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 nam 2200469Ka 4500
controlfield tag 001 002058228
008 100201s2008 xx a bm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002303
b M799f 2008
Montpetit, Marty L.
Functional remodeling of the cardiac glycome throughout the developing myocardium /
by Marty L. Montpetit.
x, 140 leaves :
ill. (some col.) ;
Also available online.
Dissertation (Ph.D.)--University of South Florida, 2008.
Includes bibliographical references (leaves 121-140).
ABSTRACT: Cell surfaces are replete with complex, biologically important glycans responsible for multiple cellular functions including cell adhesion and cellular communication. Proper protein glycosylation is essential for normal development and often pathologies are marked by altered glycosylation. Here, data showed that the auxillary subunit, [beta]1, modified voltage-gated Na channel (Na[subscript]v) gating in an isoform-specific, sialic acid dependent, and saturating manner. The regulated activity of the hundreds of glycogenes (glycosylation-associated genes) is responsible for protein glycosylation; this could result in a glycome of thousands of glycan structures. Microarray analyses indicated that glycogene expression was highly regulated throughout the heart during development. Specifically, >59% of glycogenes were significantly differentially expressed among neonatal and adult atrial and ventricular myocyQuantitative-PCR of individual genes confirmed the microarray analyses. Such substantial regulation of glycogene expression likely results in changes in glycan structures attached to cell surface proteins. To confirm this, myocyte glycan profiles were determined and compared among neonatal and adult atria and ventricles using mass spectrometry. The data predicted marked differences in glycan structures among myocyte types, indicating that the glycome is remodeled throughout the heart during development. To address the question of whether the remodeled glycome can impact cardiac function, action potentials and Na[subscript]v activity were measured and compared under conditions in which glycogene expression was regulated. That is, atrial and ventricular myocytes were isolated from control mice and from mice in which the polysialyltransferase, STX, was knocked out. STX is expressed in the neonatal atria, and is essentially absent in neonatal ventricle.Action potential waveforms and Na[subscript]v activity measured in atrial myocytes were impacted by STX expression. No changes in ventricular action potential waveform or in Na[subscript]v activity were observed; as expected since STX is not expressed in the ventricle. The magnitude of the atrial action potential and the rate of depolarization were decreased in the absence of STX. Further, Na[subscript]v gating was shifted consistently in the depolarized direction in STX knockout atrial myocytes. Together, these data indicate that the glycome is tightly controlled and regulated in the heart, and proper glycosylation is essential for normal myocyte function.
Advisor: Eric S. Bennett Ph.D.
Ion Channel Gating.
t USF Electronic Theses and Dissertations.
Functional Remodeling of the Cardiac Glycome Throughout the Developing Myocardium by Marty L. Montpetit A dissertation submitted in partial fulfillment of the requirement s of the degree of Doctor of Philosophy Department of Physiology and Biophysics College of Medicine University of South Florida Major Professor: Eric S. Bennett, Ph.D. Jahanshah Amin, Ph.D. Craig A. Doupnik, Ph.D. Bruce G. Lindsey, Ph.D. Huntington Potter, Ph.D. E. Truitt Sutton, Ph.D. Date of Approval: March 14, 2008 Keywords: glycosylation, cardiac, devel opment, excitability, sodium channel Copyright 2008, Marty L. Montpetit
DEDICATION In loving memory of my father, Louis Montpetit.
ACKNOWLEDGEMENTS This dissertation would not have been comp leted without the love and support of my family. I would like to give special thanks to my wife, Alison, for her support throughout this process and helping me to see the bigger picture of life. I would also like to thank my par ents, the late Louis Montpet it and Mary Lou and Joseph Budek, and my sister, Marla, for thei r love and continued encouragement I extend the deepest thanks to Eric S. Benne tt, Ph.D. for all the effort he has put forth to help me succeed. I could not hav e asked for a more supportive mentor. I appreciate the time invested and insight provided by my committee members: Jahanshah Amin, Ph.D., Craig Doupnik, Ph .D., Bruce Lindsey, Ph.D., Huntington Potter, Ph.D., and E. Tr uitt Sutton, Ph.D. I would like to thank all those who hav e worked with me in the lab: Jeanie Harper, Patrick Stocker, Ph.D., Dani el Johnson, Ph.D., Sarah Norring, Tara Schwetz and Barrett McCormick as well as those with whom I have collaborated: Steven R. Head, Stuart M. Haslam, Ph.D ., Timothy Gilmartin, Lana Schaffer, Simon J. North, Ph.D., Jihye JangLee, Ph.D. and Jamey D. Marth, Ph.D. Finally, I would like to extend my grat itude to my friends and family: Gerard Gole, Rickamer Hoover, Kari Bruursema, Ph.D., Iwona Misiuta, Ph.D., Kathleen and Tom Kelly, Kyle and Kristin Crawford, Andy and Kathryn Ross, and Ben and Holly Rapin.
i TABLE OF CONTENTS LIST OF FIGURES iv LIST OF TABLES vii ABSTRACT viii CHAPTER 1 INTRODUCTION 1 Cardiac remodeling impacts cardiac function 1 Regulation of glycan biosynthesis is essential for normal physiology 5 Ion transport is the basis fo r cellular communication 14 Cardiac contraction is the result of orchestrated ion channel function 15 The structure of Nav dictates channel function 17 Post-translational modifications may alter ion channel function 21 CHAPTER 2 MATERIALS AND METHODS 25 Chinese Hamster Ovary (CHO) Ce ll Culture and Transfection 25 Vector Construction and Mutagenesis 25 Electrophysiology and Data Analysis 26 Sodium Current Recordings 26 Pulse Protocols 28 Conductance-Voltage (G-V) Relationship 28 Steady-State Inactivation Curves (hinf) 28 Recovery from Inactivation 29 Measurement of Inacti vation gating kinetics 29
ii Neonatal and Adult Cardiac Tissue Isolation 30 mRNA Isolation 30 Microarray 31 Microarray Analysis 31 Quantitative PCR 33 Glycan Screening 34 Glycan Isolation 34 Glycan derivatization 35 Mass Spectrometry 36 Cardiomyocyte Isolation for Electrophysiology 36 Cardiomyocyte Electrophysiology 37 Sodium Current Recordings 37 Action Potential Recordings 37 Transgenic mouse 38 Data Analysis 38 CHAPTER 3 THE 1 SUBUNIT MODULATES Nav GATING IN AN 39 ISOFORM-SPECIFIC, SIALIC ACID-DEPENDENT MANNER Discussion 51 CHAPTER 4 GLYCOGENE EXPRES SION IS REGULATED 54 THROUGHOUT THE DEVELOPING MYOCARDIUM Core Structures 61 Terminal Structures 69 Glycan Degradation 69
iii Nucleotide Sugar Synthesis and Transporters 70 Tissue Type Comparison 70 Chamber-Specific Regulation 75 Neonatal Atria and Ventricle 75 Adult Atria and Ventricle 77 Developmental Regulation 77 Adult and Neonatal Atria 77 Adult and Neonatal Ventricle 78 Quantitative PCR verifies microarray data 78 Discussion 80 CHAPTER 5 THE GLYCOME IS REMODELED THROUGHOUT 81 THE HEART DURING DEVELOPMENT High Mannose Structures 81 Complex Structures 93 Chamber-specific glycan profile changes 93 Neonatal Atria and Ventricle 93 Adult Atria and Ventricle 94 Developmental glycan profile changes 95 Neonatal and Adult Atria 95 Neonatal and Adult Ventricle 96 Discussion 97 CHAPTER 6 THE REGULATED EXPRESSION OF A SINGLE 99 POLYSIALYLTRANSFERASE IMPACTS CARDIAC EXCITABILITY
iv The neonatal atrial action potential waveform is altered when 100 STX is absent The voltage dependence of Nav gating changes only in the 102 neonatal atria of the STX knockout Discussion 108 CHAPTER 7 FINAL DISCUSSION 111 Significance of this study 118 REFERENCES 121 ABOUT THE AUTHOR End Page
v LIST OF FIGURES Figure 1.1 Ionic basis of the cardiac action potential 2 Figure 1.2 Action potential waveforms throughout the heart 3 Figure 1.3 Roles of glycans in cellular functions 6 Figure 1.4 Overview of N-glycan biosynthesis 8 Figure 1.5 Regulation of glycan expression 11 Figure 1.6 Schematic of a ty pical cardiac action potential 16 Figure 1.7 Schematic of the volt age-gated sodium channel structure 18 Figure 1.8 Two competing theories for voltage-gated ion channel gating 20 Figure 3.1 Alpha and 1 subunit sialic acids modify channel activation 42 in an subunit dependent manner Figure 3.2 Alpha and 1 subunit sialic acids modify channel 43 inactivation in an subunit dependent manner Figure 3.3 Alpha and 1 subunit sialic acids alter channel fast 44 inactivation rates in an subunit dependent manner Figure 3.4 Alpha and 1 subunit sialic acids modify channel 45 recovery from inactivation in an subunit dependent manner Figure 3.5 1 subunit sialic acids modify channel gating parameters 47 in a saturating manner Figure 3.6 The impact 1 has on Nav gating is likely through 48 electrostatic interaction Figure 3.7 N-glycans are completely responsible for 1 effects on 50 Nav gating Figure 3.8 Model proposing the satu rating effects of sialic acids on 52 Nav gating Figure 4.1 Comparison of glycogene expression among samples 56
vi Figure 4.2A Glycosyltransferase expr ession throughout the developing 57 myocardium Figure 4.2B Glycan degradase expre ssion throughout the developing 58 myocardium Figure 4.2C Nucleotide sugar synthes is and transporter gene expression 59 throughout the developing myocardium Figure 4.3 The three basic glycosylation structures 62 Figure 4.4A Glycosylatransferase s differentially expressed 71 throughout the developing myocardium Figure 4.4B Glycan degradses diffe rentially expressed throughout 72 the developing myocardium Figure 4.4C Nucleotide sugar synthesis and transporters differentially 73 expressed throughout the developing myocardium Figure 4.5 Differential expressi on of glycogenes by category 74 Figure 4.6 qPCR validates t he GeneChip microarray data 79 Figure 5.1A The population of N-glycans is different among the four 82 myocyte types Figure 5.1B Identified low mass Nglycan structures and their 83 relative mass Figure 5.1C Identified high mass Nglycan structures and their 84 relative mass Figure 5.2A Mass spectra of t he neonatal atrial N-glycans 85 Figure 5.2B Mass spectra of the neonatal ventricular N-glycans 86 Figure 5.2C Mass spectra of the adult atrial N-glycans 87 Figure 5.2D Mass spectra of t he adult ventricular N-glycans 88 Figure 5.3 Mass spectra of masses between 1500 and 2400 m/z 89 Figure 5.4 Mass spectra of masses between 2400 and 3050 m/z 90
vii Figure 5.5 Spectra of masses above 3050 m/z 91 Figure 6.1 Expression of STX modi fies neonatal atrial, but not 101 ventricular AP waveform Figure 6.2 STX causes a change in neonatal atrial Nav activation 103 voltage, but not in ventricular Nav activation Figure 6.3 STX causes a change in neonatal atrial Nav steady state 104 inactivation, but not in ventricular Nav steady state inactivation Figure 6.4 Absence of STX causes a slowing of the neonatal atrial Nav 105 inactivation rate, but has no effect on the kinetics of ventricular Nav inactivation Figure 6.5 Absence of STX increases t he rate of recovery from fast 106 inactivation for neonatal atrial Nav to rates similar to those measured for control and knockout ventricular Nav Figure 7.1 Model proposing glyc osylation-dependent control and 116 modulation of Nav gating
viii LIST OF TABLES Table 2.1 Breakdown of probesets on GLYCOv2 32 Table 3.1 The measured gating parameters for 1sialic acid 41 Table 4.1A mRNA levels encoding proteins involved in core 63 structure synthesis Table 4.1B mRNA levels encoding pr oteins involved in sialylation 64 Table 4.1C mRNA levels encoding pr oteins involved in sulfation 65 Table 4.1D mRNA levels encoding pr oteins involved in fucosylation 66 Table 4.1E mRNA levels encoding proteins involved in glycan 67 degradation Table 4.1F mRNA levels encoding pr oteins involved in nucleotide 68 sugar synthesis and transport Table 4.2 Differential glyc ogene expression profile 76 Table 5.1 Relative percentage of glycan structures defined 92 by either structure (h igh mannose) or mass Table 6.1 Measured action potential and Nav parameters 107
ix FUNCTIONAL REMODELING OF THE CARDIAC GLYCOME THROUGHOUT THE DEVE LOPING MYOCARDIUM Marty L. Montpetit ABSTRACT Cell surfaces are replete with complex, biologically important glycans responsible for multiple cellular functions including cell adhesion and cellu lar communication. Proper protein glycosylation is essent ial for normal development and often pathologies are marked by altered glyco sylation. Here, data showed that the auxillary subunit, 1, modified voltage-gated Na+ channel (Nav) gating in an isoform-specific, sialic acid dependen t, and saturating manner. The regulated activity of the hundreds of glycogenes (glycosylation-associated genes) is responsible for protein glycosyl ation; this could result in a glycome of thousands of glycan structures. Microarray analyse s indicated that glycogene expression was highly regulated throughout the heart during development. Specifically, >59% of glycogenes were significantly differentially expressed among neonatal and adult atrial and ventricular myocytes. Quantitative-PCR of individual genes confirmed the microarray analyses. Such substantial regul ation of glycogene expression likely results in changes in gl ycan structures attached to cell surface proteins. To confirm this, myocyte glycan profiles were determined and compared among neonatal and adult at ria and ventricles using mass spectrometry. The data predicted marked differences in glycan structures among myocyte types, indicating that the glyco me is remodeled throughout the heart
x during development. To address the questi on of whether the remodeled glycome can impact cardiac function, action potentials and Nav activity were measured and compared under conditions in which glycogene expression was regulated. That is, atrial and ventricular myocytes were isolated from control mice and from mice in which the polysialyltransferase, STX, was knocked out. STX is expressed in the neonatal atria, and is e ssentially absent in neonatal ventricle. Action potential waveforms and Nav activity measured in atrial myocytes were impacted by STX expression. No c hanges in ventricular action potential waveform or in Nav activity were observed; as expected since STX is not expressed in the ventricle. The magnit ude of the atrial action potential and the rate of depolarization were decreased in the absence of STX. Further, Nav gating was shifted consistently in the depol arized direction in STX knockout atrial myocytes. Together, these data indicate that the glycome is tightly controlled and regulated in the heart, and proper glycosylation is essential for normal myocyte function.
1 CHAPTER 1 INTRODUCTION Heart disease is the leading c ause of death among U.S. citizens1. Accounting for over 27% of deaths in 2004, heart di sease caused 100,000 more deaths than any other cause1. The cardiac action potential is formed through the coordinated gating of voltage-gated ion channels. Conduction of the action potential throughout the heart leads to cardiac contra ction. Slight alterations in ion channel function (likely through cardiac re modeling) are associated with many cardiac maladies including heart failure, my ocardial infarction and hypertension. Although significant work has been devoted to understand variations in cardiac waveform, this is the first to descr ibe how a remodeled glycome could impact cardiac excitability. Cardiac remodeling impacts cardiac function Changing the expression of proteins and therefore the cellular processes in which they are involved is termed remodeling. Cellular remodeling is characteristic of normal development2-20 and pathologies21-52. Cardiac remodeling is evident in development as the human prenatal heart rate is commonly over 150 beats per minute which slows to an average of 72 beats per minute in the adult. Sympathetic inner vation and ion channel remodeling are considered responsible fo r the changing heart rate53-57. In the adult mouse
2 Figure 1.1. Ionic basis of the cardiac action potential. Figure 1.1. Schematic of human action potential waveforms in atria (blue) and ventricle (red) and the major ion current s creating these waveforms. Purple indicates the current is involved in bot h cell types. Known or presumed channels are noted to the right of each current. Figure from Pond and Nerbonne, 200158.
3 Figure 1.2. Action potential waveforms throughout the heart. Figure 1.2. Examples of the variou s action potential waveforms throughout the cardiac conduction and contractile system s. Differential expression of ion channel subunits is assumed to be responsible for these changes in action potential waveform. Action potentials are disp laced in time to re flect the temporal sequence of propagation through the heart. SA, sino-atrial ; AV, atrio-ventricular; RV, right ventricle; LV, left vent ricle. Figure from Nerbonne, 200059.
4 myocardium, ion channel expression chan ges between the atria and ventricle (Figures 1.1 and 1.2) and wit hin various locations throughout the heart (Figure 1.2) leading to different action potential (AP) waveforms58,59. Figure 1.1 shows the difference between the typical atrial and ventricular AP, the currents that constitute the AP, and the expression of the ion channels believed responsible for those currents. Figure 1.2 shows the various action potential waveforms present throughout the heart which are diffe rent due to a change in ion currents. In addition to physiological changes in heart function, cardiac pathologies may result from both electrical and structural remodeling. In atrial fibrillation, sodium currents (INa), calcium current (ICa), and the transient outwa rd potassium current (Ito) are reduced due to a decrease in the mRNA levels of channels responsible for these currents50,60. The decrease in ICa is likely responsible for shortening of the atrial action potentia l and the decrease of Ito results in loss of the ability of the heart rate to adapt to physiological c hanges. Atrial fibrillation is also accompanied by atrial enlargement, loss of myofibrils, accumulation of glycogen, alteration of mitochondrial size and shape, fragmentation of sarcoplasmic reticulum and dispersion of nuclear chromatin61,62. It is unclear whether atria are enlarged as a cause or as a re sult of atrial fibrillation63. Arrhythmias are the leading cause of death in patients with heart failure (HF)64. Recent studies indicate that AP prol ongation is a contributing factor to arrhythmias associated with HF65-69. Although the exact mechanism for AP
5 prolongation is not agreed upon, modulated K+, Na+, and Ca2+ currents have been identified65-73. Ionic currents are typically remodeled in HF through changes in the density and/or the expression of vari ous isoforms of ion channels65-73. Furthermore, glycosylation is reduced in both hamster74 and mouse73,74 models of cardiac heart failure suggesting that glycosylation machinery is altered. Regulation of glycan biosynthesis is essential for normal physiology Cell surfaces are replete with glycan stru ctures essential for proper development and normal function of living organisms with roles in protein trafficking, immunity, cell adhesion, receptor activation and endocytosis75 (Figure 1.3). Protein function may be modulated by glycans thr ough at least two mechanisms: 1) By altering the function of the proteinÂ’s c onjugate and, 2) By conferring biological activity to its conjugate. Glycans ac t as antigens on a variety of cells and activate the immune response as evidenced by the 1996 cholera pandemic of Bengal, India which was caused by Bengal 139 Vibrio cholerae76. This was the 139th identified strain of vibrio cholerae each of which had unique glycan structures. Exposure to and subsequent ant ibody formation of a single strain does not protect the host fr om any of the 138 other stra ins. Furthermore the A, B, O and AB blood types are dictated by the glycans attached with the O blood type lacking glycan structures and A and B each with unique structures. As discussed with Vibro cholerae, glycans act as antigens; therefor e, type B blood can not be administered to those with O or A types. Recent studies have
6 Figure 1.3. Roles of glycans in cellular functions. Figure 1.3. Cellular function is regulated by glycans through various mechanisms. The influence of glycans on cellular function r anges from protein folding to cellular communication. Figure from Ohtsubo and Marth, 200675.
7 reported the ability to change blood type B to O by simply altering the glycan; thus circumventing the immune respons e when, for example, type B blood is transfused into one with blood type A77. In these examples, the cell is not inherently immunoreactive, the glycans confer these attributes. In mammals, proteins and lipids are glycosylated in the endoplasmic reticulum and golgi apparatus where enzymes catal yze oligosaccaride formation from nine monosaccarides78,79. Glycosylation is non-template driven, unlike the DNA template necessary for protein synthes is, and requires expression of glycogenes that comprise 1-2% of the human genome78-82. Protein glycosylation refers to both N-glycans and O-glycans. N-linked glycosylation is attached to an asparagine residue; hence, the Â“N-linkedÂ” nomenclature. N-linked glycosylation requires the specific sequence of AsnXaa-Ser/Thr and sometimes Asn-Xaa-Cys where Xaa is any amino acid except pr oline. O-glycosylation lacks a specific conserved sequence; instead several enzymes may catalyze the first sugar residue attached to serine or threonine Glycosylation is a highly ordered process where the product of one enzyme is the substrate for the next and where catabolic glycosidase enzymes are as im portant as anabolic glycosyltransferase enzymes. N-glycan synthesis (summarized in figure 1. 4) is initiated in the cytoplasm where the first sugars are added to a lipid dolichol. This dolichol-glycan structure then translocates into the lumen of the endoplasmic reticulu m where further branching
8 Figure 1.4. Overview of N-glycan biosynthesis. Figure 1.4. Synthesis of N-glycans from initial attachment to dolichol through assembly and processing of N-linked glyc ans. Molecular defects of known CDG types are indicated where known etiologies occur. N-glycan assembly is initiated in the ER lumen by transferring two GlcN Ac residues (blue squares) to Dol-P and completed in the lumen of the golgi. Mannose (red circles), glucose (yellow triangles), fucose (grey tr iangle), galactose (green rhombus), sialic acid (pink diamonds). The mouse symbol designates a knock-out mouse of that enzyme, in red are yeast or CHO cells expressing s pecific enzymatic defects. Adapted from Marquardt and Denecke, 200382.
10 and sugar addition occurs. Eventually the glycan structure is transferred en bloc from dolichol to the asparagine residue of a newly synthesized protein. Within the endoplasmic reticulum, the glycan structures are extended and trimmed several times until the glycoprotein is transfe rred to the golgi. Fi nal processing of the glycan takes place in the golgi. The process includes addition of negatively charged sialic acid residues and when polysialyltransferase enzymes are expressed, sialic acids attach to other si alic acids (termed poly sialic acid) adding substantially more negative charges to a single structure 83. The glycome is defined as the full set of glycan structures produced in the body84,85, and is composed of thousands of gl ycan structures that perhaps is larger than the proteome75. Glycan diversity is accentuated by several factors which can be divided into two types: protein determined an d cellular factors (Figure 1.5). The protein itself c an only be N-glycosylated where specific Nlinked sequences are present and accessi ble according to tertiary protein structure (including protein phosphor ylation), and is commonly located extracellularly78,79. Sites located within the membrane, intracellularly or where the extracellular 3-dimensional structure prohibits access to those sites, will not be N-glycosylated. Also, the rate at wh ich the protein traverses through the glycosylation pathway may alter the final gl ycan structure. The second factor is cellular in nature. The repertoire of glycogenes expressed varies from cell type to cell type and, as this study indicate s, throughout developm ent. Furthermore, glycosidase and glycosyltransferase enzymes are considered to be constitutively
11 Figure 1.5. Regulatio n of glycan expression. Figure 1.5. Glycan structure expressi on is regulated through various cellular mechanisms. These include (1) glycosyltransferase and glycosidase gene transcription, (2) nucleotide sugar synt hesis and transport to the ER and golgi (sugar transporters not depicted), (3) en zymatic structure modification through phosphorylation, (4) enzyme competition for identical subs trates, (5) enzyme trafficking and access to substrates, (6) se cretion of catalytic domains resulting from proteolysis within the lumen of the golgi (7) gl ycan turnover at the cell surface by endocytosis. Figure from Ohtsubo and Marth, 200675.
12 active when expressed, yet competiti on between enzymes requiring a specific substrate further contributes to glycan diversity. Glycogene expression and glycan structure are tightly regulated in different tissues, through development and in disease states. Comelli et al. reported that bone marrow, thymus, lymph nodes, spleen, lung, testes, kidney, liver and brain all had unique glycogene expr ession and glycan populations86. Further, glycogene and glycan profiles of imm une tissues (bone marrow, thymus, lymphnodes, and spleen) were more si milar to each other than non-immune tissues (lung, testes, kidney, liver and br ain) as non-immune tissues were more similar than immune. Glycogene expression and glycan structure are also tightly regulated through development of various tissues. As s hown here and in Ishii et al., glycogene expression is altered throughout the developing myocardium and in the developing cerebral cortex, respectively4. Glycan profiles are distinct in each developmental stage of each tissue. Various disease states including Do wn syndrome, HuntingtonÂ’s disease, glaucoma, and heart failure reveal a change in glycogene expression and possibly glycan structure prof iles compared to healthy tissues28,39,74,75. Although altered glycan arrays may be present in t hese disease states, they may or may not cause, contribute to or exacerbate conditions.
13 Minor changes in glycogene expression may have a major impact on glycan structure and organism physiology75,82,85,87. The role of glycosylation is vast, spans every tissue, and is involved in numerous physiological processes. Ongoing research in glycobi ology focuses on immune responses, neuron tracking, ligand binding, and cancer i ndicating the wide range of functions of glycans in normal and pathophysiology75,86,88,89. Improper glycosylation results in patholog ies that range from mild disease to lethal 85. Common effects seem to target neuronal, cardiovascular and muscular systems. Congenital disor ders of glycosylation (CDG) are autosomal recessive disorders in which a single glycogene is mu tated or missing or there is no known cause (as shown in figure 1.4). To date, 28 unique forms of CDG have been identified, 16 N-glycosylation associated, 6 O-glycosylation associated, 4 Nand O-associated and 2 involving glycolipids90. Recently, a new category has been identified and classified as CDGs of hy perglycosylation defects. CDG tends to affect individuals differently; for exampl e, one patient of CDG-Ih was effectively treated with a low fat diet and essential oi l supplements while four others suffered fatal maladies90. With such a vast range of symptoms presented, diagnosis is difficult and with many unknown causes of death in infants, it is likely that many CDG patients are never i dentified. One common th read through the many CDGs is symptoms consistent with decreased excitability such as hypotonia and decreased metabolic activity. Also, all types of CDG have glycans with
14 reduced sialylation despite different enzymes ablated. In fact, isoelectric focusing of serum transferrin is the most common assay to diagnose CDG, testing for a decrease in te tra-, pentaand hexasialyl ated transferrins replaced by mono-, diand tri-sialylated transferrins. Chagas disease is an ailment affecting over 18 million with thousands of new cases reported each year91. Chagas disease is characterized by progressive chronic fibrotic myocarditis and degener ation of tissues innervated by the autonomic nervous system, most commonl y marked with cardiac abnormalities such as arrhythmias and cardiac insufficiency92. Trypanosoma cruzi the agent of Chagas disease, is a protozoan most commonly transmitted through insect bites, but can be transmitted th rough blood transfusions as well93. T. cruzi releases a sialidase to cleave negatively charged sialic acid residues from host cells to incorporate with itsel f. It is believed that this is the etiology of the major symptoms of Chagas. Changing the level of sialylation may contribute to cardiac arrhythmias and insufficiency possibly through modification of ion channel function. Ion transport is the basi s for cellular communication Ion transport across the membrane of ex citable tissues is essential for proper cellular and tissue function. Cellular me mbranes are essentially impermeable to ions; thus, ion transport requires assistance in the form of membrane proteins. These proteins can be divided into several groups, transporters, pumps and ion
15 channels. Transporters allow ions to move across the membrane with other solutes e.g. the sodium/glucose transporte r. Pumps require the use of energy to move ions across the membrane. Ion c hannels are water filled pores that allow ions to flow through the membrane down their electrochemical gradient when open. There are four types of ion channels, leak, ligand gated, stretch activated, and voltage gated channels. Leak channels ar e considered constitutively active (open) and contribute to maint enance of the resting membr ane potential of a cell. Stretch activated ion channels require t he membrane to physically stretch the channel to an open state while ligand gat ed channels open in response to a ligand (i.e., a neurotransmitter) binding to its' extracellu lar surface. Voltage gated ion channels gate in response to the depol arization and repolarization of the cell membrane. Cardiac contraction is the result of orchestrated ion channel function The cardiac action potential is the concer ted opening, inactivation, and closing of many types of voltage gated ion channels, the Na+/K+ ATPase pump, and possibly some ligand gated ion channels (s ummarized in Figure 1.1 and 1.6). The result of the cardiac action potential is cardiac systole. The cardiac action potential of contractile myocytes is divided into 5 distinct phases. Phase 0 is the depolarization of the cellular membr ane by opening of voltage gated sodium channels (Nav) which allows sodium ions to move down their electrochemical gradient and into the cell. Phase 1 begins at the height of ce llular depolarization and is marked by a sudden repolarization of the cell. This occurs when voltage
16 Figure 1.6. Schematic of a ty pical cardiac action potential. Figure 1.6. The cardiac action potential is shown, with ionic currents responsible for each phase listed. Figure from Keating and Sanguinetti, 200194.
17 gated sodium channels inactivate and sodium can no longer traverse the membrane and around the same time, vo ltage gated potassium channels open to allow potassium ions to exit the cell (Ito); thus, causing a short, rapid repolarization. As Ito diminishes, voltage gated calcium channels open to initiate phase 2. Influx of calcium ions is appr oximately the electrical equivalent to the efflux of potassium ions leading to a flat segment in the cardiac action potential termed the "plateau." In phas e 3, calcium channels inactivate and another population of slowly activating potassium channels open which causes the final repolarization and hyperpolar ization of the membrane. This hyperpolarization of the membrane is essential for the voltage-gated ion channels to recover from inactivation. In phase 4, mostly leak and ligand gated ion channels are open to maintain the resting membrane potential and allow more channels to return to a closed posit ion so the cell is prepared for the next action potential and resulting systole. The structure of Nav dictates channel function Voltage gated sodium channels (Nav) are transmembrane prot eins which open in response to membrane depolarization to selectively allow sodium ions to pass though95. Nav are composed of a single poly peptide chain approximately 220kD and is composed of 24 transmembrane s egments subdivided into 4 homologous domains composed of 6 tr ansmembrane segments each (F igure 1.7). Each of the 6 transmembrane segments are unique, yet have homologous segments in the other 3 domains. The S5 and S6 domains line the por e with an extracellular
18 Figure 7. Schematic of the voltage-gated sodium channel structure. Figure 7. Schematic of voltage-gated sodium channel alpha and beta subunits. (a) A characteristic alpha subunit with four homologous domains, each consisting of six alpha helical transmembr ane segments is illustrated with the 1 subunit. The S5 and S6 (shown in green) of eac h domain are considered the pore forming segments. The loop connecting the S5 and S6 dips into the pore and forms the selectivity filter (designated by white ci rcles). Note that both proteins are glycosylated (represented by ). Blue circles in the intracellular loops of domains III and IV mark the inactivati on gate IFM motif and its receptor (h, inactivation gate); P, phosphor ylation sites. (b) A hypothetical three-dimensional structure of the Nav channel -subunit compiled from electron micrograph reconstructions. Figures adapted from Yu et al. 200396.
19 S5-S6 linker that dips into the pore. Th is pore forming loop is essential to proper channel selectivity, where t he specific amino acid sequence of DEKA (aspartate, glutamate, lysine and alanine) defines the channel as sodium specific. When this sequence is changed to the calcium channel sequence of EEEE, the channel allows calcium to pass while preventing sodium entry. Alt hough the remaining structure is currently under debate, t here is consensus t hat the remaining transmembrane segments are lo cated peripherally to the pore. The S4 segment is considered to be the voltage sensor sinc e every third amino acid is a positively charged arginine or lysine and the whol e segment moves in response to membrane depolarization. Movement of the S4 segment causes a conformational change and the channel to gat e allowing sodium to enter into the cell. Also of note is the intracellular linker of domains III and IV which contains the hydrophobic amino acid sequence; IFM (isoleucine, phenylalanine and methionine) which has been implic ated in fast inactivation. The tertiary and quaternary structure of voltage gated ion channels has been the topic of recent debate within the scient ific community as an alternative hypothesis has arisen from crystallography work on the voltage gated potassium channel which shares signi ficant homology with Nav 97-100. Cartoons of both models are shown in figure 1.8. The c onventional theory, as proposed through studies from the past 25 year s, has the S4 segment lo cated within a water filled column formed by the other segments of the same domain. The S5 and S6 segments form the pore while S1-S3 surr ound the S4. The S4 segment moves
20 Figure 1.8. Two competing theories for voltage-gated ion channel gating. Figure 1.8. Cartoon depicting two models of S4 segment movement in response to a change in membrane potential ( V). (+) signs represent positively charged amino acids within the pr otein structure. Figure from Jiang et al. 2004.97
21 towards the extracellular surface in a perpendicular manner to the cellular membrane in response to membrane depolari zation. In the model proposed by Jiang et.al., the S4 segment forms a paddle-like structure with the S3 segment peripheral to the pore which rota tes through the lipid bilayer towards the extracellular surface and again causes a conformational change in the poreforming segments to open the channel97. Both theories have two important similarities. First, the positively charged amino acids, composing the S4 segment, move towards the extracellula r surface and second, this movement results in channel gating. Since the in troduction of the paddle theory in 2003, the scientific community has been vigorously debating these theories with evidence supporting the traditional theory101-108 and other data supporting the paddle theory97-100,109-113. Post-translational modifications may alter ion channel function The surface potential theory predicts that charges closely associated with the membrane adjacent to voltage gated i on channels contributes to channel gating95. The idea is based upon electrostatic attraction of the voltage sensor by negative charges closely localized to the channel. The source of these charges include charged lipids of the cell membr ane, charged amino acids of the protein itself or a closely associated protein, i ons present in the extracellular fluid and negatively charged sialic acid residues capping glycan structures.
22 Voltage gated ion channels are heavily posttranslationally modified through fatty acylation, phosphorylation, ni trosylation, sulfonation, and glycosylation. Of these posttranslational modifications, glycosylation is the highest proportion with upwards of 30% of the fi nal channel mass being glycans114-116. A fully glycosylated and sialylated channel could ha ve as many as 100 sialic acid residues attached to a single channel114-116. Each ion channel is differently glycosylated based upon number and location of potential N-linked sites and the other factors involved in N-glycosylation described above. The impact of glycosylation, particularly si alic acids, on ion channel gating has been the focus of numerous studies73,117-128. These studies report that glycosylation can directly alter gating of voltage-gated sodi um and potassium channels in an isoform s pecific manner. For example, in CHO cells, Nav1.4 gating is sialic acid sensitive whereas the gating of Nav1.5 does not change in response to the altering level of sialic acids117,118. Nav1.4 is more heavily glycosylated than Nav1.5 likely due to the number of glycosylation sites with mature glycans attached. Nav1.5 is the predominate sodium channel isoform expressed in mouse cardiac tissue and commonly is considered the cardiac isoform3. Studies have concluded that Nav1.5 is the isoform responsible for phase 0 of the cardiac action potential indicating that Nav1.5 gating initiates and propagates the cardiac action potential129. Nav1.5 has the same basic structur e as other voltage gated sodium
23 channels and is putatively heavily glyco sylated with 13 potential glycosylation sites117. A recent study of Nav function in neonatal and adult cardiomyocytes showed that neonatal ventricular Nav required a ~10mV greater depolarization to gate than does neonatal and adult atrial and adult ventricular Nav 123. Following desialylation through neuraminidase treat ment, neonatal atrial and adult atrial and ventricular Nav gated similarly to untreat ed (and neuraminidase-treated) neonatal ventricular Nav. Furthermore, investigators determined that Nav1.5 was similarly expressed throughout the developing myocardium and 1 did not contribute to the changes. Western blot analysis revealed that neonatal and adult atrial and adult ventricular Nav had higher levels of sialylation than did neonatal ventricular Nav. Regulated glycogene expression is likely responsible for the various levels of Nav glycosylation observed, and the resulting changes in Nav gating. This suggests that the ca rdiac glycome may be regulated throughout the heart during development. This study was designed to determine whether the glycome is remodeled throughout the developing myocardium and whether the remodeled glycome can affect excitability. Glycans more specifically the negatively charged sialic acid residues commonly capping glycan structur es, modulate gating of voltage gated ion channels in both a cis (glycans attached to the alpha subunit) and trans (glycans attached to an auxiliary subunit) manner as reported in chapter 3. The
24 level of glycosylation of Nav changes throughout the dev eloping myocardium and in cardiac failure74,123. The change in glycogene expression throughout the developing myocardium is described in ch apter 4 and chapter 5 illustrates the correspondingly diverse N-glycan profiles. Finally, chapter 6 suggests that cardiac excitability can be altered thr ough the regulation of a single glycogene.
25 CHAPTER 2 MATERIALS AND METHODS Chinese Hamster Ovary (CHO) Cell Culture and Transfection Pro5 and Lec2 cells were grown as described previously130. Briefly, cells were plated onto 35 mm culture dishes at 25-50 % confluence. Following a 24 h incubation, cells were then exposed to a 1 ml Opti-MEM (Invitrogen) medium containing 8 l lipofectamine (Invitr ogen) and 1-2 g DNA. Following a 5-24 h incubation at 37C in a 5% CO2 humidified incubator, t he medium was replaced with CHO medium consisting of Dulbecco 's modified Eagle's medium (DMEM; Mediatech) supplemented with 25 mM Hepes, 15% fetal bovine serum (FBS; Mediatech), and 100 U ml-1 penicillin and 100 g ml-1 streptomycin. Growing medium included the same antibiotics, 10% FBS, and alpha Minimum Essential Medium ( MEM) with (Pro5) or without (Lec 2) riboand deoxyribonucleosides (Invitrogen). Electrophysiological record ings began 68-76 h post-transfection, selecting cells expressing GFP. Vector Construction and Mutagenesis The rNav1.2 open reading frame (ORF) inserted into pRC-CMV (Invitrogen) was a gift of Dr. Alan Goldin. The hNav1.7 cDNA ORF was inserted into pcDNA3.1. Expression vectors containing hNav1.4 and hNav1.5 were as previously described118. h 1 was subcloned into the bici stronic vector, pIRES2-EGFP
26 (Clontech), to ens ure expression of 1 through visual inspection. The h 1 mutant (h 1N) was created using the GeneEditor (Promega) site-directed mutagenesis kit. h 1 was cloned into pBluescript vector (Strat agene) as a template. Each asparagine residue initiating an external N -linked consensus sequence, N X (S/T), was mutated to a serine residue through sequential mutagenesis. The constructs were sequenced to confirm successful mutagenesis. h 1N was then subcloned into pIRES2-EGFP for co-expression experiments. h 1 and h 1N were amplified using PCR with the following oligonucleotides 5'TCCGGCCACCTGGACGCCCG-3' and 5'-G CGCAGCACGCGCCGCGCAG-3'. PCR products were subcloned into pc DNA3.1/V5-His TOPO TA expression vector (Invitrogen). Both ORFs were subsequently subcloned into pEGFP-N1 (Clontech) to generate C-terminal, GFP-tagged h 1 and h 1N constructs. Electrophysiology and Data Analysis Sodium Current Recordings Sodium currents were recorded using t he whole cell patch clamp technique described previously117,118. The combination of an Ax on Instruments 200B patch clamp amplifier with a CV203 BU headstage (Axon Instrument s, Foster City, CA). Pulse acquisition software (HEKA) running on an 800 MHz Pentium III PC computer (Dell Computers) was used to generate pulse pr otocols. The resultant analog signals were digitized using t he ITC-16 analog to digital converter (Intsrutech, Great Neck, NY).
27 Whole cell patches were formed us ing techniques previously described 131. Electrodes were back-filled with electr ode solution and manipulated to close proximity to the target ce ll. Slight negative pressure was applied to the electrode and giga-seals formed between the cell and electrode tip and a short, rapid increase in negative pressure provided electr ical access to the interior of the target cell. Pulse protocol s are explained explained later. All data were recorded at least 5 minutes after attaining whole ce ll access to ensure dialysis of electrode solution. External recording solutions consisted of (in mM): 224 Sucrose, 22.5 NaCl, 4 KCl, 2.0 CaCl2, 5 glucose, and 5 Hepes. Intracellular recording (electrode) solutions contain (in mM): 120 sucrose, 60 CsF, 32.5 NaCl, and 5 Hepes. Both solutions were titrated with 1 N NaOH to pH 7.4 at room temperature. All solutions were filt ered using 0.2 m f ilters (Invitrogen) immediately prior to use. For the Ca2+ perfusion studies, the Ca2+ calcium concentration was reduced in the external solution to 0.2 mM. Seals were formed in the bath solution containing 2.0 mM Ca2+. The cells were first perfused with 2.0 mM Ca2+ bath solution and followed by perfusion the 0.2 mM Ca2+ bath solution to determine directly the shift in Va with a 10-fold change in external Ca2+ concentration. All of the data shown were recorded at least 5 min after attaining whole cell configuration to assu re complete dialysis of the intracellular solution. All of the solutions were filtered using Gelman 0.2-m filters immediately prior to use.
28 Pulse Protocols Conductance-Voltage (G-V) Relationship Pulse protocols were used as previously described 117. A holding potential of 120 mV was applied to the cell and stepped from -100 to +70 mV in 10 mV increments for 10 ms. Consecutive pulses were initiated every 1.5 s and leak subtracted using the P/4 method whic h steps negatively fr om the holding potential to eliminate any leak current. At each potential, steady-state whole-cell conductance was determined by measuring the peak current and dividing by the driving force (difference between t he membrane potential and the observed reversal potential). Single Boltzmann fits of the data determined maximum conductance and the average Va SEM were determined from this fit. Normalized data from the Bo ltzmann fits were averaged with remaining cells of the same type and an averaged conduct ance-voltage curve was determined using the following Boltzmann relation fit to the data: Fraction of maximal c onductance= [1+(exp-(V-Va/Ka))]-1, where V is the membrane potential, Va is the voltage of half activation, and Ka is the slope. Steady-State Inactivation Curves (hinf) Cells were prepulsed for 500 ms from the holding potential (-120 mV) to potentials ranging from -130 to -20 mV in 10 mV increments, followed by a +60 mV pulse for 5ms and returning to the -120 mV holding potentia l. Currents from
29 each cell were normalized to the maximal current (determined through a single Boltzmann fit), averaged with ot her cells of the same type and again fit to a single Boltzmann relationship (eq. 2) from which Vi (voltage of half inactivation) and the slope were calculated. Fraction of maximum cu rrent = [1+(exp-(V-Vi/Ki))]-1, Recovery from Inactivation Cells were held at -120 mV membrane pot ential, pulsed to +60 mV for 10 s, and stepped to the recovery potential for 1-20 ms in 1 ms increments. The potential was then stepped again to +60 mV for 10 ms. Peak currents from the two +60 mV pulses were compared to deter mine the fraction of current measured during the second pulse which represents th e fraction of channels that recovered from inactivation during the recovery pulse Fractional current was plotted as a function of the recovery time between the two test pulses of 60 mV. Single exponential functions were fit to the data to determi ne the time constants for recovery from inactivation, trec. Measurement of Inacti vation gating kinetics Inactivation gating kinetics were dete rmined from attenuating currents (90-10%) of traces used for G-V relationships which were fit to a single exponential function.
30 Neonatal and Adult Cardiac Tissue Isolation Neonatal and adult atria and ventricle ti ssue were isolated for microarray, quantitative PCR, western blot analysis and mass spectrometry. Neonatal mice and adult mice were euthanatized and whole hearts removed and placed in Dulbecco's phosphate buffered saline. Atria were gently removed from the remaining heart and ventricles were disse cted away from the base of the heart with great care taken to ensure only atri a and ventricles were removed. The remaining portions were discarded. Tissue intended for microarray and quantitative PCR studies was transferred to RNase Later (Sigma, St. Louis, MO) and incubated for minimum one hour. Tissue intended for western blot and mass spectrometry studies was snap frozen in liquid nitrogen to prevent protein degradation and stored at -80oC. mRNA Isolation Heterogeneous populations of litter-mate animals were isolated for microarray testing. Neonatal samples were each composed of 7-9 anima ls (14-18 atria or ventricles) yielding approximately 25-28 mg of tissue. Adult samples were each composed of 4 animals (2 male a nd 2 female) aged 10-12 weeks yielding approximately 25-28 mg of tissue. Tissue was homogenized by dounce and isolated following RNeasy manufacture r protocols (Qiagen). Beckman Spectrometer was used to determi ne final RNA concentration.
31 Microarray Three RNA samples each of neonatal and adul t atria and ventricle were sent to the Consortium for Functional Glycomics Gene Microarray Core E for microarray analysis. Samples were amplified and bi otin labeled using the Bioarray High Yield RNA transcript labeling kit (ENz o Life Sciences, Farmingdale, NY). Hybridization and scanning of the glyc ogene-chip, GLYCOv2, were performed according to Affymetrix's recommended prot ocols (Affymetrix, Santa Clara, CA). Microarray Analysis The GLYCOv2 chip was created by the C onsortium for Functional Glycomics and produced by Affymetrix (Affymetrix, Santa Clara, CA). This chip was designed that each probeset consists of 11 perfe ct match and 11 single base mismatch probe pairs (Table 2.1). The intensities from each perfect match were compared to corresponding mismatch pair. Invariant set normalizat ion of the data was performed using the DNA-Chip (dChip) Analyzer (www.dchip.o rg) software package for probe-level and high level analysis of gene expression microarrays. Hierachical clustering and class comparison was accomplished using Biometric Research Branch (BRB) Array Tools v3.2.2. Heatmaps were generated using the dChip program. Class comparison used a p-value cutoff of 0.05 and a multivariate permutation based false discovery rate calculation pres et at 10% with 80% confidence level.
32 Table 2.1 Breakdown of probesets on GLYCOv2 Probesets In Triplicate Probesets In Duplicate Single Probesets Total genes transcript targets Total Probesets Total Human 503 426 101 1030 2462 Total Mouse 443 363 119 925 2174 Total Other (control) 0 0 46 46 46 2001 4682
33 Quantitative PCR Three RNA samples each of neonatal and a dult atria and ventricle were reversetranscribed to cDNA using Superscript II Reverse Transcriptase (Invitrogen) following manufacturer's protocols. Brie fly, 1ug of total RNA, 100M dNTPs and 100ng of random hexameric primers (I nvitrogen) were incubated at 65oC for 5 minutes then placed on ice. First strand buffer and 10 mM dithiothreitol (DTT) were added and incubated at room temperature (25oC) for 2 minutes. Finally, 200 M Superscript II reverse transcr iptase was added to the mixture and incubated for 10 minutes at room temperaturefollowed by 42oC for 50 minutes and 70oC for 15 minutes. cDNA is ready for use in real time reactions. Primer sets were designed using Prim erQuest (IDT) and are shown below. Primer sets were tested for efficien cy and precise amplification using dilution curves. Quantitative PCR was perf ormed on 12 gene products including HPRT and -actin as controls. Each primer set was run in triplicate for each sample. SYBR Green PCR master mix (Superarray), primers and cDNA were combined in one well of a 96 well P CR plate (Rio-Rad) and cove red using RT-PCR optical tape (Bio-Rad). PCR products were detect ed in real time using the iCycler iQ detection system (Bio-Rad) with PCR conditions of 5 minutes at 95oC followed by 40 cycles of 30 sec at 95oC, 30 sec at 60oC and 30 sec at 72oC. Relative expression levels were reported using the CT method of analysis where triplicate threshold values of a si ngle gene are averaged and compared to the
34 control (either HPRT or -actin) then this CT value is compared to the CT of another sample. Quantitative PCR primer sequences ST3Gal3 CTG TGA TGA AGT GGC AGT CG CTC GCT GGA TGT TGT CTG TC ST3Gal5 AAA GTC CCA CTC CAG CCA AAG C GTG TAG CCA AGA CAA CGG CA STX AGC CAG CCT CAT CCA AAT G TAT CCT TCT CCG CAT CCA AG ST6Gal1 GAC CAG GAG TCA AGT TCA GCG T AGA AGA CAC GAC GGC ACA CT ST8Sia6 TGC TGC TCC TCC TGC GTA T TAT GTG CTG TTC CTG GTG CGT G ST6GalNAc6 AAC AAA GAG CAG CGG TCA GC GTT GCC GAG GAT AGG GAA GTA GG Versican TGG CTG TGG ATG GTG TTG TG TGC TCT GGG CTT GCT ATG AC HPRT GCA GTA CAG CCC CAA AAT GG GGT CCT TTT CAC CAG CAA GCT B-actin CCA ACC GTG AAA AGA TGA CC CCA GAG GCA TAC AGG GAC AG Glycan Screening Glycan Isolation N-glycans were isolated as previously described132. Cardiac tissue was homogenized in 0.5% SDS tris buffer and dialyzed in 12-14 kDa cut-off dialysis tubing in an ammonium hydrogen carbonat e solution (50 mM, pH 7.4) for 48 hours. Once dialyzed, samples were lyophilized. Reduction and carboxymethylation of samples were ca rried out by incubation in 0.5ml of 2 mg/mL DTT in deoxygenated tris buffer (0 .6M, pH 7.4) for 45 minutes at 37oC followed by addition of 0.5 ml of 12mg/mL i odoacetic acid in tris buffer (0.6M, pH 7.4) and incubation for 90 minutes at room temperature in the dark. Reaction was terminated by dialysis for 48 hours and the sample was lyophilized. Samples were then digested in 1 mL of a 50mM ammonium hydrogen carbonate solution (pH 8.4) with approximately 2 mg of TPCK treated bovine pancreas trypsin at
35 37oC for 16 hours. The sample was purified through a Sep-Pak C18 conditioned with 5mL methanol, 5mL 5% (v/v) acetic acid in water 5mL propan-1-ol and 30 mL of 5% (v/v) acetic acid followed by collection of 3mL of 20% (v/v) and 40% (v/v) propanol in 5% (v/v) acetic acid. These fractions are pooled and lyophilized followed by digestion with 3 units of Nglycosidase F (PNGase F) in 200 uL of 50mM ammonium hydrogen carbonate (pH 8.4) at 37oC for 20 hours. The digested sample was purified through a pre-conditioned Sep-Pak C18 (5mL methanol, 5mL 5% acetic acid, 5mL propan-1ol and 15 mL of 5% acetic acid), eluted with 5mL of 5% acetic acid. Glycan derivatization N-glycans were prepared for mass spectrom etry by chemical derivatization using the sodium hydroxide procedure133. 5 pellets of sodium hydroxide and 3mL of dry DMSO were crushed together in a glass mortar. 1mL of the resulting slurry was added to the dry sample in a glass t ube followed by 0.5mL of methyl iodide. The mixture was vigorously mixed on an aut omatic shaker for 15 minutes at room temperature. The reaction was quenched by addition of water, Permethylated N-glycans were extract ed with 1mL of chloroform and washed with 3mL of water several times. The organic phase was dried under a stream of nitrogen. Derivatized and dried glyc ans were then purified through a preconditioned Sep-Pak C18 (5mL methanol 5mL water, 5mL acetonitrile, 15 mL water) and eluted with 15%, 35%, 50% and 75% (v/v) acetonitrile in water.
36 Mass Spectrometry Mass spectrometry was performed thr ough the Consortium for Functional Glycomics and the methods were described previously132. The permethylated sample was dissolved in 10 l of methanol, and 1 l of dissolved sample was mixed with 1 l of 2,5-dihydroxybenzoic acid (20mg/mL in 70:30 (v/v) water:methanol), spotted onto a metal plate and dried under vacuum. MALDIMS and MALDI-MS/MS data were acquired using a Perseptive Biosystems Voyager-DETM STR mass spectrometer in the reflectron mode with delayed extraction and a 4800 MALDI-TOF/TOF (Applied Biosystems, Damstadt, Germany) mass spectrometer respective ly. The collision energy for MALDIMS/MS experiments was set to 1kV and argon was used as collision gas. Cardiomyocyte Isolation for Electrophysiology The cardiomyocyte isolation protocol was adapter from a method described previously134. Neonatal (postnatal day 2-3) mice we rapidly euthanized and hearts excised and placed in 0 Ca2+ TyrodeÂ’s Solution. Atria and ventricles were carefully separated and digested in 260 units Type I collagenase (Sigma, St. Louis, MO)/ mL 0 Ca2+ TyrodeÂ’s Solution at 37oC for 40 minutes. Cells were centrifuged at 160 g for 5 minutes and the supernatant replaced with fresh collagenase solution. Cells were gently triturated and incubated at 37oC for 30 minutes followed by centrifugation at 160 g for 5 minutes. The supernatant was replaced by CHO media (DMEM supplem ented with 10% Fetal Bovine Serum (Gibco) and 100 U/ml penicillin and 100mg/ml streptomycin (Gibco), triturated,
37 and incubated at 37oC for 20-40 minutes to stop diges tion. Again the cells were centrifuged for 5 minutes at 160 g and plat ed on laminin-coated 35mm dishes in fresh CHO media. Cardiomyocyte Electrophysiology Sodium Current Recordings Recording techniques were described abov e. External recording solutions consisted of (in mM): 20 NaCl, 10 TES, 5 KCl, 1 CaCl2, 5 CsCl, 10 glucose, and 100 choline chloride adjust ed to pH 7.35 with CsOH. Intracellular recording (electrode) solutions contain (in mM): 20 NaCl, 10 TES, 2 MgCl2, 2 CaCl2, 20 EGTA, and 105 CsF adjusted to pH 7.35 with CsOH. All solutions were filtered using 0.2 m filters (Invitrogen) immediatel y prior to use. Action Potential Recordings Myocytes were patched and recorded in external solution (in mM): 135 NaCl 5 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, 10 glucose, pH 7.4. Patch pipettes were filled with a solution with the following constituents (in mM): 110 K-Asp; 20 KCl, 10 NaCl, 4 ATP-Mg, 10 HEPES, pH 7.3. APs were recorded at room temperature (22-25 C) usi ng an Axopatch 200B amplif ier (Axon Instruments) and pCLAMP 9 software (Axon Instruments). APs were triggered by a 2-ms injection of a depolarizing current at a frequency of 1 Hz. Analysis of APs was performed using Clampfit 9 software (A xon Instruments, Foster City, CA).
38 Transgenic mouse The STX knockout mouse was provided through collaboration with Dr. Jamey Marth, University of California San Di ego. Neonatal mice were 2-3 days postnatal and adult mice were 10-12 weeks post-natal. Data Analysis Sodium current electrophysiological dat a were analyzed using Pulse/PulseFit (HEKA) and Sigmaplot 2001 (SSPS Inc.) softw are. Action potential data were analyzed using Clampfit (Axon) and Sigm aplot 2001 (SPSS Inc). Figures were produced using Sigmaplot 2001 (SPSS Inc), Mi crosoft Excel (Microsoft), or Corel Draw (Corel).
39 CHAPTER 3 THE 1 SUBUNIT MODULATES Nav GATING IN AN ISOFORM-SPECIFIC, SIALIC ACID-DEPENDENT MANNER Cardiac remodeling often involves modul ating ion channel expression and/or function. One mechanism of this remodeli ng likely involves regulated expression and function of Nav alpha and beta subunits. With ten identified Nav alpha isoforms, changing expression of these isoforms would modify ion currents135. The role of 1 was not conclusively establish ed with theories of the role of including nodal stabilization, cellular lo calization, functional expression, kinetics and voltage-dependence of channel gating136. Previous studies report that Nglycans alter function of some Nav alpha isoforms. Here we report 1 subunit sialic acids alter t he voltage dependence of Nav gating. The 1 subunit external domain is essentia l for correct modulation of sodium channel gating and is the site of f our potential N-glycosylation sites137. At least three of these four si tes are thought to be glycosylated in the mature protein. Publish ed reports agree that 1 causes a hyperpolarizing shift in the voltage dependence of inactivation and in several studies, activation gating was also shifted in the hyperpolarized direction by 1 138-145.
40 Voltage gated sodium channel and subunit isoforms each have a unique glycosylation signature determined through t heir differing number and location of N-glycosylation sites. In an attemp t to determine how N-glycans alter Nav function, four isoforms, the adult skeletal muscle isoform (Nav1.4), the cardiac isoform (Nav1.5), a peripheral nerve isoform (Nav1.7), and a brain isoform (Nav1.2), were expressed in the fully gl ycosylating Pro5 and reduced sialylating Lec2 cell lines. Nav1.4 and Nav1.5 were previously reported 117,118 and Nav1.2 and Nav1.7 were studied here for the first ti me. Table 3.1 and figures 3.1-3.4 indicate that Nav1.5 and 1.7 are not sensitive to sialic acids; whereas, Nav1.2 shows a small, insignificant depolarizing shift in gating when sialylation is reduced. Nav1.4 shows a significant ~14.6mV depolarizing shift in the absence of sialic acids. In an attempt to determine the role of the 1 subunit in voltage gated sodium channel gating, we co-expressed 1 with each of the four Nav isoforms. When co-expressed in the fully sialylating Pro5 cell line, 1 induced a hyperpolarizing shift in all measured gating par ameters of three of four subunits. 1 did not have an effect on Nav1.4 gating. These data genera lly agree with previously published work with most studies indicating that 1 induces a hyperpolarizing shift in the gating of various subunits137,143,146-150.
41 Table 3.1. The measur ed gating parameters for 1sialic acid Channel construct n Va (mV) Vi (mV) h (ms) trec (-120 mV) (ms) Nav1.4 + SA 9 -31.3 2.0 -71.5 3.3 2.4 0.4 1.8 0.05 Nav1.4 SA 9 -16.7 1.7a-60.8 1.5a7.9 1.1a 1.3 0.03a Nav1.4 + 1 + SA 11 -29.2 1.6 -70.0 2.3 2.9 0.5 1.7 0.02 Nav1.4 + 1 SA 9 -16.3 1.1a-63.1 1.8 b 8.1 0.2a 1.4 0.03a Nav1.5 + SA 13 -29.0 2.2 -78.7 2.5 2.8 0.4 4.0 0.1 Nav1.5 SA 10 -29.5 1.6 -79. 5 1.9 2.4 0.2 4.1 0.1 Nav1.5 + 1 + SA 11 -37.4 1.6a-86.1 3.2 b 2.0 0.1 b 5.6 0.3a Nav1.5 + 1 SA 9 28.6 0.9 -78.8 1.6 2.9 0.2 4.1 0.1 Nav1.7 + SA 10 -15.1 1.2 -70.0 2.4 4.8 0.8 5.5 0.09 Nav1.7 SA 9 -14.4 1.7 -70.0 3.7 5.2 1.2 5.5 0.06 Nav1.7 + 1 + SA 12 -23.8 1.8a-76.2 2.0 b 3.0 0.4 b 7.8 0.2a Nav1.7 + 1 SA 9 -13.4 1.3 -68.3 1.5 5.1 0.8 5.5 0.2 Nav1.2 + SA 9 -14.6 1.6 -60.5 3.1 3.3 0.7 2.5 0.08 Nav1.2 SA 10 -11.7 1.8 -62. 7 3.1 3.1 0.5 2.2 0.04 b Nav1.2 + 1 + SA 12 -20.8 0.7a-68.1 2.7 b 2.3 0.3 b 3.1 0.07a Nav1.2 + 1 SA 9 -11.9 1.0 -62.2 2.8 3.2 0.6 2.2 0.06 b hSkM1P1 + SA 10 -23.5 2.3 -70. 4 2.5 3.4 0.5 1.8 0.07 hSkM1P1 SA 8 -26.7 1.2 -68. 8 2.5 3.4 0.8 1.8 0.05 hSkM1P1 + 1 + SA 9 -32.7 2.0a-75.9 1.6 b 2.3 0.4 b 2.4 0.06a hSkM1P1 + 1 SA 11 -24.0 1.5 -69.2 1.5 3.7 0.6 1.9 0.04 Table 1. The measured gating parameters for 1sialic acid. The data are the mean parameter values S.E. h data were measured for Nav1.4, Nav1.5, and hSkM1P1 at -40mV and for Nav1.2 and Nav1.7 at -30mV. Two-tailed StudentÂ’s t test was used to determine the significance of 1 sialic acids comparing each condition with the parameter m easured for the fully sialylating subunit alone. Significance (p<0.1) demarcated wit h an (a) and highly significant ( p < 0.005) demarcated with a (b). Table from Johnson et al 2004120.
42 Figure 3.1. Alpha and 1 subunit sialic acids modify channel activation in an subunit dependent manner. Figure 3.1. Conductance-voltage ( G-V ) relationships for four voltage-gated sodium channel subunits 1 as expressed in the fully sialylating, Pro5, and reduced sialylating, Lec2, cell lines. The data are the mean normalized peak conductance (G) S.E. at a given memb rane potential and are shown as curves that are fits of the data to single Boltzmann relationships. Data are summarized in Table 1. Circles with solid lines subunit alone; squares with dashed lines + 1. Filled symbols in Pro5 cells; open symbols in Lec2 cells. A Nav1.4. B Nav 1.5. C Nav1.7. D Nav1.2. Figure adapted from Johnson et al 2004120.
43 Figure 3.2. Alpha and 1 subunit sialic acids modify channel inactivation in an subunit dependent manner. Figure 3.2. Steady stat e channel availability ( hinf) curves for the four subunits 1 sialic acid. The data are the mean normalized peak current ( ) S.E. measured during a maximally depolariz ing test pulse following a 500-ms prepulse to the plotted potentials. Li nes and symbols are identical to those described in figure 3.1. Figure from Johnson et al 2004120.
44 Figure 3.3. Alpha and 1 subunit sialic acids alter ch annel fast inactivation rates in an subunit dependent manner. Figure 3.3. The rate of fa st inactivation for the four subunits 1 sialic acid. The data are the means S.E. time constants for fast inactivation ( h) as a function of memb rane potential. Inset to C representative normalized whole cell Na+ current traces measured at Â–20 mV for Nav1.7. Note that the rate at which the current attenuates (inactivates) is much faster in the presence of 1 sialic acids, consistent with the observed shift in h along the voltage axis. The scale shown is for Na 1.7 + 1 + SA current traces, to which the other current traces were normalized. Lines and symbols are identical to those described in figure 3.1. Figure from Johnson et al 2004120.
45 Figure 3.4. Alpha and 1 subunit sialic acids modify channel recovery from inactivation in an subunit dependent manner. Figure 3.4. Time constants for re covery from fast inactivation ( rec) S.E. measured for the four subunits 1 sialic acid at three recovery potentials. Inset to C typical plot of the fractional recovery measured following a Â–120mV recovery potential for Nav1.7 1 SA. The data are the m eans S.E. fractional current measured during a second depolarizi ng test pulse following the plotted interval at Â–120 mV recovery pulses of various durations. The lines are exponential fits of t he data from which the rec were determined. Lines and symbols are identical to those described in figure 3.1. Figure from Johnson et al 2004120.
46 Due to the fact that 1 is heavily glycosylated, we tested the hypothesis that 1 sialic acids modulate Nav gating. 1 induced a hyperpolarizing shift in the gating of three of the four subunits studied when expressed in Pro5. This effect was eliminated when expressed in the essentia lly non-sialylating Lec2 cell line as all four subunits gated the same as expressed alone. 1 did not alter any gating parameter of Nav1.4 under either sialylating or non-sialylating conditions. As shown in figures 3.1-3.4, all effects of 1 on gating can be attributed to the sialic acids attached to 1, since in the absence of 1 sialic acids, Nav gating is not modulated. In addition to N-linked glycosylation, sugars can be attached to serine or threonine residues of membrane proteins termed, O-linked glycosylation. Mutagenesis of the four N-linked glycosylation sites provides a method to determine that N-linked sialic acid s are responsible for modulating Nav gating. The mutant 1 lacks all N-linked glycosylation, yet all other post-translational modifications remain. As exhibited in figure 3.5, 1N ( 1 with all N-glycosylation sites mutated resulting in no N-glycosylat ion) had no effect on gating of any of the Nav subunits previously modulated by 1. Thus, we confirm that 1 N-linked sialic acids are fully responsible for the observed shifts in Nav gating. Figures 3.1-3.4 showed that the heavily glycosylated Nav1.4 was sensitive to subunit sialic acids (cis effect), but not sensitive to 1 sialic acids (trans effect). Conversely, the putatively lesser glycosylated Nav1.2, 1.5 and 1.7 were not
47 Figure 3.5. 1 subunit sialic acids modify channel gating parameters in a saturating manner. Figure 3.5. Voltage-dependent steady st ate and kinetic gating for hSkM1P1 1 SA is shown. Circles hSkM1P1 expressed alone; squares hSkM1P1 + 1. Filled symbols in Pro5 cells; open symbols in Lec2 cells. A schematic of hSkM1P1 structure illustrates that the chimera consists of Nav1.4 with the less glycosylated Nav1.5 DIS5-S6 loop replacing the Nav1.4 DIS5-S6. A, G-V relationship. B steady state channel availability. C fast inactivation time constants. D time constants for recovery from fast inactivation. Figure from Johnson et al. 2004120.
48 Figure 3.6. The impact 1 has on Nav gating is likely through electrostatic interaction. Figure 3.6. A bar graph of the observed hyperpolarizing shifts in Va for hSkM1P1 1 SA with a 10-fold decrease in external Ca2+ concentration used to differentially screen external negative surface charges. Figure from Johnson et al. 2004120.
49 dependent on subunit sialic acids, but 1 sialic acids modified their gating. These data suggest that in this cellular syst em there may be a saturating limit to the contribution of sialic ac ids to channel gating, with Nav1.4 sialic acids possibly achieving saturation. Figure 3.6 further supports this theory by showing that gating of a less glycosylated Nav1.4 chimera, hSkM1P1 (a generous gift from Dr. A.L. George Jr.), is no longer dependent on subunit sialic acids but is sensitive to 1 sialic acids. These data suggest that by decreasing Nav1.4 sialylation below saturating levels, 1 can impact channel gating. Thus, it appears that the combined effects of cis subunit DIS5-S6 and trans 1 subunit functional sialic acids on channel gating are saturating. The surface potential theory of voltage-gated channel gating is often assigned to the phenomenon of negative external surface charges changing channel gating. It has been establishe d that increasing external Ca2+ concentrations tends to shift Nav gating to depolarized potentials. Ca2+ tends to screen the negative charges that cont ribute to the negative surface potential; thus, minimizing the external negative charge sensed by the channel gating mechanism. The voltage sensed by the channel gating mechanism becomes more negative, moving away from the vo ltage of half activation and requiring a larger depolarization to activate the channel If sialic acids contribute to this negative surface potential, c hannel gating will be more sensitive to external Ca2+ concentrations as the level of sialylation is increased. If 1 sialic acids contribute to the surface potential, co-expression of 1 with hSkM1P1 (a reduced
50 Figure 3.7. N-glycans are completely responsible for 1 effects on Nav gating. Figure 3.7. G-V relationships for hSkM1P1 (A), Nav1.5 (B), Nav1.7 (C), and Nav1.2 (D) under fully sialylated condi tions alone or co-expressed with 1 or with 1N. Filled circles with solid lines, subunit alone (n = 9Â–13 for each); filled squares with dashed lines, (n = 9Â–12 for each); filled triangles with dotted lines, co-expressed with 1N (n = 4Â–6 for each). Figure from Johnson et al 2004120.
51 glycosylated form of Nav1.4) in Pro5 cells should show the greatest sensitivity to Ca2+. Figure 3.7 clearly indica tes that the presence of 1 increases sensitivity to external Ca2+ concentrations; thus, 1 sialic acids likely contribute to the negative surface potential. Discussion To date, there have been many and subunit isoforms identified for the voltage-gated sodium channel, each with un ique glycosylation patterns that may modulate sodium current118,128. Expression of subunits and subunits are regulated over time and in disease st ates and can be processed differently among cell types possibly as a mechanism to ensure proper cellular function151155. The model in figure 3.8 suggests a scenario where Nav and 1 subunit combinations function differently and sial ylation could change the location on the curve as sialylation is changed. As proposed by this model, various subunit isoforms may function differently as a result of their level of glycosylation as ten isoforms have been identified each with a unique putative gl ycosylation signature. 1 expression, as previously described and in this study, alters some Nav channel isoforms; hence, control of 1 expression causes acute changes in f unctional sialic acids associated with an subunit possibly alteri ng the gating of the subunit.
52 Figure 3.8. Model proposing the satu rating effects of sialic acids on Nav gating. Figure 3.8. Model predicting the possible saturating effects of and 1 sialic acids on Nav gating. (A) Possible interactions between Nav1.4 (highly glycosylated) and hSkM1P1 (lesser glycosylated) subunit and 1 sialic acids. The data suggest that 1 sialic acids cannot contribut e further to the gating of Nav1.4 but do contribute to the gating of the other subunits through an apparent electrostatic mechanism. Thus, Nav1.4 shows ineffectual 1 sialic acids as distant, whereas the hSkM1P 1 illustrates that fewer subunit functional DIS5-S6 sialic acids may allow 1 sialic acids to interact more intimately with the subunit and contribute to channel gating. (B ) a theoretical G-V curve comparing contributions to Va associated with various and 1 combinations. The location of each combination is not precise but is consistent with the data shown here. Figure from Johnson et al 2004120.
53 1 expression is regulated through devel opment, commonly expressed at the highest levels after 4 weeks of age156-159. 1 is a heavily glycosylated protein; although, this level is different among the tissues in which it is expressed. If 1 glycosylation alters Nav gating, then Nav would likely be modified differently in each tissue in relation to the level of glycosyl ation. Here, we r eport the effect of 1 on Nav gating can be entirely attributed to sialic acid residues. A recent study of Nav function revealed a sialic acid dependent change in Nav activity throughout cardiac development123. These changes were independent of altered Nav protein expression as the same Nav isoform was expressed in all tissues studied and upregulation of 1 did not impact Nav function. Different levels of glycosylation, through glycogene regulation, likely is responsible for these changes and is the focus of the following two chapters.
54 CHAPTER 4 GLYCOGENE EXPRESSION IS REGULATED THROUGHOUT THE DEVELOPING MYOCARDIUM Slight changes in ion channel function may lead to devastating maladies such as myotonia, paralysis, epilepsy, long QT syndrome (LQTS) and arrhythmias associated with heart failure129,148,160-180. One possible mechanism to modulate Nav function is through alteration of the glycan structure. A previous report indicated that cardiac Nav gating is altered in a cell-specific, glycosylation dependent manner123. Furthermore, others report that glycosylation is altered in disease states, some of which present with altered excitability28,29,73-75,87,88,92. Thus, we determined that glycogene ex pression is regulated, and that the glycome is remodeled. Then, we questi oned whether and how these changes in glycosylation might alter ca rdiac excitability. Glycosylation abnormalities have been reported to occur in many disease states including heart failure. Until recently, in depth studies of gl ycogene expression in cardiac disease and non-disease states have been lacking. Here we present and compare the glycogene expression profiles for four healthy cardiac tissues; neonatal atria (NA), neonatal ventricle (N V), adult atria (AA) and adult ventricle (AV). This investigation utilized the GLYCOv2 gene chip, a customized array
55 designed by the Consortium for Functi onal Glycomics, contai ning 2174 probesets targeting 942 mouse transcripts encodi ng proteins responsible for glycan biosynthesis and glycan recognition, in cluding glycan transferases, glycan degradation proteins, pr oteins involved in nucleotide sugar biosynthesis, glycanbinding proteins, and transporters. To determine the relationship among t he four tissue types, we performed hierarchical clustering from all pr obesets, glycosyltransferases, glycan degradation proteins and nucleotide biosynt hesis (Figure 4.1). High correlation within each of the tissue types reveals mi nimal variability between replicates of each tissue type. Independent clustering of each tissue type indicates the glycogene profile is unique amon g the four tissue types. Hierarchical clustering showed that gl ycogene expression profiles are unique for each of the four tissue ty pes and these differences ar e evident when displayed as heat maps (Figure 4.2). Due to the lack of a control tissue studied, signals from each gene target are averaged and all relationships are compared to this average with red blocks indicating expression above the mean, blue blocks below the mean, and white as the mean. Despite the differences of overall glycogene expression, there are some important similarities to explore further. The patterns observed provide insight into the changing expression of glycogenes and their role in development, chamber specific expression or high expression in only one tissue type (lower in the three other tissue types).
56 Figure 4.1 Comparison of gl ycogene expression among samples. Figure 4.1. Unsupervised hierarchical clustering analysis of glycogenes in the developing murine myocardium. The dendrograms have been constructed using center correlation and average linkage. Th ree biological replicates are shown for neonatal atria (PA), neonatal ventricle ( PV), and adult ventricle (AV) and two biological replicates for adult atria (AA) (A) Overall glycogene expression. (B) Glycosyltransferases. (C) Glycan degradas es. (D) Nucleotide sugar synthesis and transporters. A. Overall D. Nucleotide Sugar Synthesis and Transport C. Glycan Degradases B. Glycosyltransferases A. Overall D. Nucleotide Sugar Synthesis and Transport C. Glycan Degradases B. Glycosyltransferases
57 Figure 4.2A. Glycosyltransferase expression throughout the developing myocardium. Figure 4.2. Heat maps displaying t he relative gene expression levels among each sample and cell type as measured from microarray data. Red indicates upregulation compared to the mean of all samples for a given glycogene whereas blue indicates downregulation a nd the individual glycogene mean is white. (A) Glycosyltransfe rases. (B) Glycan degradases (C) Nucleotide sugar synthesis and transporter genes. NV AA AV NA NV AA AV NA
58 Figure 4.2B. Glycan degradase expr ession throughout the developing myocardium. NA NV AA AV
59 Figure 4.2C. Nucleotide sugar synt hesis and transporter gene expression throughout the developing myocardium. NA NV AA AV
60 Three distinct expression patterns were identified in the data presented here. First, glycogene expression can be regulated developmentally. This is observed in figure 4.2 where the neonatal tissues ar e blue or red and adult tissues are the opposite color. Versican is developmentally regulated and is hi ghly expressed in the neonate and essentially absent in the adult. This is likely due to the importance of versican in the devel opment of the hear t and its diminished role in adults. On the contrary, ST3GalVI is also developmentally regulated, but present in the adult and absent in the neonate. Secondly, glycogene expression is also regulated in a chamber-specific manner where genes are expressed above t he mean in neonatal and adult atria and below the mean in the neonatal and adult ventri cle or vice versa. These changes indicate that the genes involved are ess ential to normal atrial or ventricular function but not both. Mannosidase-II is categorized with a chamber specific expression pattern since it is expressed at much higher levels in the atria than the ventricles. The third expression pattern identified is the high (or low) expression of a glycogene in one of the four tissue types com pared to the other three. This effect is likely a combination of the two patte rns discussed previously where the gene is developmentally expressed (i.e. only present in neonatal tissues) and in a chamber specific manner (i.e. expr essed only in the atria). The polysialyltransferase, ST8SiaII (STX), has been shown to be expressed in
61 various tissues in early developmental stages181,182, yet ST8SiaII is found to be expressed only in the neonatal atria and is absent in the neonatal ventricle as well as both adult tissues. Glycogene expression is quite different among the four tissue types and is regulated developmentally, in a chamber s pecific manner, or as a combination of the two. Furthermore, c hanging the expression patterns of different enzymes will alter glycosylation biosynthesis at va rious points throughout the process. Each glycoprotein and gl ycolipid has a common core structure which is elongated, possibly branched and eventually terminated. Figure 4.3 illustrates these core structures for Nand Oglycans and glycolipi ds. For further insight into the impact of a change in glyc ogene expression, four groups have been created: core structure synthesis termination, glycan degradation and nucleotide sugar synthesis of which some genes are represented twice; once in core structure synthesis and once in one of the other three groups. Core Structures As summarized in table 4.1A, most genes with roles in core glycan synthesis are regulated in a chamber-specific or dev elopmental manner, yet several have no significant change in expression. Most of those genes involved which have some change in expression are involved in the translocation of Nglycan structures from the cytoplasm to the lumen of the endoplasmic reticu lum. On the contrary,
62 Figure 4.3. The three basic glycosylation structures. Figure 4.3. Schematic of typical Nand O-linked gl ycans and glycolipids. Dashed lines indicate core stru ctures for each glycoconjugate. Galactose Mannose Glucose GalNAc GlcNAc Sialic Acid Fucose N S/T Cer Key: O-linked Glycolipid N-linked
63 Table 4.1A mRNA levels encoding protei ns involved in core structure synthesis. Gene Name NA NV AA AV N-Glycans ALG13 154.6 154.4 153.6 180.8 DPM1 417.1 444.4 407.9 514.4 DPM2 407.9 341.7 308.5 231.0 GPI1(PIGq) 391.6 410.8 436.3 750.0 PIG-a 76.8 70.6 54.5 50.2 Pigb 51.5 47.4 65.6 52.3 PIGF2 66.5 67.2 63.6 80.1 Putative PIG-M 151.6 153.6 159.0 149.8 Defender against cell death pr otein1 1200.01010.9 1246.5 886.4 Oligosaccharyltransferase 48 714.3 490.3 531.6 403.0 RibophorinI 1083.91023.0 869.3 722.6 RibophorinII 603.0 436.9 653.2 453.4 GNT1 473.2 517.3 464.0 448.0 GNT2 390.3 366.3 295.0 202.5 a-Mannosidase (Man2B1) 406.6 307.7 366.3 299.4 b-Mannosidase 320.1 332.0 302.6 257.5 MannosidaseII (Man2A1/ManII) 279.4 192.9 298.3 165.2 O-Glycans Galnt1(ppGalNAcT1) 623.4 598.7 826.0 1046. 5 Galnt2(ppGalNAcT2) 871.3 625.0 614.1 445.5 Galnt3(ppGalNAcT3) 12.2 12.1 12.1 12.8 Galnt4(ppGalNAcT4) 63.2 57.8 68.2 57.5 Galnt6(ppGalNAcT6) 49.2 47.7 56.5 49.9 Galnt7(ppGalNAcT7) 91.3 112.9 87.5 98.9 Glycolipids UDP-glucoseceramideglucosyltr ansferase188.6 138.2 127.2 100.4 ceramide1--galactosyltransferase 27.2 28.8 23.1 27.8 Table 4.1. Gene expression profiles of glycogenes directly involved in glycan synthesis. Intensity signals were generated with the dChip v1.3 PM-only algorithm and represent the mean of independently prepared samples. (A) Proteins involved in core structure syn thesis. (B) Sialyltransferases. (C) Sulfotransferases. (D) Fucosyltransfera ses. (B,C,D) are considered terminal glycosyltransferases. (E) Glycan degradas es; (F) Nucleotide sugar synthesis and transporters. NA, neonatal atria; NV, ne onatal ventricle; AA adult atria; AV, adult ventricle.
64 Table 4.1B. mRNA levels encoding proteins involved in sialylation. Gene Name NA NV AA AV ST3GalI 179.2142.1 195.6 160.3 ST3GalII 145.1121.9 122.4 101.7 ST3GalIII 170.2177.1 205.8 314.8 ST3GalIV 111.3126.9 132.8 108.0 ST3GalV 319.6266.6 562.4 1061.3 ST3GalVI 178.1190.5 409.8 372.6 ST6GalI 257.9139.1 198.3 82.0 ST6GalII 51.9 54.1 56.2 53.0 ST6GalNAcI 76.8 82.4 81.9 80.4 ST6GalNAcII 77.8 88.1 74.1 76.4 ST6GalNAcIII 40.1 39.9 41.8 49.7 ST6GalNAcIV 107.5118.3 85.1 94.8 ST6GalNAcV 111.0111.9 117.7 98.6 ST6GalNAcVI 259.0304.4 318.2 403.3 ST8SiaI 54.3 61.6 55.3 63.4 ST8SiaII 121.648.3 51.2 41.1 ST8SiaIII 61.9 64.3 77.4 107.9 ST8SiaIV 64.3 209.7 112.9 226.7 ST8SiaV 110.2106.0 120.4 98.5 ST8SiaVI 26.0 65.1 40.2 105.2
65 Table 4.1C. mRNA levels encoding proteins involved in sulfation. Gene Name NA NV AA AV chondroitin4-O-SulfoT1 178.2149.5212.9144.3 CS4ST-1 68.9 60.7 64.1 46.5 CS4ST-2 146.3108.2146.5103.3 CS6ST-1 66.4 64.8 70.8 71.5 Gal3ST-1 67.8 62.4 64.9 49.3 Gal3ST-2 50.7 53.0 59.3 45.9 Gal3ST-4 15.0 15.8 14.0 18.8 GlcNAc6ST-1 79.3 87.8 85.4 74.5 GlcNAc6ST-2 155.3136.2282.6195.5 GlcNAc6ST-3 34.0 36.4 36.6 35.3 GlcNAc6ST-4 79.8 105.583.4 96.6 HS2OST 354.5295.7211.2157.1 HS3OST-1 169.6107.6151.2105.1 HS3OST-3B 66.9 69.6 67.9 61.3 HS6OST-1 186.7179.1152.0155.4 HS6OST-2 53.9 55.9 43.1 45.3 HS6OST-3 58.0 56.3 55.9 58.3 KS6ST-1 284.2349.9363.0528.3 NDST1 112.771.3 90.8 54.4 NDST2 131.5125.289.0 93.6 NDST3 34.5 32.3 38.1 40.5 NDST4 43.6 43.4 43.8 43.9
66 Table 4.1D. mRNA levels encoding proteins involved in fucosylation. Gene Name NA NV AA AV Fut1 128.3 71.9 117.2 56.7 Fut2 57.6 55.8 62.3 62.9 Fut4 34.3 37.5 35.1 33.2 Fut7 79.7 80.9 77.0 72.1 Fut8 218.0 208.0 203.8 198.2 Fut9 11.0 10.7 11.7 13.0 Fut10 70.8 80.5 90.0 99.9 Fut11 216.4 166.7 208.9 130.1 Pofut1 133.8 114.6 122.5 101.3 Pofut2 302.9 240.2 233.6 161.5 Sec1 45.4 48.3 49.2 47.5
67 Table 4.1E. mRNA levels encoding proteins involved in glycan degradation. Gene Name NA NV AA AV ArylsulfataseA 259. 3 197.6 258.3 176.3 ArylsulfataseB 61. 4 63.5 61.1 68.2 alpha-GalactosidaseA 177.5 183.5 155.1 172.2 beta-Galactosidase(lactas e) 51.5 55.2 51.4 57.2 b-Galactosidase 354.0 271.2 296.5 207.8 b-Glucuronidase(Gus-s) 356.6 344.0 311.1 228.3 hexosaminidaseA 274. 1 238.5 328.4 242.5 Hyaluronidase1 63. 2 66.4 59.5 47.6 Hyaluronidase2 304. 3 267.2 206.5 181.2 a-L-iduronidase 197. 0 196.7 212.6 232.2 Acida1_4Glucosidase 481.1 311.0 644.0 515.1 AcidLipase 186.2 165.4 208.5 184.9 AcidSphingomyelinase 754.7 823.6 1308.2 1831.2 alpha-N-Acetylglucosamini dase 151.2 112.6 168.6 135.2 Asah 461.4 431.2 567.2 408.6 Cystinosis 178.3 166.8 170.2 151.6 Galactosylceramidase 37.9 38.6 34.6 40.6 Glucocerebrosidase(gba) 428.1 270.9 385.4 217.3 MPI 456.9 578.7 419.7 773.8 N-Aspartyl-b-Glucosamini dase 591.2 480.7 783.9 391.3 protectiveproteinforbeta-galac tosidase 854.5 820.4 1025.1 865.6 SialicAcidTransportProteinLAMP1 2931.23159.6 3297.9 3629.8 SialicAcidTransportProtei nLAMP2 489.3 303.0 474.3 276.8 a-Mannosidase(Man2B1) 406.6 307.7 366.3 299.4 b-Mannosidase 320.1 332.0 302.6 257.5 MannosidaseII(Man2A1/Man II) 279.4 192.9 298.3 165.2 acylneuraminatelyase 154.9 125.6 123.0 182.4 a-N-Acetyl-Galactosamini dase 245.4 177.5 237.0 143.0 GM2ActivatorProtein 248.5 215.0 269.9 199.6 Neu1 182.1 170.8 183.7 179.3 Neu2 67.6 64.6 73.1 57.7 Neu3 125.0 137.7 157.5 183.5 Galactosamine-6-Sulfat ase 109.5 88.7 96.0 74.8 IduronateSulfatase 50.2 59.6 64.9 77.0 SULF1 727.8 389.2 677.6 251.2 SULF2 864.6 939.4 695.5 401.4 N-sulfoglucosaminesulfohydr olase 66.2 68.3 72.5 72.4
68 Table 4.1F. mRNA levels encoding proteins invo lved in nucleotide sugar synthesis and transport. Gene Name NA NV AA AV CMP-sialicacid 272. 9 238.9 178.5 174.4 UDP-Galactosetransporter 154.6 147.9 173.6 156.1 UDP-galactosetransporterrelated 1157.01057.4 791.7 884.7 UDP-GlcNActransporter 107.7 115.9 90.9 81.7 CMP-N-acetylneuraminicacidsyn thase 813.5 561.8 714.7 585.7 CMP-Neu5Achydroxylase 66.6 72.1 97.6 94.0 epimerase 213.6 159.3 196.2 116.4 Fucose-1-phosphateguanylyltransfe rase 87.5 88.6 96.4 82.8 galactokinase(galK) 295.8 293.8 174.0 151.1 Galactose-1-phosphateuridylyltrans ferase 332.7 514.0 441.0 631.7 GDPfucosesynthetase 155.0 151.1 150.0 222.0 GDP-man4-6dehyd 142.7 107.5 112.4 80.0 GDP-mannosepyrophosphorylas eA 409.6 293.4 628.3 363.7 GlcNAc/ManNAckinase 348.7 372.4 270.9 297.2 GlcNAc2-epimerase 330.1 310.0 381.8 325.9 glucosamine-6phosphatedeaminase/isomerase 146.5 175.9 160.7 171.3 glucosamine-phosphateN-acetyltr ansferase 167.2 175.8 114.8 105.0 glucosephosphateisomera se 3481.43577.7 3231.83325.8 Glutamine-fructose-6phosphatetransaminase1 227.9 271.7 249.0 380.3 Glutamine-fructose-6phosphatetransaminase2 65.7 70.0 96.3 137.3 hexokinase1 915.5 963.7 648.3 737.8 ketohexokinase(fructokinas e) 134.2 121.8 155.1 171.7 Neu5Ac9-phosphatesynthase 320.7 293.4 188.2 146.7 PAPSsynthetase-1 407.2 395.7 211.2 150.9 PAPSsynthetase-2 75.2 75.6 105.2 75.9 phosphoglucomutase1 821. 8 1341.0 741.4 1767.4 phosphomannomutase 156.0 166.4 234.7 305.4 phosphomannomutase1 185.6 151.9 187.4 163.7 pyrophosphorylase 622.6 503.4 589.9 621.1 UDP-Gal-4-Epimerase 86.8 83.3 78.0 65.8 UDP-GlucoseDehydrogenas e 565.9 358.3 452.7 165.0 UDP-GlucuronicacidDecarbox ylase 144.2 130.2 142.0 153.3 uridinediphosphoglucosepyropho sphorylase2515.9 543.3 386.1 700.7
69 five of the seven O-glycan associated enzymes do not change. Of the O-glycan associated enzymes that do change (Galnt 1 and Galnt2), Galnt1 is upregulated in the atria compared to the ventri cle of both neonatal and adult tissues and Galnt2 is higher in the adult tissues. Terminal Structures Core glycan structures are elong ated, and often, branches are added contributing to variation of N-glycan st ructures that differs among cell types (Comelli et al .86 and addressed in chapter 5). Extracellular communication and any modulatory effects of glycans are likel y due not only to structural variation, but also, and possibly more importantly, to terminal re sidues. Not surprisingly, terminal glycosyltransferase expression is highly variable from tissue to tissue. Tables 4.1B, 4.1C and 4.1D illustrate these changes in sialyltransferase, sulfotransferase and fucosyltransferase expression among the four tissues. Glycan Degradation Glycan degradation enzymes are intimately involved in glycan biosynthesis as displayed in (Figure 1.4). Removal of glucose and mannose residues allows Nglycosylation to proceed in the endoplasmi c reticulum and golgi apparatus after the structure is transferred from dolic hol to asparagine of a glycoprotein. Incomplete or improper removal of t hese glucose and mannose structures leads to improper glycosylation structures and likely a pathological disorder such as
70 CDG. It should also be noted that degradases included in the GLYCOv2 chip are active at various time point s of the glycan lifetime including in the lysosome. As summarized in table 4.1E, the expressi on of glycan degradases varies among all tissues with no discernible pattern. Each tissue is the highest expresser of at least one gene and the lowest ex presser of others. Nucleotide Sugar Synthesis and Transporters Enzymes classified under the category of nucleotide sugar synthesis have a role in creating and transporting the sugars t hat glycosyltransferases add to the glycan structure. These enzymes include transporters that bring the sugars into the proper organelle, epimera ses, isomerases, synthases and other enzymes directly involved in nucleotide sugar synthes is. These proteins and their relative expression levels are summarized in tabl e 4.1F which shows varying expression of these genes and that all three expression patterns are present. Tissue Type Comparison Chamber and developmental effects can be elucidated using four comparison groups: neonatal atria (NA) vs neonatal vent ricle (NV), NA vs adult atria (AA), AA vs adult ventricle (AV), and NV vs AV. The two remaining comparison groups: NA vs AV and AA vs NV are neither c hamber specific or developmental comparisons and were excluded from this analysis.
71 Figure 4.4A. Glycosylatransferases di fferentially expressed throughout the developing myocardium. Figure 4.4. Differential expression of glycogenes directly involved in glycan synthesis; (A) glycosyltransferases, (B) glycan degradases, and (C) nucleotide sugar synthesis and transporters. Left panel shows expression relationships of all glycogenes in each category. Right panel shows glycogenes considered to be differentially expressed at p<0.01. NV AA AV NA NA NV AA AV
72 Figure 4.4B. Glycan degradses diffe rentially expre ssed throughout the developing myocardium. NA NV AA AV NA NV AA AV
73 Figure 4.4C. Nucleotide sugar synthes is and transporters differentially expressed throughout the developing myocardium. NA NV AA AV NA NV AA AV
74 Figure 4.5. Differential expres sion of glycogenes by category. Figure 4.5. Each comparison group s hows the % of genes within a type of glycogene that were found to be different ially expressed. (A) Differential expression patterns when all four tissue ty pes are compared. (B) and (C) show the chamber specific differences in neonate and adult respectively. (D) and (E) reveal the developmental changes in glycogene expression by category. Note the patterns of all comparisons are unique. AA:AV0 10 20 30 40 50 60G ly c an-transferase G ly c an Degra dati on N u c S u gar C B P:CTy pe L ecti n C BP : I-T y pe Le ctin CBP:S-Type Lectin Gly c oprote i n Not c h p at hway intra c ellular prote i n trans p ort A dhesio n Mole c ule C h emok i ne Cytokine Gr o wt h F act ors & Rece pto rs Interle u kin & Rec e ptors Proteogl y ca n% Change AA:NA0 10 20 30 40 50 60Gl y can-transfe r ase Gly c an D e gr a d a ti o n Nu c. Su g a r CBP : C-Type Lectin CBP:I-Typ e Lectin CBP: S-Ty p e L e c tin Gl y c o prot e in Notch pathway i n t ra cel l u l ar p rote i n t ra n s p o rt Ad hesion Mol e c u l e Chemoki n e Cytokine Growt h F a c t o rs & Re c e p tors I nt e rl e u k in & Recepto rs Proteoglycan% Change AV:NV0 10 20 30 40 50 60Gly c an-tran s fe ra se G l y can Deg ra da t ion Nuc. Sugar CBP:C-Type L ectin CB P: I-Ty p e Lectin CBP:S-Type L ectin Glyco p rot e in N o tc h p a thwa y in tra c e ll u la r p rotein tra ns p o rt Ad h es i on Mo l ec u le C h e m ok i n e Cyt o kine Growth Fact o rs & R eceptors In te rle u k in & Re c ep to rs P r ot eo gl y ca n% Change NA:NV0 10 20 30 40 50 60Glyc a n -tra n sfe ra s e G l y c a n D e gradation Nu c S ugar C B P :C-T y pe Lectin CBP:I-Typ e Lectin CBP:S-Type Lectin Glycoprotein Notc h pathwa y intra c e l lula r prote i n t ra n s p o rt Adhesion M olecu l e Ch e m o k ine Cy t o k i n e G r owt h F a c t or s & Re c e p t o rs In t erl e u k i n & Recept o rs P rot e o g lyc a n% Change B C D E A Overall0 10 20 30 40 50 60 70 80 90 100G lycan-transferas e G lycan Degrad ati on N uc. Sug a r CBP:C-T ype Lec tin CB P:I-Type L ec tin CBP:S -T ype Lec tin Glycoprotei n N otch pathwa y intracellular pr otein t rans po Ad hes ion Molecu le Chem okin e Cy tokin e Growth F ac tors & Recept ors Inte rl eukin & Recept or Proteoglyc a n % Change
75 Glycogene expression is largely different in each of the four comparison groups and is detailed below. Overall, 419 of the 710 genes are differentially expressed in at least one comparison group. These changes are summarized in figures 4.4 and 4.5 which reveals that t he proportion of diffe rentially expressed genes is higher than any single com parison group; indi cating that one comparison group does not encompa ss all differentially expressed genes. Surprisingly, the three gene groups directly involved in glycosylation (glycosyltransferases, glycan degradases and those involved in nucleotide sugar synthesis and transport) show that ~ 46% (110 of 239) of these genes are differentially expressed at p<0.01 (Figure 4.4) among th e four myocyte types. Glycogene expression profiles of two tissue types were compared to attempt to identify major developmental and cham ber specific changes. Table 4.2 summarizes these changes and shows overall changes between groups and the changes in specific glycogene categorie s. These data, together with the heatmap data (Figure 4.2 and 4.4), ident ify a large proportion of overall gene expression changes between tissue types. Specific details of differential expression with each relevant myocyte comparison group is discussed below. Chamber-Specific Regulation Neonatal Atria and Ventricle Expression of glycogene targets in the neonate varies between the atria and ventricle. Specifically, expression of 161 of 710 (22.7% ) gene targets are
76 Table 4.2. Differential glycogene expres sion profile NA:NV AA:AV AA:NA AV:NV Totals n % n % n % n % Glycantransferase 31 16.450 26.522 11.6 52 27.5 189 Glycan Degradation 10 25.620 51.35 12.8 18 46.2 39 Nucleotide Sugar Synthesis and Transport 8 22.215 41.710 27.8 20 55.6 36 CBP:C-Type Lectin 15 13.832 29.417 15.6 37 33.9 109 CBP:I-Type Lectin 3 18.85 31.31 6.3 4 25.0 16 CBP:S-Type Lectin 3 21.45 35.73 21.4 6 42.9 14 Glycoprotein 2 20.04 40.00 0.0 2 20.0 10 Notch pathway 7 31.85 22.74 18.2 6 27.3 22 Intracellular protein transport 0 0.0 1 14.30 0.0 1 14.3 7 Adhesion Molecule 4 50.03 37.52 25.0 4 50.0 8 Chemokine 6 9.8 8 13.19 14.8 24 39.3 61 Cytokine 1 7.1 2 14.31 7.1 5 35.7 14 Growth Factors & Receptors 45 25.356 31.530 16.9 52 29.2 178 Interleukin & Receptors 4 9.1 8 18.27 15.9 12 27.3 44 Proteoglycan 12 40.011 36.79 30.0 14 46.7 30 Table 4.2. Summary of the differentia l expression of glycogenes by category. Both the number and the overall percent of glycogenes differ entially expressed are displayed (p<0.05).
77 significantly altered (p<0. 05). Interestingly, 118 of the 161 (73%) differentially expressed gene targets are expr essed at higher levels in the neonatal ventricle. Furthermore, of the genes where expression is considered to be highly different (>1.3-fold change), the proportion re mains at approximately 73% (100 gene targets) are in the ventricle. When the p-value is decreased to 0.01, 99 genes are still considered differentially express ed. Differences in glycogene expression between atria and ventricles are evi dence of the importance of minor adjustments required for proper cellular function (discussed fu rther in chapter 6). Adult Atria and Ventricle The second chamber specific compar ison between the adult atria and ventricle shows another large change in glycogene tar get expression with 253 of the 710 targets (35.6%) altered. Within this comparison, 39.1% of the differentially expressed gene targets are up-regulated in t he atria. These figures are modified only slightly when only highly altered (>1.3-fold change) gene targets are analyzed with 36.7% (66 gene targets) of the highly differentially expressed gene targets are in the atria. Developmental Regulation Adult and Neonatal Atria Glycogene expression between n eonatal and adult atria indicates that 19.4% (p<0.05) (138 of 710 gene targets) of gene ta rgets are significant ly differentially expressed. 55.1% of these genes are up -regulation in the adult atria compared
78 to the neonatal atria. Of t he highly differentially expressed genes (>1.3 fold change), these proportions remain approx imately the same at 56% and 46% respectively. Adult and Neonatal Ventricle The largest change in glycogene target expression is found between neonatal and adult ventricle. 307 of 710 (43.2%) gene targets were significantly differentially expressed with approximately an equal nu mber of gene targets upregulated in the neonatal (154 gene targets) and adult (153 gene targets) ventricles. Among the highly altered gene targets (>1.3 fold change), neonatal ventricle has higher expressi on of 115 gene targets compared to 108 for the adult ventricle. Quantitative PCR veri fies microarray data Gene chip verification is an important process for quality control and duplication of expression levels. Here, we verified each comparison group using three distinct genes for each comparison for a total of twelve. These included sialyltransferases involved in N-glycan synthesis or O-glycan synthesis and growth factors. Tissue type comparis ons of the expression of each of the twelve glycogenes are consistent with genechip findings compared to HPRT (Figure 4.6). These data were also anal yzed by comparing expression levels to -actin as the control with sim ilar results (data not shown).
79 Figure 4.6. qPCR validates the GeneChip microarray data. Figure 4.6. Selected genes were inve stigated by qPCR analysis of RNA to validate GeneChip microarray data. Three glycogenes from each comparison group were studied, revealing that dat a from both microarray and qPCR were consistent. All genes were normalized to the endogenous control gene, HPRT. Fold change in relative expression level qPCR CT (HPRT) Microarray p<10-7p=5.09 X10-6p=1.4 X10-3p=1.4 X10-4p<10-7p<10-7p<10-7p<10-7p<10-7p<10-7ST8S ia 6 ST6Gal 1 S TX ST3Gal5 V e rsican ST3 G a l3 ST3G a l5 Ve rsic a n S T 3Gal3 ST6 GalNAc 6 ST8Sia6 STXp=3.6 X10-6p=9.54 X10-4p=8.2 X10-4p=4.31 X10-4p=3.34 X10-3p=7.06 X10-3p=1.97 X10-5p=1.23 X10-4p=7.66 X10-3p=1.83 X10-2p=5.67 X10-3NV/NA RatioAA/NANV/AVAA/AV RatioNA/NVNA/AAAV/NVAV/AA p=3 X10-4p=4 X10-7p<10-715 10 5 0 5
80 Discussion Glycogene expression is a developmentally regulated process in the myocardium with large changes in both the developing atria and ventricle. Significant differences are apparent within the devel oping atria with 19.4% of the genes differentially expressed yet the developi ng ventricle shows more than twice as many genes differentially expressed at 43.2%. Not surprisingly, glycogene expression is si gnificantly regulated between cardiac chambers. Differential expression of glycogenes between chambers at the same developmental stage range from 22.7% (neonates) to 35.6% (adults) suggesting that glycan structures are modified between chambers. Glycogene expression is differently regul ated among the four myocyte types indicating that the glycome is remodeled throughout the developing myocardium. The high levels of variation among the four comparison groups provides insight into the possible changes in glycan structure at a cellular level.
81 CHAPTER 5 THE GLYCOME IS REMODELED T HROUGHOUT THE HEART DURING DEVELOPMENT With such diversity in glycogene expression, one would expect very different populations and relative quantities of glycans synthesized by the cell. Mass spectrometry is a powerful tool fo r identifying the populations of glycans present in a given sample. The resulting spectra provide insight into changes in N-glycan profiles produced throughout t he developing heart. Complete mass spectra are shown in figures 5.1 and 5.2. High Mannose Structures The intensity patterns of the first five major high mannose structures (1579, 1783, 1987, 2192 and 2369) are almo st identical among all four tissue samples: neonatal atria, neonatal ventricle, adult atria and adult ventricle (blue peaks, Figures 5.1 and 5.3). T he high mannose structure at m/z 2369 is the most common structure of the five high mannose structures and is most common overall in three of the four tissues composing over 50% of the total glycan density. The adult atria is the exc eption with the high mannose structures composing only 23.5% of the total gl ycan population (Table 5.1). When comparing the pattern of only high m annose structures, m/z 1987 is the least
82 Figure 5.1A. The population of N-glycans is different among the four myocyte types. 0 100 150022002900360043005000 Mass ()m/z% In te n s ity50AV N-Glycans1579.1 1783.1 1987.2 1835.1 2192.1 2244.1 2039.1 2081.2 2396.0 a = 2420.0 b = 2447.1 c = 2461.1 d = 2605.2 2635.1 e = 2621.2 3026.1 2996.3 2966.3 2852.1 2839.2 2809.3 3243.2 2285.1 2652.1 3306.2 f = 3417.2 j = 3493.2 n = 3807.2 o = 3838.2 m = 3779.2 g = 3446.2 h = 3462.1 i = 3476.2 k = 3619.2 p = 3867.2 3693.2 l = 3649.2 a b c d e f ghi j kmno p l 0 100 150022002900360043005000 Mass ()m/z% In te n s ity50AA N-Glycans1579.2 1783.2 1836.2 1987.2 2039.2 a = 2069.2 b = 2081.2 2192.2 2285.2 2396.2 d = 2431.2 f = 2489.2 2605.2 2839.3 3243.3 c = 2110.2 g = 2792.6 2652.5 2243.2 2390.2 e = 2461.2 2635.2 2852.2 3026.2 2996.3 3084.3 2880.2 2966.3 2809.3 h = 2822.3 a b c d e f g h 4085.2 y = 4616.4 x = 4587.2 z = 4648.1 3867.3 t = 3837.2 s = 3807.3 r = 3777.3 p = 3680.3 o = 3650.3 3693.2 3417.2 i = 3272.3 j = 3301.3 k = 3446.3 l = 3463.3 m = 3476.3 n = 3493.4 w = 4258.3 v = 4226.2 u = 3895.2 q = 3722.2 i jk l m n o p q r s t uvwx y z 0 100 150022002900360043005000 Mass ()m/z% In te n s ity50NA N-Glycansf g h i j k m n o p q r s t uv w x y z l a b c d e 1579.5 1783.6 1987.7 2192.7 2244.7 2285.7 2396.8 2605.8 2809.9 2852.9 2966.9 2996.9 3026.9 3056.0 2822.9 2839.8 2635.8 2652.8 2489.8 2448.7 a = 2028.7 b = 2039.7 c = 2069.7 d = 2081.7 1835.7 2793.8 k = 3464.0 l = 3478.8 m = 3493.0 n = 3504.2 o = 3621.0 p = 3650.9 y = 4590.5 z = 4619.6 w = 4432.5 x = 4462.3 e = 3259.9 f = 3274.1 g = 3289.0 h = 3306.8 i = 3417.0 j = 3446.9 q = 3777.9 r = 3808.0 s = 3836.9 t = 3867.0 u = 4227.8 v = 4275.3 0 100 150022002900360043005000 Mass ()m/z% In te n s ity50NV N-Glycansn = 3271.8 o = 3288.8 p = 3305.8 q = 3415.9 r = 3445.9 t = 3492.9 u = 3503.9 s = 3462.9 w = 3619.9 v = 3550.9 x = 3632.9 y = 3647.9 z = 3660.9 = 3691.0 = 3708.0 = 3777.9 = 3808.0 = 3836.9 = 3755.0 = 3865.1 = 3895.1 = 3912.1 = 4069.1 = 4099.1 = 4116.2 = 4432.5 = 4226.2 = 4273.2 = 4462.3 =4303.3 = 4617.4 = 4565.4 = 4587.4 = 4722.5 f g h ijk mnop q r s t u v w x y z l a b c d e 1579.5 1783.0 1988.1 2192.2 2244.2 2285.3 2396.3 2605.4 2809.5 2852.5 2966.6 2996.6 3026.6 3054.7 2822.5 2839.5 g = 2635.4 h = 2652.5 f = 2489.4 e = 2448.3 a = 2029.1 b = 2040.1 c = 2070.2 d = 2081.2 1836.0 2792.5 2693.5m = 3258.8 i = 3084.7 j = 3101.7 k = 3142.7 l = 3241.8 Figure 5.1. (A) Mass Spectrometry prof iles of N-glycans in neonatal and adult atria and ventricles utilizing MALDI-TOF MS. Note th at the lower MW structures up to high mannose structures are at relatively high density for each myocyte type (blue peaks). Significant variation in comple x glycan structures among myocyte types is readily apparent (red peaks), both in relative levels and types of glycans. Predicted glycan struct ures for highlighted peaks are shown in neonatal ventricle panel. (B and C) Nglycans associated with masses reported in figure 1A. (B) Glycans with masses between 1500 and 3050 m/z, (C) structures with masses above 3050 m/z. Structures that list the mass in highlighted yellow ovals were determined using MALDI TOF/TOF analyses. Blue square, GlcNAc; green circles, mannoe; ye llow circles, glucose; red diamonds, NeuAc; light blue diamonds, NeuG c; red triangles, fucose.
83 Figure 5.1B. Identified low mass N-glycan structures and their relative mass.
84 Figure 5.1C. Identified high mass N-glyc an structures and their relative mass.
85 Figure 5.2A. Mass spectra of the neonatal atrial N-glycans. Figure 5.2. MALDI-TOF mass spectra for each tissue type. Each spectrum is an enlarged version of those displayed in fi gure 5.1 to better show detail. Masses correspond to structures in figure 5.1B and 5.1C.
86 Figure 5.2B. Mass spectra of the neonatal ventricular N-glycans.
87 Figure 5.2C. Mass spectra of the adult atrial N-glycans.
88 Figure 5.2D. Mass spectra of the adult ventricular N-glycans.
89 Figure 5.3. Mass spectra of masses between 1500 and 2400 m/z. Figure 5.3. Enlarged schematic of spectra between 1500 and 2400 m/z. The blue peaks indicate relative density of high mannose structures.
90 Figure 5.4. Mass spectra of masses between 2400 and 3050 m/z. Figure 5.4. Enlarged schematic of s pectra between 2400 and 3050 m/z. The red peaks indicate relative density of bi-ant ennary structures assigned to one of three groups.
91 Figure 5.5. Spectra of masses above 3050 m/z. Figure 5.5. Enlarged schematic of s pectra above 3050 m/z. These peaks are associated with the most comple x and highest mass structures.
92 Table 5.1. Relative percentage of glycan structures defined by either structure (high mannose) or mass. NA NV AA AV High Mannose 52.06 60.43 23.53 52.95 2400-3050 m/z 37.63 23.24 57.38 35.91 3050-5000 m/z 1.45 8.98 10.44 1.37 Table 5.1. Relative density of glycan stru ctures defined by ei ther structure (high mannose) or mass. Note the low pr oportion of high mannose glycans and high proportion of higher mass structures in the adult atria (AA) compared to all other groups. Although relative density of glyc ans in the neonatal atria (NA) and adult ventricle (AV) are similar, the mass spectra are much different as seen in figure 5.1, 5.2A and 5.2D. Neonatal ventricle, NV.
93 common while 1783 and 2192 are higher and si milar in level of expression in all tissue types. Complex Structures Complex structures, defined here as thos e structures that have been processed beyond high mannose stages and have a ma ss above 2400 m/z, are present in unique relative quantities among all four tissue types. The structures with masses between 2400 and 3050 m/z are bi -antennary complex structures (Figure 5.4) while those with masses abov e 3050 m/z are mostly triand tetraantennary (Figure 5.5). Chamber-specific glycan profile changes Neonatal Atria and Ventricle The neonatal atria and ventricle spectra are surprisingly similar in terms of glycans present or absent, yet the level of each glycan varies. The neonatal atria have higher relative levels of al most every glycan between 2400 and 3050 m/z accounting for approximately 37.6% of t he overall glycan population compared to 23.2% for the neonatal ventricle (Table 5.1). Of particular interest in this region are three sets of sialylated glycans (Figure 5.4): group 1 consists of glyc ans at 2605 and 2635 m/z, group 2 is composed of glycans at 2793, 2822 and 2852 m/z and group 3 is comprised of glyc ans at 2966, 2996 and 3026 m/z.
94 In the neonatal atria and ventricles, groups 1 and 3 decrease in relative level from low m/z to high m/z and group 2 increa ses in relative glycan level from low m/z to high m/z. The greater number of unique higher mass (above 3050 m/z) N-glycan structures are present in the neonatal ventricle with several complex structures produced by the ventricle but not the atria (Figure 5.5) In the higher mass region the relative abundance is higher in the ventricle accountin g for 8.9% in contrast to 1.4% in the neonatal atria (Table 5.1). Thus, it is likely that the neonatal atria produce the same bi-antennary structures as the ventricle but at a higher relative abundance while the neonatal ventricle produce more tr iand tetraantennary structures at higher relative levels than the neonatal atria. Adult Atria and Ventricle The population of glycans present in t he adult atria and ventricle are comparable in some manners and different in others. Note for the adult atria, the data are normalized to the 2852.2 m/ z peak whereas the other three tissue types are normalized to the high-m annose structure around 2396 m/z (Figure 5.1). Despite this change, the spectral patterns bet ween 2400 and 3050 m/z for adult atria and ventricle are very similar with adult atri a showing higher (57.4% compared to 35.9%) relative levels of each glycan (T able 5.1). Groups 1, 2 and 3 are similar in pattern in that the relative level in creases from low to high m/z; although, the
95 adult ventricle has extremely low relative expression levels of 2793 and 2822 m/z (i.e., barely greater than bas eline (Figure 5.4)). Higher mass structures are present in greater relative abundance and number of species in the adult atria than in the adul t ventricle (Figure 5.5). The most complex glycan registers wit h a minor peak at 3867.2 in the adult ventricle, yet the adult atria had a higher relative leve l of the glycan at that mass and six glycans register at a higher mass t han 3867.2. These high mass structures compose approximately 10.4% of the adult atria overall glycan population compared to 1.4% for the adult ventricle (Table 5.1). Developmental glycan profile changes Neonatal and Adult Atria In contrast to the similarities of age matched comparisons, developmental differences are more apparent. The put ative structure of middle mass, biantennary glycan structures present are si milar throughout atrial development but with much higher relative levels in the adult atria for most struct ures (Figure 5.4). One major exception is the glycan at 2809 m/z which is the highest peak above 2397 in the neonatal atria whereas this peak is relatively minor in the adult atria. Groups 1 and 3 show opposite patterns with neonatal atria decreasing relative levels from low to high m/z and adult atria increasing relative levels from low to high m/z. Surprisingly, group 2 has an identical pattern in both neonatal and adult atria with increasing relative leve ls from to low to high m/z.
96 The higher mass range of both adult and neonatal atria has similar glycan structures, yet each tissue type has some structures that are unique (Figure 5.5). The relative levels of these glyc ans is higher in the adult atria with over 10% of glycan structures with higher ma ss than 3050 m/z compared to neonatal atria with higher mass composing only appr oximately 1.4% of the total glycan structures (Table 5.1). Neonatal and Adult Ventricle The developmental changes between adult a nd neonatal ventricle are prominent in the mass spectra shown in figure 5.1. The portion of glycans in the 2400 to 3050 m/z range is much higher in the adul t ventricle (35.9%) compared to the neonatal ventricle (23.2%), but the pattern of expression is also changed (Table 5.1). Groups 1 and 3 have opposite patterns with the neonatal ventricle increasing relative level from low to high m/z and adult decreasing relative level from low to high m/z (Figure 5.4). Group 2 has an identical pattern in both adult and neonatal ventricle. In contrast, the neonatal ventricles hav e a much higher proportion (8.9%) of glycans in the higher mass range than the adul t ventricle (1.4%) (Table 5.1). The neonatal ventricle also has 37 unique high mass glycan structures compared to the adult ventricle with only 13 (Figure 5.4). This change indicates a shift towards bi-antennary glycans in the adult ventricle.
97 Discussion Regulated glycogene expr ession throughout the heart during development would lead one to expect the glycan profile to also be markedly varied among myocyte types. In fact, each tissue has a unique gl ycan profile with notable similarities among all tissues. All tissues have an identical pattern of the identified high mannose structures indicating that high mannose structure production is similar in all four tissues. The microarray dat a correlates quite well that expression levels of enzymes responsible for high mannose synthesis and pruning are comparable across the four tissue types. There are over 30 different forms of sialic acid produced in nature. In the mouse there are two common sialic acids attac hed to glycans, N-acetylneuraminic acid (NeuAc) and N-glycolylneuraminic acid (NeuG c). These sialic acids seem to be developmentally regulated in mouse myocar dium. Adult tissues seem to produce glycans that add NeuGc preferentially over NeuAc, as shown in groups 1, 2 and 3. Alternatively, this conclusion cannot be made in the neonates since the dominant peak in these groups varies between those with NeuAc (groups 1 and 3) and NeuGc (group 2). These changes are examples of the developmental modifications of glycan structures. Glycan profiles are markedly different among the four tissue types tested in two of the three glycan mass categories wit h only the high mannose patterns being similar among the four tissue types. C hanges in glycan profiles among the four
98 tissues are prevalent above 2400 m/z indica ting that the modifications in glycan structure are imposed in the latter steps of the glycosylation pathway. Biantennary glycans are similar in st ructure throughout the developing myocardium, yet the relative levels of these glycans change in both in a developmental and chamber-specific manner. Furthermore, the largest variations in the number of glycan structures are at a mass above 3050 m/z where triand tetraantennary structures are located. The glycogene expression data are consistent with these data in that th e majority of glycogenes active in the distal golgi apparatus are most commonly different ially expressed.
99 CHAPTER 6 THE REGULATED EXPRESSION OF A SINGLE POLYSIALYLTRANSFERASE IMP ACTS CARDIAC EXCITABILITY Sialic acid residues are the primary terminal residues and are added to the glycan structure through sialyltransferase activity. Through polysialyltransferase activity, the level of sialylation is gr eatly increased. Polysialyltransferase enzymes are responsible for addition of sialic acids to sialic acids creating long chains from 5 to 100 residues termed polysialic acid. Expression of the polysialyltransferase, ST8 alphaN-acetyl-neuraminide alpha-2,8sialyltransferase 2 (STX), is highly regul ated in the developing myocardium. As seen in chapter 4, STX is expressed at much higher levels in the neonatal atria compared to the neonatal ventricle, adul t atria and ventricle where STX is essentially not expressed. To as certain whether a single enzyme whose expression is regulated may affect cardiac ex citability, we re corded action potentials from control and STX knockout mice. The results of these studies indicated that Nav function may be modified by t he regulated expression of STX thereby altering atrial action potentials Here, we also question whether the regulated expression of a single sialyltr ansferase is sufficient to alter Nav gating.
100 The effects of sialic acid residues on the voltage dependenc e of voltage gated sodium channel (Nav) gating has been extensively studied 117,118,120,123. Thus far, it has been determined that sialic acid residues modulate ion channels in an alpha (pore-forming) subunit manner and th rough the expression of the auxillary subunit, 1. For Nav, the mechanism by which sialic acids modulate channel gating is through an apparent electros tatic attraction between the negative surface potential (to which the negatively ch arged sialic acid residues contribute) and the positively charged amino acid s of the channel's voltage sensors 117,118,120,123. Stocker and Bennett determined t hat the sodium channel isoform expressed in neonatal ventricles was le ss-sialylated compared to the same sodium channel isoform in the neonatal at ria, adult atria and ventricle indicating another possible mechanism for the cell to manipulate channel gating through changing the glycans attached to the proteins123. In this study, we have identified large changes in glycogene expression and glycan structure; consistent with a global mechanism by which cardiac func tion is altered by a regulated glycome. The neonatal atrial action potential w aveform is altered wh en STX is absent To question whether the regulated expre ssion of a single sialyltransferase can alter cardiac excitability, action potential waveforms were recorded from neonatal atria and ventricles of control and ST X knockout mice. Proper voltage-gated sodium, potassium and calcium channels ar e essential to initiate and propagate the action potential in cardiomyocytes. Neon atal atrial action potential recordings reveal a rate of depolarization (dV/dt) that is 65% slower in the knockout atria
101 Figure 6.1. Expression of STX modifies neonatal atrial, but not ventricular AP waveform. AP Peak Amplitude (mV) 100 110 120 130 dV/dt (normalized) 0.0 0.5 1.0 *AtriaVentricle+ STX + STXSTX STX Figure 6.1. Measured param eters of the acti on potential waveform. Bar graphs of the mean +_ S.E.M. (A) The rate of action potential depolarization. (B) The maximum AP depolarization. (C) Repres entative action potential traces from STX control (red) and kno ckout (blue) atria. A. B. C.
102 compared to the littermate control atria (F igure 6.1). The maximal depolarization of the knockout atria is also 30 mV le ss than the control atria as determined by peak amplitude (Figure 6.1). These tw o measurements tend to be associated with the portions of the action potential produced by sodium currents and are consistent with how Nav dysfunction changes excitability with STX expression. That is, the lack of STX expression in the knockout would cause the channel to gate at more depolarized potentials. These changes lead to a slower action potential depolarization rate Because the time to reach peak amplitude is increased in the absence of ST X, a higher percentage of Kv may be active at the peak of the action potential. Increased K+ currents during the rising phase of the action potential would offset Na+ currents and effectively decrease the peak amplitude. No changes in action potential waveform were observed in ventricular myocytes; as expected since STX is not expressed in the ventricle. The voltage dependence of Nav gating changes only in the neonatal atria of the STX knockout. To determine whether the absence of a singl e sialyltransferase, STX, can modify Nav gating, Na+ currents were recorded from at rial and ventricular myocytes isolated from STX knockout mice and comp ared to littermate controls. Neonatal tissues are ideal to elucidate whether STX can modulate Nav gating since STX is expressed several times more abundantly in the atria than in the ventricle and both tissues are isolated from the heart of the same animal. The Nav Va and Vi measured in the knockout atrial myo cyte were 7-9 mV more depolarized and
103 Figure 6.2. STX causes a change in neonatal atrial Nav activation voltage, but not in ventricular Nav activation. Figure 6.2. Conductance-voltage relati onships for control and STX knockout atrial and ventricular Nav. Data are the mean normalized peak conductance S.E.M. Representativ e current traces are shown to the right. Membrane Potential (mV) -80-60-40-20Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Control Atria Knockout Atria Membrane Potential (mV) -80-60-40-20Normalized Conductance 0.00 0.25 0.50 0.75 1.00 Control Ventricle Knockout Ventricle
104 Figure 6.3. STX causes a change in neonatal atrial Nav steady state inactivation, but not in ventricular Nav steady state inactivation. Prepulse Potential (mV) -120-100-80-60Normalized Current 0.00 0.25 0.50 0.75 1.00 Control Atria Knockout Atria Prepulse potential (mV) -120-100-80-60Normalized current 0.00 0.25 0.50 0.75 1.00 Control Ventricle Knockout Ventricle Figure 6.3. Steady-state availability curv es for control and STX knockout atrial and ventricular Nav. Data are the mean normalized current S.E.M.
105 Figure 6.4. Absence of STX causes a slowing of the neonatal atrial Nav inactivation rate, but has no effect on the kinetics of ventricular Nav inactivation. Membrane Potential (mV) -60-40-2002040Fast Inactivation Time Constant (ms) 0 1 2 3 4 5 6 Control Atria Knockout Atria Control Ventricle Knockout Ventricle Figure 6.4. Inactivation kinetics for control and STX knockout atrial and ventricular Nav. Data are the mean time constant of inactivation S.E.M.
106 Figure 6.5. Absence of STX increases the rate of recovery fr om fast inactivation for neonatal atrial Nav to rates similar to those m easured for control and knockout ventricular Nav. Time constant for recovery from fast inactivation (ms) 5 6 7 8 9 10 Control Atria Knockout Atria Control Ventricle Knockout Ventricle *# Figure 6.5. Recovery from inactivation kinetics of for control and STX knockout atrial and ventricular Nav. Data are the mean time constant of recovery from inactivation S.E.M. Significanc e (p < 0.05) demarcated with an *. Lack of significance demarcated with an #.
107 Table 6.1. Measured action potential and Nav parameters. Control Atria Knockout Atria Control Ventricle Knockout Ventricle dV/dt (normalized) 1.000.12 0.350.03* 1.000.41 1.020.28# Maximal Depolarization (mV) 137.24.8 111.97.5* 125.610.7 127.610.2# Nav Va (mV) -55.61.7 -48.31. 1* -53.83.2 -50.71.3# Nav Vi (mV) -105.11.2 -96.71. 4* -92.93.5 -90.01.2# Nav h (ms) 1.730.17 2.930.37* 2.850.29 2.480.24# Nav rec (ms) 9.470.57 6.710.68* 6.290.33 6.900.80# Table 6.1. The measur ed action potential and Nav gating parameters measured for control and STX knockout cardiomyocyte Nav. The data are the mean parameter values S.E.M. h data were measured at -50 mV and rec data were measured at -120 mV. Significance was determined using a two-tailed StudentÂ’s t test comparing control atria and ventricl es to STX knockout atria and ventricles, respectively. Significance (p<0.04) dem arcated with an *. Lack of significance demarcated with an #.
108 control atrial Va and Vi (figures 6.2 and 6.3). No significant difference in ventricular Nav steady state gating was observed (figures 6.2 and 6.3). The Nav kinetic gating properties are again consiste nt with the steady st ate parameters. That is, no shift in neonatal ventricular Nav gating kinetics were observed, but neonatal atrial Nav kinetics were shifted in the depolarized direction in the absence of STX (Figures 6.4 and 6.5). Collectively, these data reveal the impact of a single glycogene on Nav function. Discussion Alterations of the measured action potent ial properties are consistent with the changes measured in Nav gating in the neonatal atria. Shifting the Va to more depolarized potentials require a greater depolarization to e licit an action potential. Thus, there is a higher proportion of Nav that are closed as the membrane depolarizes which leads to the slower ra te of depolarization (dV/dt) observed. Further, the activation of some Kv channels would cause both a slower rate of depolarization and a smaller maximal depol arization as the outward delayed rectifier K+ current would counteract membr ane depolarization by the inward Na+ current. Finally, the more depolarized Vi would cause Nav to inactivate possibly before maximal depolarization; therefore, decreasing Na+ current and reducing the rate of depolarization and maximal depolarization. The glycome is remodeled between the neonat al atria and ventricle as shown in the two previous chapters. Here, we showed that the changing glycome may
109 alter cardiac excitability by modulat ing action potential waveforms and Nav function. These data are consistent with our recent study in neonatal rats which revealed a sialic acid dependent shift in Nav gating parameters in the neonatal atrial myocytes but no shift in ventricular Nav gating123. Nav of control neonatal atria which expresses STX at relatively high levels, gate at a hyperpolarized potential compared Nav gating observed in the STX knockout atrial myocytes. Consistent with the Nav data, AP waveform parameter s are altered between the control and knockout STX atrial cardio myocytes while ventricular myocytes action potentials show no significant change with STX expression. The slower rates of depolarization and lower peak amplitude seen in the STX knockout atria compared to the control at ria are consistent with another study in which cardiomyocyte glycosylation was altered73. This study indicated a role for glycosylation in heart failure as studied in a mouse model in which the muscle LIM protein (MLP) is absent. MLP is not associated with glycosylation, yet cardiomyocytes are under-glycosylated co mpared to control. MLP knockout and neuraminidase-treated myocyt es show altered action potential parameters as mentioned above as well as shifts in the voltage-dependence of gating comparable with those curr ents recorded here. Together, Nav and action potential data reveal t hat changing the expression of a single glycogene can significantly modify card iac excitability. For this study, we observed the impact of the regulated expr ession of STX on cardiac excitability,
110 yet this is only one of the >100 glycogenes that are differentially expressed in the developing heart. If a single enzyme can hav e this impact on cardiac excitability, the potential of a remodeled glycome on card iac function is likely substantial.
111 CHAPTER 7 FINAL DISCUSSION This study describes two mechanisms by which Nav function and cardiomyocyte excitability can be modulated through diffe rential glycosylation. The first mechanism, the protein-isoform specific mechanism, indicates that the cell can express combinations of pr otein isoforms that have si milar functions, but are differently glycosylated. This differentia l glycosylation affects channel function and thereby alters the action potential waveform. The second mechanism, the cell-specific glycosylation mechanism describes how the change in glycan structure through regulati on of glycogene expression in a cell-specific manner. The first portion of this project focused on Nav and the manner in which alpha subunit function can be modified by the 1 isoform. Different combinations of and subunits will likely function different ly than other combinations giving the cell the ability to slightly alter Nav function. Ten and four Nav isoforms have been identified and, as shown in chapter 3, 1 modulated each subunit function differently. Nav alpha subunits show isoform-specific s ensitivity to negatively charged sialic acid residues. When expressed in the CHO cell line, no significant shift in the
112 voltage dependence of Nav1.2 and Nav1.7 gating was observed in this study. Previously, Nav1.5 was shown to be sialic acid insensitive; whereas, Nav1.4 is sensitive to sialic acids attached to the alpha subunit. Interestingly, the 1 subunit modifies the voltage dependence of Nav gating in a subunit specific manner. Three of t he less-glycosylated Nav isoforms tested were modulated by 1 sialic acids, yet Nav1.4 was insensitive to any effect of 1 in the CHO cell expression system. Furthermore when the less-glycosylated Nav1.4 chimera was co-expressed with 1, it became sensitive to 1 modulation. Together, these data indicate that alpha and/or 1 sialylation modulates Nav gating in a saturating manner. The modulation of Nav gating by 1 was abolished when co-expressed in the non-sialylating Lec2 ce ll line; likewise, 1 was unable to modulate gating of any alpha subunit isoform when the 1 N-glycosylation sites were mutated. Thus, the effect of 1 can be attributed entirely to the glyc ans attached. We conclude that 1 modulated Nav gating in a sialic acid dependent, saturating manner. With each alpha subunit having a unique gl ycosylation signature, changing which alpha subunit is expressed could result in a sialic acid dependent shift in channel voltage dependence. Four beta subunits likely modulate alpha subunit gating through various mechanisms as we have shown here, 1 modulates gating through glycosylation, but the three rema ining beta isoforms may alter channel gating through other, glyco sylation dependent or in dependent mechanisms. A
113 recent report described this exact effect with the 2 subunit. Johnson and Bennett reported that 2 caused a sialic acid dependent hyperpolarizing shift in Nav1.5 gating while 2 caused a sialic acid independ ent depolarizing shift in the voltage dependence of Nav1.2 gating121. Co-expression of both 1 and 2 with each subunit revealed an additive effect. Nav1.5.1 .2 produced a larger hyperpolarizing shift in gating; whereas, Nav1.2.1 .2 gated like Nav1.2 alone. These differences in the manner in which 1 and 2 impact subunit gating indicates that expression of various co mbinations of alpha and beta subunits would create an array of voltages at which the channel gates. In vivo, Nav alpha and beta isoform expression ma y be upor down-regulated in response to development or pathologies, and this could result in a sialic acid dependent change in the voltage dependence of Nav gating. Theoretically, Nav gating can be modulated in hundreds of ways through unique combinations of differently glycosylated alpha and beta subunits. Nav1.5 is the primary Nav isoform expressed in both chambers of the developing heart. However, in the developing skeletal muscle, the Nav isoform changes from Nav1.5 in the neonate to Nav 1.4 in the adult183-185. The adult skeletal muscle isoform is heavily glycosylated compared to the neonatal isoform 117,118. As previously reported, the Nav1.4 gating is sialic ac id sensitive while Nav1.5 is insensitive to sialic acid modulation when expressed in CHO cells118. 1 is developmentally regulated in the rodent heart where it is highly expressed in the adult ventricle, but expressed at much lower levels in the neonatal atria and ventricle and adult atria123.
114 Expression of various combinations of and auxillary subunits may create a spectrum of channel gating parameters that are glycosylation dependent; thus, supporting the relevance of the protei n-specific mechanism by which Nav function is modulated. As described, the second mechanis m studied here questioned whether cell specific change in glycogene expression and the corresponding changes in glycan structure are relevant to Nav gating. Glycogene expression varies widely between cardiac chambers and thro ugh development with over 46% of glycosylation-associated genes differentially expressed. Comparison of neonatal and adult ventricular myocyte glycogene expression showed the highest proportion of differential expression at 43.2% while the neonatal and adult atria showed the lowest at 19.4%. Corre sponding to these changes in glycogene expression, we report large changes in N-glycan structures throughout the developing myocardium. Mass spectrometry of the cardiomy ocyte glycans throughout development showed major differences in glycan st ructure between each myocyte comparison group. All groups had similar ratios of high mannose structures However, there was marked variation throughout the developing myocardium in the more complex structures above 2400 m/z. The adult atria had the highest proportion of complex N-glycans of any myocyt e type (mass ranges of 2400-3050 and 3050-5000 m/z). Neonatal atria and adult vent ricle had comparable levels of N-
115 glycans in the 3050-5000 m/z and were the lowest relative levels of the four tissues studied. In addition, the stru ctures of these glycans varied among myocyte types, with each tissue having at least one unique glycan. Observations derived from these data support a sec ond mechanism by which glycosylation might impact cardiac function, through cell-s pecific regulation of glycosylation. That is, the GeneChip and glycan screening data indicate that the glycome is remodeled throughout the myocar dium and during development. These studies have led to a proposed m odel that predicts that gating of Nav is modulated by glycans through two mechani sms (Figure 7.1). The left panel describes the "protein isoform-mediated" mechanism which is controlled at the transcriptional level where alpha and beta subunit isoforms are expressed. Each alpha subunit has a unique glycosylation signature creating currents unique to the combination of alpha and beta isoforms. The right panel shows the model for the "cell-specific glycosylation" mechanism of modulation. Here, the glycogene pr ofile determines the glycan structures present and how sodium currents are m odulated by the remodeled glycome. The neonatal and adult atria and vent ricles express the same Nav subunit isoform yet apparently have unique glycan structures which cause the channels to gate differently. Tight control of glyc ogene expression is essential for cells to consistently produce appropriate glycans for that cell's particular function. Minor changes in glycogene expression may al ter glycan structure and therefore
116 Figure 1. Model proposi ng glycosylation-dependent co ntrol and modulation of Nav gating. Figure 1. This model describes two mechanisms by which a cell can modulate Nav gating. The protein-specific (left panel) and the cell-specific glycosylation (right panel) mechanisms together create a spectrum of possible channel gating motifs (bottom panel).
117 change the manner in which glycans medi ate cell adhesion, self vs. non-self recognition, molecular trafficking, recept or activation and even modulate cellular excitability. We find here that slight changes in glyc an structure, specifically changing sialic acid levels, could alter sodium channel function (as characterized by the G-V curves shown in figure 6.1). Hundreds of possible sodium channel GV curves would result following these s light changes in glycan composition. Combination of these two mechanis ms would create a spectrum of Nav that gate at various voltages. Although both mechanisms describe means by which sodium channel function may be modified, there may be examples in which one of the two mechanisms is dominant. In the dorsal r oot ganglion (DRG), there is an apparent change in the level of glycosylation of Nav 1.9 through development186. The DRG expression of Nav1.9 does not change, yet the level of glycosylation is greater in the neonate compared to adult. This study also s howed a functional impact of glycosylation on Nav gating with approximately a 7 mV depolarizing shift in Vi in the adult compared to neonate. When treated with neuraminidase (an enzyme that removes sialic acid residues), this voltage shift was abolished. Unlike other studies, no effects on activation were ident ified, indicating that glycosylation may impact gating of Nav in an isoform dependent manner. In this study we showed that a single po lysialyltransferase, STX, is expressed essentially only in the neonatal atria. In the absence of STX, both Nav function
118 and action potential parameters were alte red in neonatal atrial myocytes. No modulation of Nav gating or action potential wave form was observed in neonatal ventricular myocytes; consistent with the fact that STX is not expressed in the ventricle. In wild type atrial myocyt es, STX must be impor tant for proper excitability and other possible processes not studied here. This further supports the model presented above that slight al terations in glycosylation may alter channel function in a cell-specif ic manner since the same Nav subunit is expressed in the neonatal atria and ventricles. Changing the expression of a single glyco gene can have dramatic effects upon cellular excitability as shown with STX in the neonatal atria. T herefore, if one of more than 100 regulated glycogenes can alte r cardiac excitability, the potential impact of the remodeled glycome on cardiac function is considerable; not only on cardiac excitability, but on a range of cardiac processes. Significance of this study The broad role of glycans in normal and pathophysiological processes demands tight control of glycans present on cell su rfaces that may differ from cell to cell and from tissue to tissue. Glycans are e ssential to regulate protein folding, cell adhesion, molecular trafficking and clearan ce, receptor activation, endocytosis and signal transduction. Furthermore, glycans determine blood type and immunity.
119 Through understanding differential glycosy lation of cells and tissues, we better understand normal function and t he dysfunction associated with pathophysiological conditions. Drug side effects are a resu lt of the drug disrupting the normal physiological proce ss in a system not associated with the target system88,89. Improving the effectiveness of a drug and preventing side effects is a main goal in therapeutic re search. Targeting glycans may allow researchers to obtain this goal. Be cause each tissue expresses a unique population of glycans, therapeut ics can be developed that use glycans to target specific tissues for drug delivery. Altered glycans are a hallmark of the tu mor phenotype. Cancer cells overand underexpress naturally occurring glycans an d expression of glycans restricted to embryonic tissues 88. Furthermore, if the embryonic glycan forms could be modified to elicit an immune response to t he tumor, a side-effect free, effective cancer therapeutic would be developed. Cardiac arrhythmias are associated with various ion channel maladies. Several reports indicate point mutations of Nav1.5 cause aberrant inactivation leading to a persistent sodium current which can lead to LQTS by creating inward sodium current in phase 2 of the cardiac action potential 129,171. Persistent sodium current would counteract the outward rect ifying K+ current, extending the phase 2 of the cardiac action potent ial and the QT segment of ECG. By shifting the Va to more hyperpolarized potentials, Nav will open following lesser depolarizations,
120 thus limiting extension of phase 2 and LQTS. Although new methods of modifying glycans are currently under in vestigation, several methods may be used to modify sodium channel function including gene therapy which may involve either mechanism explained above. First, glycosyltransferase DNA could be delivered to the cell which could direct ly increase sialylation (via STX for example) or increase glycan branching and overall sialylation. Second, auxiliary subunits may be expressed to modify Nav function. The goal of this work was to identify and explain a model by which glycans can modulate activity of proteins (specifically Nav) and how this may influence cardiac excitability. Furthermore, we ex plained the vast changes in glycogene expression and glycan structures t hat occur throughout the developing myocardium and how the regulated expressi on of a single polysialyltransferase modulates cardiac excitability. Pathologi cal and stressor mediated (i.e. cigarette smoke) studies in all organ systems might be studied in a similar manner to better understand the changes in the glyco me and excitability that these conditions induce. Through further res earch and development of glycan focused and glycan mediated therapeutics, maladies caused by or marked by changes in glycan structures may be treated.
121 REFERENCES (1) Minino AM, Heron MP, Murphy SL Kochanek KD. Deaths: final data for 2004. Natl Vital Stat Rep. 2007;55:1-119. (2) Gosiewska A, Yi CF, Brown LJ Cullen B, Silcock D, Geesin JC. Differential expression and regulation of extracellular matrix-associated genes in fetal and neonatal fibroblasts. Wound Repair Regen. 2001;9:213222. (3) Harrell MD, Harbi S, Hoffman JF Zavadil J, Coet zee WA. Large-scale analysis of ion channel gene expre ssion in the mouse heart during perinatal development. Physiol Genomics. 2007;28:273-283. (4) Ishii A, Ikeda T, Hitoshi S et al. Developmental changes in the expression of glycogenes and the content of N-glycans in the mouse cerebral cortex. Glycobiology. 2007;17:261-276. (5) Zhou J, Law HK, Cheung CY, Ng IH, Peiris JS, Lau YL. Differential expression of chemokines and t heir receptors in adult and neonatal macrophages infected with human or avian influenza viruses. J Infect Dis. 2006;194:61-70. (6) Kistler PM, Sanders P, Fynn SP et al. Electrophysiologic and electroanatomic changes in the human atrium associated with age. J Am Coll Cardiol. 2004;44:109-116. (7) Kondo RP, Anderson RH, Kuper shmidt S, Roden DM, Evans SM. Development of the cardiac conduc tion system as delineated by minKlacZ. J Cardiovasc Electrophysiol. 2003;14:383-391. (8) Zhang JF, Robinson RB, Siegelbaum SA. Sympathetic neurons mediate developmental change in cardiac sodium channel gating through longterm neurotransmitter action. Neuron. 1992;9:97-103.
122 (9) Plotnikov AN, Sosunov EA, Patber g KW et al. Cardiac memory evolves with age in association with development of the transient outward current. Circulation. 2004;110:489-495. (10) Kaplan P, Jurkovicova D, Babus ikova E et al. Effect of aging on the expression of intracellular Ca(2+) transport proteins in a rat heart. Mol Cell Biochem. 2007;301:219-226. (11) Brunelli E, Perrotta I, Bonacci A, Tripepi S. Differential expression of aquaporin 3 in Triturus italicus from larval to adult epidermal conversion. Eur J Histochem. 2007;51:25-32. (12) Wu C, Hayama E, Imamura S, Matsuoka R, Na kanishi T. Developmental changes in the expression of volt age-gated potassium channels in the ductus arteriosus of the fetal rat. Heart Vessels. 2007;22:34-40. (13) Schug N, Braig C, Zimmermann U et al. Differentia l expression of otoferlin in brain, vestibular system, immatu re and mature cochlea of the rat. Eur J Neurosci. 2006;24:3372-3380. (14) Navarro-Tableros V, Fiordelis io T, Hernandez-Cruz A, Hiriart M. Physiological development of insu lin secretion, calcium channels, and GLUT2 expression of pancreatic rat beta-cells. Am J Physiol Endocrinol Metab. 2007;292:E1018-E1029. (15) Fry M. Developm ental expression of Na+ cu rrents in mouse Purkinje neurons. Eur J Neurosci. 2006;24:2557-2566. (16) Holtje M, Brunk I, Grosse J et al. Differential distri bution of voltage-gated potassium channels Kv 1.1-Kv1.6 in the rat retina during development. J Neurosci Res. 2007;85:19-33. (17) Wada A. Roles of volt age-dependent sodium channels in neuronal development, pain, and neurodegeneration. J Pharmacol Sci. 2006;102:253-268. (18) Wooltorton JR, Gaboyard S, Hurley KM et al. Developmental changes in two voltage-dependent sodium currents in utricular hair cells. J Neurophysiol. 2007;97:1684-1704.
123 (19) Surges R, Brewster AL, Bender RA, Beck H, Feuerstein TJ, Baram TZ. Regulated expression of HCN channels and cAMP levels shape the properties of the h current in developing rat hippocampus. Eur J Neurosci. 2006;24:94-104. (20) MacDonald SH, Ruth P, Knaus HG, Shipston MJ. Increased large conductance calcium-activated pota ssium (BK) channel expression accompanied by STREX variant downr egulation in the developing mouse CNS. BMC Dev Biol. 2006;6:37.:37. (21) Bacharova L. The structural and electrical re modeling of myocardium in LVH and its impact on the QRS voltage. Anadolu Kardiyol Derg. 2007;7 Suppl 1:37-42. (22) Benitah JP, Gomez AM, Bailly P et al. Heterogeneity of the early outward current in ventricular cells isolat ed from normal and hypertrophied rat hearts. J Physiol. 1993;469:111-138. (23) Bodi I, Muth JN, Hahn HS et al. Electrical remodeling in hearts from a calcium-dependent mouse model of hypertrophy and failure: complex nature of K+ current changes an d action potential duration. J Am Coll Cardiol. 2003;41:1611-1622. (24) Brundel BJ, Van G, I, Henning RH et al. Alte rations in potassium channel gene expression in atria of patients wit h persistent and paroxysmal atrial fibrillation: differential regulation of protein and mRNA levels for K+ channels. J Am Coll Cardiol. 2001;37:926-932. (25) Brundel BJ, Van G, I, Henning RH et al. Ion channel remodeling is related to intraoperative atrial effective re fractory periods in patients with paroxysmal and persistent atrial fibrillation. Circulation. 2001;103:684-690. (26) Brundel BJ, Henning RH, Kampi nga HH, Van G, I, Crijns HJ. Molecular mechanisms of remodeling in human atrial fibrillation. Cardiovasc Res. 2002;54:315-324. (27) Cerbai E, Barbieri M, Mugel li A. Occurrence and properties of the hyperpolarization-activated current If in ventricular myocytes from normotensive and hypertensive rats during aging. Circulation. 1996;94:1674-1681.
124 (28) Desplats PA, Denny CA, Kass KE et al. Glycolipid and ganglioside metabolism imbalances in Huntington's disease. Neurobiol Dis. 2007;27:265-277. (29) Diskin S, Kumar J, Cao Z et al Detection of diffe rentially expressed glycogenes in trabecular meshwork of eyes with primary open-angle glaucoma. Invest Ophthalmol Vis Sci. 2006;47:1491-1499. (30) Hill JA. Electrical re modeling in cardiac hypertrophy. Trends Cardiovasc Med. 2003;13:316-322. (31) Huang B, Qin D, El -Sherif N. Early down-regulat ion of K+ channel genes and currents in the postinfarction heart. J Cardiovasc Electrophysiol. 2000;11:1252-1261. (32) Huang B, Qin D, Deng L, Boutjdir M, Sherif N. R eexpression of T-type Ca2+ channel gene and current in post-infarction remodeled rat left ventricle. Cardiovasc Res. 2000;46:442-449. (33) Huang B, El-Sherif T, Gidh-Jain M, Qin D, El -Sherif N. Alterations of sodium channel kinetics and gene expression in the postinfarction remodeled myocardium. J Cardiovasc Electrophysiol. 2001;12:218-225. (34) Jiang M, Zhang M, Tang DG et al. KCNE2 protein is expressed in ventricles of different species, and changes in its expression contribute to electrical remodeling in diseased hearts. Circulation. 2004;109:1783-1788. (35) Kaprielian R, Wickenden AD, Kass iri Z, Parker TG, Liu PP, Backx PH. Relationship between K+ channel dow n-regulation and [Ca2+]i in rat ventricular myocytes following myocardial infarction. J Physiol. 1999;517 ( Pt 1):229-245. (36) Keung EC. Calcium current is in creased in isolated adu lt myocytes from hypertrophied rat myocardium. Circ Res. 1989;64:753-763. (37) Le BS, Demolombe S, Chambellan A et al. Mi croarray analysis reveals complex remodeling of cardiac ion c hannel expression with altered thyroid status: relation to cellular and integrated electrophysiology. Circ Res. 2003;92:234-242.
125 (38) Li GR, Lau CP, Ducharme A, Ta rdif JC, Nattel S. Transmural action potential and ionic current remodeling in ventricles of failing canine hearts. Am J Physiol Heart Circ Physiol. 2002;283:H1031-H1041. (39) Lockstone HE, Harris LW, Swatt on JE, Wayland MT, Holland AJ, Bahn S. Gene expression profiling in t he adult Down syndrome brain. Genomics. 2007;90:647-660. (40) Martinez ML, Heredia MP, De lgado C. Expression of T-type Ca(2+) channels in ventricular cells from hypertrophied rat hearts. J Mol Cell Cardiol. 1999;31:1617-1625. (41) McIntosh MA, Cobbe SM, K ane KA, Rankin AC. Action potential prolongation and potassium currents in le ft-ventricular myocytes isolated from hypertrophied rabbit hearts. J Mol Cell Cardiol. 1998;30:43-53. (42) Mukherjee R, Sp inale FG. L-type calcium channel abundance and function with cardiac hypertrophy and failure: a review. J Mol Cell Cardiol. 1998;30:1899-1916. (43) Mukherjee R, Hewett KW, Walker JD, Basler CG, Spinale FG. Changes in L-type calcium channel abundance and f unction during the transition to pacing-induced congestive heart failure. Cardiovasc Res. 1998;37:432444. (44) Rozanski GJ, Xu Z, Zhang K, Pate l KP. Altered K+ current of ventricular myocytes in rats with chr onic myocardial infarction. Am J Physiol. 1998;274:H259-H265. (45) Schoonderwoerd BA, Van G, I, Van Veldhuisen DJ, Van den Berg MP, Crijns HJ. Electrical and structural remodeling: role in the genesis and maintenance of atrial fibrillation. Prog Cardiovasc Dis. 2005;48:153-168. (46) Tomaselli GF, Marban E. Electr ophysiological remodeling in hypertrophy and heart failure. Cardiovasc Res. 1999;42:270-283. (47) Tomaselli GF, Rose J. Molecula r aspects of arrhythmias associated with cardiomyopathies. Curr Opin Cardiol. 2000;15:202-208.
126 (48) Tomita F, Bassett AL, Myerbur g RJ, Kimura S. Diminished transient outward currents in rat hypertrophied ventricular myocytes. Circ Res. 1994;75:296-303. (49) Van Wagoner DR, Pond AL, Mc Carthy PM, Trimmer JS, Nerbonne JM. Outward K+ current densities and Kv1. 5 expression are reduced in chronic human atrial fibrillation. Circ Res. 1997;80:772-781. (50) Van Wagoner DR, Pond AL, Lamorgese M, Rossie SS, McCarthy PM, Nerbonne JM. Atrial L-type Ca2+ curr ents and human atrial fibrillation. Circ Res. 1999;85:428-436. (51) Van Wagoner DR, Nerbonne JM. Molecular basis of electrical remodeling in atrial fibrillation. J Mol Cell Cardiol. 2000;32:1101-1117. (52) Yao JA, Jiang M, Fan JS, Zhou YY, Tseng GN. Heterogeneous changes in K currents in rat ventricles thre e days after myocardial infarction. Cardiovasc Res. 1999;44:132-145. (53) Charpentier F, Liu QY, Rosen MR Robinson RB. Age-related differences in beta-adrenergic regulation of r epolarization in canine epicardial myocytes. Am J Physiol. 1996;271:H1174-H1181. (54) Hewett KW, Rosen MR. Developm ental changes in the rabbit sinus node action potential and its res ponse to adrenergic agonists. J Pharmacol Exp Ther. 1985;235:308-312. (55) Huynh TV, Chen F, Wetzel GT, Friedman WF, Klitzner TS. Developmental changes in membrane Ca2+ and K+ current s in fetal, neonatal, and adult rabbit ventricular myocytes. Circ Res. 1992;70:508-515. (56) Osaka T, Joyner RW. Developm ental changes in calcium currents of rabbit ventricular cells. Circ Res. 1991;68:788-796. (57) Osaka T, Joyner RW. Devel opmental changes in the beta-adrenergic modulation of calcium currents in rabbit ventricular cells. Circ Res. 1992;70:104-115.
127 (58) Pond AL, Nerbonne JM. ERG prot eins and functional cardiac I(Kr) channels in rat, mouse, and human heart. Trends Cardiovasc Med. 2001;11:286-294. (59) Nerbonne JM. Molecular basis of functional voltage-gated K+ channel diversity in the mammalian myocardium. J Physiol. 2000;525 Pt 2:285298. (60) Yue L, Melnyk P, Gaspo R, Wang Z, Nattel S. Molecular mechanisms underlying ionic remodeling in a dog model of atrial fibrillation. Circ Res. 1999;84:776-784. (61) Ausma J, Wijffels M, Thone F, Wouters L, Allessie M, Borgers M. Structural changes of atrial myocardium due to sustained atrial fibrillation in the goat. Circulation. 1997;96:3157-3163. (62) Ausma J, Wijffe ls M, van EG et al. D edifferentiation of atrial cardiomyocytes as a result of chronic atrial fibrillation. Am J Pathol. 1997;151:985-997. (63) Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awak e chronically instrumented goats. Circulation. 1995;92:1954-1968. (64) Santana LF, Nunez-Duran H, Dill y KW, Lederer WJ. Sodium current and arrhythmogenesis in heart failure. Heart Fail Clin. 2005;1:193-205. (65) Kass RS, Tsien RW, Weingart R. I onic basis of transient inward current induced by strophanthidin in cardiac Purkinje fibres. J Physiol. 1978;281:209-226. (66) Koumi S, Backer CL Arentzen CE. Characterizati on of inwardly rectifying K+ channel in human cardiac myocytes. Alterations in channel behavior in myocytes isolated from patients with idiopathic dilated cardiomyopathy. Circulation. 1995;92:164-174. (67) Nuss HB, Kaab S, Kass DA, Tomaselli GF, Marban E. Cellular basis of ventricular arrhythmias and abnormal automaticity in heart failure. Am J Physiol. 1999;277:H80-H91.
128 (68) Rossow CF, Minami E, Chase EG, Murry CE, Santana LF. NFATc3induced reductions in voltage-gated K+ currents after myocardial infarction. Circ Res. 2004;94:1340-1350. (69) Thuringer D, Der oubaix E, Coulombe A, Coraboeuf E, Mercadier JJ. Ionic basis of the action potential prolongat ion in ventricular myocytes from Syrian hamsters with dilated cardiomyopathy. Cardiovasc Res. 1996;31:747-757. (70) Knollmann BC, K nollmann-Ritschel BE, Weissm an NJ, Jones LR, Morad M. Remodelling of ionic currents in hypertrophied and failing hearts of transgenic mice overexpressing calsequestrin. J Physiol. 2000;525 Pt 2:483-498. (71) Lederer WJ, Tsien RW. Tr ansient inward current underlying arrhythmogenic effects of cardiotoni c steroids in Purkinje fibres. J Physiol. 1976;263:73-100. (72) Pogwizd SM, Schlotthauer K, Li L, Yuan W, Bers DM. Arrhythmogenesis and contractile dysfunction in heart failure: Roles of sodium-calcium exchange, inward rectifier potassium current, and residual beta-adrenergic responsiveness. Circ Res. 2001;88:1159-1167. (73) Ufret-Vincenty CA, Baro DJ Lederer WJ, Rockman HA, Quinones LE, Santana LF. Role of sodium channel deglycosylation in the genesis of cardiac arrhythmias in heart failure. J Biol Chem. 2001;276:28197-28203. (74) Ma TS, Baker JC, Bailey LE. Exci tation-contraction coupling in normal and myopathic hamster hearts III: functional deficiencies in interstitial glycoproteins. Cardiovasc Res. 1979;13:568-577. (75) Ohtsubo K, Marth JD. Glycosylati on in cellular mechanisms of health and disease. Cell. 2006;126:855-867. (76) Ramamurthy T, Yamasaki S, Ta keda Y, Nair GB. Vibrio cholerae O139 Bengal: odyssey of a fortuitous variant. Microbes Infect. 2003;5:329-344.
129 (77) Gong F, Lu QS, You Y et al. [P reparation of transfu sable human universal red blood cell with recombinant alpha-galactosidase.]. Zhongguo Shi Yan Xue Ye Xue Za Zhi. 2005;13:313-316. (78) Taylor ME, Drickamer K. Introduction to Glycobiology 2nd ed. New York, NY: Oxford University Press; 2006. (79) Varki A, Cummings RD, Esko JD, Freeze HH, Hart GW, Marth JD. Essentials of Glycobiology 1st ed. New York, NY: Cold Spring Harbor Laboratory Press; 1999. (80) Kornfeld R, Kornfeld S. Assemb ly of asparagine-linked oligosaccharides. Annu Rev Biochem. 1985;54:631-664. (81) Lehle L, Strahl S, Tanner W. Protein glycosyla tion, conserved from yeast to man: a model organi sm helps elucidate congenital human diseases. Angew Chem Int Ed Engl. 2006;45:6802-6818. (82) Marquardt T, Dene cke J. Congenital disorders of glycosylation: review of their molecular bases, clinical presentations and specific therapies. Eur J Pediatr. 2003;162:359-379. (83) Kojima N, Yoshida Y, Kurosawa N, Lee YC, Tsuji S. Enzymatic activity of a developmentally regulated member of the sialyltrans ferase family (STX): evidence for alpha 2,8-sialyltransferase activity toward N-linked oligosaccharides. FEBS Lett. 1995;360:1-4. (84) Turnbull JE, Field RA. Emerging glycomics technologies. Nat Chem Biol. 2007;3:74-77. (85) Freeze HH. Genetic def ects in the human glycome. Nat Rev Genet. 2006;7:537-551. (86) Comelli EM, Head SR Gilmartin T et al. A focu sed microarray approach to functional glycomics: transcription al regulation of the glycome. Glycobiology. 2006;16:117-131.
130 (87) Haliloglu G, Topaloglu H. Glycosy lation defects in muscular dystrophies. Curr Opin Neurol. 2004;17:521-527. (88) Dube DH, Bertozzi CR. Glycans in cancer and inflammation--potential for therapeutics and diagnostics. Nat Rev Drug Discov. 2005;4:477-488. (89) Fuster MM, Esko JD. The sweet and sour of cancer: glycans as novel therapeutic targets. Nat Rev Cancer. 2005;5:526-542. (90) Jaeken J, Matthijs G. Congenita l disorders of glycosylation: a rapidly expanding disease family. Annu Rev Genomics Hum Genet. 2007;8:261278. (91) Teixeira AR, Nitz N, Guimar o MC, Gomes C, Santos-Buch CA. Chagas disease. Postgrad Med J. 2006;82:788-798. (92) Bern C, Montgomery SP, Herwaldt BL et al. Evaluat ion and treatment of chagas disease in the United States: a systematic review. JAMA. 2007;298:2171-2181. (93) Colli W. Trans-sialidase: a unique enzyme activity discovered in the protozoan Trypanosoma cruzi. FASEB J. 1993;7:1257-1264. (94) Keating MT, Sanguinetti MC. Molecular and cellular mechanisms of cardiac arrhythmias. Cell. 2001;104:569-580. (95) Hille B. Ion channels of excitable membranes 3rd ed. Sunderland, MA: Sinauer Associates; 2001. (96) Yu FH, Catterall WA. Overview of the voltage-gat ed sodium channel family. Genome Biol. 2003;4:207. (97) Jiang Y, Lee A, Chen J et al. X-ray struct ure of a volt age-dependent K+ channel. Nature. 2003;423:33-41. (98) Long SB, Campbell EB, MacKinnon R. Crystal struct ure of a mammalian voltage-dependent Shaker family K+ channel. Science. 2005;309:897-903.
131 (99) Long SB, Campbell EB, MacKinnon R. Voltage sensor of Kv1.2: structural basis of electromechanical coupling. Science. 2005;309:903-908. (100) Long SB, Tao X, Ca mpbell EB, MacKinnon R. Atomic structure of a voltage-dependent K+ channel in a lip id membrane-like environment. Nature. 2007;450:376-382. (101) Ahern CA, Horn R. Specificity of charge-carryin g residues in the voltage sensor of potassium channels. J Gen Physiol. 2004;123:205-216. (102) Ahern CA, Horn R. Focused elec tric field across the voltage sensor of potassium channels. Neuron. 2005;48:25-29. (103) Campos FV, Chanda B, Roux B, Bezanilla F. 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. 2007;104:7904-7909. (104) Chanda B, Asamoah OK Blunck R, Roux B, Bez anilla F. Gating charge displacement in voltage-gated ion channels involves limited transmembrane movement. Nature. 2005;436:852-856. (105) Cohen BE, Grabe M, Jan LY. Answers and questions from the KvAP structures. Neuron. 2003;39:395-400. (106) Posson DJ, Ge P, Miller C, Bez anilla F, Selvin PR. Small vertical movement of a K+ channel voltag e sensor measured with luminescence energy transfer. Nature. 2005;436:848-851. (107) Richardson J, Blunck R, Ge P et al. Distance measurements reveal a common topology of prokaryotic volt age-gated ion channels in the lipid bilayer. Proc Natl Acad Sci U S A. 2006;103:15865-15870. (108) Swartz KJ. Towards a structural view of gating in potassium channels. Nat Rev Neurosci. 2004;5:905-916.
132 (109) Bendahhou S, O'Reilly AO, Duclohier H. Role of hydrophobic residues in the voltage sensors of the voltage-gated sodium channel. Biochim Biophys Acta. 2007;1768:1440-1447. (110) Blanchet J, Chahine M. Accessibility of four ar ginine residues on the S4 segment of the Bacillus halodurans sodium channel. J Membr Biol. 2007;215:169-180. (111) Elinder F, Ar hem P, Larsson HP. Localizatio n of the extracellular end of the voltage sensor S4 in a potassium channel. Biophys J. 2001;80:18021809. (112) Elinder F, Nilsson J, Arhem P. On the open ing of voltage-gated ion channels. Physiol Behav. 2007;92:1-7. (113) Lu Z, Klem AM, Ramu Y. Ion conduction pore is conserved among potassium channels. Nature. 2001;413:809-813. (114) Isom LL. Sodium channel beta subunits: anything but auxiliary. Neuroscientist. 2001;7:42-54. (115) Miller JA, Agnew WS, Levinso n SR. Principal glycopeptide of the tetrodotoxin/saxitoxin binding prot ein from Electrophorus electricus: isolation and partial chemical and physical characterization. Biochemistry. 1983;22:462-470. (116) Schmidt JW, Catterall WA. Palmityl ation, sulfation, and glycosylation of the alpha subunit of the sodium channe l. Role of post-translational modifications in channel assembly. J Biol Chem. 1987;262:13713-13723. (117) Bennett E, Urcan MS, Tinkle SS, Koszowski AG, Levinson SR. Contribution of sialic acid to t he voltage dependence of sodium channel gating. A possible electrostatic mechanism. J Gen Physiol. 1997;109:327343. (118) Bennett ES. Isoform-s pecific effects of sialic acid on voltage-dependent Na+ channel gating: functional sialic acids are localized to the S5-S6 loop of domain I. J Physiol. 2002;538:675-690.
133 (119) Hagen BM, Sanders KM. Deglycosylat ion of the beta1-subunit of the BK channel changes its biophysical properties. Am J Physiol Cell Physiol. 2006;291:C750-C756. (120) Johnson D, Montpet it ML, Stocker PJ, Bennett ES. The sialic acid component of the beta1 subunit modulat es voltage-gated sodium channel function. J Biol Chem. 2004;279:44303-44310. (121) Johnson D, Bennett ES. Isoform-s pecific effects of the beta2 subunit on voltage-gated sodium channel gating. J Biol Chem. 2006;281:2587525881. (122) Johnson D, Bennett ES. Gating of the shaker potassium channel is modulated differentially by N-gl ycosylation and sialic acids. Pflugers Arch. 2007. (123) Stocker PJ, Bennett ES. Different ial sialylation modulates voltage-gated Na+ channel gating throughout the developing myocardium. J Gen Physiol. 2006;127:253-265. (124) Thornhill WB, Wu MB, Jiang X, Wu X, Morgan PT, Margiotta JF. 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. 1996;271:19093-19098. (125) Ufret-Vincenty CA, Baro DJ, Santana LF. Differential contribution of sialic acid to the function of repolarizing K(+) currents in ventricular myocytes. Am J Physiol Cell Physiol. 2001;281:C464-C474. (126) Watanabe I, Wang HG, Sutachan JJ, Z hu J, Recio-Pinto E, Thornhill WB. Glycosylation affects rat Kv1.1 pota ssium channel gating by a combined surface potential and cooperative s ubunit interaction mechanism. J Physiol. 2003;550:51-66. (127) Watanabe I, Zhu J, Sutachan JJ, Go ttschalk A, Recio-Pinto E, Thornhill WB. The glycosylation state of Kv 1.2 potassium channels affects trafficking, gating, and si mulated action potentials. Brain Res. 2007;1144:1-18. Epub;%2007 Jan 31.:1-18.
134 (128) Zhang Y, Hartmann HA, Satin J. Glycosylation influences voltagedependent gating of cardiac and skeletal muscle sodium channels. J Membr Biol. 1999;171:195-207. (129) Abriel H, Cabo C, Wehrens XH et al. Novel arrhythmogenic mechanism revealed by a long-QT syndrome muta tion in the cardiac Na(+) channel. Circ Res. 2001;88:740-745. (130) Bennett ES. Effect s of channel cytoplasmic re gions on the activation mechanisms of cardiac versus skeletal muscle Na(+) channels. Biophys J. 1999;77:2999-3009. (131) Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patchclamp techniques for high-resolution cu rrent recording from cells and cellfree membrane patches. Pflugers Arch. 1981;391:85-100. (132) Jang-Lee J, North SJ, Sutton-Smith M et al. Glyco mic profiling of cells and tissues by mass spectrometry : fingerprinting and sequencing methodologies. Methods Enzymol. 2006;415:59-86. (133) Dell A, Reason AJ. Carbohydrate analysis. Curr Opin Biotechnol. 1993;4:52-56. (134) Isenberg G, Klockner U. Calcium tolerant ventricular myocytes prepared by preincubation in a "KB medium". Pflugers Arch. 1982;395:6-18. (135) Goldin AL, Barchi RL, Caldwell JH et al. Nomenclature of voltage-gated sodium channels. Neuron. 2000;28:365-368. (136) Chen C, Westenbroek RE, Xu X et al. Mice lacking sodium channel beta1 subunits display defects in neuronal excitability, sodium channel expression, and nodal architecture. J Neurosci. 2004;24:4030-4042. (137) Isom LL, De Jongh KS, Patton DE et al. Primary structure and functional expression of the beta 1 subunit of the rat brain sodium channel. Science. 1992;256:839-842.
135 (138) Chen C, Cannon SC. Modulation of Na+ channel inactivation by the beta 1 subunit: a deletion analysis. Pflugers Arch. 1995;431:186-195. (139) Isom LL, Scheuer T, Brownstein AB, Ragsdale DS, Murphy BJ, Catterall WA. Functional co-expression of the beta 1 and type IIA alpha subunits of sodium channels in a mammalian cell line. J Biol Chem. 1995;270:33063312. (140) McClatchey AI, Cannon SC, Sla ugenhaupt SA, Gusella JF. The cloning and expression of a sodium channel beta 1-subunit cDNA from human brain. Hum Mol Genet. 1993;2:745-749. (141) McCormick KA, Isom LL, Ragsdale D, Smith D, Scheuer T, Catterall WA. Molecular determinants of Na+ channe l function in the extracellular domain of the beta1 subunit. J Biol Chem. 1998;273:3954-3962. (142) McCormick KA, Srinivasan J, Wh ite K, Scheuer T, Catterall WA. The extracellular domain of t he beta1 subunit is both necessary and sufficient for beta1-like modulation of sodium channel gating. J Biol Chem. 1999;274:32638-32646. (143) Meadows L, Malhotra JD, Stet zer A, Isom LL, Ragsdale DS. The intracellular segment of the sodium channel beta 1 subunit is required for its efficient association with the channel alpha subunit. J Neurochem. 2001;76:1871-1878. (144) Meadows LS, Malhotra J, Loukas A et al. Functional and biochemical analysis of a sodium channel beta1 subunit mutation responsible for generalized epilepsy with febrile seizures plus type 1. J Neurosci. 2002;22:10699-10709. (145) Meadows LS, Chen YH, Powell AJ Clare JJ, Ragsdale DS. Functional modulation of human brain Nav1.3 sodium channels, expressed in mammalian cells, by auxiliary beta 1, beta 2 and beta 3 subunits. Neuroscience. 2002;114:745-753. (146) Smith MR, Smith RD, Plummer NW, Meisler MH, Goldin AL. Functional analysis of the mouse Scn8a sodium channel. J Neurosci. 1998;18:60936102.
136 (147) Smith RD, Goldin AL Functional analysis of the rat I sodium channel in xenopus oocytes. J Neurosci. 1998;18:811-820. (148) Spampanato J, Escayg A, Meisle r MH, Goldin AL. Generalized epilepsy with febrile seizures plus type 2 mutation W1204R alters voltagedependent gating of Na(v)1 .1 sodium channels. Neuroscience. 2003;116:37-48. (149) Vijayaragavan K, O'Leary ME, Chahi ne M. Gating properties of Na(v)1.7 and Na(v)1.8 peripheral nerve sodium channels. J Neurosci. 2001;21:7909-7918. (150) Zimmer T, Benndorf K. The human heart and rat brain IIA Na+ channels interact with different molecu lar regions of the beta1 subunit. J Gen Physiol. 2002;120:887-895. (151) Amir R, Argoff CE, Bennett GJ et al. The role of sodium channels in chronic inflammatory and neuropathic pain. J Pain. 2006;7:S1-29. (152) Craner MJ, Klein JP, Renganathan M, Black JA, Waxman SG. Changes of sodium channel expression in experim ental painful diabetic neuropathy. Ann Neurol. 2002;52:786-792. (153) Guo F, Yu N, Ca i JQ et al. Voltage-gated s odium channel Na(v)1.1, Na(v)1.3 and beta(1) subunit were up-regulated in the hippocampus of spontaneously epileptic rat. Brain Res Bull. 2008;75:179-187. (154) Van WA, Matthews G. Expression of sodium channels Nav1.2 and Nav1.6 during postnatal development of the retina. Neurosci Lett. 2006;403:315317. (155) Waxman SG, Craner MJ, Black JA. Na+ channel expression along axons in multiple scleros is and its models. Trends Pharmacol Sci. 2004;25:584591. (156) Sashihara S, Oh Y, Black JA Waxman SG. Na+ channel beta 1 subunit mRNA expression in developi ng rat central nervous system. Brain Res Mol Brain Res. 1995;34:239-250.
137 (157) Sashihara S, Greer CA, Oh Y, Waxman SG. Cell-specific differential expression of Na(+)-channel beta 1-s ubunit mRNA in the olfactory system during postnatal development and after denervation. J Neurosci. 1996;16:702-713. (158) Sutkowski EM, Catterall WA. Beta 1 subunits of sodium channels. Studies with subunit-specific antibodies. J Biol Chem. 1990;265:12393-12399. (159) Yang JS, Bennett PB, Makita N, G eorge AL, Barchi RL. Expression of the sodium channel beta 1 subunit in rat skeletal muscle is selectively associated with the tetrodotoxin-s ensitive alpha subunit isoform. Neuron. 1993;11:915-922. (160) Abriel H, Wehrens XH, Benhor in J, Kerem B, Kass RS. Molecular pharmacology of the sodium channel mutation D1790G linked to the longQT syndrome. Circulation. 2000;102:921-925. (161) Akai J, Makita N, Sakurada H et al. A novel SCN5A mutation associated with idiopathic ventricular fibrillation without typical ECG findings of Brugada syndrome. FEBS Lett. 2000;479:29-34. (162) Bendahhou S, Cummins TR, Kwie cinski H, Waxman SG, Ptacek LJ. Characterization of a new sodium channel mutation at arginine 1448 associated with moderate Paramy otonia congenita in humans. J Physiol. 1999;518 ( Pt 2):337-344. (163) Bendahhou S, Cummins TR, Tawil R, Waxman SG, Ptacek LJ. Activation and inactivation of the voltage-gated s odium channel: role of segment S5 revealed by a novel hyperkalaemic periodic paralysis mutation. J Neurosci. 1999;19:4762-4771. (164) Bendahhou S, Cummins TR, Hahn AF, Langlois S, Waxman SG, Ptacek LJ. A double mutation in families wit h periodic paralysis defines new aspects of sodium channel slow inactivation. J Clin Invest. 2000;106:431438. (165) Bennett PB, Yazawa K, Makita N, George AL, Jr. Molecular mechanism for an inherited cardiac arrhythmia. Nature. 1995;376:683-685.
138 (166) Green DS, Hayward LJ, Geor ge AL, Jr., Cannon SC. A proposed mutation, Val781Ile, associated with hyperkalemic periodic paralysis and cardiac dysrhythmia is a benign polymorphism. Ann Neurol. 1997;42:253256. (167) Green DS, George AL Jr., Cannon SC. Human sodium channel gating defects caused by missense mutations in S6 segments associated with myotonia: S804F and V1293I. J Physiol. 1998;510 ( Pt 3):685-694. (168) Makita N, Shirai N, Wang DW et al. Cardiac Na(+) channel dysfunction in Brugada syndrome is aggravat ed by beta(1)-subunit. Circulation. 2000;101:54-60. (169) Scheffer IE, Harkin LA, Grint on BE et al. Temporal lobe epilepsy and GEFS+ phenotypes associated with SCN1B mutations. Brain. 2007;130:100-109. (170) Spampanato J, Escayg A, Meisler MH, Goldin AL. Functional effects of two voltage-gated sodium channel mu tations that cause generalized epilepsy with febrile seizures plus type 2. J Neurosci. 2001;21:7481-7490. (171) Splawski I, Timoth y KW, Tateyama M et al. Variant of SCN5A sodium channel implicated in risk of cardiac arrhythmia. Science. 2002;297:13331336. (172) Wallace RH, Scheffer IE, Bar nett S et al. Neuronal sodium-channel alpha1-subunit mutations in genera lized epilepsy with febrile seizures plus. Am J Hum Genet. 2001;68:859-865. (173) Wallace RH, Scheffer IE, Parasi vam G et al. Generalized epilepsy with febrile seizures plus: mutation of the sodium channel subunit SCN1B. Neurology. 2002;58:1426-1429. (174) Wallace RH, Hodgson BL, Grin ton BE et al. Sodium channel alpha1subunit mutations in severe myoclonic epilepsy of infancy and infantile spasms. Neurology. 2003;61:765-769.
139 (175) Wang DW, Yazawa K, George AL, Jr., Bennett PB. Characterization of human cardiac Na+ channel mutations in the congenital long QT syndrome. Proc Natl Acad Sci U S A. 1996;93:13200-13205. (176) Wang Q, Shen J, Li Z et al. Cardia c sodium channel mutations in patients with long QT syndrome, an inherited cardiac arrhythmia. Hum Mol Genet. 1995;4:1603-1607. (177) Wang Q, Shen J, Splawski I et al SCN5A mutations associated with an inherited cardiac arrhythmia, long QT syndrome. Cell. 1995;80:805-811. (178) Wehrens XH, Abriel H, Cabo C, Benhorin J, Kass RS. Arrhythmogenic mechanism of an LQT-3 mutation of the human heart Na(+) channel alpha-subunit: A comput ational analysis. Circulation. 2000;102:584-590. (179) Wehrens XH, Rossenbacker T, Jongbloed RJ et al. A novel mutation L619F in the cardiac Na+ channel SCN5A associated with long-QT syndrome (LQT3): a role for the III linker in inactivation gating. Hum Mutat. 2003;21:552. (180) Xu R, Thomas EA, Gazina EV et al. Generalized epilepsy with febrile seizures plus-associated sodium channel beta1 subunit mutations severely reduce beta subunit-mediated modulation of sodium channel function. Neuroscience. 2007;148:164-174. (181) Kitagawa H, Paulson JC. Differentia l expression of five sialyltransferase genes in human tissues. J Biol Chem. 1994;269:17872-17878. (182) Livingston BD, Paulson JC. Poly merase chain reaction cloning of a developmentally regulated member of the sialyltransferase gene family. J Biol Chem. 1993;268:11504-11507. (183) Goldin AL. Resurgence of sodium channel research. Annu Rev Physiol. 2001;63:871-894. (184) Trimmer JS, Cooperman SS, Agnew WS, Mandel G. Regulation of muscle sodium channel transcripts during development and in response to denervation. Dev Biol. 1990;142:360-367.
140 (185) Yang JS, Sladky JT, Kallen RG, Barchi RL. TTX-sensitive and TTXinsensitive sodium channel mRNA tr anscripts are independently regulated in adult skeletal muscle after denervation. Neuron. 1991;7:421-427. (186) Tyrrell L, Renganatha n M, Dib-Hajj SD, Waxman SG. Glycosylation alters steady-state inactivation of sodium channel Nav1.9/NaN in dorsal root ganglion neurons and is dev elopmentally regulated. J Neurosci. 2001;21:9629-9637.
ABOUT THE AUTHOR Marty Louis Montpetit was born September 10, 1978 in Tecumseh, Michigan. Marty graduated from Tecumseh High Sc hool in 1996 and enrolled at Grand Valley State University where he receiv ed his B.S. degree in Health Sciences. While attending Grand Valley, he joined Al pha Sigma Phi Fraternity where he held many leadership positions a nd received several scholarship commendations. Marty served as a Teac hing Assistant and, after graduation, Instructor of the Physiology Laborator y Course at Grand Valley. In 2001, he enrolled in the graduate program at the University of South Florida College of Medicine and joined the laboratory of Er ic S. Bennett, Ph.D. He earned his Masters of Science in Medical Sci ences in 2001 and was accepted as a candidate for Ph.D. study. Finally, he su ccessfully completed the requirements necessary to be awarded his Ph.D. in Medi cal Sciences from the University of South Florida College of Medicine in early 2008.