctivity filter and are repeated on all four subuni ts. As the name suggests, the selectivity filter allows for the selection of potassi um ions through the pore. Most K+ channels are

PAGE 34

17 Figure 1.7. Representation of the voltage-sensor (S4) movement. B A Figure 1.7. The S4 segment moves th rough the membrane in response to a changing membrane potential, activating the channel. A The conventional model describes the S4 segments as cylinders moving perpendicular to the membrane. B The new model portrays the S4 segments as paddles rotating through the lipid membrane. += positively charged amino acid residues within the primary protein structure. Figur e from Jiang, et al ., 200383.

PAGE 35

18 permeable to only four cations—K+, Tl+, Rb+, and NH4 +—and are blocked by Cs+ because the selectivity filter is too narrow to allow for the passage of this larger cation75. Additionally, K+ channels are considered to have a long pore, with permeant ions lining up inside the pore in single file for passage; this is known as single-file diffusion101,102. This study concentrates on two types of voltage-gated potassium channels, delayed rectifie rs and A-type channels, which can be blocked by the pharmaceutical agent s tetraethylammonium (TEA) and 4aminopyridine (4-AP), among others103-106. Delayed rectifiers and KA channels contribute to the generation of voltagegated potassium currents Delayed rectifiers, also termed Kv channels, were first called thus because they change the membrane conductance wit h a delay following depolarization107. Due to the complexity of Kv channels, a nomenclature of Kvm n where m denotes the subfamily and n is the order of discovery, has been devised (Figure 1.5). For example, Kv1.5 is the fifth member of the Shaker subfamily of vertebrate Kv channels108. Most delayed rectifiers generate a rapidly activating, slowly inactivating current, IK,slow and Iss in mouse atria and ventricles and IKur in human atria. In murine ventricular myocyt es, two distinct components of IK,slow have been discovered, IK,slow1 and IK,slow2, which are produced by different Kv channel isoforms109-112. Both partially are responsible for the repolarizing currents generated in phase 3 of the murine AP (Figure 1.8)109,113,114. Interestingly, IKur (AP phase 3) is undetectable in ventricular myocytes of most species and is

PAGE 36

19 Figure 1.8. Human and murine ventri cular action potentials are generated by similar currents. AB Figure 1.8. A variety of ionic currents produce human ( A ) and mouse ( B ) ventricular action potentials. A considerable array of cardiac K+ currents repolarize the membrane. Figure from Nerbonne, 2004115.

PAGE 37

20 considered a potential ta rget for the treatment of atrial arrhythmias116-122. IKr and IKs are human delayed rectifier potassium cu rrents principally expressed in the ventricles and involved in phases 2 and 3 of the action potential, respectively; IKr activates rapidly, inactivates even more r apidly, and exhibits inward rectification, while IKs activates slowly and displays no inward rectification (Figure 1.8)123,124. Most delayed rectifiers undergo C-type inac tivation, a conformational change in the protein structure at t he extracellular region of t he pore, namely the outermost segment of the selectiv ity filter, causing it to close (Figure 1.9B)75. A-type K+ channels (KA), a subset of Kv channels, produce a rapidly activating and inactivating transient outward current, referred to as IA or Ito (Figure 1.8). From here on, this current will be referred to by the latter term, as it is the more widely accepted designation. Two component s of transient out ward current with distinct properties have been found and are termed Ito,fast (Ito,f) and Ito,slow (Ito,s)125,126. Typically, Ito,s recovers from inactivation at a much slower rate than Ito,f. The two currents are produced by func tionally different channels, but are, however, difficult to study individually126-128. KA channels (generating both Ito,f and Ito,s) activate transiently in the s ubthreshold range of membrane potentials129-133. Typically, KA channels activate upon depolariz ation of the membrane, subsequent to a period of hyperpolarizatio n, and inactivate prior to closing. Ito channels undergo rapid N-type inactivati on, otherwise known as the ball-andchain method of inactivation. In this me thod, the N-terminus occludes the pore, causing the channel to inactivate (Figure 1.9A). The positively charged ball,

PAGE 38

21 Figure 1.9. Kv channels can inactivate through two mechanisms. Figure 1.9. Illustration of Kv channel Nand C-type inactivation. A N-type inactivation follows activation and is otherwise known as the ball and chain method of inactivation. The first 22 resi dues of the N-terminus, or ball region, occludes the pore, blocking the flow of ions. B The mechanism underlying Ctype inactivation, or slow inactivation, involves conformational changes in the extracellular mouth of the pore, closing the channe l. Figure from Rasmusson, et al ., 1998134.

PAGE 39

22 which consists of the first 22 residues of the N-terminus, is necessary for inactivation to ensue. In fact, experi ments demonstrated that merely one ball peptide per channel is adequate to inacti vate the channel, albeit at a reduced inactivation rate (decreased to 25% of the original rate)135. The next 60 amino acids constitute the chain, the length of which determines the rate of inactivation. If the chain is lengthened, t he ball locates the receptor more slowly; conversely, shortening the chain will allo w the ball to find the receptor more rapidly. The channels producing Ito inactivate at relatively hyper polarized potentials; therefore, Ito can be separated from the total outwa rd current by applying a pre-pulse potential to eliminate Ito. The resulting currents can be subtracted from currents elicited without the inactivating pr e-pulse in order to examine Ito 133. This property of Ito allows for its involvement in phase 1 of the cardiac action potential (Figure 1.8). Kv channel isoforms respons ible for production of Ito and IK,slow are expressed differentially throughout the myocardium Several different Kv channel isoforms, each with distinct timeand voltagedependent properties, generate IK,slow and Ito. This study concentrates on five Kv channel isoforms, two underlying IK,slow and three producing Ito in murine myocytes. IK,slow1 previously was shown to be produced by Kv1.5 (gene KCNA5 subfamily Shaker), due to the selective elimination of IK,slow1 from mouse ventricular myocytes when Kv1.5 is deleted136,137. Similarly, Kv1.5 encodes the atrial-selective IKur in rat, canine, and human myocytes121,122,138-140. IK,slow2 is

PAGE 40

23 Table 1.1. The diversity of Kv channels and Kv currents throughout the mammalian heart. Kv Channel Subfamily Isoform Gene Cardiac Current Tissue* Species* Kv1/Shaker Kv1.4 KCNA4 Ito,s Atrium Rabbit VentricleMouse, Rat, Rabbit, Ferret, Human Node Rabbit Septum Mouse Kv1.5 KCNA5 IKur Atrium Rat, Dog, Human IK,slow1 Atrium Mouse VentricleMouse Kv2/Shab Kv2.1 KCNB1 IK,slow2 VentricleMouse Kv4/Shal Kv4.2 KCND2 Ito,f Atrium Mouse, Rat VentricleMouse, Rat Kv4.3 KCND3 Ito,f Atrium Mouse, Rat, Dog, Human VentricleMouse, Rat, Dog, Human Table 1.1. The table indicates the Kv channel and the current produced by the channel in each area of the heart in va rious species. = Currents have been identified in these tissue types/specie s. Table modified from Nerbonne, 200015.

PAGE 41

24 generated by Kv2.1 (gene KCNB1 subfamily Shab), with similar mRNA expression levels in the mouse atria and ventricles111,113,141. A great deal of evidence points to Kv4 alpha subunits as those responsible for production of Ito,f. Kv4.3 (gene KCND3 subfamily Shal) mRNA expression levels are detectable at similar levels in mouse atria and ventricle, while Kv4.2 (gene KCND2 subfamily Shal) expression is hig her in the ventricles141. Kv4.2 and Kv4.3 were thought to associate in the mouse ventricle, sugges ting heteromeric assembly of these two isoforms produces functional ventricular Ito,f. However, recent data suggest that Kv4.3 is not required for the gener ation of mouse ventricular Ito,f, through utilization of a Kv4.3(-/-) animal model142,143. In human and canine myocardium, Kv4.2 is not expresse d; therefore, Ito,f is thought to be produced by Kv4.3 channels144,145. In contrast, Ito,s is encoded by Kv1.4 (gene KCNA4 subfamily Shaker) in mouse, rat, ferret, canine, and human ventricular myocytes and rabbit atrial myocytes; however, in the murine myocardium, Kv1.4 appears to be primarily expressed in the inte rventricular septum (Table 1.1)146-152. The differential expression of these Kv isoforms throughout the myocardium parallels the variation in AP waveform in different regions of the heart. Kv channels are post-translationally modified Voltage-gated ion channels are heavily posttranslationally modified proteins; glycosylation, fatty acylati on, phosphorylation, nitrosyl ation, and sulfonation can be attached to these channels. Impor tantly, glycosylation constitutes approximately 30% of t he mass of an ion channel153-158. Studies have suggested

PAGE 42

25 that hundreds of sugar residues per f unctional molecule are attached to Na+ channel and some Shaker K+ channel proteins154,159-164. Furthermore, greater than 1% of the human genome is involved in glycosylation, equating to about 500 proteins aiding in the addition and removal of glycans165. The Consortium for Functional Glycomics (CFG) recogni zes 503 human genes involved in glycosylation. Due to the large number of genes in volved in glycosylation, alterations in normal gene expression could cause variations in the type or amount of glycosylation added to proteins or lipids. The impact of glycosylation is significant throughout development and disease The surfaces of most types of cells are heavily decorated with various types of glycoconjugates, with glycosylation attached to proteins and lipids. The biological and physiological roles of glycosylation ar e diverse due to their complexity and ubiquitous presence. Glycosylation plays protective, stabilizi ng, organizational, and barrier roles. Further, glycan involvem ent in protein trafficking, secretion, immunity, cell adhesion, receptor activation, and endocytosis has been established166,167. In fact, the ABO blood types (A, AB, B, O) are characterized as such based on their glycan structures, which behave as antigens. In this way, type B blood given to a patient with type A blood antigens will result in an adverse reaction.

PAGE 43

26 Two general forms of glycosylation have been identified—N-lin ked and O-linked glycosylation. Unlike protei n synthesis, glycosylation is not template-driven. However, the addition of glycosylation is a highly ordered and complex process; the product of one enzyme determines the next enzyme in the synthesis reaction and, therefore, the subsequent substrate. Catabolic glycosidases and anabolic glycosyltransferases share equally import ant roles, as both are required for glycosylation of prot eins and lipids. N-glycosylation, so named because of it s attachment to an asparagine residue (Asn or N), initiates in t he endoplasmic reticulum. An oligosaccharide precursor, composed of two core N-acetylglucos amines (GlcNAc) and five mannose residues, is linked to a dolichol-phosphate on the cytosolic side of the endoplasmic reticulum (ER). The lipid-link ed precursor then is flipped across the lipid bilayer into the lumen of the ER through a process that is not wholly understood. With the additi on of mannose and glucose residues, construction of the dolichol oligosaccharide precursor is complete. A multi-subunit protein complex, termed oligosaccharyltransfe rase (OST), transfers the dolichol oligosaccharide precursor en bloc to an asparagine residue on a nascently translated protein (Figur e 1.10). The minimal sequence requirement for the acceptor asparagine and its neighboring residues is the tripeptide sequon AsnXaa-Ser/Thr (NXS/T), where X is any amino acid with the exception of proline; this is referred to as the N-glycosylation consensus sequence168-170. Following covalent attachment of the oligosacc haride precursor to the Asn residue,

PAGE 44

27 Figure 1.10. N-glycan biosynthesi s in the Endoplasmic Reticulum. Figure 1.10. Schematic of N-glycan synthesis and en bloc transfer to nascent proteins via oligosaccharyltransferase (OST) in the Endoplasmic Reticulum (ER). An oligosaccharide precursor is linked to a dolichol-phosphate on the cytosolic side of the ER before flipping across the lipid bilayer into the lumen. Once in the lumen, further processing and transfer to an asparagine residue of a newly synthesized protein occurs. Fi gure adapted from Freeze and Aebi, 2005171.

PAGE 45

28 glucosidases sequentially remove all glucose residues. At this point, improperly folded proteins are reglucosylated, wher e they are retained in the ER to be refolded or degraded. The newly glycosyl ated protein (glycoprotein) is translocated to the cis -Golgi, where the mannoses are trimmed. Formation of complex glycans occurs in the medial and trans -Golgi, where addition of sialic acid (SA) terminates the process168-170,172,173. Sialic acids, negatively charged residues at physiologic pH, can be attac hed to other sialic acid residues, forming a poly-sialic acid structure; this stru cture can contribute a substantial negative charge to the protein. Synthesis of the other form of glycosyl ation, O-glycosylation, is not as well described as N-glycan biosynthesis Although no recognized consensus sequence has been identified, some predi ctive markers have been documented; furthermore, sites of O-gl ycosylation can be identified with only 70% precision174. O-glycans are attached to serine or th reonine residues, with prolines, alanines, serines, and threonines commonly found as neighboring residues. Prolines promote the formation of -turns and -sheets and O-glycans are located largely at the predicted -turns175-183. The complexity of O-glycan biosynthesis is great, as a large number of genes (as many as 24 N-acetylgalactosaminyltransferases) can commence the process184-186. The most prevalent form of O-glycosylation, the mucin type, initiates in the ER and refers to an N-acetylgalactosamine (GalNAc) -linked to the hydroxyl group of a serine or threonine; the addition of a galactose 1-3 linked to GalNAc occurs in the Golgi

PAGE 46

29 Figure 1.11. Biosynthesis of mammalian O-glycans. Figure 1.11. The synthesis of mammalian mucin type O-glycans initiates in the Endoplasmic Reticulum with the addition of N-acetylgalacto samine (GalNAc) onto the hydroxyl group of a serine (S) or threonine (T). Enzym es responsible for each substrate are shown. Further extens ion is not shown, including addition of sialic acid. Figure adapted from Tian and Ten Hagen, 2008187.

PAGE 47

30 (Figure 1.11)184,185. O-glycosylation structures can be further extended and compounded upon, similar to that of Nglycosylation. In general, however, Oglycans have a propensity to be less br anched than their N-linked counterparts, but are commonly biantennary structures te rminated by sialic acid residues. Throughout developmental and pathological states, glycosylation, including glycogene expression, is regulated in vari ous tissue types. The brain, thymus, lymph nodes, lung, liver, kidney, sp leen, testes, and bone marrow express distinct glycogene patterns and glycosylation structures188. Moreover, glycogene profiles are unique in each stage of devel opment in the cerebral cortex and myocardium189,190. Abnormalities in glycosylation are linked to a multitude of disorders, generally targeting the neuronal, muscular, and cardiovascular systems. A collection of diseases designated Congenital Disorders of Glycosylation (CDG) are heterogeneous autosomal recessive metabolic glycosylation disorders, resulting in prot ein defects. Currently, 28 types of CDG have been identified, which re sult from alterations in both Nand O-linked glycosylation; 16 forms are associ ated with N-glycosylation, 6 with Oglycosylation, 4 with Nand O-glycosyl ation, and 2 types impact glycolipids191. All 28 CDGs result in reduced sialylation and t he simplest, most reliable indicator of CDG is isoelectric focusing analysis of se rum transferrin sialylation. Specifically, CDG-IIf, a fatal disorder, is caused by a deficiency in the CMP-sialic acid transporter, resulting in loss of sialyl-Lex antigen192. Defects in the GNE/MNK gene influence the sialylation of only O-gl ycans and mutations in this gene have

PAGE 48

31 been described in Nonaka myopathy and sialuria. Nonaka myopathy, in which Oglycan sialylation levels are decr eased by 60-80%, presents as myopathic weakness and limb atrophy193. Conversely, hypersialylation of O-glycosylation structures and increased concen trations of urinary and fibr oblastic free sialic acid occur in patients with sialuria, with symptoms of mild psychomotor delay, persistent upper respiratory tr act infections, and hepatomegaly194-196. In addition, alterations in normal glycosylation pa tterns contribute to familial tumoral calcinosis and IgA nephropathy197-205. Chagas disease, a parasitic disorder tr ansmitted through insect bites and caused by Trypanosoma cruzi ( T. cruzi ), afflicts more than 18 million people worldwide. Chronic fibrotic myocarditis, commonl y leading to cardiac arrhythmias and insufficiency, and degeneration of autonomic nervous system innervated tissues are hallmarks of Chagas disease. T. cruzi cleaves sialic acid residues from host cells via release of a trans-sialidase and subsequently utilizes the sialic acids for its own requirements; this is believed to be an origin of the symptoms of Chagas disease206-208. Due to the ubiquitous and ab undant nature of glycosylation, remodeling of Nor O-glycans (including si alic acids) could potentially affect the function of glycoproteins, particu larly the gating of ion channels. Changes in glycosylation can al ter the gating of ion channels As previously stated, voltage-gated i on channels are heavily glycosylated proteins, with the number and location of glycosylation sites varying among

PAGE 49

32 channel isoforms. For example, Nav1.5, a cardiac sodium channel, contains 13 putative N-glycosylation sites on its extracellular linkers22,209. Many Kv channel isoforms are N-glycosylated as well. Kv1.4 and Kv1.5, Shaker Kv channel isoforms, possess one N-glycosylation site on their S1-S2 extracellular linkers. Interestingly, these sites are located only six amino acids apart210. The sole Nglycosylation site for Kv2.1, a Shab Kv channel, is positioned on its S3-S4 linker (Figure 1.12); much debate surrounding the gl ycosylation of this site exists due to its location, which may or may not be lo cated within the membrane (based on the new structural model). If so it may be inaccessible to the cell’s glycosylation machinery and therefore, not N-glycosylated84,85,211. Conversely, Kv4 channel isoforms, including Kv4.2 and Kv4.3, do not contain exter nal N-glycosylation sites and cannot be N-glycosylated. However, voltage-gated ion channels have the potential to be O-glycosylated, whether a put ative N-glycosylation site is present or not; that being said, si alylation may be present on most, if not all, voltagegated channels. The surface potential theory, initially developed by Frankenhaeuser and Hodgkin and expanded upon by Hille et al postulates that negative charges on the external surface of the cell membr ane generate a surface potential. These negative charges on the extracellular memb rane surface may be due to several factors, such as sialic acids, char ged lipids and proteins, and electrolytic screening in solution. Negatively charged sia lic acids are the most relevant sugar residue influencing ion channel gating. Thes e terminally located sugar residues

PAGE 50

33 Figure 1.12. Glycosylation sites of Kv channels are located on the extracellular linkers. Kv2.1 Kv1.5 Kv1.4 Figure 1.12. Crystal structure of Kv1.2 with glycosylation sites noted. Stereo image of Kv1.2 from the extracellular side of the membrane. Arro ws point to the Kv1.4, Kv1.5 and Kv2.1 potential N-glycosylation sites. Figure modified from Long, et al ., 200584.

PAGE 51

34 are thought to contribute to the exter nal negative surface potential, which can impact the voltage experienced by the positively charged S4 segment. As previously discussed, the S4 segment cont ains positively charged amino acids at approximately every third residue and when the membrane is depolarized, the S4 segment is repelled outward, openi ng the channel. The negatively charged surface potential, to which sialic acids potentially contribut e, may affect the positively charged S4 segment through elec trostatic interactions, allowing the channel to gate at more hy perpolarized potentials. The reverse holds true as well; reducing the sialic acid content will decrease the negative surface potential sensed by the S4 voltage-sensor and thus, a greater depolarization will be required to activate the channel212,213. In addition, the gating stabilizing theory predicts that a change in the conduct ance-voltage (G-V) slope denotes an alteration in functional state stability. Thus, modifications in glycosylation can affect channel gating by shifti ng the voltage of half-activation (Va) and altering the slope75. Several previous studies suggest that alte rations in glycosylation, specifically sialylation, can impact ion channel gat ing. In Chinese Hamster Ovary (CHO) cells, gating of Nav and Kv channels is modulated by a reduction in glycosylation. For example, gating of Nav1.4 and the Shaker Kv channel isoforms, Kv1.1, Kv1.2, and the Drosophila ShB channel, is shifted to mo re depolarized potentials with a reduction in sialylation and N-glycosylation163,209,214-216. Additionally, steady-state activation and inactivation of Kv4.3 were shifted in the depolarized direction when

PAGE 52

35 expressed in CHO cells with a decreased ability to sialylate proteins. Currents recorded from cardiomyocytes were m odulated upon treatment with glycosidases as well. The Kv current Ito was decreased in amp litude and the voltage of activation was shifted to depolarized potentials in adult murine ventricular myocytes upon treatment with neuraminidas e to cleave sialic acid residues; consistently, action potential durations were prolonged in neuraminidase-treated adult ventricular myocytes217. Myocardial Nav currents also are impacted by desialylation, demonstrating depolarizing shifts in gating57,218. Additionally, data suggest that rat neonatal and adult atrial and adult ventricular Nav channels are more sialylated and thus, more affected by changes in sialylation than neonatal ventricular Nav channels218. Therefore, effects of glycosylation on ion channel gating are isoform-specific. As formerly mentioned, several glycosylat ion disorders can severely impact normal functioning of the cardiovascular system, altering cardiac conduction and rhythm. Previously, no study has det ailed the effects of changes in Kv channel glycosylation on cardiac excitability. Here, we utilized in vivo and in vitro techniques to determine whether regulated or aberrant glycosylation (concentrating on sialylation) affect cardiac Kv currents and, as a consequence, cardiac excitability. In c hapter three, ECGs, cardiac APs, and two repolarizing Kv currents (Ito and IK,slow) were examined utilizing a kno ckout mouse model deficient in a single sialyltransferase, ST3Gal-IV. ST 3Gal-IV, a sialyltransferase (ST) that is localized to the Golgi, adds 2-3 linked sialic acids to Nand O-glycosylation

PAGE 53

36 structures219. Cardiac APs and Kv currents were measured under control conditions and upon enzymatic treatment to re move specific glycan structures in chapter four; atrial action potentials and Kv currents were more sensitive to changes in glycosylation (specifically, sialylation) than their ventricular counterparts. Chapters five and six in vestigated whether alterations in glycosylation impact five cardiac Kv channels responsible for IK,slow and Ito. The data confirm the isoform-specific effects of glycosylation on Kv channel gating and suggest O-linked sialic acids can impact channel function. These results substantiate previously published data and provide novel evidence detailing the effect of K+ channel Nand O-glycosylation on cardia c excitability; this is the first study to show O-glycans modulate ion channel gating. An improved understanding of how cardiac K+ channel function is modulated advances the insight into how alterations in no rmal processes lead to pathologies.

PAGE 54

37 CHAPTER 2 MATERIALS AND METHODS Chinese Hamster Ovary Cell Culture and Transfection Pro5 and Lec2 cells were grown in minima l media and transfe cted with channel cDNA as previously described220-223. Briefly, the cells were plated at 25-50% confluence on 35 mm dishes 24 hours prio r to transfection with 1ml Opti-MEM (Invitrogen) containing 8 l of lipofectamine (Invitrogen), 2.0 g of eGFP, and 2.5 g of channel cDNA and incubated at 37C in a 5% CO2 humidified incubator. 24 hours post-transfection, the media was r eplaced with growth medium, consisting of alpha-minimum essential medium ( MEM; Invitrogen) with (Pro5) and without (Lec2) riboand deoxyribonucleosides supplemented with 10% fetal bovine serum (FBS; Hyclone), 100 U ml-1 penicillin, and 100 g ml-1 streptomycin (Mediatech). Cells were incubated at 37C for another 48 hours prior to commencing electrophysiological recordings. Vector Construction and Mutagenesis The hKv1.4, hKv1.5, rKv2.1, rKv4.2, and rKv4.3 open reading frame (ORF) inserted into either pRC-CMV or pcDNA 3.1 (Invitrogen) were a gift from Dr. Stephen Korn (NIH, formerly University of Connecticut) and Dr. Jean Nerbonne (George Washington University). Vect or construction and mutagenesis were

PAGE 55

38 performed similar to that previously described221,222. Generally, the cDNA containing the -subunit open reading frame was inserted into pcDNA3.1. Mutagenesis was completed using the Stratagene Quickchange IIXL sitedirected mutagenesis kit and mutant constr ucts were verified through sequencing (see table below, bold text = Nglycosylation site mutated). Mutagenesis Primer Sequences Mutant Forward Primer Reverse Primer Kv1.4 N354Q GGT GGG TTG TTG CAA GAT ACT TCA GCA CCC CAT C GAT GGG GTG CTG AAG TAT C TT G CA ACA ACC CAC C Kv1.5 N290Q GCC CGC CCC TGG GGC C CA A GG CAG CGG GGT CAT GG CCA TGA CCC CGC TGC C TT G GG CCC CAG GGG CGG GC Kv1.5 S292A GCC CGC CCC TGG GGC CAA CGG C GC C GG GGT CAT GG CCA TGA CCC C GG C GC CGT TGG CCC CAG GGG CGG GC Kv2.1 N283Q CCT CAC AGA ATC C CA A AA GAG CGT GCT GC GCA GCA CGC TCT T TT G GG ATT CTG TGA GG Transgenic Mice The sialyltransferase knockout strain, ST3G al-IV, was originally supplied to us in collaboration with Dr. Jamey Marth (U niversity of California San Diego)219,224. The mice were housed and cared for in the anima l care facilities at the University of South Florida, College of Medicine. All protocols have been accepted by the Institutional Animal Care and Use Committe e. Genotypes of neonatal mice (2-3 days postnatal) were determined prior to ti ssue or cell isolation by acquiring tail clips, extracting the deoxyribonucleic acid (DNA), and performing polymerase

PAGE 56

39 chain reaction (PCR) experiments using the REDExtract-N-Amp Tissue PCR Kit (Sigma). Neonatal Cardiomyocyte Isolation for Electrophysiology The cardiomyocyte isolation protocol was adapted from a previously described method225. Postnatal day 2-3 mice were eut hanized and whole hearts excised and placed in 0 Ca2+Tyrode’s Solution. Atria and ventricles were cautiously dissected and digested in 1. 5 ml of 0.08% Type I Collagenase (Sigma) per ml 0 Ca2+Tyrode’s Solution at 37C in a 5% CO2 humidified incubator for 40 and 30 minutes, respectively. Cells were centri fuged at 160 x g for 5 minutes and the supernatant was replaced with fresh collagena se solution. After gentle trituration, atrial and ventricular cells were incubated at 37C for 30 and 20 minutes, respectively, followed by centrifugation fo r 5 minutes at 160 x g. The supernatant then was replaced by Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% feta l bovine serum, 100 U ml-1 penicillin, and 100 gml-1 streptomycin. The cell samples were trit urated, incubated at 37C for 20 minutes to stop digestion, and centrifuged at 160 x g for 5 minutes. Atrial and ventricular cells were plated separately on laminincoated (Sigma) 35 mm dishes in fresh media.

PAGE 57

40 Electrophysiology and Data Analysis Whole Cell Recordings in CHO Cells The Pro5/Lec2 expression system, cell lines of Chinese Hamster Ovary cells, has been used successfully to determine the effects of sialic acid on channel gating163,214-217,220. The Pro5 cell line allows normal protein sialylation, while the Lec2 cell line produces essentially non-sial ylated proteins due to a deficiency in the CMP-sialic acid transporter226-228. The Lec2 cell line used in this study serves as a model for CDG type-IIf192,229-231. Whole cell current recordings were perfo rmed using pulse protocols, solutions, whole cell patch clamp techniques and data analyses as previously described209,220-222. An Axon 200B patch-clamp amp lifier in combination with a CV203BU headstage (Axon Instrum ents) and a Nikon TE200 inverted microscope were used. Pulse acquisi tion software (HEKA) operating on a Pentium III computer (Dell Computers) was utilized for pulse protocol generation. The ensuing analog signals were digiti zed using an ITC-16 AD/DA converter (Instrutech). All experiments were c onducted at room temperature (~22C). 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, 1 MgCl2, 2 CaCl2, 155 sucrose, 5 glucose, 10 Hepes (pH 7.3), while the intracellular solution used was (mM): 70 KCl, 65 KF, 5 NaCl, 1 MgCl2, 10 EGTA, 5 glucose, 10 Hepes (pH 7.3). For t he low divalent cation studies, the

PAGE 58

41 extracellular solution was ident ical to that listed above, wi th the exception of a 0.2 mM concentration of CaCl2 and a 0.1 mM concentration of MgCl2. Immediately prior to use, all solutions we re filtered independently with a 0.2 m Gelman filter. To ensure complete dialysis of the intr acellular solution, data was collected at least 10 minutes after attaini ng whole cell configuration. Conductance-Voltage Relationship for CHO Cell Recordings Steady state and kinetic gating parameter s were examined through the use of standard pulse protocols and solutions described by our lab and others209,216218,220-222. Cells were held at -120 mV, stepped to more depolarized potentials (-100 mV to +40 mV in 10 mV increm ents) for 100 ms, and returned to the holding potential. Consecutive pulses were stepped every 1.5 s and the data were leak subtracted using the P/4 met hod, stepping negatively from the holding potential. Steady state whole cell conduc tance values (G) were determined by measuring the peak current (I ) at each test potential (Vp) and predicting a K+ Nernst equilibrium potential (Ek= -84 mV) using Ohm’s law (G = I/(Vp – Ek)). 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 (equat ion 1, solving for maximal conductance). These single Boltzmann distributions were used to determine the average Va SEM and Ka 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

PAGE 59

42 Boltzmann relation below: Fraction of maximal conductance = [1 + exp(-(V-Va/Ka))]-1 equation (1) where V is the membrane potential, Va is the voltage of half-activation, and Ka is the slope factor. Steady-State Inactivation (hinf) Cells were pre-pulsed from -120 mV to +20 mV (10 mV increments) for 1000 ms, then stepped to +60 mV for 100 ms, and re turned to the holding potential (-120 mV). The maximum current generated by each cell was used to normalize the data for each cell to its maximum current by fitting the data to a single Boltzmann distribution (equation 2, solving for ma ximal current), from which the mean Vi SEM and Ki SEM values were determined. Fraction of maximal current = [1 + exp-((V-Vi)/Ki)]-1 equation (2) where V is the membrane potential, Vi is the voltage of half-inactivation, and Ki is the slope factor. Time Constants of Activation ( n) Activation time constants were determined by fitting the current traces used to measure the G-V relationships. Whole current traces were fitted using the Hodgkin-Huxley function of the PulseFit software suit (HEKA). The current trace data were fitted to a fourth power exponential functi on to determine n.

PAGE 60

43 Recovery from Inactivation Cells were held at -120 mV and stepped to +60 mV for 100 ms and subsequently returned to the recovery potential for various time intervals (10 – 200 ms in 10 ms increments). The potential then was stepped to +60 mV for 100 ms. Peak currents measured during the +60 mV depol arizations were compared and the fractional peak current that remains dur ing the second depolarization was plotted as a function of the recovery pulse durat ion, representative of the fraction of channels recovered from inactivation during the recovery interval. Recovery time constants were determined by fitting t he data to a single exponential function (HEKA). Cardiomyocyte Potassium Current Recordings Potassium currents were collected from neonatal myocytes using the previously described recording procedures and equipm ent. All experiments were performed at room temperature usi ng whole cell patch clamp te chniques, pulse protocols, solutions, and data analyses similar to those described by us and others21,141,218. Desialylation or deglycosylation of card iac myocytes was performed prior to recordings through 2 hour treatments at 37 C with either 0.01 units of sialidase A (Neuraminidase; Glykome) or 5 units of PNGase-F (Sigma) in base media218. Sialidase A cleaves 2-6, 2-3, and, to a lesser extent, 2-8 and 2-9 sialic acid linkages, while PNGase-F removes N-ac etylglucosamine at the asparagine residue. Myocytes were patched and re corded from in the presence of extracellular solution consisting of (mM): 130 NaCl, 5.4 KCl, 1 CaCl2, 1 MgCl2,

PAGE 61

44 0.33 Na2HPO4, 10 HEPES, 5.5 glucos e (pH 7.4). The intracellular solution was (mM): 110 K-Asp, 20 KCl, 8 NaCl, 1 CaCl2, 1 MgCl2, 10 EGTA, 10 HEPES, 4 K2ATP (pH 7.2)141. Again, all solutions were filtered using 0.2 m Gelman filters prior to use. 9 M tetrodotoxin (TTX; Tocris Coo kson) was added to the filtered extracellular solution before data collection to block Nav channels and inhibit sodium currents. Cardiomyocyte Ito and IK,slow G-V Relationships Ito and IK,slow were recorded from cardiomyocytes as previously described141,232. IK were elicited by holding at -80 mV and initiating 500 ms volt age-clamp steps from -110 mV to +50 mV in 10 mV increment s. A 100 ms pre-pulse (-60 to -70 mV) before the main activation steps was ut ilized to eliminate any remaining INa (not blocked by TTX) and record Ito and IK,slow. IK,slow was isolated by first applying an inactivating pre-pulse of -40 mV for 100 ms before the activation steps listed for recording IK in order to eliminate Ito and INa. These current traces were then subtracted from the total IK traces to isolate Ito. Data analysis was performed using PulseFit (HEKA). Action Potential Recordings from Cardiomyocytes Action potentials were recorded at room temperature (~22C) from myocytes using an Axon 200B patch-clamp amp lifier with a CV203BU headstage (Axon Instruments) and a Nikon TE200 inverted mi croscope. pClamp 9 software (Axon) operating on a Pentium III computer (De ll Computers) was ut ilized for protocol

PAGE 62

45 generation. The extracellular solution was (mM): 135 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, 10 glucose (pH 7.4). Pipette s were filled with the intracellular solution (mM): 110 K-Asp, 20 KCl, 10 NaCl, 4 ATP-Mg, 10 HEPES (pH 7.3)217. Removal of glycosylation structures pr ior to recordings was performed as described above. Action Potential Protocols Cells were held at a holding potential of -80 mV. Subsequently, APs were elicited using a 2 ms superthreshold depolarizing current at a frequency of 1 Hz from the resting membrane potential. Data were analyzed using the Clampfit 9 software (Axon Instuments). Whole Cell Homogenization Cells were rinsed with cold PBS and incubated on ice for 5 minutes in cold sodium pyrophosphate buffer with prot ease inhibitors (PI; 20 mmol/liter tetrasodium pyrophosphat e, 20 mmol/liter Na2PO4, 1 mmol/liter MgCl2, 0.5 mmol/liter EDTA, 300 mmol/ liter sucrose, 0.8 mmo l/liter benzamidine, 1 mmol/liter iodacetamide, 1.1 mol/liter leupeptin, 0.7 M pepstatin, 76.8 nM aprotinin). Cells were then homogenize d using manual tissue grinders. The homogenates were centrifuged for 10 minutes at 1000 x g in a Beckman benchtop centrifuge at 4C. The superna tant was centrifuged in an Eppendorf ultracentrifuge for one hour at 50,000 rpm (4C) after which, the pellet was resuspended in an appropriate volu me of sodium pyrophosphate buffer containing PIs. The lysates were then st ored at -80C and protein levels were

PAGE 63

46 determined using the Pierce BCA Protein Assay kit and a Beckman DU 530 spectrophotometer218. Immunoblots Immunoblot gel shift analysis was performed as described previously209,220,221. Briefly, cell homogenates (2-55 g/lane) were combined with one volume of 2x sample buffer (10% glycerol, 5% 2-mercaptoethanol, 3% sodium dodecyl sulfate, and 12.5% upper Tris buffer) and denatured fo r 3 minutes in boiling water. Samples were then run on 6.5% SDS-PA GE gels for 90 minutes at 75-110 mV and transferred onto nitrocellulose membranes using a semi-dry transfer cell (BioRad). Channel -subunits were detected through site-directed polyclonal or monoclonal antibodies. Monoclonal anti-Kv1.4, anti-Kv2.1, anti-Kv4.2, and antiKv4.3 antibodies were utilized to detect Kv1.4, Kv2.1, Kv4.2 and Kv4.3, respectively (1:1000 dilution, NeuroMab).A polyclonal anti-Kv1.5 antibody generated against the previously described233 C-terminal human Kv1.5 sequence (C)EQGTQSQGPGLDRGVQR was used (1:50 d ilution; Biosynthesis, Inc) for Kv1.5 detection. After incubat ion with primary antibody, the blot was treated with either goat anti-mouse horseradish per oxidase-conjugated (1:25,000 dilution; Jackson) or donkey anti-rabbit horser adish peroxidase-conjugated (1:4500 dilution; Amersham Biosciences) sec ondary antibodies and visualized using an enhanced chemiluminescence kit (Pierce). Deglycosylation of homogenates was performed through 3 hour treatments at 37C with 5 units of PNGase-F/10 g of protein (Sigma)218.

PAGE 64

47 Click-iT Glycoprotein Labeling and Detection Click-iT Glycoprotein Metabolic Labeli ng Reagents and Detection Kits were utilized for these reactions (Invitrogen, Mo lecular Probes). Cells were transfected with channel cDNA as previously described216,220-222. Eight hours posttransfection, the cells were replated in the presence of the tetraacetylated azido sugar of interest at a concentration of 2.5 million cells/100mm dish and incubated for 48-72 hours. Tetraacetylated azido-m odified sugars were incorporated onto protein glycan structures due to the permissi ve characteristics of oligosaccharide synthesis. N-azidoacetylgalactosamine (Ac4GalNAz) was metabolically integrated into O-glycosylation structures th rough the N-acetylgalactosamine salvage pathway. N-azidoacetylmannosamine (Ac4ManNAz) was utilized to incorporate a modified sugar into the sialic acid biosynthesis pathway. The cells were harvested and lysed. Azide-modified gl ycoprotein samples then were labeled with a biotin-alkyne and precipitated185,234-239. The precipitated sample was resolubilized and incubated on streptav idin-bound T1 Dynabeads (Invitrogen). Because of the strong interaction betw een streptavidin and biotin, this step allows for isolation of only those prot eins with the modified sugar incorporated into their glycosylation structures. Immunoblot analysis was performed on the samples and channels of interest were probed for using channel-specific antibodies, as previously described218,220.

PAGE 65

48 Cardiac Tissue Isolation Neonatal atrial and ventricular sample s were isolated for quantitative PCR experiments. The animals were euthaniz ed and whole hearts were extracted and placed in 0 Ca2+Tyrode’s solution. The atria and ventricles were carefully dissected from the heart to ensue only t hose tissues were removed and the remaining portion of the heart was discard ed. The tissue was placed in RNAlater solution (Qiagen) and incubated at room te mperature (~22C) for a minimum of one hour before storage at -80C. RNA Isolation and Reverse Transcription Cardiac samples from control and ST3Gal-IV(-/-) animals were isolated as described above for quantitative PCR experiment s. Atrial and ventricular samples were composed of 2-6 animals, yieldi ng approximately 10-25 mg of tissue for ribonucleic acid (RNA) isolation. Sa mples were homogenized using douncers and RNA was isolated at r oom temperature (~22C) according to the RNeasy protocol (Qiagen). A Beckman DU 530 Spectrophotometer was used to determine the final RNA concentration. 3 g of RNA were utilized for the reverse transcription reaction, which was perform ed following the SuperScript II Reverse Transcriptase protocol (Inv itrogen), to yield cDNA fo r real time reactions. Quantitative RT-PCR Quantitative RT-PCR was performed on 5 Kv -subunit gene products and one constitutively expressed housekeeping gene,

PAGE 66

49 hypoxanthinephosphoribosyltransfera se (HPRT), as a control. Kv1.5, Kv2.1, Kv4.2, Kv4.3, and HPRT primer sets were designed using PrimerQuest (IDT DNA), while Kv1.4 primer sets were acquire d from a previous study (see below)240. The primer sets were run in triplicate for each sample. 25 l total of SYBR Green PCR master mix (Superarray) water, cDNA, and primers were added to each well of a 96 well PCR plat e (BioRad) and covered with RT-PCR optical tape (BioRad). The PCR plate wa s then centrifuged for 5 minutes at 1000 rpm and loaded into the Chroma 4 detec tion system (BioRad). Samples were detected using the following protocol: 95C for 5 min preceding 40 cycles of 95C for 30 seconds, 61C for 30 seconds, and 30 seconds at 72C. Relative expression levels were determined by the CT analysis method. The threshold values of a single gene were averaged and compared to the control values (HPRT). The resulting CT values are then compared to another sample. Analysis was performed using Opticon Monito r (BioRad) and Microsoft Excel. Quantitative RT-PCR Primer Sequences Gene Forward Primer Reverse Primer Kv1.4 CAT TTG GTT TCC CAA TGG TC GTG GTG CAT TCC TTG TTC CT Kv1.5 TTC GCA GAG GCA GAC AAT CAG G AGG CAG AGC AAT GGT GAG GA Kv2.1 CCC AGT CTC AAC CCA TCC TCA A TGC TGC CCA TCT CCA GTT CT Kv4.2 TGA GCG GAG TCT TGG TCA TTG C TGT GCC CTT CGT TTG TCT GC Kv4.3 TTT GCC TGG ACA CTG CCT GT CAC ACT GCG GAT GAA GCG GTA T HPRT GCA GTA CAG CCC CAA AAT GG GGT CCT TTT CAC CAG CAA GCT

PAGE 67

50 Telemetric Electrocardiogram Recordings Electrocardiograms were recorded for at least 48 hours on conscious adult (1012 week) mice using a Data Exchange Matr ix, RPC-1 PhysioTel Receivers, and TA10ETA-F20 Implants from DSI. The tr ansmitters were implanted into the abdominal cavity of anesthetized animals and the leads sutured to thorax muscles in the upper right section and near the apex of the heart. Following transmitter implantation, the animals were permitted to recover for one week prior to data collection. ECGs were recor ded continuously for approximately 48 hours and stored on DVDs. Data were analyzed using Data Quest ART and Ponemah PNM-P3P-ECG software (DSI). Data Analysis and Statistics Data and statistical analyses were performed using Pulse/PulseFit (HEKA), Clampfit (Axon), Data Quest ART, Ponemah PNM-P3P-ECG, Opticon Monitor (BioRad), Microsoft Excel, and Si gma Plot (SSPS Inc.). Student’s t -tests were performed on the data produc ed by the electrophysiological recordings. A p value of <0.05 was used to determine w hether a significant difference exists.

PAGE 68

51 CHAPTER 3 THE SIALYLTRANSFERASE, ST3GAL -IV, ALTERS CARDIAC ACTION POTENTIAL WAVEFORMS AND IK Precise gating of repolarizing voltagegated potassium channels is a crucial component in producing proper cardiac ac tion potential waveform and duration. Alterations in normal Kv channel activity can modify AP repolarization, thereby changing its properties, as is observed throughout various physiological and pathological processes11,241. Glycosylation is abundant on ion channel proteins, including Kv channels157. These attached glycans are generally terminated by sialic acid, which have been shown to modulate Kv channel gating through isoform-specific mechanisms163,214,216,217. ST3Gal-IV, a sialyltransferase that adds 2-3 linked sialic acid residues to Nand O-linked glycans, is the initiating enzyme for SA addition to a galactose moiety242. ST3Gal-IV is expressed at uni form levels throughout cardiac development243. One method available to test whether SAs exert an effect on cardiac excitability was to utiliz e a ST3Gal-IV knockout (ST3Gal-IV(-/-)) mouse model. Electrocardiograms, AP waveform pa rameters, and two types of voltagegated K+ currents (IK), the transient outward Ito and the slowly inactivating IK,slow, were compared in atrial and ventricula r myocytes isolated from control and

PAGE 69

52 ST3Gal-IV(-/-) animals. A single, previous stud y conducted by our lab determined that loss of the polysialyltransferase ST8Sia2 (STX), shifts the voltagedependence of atrial Nav currents to more positive pot entials, slowing the rate of AP depolarization and decreasing peak ampl itude. Here, we found ST3Gal-IV can impact atrial Kv currents and AP duration, thus potentially altering cardiac excitability. Results Electrocardiograms suggest slowed hear t rates and altered waveforms in the ST3Gal-IV(-/-) animal To question whether cardiac excitability is impacted by aberrant sialylation that results from the loss of a single sialyltrans ferase, telemetry was utilized to record ECGs from conscious, unrestrained adult mice. Telemetry studies, in which a radio transmitter is implanted into the abdomen, were performed for approximately 48 hours; these extended reco rding durations should increase the probability of observing great er variation in heart ra te and any additional intermittent conduction problem s, if present. One limitati on to telemetry is that these experiments cannot be conducted in neonatal mice, due to the size of the animal relative to the transmitter. Howe ver, ST3Gal-IV is uniformly expressed throughout cardiac development243; thus, the impact of the loss of this sialyltransferase on neonatal cardiac exci tability might be echoed in the ECGs recorded from adult mice.

PAGE 70

53 Figure 3.1. ECGs recorded from conscious, unrestrained control and ST3Gal-IV(-/-) mice suggest altered cardiac rhythm in the knockout. 0.1 s 0.1 s 0.1 s KO1 610 0.54 BPM KO2 628 0.45 BPM Control 664 0.38 BPM 0.1 s 0.1 s 0.1 s 0.1 s 0.1 s 0.1 s KO1 610 0.54 BPM KO2 628 0.45 BPM Control 664 0.38 BPM RR P waveKO1 KO2 ControlAB C mV mV mV0.2s KO1 KO2 Control Figure 3.1. A 1 second sampling of ECGs recorded from two ST3Gal-IV(-/-) mice (KO, black and blue) and one littermate cont rol mouse (green). Note the slowed heart rate in the KOs co mpared to the control, B Enlarged view of two QRS segments from two ST3Gal-IV KOs and one littermate control. The P wave and the RR segment are lengthened in the KO compared to that of t he control.

PAGE 71

54 C Arrhythmias are noted at an in creased frequency in the ST3Gal-IV(-/-) mouse. Red arrows denote three arrhythmic beats in one ST3Gal-IV(-/-) mouse. Data were analyzed using Data Quest AR T and Ponemah PNM-P3P-ECG software (DSI).

PAGE 72

55 ECG data recorded from ST3Gal-IV(-/-) and littermate control mice suggest heart rate is slowed in the knockouts relative to the controls (Figure 3.1). However, when analyzed further, no significant di fference in heart rate was observed between knockout and control mice. This likely is attributable to the analysis software, which averages the mean hear t rate over the approximately 48 hour interval. Additionally, the P waves, i ndicating atrial depolarization, and R-R segments appear to be lengt hened in the ST3Gal-IV(-/-) mouse compared to that of the littermate control (Figure 3.1). As observed in Figure 3.1, an increase in arrhythmic events (lasting for nearly the ent ire recording duration at inconsistent frequencies) has been noted in the knocko ut animals. Three conditions of apparent conduction abnormalities were detect ed in nearly half of the ST3Gal-IV(/-) mice (four of nine). Two of the ST3Gal-IV(-/-) mice exhibited variations in the spacing between beats, while one mous e had premature arrhythmic beats (Figure 3.1C). Both conduction abnormalit ies were identified in one knockout mouse; no apparent conduction disturbances were detected in the littermate controls. Unfortunately, the software av ailable does not have the capabilities to allow for more thorough analysis of the data. Although these data cannot firmly dictate how the loss of a si ngle sialyltransferase can a ffect cardiac excitability, they do indicate alterations in conduction and rhythm.

PAGE 73

56 Action potential duration is prolonged in ST3Gal-IV(-/-) atrial myocytes, yet is unaltered in ventricular myocytes Action potential duration is a crucial component in the generation of normal cardiac conduction and rhythm. In this st udy, action potentials were elicited from control and ST3Gal-IV(-/-) neonatal atrial and ventricula r myocytes. Atrial ST3GalIV(-/-) myocytes displayed lengthened repolar ization phases, as indicated in Figure 3.2. The mean action potent ial duration (APD) at 10% (APD10), 50% (APD50), and 90% (APD90) repolarization was extende d significantly in the knockout atria (by ~50-80%) compared to t he control (Figure 3. 2, Table 3.1). The time to peak was significant ly lengthened in ST3Gal-IV(-/-) atrial myocytes, which is associated with the involvement of sodium currents in AP production. Furthermore, since repolarization rate is af fected dramatically in knockout atrial myocytes and Kv currents underlie the AP repolar ization phase, these data suggest that ST3Gal-IV(-/-) atrial K+ currents may be altered. To determine the effect of ST3Gal-IV on neonatal ventricular myocyte excitability, APs were measured from ST3Gal-IV(-/-) and control ventricular myocytes. Unlike that observed for the atria, ventricula r APD was not impacted by the absence of ST3Gal-IV; APD10, APD50, and APD90 values did not differ significantly between control and knockout ventricular myocyt es (Figure 3.3). The peak AP voltage and the time to peak were unaffected as well. Thus, the data suggest that ST3Gal-IV is not required for normal neonatal v entricular AP production (Table 3.1).

PAGE 74

57 Figure 3.2. Action potential durat ion is prolonged in ST3Gal-IV(-/-) atrial myocytes. APD90 (ms) 0 10 20 30 40 APD50 (ms) 0 5 10 15 Control NA ST3Gal-IV(-/-) NA Control NA ST3Gal-IV(-/-) NA APD10 (ms) 0 2 4 6 8 Control NA ST3Gal-IV(-/-) NA 200 ms50 mV 200 ms20 mV Control ST3Gal-IV(-/-) * A CD E B Figure 3.2. A Trains of action potentials elic ited from control (dark blue) and ST3Gal-IV(-/-) (light blue) neonatal atrial myocyt es. Note the prolongation of the ST3Gal-IV(-/-) AP repolarization phase, B Representative action potential traces from control (dark blue) and ST3Gal-IV(-/-) (light blue) atrial myocytes, C D & E Bar graphs represent c ontrol and ST3Gal-IV(-/-) atrial action potential durations at 10 % repolarization ( C APD10), 50% repolarization ( D APD50), and 90%

PAGE 75

58 repolarization ( E APD90). APD10, APD50, and APD90 are significantly lengthened in ST3Gal-IV(-/-) atrial myocytes compared to control (Table 3.1). Data are Mean SEM. n = 9-10. *=significant compared to control ( p <0.005).

PAGE 76

59 Figure 3.3. Ventricular action potential dur ation is not impacted by ST3Gal-IV. CDE AB APD50 (ms) 0 5 10 15 20 25 APD90 (ms) 0 15 30 45 60 Control NV ST3Gal-IV(-/-) NV Control NV ST3Gal-IV(-/-) NV APD10 (ms) 0 2 4 6 Control NV ST3Gal-IV(-/-) NV Figure 3.3. A Trains of action potentials elic ited from control (dark green) and ST3Gal-IV(-/-) (light green) neonatal ventricular myocytes, B Representative traces of action potentials recorded fr om control (dark green) and ST3Gal-IV(-/-) (light green) ventricular myocytes, C D & E Bar graphs represent ventricular action potential durations at 10 % repolarization ( C APD10), 50% repolarization ( D APD50), and 90% repolarization ( E APD90) recorded from control and ST3Gal-IV(-/-) ventricular myocytes. APD10, APD50, and APD90 are not significantly

PAGE 77

60 altered in ST3Gal-IV(-/-) ventricular myocytes compared to control ( p >0.1, Table 3.1). Data are Mean SEM. n = 9-13.

PAGE 78

61 Table 3.1. The measured action potential parameters for control and ST3Gal-IV(-/-) atria and ventricles. Cell Type n AP Peak V Time to Peak APD10 APD50 APD90 mV ms ms ms ms Control Atria 9 120.1 4.8 3.9 0. 1 4.4 0.2 9.0 0.8 22.3 2.6 ST3Gal-IV ( / ) Atria 10 120.1 6.3 5.0 0.2* 6.6 0.4* 14.2 1.6* 40.1 3.1* Control Ventricle 13 138.3 4.3 3.9 0.1 5.7 0.3 21.7 2.5 56.7 6.2 ST3Gal-IV ( / ) Ventricle 9 147.7 2.9 3.7 0.2 5.5 0.3 18.9 1.3 59.8 5.6 Table 3.1. Data are the mean SE M. APD: Action potential duration. Significance was determined using a twotailed Student’s ttest comparing ST3Gal-IV(-/-) action potential parameters to control. = significance ( p <0.005).

PAGE 79

62 The voltage-dependence of Ito and IK,slow activation is altere d in the atria, but is unaffected in the ventricles of ST3Gal-IV(-/-) animals As previously stated, vo ltage-gated potassium currents are responsible for the repolarization phase of the AP. Since at rial AP duration was prolonged in the absence of ST3Gal-IV (Figure 3.2), two Kv currents involved in AP repolarization were examined in control and knockout neonatal atrial myocytes. The transient portion (Ito) of IK was eliminated by application of an inactivating pre-pulse (-40 mV, 100 ms) prior to the ac tivation protocol for IK,slow. The remaining current, IK,slow, was measured and analyzed (Figures 3. 4 and 3.5). Figure 3.4 shows atrial IK,slow activates at more depolarized potentials in ST3Gal-IV(-/-) myocytes compared to control. The conductance-vo ltage relationship was shifted by a depolarizing ~14 mV in the knockout from that observed for the control. Further, the Va of IK,slow measured from ST3Gal-IV(-/-) atrial myocytes was shifted to more depolarized potentials co mpared to control Va. Thus, a stronger depolarization is necessary to activate IK,slow in the knockout atria (F igure 3.4, Table 3.2). Ito was obtained by subtracting the current traces measured wit h and without the inactivating pre-pulse (IK,slow was subtracted from IK). G-V relationships for Ito measured from control and ST3Gal-IV(-/-) atrial myocytes are shown in Figure 3.4. Similar to that of IK,slow, the G-V relationships and Va of atrial Ito were shifted to more depolarized potentials in knockout myocytes as compared to control; a positive shift of ~15 mV was observed for ST3Gal-IV(-/-) atrial Ito (Table 3.2).

PAGE 80

63 Figure 3.4. Atrial Ito and IK,slow activation are altered by loss of ST3Gal-IV. Membrane Potential (mV) -80-60-40-20020Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Ito Control Ito ST3Gal-IV(-/-) Membrane Potential (mV) -80-60-40-20020Normalized Conductance 0.00 0.25 0.50 0.75 1.00 IK,slow Control IK,slow ST3Gal-IV(-/-) 50 pA50 msC A D B Figure 3.4. A Representative atrial myocyte voltage-gated potassium current (IK) traces. IK consists of Ito and IK,slow, which can be separated experimentally, B Representative IK,slow current traces recorded from atrial my ocytes, C & D Steady-state activation of Ito and IK,slow recorded from control (dark blue) and ST3Gal-IV(-/-) (light blue) atrial myocyt es. Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Here and throughout, error bars are smalle r than the symbol if not visible. n = 5 – 7 (Table 3.2), C G-V relationships for Ito activation (circles). Note the ~15 mV shift to more depolarized pot entials in the ST3Gal-IV(-/-) myocytes compared to

PAGE 81

64 the controls, D G-V relationships for st eady-state activation of IK,slow (squares). A ~14 mV depolarizing shift in steady-state activation was observed in the ST3GalIV(-/-) atrial myocytes compared to the controls.

PAGE 82

65 Although ST3Gal-IV(-/-) ventricular myocytes demonstrated no measureable change in APD, ventricular Ito and IK,slow were examined to confirm ventricular K currents were not impacted by the absence of this sialyltransferase. As predicted, steady-state activation of Ito and IK,slow recorded from ST3Gal-IV(-/-) and control ventricular myocytes was not different significantly (Figure 3.5, Table 3.2). Therefore, these data s uggest that production of normal neonatal ventricular Ito and IK,slow was maintained in the absence of ST3Gal-IV. Discussion Changes in normal cardiac conduction may pose serious pathological consequences; therefore, the data present ed here may provide some insight into how excitability is modul ated throughout the heart. ECG recordings suggest mice deficient in the sialyltransferase ST3G al-IV demonstrate slow er heart rates than littermate controls. Parasympathetic input controls the beat-t o-beat pacing of the heart. Thus, loss of ST3Gal-IV may modul ate these inputs, resulting in the altered heart rate observed in ST3Gal-IV(-/-) mice. Broadened P waves were noted in the knockout as well, sugges ting a slowing of atrial conduction potentially through alterations in Nav or Kv channel function (Figure 3.1). While not entirely quantitative, the data indicate a trend toward altered cardiac conduction in the knockout animal, as an increased incidence of arrhythmic events was observed. The arrhythm ogenic trend noted in the ST3Gal-IV(-/-) animals may impact AP production by altering the gating of ion channels, including potassium, sodium, or hyperpol arization-activated cyclic nucleotide-

PAGE 83

66 Figure 3.5. ST3Gal-IV deficiency does not affect ventricular Ito and IK,slow. C A D B Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Ito Control Ito ST3Gal-IV(-/-) Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 IK,slow Control IK,slow ST3Gal-IV(-/-) Figure 3.5. A Representative voltage-gat ed potassium current (IK) traces recorded from ventricular myocytes. IK consists of Ito and IK,slow, which can be separated experimentally, B Representative ventricular IK,slow current traces, C & D Steady-state activation of Ito and IK,slow recorded from cont rol (dark green) and ST3Gal-IV(-/-)(light green) ventricular myocytes. No significant shift in steadystate activation was noted in ST3Gal-IV(-/-) ventricular myocytes compared to controls ( p >0.1). Data are the mean norma lized peak conductance SEM and are fit to single a Boltzm ann relationship (lines). n = 4 (Table 3.2), C G-V relationships for Ito activation (circles), D G-V relationships for steady-state activation of IK,slow (squares).

PAGE 84

67 Table 3.2. Measured gating paramete rs for atrial and ventricular Ito and IK,slow from control and ST3Gal-IV(-/-) mice. Current n Va Ka mV mV Control Atrial Ito 7 -34.5 1.7 6.7 0.8 ST3Gal-IV ( / ) Atrial Ito 5 -19.2 1.2* 8.1 0.9 Control Atrial IK,slow 7 -40.4 1.3 8.3 0.6 ST3Gal-IV ( / ) Atrial IK,slow 5 -26.8 2.9* 7.9 0.7 Control Ventricular Ito 4 -8.9 1.6 13.4 1.0 ST3Gal-IV ( / ) Ventricular Ito 4 -9.6 2.3 12.1 1.3 Control Ventricular IK,slow 4 -11.3 1.3 12.7 0.5 ST3Gal-IV ( / ) Ventricular IK,slow4 -10.0 2.0 12.8 1.3 Table 3.2. Data are the mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Signi ficance was determined using a twotailed Student’s t-test comparing ST3Gal-IV(-/-) Ito and IK,slow activation parameters to control. = significance ( p <0.005).

PAGE 85

68 gated (HCN; pacemaker) channels, by modul ating the response of the autonomic nervous system, and/or by some other unknown method. Therefore, these data only suggest that the absence of ST3G al-IV can modify cardiac conduction, without detailing the mechanism. To determine whether cellular excit ability was modulated by ST3Gal-IV expression, atrial and ventricular APs we re measured for control and knockout myocytes. Action potential durations were lengthened significantly in the ST3GalIV(-/-) neonatal atria compared to control, and yet were unaffected in the ventricle (Figures 3.2 and 3.3). Consistently, our lab reported that action potential parameters were altered in atrial myocyt es, but not ventricular myocytes, lacking the sialyltransferase STX. The previous study also found a slower time to peak and a reduced peak AP voltage in the STX(-/-) atrial myocytes189, only one of which was observed in ST3Gal-IV(-/-) atrial APs (lengthened time to peak). As with all transgenic mouse systems, the absence of one enzyme may trigger the up-regulation of another with different substr ate specificity, leading to variability in glycosylation patterns. In vivo the combination of lengthened atrial APD (likely due to changes in potassium channel function) and a possible effect on HCN channels and/or the autonomic nervous system may lead to increased frequency or severity of irregularities in cardia c conduction and rhythm. Thus, alterations in glycosylation likely impact more proce sses than atrial AP production, as discussed further in the Final Discussion.

PAGE 86

69 Half-activation voltages of the repolarizing Kv currents, Ito and IK,slow, were shifted to more positive potentials in the knockout atria as well; no significant effect of the loss of ST3Gal-IV was observed on ventricular Kv currents (Figures 3.4 and 3.5). These data are consistent wit h our ECG and AP findings; the lengthened atrial APD and time to peak corresponds to the broadened P wave on the ECG, which represents atrial depolarization. Additionally, the data presented here are consistent with previously published reports from our lab. Rat neonatal ventricular sodium channels were s hown to be less sialylated than those expressed in neonatal atria218. Recently submitted data from our lab suggest Nav currents are shifted to more positive potentials in STX(-/-) atrial myocytes189, which correlates with the slowed time to peak observed from STX(-/-) and ST3Gal-IV(-/-) atrial APs. Aberrant sialylation, through the loss of a single sialyltransferase (in this case ST3Gal-IV), different ially impacts neonatal cardiac Kv currents and APs, increasing the susceptibility to altered cardiac conduction.

PAGE 87

70 CHAPTER 4 REMOVAL OF GLYCOSYLATION PROLONGS AP DURATION AND MODULATES IK Kv channel isoforms, like many i on channels, are uniquely and heavily glycosylated proteins. Nand O-glycan st ructures typically are terminated by numerous negatively charged sialic acid re sidues. A previous report showed that adult murine ventricular Kv channels responsible for Ito, likely Kv4 channels, require a greater depolarization (~ 11-18 mV) to undergo voltage-dependent gating events upon treatment to remove sialic acid; ventricular APD was prolonged as well217. Furthermore, the data presented in chapter three show that atrial APD and IK were altered by reduced sialylat ion, produced by a deficiency in the sialyltransferase ST3Gal-IV. In addition to Kv channels, gating of cardiac Nav channels is impacted by changes in glycosylation57,218. N-glycosylation has also been shown to impact the gati ng of voltage-gated ion channels214,215. In this chapter, we questioned whether en zymatic removal of specific glycans, namely sialic acids and N-glycans, could influence neonatal cardiac Kv currents, thus modulating AP repolarization. This will allow for confirmation of the ST3GalIV(-/-) cardiomyocyte data presented in Chapt er 3 and determination of whether Nglycans modulate cardiac K+ currents and APs. Glycosidases to remove sialic

PAGE 88

71 acid residues and N-glycosylation structur es were utilized (Figure 4.1). Atrial APD was prolonged significantly under glyc osidase treated conditions compared to the untreated control, while no effect on ventricular AP duration was observed. Consistent with that seen for the APs, data indicate that a reduction in sialylation and N-glycosylation cause a large, depol arizing shift in the voltage dependence of activation for Ito and IK,slow in atrial myocytes; however, ventricular Kv currents were not affected. N-glycosylation did not exert a supplementary effect on atrial Kv currents. Thus, sialylation, but not N-glycosylation, modulates production of atrial APs and K+ currents, likely increasing susceptibility to altered cardiac rhythm. Results Action potentials are prolonged in ne onatal atrial myocytes following removal of glycosylation Changes in AP waveform and duration can lead to alterations in cardiac excitability; therefore, we investigated whether c hanges in glycosylation (via enzymatic removal of glycans) can impac t the cardiac action potential. Thus, whole cell action potentials were recor ded from neonatal atrial myocytes under conditions of full and reduced glycosylati on. Atrial myocytes incubated with neuraminidase prior to recordings, to cleave sialic acid residues, showed prolonged repolarization (F igure 4.1). Figure 4.2 summarizes the mean action potential duration at 10%, 50%, and 90% repolarization as recorded from untreated control and neuramin idase treated neonatal atrial myocytes. Measured

PAGE 89

72 Figure 4.1. Enzymatic treatment to remove glycosylation Neuraminidase PNGase-F Neuraminidase 3,63,6 Figure 4.1. Two enzymes were utilized to remove specific glycosylation residues. Sialic acids are removed by treatment with the glycosidase neuraminidase. PNGase-F cleaves N-glycosylation structur es at the asparagine residue. Figure modified from Varki, et al 2009244.

PAGE 90

73 Figure 4.2. Glycosidase treatment prolongs atrial action poten tial duration. APD50 (ms) 0 5 10 15 20 APD90 (ms) 0 15 30 45 60 Control NA Neur NA PNGase-F NA APD10 (ms) 0 2 4 6 Control NA Neur NA PNGase-F NA Control NA Neur NA PNGase-F NABCD * * * Figure 4.2. A Representative traces of ac tion potentials measured from untreated control (blue), Neuraminidase tr eated to remove sialic acid residues (red), and PNGase-F treated to cleave N-gl ycosylation (pink) atrial myocytes, B C & D Action potentials were recorded from untreated control atrial myocytes (blue) and following treatment with Neuraminidase (red) or PNGase-F (pink). Bar graphs represent action pot ential duration at 10% ( B ), 50% ( C ), and 90% ( D ) repolarization (Table 4.1). Data are the mean SEM. n =9-10. *= significant compared to control ( p <0.005)

PAGE 91

74 APDs in the neuraminidase treated myocytes were significantly longer (by ~35150%) compared to the untreated control (T able 4.1). Additionally, the peak AP voltage (maximal depolarizat ion) and the rate of rise were prolonged in neuraminidase treat ed myocytes. PNGase-F, an enzyme that removes surfac e N-glycosylation at the asparagine residue, was employed to determine whet her N-glycosylation exerts an additional effect on neonatal atrial action potential duration (Figure 4.1). APs recorded from PNGase-F treated atrial myocytes dem onstrated prolonged durations. PNGase-F treated atrial APD10, APD50, and APD90 values were ~25-85% greater than untreated control (Figure 4. 2, Table 4.1). However, no significant effect of PNGase-F treatment was observed on peak AP voltage or the time to peak. Further, PNGase-F treated atrial AP durati ons were not significantly different than those of the neuraminidase treated myocytes (Table 4.1). Ventricular action potentials are unaff ected by changes in glycosylation Action potentials recorded from neonatal ventricular myocytes also were examined under conditions of reduced glycosylation. Again, ventricular myocytes were treated with either neuraminidase or PNGase-F to remove sialylation or Nglycosylation, respectively. APDs were not significantly impacted by preincubation with either neur aminidase or PNGase-F. The peak AP voltage and rate of rise were unaffected as well, suggesting that glycosylation is not a

PAGE 92

75 Figure 4.3. Ventricular action potentia ls are not affected by removal of glycosylation. BCD A APD10 (ms) 0 2 4 6 8 APD50 (ms) 0 5 10 15 20 25 30 APD90 (ms) 0 10 20 30 40 50 60 70 Control NV Neur NV PNGase-F NV Control NV Neur NV PNGase-F NV Control NV Neur NV PNGase-F NV Figure 4.3. A Representative action potential tr aces from untreated control (green), Neuraminidase treated to remo ve sialic acid residues (red), and PNGase-F treated to cleave N-glycosylation ( purple) ventricular my ocytes, B C & D Action potential durations measured from untreated control ventricular myocytes and following treatment wit h Neuraminidase or PNGase-F. No significant impact on ventricular APD was observed upon treatment with glycosidases ( p >0.1). Bar graphs represent ac tion potential duration at 10% ( B ), 50% ( C ), and 90% ( D ) repolarization (Table 4.1). Data are the mean SEM. n =8-13.

PAGE 93

76 Table 4.1. The measured action potential parameters for control and glycosidase treated atria and ventricles. Cell Type n AP Peak V Time to Peak APD10 APD50 APD90 mV ms ms ms ms Untreated Control Atria 9 120.1 4.8 3.9 0.1 4.4 0.2 9.0 0.8 22.3 2.6 Neuraminidase Treated Atria 10 139.8 4.9* 4.5 0.2* 5.9 0.3* 17.4 2.0* 55.1 9.3* PNGase-F Treated Atria 9 124.0 11.9 4.5 0.4 5.5 0. 4* 13.8 0.9* 41.3 3.0* Untreated Control Ventricle 13 138.3 4.3 3.9 0.1 5.7 0.3 21.7 2.5 56.7 6.2 Neuraminidase Treated Ventricle 9 146.4 2.3 4.3 0.2 6.3 0.4 22.6 2.5 55.0 3.5 PNGase-F Treated Ventricle 8 136.9 6.4 4.0 0.2 6.3 0.9 22.1 2.2 55.6 2.9 Table 4.1. Data are the mean SE M. APD: Action potential duration. Significance was determined using a twotailed Student’s ttest comparing Neuraminidase or PNGase-F treated ac tion potential paramet ers to untreated control. = significance ( p <0.02).

PAGE 94

77 significant factor in production of neonat al ventricular action potentials (Figure 4.3, Table 4.1). Atrial Ito and IK,slow activation is shifted to more depolarized potentials with removal of glycosylation As previously discussed, voltage-gated pot assium currents are responsible for AP repolarization. Since atrial AP repol arization was prolonged by removal of glycosylation structures, two Kv currents involved in AP repolarization were examined under conditions of full and reduced glycosylation. IK,slow and Ito were separated experimentally, as described in Chapter 3. Figure 4.4 shows IK,slow under conditions of reduced sialylation ( neuraminidase treated) activated at more depolarized potentials than when fully sialylated. The conductance-voltage relationship for IK,slow recorded from neuraminidase treated myocytes was shifted to more depolarized potentials from t hat observed for the untreated control. Likewise, the Va was shifted by a depolarizing ~ 17 mV for neuraminidase treated myocytes, indicating a stronger depolar ization is required to activate IK,slow in the absence of sialic acids (Figure 4.4, Tabl e 4.2). Upon treatment with PNGase-F to remove surface N-glycosylation, the G-V relationship and the Va for IK,slow were shifted to more depolarized potentials by ~12 mV compared to untreated control. These data suggest that N-glycosylation structures do not exert an additional effect on atrial IK,slow activation compared to the effect with neuraminidase, consistent with that report ed for the action potentials (F igure 4.4, Table 4.2).

PAGE 95

78 Figure 4.4. Deglycosylation shifts steady-state activa tion of atrial Ito and IK,slow to more depolarized potentials. Membrane Potential (mV) -80-60-40-20020Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Ito Control Ito + Neuraminidase Ito + PNGase-F Membrane Potential (mV) -80-60-40-20020Normalized Conductance 0.00 0.25 0.50 0.75 1.00 IK,slow Control IK,slow + Neuraminidase IK,slow + PNGase-F B A Figure 4.4. Steady-state activation of Ito and IK,slow as measured from control (dark blue), neuraminidase treated (red), and PNGase-F treated (purple) atrial myocytes. Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). n = 7 – 4 (Table 4.2), A G-V relationships for steady-state activation of Ito (circles). Note the ~9-14 mV shift to more depolarized potentials in the glycosidase tr eated atrial myocytes compared to the controls, B G-V relationships for IK,slow activation (squares). A ~12-17 mV depolarizing shift in steady-state acti vation was observed in the glycosidase treated atrial myocytes compared to the controls.

PAGE 96

79 Figure 4.4 also shows the plotted G-V relationships for Ito as recorded from atrial myocytes neuraminidase or PNGase -F treatment. The G-V curve for Ito recorded from neuraminidase treated atri al myocytes was shifted to more depolarized potentials by a signifi cant ~14 mV; additionally, the Va was shifted in the depolarized direction to a similar extent (Figure 4.4, Table 4.2). Treatment of atrial myocytes with PNGase-F revealed a depolarizing shift of ~9mV in the Va compared to that of untreated control, with the G-V curve shifted accordingly (Figure 4.4, Table 4.2). Ito measured from PNGase-F tr eated atrial myocytes was not different significantly from that r eported for neuraminidase treated myocytes. Again, this is consistent wit h the AP data pres ented earlier. Glycosylation does not im pact ventricular Ito or IK,slow activation Although there was no observable effect of glycosylation on ventricular AP duration and repolarization, Ito and IK,slow were measured in neonatal ventricular myocytes under conditions of full and reduced glycosylation to determine whether IK was altered. Similarly, a r eduction in glycosylation, through neuraminidase or PNGase-F treatment, had no significant effect on activation of ventricular Ito or IK,slow (Figure 4.5, Table 4.2). The G-V relationships and Va were not significantly different in the treated myocytes (with either neuraminidase or PNGase-F) compared to untreated controls. These data suggest that glycosylation does not play a functional role in neonatal ventricular potassium current generation.

PAGE 97

80 Figure 4.5. Steady-state act ivation of ventricular Ito and IK,slow is unaffected by treatment with glycosidases. Membrane Potential (mV) -60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Ito Control Ito + Neuraminidase Ito + PNGase-F Membrane Potential (mV) -60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 IK,slow Control IK,slow + Neuraminidase IK,slow + PNGase-F B A Figure 4.5. Ito and IK,slow steady-state activation recorded from control (green), neuraminidase treated (dark red), and PNGase-F treated (purple) ventricular myocytes. No significant shift in steady-state activation was noted in neuraminidase or PNGase-F treated my ocytes compared to controls ( p >0.1). Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). n = 4 – 3 (Table 4.2), A G-V relationships for Ito activation (circles), B G-V relationships for st eady-state activation of IK,slow (squares).

PAGE 98

81 Table 4.2. Measured gating paramete rs of atrial and ventricular Ito and IK,slow under control and glycosidase treated conditions. Current n Va Ka mV mV Control Atrial Ito 7 -34.5 1.7 6.7 0.8 Neuraminidase Treated Atrial Ito 4 -20.5 0.7* 7.4 0.7 PNGase-F Treated Atrial Ito 4 -25.4 2.3* 9.7 1.4 Control Atrial IK,slow 7 -40.4 1.3 8.3 0.6 Neuraminidase Treated Atrial IK,slow 4 -23.8 1.4* 10.3 0.2* PNGase-F Treated Atrial IK,slow 4 -28.8 1.0* 10.6 0.4* Control Ventricular Ito 4 -8.9 1.6 13.4 1.0 Neuraminidase Treated Ventricular Ito 4 -10.2 1.3 13.6 1.3 PNGase-F Treated Ventricular Ito 3 -10.5 1.7 11.9 0.8 Control Ventricular IK,slow 4 -11.3 1.3 12.7 0.5 Neuraminidase Treated Ventricular IK,slow 4 -14.1 2.1 13.5 0.4 PNGase-F Treated Ventricular IK,slow 3 -10.7 2.6 12.1 0.6 Table 4.2. Data are the mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Signi ficance was determined using a twotailed Student’s t-test comparing glycosidase (Neuraminidase or PNGase-F) treated Ito and IK,slow activation parameters to untreat ed control. = significance ( p <0.005).

PAGE 99

82 Discussion Atrial action potentials are modulated by alterations in glycosylation, while ventricular action poten tials are unaffected Alterations in action potential duration potentially can impact the rhythmicity of the heart; lengthening or shortening of APD has been identified as proarrhythmic245. Therefore, this chapter exam ined the effects of enzymatic glycosylation removal on neonatal cardia c action potentials. As Figure 4.2 suggests, treatment of atrial myocyt es with neuraminidase or PNGase-F significantly and similarly lengthened acti on potential duration at 10%, 50%, and 90% repolarization. Interestingly, peak AP voltage and time to peak were significantly prolonged in neuraminidas e treated atrial myocytes, but were unaffected with PNGase-F treatment (Table 4. 1). As shown in a previous study, atrial Nav currents were impacted by loss of the polysialyltransferase ST8Sia2 (STX)189. Thus, changes in the peak AP voltage and time to peak could be caused by alterations in Nav channel gating upon removal of O-linked sialic acids, as these would not be re moved upon treatment with PNGase-F; incomplete removal of N-glycan structur es may also be responsible for these results (see Final Discussion). In addi tion, ventricular APD was not impacted significantly by removal of sialylation or N-glycan structures; however, a previous report by Ufret-Vincenty, et al found that neuraminidase treatment lengthened adult murine ventricular APs217. The inconsistent data may be the result of increased complex glycan structures in the adult ventricle compared to the neonatal ventricle, as suppor ted by data from our lab189,218. Therefore, the results

PAGE 100

83 presented here and in chapter 3 suggest that normal glycosyl ation is an essential component in neonatal atrial, but not vent ricular, AP production (Figures 4.2 and 4.3, Table 4.1). Glycosylation affects cardiac Ito and IK,slow activation in a chamber-specific mannerHere, atrial action potential repolariz ation was shown to be modulated by removal of glycosylation, with all or most of the effect due to sialylation. Since K+ channels are responsible for the AP repol arization phase, it was predicted that atrial K+ currents were affected as well. In fact, we showed voltage-dependent depolarizing shifts of ~14-17 mV in steady-state activation of atrial Ito and IK,slow with neuraminidase treatment to remove su rface sialic acids (Figure 4.4, Table 4.2). Removal of N-glycans via PNGa se-F treatment shifted atrial Ito and IK,slow activation to more depolarized potentia ls, comparable to those observed with neuraminidase treatment (~ 9-12 mV). Further, we can conclude that Nglycosylation does not exert an addi tional effect on atrial Kv channel gating, as evidenced by the production of similar Kv currents and APs upon treatment with two distinct glycosidases (neuraminidas e and PNGase-F). As discussed further in chapters five and six, only one Kv channel isoform responsible for Ito production is N-glycosylated, Kv1.4; Kv4.2 and Kv4.3 do not contain any extracellular N-glycosylation sites. Thus, N-linked sialylation of Kv1.4 may be contributing all or most of t he effect of glycosylation on Ito activation. However, Olinked sialic acids attached to Kv1.4, Kv4.2, or Kv4.3 might be responsible for the

PAGE 101

84 slightly greater effect observed on Ito measured from neuraminidase treated atrial myocytes. The same phenomenon may be occurring for IK,slow as well; O-linked sialylation could be contributing an effect on Kv1.5 and Kv2.1 activation. Ventricular Kv currents displayed no significant effect of changing glycosylation levels (Figure 4.5, Table 4.2). Neve rtheless, a shift to more depolarized potentials was observed for adult ventricular Ito activation with neuraminidase treatment in a previous study217. This inconsistency wa s not unexpected, as it has previously been shown that Nav channels are less sialylated in rat neonatal ventricular myocytes than neonatal at ria or adult atria and ventricle218. Our lab recently provided additional ev idence supporting this phenomenon (see above)189. As seen here, glycosylation im pacts the generation of neonatal cardiac Kv currents through a chamber-specific me chanism; a significant effect of changing glycosylation is observed in the neonatal atria, but not the ventricles. Further, the neonatal murine heart rate is slower than that of the adult. Physiologically, it may not be essential fo r neonatal ventricular myocytes to be as sialylated as neonatal atrial or adult atrial and ventricular myocytes. As previously discussed, sialic acids have been shown to enhance ion channel activation so, greater sialylation of adult ventricular myo cytes would result in faster conduction through the ventricles. In certain glycosyl ation-related pathologies, the differing effects of glycosylation throughout the heart could modulate normal conduction and rhythmicity, leading to potentia lly life-threatening arrhythmias.

PAGE 102

85 Thus, we showed that changes in glycosyl ation, mainly sialylation, modulate cardiac K+ channel function and have chamber-specif ic effects. However, to fully appreciate the impact of glycosylation on cardiac excitability, we must understand how gating of thos e channels involved in IK production, and thus AP repolarization, is affected by changes in glycosylation. As previously discussed, several Kv channel isoforms generate cardiac Ito (Kv1.4, Kv4.2, and Kv4.3) and IK,slow (Kv1.5 and Kv2.1). The following two chapters detail the results of altered glycosylation on Kv1.4 and Kv1.5 (Chapter 5) and Kv2.1, Kv4.2, and Kv4.3 (Chapter 6) gating. Discussion of the effect of glycosylation/sialylation on gating of Kv channel isoforms was divided in to two chapters based on the channel subfamily and the type of glycosylation attached to each channel (Nor Oglycosylation).

PAGE 103

86 CHAPTER 5 UNIQUE MODULATION OF Kv1.4 AND Kv1.5 GATING BY N-GLYCANS The voltage-gated potassium channel family is comprised of a large and diverse set of ion channels that serve a va riety of functions throughout the body246,247. Various sets of Kv channel isoforms are responsib le for the repolarization phase of skeletal muscle, neuronal, and cardiomyocyte action potentials75.Slight changes in the type, relative density, or activity of Kv channel isoforms that occur following a remodeling process can lead to altered action potential repolarization. Such remodeling is observed through physiological processes such as development and aging, as well as through pathologies such as Long QT Syndrome, deafness, and epilepsy58,59,61,64. Posttranslational modifications make up a significant portion of the mass of ion channels, with glycosylation contributing up wards of 30% of a channel’s total mass157. Analyses of samples purified from br ain suggested that over 100 sialic acid residues per functional molecule are attached to Na+ channel160,164 and some Shaker K+ channel proteins154. Previous studies showed that N-glycans can modulate Nav and Kv channel activity through isoform-specific mechanisms214,220,248. All reported data suggest that si alic acids contribute to the full effect of glycans on Nav channel gating through electrostatic

PAGE 104

87 mechanisms209,220-222. On the other hand, previous st udies indicated that gating of Shaker Kv channel isoforms, Kv1.1163,214, Kv1.2249, and the Drosophila ShB channel216, were modulated by N-glycans and/or sialic acids through multiple mechanisms that include electrostatic me chanisms (surface potential theory) and glycan-dependent effects on the stability of channel proteins among functional states (gating stabilizi ng theory). Here, we questioned whether N-glycans and sialic acids alter the gating of two homologous Shaker K+ channel isoforms responsib le in part for cardiac IK, Kv1.4 and Kv1.5. These two isoforms share a great deal of sequence homology, including a similar location for the sole N-glycosylation site positioned on the S1S2 linkers (Figure 5.1). The data indicate that a reduction in sialylation caused a large, depolarizing shift in the voltage-dependence of activation for Kv1.5, while Kv1.4 gating was unaffected. Unlike previous studies of Shaker Kv channel isoforms214,216,250, N-glycosylation did not exert an additional effect on the gating of either isoform. Further, sialic acids modulated Kv1.5 gating solely through electrostatic mechanisms. Results Cell lines of Chinese Hamster Ovary cells were utilized to test whether and how differential glycosylation impacts gating of two homologous Kv1 isoforms; Pro5 cells are fully sialylating, while Lec2 ce lls are essentially non-sialylating due to a deficiency in the CMP-sialic acid transporter226. This deficit serves as a model for

PAGE 105

88 Figure 5.1. Kv1.4 and Kv1.5 contain one putative N-glycosylation site. NCNXS/TS1 S2hKv1.4 ETLPEFR DDRDLVMALSAGGHG G LL NDT S A P HLENSGHTIFN DPF hKv1.5 ETLPEFR DERE LLRHPPAPHQP PAPAPGA NGS G VMAPP S G P TVAPLLPRTLA DPF Figure 5.1.Schematic of a Kv1 channel -subunit. The amino acid sequences of the S1-S2 linker for human Kv1.4 and Kv1.5 are shown above. The Nglycosylation consensus sequences are underlined in red and conserved sequences are shown in black.

PAGE 106

89 one form of Congenital Disorders of Glycosylation, CDG type-IIf,192 and was used by several investigators, including our laboratory, to question how sialic acids modulate ion channel gating163,214,216,217,220,221. We predicted that Kv1.4 and Kv1.5 gating would be similarly impacted by t he N-glycans attached to the homologous positions along the channel structure. Ho wever, the data indicate unique and different effects of N-glycans on Kv1.4 and Kv1.5 gating. N-linked sialic acids account for th e full effect of glycans on Kv1.5 gating N-linked glycans impact Kv1.5 gating Kv1.5 is partially responsible for the pr oduction of a slowly inactivating current (IK,slow) involved in action potential repol arization in cardiomyocytes and neurons251-253. The data presented in chapters three and four suggest that changes in glycosylation, most notably sialylation, shift the voltage-dependence of atrial IK,slow activation to more depolarized potentials, effectively lengthening AP repolarization and duration. To dete rmine whether N-glycosylation impacts Kv1.5 gating, a Kv1.5 N-glycosylation mutant (Kv1.5N290Q) was generated by mutating the asparagine residue (N) that initiates the potential N-glycosylation consensus sequence to a glutamine (Q). Kv1.5N290Q expressed currents similar to that produced by the wild-type channel However, the conductance-voltage relationship for Kv1.5N290Q was shifted along the voltage axis by a depolarizing ~18 mV compared to the wild-type Kv1.5 G-V relationship, suggesting that Nglycosylation modulates Kv1.5 channel activation (Figure 5.2B, Table 5.1). To confirm that lack of N-glycosylation caused by the N to Q mutation was

PAGE 107

90 responsible fully for the observed effect on channel gating, the serine (S) residue of the N-glycosylation site was mutated to an alanine (A) in order to generate a second N-glycosylation mutant Kv1.5 channel (Kv1.5S292A). The G-V relationship and the half-activation voltage for the Kv1.5S292A channel were not significantly different from that measured for the Kv1.5N290Qchannel (Figure 5.2B, Table 5.1). These data indicate that mutation of the asparagine residue or the serine residue that comprise the N-glycosylation c onsensus sequence impacted channel gating identically, allowing one to infer that the mutation itse lf did not significantly influence Kv1.5 gating, but rather the lack of N-glycosylation was responsible for modulated Kv1.5 gating. The data indicate that N-glycosylation promotes activation of Kv1.5 by causing a significant sh ift in the voltage at which the channel activates. N-linked sialic acids account fo r the full effect of sugars on Kv1.5 gating To determine whether Kv1.5 sialic acids contribute to the effect of N-glycans on channel gating, Kv1.5 activation was studied under conditions of full and reduced sialylation by expressing the channel in Pro5 (full sialylation) and Lec2 (reduced sialylation) cell lines. As shown in Figure 5.2, Kv1.5 under conditions of reduced sialylation activated at more depolariz ed potentials than the fully sialylated channel. The G-V relationship for the less sialylated Kv1.5 was shifted by a depolarizing ~18 mV from that observed fo r the fully sialylated channel. Likewise, the Va was shifted to more depolarized poten tials for the less sialylated channel, indicating a stronger depolarizatio n is required to activate Kv1.5 in the absence of

PAGE 108

91 Figure 5.2. N-linked sialic acids account for the full effect of glycans on Kv1.5 activation. Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv1.5 + SA Kv1.5 SA Kv1.5N290Q + SA Kv1.5N290Q SA Kv1.5S292A + SA Membrane Potential (mV) -2002040n (ms) 0 5 10 15 20 25 30 35 A B C 1 nA50 ms2 nA50 ms

PAGE 109

92 Figure 5.2. A Typical Kv1.5 + SA (Pro5) whole cell current trac es. SA: Sialic acid. B & C Voltage-dependence of Kv1.5 SA, Kv1.5N290Q SA and Kv1.5S292A + SA activation, B Conductance-voltage (G-V) rela tionships. Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Black Circles: Kv1.5 + SA; Gray Squares: Kv1.5 – SA; Black Upright Triangles: Kv1.5N290Q + SA; Gray Inverted Triangles: Kv1.5N290Q – SA; Black Diamonds: Kv1.5S292A + SA. n = 3-11. (Table 5.1), C Activation time constants ( n). Data are mean SEM. Lines ar e non-theoretical point to point. Symbols are as listed in panel B Inset: Typical current traces measured during a 0 mV test potential. Peak currents we re normalized for comparison. The scaling factors used to normalize the current s were 2.63 to 23.75. Solid Black: Kv1.5 + SA; Solid Gray: Kv1.5 – SA; Dotted Black: Kv1.5N290Q + SA; Dotted Gray: Kv-1.5N290Q – SA; Dashed Black: Kv1.5S292A + SA. n = 3-11.

PAGE 110

93 sialic acids (Figure 5.2B, Table 5.1). As described, the gl ycosylation-deficient mutant channels, Kv1.5N290Q and Kv1.5S292A, activated at potentials ~18 mV more depolarized than fully sialylated Kv1.5, with the corresponding G-V relationships nearly identical to Kv1.5 under conditions of reduced si alylation. Further, the G-V curve for Kv1.5N290Q was not altered when studied under conditions of reduced sialylation and was nearly identical to the G-V curves measured for Kv1.5 under conditions of reduced sialylation (Figure 5. 2B).Together, these data indicate that N-linked sialic acids fully account for the effect of N-glycans on Kv1.5 gating – a novel finding among Shaker family Kv channel isoforms. Sialic acids similarly modulated Kv1.5 activation rate, with channels activating more slowly under conditions of reduced sial ylation (Figure 5.2C). Note the time constants of activation were shifted al ong the voltage axis by approximately 18 mV, consistent with the magnitude of the shift in Va observed (Figure 5.2). The activation rates for the Kv1.5N290Q and Kv1.5S292A mutant channels when studied under conditions of full or reduced sialyl ation were nearly identical and did not significantly deviate from the activation rates for Kv1.5 under conditions of reduced sialylation (Figure 5.2C). Thus, t he data indicate that sialic acids impact Kv1.5 gating, causing the channel to activa te at more hyperpolarized potentials; this will effectively increase the activity of Kv1.5 at a given memb rane potential.

PAGE 111

94 Table 5.1. Gating parameters measured for Kv1.5. Construct n Va Ka mV mV Kv1.5 + SA 11 -27.1 3.1 9.3 0.7 Kv1.5 – SA 11 -8.9 1.8* 10.4 0.3 # Kv1.5 N290Q + SA 5 -7.9 3.0* 8.3 0.3 # Kv1.5 N290Q – SA 4 -8.9 2.6* 9.5 0.5 # Kv1.5 S292A + SA 4 -7.1 1.5* 9.8 0.4 # Table 5.1. Data are mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Significance was te sted using a two-tailed Student’s t-test to compare gating parameters as ex pressed under conditions of reduced glycosylation with control conditions.*= significance ( p <0.0005). # = not significantly different from control ( p >0.1).

PAGE 112

95 Figure 5.3. Kv1.5 is N-glycosylated. Figure 5.3. Immunoblot of Kv1.5 and Kv1.5N290Q. Samples were fractionated on a 6.5% SDS gel and pr obed with an anti-Kv1.5 polyclonal antibod y directed against a C-terminal sequence (1:50 dilution, s ee Materials and Methods). Lane 1: Molecular weight markers, Lane 2: wild-type Kv1.5, Lane 3: Kv1.5N290Q. Protein loaded per lane: wild-type Kv1.5, 55 g; Kv1.5N290Q, 50 g.

PAGE 113

96 Kv1.5 is N-glycosylated Immunoblot analysis verified that Kv1.5 expressed in CHO cells is N-glycosylated (Figure 5.3). Note a band of lower pr edicted molecular weight (MW) for the mutant Kv1.5N290Q channels compared to the band observed for wild-type Kv1.5. Kv1.5 sialic acids apparently modulate channel gating by contributing to the external negative surface potential Negatively charged sialic acids are typi cally the terminal residue attached to external glycans254. These negative charges may c ontribute to the external surface potential that impacts channel gati ng, as outlined in the surface potential theory212. If so, the fully sialylated channel should activate at more hyperpolarized potentials than the less si alylated channel, as observed for Kv1.5. Further, the presence of external diva lent cations would screen the surface potential and reduce the impac t of the negatively charged sialic acid residues on the voltage-dependence of channel gating212,213. Here, the G-V re lationships for Kv1.5 expressed in Pro5 and Lec2 cells were studied at two different external divalent cation concentrations (Figur e 5.4A and C). The data show that Kv1.5 under fully sialylating conditions (in Pr o5 cells) was sensitive to changes in extracellular divalent cati ons; the G-V relationship at the lower [divalent cationo] shifted by a hyperpolarizing ~13 mV (Fi gure 5.4A). Changes in [divalent cationo] had no significant effect on Kv1.5 gating when expresse d in the non-sialylating (Lec2) cell line (Figure 5.4C). Extracellu lar divalent cations similarly modulated Kv1.5 activation rate, with fully sialylated channels activating more rapidly under

PAGE 114

97 Figure 5.4. Sialic acids modulate Kv1.5 gating through apparent electrostatic mechanisms. Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv15 + SA Kv15 + SA Low [CationO 2+] Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 000 025 050 0.75 100 Kv1.5 SA Kv1.5 SA Low [CationO 2+] Membrane Potential (mV) -40-2002040n (ms) 0 2 4 6 8 10 Membrane Potential (mV) -40-2002040n(ms) 0 2 4 6 8 10 5 nA20 ms 1 nA50 ms Va~12 mV Va~1 mVA C B D Figure 5.4. Kv1.5 expressed under conditions of full and reduced sialylation at 3 mM (normal, filled symbols) and 0.3 mM (low, open symbols) external divalent cation concentrations (see Materials and Methods). Black Circles: Kv1.5 + SA; Gray Squares: Kv1.5 SA. n = 3-4, A & C. Conductance-voltage (G-V) relationships of Kv1.5 expressed in fully sialylating (Pro5, A ) and reduced

PAGE 115

98 sialylating (Lec2, C ) cells at normal and low external divalent cation concentrations. Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Va: Change in the voltage of halfactivation under each condition, B & D Activation time constants ( n) for Kv1.5 in the presence (Pro5, B ) and absence (Lec2, D ) of sialylation in the presence of normal and low external divalent cati on concentrations. Data are mean SEM. Lines are non-theoretical point to point. Insets: Current trac es measured during a -20 mV test pulse. Peak currents were normalized for comparison. The scaling factors used to normalize the currents we re 1.02 to 1.13. Solid: normal [divalent cationo]; Dotted: low [divalent cationo].

PAGE 116

99 conditions of reduced extracellular [divalent cationo] (Figure 5.4B). No significant effect of changing external divalent cation concentration was observed for Kv1.5 activation when expressed in the non-sial ylating cell line (Figure 5.4D). These data suggest that N-linked Kv1.5 sialic acids contribute to the external negative surface potential and thereby increase Kv1.5 activity through electrostatic mechanisms. Kv1.4 gating is not modulated by N-glycosylation or sialylation Sialic acids do not affect Kv1.4 gating Kv1.4 is an A-type Kv channel involved in action pot ential repolarization (in the heart, it is responsible for production of Ito,s) and is expressed in many regions of the brain and heart255-259. Figure 5.5A shows the normalized, average G-V relationships for Kv1.4 expressed in Pro5 and Lec2 cells. Note that the G-V relationships and the Va for Kv1.4 under conditions of full and reduced sialylation were nearly identical (Figure 5.5A, Table 5.2). The activation data were fit to a single Boltzmann relation in Figure 5.5A, but the fit is not ideal, indicating multiple gating states. A double Boltzmann relation wa s utilized as well; however, the fit nearly was identical to that of the single Boltzmann. Further, there was no significant difference in the voltage depend ence of steady-state inactivation for Kv1.4 as expressed in Pro5 and Lec2 cells (F igure 5.5B). Therefore, sialylation (in CHO cells) does not alter Kv1.4 channel gating.

PAGE 117

100 Figure 5.5. Kv1.4 gating is not impacted by si alylation or N-glycosylation. Membrane Potential (mV) -100-80-60-40-20020Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv1.4 + SA Kv1.4 SA Kv1.4N354Q + SA B Pre-Pulse Potential (mV) -100-80-60-40-200Normalized Current 0.00 0.25 0.50 0.75 1.00 C 20 ms50 nAA 2 nA20 ms

PAGE 118

101 Figure 5.5. A Typical Kv1.4 whole cell current traces as expressed in Pro5 cells. B & C Voltage-dependence of Kv1.4 SA and Kv1.4N354Q + SA. B G-V relationships. Data are the mean no rmalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Black Circles: Kv1.4 + SA (Pro5); Gray Squares: Kv1.4 – SA (Lec2); Black Upright Triangles: Kv1.4N354Q + SA. n = 3-7. C Steady-state channel availability (i nactivation). Data are the mean normalized peak current SEM measured during a 100 ms pulse to +60 mV, following a 1000 ms pre-pulse to the pl otted potentials. Curves are single Boltzmann distribution fits to the data. Symbols as listed in panel B Inset: Kv1.4 inactivation (hinf) current traces. n = 3-7. (Table 5.2).

PAGE 119

102 Table 5.2. The measured gating parameters for Kv1.4. Construct n Va Ka Vi Ki mV mV mV mV Kv1.4 + SA 6 -29.3 2.8 14.6 0.8 -53.6 1.1 -4.0 0.5 Kv1.4 – SA 7 -26.3 1.4 15.3 0.8 -52.5 1.6 -4.3 0.9 Kv1.4 N354Q + SA 3 -28.4 3.5 15.2 1. 3 -52.7 0.3 -4.7 0.1 Table 5.2. Data are mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Vi: Voltage of half-inactivation. Ki: Boltzmann inactivation slope factor. Significance was tested us ing a two-tailed Student’s t-test to compare gating parameter s as expressed under c onditions of reduced glycosylation with control conditions. No parameter was significantly different from control ( p >0.05).

PAGE 120

103 N-glycosylation does not alter gating of Kv1.4 Like most Kv1 channels, Kv1.4 contains an N-glycosylation consensus sequence on its S1-S2 linker260. To determine whether N-glyc ans other than sialic acids modulate Kv1.4 gating, a mutant Kv1.4 (Kv1.4 N354Q) was generated by replacing the asparagine residue (N) of the Nglycosylation consensus site with a glutamine residue (Q). Kv1.4 N354Q produced currents similar to those of the wildtype channel. Gating of the wild-type Kv1.4 and the mutant Kv1.4 N354Q were nearly identical, with no significant differ ence observed for steady-state activation or inactivation (Figure 5.5, Table 5.2). It is of interest to note that the Kv1.4N354Q activation data did not fit well to eit her a single or double Boltzmann relation, similar to that of the wild-type channel. The N-glyco sylation mutant does not likely comprise a heterogenous population, eliminating the possibility that a differentially glycosylated collection of c hannels is responsible for the unusual fit of the data. Thus, neither N-glyca ns nor sialic acids modulate Kv1.4 gating. Kv1.4 is N-glycosylated and sialylated To determine whether Kv1.4 expressed in CHO cells is N-glycosylated and sialylated, immunoblot gel shift analysis un der varying conditions of glycosylation was performed (Figure 5.6). Note that only a single band of lower MW was detected for the mutant Kv1.4 N354Q lysate compared to the double banding pattern observed for the Kv1.4 Pro5 and Lec2 lysates. The lower band of the Kv1.4 lysate, presumably a lesser gl ycosylated (possibly, non-glycosylated) Kv1.4, is of similar MW to the non-glycosylated Kv1.4 N354Q mutant channel. The

PAGE 121

104 Figure 5.6. Kv1.4 is N-glycosylated and sialylated. Figure 5.6. Immunoblot of Kv1.4 and Kv1.4N354Q. Samples were fractionated on a 6.5% SDS gel and pr obed with an anti-Kv1.4 monoclonal antibody (1:1000 dilution, see methods). Lane 1: Mole cular weight markers, Lane 2: Kv1.4N354Q, Lane 3: Kv1.4 + SA (Pro5), Lane 4: Kv1.4 – SA (Lec2). Prot ein loaded per lane: Kv1.4N354Q, 15 g; Kv1.4 + SA, 20 g; Kv1.4 – SA, 2 g.

PAGE 122

105 Kv1.4 lysate as expressed in the non-sial ylating Lec2 cell line showed two bands, one that matched the lower MW of the wild type Kv1.4 and one that was of lower MW than the upper band obs erved for the Pro5 Kv1.4 lysate. Together, these data indicate that Kv1.4 expressed in CHO cells is Nglycosylated and sialylated. Discussion As indicated in Figure 5.1, Kv1.4 and Kv1.5 each contain an N-glycosylation consensus site along the S1-S2 linker261. Alignment of Kv1.4 and Kv1.5 S1-S2 linkers indicate that the Kv1.5 N-glycosylation site is located six amino acids Cterminal to the location of the Kv1.4 N-glycosylation site. Figures 5.3 and 5.6 confirm that Kv1.4 and Kv1.5 are N-glycosylated. He re, we show that Nglycosylation of two Kv1 subfamily members, Kv1.4 and Kv1.5, differently and uniquely (for Shaker family Kv channels) modulate channel function. N-linked sialic acids account for the full effect of glycosylation on Kv1.5 gating – a novel finding for modulat ion of Shaker potassium channel function Sialic acids have a large impact on Kv1.5 gating. The G-V relationship and the Va for Kv1.5 were shifted by an identical ~18 mV under conditions of reduced sialylation and with removal of N-glycans (Figure 5.2B, Table 5.1). This ~18 mV shift in voltage dependence of channel acti vation is the largest effect of Nglycans on a Kv1 isoform reported to date. The ti me constants of activation were shifted in the same manner as the G-V re lationships and were nearly identical for

PAGE 123

106 Kv1.5 under conditions of reduced sialyl ation and the non-glycosylated mutant channels, suggesting that sialic acids modulate Kv1.5 gating through electrostatic interactions (Figure 5.2C). This wa s confirmed by the charge screening experiment outlined in Figure 5.4. These data indicate that the full effect of glycans on Kv1.5 gating is imposed by N-linked sialic acids. The surface potential theory pr edicts that the electric field experienced by the voltage sensor(s) of ion channels is alte red by external negative charges that contribute to a surface potential. External divalent cations act to screen these negative charges, effectively neutralizing t heir impact on the voltage sensed by the channel gating mechanism. Thus, a gr eater depolarization is required to activate the channel, producing a positive shift in the Va. Assuming a homogenous electric field, changes in the external surface potential should impact all voltage-dependent gating parameters equally and shift the G-V curve without impacting the slope of the curve212, as observed here for Kv1.5 (Figures 5.4A and 5.4C). In addition to an impact on surface potential, previous reports suggested that glycosylation can alter state stability214,216. The gating stabilizing theory, summarized by Hille, et al75, predicts a change in G-V relationship slope for any change in state stability, as reported previously by Watanabe, et al214. Such an effect was not observed in this study. The G-V slope factor (Ka) measured for the two non-glycosy lated mutant channels and Kv1.5 expressed under conditions of reduced sialylation were not significantly di fferent than the Ka measured for Kv1.5 (Figure 5.2B). Further, the ac tivation time constants for the

PAGE 124

107 Kv1.5 N-glycosylation mutants also were not significantly different from that of Kv1.5 under conditions of reduced sialylation (Figure 5.2C). Therefore, the effect of sialic acids on Kv1.5 channel activation is achieved solely through electrostatic mechanisms (surface potential theory) The data do not support an additional mechanism by which sialic acids or other N-glycans contribute to Kv1.5 gating (gating stabilizing theory).This finding is unique among Shaker-like potassium channel isoforms. Previ ous studies of the Drosophila Shaker K+ channel, ShB, and the mammalian Kv1.1 and Kv1.2 channels reported addi tional effects of Nglycans and/or sialic acids on Kv1 channel function, consistent with a contribution of N-glycans to channel state stability163,214,216. The data shown here provide the first evidence that gating of a Shaker family potassium channel isoform, Kv1.5, is modulated by N-linked sialic acids ac ting through an electrostatic mechanism that is responsible for the full effect of glycans on Kv1.5 gating. N-glycosylation does not affect Kv1.4 voltage-dependent gating In this study, reduced Kv1.4 glycosylation, achieved through a reduction in channel sialylation or mutagenesis to remo ve the full N-glycan structure, had no effect on Kv1.4 gating (Figure 5.5). Interestingl y, a previous study by Johnson and Bennett examined the impact of sial ylation and N-glycosylation on a similar channel, ShB216. Kv1.4 and ShB are the only two Shaker family K+ channel isoforms that undergo N-type inactivation. Both are N-glycosylated, but ShB has two potential N-glycosylation sites loca ted within the S1-S2 linker, while Kv1.4 has only one site262,263. Unlike Kv1.4, a reduction in ShB si alylation resulted in

PAGE 125

108 depolarizing shifts in steady-state and kinet ic activation and inactivation of ShB. Further, N-linked sugars other than si alic acids were shown to exert an additional, state stabilizing, effect on channel gating 216.The distinct effects of Nglycans on Kv1.4 and ShB gating may be due to the number and location of putative N-glycosylation sites. That is, 1) The presence of two N-glycosylation sites on the ShB S1-S2 linker may vary the position of the glycans and glycan structures relative to the voltage sens or and/or 2) Two glycan structures per subunit could lead to a mature K+ channel that is potentially twice as sialylated. Both possible outcomes could lead to a greater effect of N-glycans on channel gating. Further, the ShB S1-S2 linker is si gnificantly shorter than are any of the mammalian Kv1 isoforms, and the ShB N-glycosylation sites are located near the middle of the linker. The Kv1.4 N-glycosylation site can be found closer to the S2 transmembrane segment (Figure 5.1). These slight alterations in linker length and position of the N-glycosylat ion site(s) may account for the differing effects of glycosylation on rapidly inactivating Shaker family channel gating. Additionally, the report published by Johnson and Bennett did not distinguish the individual effects of the two N-glycosylation sites, as one may contribute all or most of the effect. Summary Together, these data demonstrat e that members of the Kv1 channel subfamily are affected by glycosylation through dist inct mechanisms. The unique effects of

PAGE 126

109 glycans on Kv1 family isoforms are relevant to changes in AP repolarization that occur with regulated expression and glycosylation of each Kv1 isoform.

PAGE 127

110 CHAPTER 6 O-LINKED SIALIC ACIDS IMPACT Kv CHANNEL GATING Precise gating of Kv channels is a crucial component in producing proper neuronal, cardiac, and skeletal muscle ac tion potential waveform and duration. Ion channel activity can be modulated by varying types and levels of posttranslational modifications, including gl ycosylation. The number and location of glycosylation sites vary among channel ty pes and between isoforms within a channel subfamily, including the Kv channel family. Glycans are attached to channel proteins putatively thr ough Nand O-linked glycosylation168,175. Several studies, including chapter five, detailed t he isoform-specific effects of N-glycans on Kv 163,214-216,264 and Nav 209,220,221 channel function. Such studies were possible because an external consensus site for N-glycosylation has been identified. On the other hand, little is known about the role of O-glycosylation on voltagegated ion channel activity. The enzymes re gulating O-glycosylation are complex, and O-glycosylation does not have a recognized consensus sequence187. The most prevalent form of O-glycosylation is the mucin-type, in which an Nacetylgalactosamine is bound to hydroxyl groups of serine or threonine side chains175,184. Lectin binding studies, utilizing the lectin Helix pomatia agglutinin to bind GalNAc, can identify O-glycosylation st ructures, but require extensive prior

PAGE 128

111 digestion with numerous glycosidase s and are, therefore, inefficient265,266. However, Cu(I)-catalyzed cycloaddition (Click) chemistry was employed to identify the appearance and presence of O-glycosylation during zebrafish development 236. All previous efforts to study the impact of glycosylation on Kv channel function concentrated on determining effects of gener ic glycosylation or N-glycosylation, including chapter five of this report163,214-217,264. No previous studies questioned the role of O-glycans in Kv channel function. Kv4.2 and Kv4.3 are members of the Shal subfamily of Kv channels and are rapidly activating and inactivating (A-t ype) channels that produce cardiac Ito,f 75. Activation of atrial Ito was shifted to more depolar ized potentials with reduced glycosylation, as discussed in chapters three and four. The putative external sequences of Kv4.2 and Kv4.3 do not contain potentia l N-glycosylation sites. Therefore, neither isofo rm can be N-glycosylated. Kv2.1 is a delayed rectifier that contributes to the gener ation of cardiac IK,slow (see Chapters 3 and 4 for further details on the impact of glycosylation on IK,slow). Kv2.1 contains one Nglycosylation site, located on the S3-S4 linker. Based on recent structural models, this site may be inaccessible to the glycosylation mach inery of the cell; consistently, brain Kv2.1 was shown to not be N-glycosylated84,85,211.

PAGE 129

112 In this study, we investigated whether and how sialic acids modulate gating of Kv2.1, Kv4.2, and Kv4.3, and probed for the presenc e of O-glycosylation and sialylation. The data show that each isofo rm is uniquely modulated by sialic acid residues that are attached to the channel through O-linkages. Results Negatively charged sialic acids (at physiol ogical pH) are typically located as terminal residues on Nand O-glycans, and were shown to impact gating of various voltage-gated ion channels differentially163,209,214-217,220,221,264. However, all previous studies concentrated on the effect of generic or N-linked sugars on ion channel gating without ex amining the possible specif ic effects of O-glycans on channel function. Unlike N-glycosylat ion, there is no known consensus sequence for O-glycosylation; however, predictive software (OGPET v1.0) and the Kv channel crystal structure can be utilized to estimate the location and number of possible O-gl ycosylation sites. Several Kv isoforms, including Kv2.1, Kv4.2, and Kv4.3, are not likely Nglycosylated or do not contain N-glycosyl ation consensus sequences; however, potential O-glycosylation site(s) can be identified on and limited to the Kv4.3 S1S2 linker and the S5-S6 linker for Kv2.1, Kv4.2, and Kv4.3. These three isoforms were expressed in CHO cell lines diffe ring in their ability to sialylate (see Materials and Methods and Chapter 5 Results). Kv channel O-glycosylation and sialylation were confirmed using Click chem istry. The data indica te for the first

PAGE 130

113 Figure 6.1. Sialic acids modulate Kv4.2 and Kv4.3 activation. Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv4.2 + SA Kv4.2 SA Membrane Potential (mV) -80-60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv4.3 + SA Kv4.3 SA AB 1 nA50 ms2 nA20 ms Figure 6.1. Conductance-voltage (G-V) relationships for Kv4.2 and Kv4.3 under conditions of full (Pro5) and reduced sial ylation (Lec2). Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Insets: Typical whole cell Kv4.2 and Kv4.3 current traces. SA: Sialic acid, A Kv4.2. Black Circles: Kv4.2 + SA; Gray Circles: Kv4.2 – SA. n = 8-11, B Kv4.3. Black Squares: Kv4.3 + SA; Gray Squares: Kv4.3 – SA. n = 10-13 (Table 6.1).

PAGE 131

114 time that O-linked sialic acids modulate Kv channel gating through isoformspecific mechanisms. Sialic acids modulate Kv4.2 and Kv4.3 activation Kv4.2 and Kv4.3 are members of the Shal subfamily of Kv channels and are rapidly activating and inactivati ng (A-type) channels. Neither Kv4.2 or Kv4.3 contain an external N-glycosylation site75. To determine whether Kv4.2 and Kv4.3 sialylation modulated channe l gating, the two isof orms were expressed individually in CHO cells under conditions of full and reduced sialylation, and whole cell K+ currents were characterized. Figure 6.1 shows the conductancevoltage relationships measured for each is oform as expressed in Pro5 (+ SA) and Lec2 (-SA) cells. Note that the G-V re lationship was shifted significantly to more depolarized potentials for both isof orms under conditions of reduced sialylation. The voltages of half-activation for Kv4.2 and Kv4.3 under conditions of reduced sialylation were shifted to more depolarized potentials by 16 mV and 8 mV, respectively, (Table 6.1). Kv4.2 and Kv4.3 fast inactivation and recovery from fast inactivation are not altered by changes in channel sialylation Kv4.2 and Kv4.3 are rapidly activating and inac tivating (A-type) channels that undergo N-type fast inactivation. N-type i nactivation occurs when the N-terminus of the protein occludes t he pore from the cytoplasmic face, effectively blocking the flow of ions75. For Kv4.2 and Kv4.3, channel sialylation had no measurable

PAGE 132

115 Figure 6.2. Kv4.2 and Kv4.3 fast inactivation and recovery from fast inactivation are not affected by sialic acids. Pre-Pulse Po tential (mV) -120-100-80-60-40-200Normalized Current 0.00 0.25 0.50 0.75 1.00 Kv4.2 + SA Kv4.2 SA Pre-Pulse Potential (mV) -100-80-60-40-200Normalized Current 0.00 0.25 0.50 0.75 1.00 Kv4.3 + SA Kv4.3 SA Interpulse Duration (ms) 050100150200I/I0 0.0 0.2 0.4 0.6 0.8 1.0 Interpulse Duration (ms) 050100150200I/I0 0.0 0.2 0.4 0.6 0.8 1.0 A CD B 02 nA20 ms 2 nA20 ms

PAGE 133

116 Figure 6.2. Fast inactivation and reco very from fast inactivation for Kv4.2 and Kv4.3 expressed under conditions of full and reduced sialylation. Data are mean SEM, A, B : Steady-state channel availability (inactivation). Curves are single Boltzmann distribution fits to the data ( lines). Insets represent inactivation (hinf) current traces recorded from Pro5 cells, C, D : Recovery from fast inactivation at a -140 mV recovery potential. Data are mean SEM fractional current. Curves are single exponential fits to the data, A C : Black Circles: Kv4.2 + SA; Gray Circles: Kv4.2 – SA, B D : Black Squares: Kv4.3 + SA; Gray Squares: Kv4.3 SA. n = 8-13 (Table 6.1).

PAGE 134

117 Table 6.1. The measured gating parameters for Kv4.2 and Kv4.3 sialic acid. Construct n Va Ka Vi Ki mV mV mV mV Kv4.2 + SA 8 -20.7 1.7 15.6 0. 9 -66.2 2.8 -7.8 0.6 Kv4.2 – SA 11 -4.5 2.2* 16.9 0. 6 -67.2 2.6 -6.7 0.4 Kv4.3 + SA 10 -15.4 2.2 12.3 0. 5 -51.3 2.1 -5.9 0.9 Kv4.3 – SA 13 -7.8 1.5* 15.4 0. 7* -57.3 2.5 -6.1 0.4 Table 6.1. Data are the mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Vi: Voltage of half-inactivation. Ki: Boltzmann inactivation slope factor. Si gnificance was tested using a two-tailed Student’s t-test to compare gating parameters under conditions of reduced sialylation (Lec2, SA) with control c onditions (Pro5, + SA). = significance ( p <0.005).

PAGE 135

118 effect on steady-state fast inactivation or on the recovery from fast inactivation (Figure 6.2, Table 6.1). Kv2.1 is not N-glycosylated Kv2.1 is a delayed rectifier channel that produces a slowly inactivating current and contains one N-glycosylation site located on the S3-S4 linker75. Previously, it was found that Kv2.1 is not N-glycosyl ated in the brain211. To determine whether Kv2.1 is N-glycosylated in Chin ese Hamster Ovary cells, a Kv2.1 N-glycosylation mutant (Kv2.1N283Q) was generated by mutating the asparagine residue (N) that initiates the potential N-glycosylation cons ensus sequence to a glutamine (Q). In addition, fully glycosylated Kv2.1 channel protein was treated with the glycosidase PNGase-F to remove the full N-glycan structure, as previously performed by our lab218. Immunoblot analyses showed the electrophorectic mobilities of Kv2.1, Kv2.1N283Q, and Kv2.1 treated with PNGase-F were nearly identical, indicating that Kv2.1 expressed in CHO cells is not N-glycosylation (Figure 6.3). Sialic acids modulate Kv2.1 activation To determine whether sialic acids impact Kv2.1 channel function, channel gating was studied under conditions of full and reduced sialylation by expressing Kv2.1 in Pro5 (+ SA) and Lec2 (SA) cell lines. Under conditions of reduced sialylation, Kv2.1 activated at more depolarized potentia ls than the fully sialylated channel

PAGE 136

119 Figure 6.3. Kv2.1 is not N-glycosylated. 100 kDKv2.1N283QKv2.1 + PNGase-F Kv2.1 Figure 6.3. Immunoblot of Kv2.1N283Q and untreated and PNGase-F treated Kv2.1.Samples were fractionated on a 6. 5% SDS gel and probed with an antiKv2.1 monoclonal antibody (1: 1000 dilution, see Materials and Methods). Lane 1: Kv2.1N283Q, Lane 2: Kv2.1 + PNGase-F, Lane 3: Kv2.1. Molecular weight marker is denoted to the left of lane 1. 5 g of protein loaded per lane.

PAGE 137

120 Figure 6.4. Sialic acids modulate Kv2.1 activation. Membrane Potential (mV) -2002040 n (ms) 5 10 15 20 25 30 35 Membrane Potential (mV) -60-40-2002040Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Kv2.1 + SA Kv2.1 SA AB 0.5 nA40 ms 2 nA50 ms Figure 6.4. A G-V relationships for Kv2.1 under conditions of full and reduced sialylation. Data are the mean normalized peak conductance SEM and are fit to a single Boltzmann relationship (lines). Black Triangles: Kv2.1 WT + SA (Pro5); Gray Triangles: Kv2.1 WT – SA (Lec2). Inset: Typical whole cell Kv2.1 current traces. n = 12-14 (Table 6.2), B Activation time constants ( n) for Kv2.1 in the presence and absence of sialylation. Data are mean SEM. Lines are nontheoretical point to point. Symbols are the same as in A Dotted line represents the data from Kv2.1 expressed in Lec2 cells shifted by the Va. Inset: Current traces during a -20 mV test potential. The scaling factor used to normalize the currents was 1.85. Black traces: Kv2.1 + SA; Gray traces: Kv2.1 – SA. n = 12-14.

PAGE 138

121 Table 6.2. Gating parameters measured for Kv2.1. Construct n Va Ka mV mV Kv2.1 + SA 14 -12.9 1.9 13.6 03 Kv2.1 – SA 12 -3.7 1.8* 13.3 0.4 Table 6.2. Data are mean SEM. Va: Voltage of half-activation. Ka: Boltzmann activation slope factor. Significance was determined using a two-tailed Student’s t-test to compare gating parameters under conditions of reduced sialylation (Lec2, SA) with control condition s (Pro5, + SA).* = significance ( p <0.005).

PAGE 139

122 (Figure 6.4). The G-V relations hip for the less sialylated Kv2.1 was shifted by a significant depolarizing ~9 mV from t he G-V relationship measured for fully sialylated channels. Likewise, the Va was shifted to more depolarized potentials for the less sialylated channel, indicati ng a stronger depolarization is required to activate Kv2.1 in the absence of sialic acid s (Figure 6.4, Table 6.2). Sialic acids impose an appare nt electrostatic effect on Kv2.1, Kv4.2, and Kv4.3 channel gating Sialic acids are negatively charged resi dues at physiological pH that may contribute to the external negative surface potential, impacting voltagedependent channel gating. With increased n egative surface charge, the channel would activate at less depolarized potentials. T hat is, if sialic acids contribute to a negative surface potential, then channels with greater levels of sialic acids would be predicted to activate at less depolariz ed potentials than channels with reduced levels of sialylation. The presence of ex ternal divalent cations will screen this surface potential and reduce the impact of the negatively charged sialic acid residues on channel gating212,213. Here, the Va for Kv4.2, Kv4.3, and Kv2.1 as expressed under conditions of full and reduced sialylation were measured and compared at two different divalent cati on concentrations. The data show the Va for each isoform as expressed under c onditions of full sialylation was more sensitive to changes in external di valent concentrations, while the Va for the isoforms expressed in Lec2 cells were mi nimally affected by changes in external divalent concentrations (Figure 6.5). Thes e data are consistent with sialylation

PAGE 140

123 Figure 6.5. Sialic acids modulate Kv2.1, Kv4.2, and Kv4.3 activation through electrostatic mechanisms. Va with 10-fold Increase in [CationO 2+] (mV) -5 0 5 10 15 20 Kv4.2 + SA Kv4.2 SA Kv4.3 + SA Kv4.3 SA Kv2.1 + SA Kv2.1 SA Kv4.2Kv4.3Kv2.1 14 mV 6 mV 7 mV* * Figure 6.5. The Va for Kv2.1, Kv4.2, and Kv4.3 expressed under conditions of full and reduced sialylation were measured and compared at normal and low (10-fold reduction) external divalent cation concentrations. Data are mean SEM hyperpolarizing shifts in Va measured. Note the signifi cantly larger shift in Va in the presence of sialylation. Solid bars: + SA; Patterned Bars: – SA. n = 3-4. Significance (*) tested using two-ta iled Student’s t-te st comparing Va shifts with a

PAGE 141

124 tenfold increase in external divalent cation concentration as measured in Pro5 versus Lec2 cells ( p <0.005).

PAGE 142

125 altering the activation of Kv4.2, Kv4.3, and Kv2.1, through electrostatic mechanisms. Kv2.1, Kv4.2, and Kv4.3 channels are O-glycos ylated and sialylated Data shown here indicate that sialic acids attached to three Kv isoforms differentially impact channel activation. Prev ious efforts to determine how glycans modulate channel function typically ques tioned how N-linked glycans impact channel function. There have been no previous studies of putat ive effects of Oglycans on voltage-gated ion channel function. However, Kv4.2 and Kv4.3 do not contain N-glycosylation sites, and we show here that Kv2.1 is not N-glycosylated as expressed in CHO cells. Thus, the data suggest that sialic acids attached through O-linkages are responsible for the observed effects on ch annel gating. O-glycosylation does not have a recognized consensus sequence; however, it is known to be attached to serines or thr eonines with prolines, alanines, serines, and threonines among t he neighboring residues176-182. Thus, numerous potential O-glycosylation sites are located on each Kv channel (i.e., many externally located serine and threonine residues), making removal of all potential Oglycosylation sites through mutagenesis near ly impossible. To overcome this limitation, Click chemistry was employed to identify whether Kv2.1, Kv4.2, and Kv4.3 were O-glycosylated and sialylated185,235,237,267. Tetraacetylated azidomodified sugars, Ac4GalNAz (O-glycosylation) or Ac4ManNAz (sialylation), were incorporated into protein glycan struct ures and bound with biotin-alkyne (Figure 6.6). Because of the strong interacti on between streptavidin and biotin, the

PAGE 143

126 Figure 6.6. Metabolic incorporation of labeled sugars utilizing Cu(I)catalyzed cycloaddition. Input Output A B C

PAGE 144

127 Figure 6.6. A Azide-modified sugars are metabol ically incorporated into cell surface protein glycan structures through the permissive nature of the oligosaccharide biosynthesis pathway, B An azide-modified Nacetylmannosamine (ManNAz) is incorporat ed into the sialic acid biosynthetic pathway generating an azide-modified sia lic acid residue. An azido analog of Nacetylgalactosamine (GalNAz) can be me tabolically introduced at the core position of mucin-type O-linked glycoproteins, C The Cu(I)-catalyzed cycloaddition reaction occurs between an azide (attached to the protein) and a biotin-alkyne to form a stable triazole conjugate, which can be detected via a steptavidin column. Figures adapted from Prescher and Bertozzi, 2005268 and www.Invitrogen.com 2008.

PAGE 145

128 biotinylated samples were incubated wit h streptavidin-bound magnetic beads to isolate only those proteins with the modified sugar inco rporated into the protein glycan structures (see Materials and Me thods). Immunoblot analysis of the Ac4GalNAz and Ac4ManNAz samples detected bands at the appropriate molecular weights for Kv2.1, Kv4.2, and Kv4.3 under both conditions (Figure 6.7). These data indicate that each Kv isoform tested is O-glyco sylated and sialylated, supporting the hypothesis that O-linked sialic acids are re sponsible for the effect of sialic acids on Kv2.1, Kv4.2, and Kv4.3 gating. Discussion O-linked sialylation modulates the gating of Kv2.1, Kv4.2, and Kv4.3 through apparent electrostatic mechanisms Here, we examined the role of Olinked sialic acids on gating of Kv2.1, Kv4.2, and Kv4.3. Figures 6.1 and 6.4 show that a reduction in sialic acids causes a significant depolarizing shift in the G-V re lationship for all three isoforms studied. Kv4.2 and Kv4.3 do not contain an N-glycosylati on consensus site within their extracellular amino acid sequence and therefore, cannot be N-glycosylated. Kv2.1 contains one N-glycosylation site on it s S3-S4 linker, but as predicted, the data shown in figure 6.3 indicate that th is site is not N-glycosylated on channels expressed in CHO cells, c onsistent with previous studies in COS-1 cells and rat brain211. Several possible explanations for the lack of glycosylation of the potential Kv2.1 site exist. A likely explanation is that the N-glycosylation site, located on the S3-S4 linker, is not acce ssible to the glycosylation machinery,

PAGE 146

129 Figure 6.7. Kv2.1, Kv4.2, and Kv4.3 channels are O-glycosylated and sialylated. ControlManNAz GalNAz100 75 75 ControlManNAzGalNAz 75 ControlManNAzGalNAz Kv2.1 A Kv4.2 B Kv4.3 C Figure 6.7. Click chemistry confirm ed the presence of O-glycosylation and sialylation attached to Kv2.1, Kv4.2, and Kv4.3. Immunoblots of Kv2.1, Kv4.2, and Kv4.3 lysates. Samples were fractionat ed on a 6.5% SDS gel and probed with anti-Kv2.1, anti-Kv4.2, or anti-Kv4.3 monoclonal antibodies (1:100-1000 dilution, see Materials and Methods).Lanes 1: Control samples, Lanes 2: Ac4ManNAz labeled (sialic acid) samples, and Lanes 3: Ac4GalNAz-labeled (O-glycosylated) samples. Molecular weight marker is denoted to the left of lane 1. A Kv2.1, B Kv4.2, C Kv4.3.

PAGE 147

130 thereby preventing addition of glyc osylation (as suggested by the Kv channel crystal structures)84,85. Although Kv2.1, Kv4.2, and Kv4.3 are not N-glycosylated we observed isoformspecific effects of sialylation on channel gating. A reduction in sialylation caused rightward, depolarizing shifts in the GV relationships that were unique in magnitude for each isoform (Figures 6.1 and 6.4, Tables 6.1 and 6.2). Sialic acids modulate activity of each isoform through apparent electrostatic mechanisms. The data indicate that negat ively charged sialic acids modulate activation of Kv2.1, Kv4.2, and Kv4.3 through electrostatic mec hanisms. Based on the surface potential theory, negative charges on t he outer surface of the membrane generate a surface potential. Specifically with a reduction in external negative charges, a hyperpolarization would be s ensed by the channel gating mechanism, and thus, channel activation would require a greater depolarization212,213. Extracellular divalent cations should act to screen the effects of negatively charged sialic acid residues on surfac e potential, effectively reducing the negative surface potential. We found t hat channels expressed in the fully sialylating Pro5 cells were more sensitiv e to changes in extracellular divalent cation concentration than channels expre ssed in the reduced sialylating Lec2 cells (Figure 6.6). The data indica te that sialic acids modulate Kv2.1, Kv4.2, and Kv4.3 activation through electrostatic mechanisms 212,213. Further, changes in

PAGE 148

131 sialylation shifted the G-V curve for each isoform nearly linearly, with little or no effect on the slope. This suggests that changes in Kv2.1, Kv4.2, and Kv4.3 sialylation primarily alter the negative surf ace potential, with little to no effect on the stability of channels among functional st ates. This is distinct from that previously described for Kv1 isoforms, in which N-glycans conferred stabilizing effects on Kv1 channel states214,215. Steady-state inactivation and reco very from inactivation of Kv4.2 and Kv4.3 were not significantly impacted by changes in channel sialylation (Figure 6.2, Table 6.1). Most previous studies showed that the surface char ge effects of sialic acids on channel voltage-dependent gati ng similarly affect all voltage-dependent gating mechanisms, suggesting that the surfac e potential produced by sialic acids contributes uniformly to the electric field sensed by each channel gating mechanism216,217. Typically, with a linear change in channel activation along the voltage axis, as observed in Figure 6.1A, there is a consistent shift in channel steady-state inactivation voltage dependence. In this study, we show that sialic acids shift only the G-V relationship for Kv4.2 and Kv4.3, with no effect on inactivation or recovery from inactivation. The data also indicate that sialic acids modulate activation voltage by contributi ng to the negative surface potential. Together, these data suggest that sialic acids affect channel gating through electrostatic mechanisms, but do not c onfer a uniform effect on all voltagedependent gating mechanisms. One likely expl anation is that sialic acids impose a unique, inhomogeneous effect on the el ectric field sensed by the channel

PAGE 149

132 gating mechanisms (for Kv4.2 and Kv4.3 isoforms), such that only activation is affected. Such a phenomenon has not been previously described for Kv channel gating. A single previous report indicated that sialic acids impact gating of Kv4.3217. However, the study did not determine the si alic acid linkages responsible for the effect on channel gating. Our data are in general agreement with the previous report; Kv4.3 channel activation voltage was sh ifted to depolarized potentials under conditions of reduced sialylation. The preceding report indicated a small, but significant, shift in channel inactivati on voltage that we did not observe. There are several possible reasons for the rela tively minor differences between our data and that previously described, includi ng the use of different external and internal solutions to record Kv channel currents. It was shown in the previous study that sialolipids contri bute at most 3-4 mV to Kv4.3 channel gating. We measure a much larger sia lic acid-dependent shift for Kv4.3 gating (~ 8 mV) and an even greater shift for Kv4.2 gating (~16 mV), indicating that most of the effect of sialic acids on Kv4 gating is caused by O-linked channel sialic acids. In addition, while Kv4.2 and Kv4.3 are homologous proteins, changes in channel sialylation have a much greater impact on Kv4.2 (the Va was shifted by 16 versus 8 mV). This suggests that O-linked sialic acids attached to Kv4.2 and Kv4.3 impact channel activation differently. That is, Kv4.2 and Kv4.3 O-linked sialic acids apparently are different in number or location relative to the channel

PAGE 150

133 activation mechanisms. Significant furt her studies are required to determine whether changes in the number and/or locati on of O-linked sialic acid impact Kv4 channel activation. Summary We showed that Kv2.1, Kv4.2, and Kv4.3 are O-glycosylated and sialylated in CHO cells using Click chemistry (Figur e 6.6) and that O-linked sialic acids modulate gating of each isoform uniquely (Figures 6.1 and 6.4). No previous works link Kv channel O-glycosylation to channel gating. Therefore, the data presented in this study reflect a novel finding that Kv channel gating is modulated by O-linked sialic acids. Implications of this study are broad, given the abundant and ubiquitous nature of glycosylation.

PAGE 151

134 CHAPTER 7 FINAL DISCUSSION During arrhythmic episodes, cardiac c onduction and rhythm are altered. ECGs and APs measure excitability in the heart and cardiomyocytes; from these, we can pinpoint which currents are modulat ed under developmental and pathological conditions. As previously discussed in detail, Kv currents are responsible for AP repolarization, returning the membrane potential to hyperpolarized potentials. Post-translational modification of Kv channels may be one mechanism by which channel function, and thus AP repolarization, can be altered. Therefore, the aim of this study was to determine whether aberrant glycosylation, via reductions in sialylation, have a direct effect on ca rdiac excitability by impacting cardiac Kv currents and AP repolarization; a k nockout mouse model, treatment with glycosidases, and a heterologous expressi on system were utilized to perform these studies. ST3Gal-IV deficient mice sh ow alterations in ECGs The knockout mouse strain used in this study, ST3Gal-IV(-/-) mice, does not demonstrate an obvious, di stinct phenotype. However, a previous study determined that mice lacking ST3Gal-I V have a deficiency in Von Willebrand factor (VWF), a plasma si aloglycoprotein with 24 putativ e glycosylation sites that

PAGE 152

135 stabilizes coagulation factor VIII Because of this, ST3Gal-IV(-/-) mice showed an increased bleeding time, a subs tantial decrease in platel ets (30% of normal), and can develop thrombocytopenia due to elev ated platelet clearance. The observed effects are likely the result of a decreased VWF half-life caused by a reduction in sialylation that exposes galactose linkages, which have been implicated as asialoglycoprotein receptor ligands219. These results suggest that a deficiency in ST3Gal-IV can have a multitude of effe cts on normal physiologic processes. Mice with a knockout of the sialyltr ansferase ST3Gal-IV were utilized for telemetric analysis of ECGs. ST3Gal-IV(-/-) mice demonstrated slower heart rates and showed putative changes in atrial conduction. This was observed as broadened P waves and R-R s egments on the ECG (Figure 3.1). An increased incidence of arrhythmic beats and greater variation between beats was noted in the ST3Gal-IV(-/-) animals, suggesting that the lo ss of a single sialyltransferase can promote arrhythmogenic conditions ; conduction abnormalities were not detected in the littermate controls. Howe ver, the ECG data does not give a clear picture of exactly how cardiac excit ability is impacted by ST3Gal-IV. Nonetheless, a reduction in sialic acids attached to ion channels may be one mechanism, as channel proteins typically are glycosylated/sialylated heavily. For example, the cardiac sodium channel, Nav1.5, possesses 13 putative, extracellular N-glycosylation sites; c hanges in sialylation have been shown to alter Nav1.5 function218. In addition, pacemaker activity may be impacted by aberrant sialylation of HCN channels, which have one putative N-glycosylation

PAGE 153

136 site on the S5-S6 linker269. The potential alteration in HCN activity likely will modulate pacemaker current (If) in the SA node, thus affecting normal cardiac rhythm. Changes in glycosylation, namely si alylation, may alter the contribution of the autonomic nervous system as well. Parasympathetic input contributes to beat-to-beat pacing of cardiac rhythm14; therefore, reduc ed glycosylation might increase vagal activity, which would decrease heart rate and (potentially) the susceptibility to arrhythmia s. On the other hand, impaired N-glycosylation is one mechanism by which neural norepinephri ne transporter (NET) density is reduced at sympathetic nerve endings270. This will lead to dimi nished norepinephrine (NE) uptake, which is a salient feature of cardiac failure271,272. Thus, the observed susceptibility to arrhythmic beats in ST3Gal-IV(-/-) mice, as deduced from ECG analysis, may be the result of multiple sialic acid-dependent effects. Regulated and aberrant gl ycosylation modulates atri al, but not ventricular, action potentials In order to evaluate the e ffects of ST3Gal-IVdeficien cy more thoroughly, action potentials were recorded fr om isolated neonatal atrial and ventricular myocytes. In this way, any potential influence of the autonomic nervous system and pacemaker cells should be eliminated. All experiments were performed in the neonate, as data suggested that sodium channels are less sialylated in neonatal ventricles compared to neonatal atria218.

PAGE 154

137 Action potentials measured from ST3Gal-IV(-/-) atrial myocytes were prolonged significantly compared to control (Figur e 3.2). Treatment of control atrial myocytes with neuraminidase or PNGase -F to cleave sialic acids or Nglycosylation, respectively, lengthened action potential duration similar to that of the knockout (Figures 4.1 and 4.2). Specifically, the action potential durations at 10%, 50%, and 90% repolarization were lengthened by ~25-150% in ST3GalIV(-/-), neuraminidase treated, and PNGase-F tr eated atrial myocytes (Tables 3.1 and 4.1). A reduction in sialylation slowed the AP ri sing phase in atrial myocytes. The time to peak was prolonged in ST3Gal-IV(-/-) and neuraminidase treated atrial myocytes, which suggests an effect on Nav channel function. However, time to peak was unaffected with PNGase-F treatme nt (Table 4.1). This difference could be achieved by contribution of O-linked si alic acids, which are not removed upon PNGase-F treatment, to AP production. Incomplete removal of N-glycans with PNGase-F treatment may also account for the similar time to peak as control. Chapter six shows O-linked sialylation alters gating of Kv channel isoforms expressed in an in vitro system; therefore, the slight variation in AP parameters between ST3Gal-IV(-/-), neuraminidase, and PNGase-F treated myocytes could be a modulatory effect of O-linked sialic acids (Tables 3.1 and 4.1).A previous study from our lab report ed similar effects of al tered sialylation on AP parameters; the rate of depolarizat ion and action potential duration were lengthened significantly in STX(-/-) atrial myocytes189. The extended atrial AP

PAGE 155

138 repolarization phase observed in ST3Gal-IV(-/-), STX(-/-), and glycosidase treated myocytes suggests Kv channel activity is modulated as well. A rightward shift in the voltage-dependence of activation would del ay initiation of AP repolarization, thus lengthening the r epolarization phase. Interestingly, ventricular AP paramet ers were not affected in ST3Gal-IV(-/-) or STX(-/-) mice; AP parameters also were not al tered significantly in glycosidase treated ventricular myocytes (Figures 3. 3 and 4.3). Human ventricular APDs, due to a lengthened plateau phase, are longer than atrial APDs. In the neonatal mouse, the same holds true. If volt age-gated ion channels expressed in the neonatal ventricle are less sialylated t han the atria, as shown for sodium channels218, the chamber-specific differences in APD may be accounted for (at least partially in the neonate) by changes in sialylation. Together, these data suggest the presence of normal glycosylat ion structures is not required for production of typical neonatal ventricular ac tion potentials. However, a previously published study determined that adult v entricular myocytes treated with neuraminidase to remove sialylat ion displayed lengthened APDs and an increased incidence of ear ly after depolarizations217. One possible explanation for the conflicting data is adult ventricl e has a greater portion of complex, biantennary glycans than neonatal ventricle189,218. Furthermore, N-glycosylation structures from adult atri al and ventricular tissue have a significantly larger proportion of sialic acids than do neonatal cardiac samples189. Hence, adult ventricular action potentials may be more sensitive to changes in glycosylation

PAGE 156

139 than their neonatal counterparts. Reduced sialylation of neonatal ventricular myocytes should slow conduction of APs through the ventricles of the neonatal animal compared to the adult. With reduced glycosylation, chang es in activation of repolarizing Kv currents are consistent with the effects reported for APs Since alterations in glycosylation demonstr ated a significant effect on atrial AP repolarization rate, Kv currents that underlie the repolarization phase were recorded from ST3Gal-IV(-/-), glycosidase treated (neuraminidase or PNGase-F), and control atrial and ventricular myo cytes. Steady-state activation of Ito and IK,slow, two components of IK, was shifted to more depolarized potentials in knockout and glycosidase treated atrial myocytes, observed in Figures 3.4 and 4.4 as depolarizing shifts in the G-V relationships and the Va. Furthermore, treatment of atrial myocytes with PNGase-F did not result in a further affect on Ito or IK,slow activation compared to ST3Gal-IV(-/-) and neuraminidase treated atrial myocytes, suggesting that N-glycosylati on does not contribute an additional effect and sialic acids exert t he entire effect of sugars on Ito and IK,slow (Figure 4.4). These data support the atrial AP findings and our Kv1.5 data in the heterologous expression system as this isoform demon strated no significant effect of N-glycosylation (other than si alic acids) on channel gating (Chapter 5). Treatment with neuramin idase produced a slightly greater depolarizing shift in the Va of atrial IK,slow compared to PNGase-F treatment Data from chapters five and six suggest that O-link ed sialic acids alter Kv2.1 gating, while N-linked

PAGE 157

140 sialylation is responsible for the en tire effect of sialic acids on Kv1.5 channel function. Thus, O-linked sialic acids attached to Kv2.1 may be responsible for the additional effect on atrial IK,slow activation observed with neuraminidase treatment. Since Kv4.2 and Kv4.3 are not N-glycosylated (Chapt er 6), the effects of Nglycosylation on Ito activation are likely due to N-linked SAs attached to Kv1.4. Although no effect of a reduction in glycosylation was observed on Kv1.4 gating in CHO cells (Chapter 5) it is possible that in vivo channels may be impacted differently by changes in glycosylation, as previously found for cardiac Nav1.5 channels218,220. This may be because the channels are more sialylated in vivo than in CHO cells. Chapter six shows that O-linked sialic acids exert an effect on Kv4.2 and Kv4.3 activation. Hence, the sli ghtly larger shift in the Va for atrial Ito following neuraminidase treatment (com pared to PNGase-F treatment) may be due to O-linked sialic acids altering Kv4.2 and/or Kv4.3 gating. Consistently, a study from our lab showed that steady-state and kinetic activation of murine atrial Nav currents was shifted to more depolar ized potentials (by ~7 mV) in STX(-/-) neonatal myocytes189. Our lab also published that activation of Nav currents recorded from rat neonatal atrial myocyt es was shifted to more depolarized potentials upon treatm ent with neuraminidase218. The alteration in Nav currents measured from STX(-/-) and rat neonatal atrial my ocytes correlates with the slowed time to peak observed in STX(-/-) and ST3Gal-IV(-/-) atrial APs. Neonatal ventricular Ito and IK,slow were not impacted by changes in glycosylation. That is, ST3Gal-IV(-/-) and glycosidase treated ventri cular myocytes produced Kv

PAGE 158

141 currents similar to those recorded from control (Figures 3.5 and 4.5). Since we observed no significant altera tion in ventricular repolariz ation, these results were predicted. A study by our lab cons istently showed no affect on neonatal ventricular Nav currents collected from STX(-/-) mice189. On the other hand, published data from our lab also sugges ts that treatment with neuraminidase caused depolarizing shifts in steady-sta te activation and inactivation of Nav currents recorded from adult rat ventricula r myocytes. Similarly, Ufret-Vincenty, et al found that sialic ac ids modulate murine Ito, but not IK,slow, in adult ventricular myocytes. These contradictory results are not surprising given that neonatal ventricle has less complex glycan struct ures (and likely fewer sialic acid residues) than adult ventricle189,218. The differences in half-activation voltage of Ito and IK,slow between the atria and ventricle not accounted for by changes in glycosylation are likely due to differential expression of Kv channel isoforms. For example, Kv1.5, which demonstrated t he largest shift in Va upon a reduction in sialylation as expressed in CHO cells, is expressed at greater levels in the atria than the ventricle273. In all, glycosylation affects the production of neonatal cardiac potassium currents and, thus, ac tion potentials through chamber-specific mechanisms. Kv channel activation is modulated by reduced glycosylation in isoformspecific manners To fully appreciate the glycosylation-specific effects on cardiac K+ currents and excitability, the specific channel isof orms underlying the production of Ito and

PAGE 159

142 IK,slow must be examined under conditions of full and reduced glycosylation. Briefly, Kv1.4, Kv4.2, and Kv4.3 produce Ito, which perpetuates initial AP repolarization; IK,slow is generated by Kv1.5 and Kv2.1. For functional studies, however, these channel isoforms were cla ssified by sub-family and by the type of glycosylation attached to each channel. The Kv1 isoforms each contain one putative N-glycosylation site on the S1-S2 linker and, as discussed in Chapter 5, Kv1.4 and Kv1.5 are N-glycosylated and sialylated. Kv4.2 and Kv4.3 do not have an extracellular N-glycosylation site and, therefore, are not N-glycosylated. One putative N-glycosylation site is located on the S3-S4 linker of Kv2.1; however, recent studies suggest this site is i naccessible to the cell’s glycosylation machinery84,85 and that Kv2.1 expressed in CHO cells (Chapter 6) and the mammalian brain is not N-glycosylated211. Thus, the Kv isoforms investigated in this study are categorized by the pres ence or absence of N-glycosylation. Gating of two Kv1 isoforms is modulated un iquely by glycosylation Gating parameters of tw o homologous Shaker K+ channels, Kv1.4 and Kv1.5, were studied under conditions of full and reduced glycosylation. N-linked sialic acids contribute to the voltagedependence of channel gating for Kv1.5 through electrostatic mechanisms (Figures 5. 2 and 5.4), but have no affect on Kv1.4 gating (Figure 5.5). These data are novel in several respects, including: 1) Kv1.4 is the only member of the Kv1 subfamily in which the attached N-glycans have no measurable effect on channel gating, 2) Wi th reduced sialylation/glycosylation,

PAGE 160

143 Kv1.5 exhibited the largest depolarizing sh ift in voltage-dependence of activation of any member of the Kv1 subfamily, and 3) Kv1.5 is the only mem ber (to date) of the Shaker Kv channel family for which N-linked sialic acids account fully for the impact of N-glycans on channel function, produced solely by an electrostatic effect. Because Kv1.5 produces an important repol arizing current in the human atria, IKur, altered Kv1.5 channel function, whether th rough regulated expression or glycosylation, could lead to signi ficant changes in AP repolarization122. O-linked sialic acids alter activation of Kv2.1, Kv4.2, and Kv4.3 In chapter six, we showed that reduced si alylation, likely O-lin ked, caused unique depolarizing shifts in steady-state activation voltage for three Kv channel isoforms that are not N-glycosylated (Figures 6.1 and 6.4). An electrostatic mechanism is apparently responsible for thes e effects, with O-linked sia lic acids contributing to the external negative surface potential (Fi gure 6.5). Inactivation and recovery from inactivation were not impacted signifi cantly by reduced sialylation of the two rapidly inactivating isoforms, Kv4.2 and Kv4.3. This suggests a non-homogenous influence of Kv4 sialic acids on the electric field sensed by the activation, inactivation, and recovery gating mechani sms (Figures 6.1 and 6.2). Together, the data indicate that O-linked si alic acids modulate gating of Kv channels that are not N-glycosylated, and that this modulation is unique for each isoform. This is the first study to report direct ef fects of O-glycans on ion channel gating.

PAGE 161

144 Physiological and pa thological relevance Over 1% of the human genome is implicat ed in glycosylation of proteins and lipids, which occurs through a highly regulated process. Due to the abundance and complexity of glycosylation, proper regulation is vital to normal protein function. Protection, stabilization, and organization are among the responsibilities of glycosylation. For instance, glycan st ructures act as antigens, determining blood type. Aberrant glycosylation also can resu lt in pathological consequences or exacerbate another condition. As discuss ed in the introduction, Congenital Disorders of Glycosylation are multi-symptom disorders in which the assembly or processing of glycans is altered by mu tation. In CDG type-Ia (CDG-Ia), the central nervous system (CNS) is traditionally the main site affected, as noticeable mental and neuromuscular retardation per sists; however, patients also tend to exhibit cardiac failure274. The results presented here suggest that alterations in glycosylation can impact cardiac ex citability by modulating the Kv currents involved in AP repolarization. Thus, abno rmal or reduced glycosylation in CDG-Ia patients may alter ion channel acti vity, in particular that of Kv channels, changing the AP waveform and leading to cardiac arrhythmias and failure. Additionally, glycosylation has been shown to enhance ion channel function; therefore, a decrease in the number of glycans could dimi nish the activity of ion channels as well as other proteins. For example, ner ve conduction velocity is slowed in CDG-Ia patients274. This could be due to changes in the activity of Nav channels,

PAGE 162

145 decreasing the excitability of the cell. Furthermore, altered function of glycoproteins involved in myelination could display a similar outcome by lessening the amount of myelin attac hed to an individual neuron. Together with the effect of reduced glycosylation on card iac ion channels, the alteration in conduction velocity could manifest the pr esented cardiac failure. Some of the cardiac maladies detected in patients wit h the sialic acid-cleaving parasitic disorder, Chagas disease, may be generated in a similar manner. As detailed in the introduction, Chagas dise ase is a parasitic disorder caused by T. cruzi which afflicts approximately 18 milli on people worldwide. Initial stages of host cell invasion and host-parasite interactions are mediated by trans -sialidase, an enzyme that cleaves and transfers SA residues from host cells, and SA acceptors on the parasitic cell surface208,275. Significant quantities of trans sialidase are released into the cytosol of host cells and, eventually, into the extracellular space275. Thus, sialic acids from glycoproteins, such as Kv channels, and glycolipids may be removed as a re sult. This likely would decrease the activity of cardiac ion channels and alte r AP production (as described in chapters three and four), leading to cardiac conduc tion and rhythm disruptions. In fact, atrioventricular and intraventricular blocka ges are common in mice infected with T. cruzi276. Chronically, pati ents infected with T. cruzi typically develop myocarditis and/or dilated congestive cardiomyopathy, with one hallmark of the disease being ventricular apical aneur ysms. Because of the destruction of

PAGE 163

146 conduction tissues and potentially desia lylation of ion channels, cardiac arrhythmias typically arise277. Dystroglycan, a membrane protein that bi nds to components of the extracellular matrix (ECM), contains an estimated 50 O-glycosylation sites (many of which likely are sialylated) in the middle third of the molecule, constituting approximately 50% of the molecular weight278. When O-glycosylation is removed from -dystroglycan, laminin binding is inhi bited, dissociating the ECM from the muscle membrane279. Glycosylation-dependent congenit al muscular dystrophies (CMDs) are defined by altered -dystroglycan glycosylation leading to wasting of skeletal muscles. However, cardiac muscle can be severely affected as well. Atrial fibrillation or flutter was obs erved in patients with myotonic muscular dystrophy type 1 (MD1), Emery-Drei fuss muscular dystrophy (EDMD), Duschenne muscular dystrophy (DMD), and Becker muscular dystrophy (BMD) in particular. Consistent ly, our data suggest that hypoglycosylation/hyposialylation resu lts in altered atrial conduction, demonstrated by a broadened P wave on t he ECG, prolonged atrial APD, and changes in the voltage-dependence of two repolarizing atrial Kv currents. Thus, the secondary effect of altered cardiac ex citability in muscular dystrophy patients may be due to changes in O-linked sialic acids attached to cardiac potassium channels. Variations in Kv channel function via reduced glycosylation/sialylation also may contribute to slowing of neural conduction, as a host of brain defects are associated with certain types of muscular dystrophy278.

PAGE 164

147 Alterations in glycosylation have been link ed to generation or exacerbation of various other diseases. For example, Ig A nephropathy (IgAN) is characterized by aggregation of hypoglycosylated and hyposia lylated IgA1, which forms immune complexes that promote glomerular inflammation200,202. Mutations in the glycosyltransferase PPGalNAc-T3 are res ponsible for some forms of familial tumoral calcinosis due to production of under-glycosylated FGF23 (a phosphate regulating hormone)197. This leads to prematurel y cleaved FGF23 and improper control of phosphate levels280. Additionally, changes in glycosylation have been associated with oncogenic acti vation and cancer progression281. In fact, knockout of the Mgat5 gene, encoding the glycosyltr ansferase GlcNAc-T5, results in a reduced metastatic response282. Based on the data presented in this study, individuals with these disorders may be mo re likely to develop cardiac conduction abnormalities, as reduced glycosylati on of other proteins, such as Kv channels, could result. Since glycosylation is a ubiquitous and abundant modification of lipids and proteins, alterations in specific aspects of glycosylation, as in the aforementioned diseases, could have seco ndary effects, such as cardiac arrhythmias, that have not yet been linked to a particular disorder. Treatment with glycosidases or glycosyltransferases could ac t to slow the progression of cancer in a less toxic manner than current therapi es or reverse or lessen the symptoms associated with familial tumoral calcinos is. Furthermore, this may prevent the potential development of secondary sympt oms, such as cardiac failure or arrhythmias.

PAGE 165

148 Chronic consumption of la rge quantities of alcohol is associated with atrial fibrillation, congestive heart fa ilure, and other cardiomyopathies283,284. Changes in SA expression have been observed in pa tients consuming large amounts of ethanol (alcoholics). In fact, SA concentra tions were higher significantly in alcoholics than in healthy controls. Howe ver, there was no distinction between those alcoholics with or without liver disease285. Ethanol has been shown to alter protein glycosylation and decrease t he activity of sialyltransferases286,287; therefore, alcoholics ma y have reduced addition of SA to cardiac channel proteins, which, again, may alter ioni c currents involved in AP production and increase the susceptibility to arrhythmia s. Additionally, low density lipoproteins (LDL) are glycolipids containing terminal sialic acid residues. Although the significance of SAs on lipoproteins has not been clearly identified, SAs have been associated with lipoprotein solubilit y, receptor binding and uptake, and cholesterol efflux, among others288. In fact, sialic acids may play a role in early atherogenesis, as a previous study show ed that desialylation of LDLs induced cholesteryl ester accumulation in human aortic smooth muscle289. As previously mentioned, voltage-gat ed ion channels contribute to action potential initiation and pr opagation in skeletal muscle, neuronal, and cardiac cells. Often, the activities of several Kv channel isoforms are responsible for AP repolarization; the AP waveform is m odulated (remodeled) through changes in the expression, density, and/or distribution of specific Kv isoforms. Changes in posttranslational modifications, such as glycosylation, can potentially impact

PAGE 166

149 gating of Kv channels, leading to changes in AP waveform217,290,291. Thus, Kv channel activity would be modulated by changes in glycosylation and/or sialylation that result from remodeling of Kv isoforms (subunit-specific differential glycosylation). For example, as Kv1.4, Kv1.5, Kv2.1, Kv4.2, and Kv4.3 expression and distribution are regulated, AP repolariz ation would be altered by the changes in Kv channel activity. Because sialylation modulates Kv channel function uniquely for each isoform, r egulated expression of these Kv isoforms would lead to changes in the contribution of sialic acid to AP repolarization. In this incidence, no significant modification of Kv channel mRNA expression was noted in ST3GalIV(-/-) atria and ventricles compared to control. These findings confirm that modified expression of Kv channels in knockout cardiac tissue does not occur and thus, cannot account for the observed changes in APD or Kv currents. In addition to the relevant isoform-specif ic effects of glycosylation on channel function, AP repolarization may be modulat ed through a cell-specific process of regulated glycosylation, in which the glycosyl ation structures attached to a single Kv channel type might vary, thereby altering Kv channel activity. The data presented here suggest that Kv1.5, Kv2.1, Kv4.2, and K4.3 activity is modulated by N-and O-linked sialic acids. Our lab previously showed that Nav1.5 is less sialylated in the neonatal ventricles than t he neonatal atria, and this difference in Nav1.5 sialylation is responsible fo r the differences in channel gating observed218.Kv channels, including those exam ined here, may follow a similar sialylation pattern in atria and ventricles. This is consistent with our previously described data, in which action potential duration was shortened in the neonatal

PAGE 167

150 atria compared to the ventricles because of a sialic acid-dependent increase in Kv channel activity in the atria. Therefore, as cardiac glycosylation is regulated, Kv channel sialylation levels would change, modulating Kv channel gating. This cell-specific change in Kv isoform gating would hav e a direct effect on AP repolarization. As described in this study, altered gl ycosylation differentially modulates Kv channel function. In addition to changes in expression and/or glycosylation, mutations in the or -subunit of channel proteins, namely Nav1.5, KvLQT1, hERG, and KCNE1 and 2, can result in al tered channel function as well. This can modulate the cardiac AP, leading to such channelopathies as Long QT Syndrome and Brugada Syndrome (s ee Introduction for details)292. However, Nav1.5, KvLQT1, hERG, and KCNE1 and 2 contai n extracellular N-glycosylation sites22,209,293-296; thus, the normal activity of these channels could be altered by changes in glycosylation, pr oducing similar losses or gains of function as that observed in the currently identified c hannelopathies. Furthermore, changes in glycosylation of mutant channel proteins could act to exacerbate or correct the observed pathological cons equence of the mutation. Kv1.5 encodes the ultra rapid delayed rectifier current, IKur, in human atria122. Due to the specificity of IKur, as it is not expressed in human ventricular myocytes or Purkinje fibers, modulation of IKur and Kv1.5 is considered a potential target for the treatment of atrial arrhythmias297. Therefore, considering the results from this

PAGE 168

151 study, altering sialylation of Kv1.5 could be explored as a potential therapy to control atrial fibrillation. The data presented here suggest that Kv2.1, Kv4.2, and Kv4.3 activity is modulated by O-linked sialic ac ids. As formerly discussed, Kv4.2 and Kv4.3 are responsible for producing a rapidly acti vating and inactivating cardiac current, Ito, involved in early action potential repolarizat ion (phase 1). A previous report indicated that changes in murine ventricular Ito following treatment with an enzyme to remove sialic acids may c ontribute to the ex tended AP duration and increase in early after depolarizations observed217. In diabetic myocytes, the expression of Kv4 channels was shown to be reduced, while Kv1.4 expression was increased298,299. This may be a compensatory effect of diabetes, as Kv1.4 channel gating was shown in this study to be unaltered by changes in glycosylation (as expressed in CHO cells). Furthermore, Kv2.1, which produces a slowly inactivating current, aids in regul ation of excitabili ty in cortical and hippocampal pyramidal neurons by acting as a suppressor during periods of hyperexcitability300,301. Therefore, alterations in Kv2.1 sialylation may disrupt this suppressive role and increase the frequency and/or duration of hyperexcitability in these neurons. Many potential implicati ons of this study exist beyond that of the cardiovascular system.

PAGE 169

152 Future Directions Here, we showed that aberrant gl ycosylation modulates gating of Kv channels, altering cardiac K+ currents and AP repolarization; in turn, cardiac excitability may be impacted. However, much more re mains to be done. To acquire a more comprehensive understanding, analyses of t he contribution of O-glycans must be further elucidated. Although O-glycosidas e treatment is harsh on the cells (because of pre-digestion with multiple enz ymes) and somewhat ineffective, it is the only means available to investigat e the effect of O-glycosylation on AP production and Kv channel function. Based on the Kv channel data presented here, we would predict that removal of O-glycans would lengthen atrial APD and time to peak and shift the voltage-dependence of atrial Kv currents to more depolarized potentials, without an effect on ventricular APs or Kv currents. Additionally, examination of sodium and calcium currents from murine neonatal atria and ventricles should be performed using ST3Gal-IV(-/-) and glycosidase treated (neuraminidase, PNGase-F, and O-glycosidase) myocytes. APs and potassium, sodium, and calcium currents measured from adult atria and ventricle should be tested under the af orementioned experimental conditions. Previous studies from our lab suggest that adult at ria and ventricles contain more complex glycosylation than neonatal ventricular myocytes189; therefore, a larger effect of glycosylation may be observed on APs and currents recorded from adult myocytes.

PAGE 170

153 These experiments also could be perfo rmed under pathological conditions, using mouse models of diabetes, heart failure, or (inducible) myocardial infarction, for example. Another possible project would be to investigate the effect of changing glycosylation on neuronal and skeletal muscle conduction pathways throughout development and with the onset of disease. All of these experiments, including those conducted in the heterologous ex pression system, should be performed under more physiologic temperature conditions—37C, as opposed to 22C. Previous studies have found that changes in temperatur e can alter the gating properties of ion channels302,303. For example, Kv2.1 currents recorded from pancreatic -cells at 22C displayed typical slow inactivation rates; however, when the temperature was increased to 3235C, these currents exhibited faster activation and inactivation kinetics, re sembling those currents produced from Kv1.4302. Currents measured from hypoglyco sylated cardiomyocytes, neurons, skeletal muscle cells, and/or transiently transfected CHO cells could be examined under hypoxic, acidotic, and ischemic conditions as well. Ito amplitude was decreased in ischemic canine ventri cular myocytes, as reported in a prior study; furthermore, separately induced co nditions of hypoxia and acidosis reduced Ito amplitude304. Therefore, it would be interesting to observe whether reduced glycosylation further diminished potassium, sodium, and calcium current amplitude in ischemic excitable cells. In addition, ECGs should be evaluated in control and ST3Gal-IV(-/-) mice with an induced pathology, such as the ones mentioned above. Pharmacologic block or

PAGE 171

154 stimulus of vagal inputs in ECG-monitored control and ST3Gal-IV deficient mice may provide a clearer under standing of the impact of glycosylation on the beatto-beat control of the heart via the aut onomic (parasympathetic) nervous system. In humans, atrial fibrillation is prevalent in the elderly populati on; therefore, ECGs recorded from aged control and ST3Gal-IV(-/-) mice could be compared to determine whether the incidence of atrial fi brillation is increased with a deficiency in ST3Gal-IV. Telemetry allows for t he recording of not only cardiac conduction (ECGs), but also activity level and blood pr essure. It would be of interest to note whether either of thes e parameters is altered in healthy or diseased ST3Gal-IV(-/-) mice. Optical mapping of the my ocardium should be performed on ST3Gal-IV(-/-) and littermate control mice as well as those with induced pathologies to further elucid ate the impact of glycosylati on on cardiac excitability. Modeling software may provide some insight into the mechanisms contributing to the ECG and AP data and may prove useful in determining the next appropriate experimental step. Previously published reports have utilized a technique similar to that of Click chemistry to examine O-glycosylation in vivo236,237. In this way, specific azidemodified sugars can be incorporated in to endogenous glycosylation structures via tail vein injection. This would allow fo r detection of specific sugar residues on certain proteins. Thus, glycosylati on patterns throughout development or pathological progression could be observed on various glycoproteins, such as ion channels and extracellular matrix proteins The patterning could be examined in

PAGE 172

155 tissues other than the heart, for example in the brain, skeletal muscle, and/or liver. The ST3Gal-IV(-/-) and STX(-/-) mouse models also could be utilized to understand whether compensatory mechanisms exist to attach some sialic acid, albeit at a much lesser am ount, to certain proteins. Although treatment with O-glycosidases is not optimal, these experiments should be performed on Kv2.1, Kv4.2, and Kv4.3 to once again confirm that Oglycosylation exerts an effect on channe l activation. Software capable of predicting potential O-glycosylat ion sites and the current Kv channel crystal structure could be utilized to limit the number of potent ial O-glycosylation sites. Based upon these data, mutagenesis of the potential sites could be performed to allow for confirmation of t he data presented in this study (via western blots and electrophysiologic studies). These experim ents would allow for a more complete appreciation of how O-glycosylation and si alylation are regulated (remodeled) under pathological and norma l developmental states. Summary Here, we show that aberrant glycosylat ion alters cardiac excitability, with changes in sialylation primarily respons ible for the observed effects. ECGs suggested a trend toward increased arrhythmias in ST3Gal-IV(-/-) mice; broadened P waves were exhibited as well. Consistently, atrial APs and Kv currents showed glycosylation-dependent changes in duration and gating, respectively; ventricular APs and Kv currents were not impacted by changes in

PAGE 173

156 glycosylation. Expression of Ito and IK,slow generating channels in differentially sialylating cell lines showed depolarizing shifts in activation with reduced Nand O-glycosylation, supporting our result s in the atrial myocytes. Changes in glycosylation have been shown to affect multiple systems, including the cardiovascular system. The ubiquit ous and complex nature of glycosylation makes it a viable modulator of protein function. Glycosylation enhances ion channel activity, including that of heavily glycosylated Kv channels. Reduced glycosylation, specifically sialylation, of atrial Kv currents leads to depolarizing shifts in activation, promoting arrhythmic ity (at least partially) by lengthening the AP. As glycosylation a nd/or expression of Kv channels is remodeled throughout development or with the progression of pathologies, changes in the functional effect of Kv currents may occur. Thus, the exci tability of cardiac, neuronal, and skeletal muscle tissues may be modulat ed. Changing protein glycosylation structures has been largely overlooked as treatment for pathologies such as atrial fibrillation. However, if by alte ring the amount or type of glycans, one might enhance or diminish protein activity level and, thus, ameliorate the symptoms of associated disorders. Unfortunately, th is therapy may only be effective if localized or given to patients with global glycosylation disorders, as unsolicited side effects may result due to t he abundance of cell surface glycans.

PAGE 174

157 REFERENCES 1. Rosamond W, Flegal K, Friday G, Furie K, Go A, Gr eenlund K, Haase N, Ho M, Howard V, Kissela B, Kitt ner S, Lloyd-Jones D, McDermott M, Meigs J, Moy C, Nichol G, O'Donnell CJ, Roger V, Rumsfeld J, Sorlie P, Steinberger J, Thom T, Wasserthi el-Smoller S, Hong Y 2007. Heart disease and stroke statistics--2007 update: a report fr om the American Heart Association Statistics Co mmittee and Stroke Statistics Subcommittee. Circulation 115:e69-171. 2. Brundel BJ, Henning RH, Kampi nga HH, van G, I, Crijns HJ 2002. Molecular mechanisms of remodeling in human atrial fibrillation. Cardiovasc Res 54:315-324. 3. Huang B, El-Sherif T, Gidh-Jain M, Qin D, El-S herif N 2001. Alterations of sodium channel kinetics and gene expression in the postinfarction remodeled myocardium. J Cardiova sc Electrophysiol 12:218-225. 4. Huang B, Qin D, El-Sherif N 2000. Early down-regulation of K+ channel genes and currents in the pos tinfarction heart. J Cardiovasc Electrophysiol 11:1252-1261. 5. Huang B, Qin D, Deng L, Boutjdir M, Sherif N 2000. Reexpression of Ttype Ca2+ channel gene and current in post-infarction remodeled rat left ventricle. Cardiovasc Res 46:442-449. 6. Li GR, Lau CP, Ducharme A, Tard if JC, Nattel S 2002. Transmural action potential and ionic current remodeling in ventricles of failing canine hearts. Am J Physiol Heart Circ Physiol 283:H1031-H1041. 7. Schoonderwoerd BA, van G, I, Van Veldhuisen DJ, van den Berg MP, Crijns HJ 2005. Electrical and structur al remodeling: role in the genesis and maintenance of atrial fibrillat ion. Prog Cardiovasc Dis 48:153-168. 8. Tomaselli GF, Rose J 2000. Molecu lar aspects of arrhythmias associated with cardiomyopathies. Curr Opin Cardiol 15:202-208. 9. Van Wagoner DR, Pond AL, Lam orgese M, Rossie SS, McCarthy PM, Nerbonne JM 1999. Atrial L-type Ca2+ currents and human atrial fibrillation. Circ Res 85:428-436.

PAGE 175

158 10. Van Wagoner DR, Pond AL, McCa rthy PM, Trimmer JS, Nerbonne JM 1997. Outward K+ current densities and Kv1.5 expression are reduced in chronic human atrial fibril lation. Circ Res 80:772-781. 11. Van Wagoner DR, Nerbonne JM 2000. Molecular basis of electrical remodeling in atrial fibrillation. J Mol Cell Cardiol 32:1101-1117. 12. Mukherjee R, Hewett KW, Walk er JD, Basler CG, Spinale FG 1998. Changes in L-type calcium channel abundance and function during the transition to pacing-induced congesti ve heart failure. Cardiovasc Res 37:432-444. 13. Nerbonne JM, Kass RS 2005. Molecular physiology of cardiac repolarization. Ph ysiol Rev 85:1205-1253. 14. Berne RM, Levy MN, Koeppen BM, Stanton BA 2004. Physiology, Fifth ed., St. Louis: Mosby-Elsevier, Inc. 15. Nerbonne JM 2000. Molecular bas is of functional voltage-gated K+ channel diversity in the mammalian my ocardium. J Physiol 525 Pt 2:285298. 16. Pond AL, Nerbonne JM 2001. ERG proteins and functional cardiac I(Kr) channels in rat, mouse, and human heart. Trends Cardiovasc Med 11:286-294. 17. Cerbai E, Barbieri M, Mugelli A 1996. Occurrence and properties of the hyperpolarization-activated current If in ventricular myocytes from normotensive and hypertensive rats during aging. Circulation 94:16741681. 18. Zhang JF, Robinson RB, Siege lbaum SA 1992. Sympathetic neurons mediate developmental change in card iac sodium channel gating through long-term neurotransmitter action. Neuron 9:97-103. 19. Plotnikov AN, Sosunov EA, Patberg KW, Anyukhovsky EP, Gainullin RZ, Shlapakova IN, Krishnamurthy G, Danilo P, Jr., Rosen MR 2004. Cardiac memory evolves with age in associati on with development of the transient outward current. Circulation 110:489-495. 20. Sanders P, Morton JB, Kistler PM Spence SJ, Davidson NC, Hussin A, Vohra JK, Sparks PB, Kalman JM 2004. Electrophysiological and electroanatomic characterization of the atria in sinus node disease: evidence of diffuse atrial rem odeling. Circulation 109:1514-1522. 21. Brunet S, Aimond F, Li H, Guo W, Eldstrom J, Fedida D, Yamada KA, Nerbonne JM 2004. Heterogeneous expr ession of repolarizing, voltagegated K+ currents in adult mouse v entricles. J Physiol 559:103-120.

PAGE 176

159 22. Harrell MD, Harbi S, Hoffman JF Zavadil J, Coet zee WA 2007. Largescale analysis of ion channel gene ex pression in the mouse heart during perinatal development. Ph ysiol Genomics 28:273-283. 23. Wu C, Hayama E, Imamur a S, Matsuoka R, Nakanishi T 2007. Developmental changes in the expr ession of voltage-gated potassium channels in the ductus arte riosus of the fetal rat. Heart Vessels 22:34-40. 24. Fry M 2006. Developmental expressi on of Na+ currents in mouse Purkinje neurons. Eur J Neurosci 24:2557-2566. 25. Bacharova L 2007. The struct ural and electrical remodeling of myocardium in LVH and its impact on the QRS voltage. Anadolu Kardiyol Derg 7 Suppl 1:37-42. 26. Benitah JP, Gomez AM, Bailly P, Da Ponte JP, Berson G, Delgado C, Lorente P 1993. Heterogeneity of the early outward current in ventricular cells isolated from normal and hypertr ophied rat hearts. J Physiol 469:111138. 27. Hill JA 2003. Electrical re modeling in cardiac hypertrophy. Trends Cardiovasc Med 13:316-322. 28. Le BS, Demolombe S, Chambellan A, Bellocq C, Aimond F, Toumaniantz G, Lande G, Siavoshian S, Baro I, Pond AL, Nerbonne JM, Leger JJ, Escande D, Charpentier F 2003. Microa rray analysis reveals complex remodeling of cardiac ion channel expr ession with altered thyroid status: relation to cellular and integrated electrophysiology. Circ Res 92:234-242. 29. McIntosh MA, Cobbe SM, Kane KA, Rankin AC 1998. Action potential prolongation and potassium currents in le ft-ventricular myocytes isolated from hypertrophied rabbit hearts J Mol Cell Cardiol 30:43-53. 30. Tomita F, Bassett AL, Myerburg RJ, Kimura S 1994. Diminished transient outward currents in rat hypertrophi ed ventricular myocytes. Circ Res 75:296-303. 31. Charpentier F, Liu QY, Ros en MR, Robinson RB 1996. Age-related differences in beta-adrenergic regula tion of repolarization in canine epicardial myocytes. Am J Physiol 271:H1174-H1181. 32. Hewett KW, Rosen MR 1985. Devel opmental changes in the rabbit sinus node action potential and its res ponse to adrenergic agonists. J Pharmacol Exp Ther 235:308-312. 33. Huynh TV, Chen F, Wetzel GT, Friedman WF, Klitzner TS 1992. Developmental changes in membrane Ca2+ and K+ currents in fetal, neonatal, and adult rabbit ventricula r myocytes. Circ Res 70:508-515.

PAGE 177

160 34. Osaka T, Joyner RW 1992. Developmental changes in the betaadrenergic modulation of calcium current s in rabbit ventricular cells. Circ Res 70:104-115. 35. Osaka T, Joyner RW 1991. Devel opmental changes in calcium currents of rabbit ventricular cells. Circ Res 68:788-796. 36. Limmer D, O'Keefe MF 2005. Em ergency Care, 10th ed., Upper Saddle River: Prentice Hall. 37. Feinberg WM, Blackshear JL, L aupacis A, Kronmal R, Hart RG 1995. Prevalence, age distribution, and gender of patients with atri al fibrillation. Analysis and implications. Arch Intern Med 155:469-473. 38. Hordof AJ, Edie R, Malm JR, Hoffman BF, Rosen MR 1976. Electrophysiologic properties and response to pharmacologic agents of fibers from diseased human at ria. Circulation 54:774-779. 39. Boutjdir M, Le Heuzey JY, Lav ergne T, Chauvaud S, Guize L, Carpentier A, Peronneau P 1986. Inhom ogeneity of cellular refractoriness in human atrium: factor of arrhythmia? Pa cing Clin Electrophysiol 9:1095-1100. 40. Yue L, Feng J, Gaspo R, Li GR Wang Z, Nattel S 1997. Ionic remodeling underlying action potential changes in a c anine model of atrial fibrillation. Circ Res 81:512-525. 41. Bosch RF, Zeng X, Grammer JB, P opovic K, Mewis C, Kuhlkamp V 1999. Ionic mechanisms of electrical rem odeling in human atri al fibrillation. Cardiovasc Res 44:121-131. 42. Gaspo R, Bosch RF, Talajic M, Nattel S 1997. Functional mechanisms underlying tachycardia-induced sustained at rial fibrillation in a chronic dog model. Circulation 96:4027-4035. 43. Le Grand BL, Hatem S, Der oubaix E, Couetil JP, Coraboeuf E 1994. Depressed transient outward and calciu m currents in dilated human atria. Cardiovasc Res 28:548-556. 44. Yue L, Melnyk P, Gaspo R, Wang Z, Nattel S 1999. Molecular mechanisms underlying ionic remodel ing in a dog model of atrial fibrillation. Circ Res 84:776-784. 45. Feng J, Wible B, Li GR, Wang Z, Nattel S 1997. Antisense oligodeoxynucleotides directed against Kv1.5 mRNA specifically inhibit ultrarapid delayed rectifier K+ curr ent in cultured adult human atrial myocytes. Circ Res 80:572-579.

PAGE 178

161 46. Nattel S, Maguy A, Le BS, Y eh YH 2007. Arrhythmogenic ion-channel remodeling in the heart: heart failure, myocardial infarction, and atrial fibrillation. Physiol Rev 87:425-456. 47. Armoundas AA, Wu R, Juang G, Mar ban E, Tomaselli GF 2001. Electrical and structural remodeling of the fail ing ventricle. Pharmacol Ther 92:213230. 48. Schoen FJ 1999. Robbin's Pathologi c Basis of Disease, Philadelphia: WB Saunders Co. 49. Marionneau C, Brunet S, Flagg TP Pilgram TK, Demolombe S, Nerbonne JM 2008. Distinct cellular and molecular mechanisms underlie functional remodeling of repolarizing K+ current s with left ventricular hypertrophy. Circ Res 102:1406-1415. 50. Koumi S, Backer CL, Arentzen CE 1995. Characterization of inwardly rectifying K+ channel in human cardiac myocytes. Alterations in channel behavior in myocytes isolated from patients with idiopathic dilated cardiomyopathy. Circulation 92:164-174. 51. Nuss HB, Kaab S, Kass DA, Tomaselli GF, Marban E 1999. Cellular basis of ventricular arrhythmias and abnorma l automaticity in heart failure. Am J Physiol 277:H80-H91. 52. Thuringer D, Deroubaix E, Coul ombe A, Coraboeuf E, Mercadier JJ 1996. Ionic basis of the action potential prolongation in ventricular myocytes from Syrian hamsters with dilate d cardiomyopathy. Cardiovasc Res 31:747-757. 53. Knollmann BC, Knol lmann-Ritschel BE, Weissm an NJ, Jones LR, Morad M 2000. Remodelling of ionic currents in hypertrophied and failing hearts of transgenic mice overexpressing ca lsequestrin. J Physiol 525 Pt 2:483498. 54. Lederer WJ, Tsien RW 1976. Tr ansient inward current underlying arrhythmogenic effects of cardiotonic st eroids in Purkinje fibres. J Physiol 263:73-100. 55. Pogwizd SM, Schlotthauer K, Li L, Yuan W, Bers DM 2001. Arrhythmogenesis and contractile dysf unction in heart failure: Roles of sodium-calcium exchange, inward rect ifier potassium current, and residual beta-adrenergic responsivene ss. Circ Res 88:1159-1167. 56. Ma TS, Baker JC, Bailey LE 1979 Excitation-contraction coupling in normal and myopathic hamster hearts III: functional deficiencies in interstitial glycoproteins. Cardiovasc Res 13:568-577.

PAGE 179

162 57. Ufret-Vincenty CA, Baro DJ Lederer WJ, Rockman HA, Quinones LE, Santana LF 2001. Role of sodium channel deglycosylation in the genesis of cardiac arrhythmias in heart fa ilure. J Biol Chem 276:28197-28203. 58. Abriel H, Cabo C, Wehrens XH, Rivolta I, Motoike HK, Memmi M, Napolitano C, Priori SG, Kass RS 2001. Novel arrhythmogenic mechanism revealed by a long-QT syndrome mutation in the cardiac Na(+) channel. Circ Res 88:740-745. 59. Bennett PB, Yazawa K, Makita N, George AL, Jr. 1995. Molecular mechanism for an inherited cardiac arrhythmia. Nature 376:683-685. 60. Clancy CE, Kass RS 2005. Inherited and acquired vulnerability to ventricular arrhythmias: cardiac Na + and K+ channels. Physiol Rev 85:3347. 61. Kubisch C, Schroeder BC, Friedric h T, Lutjohann B, El-Amraoui A, Marlin S, Petit C, Jentsch TJ 1999. KC NQ4, a novel potassium channel expressed in sensory outer hair cells is mutated in dominant deafness. Cell 96:437-446. 62. Meadows LS, Malhotra J, Loukas A, Thyagarajan V, Kazen-Gillespie KA, Koopman MC, Kriegler S, Isom LL, Ragsdale DS 2002. Functional and biochemical analysis of a sodi um channel beta1 subunit mutation responsible for generalized epilepsy with febrile seizures plus type 1. J Neurosci 22:10699-10709. 63. Monaghan MM, Menegola M, Vacher H, Rhodes KJ, Trimmer JS 2008. Altered expression and localization of hippocampal A-type potassium channel subunits in the pilocarpineinduced model of temporal lobe epilepsy. Neuroscience 156:550-562. 64. Smart SL, Lopantsev V, Zhang CL Robbins CA, Wang H, Chiu SY, Schwartzkroin PA, Messing A, Tempel BL 1998. Deletion of the K(V)1.1 potassium channel causes epile psy in mice. Neuron 20:809-819. 65. Spampanato J, Escayg A, Meisler MH, Goldin AL 2003. Generalized epilepsy with febrile seizures plus type 2 mutation W1204R alters voltagedependent gating of Na(v)1.1 sodium channels. Neuroscience 116:37-48. 66. Wallace RH, Wang DW, Singh R, Sc heffer IE, George AL, Jr., Phillips HA, Saar K, Reis A, Johnson EW, Sutherland GR, Berkovic SF, Mulley JC 1998. Febrile seizures and genera lized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat Genet 19:366-370.

PAGE 180

163 67. Nerbonne JM, Nichols CG, Schw arz TL, Escande D 2001. Genetic manipulation of cardiac K(+) channel function in mice: what have we learned, and where do we go from here? Circ Res 89:944-956. 68. Moss AJ, Kass RS 2005. Long QT syndrome: from channels to cardiac arrhythmias. J Clin Invest 115:2018-2024. 69. Roden DM, Balser JR 1999. A pl ethora of mechanisms in the HERGrelated long QT syndrome. Genetics meets electrophysiology. Cardiovasc Res 44:242-246. 70. Splawski I, Shen J, Timothy KW Lehmann MH, Priori S, Robinson JL, Moss AJ, Schwartz PJ, Towbin JA, Vincent GM, Keating MT 2000. Spectrum of mutations in longQT syndrome genes. KVLQT1, HERG, SCN5A, KCNE1, and KCNE2. Circulation 102:1178-1185. 71. Roden DM, Viswanathan PC 200 5. Genetics of acquired long QT syndrome. J Clin Invest 115:2025-2032. 72. Brugada J, Brugada R, Brugada P 1998. Right bundle-branch block and ST-segment elevation in leads V1 th rough V3: a marker for sudden death in patients without demonstrable stru ctural heart disease. Circulation 97:457-460. 73. Chen Q, Kirsch GE, Zhang D, Br ugada R, Brugada J, Brugada P, Potenza D, Moya A, Borggrefe M, Breit hardt G, Ortiz-Lopez R, Wang Z, Antzelevitch C, O'Brien RE, Schulz e-Bahr E, Keating MT, Towbin JA, Wang Q 1998. Genetic basis and molecular mechanism for idiopathic ventricular fibrillation. Nature 392:293-296. 74. Grant AO, Carboni MP, Neplioue va V, Starmer CF, Memmi M, Napolitano C, Priori S 2002. Long QT syndrom e, Brugada syndrome, and conduction system disease are linked to a single sodium channel mutation. J Clin Invest 110:1201-1209. 75. Hille B 2001. Ion Channels of Excitable Membranes, Third ed., Sunderland: Sinauer Associates, Inc. 76. Hagiwara S, Yoshida S, Yosh ii M 1981. Transient and delayed potassium currents in the egg cell me mbrane of the coelenterat e, Renilla koellikeri. J Physiol 318:123-141. 77. Schoppa NE, McCormack K, Tanouye MA, Sigworth FJ 1992. The size of gating charge in wild-type and mut ant Shaker potassium channels. Science 255:1712-1715.

PAGE 181

164 78. Seoh SA, Sigg D, Papazian DM Bezanilla F 1996. Voltage-sensing residues in the S2 and S4 segments of the Shaker K+ channel. Neuron 16:1159-1167. 79. Aggarwal SK, MacKinnon R 1996. Contribution of t he S4 segment to gating charge in the Shaker K+ channel. Neuron 16:1169-1177. 80. Jiang Y, Ruta V, Chen J, Lee A, MacKinnon R 2003. The principle of gating charge movement in a vo ltage-dependent K+ channel. Nature 423:42-48. 81. Purves D, Augustine GJ, Fitzpatrick D, Hall WC, LaMantia A-S, McNamara JO, White LE 2008. Neur oscience, Fourth ed., Sunderland: Sinauer Associates, Inc. 82. Yang N, Horn R 1995. Evidence for voltage-dependent S4 movement in sodium channels. Neuron 15:213-218. 83. Jiang Y, Lee A, Chen J, Ruta V, Cadene M, Chait BT, MacKinnon R 2003. X-ray structure of a voltage-dependent K+ channel. Nature 423:33-41. 84. Long SB, Campbell EB, MacKi nnon R 2005. Crystal structure of a mammalian voltage-dependent Shaker family K+ channel. Science 309:897-903. 85. Long SB, Tao X, Campbell EB, MacK innon R 2007. Atomic structure of a voltage-dependent K+ channel in a lip id membrane-like environment. Nature 450:376-382. 86. Long SB, Campbell EB, MacKinnon R 2005. Voltage sensor of Kv1.2: structural basis of electromechani cal coupling. Science 309:903-908. 87. Ahern CA, Horn R 2004. Specificit y of charge-carrying residues in the voltage sensor of potassium c hannels. J Gen Physiol 123:205-216. 88. Campos FV, Chanda B, Roux B, Be zanilla F 2007. Two atomic constraints unambiguously position the S4 segment relative to S1 and S2 segments in the closed state of Shaker K channel. Proc Natl Acad Sci U S A 104:79047909. 89. Chanda B, Asamoah OK, Blunck R, Roux B, Bezanilla F 2005. Gating charge displacement in voltage-ga ted ion channels involves limited transmembrane movement. Nature 436:852-856. 90. Cohen BE, Grabe M, Jan LY 2003 Answers and questions from the KvAP structures. Neuron 39:395-400.

PAGE 182

165 91. Posson DJ, Ge P, Miller C, Bezanilla F, Selvin PR 2005. Small vertical movement of a K+ channel voltag e sensor measured with luminescence energy transfer. Nature 436:848-851. 92. Richardson J, Blunck R, Ge P, Selvin PR, Bezanilla F, Papazian DM, Correa AM 2006. Distance measurements reveal a common topology of prokaryotic voltage-gated ion channels in the lipid bilayer. Proc Natl Acad Sci U S A 103:15865-15870. 93. Swartz KJ 2004. Towards a stru ctural view of gating in potassium channels. Nat Rev Neurosci 5:905-916. 94. Bendahhou S, O'Reilly AO, Duclohier H 2007. Role of hydrophobic residues in the voltage sensors of the voltage-gated sodium channel. Biochim Biophys Acta 1768:1440-1447. 95. Elinder F, Arhem P, Larsson HP 2001. Localizatio n of the extracellular end of the voltage sensor S4 in a pot assium channel. Biophys J 80:1802-1809. 96. Elinder F, Nilsson J, Arhem P 2007. On the openi ng of voltage-gated ion channels. Physiol Behav 92:1-7. 97. Lu Z, Klem AM, Ramu Y 2001. I on conduction pore is conserved among potassium channels. Nature 413:809-813. 98. Heginbotham L, MacKinnon R 19 92. The aromatic binding site for tetraethylammonium ion on potassi um channels. Neuron 8:483-491. 99. Heginbotham L, Lu Z, Abramson T, MacKinnon R 1994. Mutations in the K+ channel signature sequence. Biophys J 66:1061-1067. 100. Heginbotham L, Abra mson T, MacKinnon R 1992. A functional connection between the pores of distant ly related ion channels as revealed by mutant K+ channels. Science 258:1152-1155. 101. HODGKIN AL, Keynes RD 1955. T he potassium permeability of a giant nerve fibre. J Physiol 128:61-88. 102. Hille B, Schwarz W 1978. Potassi um channels as multi-ion single-file pores. J Gen Physiol 72:409-442. 103. Armstrong CM 1971. Interaction of tetraethylammonium ion derivatives with the potassium channels of gian t axons. J Gen Physiol 58:413-437. 104. Armstrong CM 1969. Inactivati on of the potassium conductance and related phenomena caused by quaternar y ammonium ion injection in squid axons. J Gen Physiol 54:553-575.

PAGE 183

166 105. Armstrong CM, BINSTOCK L 1965. ANOMALOUS RECTIFICATION IN THE SQUID GIANT AXON INJECTED WITH TETRAETHYLAMMONIUM CHLORIDE. J Gen Physiol 48:859-872. 106. Pelhate M, Pichon Y 1974. Proceedings: Selective inhibition of potassium current in the giant axon of the cockroach. J Physiol 242:90P-91P. 107. HODGKIN AL, Huxley AF, Katz B 1949. Ionic currents underlying activity in the giant axon of the squi d. Arch Sci Physiol 3:129-150. 108. Chandy KG 1991. Simplified gene nomenclature. Nature 352:26. 109. Li H, Guo W, Yamada KA, Ner bonne JM 2004. Selective elimination of I(K,slow1) in mouse ventricular my ocytes expressing a dominant negative Kv1.5alpha subunit. Am J Physiol H eart Circ Physiol 286:H319-H328. 110. Xu H, Guo W, Ner bonne JM 1999. Four kinetically distinct depolarizationactivated K+ currents in adult mouse ventricular myocytes. J Gen Physiol 113:661-678. 111. Xu H, Barry DM, Li H, Brunet S, Guo W, Nerbonne JM 1999. Attenuation of the slow component of delayed rectification, action potential prolongation, and triggered activity in mice expressing a dominantnegative Kv2 alpha subunit. Circ Res 85:623-633. 112. Zhou J, Jeron A, London B, Han X, Koren G 1998. Characterization of a slowly inactivating outward current in adult mouse ventricular myocytes. Circ Res 83:806-814. 113. Zhou J, Kodirov S, Murata M, Buckett PD, Nerbonne JM, Koren G 2003. Regional upregulation of Kv2.1-encoded current, IK,slow2, in Kv1DN mice is abolished by crossbreeding with Kv2DN mice. Am J Physiol Heart Circ Physiol 284:H491-H500. 114. Kodirov SA, Brunner M, Nerbonne JM, Buckett P, Mitchell GF, Koren G 2004. Attenuation of I(K,slow1) and I(K,slow2) in Kv1/Kv2DN mice prolongs APD and QT intervals but does not suppress spontaneous or inducible arrhythmias. Am J Physiol Heart Circ Physiol 286:H368-H374. 115. Nerbonne JM 2004. Studying card iac arrhythmias in the mouse--a reasonable model for probing mec hanisms? Trends Cardiovasc Med 14:83-93. 116. Boyle WA, Nerbonne JM 1991. A novel type of depolarization-activated K+ current in isolated adult rat atrial myocytes. Am J Physiol 260:H1236H1247.

PAGE 184

167 117. Sanguinetti MC, Bennett PB 2003. An tiarrhythmic drug target choices and screening. Circ Res 93:491-499. 118. Wang Z, Fermini B, Nattel S 1994. Rapid and slow components of delayed rectifier current in human atrial my ocytes. Cardiovasc Res 28:1540-1546. 119. Wang Z, Fermini B, Nattel S 1993. Delayed rectifier outward current and repolarization in human atrial myocytes. Circ Res 73:276-285. 120. Yue L, Feng J, Li GR, Nattel S 1996. Characterization of an ultrarapid delayed rectifier potassium channel invo lved in canine atrial repolarization. J Physiol 496 ( Pt 3):647-662. 121. Boyle WA, Nerbonne JM 1992. Two f unctionally distinct 4-aminopyridinesensitive outward K+ currents in ra t atrial myocytes. J Gen Physiol 100:1041-1067. 122. Wang Z, Fermini B, Nattel S 1993. Sustained depolarization-induced outward current in human atrial myo cytes. Evidence for a novel delayed rectifier K+ current similar to Kv 1.5 cloned channel currents. Circ Res 73:1061-1076. 123. Sanguinetti MC, Jurkiewicz NK 1991. Delayed rectifier outward K+ current is composed of two currents in gui nea pig atrial cells. Am J Physiol 260:H393-H399. 124. Sanguinetti MC, Jurkiewicz NK 1992. Role of external Ca2+ and K+ in gating of cardiac delayed rectifier K+ currents. Pfluger s Arch 420:180-186. 125. Xu H, Guo W, Nerbonne JM 1999. F our kinetically distinct depolarizationactivated K+ currents in adult mouse ventricular myocytes. J Gen Physiol 113:661-678. 126. Barry DM, Nerbonne JM 1996. Myocardial potassium channels: electrophysiological and molecular di versity. Annu Rev Physiol 58:363394. 127. Xu H, Guo W, Nerbonne JM 1999. F our kinetically distinct depolarizationactivated K+ currents in adult mouse ventricular myocytes. J Gen Physiol 113:661-678. 128. Guo W, Xu H, London B, Nerbonne JM 1999. Molecular basis of transient outward K+ current diversity in mous e ventricular myocytes. J Physiol 521 Pt 3:587-599. 129. Hagiwara S, KUSANO K, Sait o N 1961. Membrane changes of Onchidium nerve cell in potassium-rich media. J Physiol 155:470-489.

PAGE 185

168 130. Nakajima S 1966. Analysis of K inactivation and TEA action in the supramedullary cells of puffe r. J Gen Physi ol 49:629-640. 131. Neher E 1971. Two fast trans ient current components during voltage clamp on snail neurons. J Gen Physiol 58:36-53. 132. Connor JA, Stevens CF 1971. Vo ltage clamp studies of a transient outward membrane current in gastropod neural somata. J Physiol 213:2130. 133. Connor JA, Stevens CF 1971. Inward and delayed outward membrane currents in isolated neural somata under voltage clamp. J Physiol 213:119. 134. Rasmusson RL, Morales MJ, Wang S, Liu S, Campbell DL, Brahmajothi MV, Strauss HC 1998. Inactivation of voltage-gated cardiac K+ channels. Circ Res 82:739-750. 135. MacKinnon R, Aldrich RW, Lee AW 1993. Functional stoichiometry of Shaker potassium channel inac tivation. Science 262:757-759. 136. London B, Guo W, Pan X, Lee JS, Shusterman V, Rocco CJ, Logothetis DA, Nerbonne JM, Hill JA 2001. Targeted replacement of KV1.5 in the mouse leads to loss of the 4-am inopyridine-sensitive component of I(K,slow) and resistance to drug-i nduced qt prolongation. Circ Res 88:940946. 137. Brunner M, Guo W, Mitchell GF Buckett PD, Nerbonne JM, Koren G 2001. Characterization of mice with a combined suppression of I(to) and I(K,slow). Am J Physiol Hear t Circ Physiol 281:H1201-H1209. 138. Xu H, Guo W, Nerbonne JM 1999. F our kinetically distinct depolarizationactivated K+ currents in adult mouse ventricular myocytes. J Gen Physiol 113:661-678. 139. Zhou J, Jeron A, London B, Han X, Koren G 1998. Characterization of a slowly inactivating outward current in adult mouse ventricular myocytes. Circ Res 83:806-814. 140. Fiset C, Clark RB, Larsen TS Giles WR 1997. A rapidly activating sustained K+ current modul ates repolarization and excitation-contraction coupling in adult mouse ventricle. J Physiol 504 ( Pt 3):557-563. 141. Brouillette J, Clark RB, Giles WR Fiset C 2004. Functional properties of K+ currents in adult mouse ventricula r myocytes. J Physiol 559:777-798.

PAGE 186

169 142. Guo W, Li H, Aimond F, Johns DC, Rhodes KJ, Trimme r JS, Nerbonne JM 2002. Role of heteromultimers in t he generation of myocardial transient outward K+ currents. Circ Res 90:586-593. 143. Niwa N, Wang W, Sha Q, Marionneau C, Nerbonne JM 2008. Kv4.3 is not required for the generation of functi onal Ito,f channels in adult mouse ventricles. J Mol Cell Cardiol 44:95-104. 144. Kong W, Po S, Yamagishi T, Ashen MD, Stetten G, Tomaselli GF 1998. Isolation and characterization of the human gene encoding Ito: further diversity by alternative mRNA spli cing. Am J Physiol 275:H1963-H1970. 145. Dixon JE, Shi W, Wang HS, McDona ld C, Yu H, Wymore RS, Cohen IS, McKinnon D 1996. Role of t he Kv4.3 K+ channel in ventricular muscle. A molecular correlate for the transient outward current. Circ Res 79:659-668. 146. Guo W, Xu H, London B, Nerbonne JM 1999. Molecular basis of transient outward K+ current diversity in mous e ventricular myocytes. J Physiol 521 Pt 3:587-599. 147. Brahmajothi MV, Campbell DL, Rasmusson RL, Morales MJ, Trimmer JS, Nerbonne JM, Strauss HC 1999. Distinc t transient outward potassium current (Ito) phenotypes and distributi on of fast-inactivating potassium channel alpha subunits in ferret left ventricular myocytes. J Gen Physiol 113:581-600. 148. Wettwer E, Amos G, Gath J, Ze rkowski HR, Reidemeister JC, Ravens U 1993. Transient outward current in human and rat ventricular myocytes. Cardiovasc Res 27:1662-1669. 149. Wettwer E, Amos GJ, Posival H, Ravens U 1994. Transient outward current in human ventricular myocytes of subepicardial and subendocardial origin. Circ Res 75:473-482. 150. Campbell DL, Qu Y, Rasmusso n RL, Strauss HC 1993. The calciumindependent transient outward potassium current in isolated ferret right ventricular myocytes. II. Closed stat e reverse use-dependent block by 4aminopyridine. J Gen Physiol 101:603-626. 151. Furukawa T, Myerburg RJ, Furu kawa N, Bassett AL, Kimura S 1990. Differences in transient outward currents of feline endocardial and epicardial myocytes. Circ Res 67:1287-1291. 152. Litovsky SH, Antzelevitch C 1988. Tr ansient outward current prominent in canine ventricular epicardium but not endocardium. Circ Res 62:116-126.

PAGE 187

170 153. Miller J, Agnew W, Levinson S 1 983. Principal glycopeptide of the tetrodotoxin/saxitoxin binding prot ein from Electrophorus electricus: isolation and partial chemical and physi cal characterization. Biochemistry 22:462-470. 154. Rehm H 1989. Enzymatic deglyco sylation of the dendrotoxin-binding protein. FEBS Lett 247:28-30. 155. Rogero O, Tejedor FJ 1995. Im munochemical characterization and developmental expression of Shak er potassium channels from the nervous system of Drosophila. J Biol Chem 270:25746-25751. 156. Santacruz-Toloza L, H uang Y, John SA, Papazian DM 1994. Glycosylation of shaker potassium c hannel protein in insect cell culture and in Xenopus oocytes. Biochemistry 33:5607-5613. 157. Schmidt JW, Catterall WA 1987. Palm itylation, sulfati on, and glycosylation of the alpha subunit of the sodium channel. Role of post-translational modifications in channel assembly. J Biol Chem 262:13713-13723. 158. Jindal HK, Folco EJ, Liu GX, Kor en G 2008. Posttranslational modification of voltage-dependent potassium c hannel Kv1.5: COOH-terminal palmitoylation modulates its biological properties. Am J Physiol Heart Circ Physiol 294:H2012-H2021. 159. Barchi RL 1988. Probing the molecu lar structure of t he voltage-dependent sodium channel. Annu Re v Neurosci 11:455-495. 160. Catterall WA 1988. Structure and function of volt age-sensitive ion channels. Science 242:50-61. 161. Rehm H, Newitt RA, Tempel BL 1989. Immunological evidence for a relationship between the dendrotoxin-b inding protein and the mammalian homologue of the Drosophila Shaker K+ channel. FEBS Lett 249:224-228. 162. Scott VE, Parcej DN, Keen JN, Findlay JB, Dolly JO 1990. Alphadendrotoxin acceptor from bovine brai n is a K+ channel protein. Evidence from the N-terminal sequence of it s larger subunit. J Biol Chem 265:20094-20097. 163. Thornhill WB, Wu MB, Jiang X, Wu X, Morgan PT, Margiotta JF 1996. Expression of Kv1.1 delayed rectifie r potassium channels in Lec mutant Chinese hamster ovary cell lines reveal s a role for sialidation in channel function. J Biol Chem 271:19093-19098. 164. Trimmer JS, Agnew WS 1989. Molecu lar diversity of vo ltage-sensitive Na channels. Annu Rev Physiol 51:401-418.

PAGE 188

171 165. Jaeken J 2003. Congenital disorders of glycosylation (CDG): It's all in it! Journal of Inherited Metabolic Disease 26:99-118. 166. Helenius A, Aebi M 2001. Intracellular functions of N-linked glycans. Science 291:2364-2369. 167. Ohtsubo K, Marth JD 2006. Glycosylat ion in cellular mechanisms of health and disease. Cell 126:855-867. 168. Kornfeld R, Kornfeld S 1985. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 54:631-664. 169. Geetha-Habib M, Noiva R, Kapl an HA, Lennarz WJ 1988. Glycosylation site binding protein, a component of oligosaccharyl transferase, is highly similar to three other 57 kd luminal proteins of the ER. Cell 54:1053-1060. 170. Jenkins N, Parekh RB, James DC 1996. Getting the glycosylation right: implications for the biotechnology i ndustry. Nat Biotechnol 14:975-981. 171. Freeze HH, Aebi M 2005. Altered gl ycan structures: the molecular basis of congenital disorders of glycosylation. Curr Opin Struct Biol 15:490-498. 172. Spiro MJ, Spiro RG 1991. Potential regul ation of N-glycosylation precursor through oligosaccharide-lipid hydrol ase action and glucosyltransferaseglucosidase shuttle. J Biol Chem 266:5311-5317. 173. Orlean P 1992. Enzymes that rec ognize dolichols participate in three glycosylation pathways and are required for protein secretion. Biochem Cell Biol 70:438-447. 174. Hansen JE, Lund O, Tolstrup N, G ooley AA, Williams KL, Brunak S 1998. NetOglyc: prediction of mucin type O-glycosylation sites based on sequence context and surface accessibi lity. Glycoconj J 15:115-130. 175. Van den SP, Rudd PM, Dwek RA, Opdenakker G 1998. Concepts and principles of O-linked glycosylation. Crit Rev Biochem Mol Biol 33:151208. 176. O'Connell BC, Hagen FK, Tabak LA 1992. The influence of flanking sequence on the O-glycosylation of thr eonine in vitro. J Biol Chem 267:25010-25018. 177. Gooley AA, Williams KL 1994. Towards characterizing O-glycans: the relative merits of in vivo and in vi tro approaches in seeking peptide motifs specifying O-glycosylation si tes. Glycobiology 4:413-417.

PAGE 189

172 178. Chou KC, Zhang CT, Kezdy FJ, P oorman RA 1995. A vector projection method for predicting the specificit y of GalNAc-transferase. Proteins 21:118-126. 179. Elhammer AP, Poorman RA, Brow n E, Maggiora LL, Hoogerheide JG, Kezdy FJ 1993. The specificity of UDP-GalNAc:polypeptide Nacetylgalactosaminyltransferase as in ferred from a database of in vivo substrates and from the in vitro glycosylation of proteins and peptides. J Biol Chem 268:10029-10038. 180. Hill HD, Jr., Schwyzer M, Steinm an HM, Hill RL 1977. Ovine submaxillary mucin. Primary structure and peptide substrates of UDP-Nacetylgalactosamine:mucin transfera se. J Biol Chem 252:3799-3804. 181. Wang Y, Agrwal N, Eckhardt AE, Stevens RD, Hill RL 1993. The acceptor substrate specificity of porcine s ubmaxillary UDP-GalNAc:polypeptide Nacetylgalactosaminyltransferase is dependent on the amino acid sequences adjacent to serine and threonine residues. J Biol Chem 268:22979-22983. 182. Wilson IB, Gavel Y, von HG 1991. Amino acid distributions around Olinked glycosylation sites. Bi ochem J 275 ( Pt 2):529-534. 183. Nehrke K, ten Hagen KG, Hagen FK Tabak LA 1997. Charge distribution of flanking amino acids inhibits Oglycosylation of several single-site acceptors in vivo. Glycobiology 7:1053-1060. 184. Hanisch FG 2001. O-glycosylation of the mucin type. Biol Chem 382:143149. 185. Hang HC, Yu C, Kato DL, Be rtozzi CR 2003. A metabolic labeling approach toward proteomic analysis of mucin-type O-linked glycosylation. Proc Natl Acad Sci U S A 100:14846-14851. 186. ten Hagen KG, Tran DT, Gerken TA, Stein DS, Zhang Z 2003. Functional characterization and expression a nalysis of members of the UDPGalNAc:polypeptide N-acetylgalactosa minyltransferase family from Drosophila melanogaster. J Biol Chem 278:35039-35048. 187. Tian E, ten Hagen KG 2008. Recent insights into the biological roles of mucin-type O-glycosylation. Glycoconj J. 188. Comelli EM, Head SR, Gilmartin T, Whisenant T, Haslam SM, North SJ, Wong NK, Kudo T, Narimatsu H, Esko JD, Drickamer K, Dell A, Paulson JC 2006. A focused microarray ap proach to functional glycomics: transcriptional regulation of the glycome. Glycobiology 16:117-131.

PAGE 190

173 189. Montpetit ML, Stocker PJ, Schw etz TA, Harper JM, Norring SA, Schaffer L, North SJ, Jang-Lee J, Gilmartin T, Head SR, Haslam SM, Dell A, Marth JD, Bennett ES 2009. Regulated Glycosylation Modulates Cardiac Electrical Signaling., S ubmitted for publication. 190. Ishii A, Ikeda T, Hitoshi S, Fujimoto I, Torii T, Sakuma K, Nakakita S, Hase S, Ikenaka K 2007. Developmental changes in the expression of glycogenes and the content of N-glycans in the mouse cerebral cortex. Glycobiology 17:261-276. 191. Jaeken J, Matthijs G 2007. Congenital disorders of glycosylation: a rapidly expanding disease fam ily. Annu Rev Genomics Hu m Genet 8:261-278. 192. Martinez-Duncker I, Dupre T, Piller V, Piller F, Candelier JJ, Trichet C, Tchernia G, Oriol R, Mollicone R 2005. Genetic complementation reveals a novel human congenital disorder of glycosylation of type II, due to inactivation of the Golgi CMP-sial ic acid transporter. Blood 105:26712676. 193. Kanagawa M, Saito F, Kunz S, Yoshida-Moriguchi T, Barresi R, Kobayashi YM, Muschler J, Dumanski JP, Mic hele DE, Oldstone MB, Campbell KP 2004. Molecular recognition by LARGE is essential for expression of functional d ystroglycan. Cell 117:953-964. 194. Leroy JG, Seppala R, Huizing M, Dacremont G, De SH, Van Coster RN, Orvisky E, Krasnewich DM, Gahl WA 2001. Dominant inheritance of sialuria, an inborn error of feedback i nhibition. Am J Hum Genet 68:14191427. 195. Fontaine G, Biserte G, Montreuil J, Dupont A, Farriaux JP 1968. [Sialuria: an original metabolic disorder]. Helv Paediatr ActaSuppl-32. 196. Thomas GH, Reynolds LW, Miller CS 1985. Overproduction of Nacetylneuraminic acid (sialic acid) by sialuria fibroblasts. Pediatr Res 19:451-455. 197. Topaz O, Shurman DL, Bergman R, Indelman M, Ratajcza k P, Mizrachi M, Khamaysi Z, Behar D, Petronius D, Friedman V, Zelikovic I, Raimer S, Metzker A, Richard G, Sprecher E 2004. Mutations in GALNT3, encoding a protein involved in O-linked glyco sylation, cause familial tumoral calcinosis. Nat Genet 36:579-581. 198. Ichikawa S, Guigonis V, Imel EA, Courouble M, Heissat S, Henley JD, Sorenson AH, Petit B, Lienhardt A, Econs MJ 2007. Novel GALNT3 mutations causing hyperostosis-hy perphosphatemia syndrome result in low intact fibroblast growth factor 23 concentrations. J Clin Endocrinol Metab 92:1943-1947.

PAGE 191

174 199. Barbieri AM, Filopanti M, Bua G, Beck-Peccoz P 2007. Two novel nonsense mutations in GALNT3 gene are responsible for familial tumoral calcinosis. J Hum Genet 52:464-468. 200. Hiki Y, Kokubo T, Iwase H, Masaki Y, Sano T, Tanaka A, Toma K, Hotta K, Kobayashi Y 1999. Underglycosylati on of IgA1 hinge plays a certain role for its glomerular deposition in IgA nephropathy. J Am Soc Nephrol 10:760-769. 201. Hiki Y, Odani H, Takahashi M, Ya suda Y, Nishimoto A, Iwase H, Shinzato T, Kobayashi Y, Maeda K 2001. Ma ss spectrometry proves under-Oglycosylation of glomerular IgA1 in IgA nephropathy. Kidney Int 59:10771085. 202. Tomana M, Novak J, Julian BA, Matousovic K, Konecny K, Mestecky J 1999. Circulating immune complexes in IgA nephropathy consist of IgA1 with galactose-deficient hinge regi on and antiglycan antibodies. J Clin Invest 104:73-81. 203. Coppo R, Amore A 2004. Aberr ant glycosylation in IgA nephropathy (IgAN). Kidney Int 65:1544-1547. 204. Hiki Y, Tanaka A, Kokubo T, Iwas e H, Nishikido J, Hotta K, Kobayashi Y 1998. Analyses of IgA1 hinge glycopepti des in IgA nephropathy by matrixassisted laser desorption/ionization time -of-flight mass spectrometry. J Am Soc Nephrol 9:577-582. 205. Allen AC, Bailey EM, Brenchley PE, Buck KS, Barratt J, Feehally J 2001. Mesangial IgA1 in IgA nephropathy ex hibits aberrant O-glycosylation: observations in three pati ents. Kidney Int 60:969-973. 206. Bern C, Montgomery SP, Herwaldt BL, Rassi A, Jr., Marin-Neto JA, Dantas RO, Maguire JH, Acquatella H, Morillo C, Kirchhoff LV, Gilman RH, Reyes PA, Salvatella R, Moore AC 2007. Evaluation and treatment of chagas disease in the United Stat es: a systematic review. JAMA 298:2171-2181. 207. Teixeira AR, Nitz N, Guimaro MC, Gomes C, Santos-Buch CA 2006. Chagas disease. Postgrad Med J 82:788-798. 208. Colli W 1993. Trans-sialidase: a unique enzyme activity discovered in the protozoan Trypanosoma cruz i. FASEB J 7:1257-1264. 209. Bennett E, Urcan MS, Tinkle SS, Koszowski AG, Levinson SR 1997. Contribution of sialic acid to the voltage dependence of sodium channel gating. A possible electrostatic mec hanism. J Gen Physiol 109:327-343.

PAGE 192

175 210. Stuhmer W, Ruppersberg JP, Schr oter KH, Sakmann B, Stocker M, Giese KP, Perschke A, Baumann A, Pongs O 1989. Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J 8:3235-3244. 211. Shi G, Trimmer JS 1999. Different ial asparagine-linked glycosylation of voltage-gated K+ channels in mammalian brain and in transfected cells. J Membr Biol 168:265-273. 212. Frankenhaeuser B, HODGKIN AL 1957. The ac tion of calcium on the electrical properties of squi d axons. J Physiol 137:218-244. 213. Hille B, Ritchie JM, Strichartz GR 1975. The effect of surface charge on the nerve membrane on the action of tetrodotoxin and saxitoxin in frog myelinated nerve. J Physiol 250:34P-35P. 214. Watanabe I, Wang HG, Sutachan JJ, Z hu J, Recio-Pinto E, Thornhill WB 2003. Glycosylation affects rat Kv1. 1 potassium channel gating by a combined surface potential and co operative subunit interaction mechanism. J Physiol 550:51-66. 215. Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB 2007. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated ac tion potentials. Brain Res 1144:1-18. 216. Johnson D, Bennett ES 2008. Gating of the shaker potassium channel is modulated differentially by N-glycosylation and sialic acids. Pflugers Arch 456:393-405. 217. Ufret-Vincenty CA, Baro DJ, Sant ana LF 2001. Differential contribution of sialic acid to the function of repol arizing K(+) currents in ventricular myocytes. Am J Physiol Cell Physiol 281:C464-C474. 218. Stocker PJ, Bennett ES 2006. Differ ential sialylation modulates voltagegated Na+ channel gating throughout the developing myocardium. J Gen Physiol 127:253-265. 219. Ellies LG, Ditto D, Levy GG, Wa hrenbrock M, Ginsburg D, Varki A, Le DT, Marth JD 2002. Sialyltransferase ST3Gal-IV operates as a dominant modifier of hemostasis by conceali ng asialoglycoprotein receptor ligands. Proc Natl Acad Sci U S A 99:10042-10047. 220. Bennett ES 2002. Isoform-specific effects of sialic acid on voltagedependent Na+ channel gating: functional sialic acids are localized to the S5-S6 loop of domain I. J Physiol 538:675-690.

PAGE 193

176 221. Johnson D, Montpetit ML, Stocke r PJ, Bennett ES 2004. The sialic acid component of the beta1 subunit modulat es voltage-gated sodium channel function. J Biol Chem 279:44303-44310. 222. Johnson D, Bennett ES 2006. Isofo rm-specific effects of the beta2 subunit on voltage-gated sodium channel gating. J Biol Chem 281:25875-25881. 223. Stanley P, Caillibot V, Siminovitch L 1975. Sele ction and characterization of eight phenotypically distinct lines of lectin-resistant Chinese hamster ovary cell. Cell 6:121-128. 224. Angata K, Long JM, Bukalo O, Lee W, Dityatev A, Wynshaw-Boris A, Schachner M, Fukuda M, Marth JD 2004. Sialyltransferase ST8Sia-II assembles a subset of polysialic ac id that directs hippocampal axonal targeting and promotes fear behav ior. J Biol Chem 279:32603-32613. 225. Isenberg G, Klockner U 1982. Calc ium tolerant ventricular myocytes prepared by preincubation in a "KB medium". Pflugers Arch 395:6-18. 226. Deutscher SL, Nuwayhid N, St anley P, Briles EI, Hirschberg CB 1984. Translocation across Golgi vesicle membranes: a CHO glycosylation mutant deficient in CMP-sialic acid transport. Cell 39:295-299. 227. Stanley P 1989. Chinese hamst er ovary cell mutants with multiple glycosylation defects for production of glycoproteins with minimal carbohydrate heterogeneity. Mo l Cell Biol 9:377-383. 228. Stanley P 1985. Membrane mutants of animal cells: rapid identification of those with a primary defect in glyco sylation. Mol Cell Biol 5:923-929. 229. Barresi R, Michele D, Kanagawa M, Harper H, Dovico S, Satz J, Moore S, Zhang W, Schachter H, Dumanski J, Cohn R, Nishino I, Campbell K 4 A.D. LARGE can functionally bypass -dystroglycan glycosylation defects in distinct congenital muscula r dystrophies. Nat Med 10:696-703. 230. Dobyns W, Pagon R, Armstrong D, Curry C, Greenberg F, Grix A, Holmes L, Laxova R, Michaels V, Robinow M, et al 1989. Diagnostic criteria for Walker-Warburg syndrome. Am J Med G enet 32:195-210. 231. Jaeken J 2003. Congenital disorders of glycosylation (CDG): It's all in it! Journal of Inherited Metabolic Disease 26:99-118. 232. Trepanier-Boulay V, Lupien MA, St -Michel C, Fiset C 2004. Postnatal development of atrial repolarization in the mouse. Cardiovasc Res 64:8493.

PAGE 194

177 233. Fedida D, Eldstrom J, Hesketh JC, Lamorgese M, Castel L, Steele DF, Van Wagoner DR 2003. Kv 1.5 is an important co mponent of repolarizing K+ current in canine atrial myocytes. Circ Res 93:744-751. 234. Dube DH, Prescher JA, Quang CN, Bertozzi CR 2006. Probing mucin-type O-linked glycosylation in living anim als. Proc Natl Acad Sci U S A 103:4819-4824. 235. Laughlin ST, Agard NJ, Baskin JM, Carrico IS, Chang PV, Ganguli AS, Hangauer MJ, Lo A, Prescher JA, Berto zzi CR 2006. Metabolic labeling of glycans with azido sugars for visua lization and glycopr oteomics. Methods Enzymol 415:230-250. 236. Laughlin ST, Baskin JM, Amacher SL Bertozzi CR 2008. In vivo imaging of membrane-associated glycans in developing zebrafish. Science 320:664-667. 237. Prescher JA, Dube DH, Bertozzi CR 2004. Chemical remodelling of cell surfaces in living animals. Nature 430:873-877. 238. Saxon E, Bertozzi CR 2000. Ce ll surface engineering by a modified Staudinger reaction. Science 287:2007-2010. 239. Saxon E, Luchansky SJ, Hang HC, Yu C, Lee SC, Bertozzi CR 2002. Investigating cellular metabolism of synthetic azi dosugars with the Staudinger ligation. J Am Chem Soc 124:14893-14902. 240. Han P, Lucero MT 2006. Pituitar y adenylate cyclase activating polypeptide reduces expression of Kv1.4 and Kv4. 2 subunits underlying A-type K(+) current in adult mouse olfactor y neuroepithelia. Neuroscience 138:411419. 241. Trepanier-Boulay V, Lupien MA, St -Michel C, Fiset C 2004. Postnatal development of atrial repolarizatio n in the mouse. Cardiovasc Res 64:8493. 242. Sasaki K, Watanabe E, Kawashim a K, Sekine S, Dohi T, Oshima M, Hanai N, Nishi T, Hasegawa M 1993. Expression cloning of a novel Gal beta (1-3/1-4) GlcNAc alpha 2,3-sialyl transferase using lectin resistance selection. J Biol Chem 268:22782-22787. 243. Kono M, Ohyama Y, Lee YC, Ha mamoto T, Kojima N, Tsuji S 1997. Mouse beta-galactoside alpha 2,3-sialyl transferases: comparison of in vitro substrate specificities and tiss ue specific expression. Glycobiology 7:469-479.

PAGE 195

178 244. Varki A, Cummings RD, Esko JD Freeze HH, Stanley P, Bertozzi CR, Hart GW, Etzler ME 2009. Essentials of Glycobiology, Second ed., Cold Spring Harbor: Cold Spring Harbor Laboratory Press. 245. Odening KE, Nerbonne JM, Bode C, Zehender M, Brunner M 2009. In vivo effect of a dominant n egative Kv4.2 loss-of-function mutation eliminating I(to,f) on atrial refractoriness and atri al fibrillation in mice. Circ J 73:461467. 246. Jan LY, Jan YN 1997. Cloned pot assium channels from eukaryotes and prokaryotes. Annu Rev Neurosci 20:91-123. 247. Coetzee WA, Amarillo Y, Chiu J, Chow A, Lau D, McCormack T, Moreno H, Nadal MS, Ozaita A, Pountney D, Saganich M, Vega-Saenz de ME, Rudy B 1999. Molecular diversity of K+ channels. Ann N Y Acad Sci 868:233-285. 248. Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB 2007. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated ac tion potentials. Brain Res 1144:1-18. 249. Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB 2007. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated ac tion potentials. Brain Res 1144:1-18. 250. Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB 2007. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated ac tion potentials. Brain Res 1144:1-18. 251. Feng J, Wible B, Li GR, Wang Z, Nattel S 1997. Antisense oligodeoxynucleotides directed against Kv1.5 mRNA specifically inhibit ultrarapid delayed rectifier K+ curr ent in cultured adult human atrial myocytes. Circ Res 80:572-579. 252. Fedida D, Eldstrom J, Hesketh JC, Lamorgese M, Castel L, Steele DF, Van Wagoner DR 2003. Kv 1.5 is an important co mponent of repolarizing K+ current in canine atrial myocytes. Circ Res 93:744-751. 253. Kues WA, Wunder F 1992. He terogeneous Expression Patterns of Mammalian Potassium Channel Genes in Developing and Adult Rat Brain. Eur J Neurosci 4:1296-1308. 254. Varki A 1992. Diversity in t he sialic acids. Glycobiology 2:25-40. 255. Guo W, Xu H, London B, Nerbonne JM 1999. Molecular basis of transient outward K+ current diversity in mous e ventricular myocytes. J Physiol 521 Pt 3:587-599.

PAGE 196

179 256. Maletic-Savatic M, Lenn NJ, Trimme r JS 1995. Differential spatiotemporal expression of K+ channel polypep tides in rat hippocampal neurons developing in situ and in vitro. J Neurosci 15:3840-3851. 257. Lujan R, de Cabo de la Vega, Domi nguez del TE, Ballesta JJ, Criado M, Juiz JM 2003. Immunohistochemical localization of the voltage-gated potassium channel subunit Kv1.4 in t he central nervous system of the adult rat. J Chem Neuroanat 26:209-224. 258. Roberds SL, Tamkun MM 1991. Cloning and tissue-specific expression of five voltage-gated potassium channel cD NAs expressed in rat heart. Proc Natl Acad Sci U S A 88:1798-1802. 259. Stuhmer W, Ruppersberg JP, Schr oter KH, Sakmann B, Stocker M, Giese KP, Perschke A, Baumann A, Pongs O 1989. Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J 8:3235-3244. 260. Stuhmer W, Ruppersberg JP, Schr oter KH, Sakmann B, Stocker M, Giese KP, Perschke A, Baumann A, Pongs O 1989. Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J 8:3235-3244. 261. Stuhmer W, Ruppersberg JP, Schr oter KH, Sakmann B, Stocker M, Giese KP, Perschke A, Baumann A, Pongs O 1989. Molecular basis of functional diversity of voltage-gated potassium channels in mammalian brain. EMBO J 8:3235-3244. 262. Santacruz-Toloza L, H uang Y, John SA, Papazian DM 1994. Glycosylation of shaker potassium c hannel protein in insect cell culture and in Xenopus oocytes. Biochemistry 33:5607-5613. 263. Schwarz TL, Tempel BL, Papazian DM, Jan YN, Jan LY 1988. Multiple potassium-channel components are produced by alternative splicing at the Shaker locus in Drosophila. Nature 331:137-142. 264. Zhu J, Watanabe I, Poholek A, Koss M, Gomez B, Yan C, Recio-Pinto E, Thornhill WB 2003. Allowed N-glycosyl ation sites on the Kv1.2 potassium channel S1-S2 linker: implications fo r linker secondary structure and the glycosylation effect on channel f unction. Biochem J 375:769-775. 265. Lis H, Sharon N 1998. Lectins: Ca rbohydrate-Specific Proteins That Mediate Cellular Recogniti on. Chem Rev 98:637-674. 266. Hammarstrom S, Murphy LA, Gold stein IJ, Etzler ME 1977. Carbohydrate binding specificity of four N-acetyl -D-galactosamine"specific" lectins: Helix pomatia A hemagglutinin, soy bean agglutinin, lima bean lectin, and Dolichos biflorus lectin. Biochemistry 16:2750-2755.

PAGE 197

180 267. Tornoe CW, Christensen C, Mel dal M 2002. Peptidotriazoles on solid phase: [1,2,3]-triazoles by regiospec ific copper(i)-catalyzed 1,3-dipolar cycloadditions of terminal alkynes to azides. J Org Chem 67:3057-3064. 268. Prescher JA, Bertozzi CR 2005. Chemistry in living systems. Nat Chem Biol 1:13-21. 269. Much B, Wahl-Schott C, Zong X, Schneider A, Baum ann L, Moosmang S, Ludwig A, Biel M 2003. Role of subunit heteromerization and N-linked glycosylation in the formation of f unctional hyperpolariz ation-activated cyclic nucleotide-gated channel s. J Biol Chem 278:43781-43786. 270. Liang CS 2007. Cardiac sympathetic nerve terminal function in congestive heart failure. Acta Pharmacol Sin 28:921-927. 271. Chidsey CA, Braunwald E 1966. Sy mpathetic activity and neurotransmitter depletion in congestive heart failu re. Pharmacol Rev 18:685-700. 272. Bohm M, La RK, Schwinger RH, Erdmann E 1995. Evidence for reduction of norepinephrine uptake sites in the failing human heart. J Am Coll Cardiol 25:146-153. 273. Trepanier-Boulay V, Lupien MA, St -Michel C, Fiset C 2004. Postnatal development of atrial repolarization in the mouse. Cardiovasc Res 64:8493. 274. Leroy JG 2006. Congenital disorder s of N-glycosylation including diseases associated with Oas well as N-glyco sylation defects. Pediatr Res 60:643656. 275. Schenkman S, Eichinger D, Pere ira ME, Nussenzweig V 1994. Structural and functional properties of Trypanos oma trans-sialidase. Annu Rev Microbiol 48:499-523. 276. Bustamante JM, Rivarola HW, Fret es R, Paglini-Oliva PA 2005. Weekly electrocardiographic pattern in mice infected with two different Trypanosoma cruzi strains. Int J Cardiol 102:211-217. 277. Tanowitz HB, Machado FS, Jelicks LA, Shirani J, de Carvalho AC, Spray DC, Factor SM, Kirchhoff LV, Weiss LM 2009. Perspectives on Trypanosoma cruzi-induced heart disease (Chagas disease). Prog Cardiovasc Dis 51:524-539. 278. Martin PT 2007. Congenital mu scular dystrophies involving the Omannose pathway. Curr Mol Med 7:417-425. 279. Ervasti JM, Campbell KP 1991. Memb rane organization of the dystrophinglycoprotein complex. Cell 66:1121-1131.

PAGE 198

181 280. Kato K, Jeanneau C, Tarp MA, et -Pages A, Lorenz-De piereux B, Bennett EP, Mandel U, Strom TM, Claus en H 2006. Polypeptide GalNActransferase T3 and familial tumoral calcinosis. Secretion of fibroblast growth factor 23 requires O-glycosylat ion. J Biol Chem 281:18370-18377. 281. Hakomori S 1996. Tumor malignancy defined by aberrant glycosylation and sphingo(glyco)lipid metabo lism. Cancer Res 56:5309-5318. 282. Granovsky M, Fata J, Pawling J, Muller WJ, Khokha R, Dennis JW 2000. Suppression of tumor grow th and metastasis in M gat5-deficient mice. Nat Med 6:306-312. 283. Balbao CE, de Paola AA, Fenelon G 2009. Effects of alcohol on atrial fibrillation: myths and truths. T her Adv Cardiovasc Dis 3:53-63. 284. Correale M, Laonigro I, Altomare E, Di BM 2009. [Alcohol-induced cardiac disease]. G Ital Cardiol (Rome) 10:18-27. 285. Romppanen J, Punnonen K, Anttila P, Jakobsson T, Blake J, Niemela O 2002. Serum sialic acid as a marker of alcohol consumption: effect of liver disease and heavy drinking. Alco hol Clin Exp Res 26:1234-1238. 286. Hale EA, Raza SK, Ciecierski R G, Ghosh P 1998. Deleterious actions of chronic ethanol treatment on the glycosylation of rat brain clusterin. Brain Res 785:158-166. 287. Malagolini N, Dall'Olio F, Serafini-Cessi F, Cessi C 1989. Effect of acute and chronic ethanol administration on ra t liver alpha 2,6-sialyltransferase activity responsible for sialylation of serum transferrin. Alcohol Clin Exp Res 13:649-653. 288. Millar JS 2001. The sialylation of plasma lipoproteins. Atherosclerosis 154:1-13. 289. Tertov VV, Kaplun VV, Sobenin IA Boytsova EY, Bovin NV, Orekhov AN 2001. Human plasma trans-sialidase causes atherogenic modification of low density lipoprotein. At herosclerosis 159:103-115. 290. Sutachan JJ, Watanabe I, Zhu J, Go ttschalk A, Recio-Pinto E, Thornhill WB 2005. Effects of Kv1.1 channel glycosylation on C-type inactivation and simulated action potentia ls. Brain Res 1058:30-43. 291. Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB 2007. The glycosylation state of Kv1.2 potassium channels affects trafficking, gating, and simulated ac tion potentials. Brain Res 1144:1-18. 292. Brugada J, Brugada R, Brugada P 2007. Channelopathies: a new category of diseases causi ng sudden death. Herz 32:185-191.

PAGE 199

182 293. Petrecca K, Atanasiu R, Akhavan A, Shrier A 1999. Nlinked glycosylation sites determine HERG channel surfac e membrane expression. J Physiol 515 ( Pt 1):41-48. 294. Takumi T, Ohkubo H, Nakanish i S 1988. Cloning of a membrane protein that induces a slow voltage-gated potassium current. Science 242:10421045. 295. Blumenthal EM, Kaczmarek LK 1992. Structure and regulation of the MinK potassium channel. Neur ochem Res 17:869-876. 296. Abbott GW, Sesti F, Splawski I, Buck ME, Lehmann MH, Timothy KW, Keating MT, Goldstein SA 1999. MiRP1 forms IKr potassium channels with HERG and is associated with card iac arrhythmia. Cell 97:175-187. 297. Sanguinetti MC, Bennett PB 2003. An tiarrhythmic drug target choices and screening. Circ Res 93:491-499. 298. Nishiyama A, Ishii DN, Backx PH, Pulford BE, Birks BR, Tamkun MM 2001. Altered K(+) channel gene expr ession in diabetic rat ventricle: isoform switching between Kv4.2 and Kv1.4. Am J Physiol Heart Circ Physiol 281:H1800-H1807. 299. Qin D, Huang B, Deng L, El-Adaw i H, Ganguly K, Sowers JR, El-Sherif N 2001. Downregulation of K(+) channel ge nes expression in type I diabetic cardiomyopathy. Biochem Biophys Res Commun 283:549-553. 300. Colbert CM, Pan E 1999. Arachid onic acid reciprocally alters the availability of transient and sust ained dendritic K(+) channels in hippocampal CA1 pyramidal neurons. J Neurosci 19:8163-8171. 301. Du J, Haak LL, Phillips-Tansey E, Russell JT, McBain CJ 2000. Frequency-dependent regulati on of rat hippocampal somato-dendritic excitability by the K+ channel subunit Kv2.1. J Physiol 522 Pt 1:19-31. 302. Macdonald PE, Salapatek AM, W heeler MB 2003. Te mperature and redox state dependence of native Kv2.1 current s in rat pancreatic beta-cells. J Physiol 546:647-653. 303. Rodriguez BM, Sigg D, Bezanill a F 1998. Voltage gating of Shaker K+ channels. The effect of temperature on ionic and gating currents. J Gen Physiol 112:223-242. 304. Du Z, Chaoqian X, Shan H, Lu Y, Ren N 2007. Functional impairment of cardiac transient outward K+ current as a result of abnormally altered cellular environment. Clin Exp P harmacol Physiol 34:148-152.

PAGE 200

ABOUT THE AUTHOR Tara Ashley (Munn) Schwetz was born Ma y 4, 1983 in Lakeland, Florida. She graduated from George W. Jenkins Senior High School in 2001. Tara enrolled at Florida State University in the Honor s Liberal Studies program, where she received an ACS James R. Fisher Fellowship and an Honors in the Major Thesis Grant, among others. At FSU, she worked in the laboratory of Timothy Logan, Ph.D. and served as a teaching assistant. In 2005, she graduated Magna Cum Laude with Honors with a B.S. in Biochem istry. That year, Tara joined the laboratory of Eric Bennett, Ph .D. at the University of South Florida. She was awarded an AHA Pre-doctoral Fellowshi p and a Genshaft Family Doctoral Fellowship while at USF. She earned her M.S.M.S. and was accepted as a doctoral candidate in 2007. Tara was confe rred with a Ph.D. in Medical Sciences from the University of South Florida, College of Medicine in 2009.